Journal UII Yogya

i
Vol.1, No. 1, Mei 2011 ISSN 1979-7303

DAFTAR ISI

Daftar Isi i

Comparison of the Behaviour of Pre-Crack and Intact Partially Saturated Stiff Clay
Subjected to Plane Strain Testing
Miftahul Fauziah ........................................................................................................................... 1

Compressive Strength Assessment of Concrete Structures from Small Core
by Point Load Test
Achfas Zacoeb and Koji Ishibashi ................................................................................................ 10

Essential Strategies to Promote Earthquake Safer Non-Engineered Housing:
An Indonesian Perspective
Setyo Winarno .............................................................................................................................. 19

The Implication Of Peatland Built Environment In Urban Drainage System:
Case Study Of Sungai Merah Area, Sibu, Sarawak
Frederik Josep Putuhena, Ting Sie Chun, and Salim Said ........................................................... 29

Perbandingan Kekakuan Struktur Lantai Komposit Kayu Glugu-Beton
dengan Berbagai Jenis Penghubung Geser
Susastrawan .................................................................................................................................. 38

Masonry Unit Utilizing Aggregate from Construction Demolition Bound with Asphalt
I Nyoman Arya Thanaya .............................................................................................................. 47

Pengembangan Peta Hazard Kegempaan untuk Pulau Jawa dan Perbedaannya
dengan Hasil-Hasil Studi Terdahulu
Lalu Makrup, Sarwidi, dan Susilo ................................................................................................ 58

Shoreline Change Model Using the Epr Method and the Simulation of
Coastal Vulnerability in Sambas District-West Kalimantan
M. Meddy Danial, Rustamaji, and Eka Priadi .............................................................................. 68

Time-History Response of 2-D Timber Frame Reinforced with Wooden Panel
Ali Awaludin ................................................................................................................................ 75

Wavelet-Spectrogram Analysis of Surface Wave Technique: A Novel Procedure
for Non-Destructive Measurement on Pavement Surface Layer
Sri Atmaja P. Rosyidi .................................................................................................................. 83

Analisis Antrian Akibat Daerah Sempit
di Jalan Bebas Hambatan
Sukarno ......................................................................................................................................... 93

Fauziah: Comparison of The Behaviour of Precrack and Intact Partially Saturated Stiff Clay ...Testing
Comparison of the Behaviour of Pre-Crack and Intact Partially
Saturated Stiff Clay Subjected to Plane Strain Testing

Miftahul Fauziah

Civil Engineering Departement, Islamic University of Indonesia, Yogyakarta, Indonesia,
email: miftahul.fauziah@uii.ac.id

Abstract: The present modeling of soil is based on principles of continuum mechanics. The
existence of cracks bring soils to be non uniform condition and therefore not amendable to analysis
by continuum mechanics. On the other hand, there has been considerable interest in the application
of fracture mechanics to replicate their behaviour. Partially saturated soil behaviour is different
from those of fully saturated soil because of the influence of suction. Results obtained with the
strength theory of saturated soil could not be directly applied to solve the partially saturated soil
problems. Moreover, the mechanical properties of soil that usually conducted using triaxial
apparatus rather than biaxial device will be addressed according to the fact that geotechnical field
problems are often trully or close to plane strain situation. This paper will disseminate the result of
some experimental testing on the on the properties of pre-crack and intact partially saturated
compacted kaolin clay specimen under plane strain condition. Two types of intact specimen and
pre-crack specimen had tested under different matric suction and net normal stress. The results
demonstrated that a higher failure stress and compressive strength was reached by the intact
specimen than the pre-crack specimen. Shear strength of the intact specimens were higher along
their axial strain than those of pre-crack specimens. It was also shown that a pronounce failure
stress was exhibited by pre-crack specimen rather than intact specimen.

Keywords: Pre-crack, partly saturated, matric suction, and plane strain.

INTRODUCTION

The present modeling of soil is based on
principles of continuum mechanics in spite of
the fact that discontinuities are known to
develop when such geological materials are
subject to loading. However, in the case of
strong rock, there has been considerable
interest to account for such discontinuities
using fracture mechanical approach (Ingraffea,
1987). Many studies have been conducted on
detailed aspect of such discontinuities, but
these are of limited practical application in an
actual situation. The existence of cracks and
fissures, which are the results of mechanical,
thermal and volume-change-induced stresses,
such soils are non uniform and therefore not
amendable to analysis by continuum
mechanics. On the other hand, fracture
mechanical theory may be used to advantage
to replicate their behaviour. Atkinson and
Bransby (1982) proposed the conventional
failure criteria for soils which might be partly
appropriate to yield-dominant behaviour, but
not this category of brittle fracture. In practice,
there is the possibility that soil behaves more
like a brittle material. The soil ruptures
suddenly under compressive loading like soft
rock, starting from the weakest fracture in it.
A basic premise of fracture theory is that crack
like imperfections are inherent in engineering
materials. These defects have the tendency to
make stresses higher, which eventually trigger
off fractures when a material body is subjected
to a critical load or undergoes damage under
cyclic loading. This present state of fracture
mechanical theory has been summarized by
Anderson (2004). Lo et al. (2005) modeled
brittle overconsolidated clay accordingly and
thereby provided a rational basis for the
prediction of such soil behaviour.
The behaviour of partially saturated soil is
different from those of fully saturated soil
because of the influence of suction. It

has been
1
Jurnal Rekayasa Sipil (JRS), Vol. 1, No. 1, Mei 2011: 1-9
observed that several stability problems,
involving soils used as construction materials
are associated with variability of water content
that occur periodically in nature and
consequently in changes of matric suction.
Although

soils are generally assumed fully
saturated below the groundwater table,

they
may be semi saturated near the state of full

saturation under certain conditions. The
partially saturated soil can be the results of
fluctuation of the ground water table being the
consequence of natural processes or human
activity. Partially saturated soils form the
largest category of materials that cannot be
classified by classical saturated soil mechanics
concepts. The strength theory of saturated soil
could not be directly applied to solve the
problems of partially saturated soil.
Partially saturated soil is generally
characterized by three phases, soil solids,
water, and air. Fredlund and Morgenstern
(1977) was introduced an additional
independent phase, a so called the airwater
interface or contractile skin. Based on multi
phase continuum mechanics, a theoretical
stress analysis of an partially saturated soil has
been presented (Fredlund and Morgenstern,
1977; Fredlund and Morgenstern, 1976). The
analysis concluded that any two of three
possible normal stress variables can be used to
describe the stress state of an unsaturated soil.
This is in contrast to saturated soil, where it is
possible to relate the mechanical properties of
the soil to the effective stress only.
The mechanical properties of soil is routinely
interpreted from conventional triaxial testing
or axisymmetric conditions; whereas, testing
of soil using the plane strain device would be
more useful information, as more geotechnical
field problems such as landslide problems,
failure of soils beneath shallow foundations,
and failure of retaining structures are truly or
basically occur in these situations. It was
reported by Mochizuki et al. (1993) that when
soil is tested under plane strain conditions, it,
in general, exhibits a higher compressive
strength and lower axial strain. The plane
strain testing on the behaviour of fined grained
sands had been reported (Alshibli and Akbas,
2007, Alshibli et al., 2004; Alshibli and Sture,
2000; Bizzarri, 1995; Han and Vardoulakis,
1991; Hans and Drescher, 1993; Lee, 1970;
Marach et al., 1984; and Mochizuki et al,
1993). However, the plane strain testing of
clay soils have only been initiated recently
(Fauziah and Nikraz, 2008; Fauziah and
Nikraz, 2007; Lo et al., 2000; Drescher et.al
(1990)) and published data of such tests
especially for brittle clay material is very
limited.
This paper will disseminate the result of
experimental study on the behaviour of pre-
crack partially saturated clay specimens
compare to the intact specimen by the use of
plane strain apparatus, although the behaviour
of overconsolidated clay (Fauziah and Nikraz,
2007) and fracture characteristics of brittle
clay may also be determined by this test
apparatus. Discription of the apparatus,
specimen preparation, testing method,
procedures and data analysis will be presented
in the following discussion. Some results of
the testing will be compared with the known
soil behaviour and previous working.

EXPERIMENTAL PROGRAM

Material and Specimen Preparation
Plane strain experiments have been performed
on remoulded kaolin clay specimens. The
material used in this study was commercial
kaolin clay, which is a product of UNIMIN
PTY LTD, Australia, with a specific gravity
G
s
=2.6, liquid limit LL of 53.5 %, plasticity
index PI of 22.74 % and plastic limit PL of
30.76 %.
Firstly, a kaolin clay sample was slurried to a
uniform consistency of 1 times its liquid
limit using an electrical soil mixer. This slurry
was obtained by mixing 8 kg of kaolin powder
with 6 kg of water using the electric mixer for
about 2 hours. A lubricated steel cylindrical
mould with the height of 600 mm and 150 mm
in diameter was used to consolidate the slurry
using a hydraulic tester in over a period of one
to two weeks which the maximum of 300 kPa
was applied in three stages. Two circular
perspexes were placed at both ends of the
slurry in the mould to apply the pressure
evenly to the slurry. The slurry was allowed to
2
Fauziah: Com
consolidate
pressure w
being squ
perspexes
pressure w
further s
consolidati
was then
lubricated
specimen
plastic film
faces with
water pres
the dehum
tested.
In the pla
membrane
specimen
The rectan
on top of
assembly a
The rubbe
the porous
HAED and
rings and
specimen,
the top as
pedestal o
apparatus.
slipped ov
use of a s
rigid persp
secured by
set up. Th
laser senso

Plane Stra
A modifie
was used f
of partiall
shows the p
The plane
cell, with
internal di
An initial
of 36 mm,
72 mm, so
on the bas
laterally by
its out-of
mparison of The
e by its o
was applied
ueezed out
and the m
was applie
settlement
ion process
extruded
formers.
and the fo
m sheeting
h liquid wa
ssure, the sp
midifier for a
ane strain
was first
with the ai
ngular porou
the specim
and the high
r membran
s plate, the
d secured b
rectangula
together w
ssembly we
of the pla
The rubb
ver the pede
set of O-rin
pex plates
y the use of
he pressure
or set-up we
ain Equipm
ed convent
for the purpo
ly saturate
plane strain
strain arra
the height
iameter and
prismatic s
, height of
that the asp
se pedestal,
y two rigid
plane move
Behaviour of P
own weigh
d to preve
t between
mould. A h
ed when th
change.
s had comp
from the
Followin
ormer were
after seali
ax. To obta
pecimens w
at least 2 d
test set-up
t placed o
id of a sle
us plate wa
men followe
h air-entry
ne was next
e top assem
by the use o
ar Perspex
with the por
ere then pla
ane strain
er membra
estal and se
ngs and cla
were then
clamp set f
cell, top
ere then inst
ment
tional triax
ose of plane
ed specime
n test arrang
angement i
t of 300 m
d 30 mm w
specimen w
72 mm, and
pect ratio is
where it w
perspex pla
ement whic
Precrack and In
ht and sm
ent the slu
the circu
higher verti
here was
Once t
leted, the s
mould in
ng this, t
e wrapped
ing both th
ain even po
were placed
days until it
p, the rubb
over the t
eeve stretch
as then plac
ed by the t
disc (HAED
t slipped ov
mbly and t
of a set of
clamp. T
rous plate a
aced over t
compressi
ane was ne
ecured by t
amp set. Tw
n placed a
for biaxial t
assembly a
talled.
xial appara
e strain testi
en. Figure
gement.
s placed in
mm, 200 m
wall thickne
with the wid
d thickness
s 2 was plac
was restrain
ates to restra
ch make ,
ntact Partially S
mall
urry
ular
cal
no
the
soil
nto
the
in
heir
ore
d in
t is
ber
test
her.
ced
top
D).
ver
the
O-
The
and
the
ion
ext
the
wo
and
test
and
atus
ing
1
n a
mm
ess.
dth
of
ced
ned
ain
, .
Th
pr
A
ce
in
the
19
19
the
2
fri
wo
ma

F
Saturated Stiff C
herefore, o
rincipal stre
30 mm
entrally with
Figure 2, t
e specimen
941; Rowe a
964; Bishop
e specimen
or more, so
iction and
ould be neg
aintain cons
Figure 1.
igure 2. A s
conditio
72 mm
c
o
3

Clay ...Testing
only major
esses acting
diagonal p
hin the inta
to simulate
n. Previous
and Barden
p and Green
n aspect rati
o that the ef
the restrai
gligible and
stant strengt

Plane strain

schematic d
on of the pre

c
2
=0
o
1

o
1

() and
on the soil
precrack w
act specimen
e the discon
s researche
n, 1964; Lee
n, 1965) sug
io (height to
ffect of loa
int of load
d the specim
th.
n test arrang
diagram of p
e-crack spec
o
1

36 m
minor ()
l specimen.
was formed
n as shown
ntinuities in
rs (Taylor,
e and Seed,
ggested that
o width) of
ading platen
ding frame
men would
gement
plane strain
cimen
o
3

mm

)
.
d
n
n
,
,
t
f
n
e
d
3
A high air-entry disc (HAED) was used as the
interface between the partly saturated soil and
the pore water pressure measuring system to
prevent any air from passing through the disc
into the measuring system. All surfaces which
are in contact with the specimen were greased
to avoid the likelihood of scratching and
reduced friction. An LVDT (Linear Vertical
Displacement Transducer) was used to
measure the axial displacements of the test
specimen. The global volume change of the
water saturated soil specimen was monitored
by the use of an automatic volume change unit
which is connected to the back-pressure line.
A data acquisition system consisting of an
MPX 300 data logger and a set of
microcomputer were used to record the
displacement, loads, pressure and volume
change reading of the specimen. WINHOST
2.0 package software was used to convert
digital bit data from the ADU (Analogue
digital Unit) to engineering units based on the
calibration of the relevant measuring unit,
which was done before running the plane
strain test. A more detail discription of this
biaxial compression device can be found in
previous working reported by Fauziah and
Nikraz (2008) and Fauziah and Nikraz (2007).

Testing Method and Procedure
Two types of intact specimen and pre-crack
specimen have been tested under plane strain
condition. The pre-crack specimens were
formed a 30 mm diagonal precrack in the
center of the intact specimen to simulate the
discontinuities in the specimen. Two specimen
of IM (intact specimen) and PCM (Pre-crack
specimen) were tested under net normal stress
of 0 and maximum matric suction of 500 kPa
and two specimen of IN (intact specimen) and
PCN (Pre-crack specimen) were tested under
matric suction of 0 and maximum net normal
stress of 800 kPa. The specimen was first
saturated until the B-value of the specimen
reached the value of 0.95-0.98, followed by
matric suction and net normal stress applied
and loading compression processed.
An initial net normal stress of 0 and matric
suction of 10 were applied to the IM and
PCM specimens by set a cell pressure of 600
kPa, back pressure of 590 kPa and pore
pressure of 600 kPa. The volume change of
the soil skeleton was monitored continuously
by the laser sensors. The change in volume of
water in the specimen was also monitored
continuously by the volume change gauge
which was connected to the back-pressure
line. Once the changes in the soil and water
volumes had ceased, the test specimen was
presumed to have fully consolidated under a
matric suction of 10 kPa. The matric suction
was next increased to 20 kPa by reducing the
back pressure to 580 kPa. The corresponding
changes in the soil skeleton and water volumes
were monitored continuously until they had
ceased, at which stage the corresponding void
ratio and the water content of the test
specimen were computed from the cumulative
changes in soil skeleton and water volumes.
The entire procedure, that is from increasing
the matric suction to the desired value up to
flushing out the air bubbles, was repeated for
matric suctions of 50, 100, 200, 300, and 500
kPa respectively, which were applied by
reducing the back pressure accordingly.
Another test apparatus were set up for IN and
PCN specimens similarly as before, except
that the head was replaced with a porous disc.
A cell pressure of 110 kPa, back-pressure of
100 kPa and pore-air pressure of 100 kPa were
then applied to the specimen to provide an
initial net normal stress of 10 kPa and matric
suction of 0. The changes in soil skeleton and
water volumes were then monitored
continuously and when these changes had
ceased, the total changes in soil and water
volumes were noted. The net normal stress
was first increased to 20 kPa by increasing the
cell pressure to 120 kPa. Thereafter, the entire
above procedure, starting from applying the
net normal stress up to when the changes in
soil and water volumes ceased, was repeated
for net normal stresses of 50, 100, 200, 300,
500 and 800 kPa.
The specimen was then compressed by
elevating the base of the confining pressure
cell at a constant velocity of 0.08 mm/m with
the drainage line closed at at net normal stress
of 0 kPa (NNS=0kPa) and matric suction of
500 kPa (MS=500 kPa) for IM and PCM
4
specimens and at matric suction of 0 and net
normal stress of 800 kPa for IN and PCN
specimens. This loading rate was deduced
based on the permeability of adopted kaolin
clay suggested by Bishop and Henkel (1962).
The data were recorded at 3 minute interval
test and it was terminated at the axial strain of
about 20 % or sooner. The specimen was then
taken out immediately for the purpose of
moisture content test. In the analysis of the
behaviour of the brittle unsaturated clay, the
pore pressure parameter would be required in
order to the determined the pore pressure
increments and the matric suction. The pore
pressure parameters were deduced from the
volumetric deformation coefficient, which was
obtained by laboratory testing. This procedure
was adopted from Fredlund and Rahardjo
(1993), although adapted to biaxial conditions.

RESULTS AND DISCUSSIONS

Figure 3 and Figure 4 present the strain
softening response of the intact and pre-crack
specimen. The peak stress and strain of the
specimen is summarised in Table 1.

Table 1. The peak stress vs axial strain
Specimen
name
Peak
stress
(kPa)
Vertical
strain (%)
IN 159.18 3.64
PCN 125.34 2.73
IM 228.39 3.87
PCM 124.52 3.33

In general, the shear stress curves increase
monotonically with the increasing of vertical
strain until they reach peak stresses followed
by strain softening behaviour. According to Lo
et.all (2005), this is the typical phenomenon of
specimen of brittle, hard partly saturated soil
specimen and demonstrated elastic failure
only.
As can be seen in Figure 3 and Figure 4, the
shear strength of the intact specimen of IM
and IN were higher along the vertical strain
than that the specimen containing
discontinuities or precrack of PCM and PCN
specimen. The highest failure stress of
228.395 kPa was reached by the intact
specimen of IM, and the lowest failure stress
of 124.519 kPa was derived by the pre-crack
specimen of PCM. The presence of a fissure or
discontinuity makes the soil weaker as the
effective area offering resistance to shear is
reduced. The shear strength along a surface of
discontinuity is thereby less than that of the
intact material. It is also shown from the graph
and Table that pronounced peak shear
strength was occurred to the precrack
specimen than the intact specimen. Similar
observation of pre-crack overconsolidated clay
had been reported elsewhere (Lo et al., 2000).

Figure 3. Stress-strain of specimen under Net
Normal Stress (NNS)= 0, and Matric Suction
(MS )= 500 kPa

Figure 4. Stress-strain of specimen under
Matric Suction (MS )=0 and Net Normal
Stress (NNS)= 800 Kpa

Figure 5 shows the constitutive surface of void
ratio versus log net normal stress and log
matric suction of the intact specimen, while
0
50
100
150
200
250
300
0 1 2 3 4 5 6 7 8
axial strain (%)
D
e
v
i
a
t
o
r
i
c

s
t
r
e
s
s

(
k
P
a
)
IM (Intact)
PCM (Pre-crack)
0
50
100
150
200
0 1 2 3 4 5 6 7 8
axial strain (%)
D
e
v
i
a
t
o
r
i
c

s
t
r
e
s
s

(
k
P
a
)
IN (Intact specimen)
PCN (Pre-crack)
5
Figure 6 plots the constitutive surface of void
ratio versus log net normal stress and log
matric suction of the pre-crack specimen. The
slope of the intersection curves are the volume
change index C
m
and water content index D
m
respectively for the case that net normal stress
set to zero, whereas the slope of the
intersection curves are the volume change
index C
t
and water content index D
t
when the
matric suction set to zero. C
t
is the slope of the
consolidation curve and is equal to the
compressive index of a saturated soil, while
C
m
is the slope of the shrinkage curve.

Figure 5. Constitutive surface of void ratio of
intact specimen

Figure 6. Constitutive surface of void ratio of
pre-crack specimen

The change in void ratio of the IM and PCM
specimen in relation with matric suction
depicted in Figure 7, while void ratio changes
of the IN and PCN specimen in connection
with net normal stress changes presented in
Figure 8. It can be shown from the graphs that
the curves went down with the increasing of
either matric suction or net normal stress for
all of the specimens. The value of the volume
change index of the IM and PCM specimens
were calculated as 0.011and 0.016
respectively. The value of the C
t
of the IN and
PCN specimens are 0.219 and 0.257
respectively. The higher slope of pre-crack
curves of IM and IM than that of intact curve
of PCM and PCN indicated that the intact
specimen had higher compressive strength
than that of pre-crack specimen either tested
under net normal stress of zero and matric
suction of 500 Kpa or under zero matric
suction and net normal stress of 800 kPa. This
is in consistent with the observation reported
by Lo et al (2000).

Figure 7. Void ratio versus matric suction of
specimen under NNS=0 and MS=500 Kpa

Figure 8. Void ratio versus Net normal stress
of specimen under MS=0, NNS=800 Kpa
0.30
0.50
0.70
0.90
1.10
1.30
1.50
1.70
1.00 1.50 2.00 2.50 3.00
Log (matric suction, net normal strees)
v
o
i
d

r
a
t
i
o

(
%
)
Log MS Log NNS
0.30
0.50
0.70
0.90
1.10
1.30
1.50
1.70
1.00 1.50 2.00 2.50 3.00
v
o
i
d

r
a
t
i
o

(
%
)
Log MS Log NNS
1.170
1.175
1.180
1.185
1.190
1.195
1.0 1.5 2.0 2.5 3.0 3.5
Log Matric suction (kPa)
V
o
i
d

r
a
t
i
o

(
%
)
IM (Intact)
PCM (Pre-crack)
0.800
0.900
1.000
1.100
1.200
1.300
1.0 1.5 2.0 2.5 3.0 3.5
Log Net Normal stress (kPa)
v
o
i
d

r
a
t
i
o

(
%
)
PCN (Pre-crack)
6
The Constitutive surface of water content of
intact specimen and pre-crack specimen
plotted in Figure 9 and Figure 10, respectively.
The constitutive surface of water content is
defined by the compressive index D
m
and D
t

corresponding to the matric suction and net
normal stress respectively. The value of
compressive index obtained by determining
the gradient of the linear portion of the curve
of the water content against the log of matric
suction for D
m
and the log of net normal stress
for D
t.
. As can be seen from the graphs, the
curves went down with the increasing of either
matric suction or net normal stress.

Figure 9. Constitutive surface of water content
of intact specimen

Figure 10. Constitutive surface of water
content of pre-crack specimen

The curves also indicated that the higher
gradient value of 0.071 and 0.068 were
reached by precrack specimens of PCM and
PCN respectively, than that of the gradient
value of intact specimens of IM and IN which
were 0.55 and 0.056 respectively. The changes
in water content of the specimen in relation
with matric suction depicted in Figure 11,
while water content changes versus net normal
stress plotted in Figure 9. It were clearly
demonstrated from the graphs that under the
same net normal stress and matric suction the
intact specimen had lower gradient than the
pre-crack specimen.

Figure 11. Water content change of specimen
under NNS=0 and MS=500 Kpa

Figure 12. Water content change of specimen
under MS=0, NNS=800 Kpa

Similar to the curve of the void ratio changes
in Figures 8-9, as well as water content
changes in Figures 11-12 and consistent with
the stress-strain behaviour of the specimens
shown in Figures 4-5, the existance of the
crack or discontinuities on the specimen not
only weaken its shear strength as well as its
compressive strenght but they also quicken the
failure of the specimen.
0.10
0.15
0.20
0.25
0.30
0.35
1.00 1.50 2.00 2.50 3.00
w
a
t
e
r

c
o
n
t
e
n
t

(
%
)

Log MS Log NNS
0.10
0.15
0.20
0.25
0.30
0.35
1.00 1.50 2.00 2.50 3.00
w
a
t
e
r

c
o
n
t
e
n
t

(
%
)

Log MS Log NNS
0.12
0.15
0.18
0.21
0.24
0.27
1.0 1.5 2.0 2.5 3.0 3.5
Log Matric suction (kPa)
W
a
t
e
r

c
o
n
t
e
n
t

(
%
)
IM (Intact)
PCM (Pre-crack)
0.15
0.17
0.19
0.21
0.23
0.25
0.27
0.29
1.0 1.5 2.0 2.5 3.0 3.5
Log Net Normal stress (kPa)
W
a
t
e
r

c
o
n
t
e
n
t

(
%
)
PCN (Pre-crack)
7
CONCLUSSION

Based on the data obtained and the analysis
carried on the following conclussion might be
drawn from this experimental study.
1. The specimens tested demonstrated the
typical of brittle hard clay and exhibit
elastic only behaviour .
2. Shear strength of the intact specimens
were higher along their axial strain than
that the pre-crack specimens.
3. Pronounced and lower value of peak shear
strength were occurred to the precrack
specimen than that the intact specimen.
4. The compressive strength of the pre-
crack specimens were lower than the
intacts specimen indicated by their higher
void ratio change index as well as their
water change index than the intact
specimens.

ACKNOWLEDGEMENT

The author would like to give
acknowledgment to all those involved in the
project, especially to Professor Hamid Nikraz
from Curtin University of Technology,
Professor Kwang Wei Lo from National
University of Singapore and Dr Min Min Zhao
for their valuable advice.

REFERENCES

Alshibli, K. A. and Akbas, I. S. (2007). Strain
Localization in Clay: Plane Strain Versus
Triaxial Loading Conditions.
Geotechnical Geological Engineering.
No. 25, p. 45-55.
Alshibli, K. A., Godbold, D. L., and Hoffman,
K. (2004). The Louisiana Plane Strain
Apparatus for Soil Testing. Geotechnical
Testing Journal, ASTM, Vol. 27, no. 4, p.
337-346.
Alshibli, K. A. and Sture, S. (2000). Shear
Bands Formation in Plane Strain
Experiments of Sand. Journal of
Geotechnical and Geoenvironmental
Engineering, ASCE, No. 21167.
Atkinson, J. H. and Bransby, P. L. (1982).
The Mechanics of Soils: an Introduction
to Critical State Soil Mechanics. 1
st
ed.,
McGraw-Hill, New York.
Anderson, T. L. (2004). Fracture Mechanics:
Fundamental and Aplications. 3
rd
ed.,
CRC Press, London.
Bishop, A. W. and Henkel, D. J. (1962). The
Measurement of Soil Properties in The
Triaxial Test. Arnold, London.
Bizzarri, A., Allersma, H. G. B., and Koehorst,
B. A. N. (1995). Preliminary Tests on
Soft Clay with a Biaxial Apparatus.
Proceedings of the 1995 International
Symposium on Compression and
Consolidation of Clayey Soils, Part 1 (of
2).
Drescher, A., Vardoulakis, I., and Han, C.
(1990). A Biaxial Apparatus for Testing
Soils. Geotechnical Testing Journal,
GTJODJ, Vol. 13, p. 226-234.
Fauziah, M. and Nikraz, H. (2007). Biaxial
Testing of Overconsolidated Clay.
Proceeding of The 1
st
International
Conference of European Asian Civil
Engineering Forum, Pelita Harapan
University, Jakarta.
Fauziah, M. and Nikraz, H. (2007). Stress-
strain Behaviour of Overconsolidated
Clay Under Plane Strain Condition.
Proceeding of 10
th
Australia New
Zealand Conference on Geomechanics,
AGS, Brisbane, Australia, p. 148-153.
Fauziah, M. and Nikraz, H. (2008). The
Behaviour of Unsaturated Compacted
Clay Under Plane Strain Condition.
Proceeding of the 3rd International
Conference on Evaluation, Monitoring,
Simulation, Management and
Remediation of the Geological
Environment and Landscape, WIT Press
UK, The New Forest UK, p. 77-85.
Fauziah, M. and Nikraz, H. (2008). Plane
Strain Testing on Properties of
Unsaturated Compacted Clay.
Proceeding of Geo-Chiang Mai 2008, An
International Conference on Geotechnical
Engineering, Chiang Mai, Thailand, p.
157-164.
Fredlund, D. G. and Rahardjo, H. (1993).
Soil Mechanics for Unsaturated Soil.
John Willey & Sons, Inc.
8
Fredlund, D. G. and Morgenstern, N. R.
(1976). Constitutive Relation for Volume
Change in Unsaturated Soil. Canadian
Geotechnique Journal, Vol. 17, No. 3, p
261-276.
Fredlund, D. G. and Morgenstern, N. R.
(1977). Stress State Variables for
Unsaturated Soils. Journal of
Geotecnical Engineering, ASCE, Vol.
103, No. 5, p 447-466.
Green, G. E. and Reades, D. W. (1975).
Boundary Conditions, Anisotropy and
Sample Shape Effects on the Stress-strain
Behaviour of Sand in Triaxial
Compression and Plain Strain,
Geotechnique, Vol. 25, No. 2, p. 333-356.
Han, C. and Drescher, A. (1993). Shear
Bands in Biaxial Tests on Dry Coarse
Sand. Soils and Foundations, Japanese
Society of Soil Mechanics and Foundation
Engineering, Vol. 33, No. 1, p. 118-132.
Han, C. and Vardoulakis, I. G. (1991). Plane
Strain Compression Experiments on
Water-saturated Fine-grained Sand.
Geotechnique, Vol. 41, No. 1, p. 49-78.
Ingraffea, A. R. (1987). Theory of Crack
Initiation and Propagation in Rock. In
Fracture Mechanics of Rocks, ed by B.K.
Atkinson, Academic Press, London, p. 71-
110.
Lee, K. L. (1970). Comparison of Plane
Strain and Triaxial Tests of Sand.
Journal of the Soil Mechanics and
Foundation Division, ASCE, Vol. 96, No.
3.
Lee, K. L. and Seed, H. B. (1964).
Discussion on Use of Free End in
Triaxial Testing on Clays. ASCE, Vol.
91, No. 6, p. 173-177.
Lo, K. W., Mita, K. A., and Thangayah, T.
(2000). Plane Strain Testing of
Overconsolidated Clay. Research
Report, Department of Civil Engineering,
National University of Singapore.
Lo, K. W., Nikraz, R. H., Thangayah, T., and
Zhao, M. M. (2005). An Elastoplastic
Shear Fracture Model for Soil and Soft
Rock. Proc of the 11th International
Conference on Fracture at
www.icf11.com/proceeding/TOPIC/topic.
htm
Marach, N. D., Duncan, J. M., Chan, C. K.,
and Seed, H. B. (1984). Plane Strain
Testing of Sand. Laboratory Shear
Strength of Soil, ASTM STP 740, p. 294-
302.
Mochizuki, A., Min, C., and Takahashi, S. A.
(1993). A Method for Plane Strain
Testing of Sand. Journal of Japanese
Geotechnical Society, No. 475, p. 99-107.
Rowe, P. W. and Barden, L. (1964).
Importance of Free Ends in Triaxial
Testing. ASCE, Vol. 90, No.1, p. 1-27.
Taylor, D. W. (1941). 7
th
Progress Report on
Shear Strength to US Engineers.
Massachusetts Institute of Technology.
Viggiani, G., Finno, R. J., and Harris, W. W.
(1994). Experimental Observations of
Strain Localisation in Plane Strain
Compression of a Stiff Clay. In
Localisation and Bifurcation Theory for
Soils and Rocks, Chambon et.al., Eds.,
Balkema, Rotterdam, p. 189-198.
9
Compressive Strength Assessment of Concrete Structures from Small Core
by Point Load Test

Achfas Zacoeb
1
and Koji Ishibashi
2

1
Department of Civil Engineering, Brawijaya University, Indoneia, email: zacoeb_a@yahoo.com,
2
Professor, Department of Civil
Engineering, Saga University, Japan, email: isibasit@cc.saga-u.ac.jp

Abstract: To assess a compressive strength from existing concrete structures by core drilling are
usually gathered with a diameter specimen of 100 mm or three times of maximum coarse aggregate
size and examined by uniaxial compressive strength (UCS). It is relatively difficult to gather a large
sized core, and a pit place will be limited by main members. To get an alternative solution with
smaller specimen, point load test (PLT) has been selected which is a simple test and widely
accepted in rock materials research, but relatively new in concrete. The reliability of PLT is
examined by extracting a lot of core drilled specimen from ready mixed concrete blocks with
maximum coarse aggregate size, G
max
of 20 mm in representative of architectural structures and 40
mm in representative of civil structures on the range of concrete grade from 16 to 50.
Compressive strengths were classified into general categories, conversion factors were determined,
and scattering characteristics were also investigated. The relationship between point load index
(I
S
and compressive strength of concrete core specimen (f
cc
) can be written as linear approximation
as f
cc
= k.I
S
C.

Keywords: Strength assessment, standard specimen, small core, point load index, and linear
approximation.

INTRODUCTION

One of the reliable tests for assessing insitu
strength of concrete is coring. Coring may
prove expensive and the holes have to be
backfilled, but the resulting data are usually
accepted as the best evidence of the condition
of the concrete in place. It is established in JIS
A1107 (1993) that a core drilled specimen
diameter of 100 mm or three times of
maximum coarse aggregate size from a
concrete structure member should be taken for
performing strength evaluation. Small cores
are often used as substitutes for large cores to
test concrete strength. They have the
advantages of being easily drilled and cut,
minimum damage to structures, and a lower
capacity machine is needed (Ruijie, 1996).
The main parameter for characterizing a
concrete in engineering practice is
compressive strength. Ibragimov (1989), the
maximum aggregate size is considerable as
played role for affecting the properties of
concrete. The standard laboratory test ussualy
requires a standard specimens, so indirect test
are needed. The PLT is intended as an index
test for the strength classification of rock
materials, but it may also be widely used to
predict other material strength parameter. It is
an attractive alternative method, because it can
provide similar data at a lower cost, a simple
preparation of specimen, and possibility on
field application ISRM (1985), in order to
estimate UCS indirectly, index-to-strength
conversion factors are developed. Richardson
(1989) conducted a point load tests of cast
specimens with various diameters. Zacoeb et
al. (2007) showed a strong correlation
between point load index of
core drilled
specimen (I) and compressive strength of
concrete core (f). Ishibashi et al.
(2008)
investigated the influence of h/d ratio and
maximum aggregate size (G
max
) on concrete
core specimen by using PLT. Many research
works had been conducted to acknowledge
with regard to PLT and resulted in widely used
10
Zacoeb, Ishibashi: Compressive Strength Assessment of Concrete Structures From Small Core by Point Load Test
and other parameters. However, more
experimental works helps to substantiate the
existing correlation.

FINITE ELEMENT ANALYSIS

Broch, et al. (1972) started with a simple
formula taking an idealized failure plane of
diame-tric core sample as shown in Figure 1.
From this figure can be taken into account as
conceptual model for derivation on point load
index equation as:

I
S
=
P
n
2

where, I
S
: point load index (MPa)
P : load (N)
D : diameter of specimen (mm)

By taking the circular area of the core into
account, an argument can be made that
Equation 1 should be written as:

I
S
=
4P
nJ
2

The user of this test soon noticed, that the
results of a diametric test were about 30%
higher than those for an axial test using the
same specimen dimensions. Broch et al. (1972)
and ISRM (1985) suggested acknowledge this
difference by applying a size correction and
introducing the equivalent core diameter of D
e
.
Hence, the Equation 2 can be rewritten as:
I
S
=
P
c

The variations of IS with specimen size and
shape lead to introduce a reference index I
S(50)
which corresponds to the I
S
of a diametrically
loaded rock core of 50 mm diameter (Broch et
al.1972). Accordingly, initial I
S
values are re-
duced to I
S(50)
by size correction factors deter-
mined from empirical curves as a function of
d. It is indicated that the considerably larger
shape effect should be avoided by testing spe-
cimens with specified geometries. ISRM
(1985) proposed a new correction function
which ac-counts for both size and shape
effects by utiliz-ing the concept of equivalent
core diameter (D
e
). This function, known as
geometric correc-tion factor F is given by:

I
S(50)
= FI
S

where F : the geometric correction factor

F = _
c
Su
]
0.45

The unique point load index can be obtained
by applying a size correction for the specimen
as point load index of I
S
varies with core
specimen diameter of D
e
. The size-corrected
point load index of I
S(50)
for each specimen is
defined as the value of IS that would have
been measured on a standard specimen
diameter of D
e
=50 mm. In the case of
specimen diameter of DeLABORATORY
WORKS other than 50 mm, size correction
must be calculated by using of Equation 5.

LABORATORY WORKS

The concrete block for core specimen
extracting were sized of 300 mm x 300 mm x
600 mm made from ready-mixed normal
concrete with typical slump range value from
8 to 12 cm for most application as workability
control and divided into two groups as shown
in Table 1. For curing, all concrete block
specimens were covered with plastic sheets
and the humidity was set for about a week.
Commonly in Japan, for architectural
Figure 1. Specimen diametric of PLT
on cores
(2)
(1)
(4)
(5)
(3)
Idealized failure plane
Loading direction
d
Central axial
direction
h/2 h/2
11
structures such as building construction is
using the maximum coarse aggregate size,
G
max
of 20 mm. While for civil structures such
as pier, abutment, bridge deck and check dam
is using the maximum coarse aggregate size,
G
max
of 40 mm.

Table 1. Group of concrete block
Group
Gmax
(mm)
Grade Cement Type
I 20
16
OPC (Ordinary
Port- land
Cement)
21
24
36
50
II 40
16
PBSC
21
24
30

Core specimen diameter of 125, 100, 50 and
35 mm were extracted from the above men-
tioned concrete block with the electric core
pull-ing out machine. The wet type that used
by flowing some water during the core drilled
process is applied, and the extraction speed
was assumed to be about 4 cm/min. The direc-
tion of extraction is considered as the direction
of concrete placing as vertical direction in
assumption of practical work in construction.
The situation of core specimen extraction is
shown in Figure 2.

The core specimen with h/d ratio of 1.5 and
2.0 were selected as core specimen of PLT in
this study. In order to establish a specific h/d
ratio, core specimens were cut both ends with
a concrete cutting machine to become a fixed
height (h). The total number collection of each
specimen is shown in Table 2.

Table 2. Total number of core specimens
Group Grade d (mm) h/d Total
I
16
35
1.5 135
2.0 135
50
1.5 105
2.0 99
21
35
1.5 90
2.0 90
50
1.5 60
2.0 60
24
35
1.5 66
2.0 66
50
1.5 59
2.0 58
36
35
1.5 123
2.0 126
50
1.5 111
2.0 108
II
16
35
1.5 126
2.0 126
50
1.5 85
2.0 87
21
35
1.5 154
2.0 138
50
1.5 82
2.0 79
24
35
1.5 113
2.0 113
50
1.5 87
2.0 86
30
35
1.5 157
2.0 172
50
1.5 113
2.0 108

The specimen in PLT is taken and loaded
between two hardened steel cones. The system
consists of a small hydraulic pump, a hydraulic
jack, a pressure gauge and interchangeable
Concrete placing
direction
Figure 2. Core specimen extractions
12
testing frame of very high transverse stiffness.
Spherically truncated, conical platens of the
standard geometry shown in Figure 3 are to be
used with the cylinder area of 14.52 cm
2
. The
platens should be of hard material such as
tungsten carbide or hardened steel so that they
remain undamaged during testing (ISRM,
1985).

The core specimen in this study is gradually
loaded by activating the hand pump until
failure and determined this load as P. The
S
was calculated by using
Equation (3) and for core specimen diameter
of 35 mm was corrected to the standard core
diameter as point load index of I
S(50)
for core
specimen diameter of 50 mm by using
Equation (4). The examination is conducted
by using PLT machine with oil pressure
cylinder type and maximum load capacity of
98 kN as shown in Figure 4.

RESULTS AND DISCUSSION

Compressive Strength
From a concrete block is extracted a core
specimen with diameter of 100 mm as
minimum requirement and 125 mm as three
times of maximum coarse aggregate size (JIS
A1107, 1993), cut both ends of the core with a
concrete cutting machine, end face polished,
processed it to become specific h/d of 2.0 and
examined by UCS test. The mean value of
compressive strength of concrete core, f
cc
is
shown in Table 3, and assumed these values as
reference on this study.

Table 3. Compressive strength
Group Grade
Age
(days)
f
cc

(MPa)
I
16 161 15.6
21 337 35.4
24 73 31.5
36 177 42.9
50 78 51.5
II
16 188 21.6
21 173 22.4
24 532 34.4
30 118 32.2

Point Load Index
The mean values of PLT were computed for
both diameter sizes as shown in Table 4.
Scattering characteristics were also
investigated by mentioning of CoV (coefficient
of variation). The CoV is the degree to which a
set of data points varies. When assessing
precision, the lower of CoV percentage, the
better of precision between replicates. For
Group I, the level of CoV is almost same or
less than that on actuality of ready-mixed
concrete product (Saga, 2008) from 10 to
15%. It can be stated that the test results are
Figure 3. Point load cone platen
Figure 4. Setup of PLT
13
satisfy enough. For group II, the CoV is bigger
than the requirements (except for concrete
grade of 30).

Table 4. Point load index
Group Grade
d
(mm)
h/d
I
S

(MPa)
CoV
(%)
I
16
35
1.5 2.27 11.4
2.0 2.32 13.3
50
1.5 1.87 10.7
2.0 1.93 10.8
21
35
1.5 3.24 11.4
2.0 3.31 12.4
50
1.5 2.57 9.7
2.0 2.61 11.1
24
35
1.5 3.21 12.4
2.0 3.28 13.1
50
1.5 2.71 8.1
2.0 2.77 9.3
36
35
1.5 3.61 11.9
2.0 3.69 13.5
50
1.5 3.06 10.4
2.0 3.02 10.9
50
35
1.5 3.95 8.1
2.0 4.05 9.6
50
1.5 3.27 8.2
2.0 3.34 8.9
II
16
35
1.5 2.20 26.1
2.0 2.30 31.9
50
1.5 1.85 18.7
2.0 1.95 18.1
21
35
1.5 2.52 23.0
2.0 2.55 24.3
50
1.5 2.12 21.0
2.0 2.00 18.0
24
35
1.5 2.90 26.3
2.0 2.92 27.0
50
1.5 2.31 18.2
2.0 2.43 18.2
30
35
1.5 2.80 24.7
2.0 2.85 19.7
50
1.5 2.43 8.5
2.0 2.47 7.6
For Group I, the level of CoV for h/d of 1.5 is
smaller than h/d of 2.0. It can be stated that h/d
ratio of 1.5 is better than h/d of 2.0 for making
a PLT specimens from core drilled extraction.
While for core specimen diameter, d is better
using 50 mm than 35 mm, because the level of
CoV is also smaller. Beside this reason, it is
also fulfilled with the standard core diameter
requirements of 50 mm. For all groups, it is
possible and acceptable for using a core
diameter of 50 mm and h/d ratio of 2.0 as PLT
specimen with results in the range of CoV
from 8 to 18%. Application of PLT for small
diameter of core specimen is not suggested for
d/G
max
ratio below 1.25, considering the CoV
results for G
max
of 40 mm and d of 35 mm are
larger than 20%.

Correlation Between Point Load Index and
Compressive Strength
Point load index of core specimen diameter of
50 mm, I
S(50)
is determined as standard value.
Hence, the value of different core specimen
diameter, I
S(35)
should be corrected in order to
show a relationship with I
S(50)
by using
Equation (4) and (5). By correcting the point
load index of I
S(35)
and assuming as standard
core specimen diameter of 50 mm, will add the
number of data for analysis of point load index
I
S(50)
. The new result for this combination is
shown in Table 5 corresponding with the
compressive strength of concrete core (f
cc
) for
each grade.

Table 5. Point load index and compressive
strength
Group Grade
fcc
(MPa)
I
S
of h/d
1.5 2.0
I
16 15.6 1.86 1.93
21 35.4 2.57 2.61
24 31.5 2.71 2.77
36 42.9 3.07 3.03
50 51.5 3.27 3.34
II
16 21.6 1.85 1.95
21 22.4 2.12 2.00
24 34.4 2.31 2.43
30 32.2 2.43 2.47
14

Figure 5. Second order of polynomial regression
Figure 6. Linear regression
Figure 7. Linear approximation for I
S(50)
to f
cc

15
The correlation between point load index,
I
S(50)
and compressive strength of concrete
core, f
cc
for both of groups is shown
graphically in Figure 5 and 6. It is clearly
evident to show the correlation by proposing a
second order of polynomial and linear
regression, respectively. It is proven by
showing the square value of correlation
coefficient which judges the effectiveness of a
second order of polynomial approximation
curve for h/d = 1.5 is thought to be similar
for h/d = 2.0. Linier regression also showed
the same trend of effectiveness except for
group II.
JIS A5308 (2003) gives the compressive
strength range of ready-mixed concrete in field
application from 18 to 45 MPa. Hence, the
application of PLT for estimating insitu
strength of concrete structure should be
confirmed in this range. So, the correlation
was limited to this range for core specimen
diameter of 35 and 50 mm as shown in Figure
7. When using the linear regression as shown
in Figure 6, the approximation line does not
intercept in the origin point. However, Figure
7 shown that the fitted curve will pass through
the origin which aims to establish the relation
of the whole area would be overestimated. It is
preferable to using linear approximation than
other modes in order to minimize the standard
for the assessment of risk.

The New Geometric Correction Factor
By considering the Equation (4) and (5) were
proposed for rock specimen, so it is not
suitable for concrete regarding the issue of
homogeneity. The previous section already
mentioned that maximum coarse aggregate
size in concrete will affect the results of point
load index. A new correction factor of F is
proposed by following the format of previous
Equation as:

I
S(50)
= _
J
Su
]
x
I
S

The value of X can be generated by using data
from group I for core specimen diameter of
35mm. The selection of this data was
considered more reliable by showing a lower
CV. The solution is simple, because the nature
of linear approximation as the origin. The
exponent value of X is calculated as 0.53 with
coefficient of correlation is 0.982. Finally, the
expression geometric correction factor for
concrete core specimen is given by:

F = _
c
Su
]
0,53

Table 6 shows the absolute relative error be-
tween experimental and estimation values for
S(35)
to become standard
S(50)
by using Equation (8).
The results are satisfied enough by showing a
value of absolute relative error less than 5% in
the case of d of 35 mm and G
max
of 20 mm.

Table 6. Experimental and estimation of I
S(50)

f
cc

(MPa)
h/d
Point load index (MPa)
Error
(%)
I
S(35)
I
S(35)
a
I
S(50)
b

15.6
1.5 2.27 1.88 1.87 0.53
2.0 2.32 1.92 1.93 0.52
31.5
1.5 3.24 2.68 2.57 4.28
2.0 3.31 2.74 2.61 4.98
35.4
1.5 3.21 2.66 2.71 1.85
2.0 3.28 2.72 2.77 1.81
42.9
1.5 3.61 2.99 3.06 2.29
2.0 3.69 3.06 3.02 1.32
51.5
1.5 3.95 3.27 3.27 0.00
2.0 4.05 3.35 3.34 0.30
*
)
a = estimation
b = experimental

Recalculation procedure is conducted by using
a new Equation (8) for correcting point load
index of core specimen diameter of 35 mm and
performing linear regression analysis to
propose a formula of compressive strength
esti-mation for equivalent core diameter of 50
mm as f
cc
= k.I
S
- C as shown in Table 7. The
coefficient of correlation, R
2
also shows an
improvement in strong relationship between
I
S(50)
and f
cc
.
(6)
(7)
16
In consideration with the standard compressive
strength that used as linear approximation for
index-to-strength conversion factor k. The k
value is calculated by divided the
compressive strength (f
cc
) with point load
index (I
S
). While C is constant depend on the
linear regression equation.

Table 7. Formula of estimation
Group h/d Formula of Estimation R
2

I
1.5 f
cc
= 24.4I
S
30.3 0.953
2.0 f
cc
= 24.9I
S
32.7 0.953
II
1.5 f
cc
= 20.8I
S
16.7 0.928
2.0 f
cc
= 22.3I
S
22.0 0.979

CONCLUSION

Based on this study, it can be concludedthat an
approximation curve showed a strong corre-
lation between point load index (I
S
) and com-
pressive strength (f
cc
) for core diameter of 35
and 50 mm with height and diameter ratio of
h/d=1.5 and 2.0. In addition for reference
index I
S(50)
, it can really deal with a linear
approximation. Considering the issue of
homogeneity that concrete is composite
material, a new correction factor is proposed
for core specimen diameters differ from 50
mm as:

F = _
c
Su
]
0,53

To estimate a concrete compressive strength
can be conducted with proposed equation as
f
cc
= k.I
S
- C.
There is a prospect that PLT can be applied as
indirect method to estimate a compressive
strength on concrete structure. Application of
PLT for insitu concrete compressive strength
estimation should be confirmed in the range of
compressive strength of ready-mixed concrete
product from 18 to 45 MPa. Considering the
maximum coarse aggregate size of G
max
in
concrete, new criterion is proposed by
determining the minimum value of d/G
max
ratio
should not less than 1.25.
ACKNOWLEDGEMENTS

The research work for this study was
conducted by the first author during Doctor
Course in the Laboratory of Structural
Engineering and Mechanics, Department of
Civil Engineering, Saga University, Japan
under supervision of the second author.
Thanks to the Government of Japan for
financial support trough the Monbukagakusho
scholarship.

REFERENCES

Broch, E. and Franklin, J. A. (1972). The
Point Load Strength Test. International
Journal of Rock Mechanics and Mineral
Sciences, Vol. 9(6), p. 669-676.
ISRM Commission on Testing Methods.
(1985). Suggested Method for
Determining Point Load Strength.
International Journal of Rock Mechanics,
Mineral Sciences and Geomechanics,
Abstract Vol. 22, p. 51-60.
Ibragimov, A. M. (1989). Effect of the
Maximum Size of Coarse Aggregate on
the Main Parameters of Concrete.
Journal of Power Technology and
Engineering, Vol. 23, p. 141-144.
Ishibashi, K., Zacoeb, A., and Ito, Y. (2008).
Influence of Coarse Aggregate Size on
The Estimation of Compressive Strength
of Concrete by Point Load Testing.
Journal of Structures and Materials in
Civil Engineering, Japan, Vol. 24, p. 108-
115.
Japanese Industrial Standard. (1993). JIS
A1107: Method of Sampling and Testing
for Compressive Strength of Drilled Cores
of Concrete. Referenced JIS Standard,
Japan.
Japanese Industrial Standard. (2003). JIS
A5308: Ready-mixed Concrete.
Referenced JIS Standard, Japan.
Richardson, D., N. (1989). Point Load Test
for Estimating Concrete Compressive
Strength. ACI Materials Journal, Vol.
86(4), p. 409-416.
Ruijie, K. L. (1996). The Diameter-
compression Test for Small Diameter
(8)
17
Cores. Journal of Materials and
Structures, Vol. 29(1), p. 56-59.
Saga Fresh Concrete Industrial Union. (2008).
Ready-mixed Concrete Products in Saga
Prefectures, Fiscal Year of 2007. Quality
Audit Report, Japan.
Zacoeb, A., Ishibashi, K., and Ito, Y. (2007).
Estimating the Compressive Strength of
Drilled Concrete Cores by Point Load
Testing. Proceeding of the 29th JCI
Annual Meeting, Sendai, Japan, July 11-
13, 2007, p. 525-530.

18
Essential Strategies To Promote Earthquake Safer Non-Engineered Housing: An Indonesian Perspective
Essential Strategies to Promote Earthquake Safer Non-Engineered
Housing: An Indonesian Perspective

Setyo Winarno

Senior lecturer at Islamic University of Indonesia, Yogyakarta, Indonesia,
email: setyo_winarno@ftsp.uii.ac.id

Abstract: Current evidence suggests that the most effective measure to protect peoples lives
against major earthquakes is to build safer houses, particularly on non-engineered residential
houses. Promoting earthquake safer housing requires a thorough understanding and synergetic
efforts on both controlling by seismic codes and regulation and addressing the local practices,
opportunities, limitations, and problems. This paper explores some appropriate strategies to promote
earthquake safer housing by improving non-engineered construction practices that ensures safety
against disasters. Information and data were gathered in Indonesia through a direct observation to
the stricken areas following the 2006s Yogyakarta earthquake, a series of in-depth interviews with
multidisciplinary stakeholders, and also two workshop events. This study found that promoting
safer non-engineered housing requires a comprehensive strategy that includes technical and non-
technical measures. In the top priority, the role of government is extremely crucial in tackling the
compelling problem by institutionalisation of seismic risk reduction activities at national and local
levels. For grass root communities, wide ranging reform in non-engineered construction practice
should be delivered by a locally adapted technique, culturally accepted and compatible, local
resource-based, not burdensome, and less bureaucratic. This fits into the existing community
structures without any friction, value everyones unique contribution, and also break powerful
psychological barriers as well as emphasize the importance of sustainability. If people perceive that
the implementation of the codes is practically possible and achievable and they can control it, this
strategy will have a tremendous effect on how well they can cope with change. In all, the most
challenging part of successful promotion of earthquake safer housings is not finding the tools, but
realizing and accepting that a seismic event is simply a real fact for all people who live in a seismic
prone area. Therefore it is better to grasp it rather than denying or ignoring it.

Keywords: Earthquake, essential strategies, and non-engineered housings.

INTRODUCTIONS

Recently, strong earthquakes have occurred
throughout the world. Protracted study of the
true nature of the seismic risk has shown that
most of the loss of life has occurred due to the
collapse of non-seismic resistant buildings
made of heavy materials, classified as 'non-
engineered' - in simple terms: unsystematically
designed and poorly built structures, in both
developing and developed countries (Mansouri
et al., 2002; Corpuz, 1990; Lee et al., 2003;
Sarwidi, 2001). An example can be seen when
a severe ground shaking hit Yogyakarta and
Central Java, Indonesia in May 2006, and left
over 150,000 private houses, mostly low-rise
non-engineered houses, totally destroyed and
claimed to around 6.000 fatalities
(BAPPENAS, 2006). The strong correlation
between the large number of deaths and the
collapse of non-engineered buildings suggests
it is imperative to improve seismic resistance
for both existing and new non-engineered
housing.
While seismic-related building codes for
simple housing have been available for use in
Indonesia for many years (Boen, 1978), as in
other countries, current evidence has shown
that earthquakes still continue to cause tragic
events with high death tolls. Obviously, there
is a broad gap between the existence of
appropriate building codes and their effective
19
widespread application (Comartin et al., 2004).
It is clear that unless something is done
quickly to significantly implement seismic
codes in Indonesia, earthquakes will continue
to tragically inflict greater human and
economic losses. Based on the fact that non-
engineered buildings as the majority of
building stocks (Winarno and Sarwidi, 2010),
the existence of seismic codes, the reluctance
of grass root people to implement seismic
codes, and the growing concern for the seismic
safety, this paper reports on research that has
explored fundamental strategies to promote the
implementation of seismic codes on non-
engineered construction in Indonesia. This
paper elaborates a subject that has not yet been
sufficiently researched in Indonesia. By better
understanding the varied facets of the
particular construction type, the problems in
the non-engineered construction practice for
the implementation of seismic codes can be
better resolved. Hence, non-engineered
buildings without seismic codes will
eventually be replaced by more reliable
constructions built with seismic resistant
attribute. The findings presented herein should
not be seen as conclusive, but rather for the
development of practical initiatives to
encourage ways in which the various
disciplines can become more involved. In this
paper, the criterias of earthquake safer non-
engineered houses focus on self-construction
of non-engineered houses under support,
supervision, and training from engineer or
expert and also the use of local labour and
material.
The overarching aim of the research study was
to examine the essential strategies to promote
earthquake safer non-engineered housing
within the context of current practice in
Indonesia. In sequence with the aim of the
research, the objectives of the research were
(1) to examine best practice of a range of
stakeholders regarding sustainable and
effective dissemination of implementation of
seismic codes at grass root level, and (2) to
study the potential role from various
stakeholders as agents of change.
Through understanding these important
strategies, all stakeholders can take a part to
prevent a widespread persistence of grass root
communities not to implement seismic codes.
Indeed, the implication of a general
assessment of such grass root communities is
that the manual of seismic features is
understood in different ways and not through
standard curricula. Therefore, all perspectives
need to be involved deliberately to resolve this
problem.

METHODOLOGY

Based on wider literature, involvement of
multidisciplinary stakeholders in the seismic
risk reduction of non-engineered buildings is
imperative to ensure sustainability. Those
Figure 1. NVivo screen display of the nodes created from the interview and workshop events
20
stakeholders includes nine types of
multidisciplinary stakeholders, i.e.
researchers/scientists, small and medium
contractors, foremen, government officials,
businessmen, educators, Non Government
Organizations (NGOs), community leaders,
and reporters (IUDMP, 2001; CEEDEDS,
2004; SCEC, 2002; GREAT, 2001; Dixit,
2003). Thus, the above stakeholders were
selected as respondents in the research. The
appropriateness of each respondent was
determined by their role, responsibilities and
normal activities within their own organization
and the level of experience in the specific
subject. Information and data were collected
through a direct observation to the stricken
areas following the 2006s Yogyakarta
earthquake, a series of in-depth interviews
with multidisciplinary stakeholders, and also
two workshop events.
Firstly, field investigations in regions damaged
by the Yogyakarta earthquake were
undertaken during the reconstruction stage to
obtain a more enhanced understanding of the
problems involved. Through this investigation,
nine respondents, representing a mix of nine
types of stakeholders, were chosen as
interviewees. The interview method was
appropriately chosen in this study to collect
factual in-depth information, opinions, and the
story behind the respondents experiences.
Then, the interview findings were augmented
via two workshop events conducted in
Yogyakarta City (with 13 participants) and
Bengkulu City (with 12 participants), which
were attended by key people and at the leading
edge of decision making within each city, in
order to ensure that individual's assessment
was factual and logically sound. The workshop
events took place under the name Workshop
on Seismic Risk Management of Non
Engineered Buildings in collaboration with
Master Program in Civil Engineering,
Universitas Islam Indonesia, Yogyakarta and
Civil Engineering Department, Universitas
Bengkulu. The overall data collected was
analyzed with the assistance of NVivo
software programme. The software helped to
code the data and identify themes and/or
patterns generated.
FINDINGS

During the interview and workshop events,
three guiding principles for a sustainable and
successful dissemination process emerged,
which may have been greatly neglected in the
past. The above three principles confirm that:
(1) the government role is the backbone of the
dissemination initiative mentioned by 8
interviewees and 22 workshop participants, (2)
the dissemination channel through the existing
social bond is imperative mentioned by 8
interviewees and 16 workshop participants,
and (3) the message should convince people
that they have control over the implementation
of seismic codes - mentioned by 6
interviewees and 17 workshop participants. If
the three principles are applied rigorously, the
dissemination process will obtain great
achievement. This is truly beyond technical
capacity. However, sometimes the key
government officials and the disseminators,
who are usually from a wealthy background,
do not persevere in conducting dissemination
to the low-medium income population, as they
are not on the poor peoples side and also do
not gain any clear financial and political
benefit. Figure 1 depicts the research findings
with the assistance of NVivo software.
Overall, the findings from the research data
address each of the objectives stated
previously, as follows:

The Current State of Non-Engineered
Construction Practice and Guiding
Principles of Sustainable and Effective
Dissemination in the Grass Root Initiatives
Field investigation and interviews with local
practitioners undertaken in the stricken area
after the Yogyakarta earthquake revealed some
inadequate construction materials. The quality
of materials, in fact, rested largely on the
knowledge of the local builders, foremen,
masons, and carpenters. They continued to
practice the traditional approach and were
often reluctant to accept new techniques. They
felt that what they did was in line with their
belief and there was no strict regulation which
confronted their practices. The resulting
21
mortar and concrete were generally of poor
quality. All too often, there was incorrect
beam-column connection detail.
Interviewees from government staff,
researchers/scientists, contractors, and
foremen revealed that information on building
using seismic codes with seismic codes has
been available for public. In reality, the
seismic codes were not implemented widely,
particularly within non-engineered buildings.
From the point of view of contractors and
foremen, it was true that there were regulations
with limited enforcement and no
accountability. The government has not been
able to implement even the existing seismic
codes because of a lack of suitable
implementation mechanism and limited
resources for building inspection and control.
Also, the respondents who were represented
by the community leaders, NGOs, teachers,
and businessmen said that the building code
was not disseminated down through to the
grass root communities. To overcome the
limited resources on government, obviously,
this requires a comprehensive strategy that
includes various multidiscipline interventions.
Combining data from interviews and
workshop event, three guiding principles of
sustainable and effective dissemination in the
grass root initiatives emerged to break a
widespread reluctance of grass root
communities to implement seismic codes, as
follows.
Government should enact as a proactive
backbone of the dissemination initiative
To disseminate the benefits of seismic codes in
real construction successfully and
continuously, almost all interviewees and
workshop participants mentioned the
importance of the governments role, and even
mentioned it as the most important factor.
Although the government is not necessarily
solely resposible for disseminating information
regarding seismic codes, and there is a new
opportunity for multidisciplinary involvement
in this matter, nevertheless, the government
itself should act as a proactive backbone of the
dissemination initiative. The ability of
government to listen, respect, encourage, and
motivate grass root programs initiated by
community groups is indispensable. At the
same time, the government should generate
campaigns or dissemination programs down to
the community members to motivate people to
gradually prepare for the next disaster,
highlighting a shared risk and also enforcing
seismic codes. The governments role is to be
the focal point for law enforcement
(regulation), communication, coordination,
programme monitoring, delivery of
knowledge, and accountability. Also,
government should institutionalise seismic risk
reduction activities at national and local level.
It is believed that, if the government makes a
huge effort to rigorously enforce seismic codes
today in many model houses so people see
what they look like and feel very proud, many
more model houses would soon be replicated
everywhere. When a strong earthquake strikes,
the government will make only a little effort in
response because the blossom of the
implementation of seismic codes thus the
collapse of buildings/houses would be rare,
even avoided. Furthermore, it would ensure
faster recovery and reconstruction.
The dissemination method should utilize
existing social bonds and/or indigenous
methods
Three out of nine interviews and ten workshop
participants stated the importance of social
bonds and indigenous methods when dealing
with communal grass root dissemination. This
finding argues that when government staff,
researchers, scientists, and/or other disaster
management experts gather to elucidate the lay
people, this process will be effective if it uses
or merges into the formal or informal existing
tradition of community meetings. Innovative
initiatives, new synergies, and networks are
easily absorbed over those already established.
This will fit into the existing community
structures without any friction and also value
everybodys unique contribution.
People will be comfortable, happier, and less
worried about being involved in the
dissemination process if they are among
people they have worked with in the past and
with whom they have developed a long-term
22
relationship and have similar beliefs. This
approach also emphasizes the importance of
sustainability. This process breaks powerful
psychological barriers and continues to build
up trust amongst them and also acts as
encouragement for others who have not been
involved in any initiatives in the past. It is a
better and less expensive method for including
communal actions.
In addition, the existing social bond has
shaped the communitys capacity in their daily
lives. Through this bond, community members
can better adjust their own practices to
improve performance. If this dissemination
mechanism is well done and streamlined over
time before disaster strikes, obviously this
channel will be sustainable and will also work
well in the critical stages of disaster response.
Perhaps, the current dissemination method
initiated by government officials uses
government standards which are often not
flexible enough to adjust to various peoples
needs. The government tends to be just
interested in promoting their own status, and
the use of existing social bonds is often greatly
neglected.
The above discussion is about channeling
dissemination by giving a talk or elucidation to
grass root communities by visiting the
community group. On the other hand, the
dissemination technique by inviting
individuals to a meeting, discussion, or
seminar is very suitable for people from
medium-top management in organizations who
have experienced better education because
they have an appropriate level of expertise and
knowledge. This is why the dissemination to
the grass root community by inviting them to
a formal forum, such as a discussion in
government offices, always fails to motivate
them into concrete action. This is likely to be
because the prescriptive mechanisms are not
compatible with their beliefs.
For grass root communities, wide ranging
reform in non-engineered construction practice
should be by a locally-adapted technique,
culturally accepted and compatible, local
resource-based, not burdensome, and less
bureaucratic. This principle is closely related
to the idea of a sense of place, meaning that
feeling is attached to a place, and feeling
secure in all the things that make life truly
meaningful is important for peoples identity
as mentioned by Covenry and Dutson (2006).
The message should be achievable under
control of the people
After the correct dissemination channel is
diligently decided, here, two interviews and
three workshop participants elaborated that the
message about the implementation of seismic
codes should be achievable under the control
of the people. This finding underpins what
Lustig (1997) has found: that for a disaster-
management system to be sustainable, it
should be designed not only to convey the
message to the members of the disaster-prone
community that they are in control, but also
that the system is actually under their control.
Although the seismic event itself, when it
occurs and how big the magnitude will be, is
uncontrollable, most of the components of
seismic risk people face are absolutely
controllable. Community members who live in
a seismic hazard prone area can control the
suffering from seismic tremor, for example,
through implementing such codes in their
houses/buildings and making any other
appropriate preparedness. If people rationalize
that the implementation of seismic codes in
their own houses is the most effective strategy
for minimizing losses, and at the same time,
they perceive it is practical, possible and
achievable and they can control it, this strategy
will have a tremendous effect on how well
they can cope with it.
It is an unwise solution to force people to
leave their beloved homeland because of the
high level of uncertain but inevitable threats of
an earthquake event, as an inconvenient truth.
Thus, the dissemination message should
convince people to devote themselves to living
harmoniously with seismic risk through
matters they perceive they can control. The
ideal solution is to give people a better
understanding about seismic hazard and risk in
a reasonable and rational manner and then
convinces them they can cope and control it
with the proper implementation of seismic
codes. As a result, people become happier and
23
dont worry about living in seismic prone
areas because they have some control over
events. Better understanding of seismic hazard
and risk will encourage people to implement
seismic codes voluntarily without coercion.
Although disseminating new ideas to people is
not welcomed at first, even fought against,
efforts must be made to present persuasive
arguments of the soundness of the protective
and cost-effective measures.
It is important to educate people that the
implementation of seismic codes is simple,
economically feasible, achievable, and
culturally acceptable to obtain their sense of
control over their destinies. If people have a
sense of control and are clever enough to
implement seismic codes properly, then they
can also achieve a sense of ownership. This
sense of ownership is really needed, not only
to encourage and maintain actions voluntarily
in order to generate a culture of prevention,
but also to make community members feel part
of the effort. As a result, people will actively
make contributions to reduce vulnerability.
Figure 2. Putting multidisciplinary stakeholders together as agents of change to share the
seismic risk of non-engineered buildings
24
Through a greater sense of ownership, more
people tend to recognize that reducing seismic
risk needs shared responsibility and shared
effort.
In conclusion, the above finding suggests three
guiding principles for a sustainable and
successful dissemination process, which may
have been greatly neglected in the past. If the
three principles are applied rigorously, the
dissemination process will obtain great
achievement. This is truly beyond technical
capacity.

A Map of Multidisciplinary Stakeholders as
Agents of Change
This section maps out multidisciplinary
stakeholders as agents of change, highlighting
a shared risk and responsibility and focusing
on grass root initiatives with local wisdom, as
later elaborated in Figure 2. Based on the data
collected and the general facts, many forms of
people and organizations can act as agents of
change and have unique roles, which are
highly varied due to the their nature of duty of
care. Therefore, it is important to depict the
influence level of many stakeholders as agents
of change in seismic risk reduction of non-
engineered buildings. The principal aim of this
depiction is to give a better understanding that
many people or organizations have an
important stake in steering the seismic risk
management process. It is hoped that a full
spectrum of tasks and activities and broad
banded partnerships in relation to the
implementation of seismic codes to non-
engineered buildings are well done and
assimilated continuously, integratively, and
harmoniously by a new configuration of
various people and organizations to ensure a
new generation of sustainable seismic risk
reduction.
Most arguments in the interview and workshop
events state directly that the governments role
greatly influences the successful seismic risk
reduction of non-engineered buildings under
their principal aim of public safety in general.
In particular, the top management in local
government, together with the legislative
council, has the authority and the power to
give approval or disapproval for all
development related issues. The top local
authority can communicate with different
department heads and can make them act on
issues of seismic risk which need to be
accomplished. Essentially, local government
should rigorously disseminate and
communicate to their local seismic risks and
the benefit of seismic codes by educating
people to implement them voluntarily; at the
same time, they should enforce the codes
through effective regulation. In addition, this
includes rules for the control of development,
land use regulations, and suitable compliance
mechanisms for building construction. The
quality of their leadership is an indispensable
component of success. They serve as a source
of hope. Here, the role of the government
appears to be as major contributors to the
successful seismic risk reduction of non-
engineered buildings. Finally, the role of the
government (together with the role of the
legislative council) and their concrete actions
in disaster reduction remain the biggest and
top challenge for effective seismic risk
reduction today.
On the other side, sometimes, difficulties
appear when local builders have to change
their traditional approach to fall in line with a
new technique of seismic feature
implementation. Therefore, changing the
practice of builders is very necessary to make
more seismic resistant buildings/houses and, as
a result, they constitute the second agent of
change.
Researchers and scientists can be both the first
and second agents of change. For example, the
researcher/scientist role could be as the most
important of the availability of earthquake data
and development of seismic codes, and they
are the first agent of change in this sense. The
researcher who conducts the cost-effectiveness
of seismic feature implementation can perform
at the second level after the government role to
support law enforcement of seismic codes.
Researchers and scientists can inspire many
aspects of good seismic risk management of
non-engineered buildings, such as the
development of seismic codes suited to the
specific area, introducing earthquake facts for
25
land use planning and a seismic awareness
program. Moreover, social and psychology
researchers can innovate on how to deliver
information, leading to a change in attitude
and a change in behaviour. They also can
serve as a source of new knowledge and
inspiration for those implementing the
changes.
When discussing the effectiveness of social
bonds and how to break psychological barriers
of community perception to achieve a feeling
of control (as mentioned in the previous
findings), it is clear that the role of community
leader is very important to explain the whole
thinking process. In certain cases, the
construction process of residential
houses/buildings is directly under the guidance
of the owner, who often does not have
sufficient skill in the concept of seismic codes
or neglects the workmanship of the builders
due to the time completion pressure and the
overwhelming need for steel bar
reinforcement. Again, in this situation, the
community leader can contribute to the
education of those people and also act as an
intermediary between the government and the
house owner when government staffs enforce
seismic codes during the construction process
of the residential house. In certain cases, the
house owner often acts as a self-builder due to
limited funding available. Here the community
leaders (who also represent the house owner)
are the third agent of change.
The last agent of change belongs to the groups
of educators, NGOs, businessmen, reporters,
and others since many respondents comment
appeared to illustrate aspects of non technical
intervention, where the involvement of many
parties related to non-technical intervention
(excluding government officials) was
reasonably less important.
Obviously, the picture of the agents of change
in Figure 1 is merely a general assessment as
an indicative explanation drawn from the
research data. Probably, certain cases will
illustrate different patterns of the degree of the
influence within larger parties as condition
changes over time. Subsequent investigation is
truly needed to explore further for better
understanding and to validate against any
misunderstanding or bias due to strong
personal opinions.

CONCLUSIONS

In order to reduce seismic risk of non-
engineered buildings, the proper
implementation of seismic codes is imperative,
because flawed construction practice often lies
at the heart of implementation of seismic
codes. Thus, the implementation should be an
integral component of the total construction
process in the first place. Realistically, a
combination of skilled non-engineered
construction participants and the proactive
involvement of government and non-
government organizations as well as many
other parties (technical and non-technical
intervention) are indispensable in promoting
safer non-engineered housing. It needs to be
controlled by seismic codes and regulation. In
the top priority, the role of government is
extremely crucial in tackling the compelling
problem by institutionalisation of seismic risk
reduction activities at national and local level.
There are three principles for a sustainable and
effective way towards the dissemination of
seismic codes for non-engineered construction
practice for the grass roots level: (1) the
government should act as a proactive
backbone of the dissemination initiative
mentioned by 8 interviewees and 22 workshop
participants, (2) dissemination mechanisms
should use the existing social bond and/or
indigenious method mentioned by 8
interviewees and 16 workshop participants,
and (3) dissemination message should
convince the communities that the
implementation of seismic codes is easily
achievable under their control - mentioned by
6 interviewees and 17 workshop participants.
The successful promotion of safer non-
engineered structures needs to sensitise policy
makers (government) toward seismic risk,
followed by the involvement of many
technical and non-technical actors, such as
researchers, scientists, contractors, foremen,
masons, carpenters, businessmen, educators,
NGOs, community leaders, reporters, and
others. Wider recognition is needed that
26
building a culture of disaster prevention should
become everybodys duty of care on a daily
basis to ensure sustainability.
Underlying the research finding, the most
challenging part of successful seismic risk
management of non-engineered buildings is
not finding the tools, but realizing and
accepting that a seismic event is simply a real
fact for all people who live in a seismic prone
area. We cannot predict when an earthquake
will occur or how big the magnitude will be.
Therefore it is better to grasp it rather than
denying or ignoring it. A balance should be
struck between communicating local seismic
risk and the importance of seismic codes to the
people, and enforcing them to implement the
codes; whilst some intense regulation is
imperative, it should be non-prescriptive.

ACKNOWLEDGEMENT

This research would not have been possible
without the support of the Universitas Islam
Indonesia. We thank to Professor Dr. Sarwidi,
and Professor Dr. Widodo for their tireless
contributions to this study.

REFERENCES

BAPPENAS/National Development Planning
Agency. (2006). Preliminary Damage
and Loss Assessment: Yogyakarta and
Central Java Natural Disaster. The
Consultative Group, Jakarta, Indonesia.
Boen, T. (1978). Detailers Manual for Small
Buildings in Earthquake Areas (Manual
Bangunan Tahan Gempa (Rumah
Tinggal)). Yayasan Lembaga
Penyelidikan Masalah Bangunan,
Bandung, Indonesia.
CEEDEDS/The Center for Earthquake
Engineering, Dynamic Effect, and
Disaster Studies. (2004). The Manual of
Earthquake Resistant Building. Project
Report between CEEDEDS and Japan
Government, Yogyakarta.
Comartin, C., Brzev, S., Naeim, F., Greene,
M., Blondet, M., Cherry, S., DAyala, D.,
Farsi, M., Jain, S. K., Pantelic, J., Samant,
L., and Sassu, M. (2004). A Challenge to
Earthquake Engineering Professionals.
Earthquake Spectra, Earthquake
Engineering Research Institute, November
2004, Vol. 20, No. 4.
Corpuz, A. (1990). Some Implications of the
July 16, 1990 Earthquake on Urban and
Regional Planning in the Philippines.
School of Urban and Regional Planning,
University of the Philippines, Philippines
at http://www.phivolcs.dost.gov.ph/Earth-
quake/1990LuzonEQ_Monograph/pp287/
pp290_291.html; accessed on 14 October
2004.
Covenry, I. and Dutson, T. (2006). Sense of
Place in Northern England. International
Journal of Biodiversity Science and
Management, Vol. 4, No. 2, p. 213-217.
Dixit, A. M. (2003). The Community Based
Program of NSET for Earthquake Disaster
Mitigation. The International Conference
on Total Disaster Risk Management,
Nepal.
GREAT/Gujarat Relief Engineering Advice
Team. (2001). Repair and Strengthening
Guide for Earthquake Damaged Low-Rise
Domestic Buildings in Gujarat, India.
GREAT Publication, India.
IUDMP/Indonesian Urban Disaster Mitigation
Project. (2001). Increasing the Safety of
Indonesian Cities from Earthquake
Disaster Threat. Asian Disaster
Preparedness Center.
Lee, G. and Loh, C. H. (2003). Human and
Institutional Perspective of the 921
Earthquake in Taiwan: Lessons Learned.
at http://mceer.buffalo.edu/publications/
reports/docs/00-0003; accessed on 29
October 2004.
Lustig, T. (1997). Sustainable Management
of Natural Disaster in Developing
Countries (in Fundamental Risk Analysis
and Risk Management edited by Vlosta
Molak). Lewis Publisher, USA.
Mansouri, B., Aghda, F., and Safari. (2002).
Preliminary Earthquake Reconnaissance
Report on the June 22, 2002 Changureh
(Avaj), Iran Earthquake. at
http://www.iiees.ac.ir/English/Publication/
eng_publication_journal_12.html;
accessed on 8 November 2004.
27
Sarwidi. (2001). Reconnaissance Team
CEEDEDS. The Center for Earthquake
Engineering, Dynamic Effect, and
Disaster Studies, Yogyakarta.
SCEC/Southern California Earthquake Center.
(2002). Earthquake as Extreme Events.
Extreme Events Workshop, California.
Winarno, S. and Sarwidi. (2009). Earthquake
Vulnerability of Redential Houses in
Yogyakarta City. Proceedings of 1st
International Conference on
Rehabilitation and Maintenance in Civil
Engineering (ICRMCE), Solo, Indonesia.
Zacoeb, A., Ishibashi, K., and Ito, Y. (2007).
Estimating the Compressive Strength of
Drilled Concrete Cores by Point Load
Testing. Proceeding of the 29th JCI
Annual Meeting, Sendai, Japan, July 11-
13, 2007, p. 525-530.
28
Putuhena, et al.: The Implication of Peatland Built Environment in Urban Drainage System: Case Study... Sarawak.
The Implication of Peatland Built Environment in Urban Drainage System:
Case Study of Sungai Merah Area, Sibu, Sarawak

Frederik Josep Putuhena
1
, Ting Sie Chun
2
, and Salim Said
3
1
Universiti Malaysia Sarawak, Malaysia, email: fjputuhena@feng.unimas.my,
2
Universiti Malaysia Sarawak, Malaysia, email:
sie_chun@hotmail.com,
3
Universiti Malaysia Sarawak, Malaysia, email: ssalim@feng.unimas.my

Abstract: The main purpose of this project is to conduct an analysis to the flooding of Sungai
Merah residential area, Sibu, Sarawak. The total catchment area is approximately 13.74 ha with
relatively flat and typically of a lowlying floodplain which covered with peat. Poor maintenance,
severe conditions and undersized culvert on the current drainage system in Sungai Merah residential
area causes flooding to the area in addition to the waterlogged soil. Postflood forensic analysis of
Sungai Merah residential area presented an approach using InfoWorks Collection System (CS),
coupled with its embedded Geographic Information System (GIS) applications, to identify
hydrology of the drainage pattern. The digital map featuring Sungai Merah area was used to create
GIS map using ArcMap from Arcgis 9. Hydrology and hydraulic data collected is used for model
calibration and verification for 2.4 km-long of drainage network in Sungai Merah residential area
during December 26, 2008 and February 26, 2009 flood events. InfoWorks CS is satisfactorily
capable of providing a clear picture of flood event through model simulation. Besides, excess
rainfall do not influence the surface-runoff of peatlands which been verified by InfoWorks CS.
From the study, the root of flood in Sungai Merah area is triggered by the backwater from Seduan
River and the current existing drainage is fail to manoeuvre the current situation. Furthermore,
InfoWorks CS demonstrates its capability for flood inundation modelling on Sungai Merah area.
Thus, it is recommended that further improvement on the drainage network and floodplain in Sungai
Merah residential area to create a better understanding of the drainages flowpath.

Keywords: Peatland, urban drainage, flood, and InfoWorks CS.

INTRODUCTION

Peat soils are identified as soft soils with
more than 75% organic matter content
(Tang 2009). Compared to mineral soils, peat
has a much higher infiltration capacity,
drainable pore space and hydraulic
conductivity (up to 30 m per day) of the
surface (approximately first 1 metre) of peat
layer but have lower capillary rise, bulk
density and plant-available water (Wsten et
al., 2003). In Malaysia, 1.7 million hectares
or 63% of peat swamp are in delta and
coastal plains of Sarawak. In Sarawak,
approximately 1.66 million hectares or 13%
of the states total land area are covered with
peatlands (Singh and Bujang, 2003). The
aerial extent of peat swamp forest in Sarawak
is shown in Figure 1. About 90% of these peat
areas are classified as deep peat with peat layer
of more than 1 m in depth (Tie and Kueh,
1979).
Figure 1. Peat swamps in Sarawak (Staub
et al., 2000)
29
The lowland peat swamps in Sarawak are
purely rain-fed and waterlogged most times of
the year. Drainage is the leading principle of
the water management practice in order to
curb flooding by evacuation of excess rainfall
within a period of time which is mainly
removed by surface runoff. Furthermore,
variation in tidal level may have a direct and
consequential bearing on the water level in
drainage. Therefore, forensic investigation is
the key to identify the causes of the floods in
peatland.
BACKGROUND OF THE PROJECT
AREA

Sibu Town is the capital of Sibu District In

Sungai Merah, Sibu, Sarawak which is
about 5 km from Sibu Town.
Sungai Merah is located on Seduan
catchement with an area of 116.6km
2
. Water
level measuring gauges are also located at
Sungai Merah, which record water level
continuously in every interval of 30 minutes.
The project area is fairly flat and relatively
Sibu Town
Seduan
Igan River
Outlet of study
Sungai Merah Residential Area
Rajang River
Inlet of study area
Figure 2. Location of the study area.
30
lowying with ground levels ranging from
below 2.0 to 3.0 MSL. Its locality is located in
the central part of Seduan catchement. The
peats properties covers Sungai Merah
residential area with a shear stress of 4.87 kPa,
compressibility index of 1.045 and bulk
density of

0.08g/m (Ting 2009). The shocking
floods in Sibu division, Sarawak with its
confluence of three major rivers namely the
Rajang River, the Seduan River and the Igan
River. The research site (Figure 2) is located
at Sungai Merah December 26, 2008 and
February 26, 2009 had struck the middle and
lower part of Seduan catchement due to
extreme rainfall with addition to high tidal
events.
The urban catchment of Sungai Merah covers
an area of 13.74 ha of residential area. Sungai
Merah residential area is facing frequent flood,
whereby water rises and subsides in matter of
hour. Flood might be caused by poor
maintenance of existing drainage or from earth
drain which has waterlogged soil. Figure 3
highlights a few conditions of existing river
and drainage.
The objectives of conducting this study are
the reconstruction of December 2008 flood to
analyze and comprehend floods behaviors in
Sungai Merah area, studying the hydrology
of the drainage pattern in Sungai Merah area
and lastly engaging a comprehensive analysis
of the performances on the exisiting peatlands
drainage system for future development
purposes.
RESEARCH METHODOLOGY

The research involved data collection, model
development, calibration, verification and
analysis (Figure 4). Nowadays, technology
such as computer software namely Info
Works, MOUSE, MIKE and the SWMM
Models are used to develop computer models
in order to study on the drainage or sewer
system and river flow.
InfoWorks Collection System (CS) is used to
develop 1 Dimensional Urban Drainage
System (Figure 5) with its embedded GIS
applications modeling Sungai Merah
catchment. InfoWorks CS supports the input
of all network data from models such as
HydroWorks, DHI/MOUSE and SWMM
(InfoWorks CS Technical Review, 2009).
Data required in the model development
comprises of hydrology data such as the
rainfall data and tide data from the Department
of Irrigation and Drainage (DID), Sarawak.
Poor Maintenance
Severe Inundation
Water Logging
Figure 3. Current condition of existing
drainage and contributing drainage system
31
Apart from that, catchment characteristic,
soil type and hydraulic data (water level,
discharge and channel manning roughness
coefficient) is taken into consideration.
The United States Soil Conservation Service
(SCS) Method is preferred for rainfall runoff
analysis in the model simulation. This analysis
is being adopted because of its simplicity and
applicability to those watersheds of ungauged
catchment with minimum hydrologic
information of soil type, land use and
treatment, surface condition, and antecedent
moisture condition (AMC) (Leow et al., 2008).
Hydrodynamic of Sungai Merah residential
drainage network is simulated using numerical
modeling. The hydrodynamic modeling used
to predict the flood depth and performance of
the existing drainage system.

MODEL BUILD-UP

Sungai Merah catchment and its floodplan
7.68 ha were modeled with 1.3 km-long
drainage network using InfoWorks CS. A 1:50
000 scaled digital topographical map of 5m
contour intervals featuring Sungai Merah area
which is requested from the Land and Survey
Department, Sarawak. In addition, 1:10 000
scaled Seduan River sounding is bought from
Sarawak Marine Department and digitized
using ArcGIS 9.
The Sungai Merah Catchment modeling first
took place with ground modeling of Sungai
Merah residential area using Geographic
Information System (GIS) technology. GIS
later dis played terrain features of Triangulated
Irregular Network (TIN) dataset. The digital
map featur ing Sungai Merah area was used to
create GIS map using ArcMap from ArcGIS 9.
The TIN model was exported to InfoWorks CS
as a Digital Terrain Model (DTM) where the
ground surface information was used for
network build ing. The resulting TIN
consisted of approximately 125,000 triangles
based on contourand mass points.
InfoWorks CS version 9 was used in this
study. The model is a hydrology and hydraulic
mode, which has its capability in urban
drainage components (Liew et al., 2008).
Sustainable Drainage System (SUDS)
components were also included in the model
necessary in small scale residential area.
Firstly, DTM is exported into InfoWorks CS
through its implanted GIS tool-Geographical
Plan and later used to generate and display
ground level contours, drainage, roads,
buildings and ultility of Sungai Merah
catchment. The sub-catchment of each
compartment of housing areas was delineated
based on available contour lines and inventory
data. Nodes and conduits are then digitally
placed along the existing drainage to form
drainage network.


Calibration and Verification
The model was calibrated against flood
level of a selected historical flood event
on February 26, 2009 whereas model
verification event on December 26, 2008 is
tabulated in Table 1 and illustrated in Figure 6
and Figure 7. Additionally, runoff from the
Figure 4. Flowchart of research
methodology
32
sub-catchment is calibrated and verified for
both events as listed in Table 1 and shown in
Figure 8 and Figure 9. The calibration method
was made by adjusting certain network
component to obtain the effect of the model.
This scenario was due to lack of information
on the areas and flood records. Relevant
information such as flood depth was obtained
from local authorities. Among the adjusted
parameters are the SCS Curve Number
(CN=87), Manning coefficients, (=0.03)
and catchement slope. The selected parameters
Figure 5. Schematic diagram of model building
Model Development, Calibration, and Verification Reconstruction of December 2008
Digital Model of scale 1:50 000

TIN model featuring Sungai Merah
catchment ground level
TIN surface model of Sungai
Merahs houses, drainage, road and
DTM of Sungai Merah
residential Area
ArcGIS 9 Application

33
of certain network component can be adjusted
by percentage to view the effect of changes
on the model.

Peatlands Implication on Drainage
Network
Two different types of flood scenarios were
simulated in Sungai Merah drainage system
namely internal flood and external flood.
Internal flood is the scenario, where heavy
rainfall was occurred and there was no tide
(normal water level). The flow and the
velocity of the existing drainage network were
observed whether the drain can cater for the
amount water runoff during high intensity
storms. External flood is the scenario, where
heavy rainfall is occurred in coincidence with
the high tide.
The peatlands are abundant with extensive
waterlogged soils (Gore, 1983). Research done
by Ong and Yogeswaran, 1991, stated that the
high infiltration, water holding and
transmitting capacity of peat influences the
relatively small fluctuations in the water table.
Besides, as external flood, it is influenced by
King Tide levels along Seduan River with
extreme rainfall intensity.

Table 1. Results of model calibration and
verification
Event
Peak Water Level (meter)
Observed Simulated Differences
Feb. 26,
2009
0.55

0.48

0.07

Dec. 26,
2008
0.60

0.53

0.07

Event
Peak Runoff (m
3
/s)

Calculated Simulated Differences
Feb. 26,
2009
5.13x10
-4

5.37x10
-5

2.3x10
-5

Dec. 26,
2008
9.48x10
-3

9.50x10
-3
2.0x10
-5

From the simulation that has been carried out,
results demonstrate that the most critical
Figure 6. Calibration result for Sungai Merah area during February 26, 2009
Figure 7. Verification result for Sungai Merah
area during December 26, 2008 (Long section)
(a) Calibration and verification point
(During normal flow)
(b) Observed and simulated water level
during February 26, 2009 (Long section)
Simulated Level:
0.48m
Observed Level:
0.55m

34
scenario is when the heavy rainfall and
fluctuation of water levels in Seduan River
coincided. This will affect the water level in
the drainage and causes flooding to the Sungai
Merah residential area. In the nut shell,
flooding in Sungai Merah lowlying areas
depends much on the tide influences.
The simulation results (Table 2) offers a way
to identify the problems, sources of flooding
and potential problems to Sungai Merah
resident. The results were taken from the
culvert in Figure 6(a) whereby the inlet and
outlet is connected to the earth drain.

The simulation results (Table 2) offers a way
to identify the problems, sources of flooding
and potential problems to Sungai Merah
resident. The results were taken from the
culvert in Figure 6(a) whereby the inlet and
outlet is connected to the earth drain.
During internal flood, the 0.21 m
3
/s of flow
in drain with 0.156m/s in velocity is higher
than external flood of which the flow is 0.14
m
3
/s with a velocity of 0.11 m/s as illustrated
in Figure 10 and Figure 11.
This study strongly suggests that the
influences of backwater effect from the
Seduan River significantly affect the flood
in Sungai Merah area. Thus, present studies
by Ritzema and Wosten 2002, confirmed that
in the lower-lying areas, drainage may be
possible only during low-tides.

Furthermore, Ritzema and Wosten (2002) and
Holden et al. ( 2006) added that excess
rainfall will not be removed as surface runoff
Figure 8. Model calibration (Runoff event
on February 26, 2009)
Figure 9. Model verification (Runoff event
on December 26, 2009)
Figure 10. Comparison of flow for internal
and external flood in December 2008
Figure 11. Comparison of velocity for
internal 2008
35
but mainly as groundwater runoff for
peatlands area. The surface-runoff (Figure 12)
of Sungai Merah area is tabulated in Table 3.
The result is verified using InfoWorks CS
as shown in Figure 13. In addition, the
model demonstrates that the Sungai Merah
area will still be flooded due to internal flood
resulting from excessive rainfall which is not
influence by high tide. The impact from high
tide will make the flood in Sungai Merah area
more dreadful. Hence, fundamental aspect of
flood in Sungai Merah area is triggered by the
Seduan Rivers backwater as demonstrated
from the comparison results of the two events
as stated in Table 2.

Table 3. Comparison of surface-runoff of
internal and external flood
Events
Internal
flood
External flood
Surface-Runoff
(m
3
/s)

0.011

0.011

Therefore, the results simulated using Info-
Works CS is believed to be sufficiently
strong to provide information of flood in
Sungai Merah area. In this study, InfoWorks
CS has successfully identified the implication
of peatlands on flood. Hence, further
improvement on the drainage network system
in the Sungai Merah area can be studied for a
better implement.

CONCLUSION

The simulation demonstrates its capability for
flood study on urban storm drainage in peat
lowlands area. The current results of
simulation give a clearer picture of the
drainage flows, velocities, surface-runoffs and
flood depths of an area along with tide level
from the river channel. Further improvement
on current existing drainage system in Sungai
Merah residential area needs to be
reconsidered in order to endure extreme
rainfall and King Tide from Seduan River. The
created model also allows better understanding
of the drainages flowpath for future
Events
Internal flood (influence by heavy
rainfall)
External flood(influence by heavy
rainfall and high tide)
Peak Flow (m3/s) 0.21 0.14
Peak Velocity (m/s) 0.156 0.111
Average Flood Depth (m) 0.31 0.44
Table 2. Comparison of flow, velocity, surface-runoff and average flood depth for internal and
external flood
Figure 12. The location of the surface-
runoff in Sungai Merah area
Figure 13. Comparison of surface-runoff for
internal and external flood in December 2008
Catchment area

Surface- runoff
36
development and research work. However,
simulation on larger study area, where more
observed hydrological flood data are available
is recommended for better understanding on
the drainage behavior on the peatlands
environment.

REFERENCES

Gore, A. J. B. (1983). Introduction In: Gore,
A.J.B. (Ed.), Ecosystems of the World.
Mires: Swamp, Bog, Fen, and Moor.
General Studies, Elsevier, Amsterdam,
The Netherlands, Vol. 4A, p. 134.
Holden, J., Evans, E. G., Burt, T. P., and
Horton, M. (2006). Impact of Land
Drainage on Peatland Hydrology.
Journal of Environmental Quality, p.
1764-1777.
Infoworks CS Technical Review. (2009).
Available online from:
http://www.wallingfordsoftware.com/prod
ucts/infoworks_cs/infoworks_cs_technical
_review.pdf; accessed on 31 October 2009.
Leow, C. S., Rozi, A., Zakaria, N. A., Ghani,
A. A., and Chang, C. K. (2008).
Modelling Urban River Catchment: a
Case Study in Malaysia. Proceeding of
the Institution of Civil Engineers, Water
Management (162) 2009, p. 27-36.
Liew, Y. S., Selamat, Z., and Ghani, A. A.
(2009). Urban Stormwater Drainge
System Study Using Numerical
Modelling. International Conference on
Water Resources (ICWR 2009) 26-27 May
2009, Bayview Hotel, Langkawi, Kedah,
Malaysia.
Majewski, W. (2008). Urban Flash Flood In
Gdansk-2001; Solutions and Measures for
City Flood. International Journal of
River Basin Management, Vol. 6, No. 4, p.
357-367.
Ong, B. Y. and Yogeswaran, M. (1991).
Peatland as a Resource for Water Supply
in Sarawak. Proceedings of the
International Symposium on Tropical
Peatland, Kuching, Sarawak.
Ritzema, H. and Wsten, H. (2002).
Hydrology of Borneos Peat Swamps.
STRAPEAT-Status Report Hydrology.
Singh, H. and Bujang, B. K. H. (2003). Basic
Engineering Geology for Tropical
Terrain. Universiti Malaysia Sarawak.
Staub, J. R., Among, H. L., and Gastaldo, R.
A. (2000). Seasonal Sediment Transport
And Deposition in the Rajang River Delta,
Sarawak, East Malaysia. Sedimentary
Geology, V. 133, p. 249-264.
Tang, V. C. K. (2009). Sustainable
Construction on Soft Soils in Sibu: A
Practical Perspective. Engineering
Seminar on Peat: Soft Soils:Challenges
and Sustainable Solutions, October 15-16,
2009, Sibu, Sarawak.
Tie, Y. L. and Kueh, H. S. (1979). A
Review of Lowland Organic Soils of
Sarawak. Technical Paper, Soils Branch,
Department of Agriculture Sarawak, No.
4, pp. 34.
Ting, S. C. (2009). Relationship between
Road Construction Methods And Post-
Construction Rate of Settlement: Case
Study on Kota Samarahan and Sibu
Roads. Degree Thesis, Universiti
Malaysia Sarawak.
Wsten, J. H. M., Ritzema, H. P., Chong, T.
K. F., and Liong, T. Y. (2003). Potentials
for Peatland Development. In: Chew, Dr.
D. and Sim Ah Hua (Ed.). Integrated
Peatland Management for sustainable
development. A Compilation of Seminar
Papers, Sarawak Development Institute,
Sarawak, Malaysia, p. 233-242.
37
Perbandingan Kekakuan Struktur Lantai Komposit Kayu Glugu-Beton
dengan Berbagai Jenis Penghubung Geser

Susastrawan
Jurusan Teknik Sipil, Universitas Islam Indonesia, Yogyakarta,
email: susaterawan@ftsp.uii.ac.id

Abstract: There are several types of known composite structures, one of them is wood-concrete
composite structure. It is relatively good building structure and cheap since wood materials are
largely available in most markets and also cheaper if compared with steel materials. Coconut palm
wood as part of wood-concrete composite structure has shown a good functioning. In order to make
wood and concrete work together as a composite structure, needs shear connector as fastening
devices are required such as nails, screws, or wood nails. This research investigates the effect
shear connector as fastening devices at a composite structure. The result of laboratory experiment
has indicated that wood-concrete structure using nails shear connector has the highest stiffness
compared with similar structure employing other shear connector as fastening devices.

Keywords: Shear connector and composite.

PENDAHULUAN

Salah satu persoalan pemukiman di kota-kota
besar adalah pengadaan rumah tinggal untuk
masyarakat, khususnya masyarakat
berpenghasilan menengah kebawah. Ada
kecenderungan bahwa orang akan berusaha
untuk tinggal/ bermukim di tempat yang dekat
dengan dimana dia bekerja. Sebagai akibatnya
muncul pemukiman yang kumuh. Mengingat
lahan didaerah perkotaan sangat terbatas dan
relatif mahal maka dapat ditempuh cara
dengan membangun rumah susun yang sangat
sederhana. Salah satu cara untuk dapat
membangun rumah susun yang relative murah
dan aman adalah dengan membangun rumah
susun dengan lantai Komposit Kayu-Beton.
Dengan metode ini harga bangunan dapat
dihemat sampai pada batas yang optimal dan
struktur masih cukup aman.
Penelitian sejenis yang pernah dilakukan Nor
Intang Setyo H. dan Gathot Heri Sudibyo, Juli
2005 menggunakan Bambu-Keruing pada
lantai beton dan Suwandoyo Siddiq dan Siti
Zubaidah Kurdi menggunakan Kayu Kamper
dan Kruing pada lantai beton.
Pada penelitian ini akan diteliti Komposit
Kayu-Beton, dengan bahan kayu Glugu dan
sebagai penghubung geser mengunakan:
pasak, baut dan paku. Dari ketiga macam
penghubung ini akan diteliti mana yang
terbaik.
Lantai komposit kayu-beton ini diharapkan
dapat memberikan keuntungan, diantaranya :
lebih ringan, kedap air dan suara, bersifat
dekoratif, mudah dipasang, beaya lebih ringan,
sedang kerugiannya struktur ini hanya efektif
untuk menahan momen positif saja dan perlu
perawatan yang baik agar kayunya tidak
mudah lapuk.

LANDASAN TEORI

Aksi Komposit
Badan LITBANG PU Dep. PU menyatakan,
struktur komposit merupakan struktur yang
terdiri dari dua jenis bahan konstruksi bahan
yang berbeda yang disatukan dengan bagian
Gambar 1. Struktur balok non komposit
Terjadi slip Terjadi slip
38

Susastrawan: Perbandingan Kekakuan Struktur Lantai Komposit Kayu Glugu-Beton
penyambung yang lebih dikenal dengan
penghubung geser (shear connector).
Penghubung geser dipasang untuk meng-
hubungkan dua bahan tersebut sehingga secara
bersama-sama dapat memikul beban yang
bekerja pada struktur.
Untuk memahami aksi komposit, pertama-
tama ditinjau aksi komposit tidak sempurna
(non-komposit) seperti yang disajikan pada
Gambar 1.
Gesekan diantara pelat beton dan balok kayu
diabaikan seperti terlihat pada Gambar 1. Dua
element struktur, pelat beton dan balok kayu
masing-masing memikul sebagian beban
secara terpisah. Permukaan bawah beton
mengalami perpanjangan akibat deformasi
lentur, sedangkan permukaan atas kayu akan
mengalami perpendekkan akibat deformasi
tekan. Diagram regangan yang bekerja pada
struktur tidak komposit disajikan dalam
Gambar 2.
Dengan meninjau Gambar 2, terlihat bahwa
kasus ini terdapat dua garis netral. Garis netral
pertama berada pada titik berat pelat beton dan
garis netral yang kedua terletak pada titik berat
balok kayu. Selanjutnya ditinjau struktur
bekerja komposit sempurna, maka slip antara
pelat beton dengan balok kayu tidak akan
terjadi. Konsep analisis penampang komposit
penuh didasarkan pada dua kondisi, yaitu
kondisi elastis dan kondisi plastis. Kondisi
elastis adalah kondisi dimana baik pelat beton
maupun balok kayumasih berada dalam batas-
batas elastis. Penelitian ini dibatasi pada
kondisi elastis. Beberapa batasan dalam
analisis struktur komposit ini diantaranya
adalah:
1. Defleksi vertikal mempunyai nilai yang
sama untuk kedua elemen, hal ini berarti
tidak ada gap antara pelat beton dengan
balok kayu.
2. Penampang tetap rata baik sebelum maupun
sesudah dibebani, deformasi geser antara
dua elemen diabaikan;
3. Jarak antara penghubung geser adalah
sama;
Sistem komposit sempurna tidak akan terjadi
gelincir relatif diantara pelat beton dan balok
kayu seperti yang disajikan pada gambar 3.
Gaya-gaya horisontal (geser) terjadi dan
bekerja pada permukaan bawah slab tersebut,
sementara gaya-gaya tersebut juga bekerja
pada permukaan atas balok kayu.
Pada kasus komposit sempurna, diantara pelat
beton dan balok kayu tidak akan terjadi
gelincir dan diagram tegangan yang disajikan
pada Gambar 3. Terjadilah sumbu netral
tunggal yang terletak di bawah sumbu netral
pelat beton dan diatas garis netral balok kayu.

Penghubung Geser
Dalam tinjauan balok lentur hendaknya
dipertimbangkan pula, bahwa pada saat yang
sama balok mengalami gaya horizontal yang
timbul akibat leturan. Gaya horizontal yang
timbul antara pelat beton dan balok kayu
selama pembebanan harus ditahan agar aksi
komposit bekerja secara monolit (lihat Gambar
4 dan 5).
Dalam keadaan terbebani, antara elemen kayu
dan beton diberi shear connector sebagai
alat penyambung geser, agar kedua element
tersebut mengalami deformasi yang sama.
Adapun jenis-jenis alat penyambung geser
yang akan diteliti adalah paku, baut, dan
pasak.
Tegangan geser yang timbul dapat dilihat
keseimbangan elemen pp
1
nn
1
pada Gambar 5
yang dipotong dari sebuah balok antara dua
Gambar 3. Struktur balok komposit sempurna
Gambar 2. Diagram tegangan penampang
balok non komposit
P
T
T
C
C

beton
kayu
39
penampang berdekatan mn dan m
1
n
1
yang
terpisah oleh jarak dx. Permukaan atas pada
bagian yang diarsir bekerja tegangan-tegangan
geser horizontal () pada pemukaan-
permukaan ujung elemen bekerja tegangan-
tegangan lentur normal (
x
) yang dihasilkan
momen-momen lentur (Singer dan
Pytel,1985).

Ditinjau sebuah elemen balok seluas dA pada
jarak y dari sumbu netral pada gambar 5.
uaya noimal =
x
uA (1)
Tegangan lentur,
x
=
My
I
(2)
Gaya-gaya elemental melalui luas permukaan
pn,
F
1
= ]
My
I
uA (3)
Dengan cara yang sama p
1
n
1
,
F
2
= ]
(M+dM)y
I
uA (4)
Gaya horizontal F
3
yang bekerja pada
permukaan pp
1
adalah,
F
3
= b ux (5)

Gaya-gaya F
1
,F
2
,F
3
harus berada dalam
keseimbangan statis dalam arah sumbu x
maka,

F
3
= F
2
F
1
(6)
b u
x
= ]
(M+dM)y
I
uA -]
My
I
uA
=
dM
dx
[
1
Ib
*]yJA
=
VQ
Ib
(7)
Aliran geser,
v
=
VQ
I
(8)
Hubungan antara momen lentur dan tegangan
lentur, dengan asumsi:
1. Bidang penampang balok, bidang semula,
tetap bidang.
2. Balok homogen dan mengikuti hukum
Hooke.
3. Modulus elastisitas tarik dan tekan sama.
4. Balok lurus dan penampang tetap.

Hubungan MomenKelengkungan

Persamaan differensial defleksi balok

d
d
g
y
( b )
Gambar 6. Sistem Balok Terlentur
Gambar 4. Diagram Regangan dan
Tegangan penampang komposit sempurna
Gambar 5. Diagram tegangan elemen balok
beton
F
C
kayu
T
F
h
dA
F
d
x

m m
n n
p
1

p
y
y
1

c
( a )
dx
P
R
2 R
1
d b
a
k
h
f
e
c
b
c
a

40

Susastrawan: Perbandingan Kekakuan Struktur Lantai Komposit Kayu Glugu-Beton
Gambar 6 (a) memperlihatkan dua penampang
yang berdekatan, ab dan cd, dipisahkan oleh
jarak dx. Penampang ab dan cd berotasi
relative satu sama lain sebesar d, seperti
nampak pada gambar 6 (b), tetati tetap lurus.
Serat ac akan memendek dan serat bd akan
memanjang. Serat ef tetap tidak berdeformasi
karena terletak di garis netral. Kemudian dari
titik f ditarik garis sejajar ab, sehingga
diperoleh garis cd dan bd=gh=ac=dx.
Kemudian ditinjau serat gh sejauh y dari garis
netral. Sebelum deformasi gh=ef=dx, setelah
deformasi serat gh bertambah sebesar hk.
Oef dan fhk sebangun, sehingga dapat
diperoleh perbandingan sebagai berikut :
cI
Oc
=
hk
hI
(9)
Oe = ,ef = gh dan hf = y, maka:
gh
p
=
hk
y
(10)
Sehingga diperoleh ;

y
p
=
hk
gh
(11)
hk
gh
= F, maka:

y
p
= F (12)
Dari Hukum Hooke :
= EF = E
y
p
= [
E
p
y (13)
Persamaan diatas menunjukkan bahwa
tegangan pada setiap serat bervariasi langsung
dengan kedudukan y dari garis netral dan
karena modulus elastisitas E dianggap sama
untuk tarik dan tekan serta nilai tidak
tergantung dari kedudukan serat y. Hal ini
berlaku untuk kondisi tegangan masih di
bawah batas proporsional, karena hukum
Hooke hanya berlaku pada tegangan di bawah
batas proporsional.
Untuk menurunkan rumus lentur, digunakan
kondisi keseimbangan yang terjadi pada suatu
tampang. (lihat Gambar 7):
Fx = 0, maka M
luar
= M
dalam
(14)
dM = (
x
dA)y (15)
Keseimbangan momen terhadap sumbu Z
harus sama dengan nol ( Mz = 0 ). Hal ini
berarti momen lentur diimbangi oleh tahanan
momen, yaitu M = Mr. Tahanan momen
terhadap sumbu netral adalah = y(xdA),
sehingga total tahanan momen adalah :

dM = (
x
dA)y (16)
H = ]y(o
x
JA) (17)
Karena
x
= = (E/)y, maka:
H =
L
p
]y
2
JA (18)
Karena ]y
2
JA didefinisikan sebagai I, maka:

H =
LI
p
(19)
Atau:

1
p
=
M
LI
(20)
Tan = dy/dx, karena sudut sangat kecil
sekali maka tan = , sehingga:
=
dy
dx
(21)
Gambar 7. Tegangan pada tampang balok
Permukaan
netral
Y
xz
dA
xy
dA
x
dA
X
Z
z
y
41
Dan,

d
dx
=
d
2
y
dx
2
(22)
Dari gambar 1b diperoleh :
ef = d dan ef = gh = bd = dx, maka:
dx = d atau :
1
p
=
d
dx
(23)
Sehingga:

1
p
=
d
2
y
dx
2
(24)
Maka:

M
EI
=
d
2
y
dx
2
(25)
Penyelesaian persamaan deferensial diatas:
(Chajes, 1974)

Dari gambar 8 diatas dapat diperoleh :

fi =
I
+h
-I
-h
2h
(26)

2
f
I
= (f
I
) =
(I
+h2
-I
-h2
)
h

=
I
+h
-I
h
-
I
-I
-h
h
h

2
f
I
=
I
+h
-2I
+I
-h
h
2
(27)
Dari persamaan batang lentur diatas akan
dapat diperoleh:
M
EI
=
I
+h
-2I
+I
-h
h
2
(28)

M
EI
=
y
1
-2y
2
+y
3
h
2
(29)

TEST LABORATORIUM DAN
PEMBAHASAN

Pengujian Geser terhadap Alat Sambung
Geser
Pada penelitian ini dipakai baut diameter 10
mm, pasak dg kayu bambu diameter 10 mm
dan paku diameter 4,57 mm dengan panjang
masing-masing 12 cm. Masing-masing benda
uji terdiri dari 12 buah penghubung geser(lihat
Gambar 9). Dari hasil test laboratorium
diperoleh hasil bahwa penghubung geser
dengan paku memberikan hasil yang paling
baik (paling kaku) dibandingkan baut dan
pasak (lihat Gambar 10).
Gambar 10 menunjukkan bahwa masing-
masing jenis penghubung geser dapat menahan
beban sebagai berikut, penghubung geser paku
Pmax = 3000 kg, penghubung geser pasak
Pmax = 2000 kg dan penghubung geser baut
Pmax = 2400 kg. Dengan demikian maka
kekuatan penghubung geser adalah:
- Baut: Pmax = 2400/12 = 200kg/buah
- Pasak: Pmax = 2000/12 = 166,67 kg/buah
- Paku: Pmax = 3000/12 = 250 kg/buah
Pemindahan gaya pada penghubung geser
paku dapat berlangsung lebih baik dari pada
baut dan pasak. Hal ini disebabkan karena
paku-paku tersebut dipukulkan begitu saja
kedalam lubang yang lebih sempit yang dibuat
sebelumnya, sehingga paku dapat menyatu
dengan kokoh pada kayu (lihat Wiryomartono,
1976).
Selanjutnya dibuat benda uji balok komposit
yang terdiri dari balok kayu Glugu ukuran
6x12 cm2 dan pelat beton ukuran lebar = 40
cm dan tinggi 6 cm dengan penghubung geser
masing-masing Baut, Pasak dan Paku. Bentang
masing-masing benda uji adalah sebesar 3,4 m,
seperti nampak pada Gambar 11.
Gambar 8. Model Finite Difference
x= i
f(i-h)
f(i)
f(i+h) h)
f(x)
f(x)
X
h h
42

Susastrawan:

P
kayu
Perbandingan
( a ) Baut
P

beton
Kekakuan Stru
Gambar 10
G
G
A
P
kayu
ktur Lantai Kom

0. Grafik da
Gambar 9. U
rafik Be
Alat Sam
P
kayu
mposit Kayu Gl

aya dukung
Uji Geser A
eban-Def
mbung G
( b ) Pasak
P
P
beton
lugu-Beton
alat penghu
lat Sambun
fleksi
Geser
P
kayu
ubung geser
ng
P
kayu
r
( c ) Pak
P
P
beton
ku
P
kayu
43

Gamb
(
Gam
(
Pengguna
bar 12. Graf
alat sambun
mbar 13. Gra
(alat sambu
aan Alat Sa
fik Beban-D
ng geser bau
afik Beban-
ung geser pa
Gambar
( b )
Penghubu
geser bau
ambung Ge

Defleksi
ut)
Defleksi
asak)
11. Struktur
dial
Pe
ge
ung
ut
J
eser pada S

340 cm
r Balok Kom
(a)
P
( c )
enghubung
eser pasak
Jurnal Rekayas
Struktur Ko

mposit Kay
Gambar
(ala
Gamb
(
Peng
gese
sa Sipil (JRS), V
omposit Ka
yu-Beton
r 14. Grafik
at sambung
bar 15. Dim
kompo
t
w
b
tr

( d )
hubung
er paku
Vol. 1, No. 1, M
ayu-Beton
k Beban-Def
geser paku
ensi tampan
osit
kayu
beton
Mei 2011: 38-46
fleksi
u)
ng
t
h
w
6 44

Susastrawan:
Dari uji ba
fc = 20 M
MOR = 52

Dengan dim
b = 40 cm
b
w
= 6 cm
h
w
= 12 cm
t = 6 cm
h = 18 cm
maka dap
teoritis ada
u
= 0,493

Tabe
Penghubu
y
geser
Baut
Pasak
Paku

Dari hasil
pada Gamb
tampak b
Perbandingan
ahan beton d
MPa, maka F
2,395 MPa,
mensi tamp
m dan b
trans
=
m
m
pat dihitung
alah:
34 rad/m
l 1. Kekaku
ung Beban
Py
(kN)
1
40
2 22
1
30
2 22
1
38
2 38
test labora
bar 12, 13,
bahwa ba
Gamb
Kekakuan Stru
dan kayu di
Fc = 17 MPa
maka Fw =
pang (lihat G
= 171,2 cm
g kelengku
uan Balok K
n
)
Defleks
y
(mm)
55.28
48,61
42,8
30,25
36,6
31,74
atorium yan
14, dan 16
alok komp
bar 16. Graf
ktur Lantai Kom
iperoleh ha
a.
= 41,92 MPa
Gambar 15)
ungan ultim
Komposit
si Kekakua
Py/y
(kN/m)
723,59
452.58
700,93
727,27
1038,25
1197,22
ng ditunjukk
serta Tabel
posit deng
fik hubunga
mposit Kayu Gl
asil:
a.
:
mit
an

)
9
8
3
7
5
2
kan
l 1,
gan
pe
ya
B
P
P

Ap
ya
de
nil
nil
ult
Ba
pa
dib
ge
ka
de
leb
lai
saj
ka
an Beban-de
lugu-Beton
enghubung g
ang paling b
Tabel 2. K
Penghubung
Geser
Baut
1
2
Pasak
1
2
Paku
1
2
pabila diliha
ang terjadi
engan peng
lai kelengku
lai kelengku
timet teoriti
alok komp
aku mempun
banding bal
eser baut d
arena pemin
engan pengh
bih baik da
in, karena p
ja, sehingg
ayu (Wiryom
efleksi komp
geser paku
besar.

Kelengkunga
g
Mom
Maksim
(kNm
21,6
12,9
18,3
12,9
28,0
28,0
at dari perh
(lihat Tabe
ghubung ge
ungan yang
ungan terse
is = 0,4934.
osit denga
nyai kekaku
lok kompos
dan pasak.
ndahan gaya
hubung ge
ari pada pe
paku-paku t
ga terikat/m
martono, 19
posit Kayu-
mempunya
an Balok K
men
mum
m)
Kel
60
96
36
96
08
08
hitungan kel
el 2), balok
eser paku m
g paling ke
ebut << kel
.
an penghub
uan yang p
sit dengan p
Hal itu
a pada balo
ser paku b
enghubung
ersebut dipu
melekat er
976).
-Beton
ai kekakuan
omposit
lengkungan
(1/m)
0,0222
0,0321
0,0633
0,0777
0,0030
0,0041
lengkungan
k komposit
mempunyai
cil, dimana
lengkungan
bung geser
paling besar
penghubung
disebabkan
k komposit
berlangsung
geser yang
ukul begitu
rat dengan

n
n
t
i
a
n
r
r
g
n
t
g
g
u
n
45
SIMPULAN

1. Komposit Kayu Glugu-Beton dengan
penghubung Geser Paku menghasilkan
struktur yang paling kaku dibandingkan
dengan struktur komposit Kayu-Beton
dengan penghubung baut dan pasak.
2. Pelaksanaan pembuatan komposit kayu-
beton mudah. Karena mudah tentunya juga
menjadi lebih murah.

SARAN

1. Alat penghubung geser mempunyai variasi
yang cukup banyak, sehingga diperlukan
penelitian lebih lanjut untuk penghubung
geser yang lain.
2. Perlu diteliti lebih lanjut struktur komposit
kayu-beton dengan menggunakan kayu
yang mempunyai kelas awet lebih tinggi,
misal dengan kayu jati atau kayu
Kalimantan.

DAFTAR PUSTAKA

Chayes and Alexander. (1974). Principles of
Structural Stability Theory, Prentice-
Hall, Inc, Englewood Cliff, N.J.
Singer, F. L., Pytel, A., and Sebayang, D.
(1985). Teori Kokoh - Strength of
Materials, Erlangga.
Setyo, N. I. H. and Sudibyo, G. H. (2005).
Balok Komposit (Glulam) Bambu
Keruing pada Lantai Beton. Media
Teknik Sipil.
Wiryomartono, S. (1976). Konstruksi Kayu.
Fakultas Teknik UGM, Jilid 1.
Siddiq, S. and Kurdi, S. Z. (1995). Struktur
Lantai Komposit Kayu-Beton. Majalah
Ilmiah Teknik, ISSN: 0854-2279, Jilid 1.
46
Thanaya: Masonry Unit Utilizing Aggregate From Construction Demolition Bound with Asphalt
Masonry Unit Utilizing Aggregate from Construction
Demolition Bound with Asphalt

I Nyoman Arya Thanaya
Udayana University, Denpasar, Bali, Indonesia,
email: aryathanaya@ymail.com

Abstract: In line with increasing pressures to reduce the exploration of natural aggregates,
utilization of waste aggregates materials for construction industry had been encouraged. One
alternative material that can be used as masonry unit is aggregate from construction demolition
with asphalt as the binder (an alternative to cement). Construction demolition may available due to
various reasons such as demolition for reconstruction, war, natural disasters etc. The investigation
was carried out in the United Kingdom (UK), which is applicable else where in the world especially
in oil producing countries where large amount of asphalt is available. The as phalt used as the
binder was 100pen grade asphalt. The objective of the investigation was to produce masonry unit
with performance equal to concrete block commonly used in the United Kingdom with
compressive strength between 2.810 MPa. This requirement is slightly higher than the minimum
compressive strength of 2.5 MPa in line with Indonesian standard. The specific creep strain
targeted was less than 100 microstrain. The masonry unit requires suitable particles size proportion,
in order to obtain the expected results: stable during handling with low compaction effort with
satisfactory compressive strength and to meet demand in using minimum bitumen content. The
materials were proportioned, hot mixed, compacted than heat cured. The masonry unit requires
sufficient heat curing. It was found that construction demolition materials (CDM) were suitable to
be used for making masonry unit with asphalt as the binder. Compaction level of 2 MPa and curing
regime of 200C for 24 hours were sufficient and gave an overall satisfactory result. It was also
found that the masonry unit volume stability was affected by relative humidity (moisture). Its
volumetric movement due to moisture was not fully reversible, but highly reversible on thermal
exposure. The compressive strengths of the unit produced using locally available construction
demolition material, well meet the Indonesian standard, i.e. minimum of 25 kg/cm2.

Keywords: Masonry, construction, demolition, aggregate, and asphalt.

INTRODUCTION

From masonry Each house on average requires
Demands on aggregates fot construction
approximately 200 m of building block
industries worldwide continue to increase. For
example, aggregate demand of the United
Kingdom (UK) raise from a level of 270
million tonnes in 1989, to a predicted demand
of 420-490 million tonnes by 2011 (Whitbread
et al., 1991).
Meanwhile there has been an increasing
work resulting in approximately 350 million
blocks being manufactured each year. Waste
or by product materials such as steel slag,
crushed glass, and coal fly had been
incorporated into masonry building blocks
(Forth et al., 2006).
Another alternative material that can be used
is construction demolition waste (CDW)
from
pressure to reduce the exploration of
natural building or road pavement (Craighill et
al. aggregates).

This situation encourages the
utilization of waste and secondary aggregate
materials for construction industry.

Currently, 160,000 new homes are built achear
in the UK of which 90% are constructed
(Forth et al., 2006).
Around 17% of the total UK waste arises from
the construction and demolition industries
(DoE, 1994). Although a large proportion of
47
the waste
most of it
such as ac
only 4% is
(Humphrey
There are
materials,
perceived
strength.
confidence
materials,
identifying
appropriate
disseminat
publication
of the agg
replaced b
CDW ag
affecting th
al., 2004)

The objec
within this
with perfo
blocks co
Kingdom
between 3
of 2.8 M
creep strai
1985).
The applic
when large
demolition
inevitable
Figure
Was
created is
t is used
ccessroads
s used to rep
ys et al., 19
e many po
however th
risks inv
There is
e in the
which can
g, underta
e demons
ting the
ns and semi
regate for c
by up to a
ggregates
he compres
tive of the
s paper is
ormances at
ommonly
(UK) wit
.57 MPa (S
MPa (BS60
in less than
cation of t
e amount C
n of old b
natural d
e 1. The Con
ste (CDW) a
recycled i
for low gr
within land
place prima
94).
otential use
hey are de
olved due
a need
use of CD
n only be
aking and
stration p
e result
inars. The c
concrete mi
approximate
without
ssive streng
e investigat
to produc
t least equa
used in
h compres
Sear 2005)
0731981),
n 100 micro
this researc
CDW is av
buildings, a
disasters, su
nstruction D
aggregates m
in some w
rade purpo
dfill sites a
ary aggrega
es for CD
eterred by
to its l
to increa
DW recyc
achieved
d monitori
projects, a
ts throu
coarse fracti
ixtures can
ely 20% w
significan
gth (Soutsos
tion describ
ce CDWblo
al to concr
the Uni
ssive streng
or minimu
and speci
ostrain (Tap
ch is possi
ailable due
and/or due
uch as ea
Demolition
material
way,
ses
and
ates
DW
the
ow
ase
led
by
ing
and
ugh
ion
be
with
ntly
s et
bed
ock
rete
ted
gth
um
ific
sir,
ble
e to
to
arth
qu
att
esp
lot
ca
av
are
pro
sel
bit
(T
Th
tes
an

M

M
Th
co
wo
ma
bu
thi
tha
cla
rec
pro
as
siz
ma
Th
gra
bit
Jurnal Rekaya
uake, flood,
tractive in
pecially in
t of asphalt
n be used
vailable. Th
eas can be
oviding b
lves.tumen.
tumen ha
Thanaya et a
he propertie
sted in line
nd are given
METHODS
Materials
he aggrega
onstruction
ould be of
ay come fro
uildings or c
is investiga
at consisted
ay brick ma
claim aspha
oportion w
can be see
ze distribut
aximum par
he bitumen
ade bitume
tumen.
Figure 2.
Waste (
100
90
80
70
60
50
40
30
20
10
0
0.01
asa Sipil (JRS),
cyclone, et
areas clo
oil produci
t (oil distil
as an altern
he people in
trained to m
building m
Utilizatio
ad given
al., 2006).
es of the C
e wit the B
n in Table 1.
ate used
demolition
non homog
om demoliti
construction
ation was t
d of a mixtu
asonry, san
alt pavemen
with minor c
en in Figure
tion is show
rticle size o
used as th
en. This is
The Constr
(CDW) agg
original
grad
0.1
Vol. 1, No. 1, M
tc. This wo
ose to oi
ing countrie
llation resid
native binde
n unfortuna
make CDW
materials
on of so
satisfactor
CDW aggre
BS 812 (BS
.
is aggreg
waste (CD
geneous ma
ion of vario
ns. The CD
taken from
ure of broke
nd cement m
nt (RAP) in
clay brick
e 1. The CD
wn in Figu
f 10 mm.
he binder w
relatively
ruction Dem
gregates mat
1 10
Mei 2011: 47-57
rks is more
il refinery,
es, where a
due) which
er is widely
ate disaster
Wblocks for
for them
fter grade
ry results
egates were
S812 1995),
gate from
DW), which
aterial as it
ous types of
W used for
a supplier
en concrete,
mortar, and
n a random
component
DW particle
ure 2, with
was 100 pen
soft grade
molition
terial
0 100
7
e
,
a
h
y
r
r
m
e
s
e
,
m
h
t
f
r
r
,
d
m
t
e
h
n
e
48
C
l
t
i
P
i
Referring to Table 1, the properties the CDW
aggregates were found very comparable with
commonly aggregates available for building
industries.
However, as the CDW is processed from
waste materials, therefore the homogeneity
of the aggregate component may not be
consistent.

Table 1. Properties of the CDW aggregates

Initial Trials
Building block units bound with asphalt can
be produced with either using continuous or
gap aggregate grading. When using
continuous aggregate grading, higher
compaction effort commonly needed (Forth et
al., 2006).
Within this investigation gap aggregate
gradation was selected in order to enable the
use of lower compaction effort. As there is no
standard specifically available for building
block aggregate grading bound with asphalt,
Hot Rolled Asphalt (HRA) aggregate
gradation was referred to as an initial
reference, then modified based on trials
results.
During the initial trials, the CDW material
particles sizes were graded (sieved) into coarse
fraction of 105mm and 52.36mm, and fine
fraction of all passing 2.36 mm. Initially the
aggregate grading used for the CDWblock was
the same as the Bitublocks (similar block
incorporating various waste aggregate
materials) previously produced with a gap
graded aggregate grading (Forth et al. 2008),
but with maximum particle size of 10mm. The
aggregate composition was: 40% coarse
fraction, 50 % fine fraction, and 10 % coal
fly ash filler as shown in Figure 3, which is
completed with a hot rolled asphalt (HRA)
grading of the BS 594 (BS5941 2005), for a
general comparison, and for a better
appreciation on the aggregate composition.

The bitumen used as the binder is a
commercially valuable by product material
from crude oil refinery industry, therefore the
CDWblocks were produced with lowest
bitumen content possible, that can give
sufficient or adequate bitumen coating with
satisfactory shape stability during handling
and satisfactory performances when
compacted at low compaction effort.
Minimum bitumen content of 5 % by weight
of total mixture was initially tried. The static
compaction effort was 1 MPa for 1 minute.
The degree of bitumen coating was
satisfactory; however the shape of the blocks
was not stable. The corner sides or edges
were easily taken off. The surface texture was
rather smooth, and some aggregate particles
on the edge sides of the samples were taken
off. This had caused some part of the samples
had an open texture, therefore water
absorption after 24 hour immersion was found
high (6.9%), as shown in Tables 2b (Mix A).

Gradation Modification and the Properties
of the CDWblock
The specific gravity of the CDW materials
were not the same with the Bitublock
previously made (Forth et al., 2008), so it
affects the volumetric composition of the
materials. This was found to give effect of the
Properties Unit
Coarse CDW
(> 2.36mm)
Fine CDW
(< 2.36mm)
Density (bulk) g/cm3 2.415 2.341
Density (ssd) g/cm3 2.478 2.433
Density (app) g/cm3 2.569 2.578
Water abs. % 2.5 3.9
BS594 upper limit
BS594 lower limit
Bitublock grading
CDWblock grading
100
90
80
70
60
50
40
30
20
10
0
0,01 0.1 1 10 100
Figure 3. The CDW block aggregate
grading compared with the Bitublock
grading and the BS594
49
Mix
Comp. effort
(MPa)
Water Abs*
(%)
Comp. Strength
(MPa)
uncured cured
Mix A 1 6.9 2.0 8.1

Mix B
1 3.3 2.8 10.2
2 2.7 3.8 17.5
4 1.6 5.4 25.8

Mix Comp. effort
(MPa)
Density
(g/cm
3
)
Porosity
(%)
IRS
(kg/m
2
.min)
Mix A 1 1.925 15.1 0.105

Mix B
1 1.872 17.5 0.028
2 1.950 14.0 0.021
4 1.992 12.2 0.018
samples shape stability (compactness). In
order to improved the CDWblock shape
stability, the aggregate grading was slightly
modified. The max particle size used was
10mm, instead of 14 mm as used for the
Bitublock. The filler content was reduced from
10 % to 5%, but increasing the fine fraction
from 50 % to 55 %, where the coarse fraction
remains at 40 % (Figure 3). This gradation
modification was found to give more compact
samples with surface texture neither too
smooth nor too rough. Overall, the aggregate
gradation of the mix became coarser, hence
theoretically it has lower total surface area.
Even at the minimum bitumen content of 5
% as initially tried the asphalt film thickness
would increase.
In order to improve impermeability, asphalt
film need to be made thicker. For this
reason the bitumen content was increase from
5 % to 5.5 %, in addition to the reduction of
aggregate surface area as mentioned above.
The shape stability during handling was found
satisfactory and the surface texture of the
newly produced sample was found neither too
smooth nor too rough.
By visual observation, the asphalt coating was
also satisfactory, and theoretically with thicker
asphalt film. This was considered necessary in
order obtain satisfactory overall performances.
The increase of the bitumen content was also
done for anticipating the variation in quality of
the CDW materials which are very likely of
various absorption properties, as they are
indeed a waste material. CDWblock with
modified gradation (Mix B in Tables 2a and
2b), were then produced with compaction
effort of 1, 2, and 4 MPa, with bitumen
content of 5.5 % by weight of total mixture.

Table 2a. The density, porosity, and IRS
of the CDWblocks

Parameter that can provide an indication of the
effect of the unit on the sand cement mortar.
Units with high IRS require very plastic
mortar (high water/cement ratio), while units
with lower IRS need stiffer mortar (Vekey,
2001).
In addition to the data presented in Table 2,
the compressive strength of Mix B are also
plotted in a graph as shown in Figure 4.

Table 2b. The density, porosity, and IRS of the
CDWblocks

Heat Curing
Asphalt is a viscoelastic material. It has
viscous and elastic component. The viscous
component of the asphalt would cause creep
when loaded. It had previously been found
that curing regime played a very significant
role for hardening the asphalt due to the
evaporation of the volatile component and
increasing the asphaltene component of the
asphalt (Forth et al., 2006, Whiteoak, 1991),
hence can reduce creep deformation due to
static load (Forth et al. 2006).
It had been previously investigated that when
using a 50 pen bitumen and cured in oven
at 160 C, the curing duration required to
Figure 4. Compaction level vs compressive
strength 24 hours immersion in water
Cured
Un-Cured
0 1 2 3 4 5
30
25
20
15
10
5
0
50
Thanaya: Mas
satisfy cr
(Thanaya
curing du
samples w
hours, and
satisfactory

Initial Rat
IRS test w
sample in
second. Th
sample wa
area in co
IRS is a p
performanc
grading
compared
content, a
bitumen co
B, at 1 MP
lower poro
absorption
thinner. Th
was better
coarser an
logical, th
improved
regime had
improve t
samples.
The water
a reasonab
5.5% bitum
more imp
found rela
values of c
kg/m2/min
CDWblock
lower wate

Expansion
This test
expansion
different
equipment
with its su
points. Tw
each of t
vertical (V
sonry Unit Util
eep perfor
et al., 200
uration, in
were cured
d the samp
y creep resi
te of Suctio
was carried
n 3mm de
he weight o
as then calcu
ontact with
pareferring t
ce of Mix
which c
with Mix
and also
ontent. As
Pa compac
osity than M
was higher
he compres
than Mix A
nd of thick
hat the perf
at higher c
d been exp
the compr
absorption
ble level,
men conten
permeable.
atively low
clay brick fo
n). This
ks require s
er cement ra
n/Shrinkag
was carried
and shrink
environm
used was
upporting f
wo Demec
the four si
V) direction,
lizing Aggregat
rmance wa
06). In ord
this inve
d at 200C
ple was fo
istant.
on (IRS)
d out by im
epth of w
of water ab
ulated and d
h water (B
to Tables 2
A was af
contained
x B with
affected b
s Mix A fi
tion level M
Mix B. How
r as the bitu
ssive streng
A, as Mix B
ker bitume
formance o
compaction
perienced to
ressive stre
of Mix B w
which indi
nt, the sam
The IRS
er than the
ound in the
indicates
sand cemen
atio (Vekey
ge Test
d out by m
kage of th
ment con
a 50 mm
facilities su
points we
ides of the
as shown in
te From Constr
as 72 ho
der to redu
estigation
C just for
ound to ga
mmersing
water for
bsorbed by
divided by
S3921 198
2a and 2b,
ffected by
10% fil
5 % fil
by its low
iner than M
Mix A was
ever the wa
umen film w
gth of Mix
B was sligh
en film. It
of Mix B w
level. Curi
o significan
ength of
was overall
icated that
mples beca
values w
e typical I
UK (0.252
s that
nt mortar w
, 2001).
measuring
he samples
dition. T
Demec gau
uch as dem
ere pasted
e samples
n Figure 5.
ruction Demol
urs
uce
the
24
ave
the
60
the
the
85).
the
its
ller
ller
wer
Mix
of
ater
was
x B
htly
is
was
ing
ntly
the
l in
at
ame
were
RS
2.0
the
with
the
at
The
uge
mec
on
in

F

W
for
3.5
CD
the
the
kg
be
asp
su
ex
us

Cr
Th
res
to
me
6)
cre
cre

lition Bound wi
Figure 5. Th
m
Where comm
r concrete b
57 MPa (Se
DWblock w
e range of I
e United
g/m
2
/min).
ecause the a
phalt which
ggest that
xperiment w
e stiffer mo
reep Test
his test wa
sistance of
static load
eans of a s
, with 1 M
eep stress)
eep test (Ta
Figure 6. A
ith Asphalt
he CDWblo
measuring e
mon compr
blocks foun
ear, 2005).
were found
IRS values f
Kingdom
Low IRS
aggregates w
h has hydro
the CDW
would requir
ortar.
as done in
the sample
d. The sam
simple arm
MPa stress.
is common
apsir, 1985)
Arm load ma
test loa
ock sample a
equipment
ressive stre
nd the UK
The IRS va
somewhat
for clay bric
m (between
values wer
were evenly
phobic char
Wblock test
re or more
order to e
es to defor
mples were
load mach
. This stre
nly applied
.

achine used
ading
and strain
ength value
is between
alues of the
lower than
ck found in
n 0.252.0
re obtained
y coated by
racter. This
ted in this
suitable to
evaluate the
rmation due
e loaded by
hine (Figure
ss (specific
in masonry
for creep

e
n
e
n
n
0
d
y
s
s
o
e
e
y
e
c
y
51
Expansion Due to Moisture Absorption
Before doing creep test, the cured samples
were tested for their expansion at room
environment condition (210.5 C and 46 %
relative humidityRH). It was found that the
samples expanded then stabilized after about 7
days. The samples with higher compaction
level gave lower expansion as shown in Figure
7.


Compressive Strength and IRS
Referring to Figure 4, it is revealed that the
cured compressive strength of the samples
were satisfactory, i.e well exceeded 7 MPa, In
order to evaluate the performance of the
samples at different relative humidity, a
further volume stability test was done. Two
samples were initially conditioned at room
environment (at a different room from
previous experiment). The temperature was
relatively constant at 21.0 0.5 C, but the
humidity fluctuated at 622% RH.
The samples were also conditioned at 12%RH
and 85%RH which were carried out by using
desiccators filled with lithium chloride and
potassium chloride hygrostatic solution
respectively.
The expansion reading was taken at certain
time interval until the expansion stabilized.
Then the conditioning was changed. The
results are shown in Figure 8 where the
starting of conditioning changes are coded
from A to E.
The samples were initially left overnight at
room environment with 622% RH before the
first strain reading was noted. After
conditioning at room environment, the
samples were then conditioned in a desiccator
with 12% RH, using lithium chloride
hygrostatic solution (start of conditioning A).
The samples gradually shrunk then stabilized
at +100 microstrain. Starting on day 11
th
, the
samples were taken out from the desiccator
and left at room environment (622% RH),
(start of conditioning B). The samples slowly
expanded then stabilizedat 20 microstrain.
Starting from day 24
th
the samples were put
back into a different desiccator with 85%RH,
using potassium chloride hygrostatic solution
(start of conditioning C). The samples
expanded (towards negative strain values) then
stable at -340 microstrain. Starting for day 44
th
and then the following days, the samples were
consecutively condition until stabilized at
room environment (start of conditioning D),
then at again at 12% RH (start of conditioning
E). cules onto the surface of the particles
reduces the surface energy on the capillary
system, hence reducing the balancing internal
compressive stress leading to volume increase
or swelling (Domone, 1994).
This is also described by Neville (Neville
1991), that during water adsorption, the
water molecules act against cohesive forces
and tend to force the cement gel particles
further apart. The ingress of water also
decreases surface tension, and results in
swelling.
It was also observed that the samples did not
crack which indicates that the expansion was
not excessive. The expansion of the unit would
be neutralized by the shrinkage of the sand
cement mortar joints in wall construction. The
expansion can also give a prestressed
condition to the wall structure which can
improve the ability of the wall to receive
horizontal load.

Thermal Expansion
Thermal expansion test was carried out by
conditioning the samples in oven at 70 C for
3 hours. This time was sufficient to generate
the targeted heat on the core of the samples
0 1 2 4 6 8 10 12 14 16
Figure 7. The vertical expansion test results
1 MPa
2 MPa
3 MPa
-25
-50
-75
-100
-125
-150
-175
0
52
(tested using a thermocouple inserted into one
of the samples).

Figure 8. Average movement of the samples
(sp) in vertical (v) direction

Figure 9. The strain profile of the samples (sp)
during heating at 70 C for 3 hours in vertical
(v) direction

This further test confirmed that the volume
stability of the samples was affected by
changes in relative humidity. Conditioning to
lower relative humidity caused the samples to
shrink and vice versa. However, the magnitude
of expansion and/or shrinkage was found not
proportional to the changes in RH. The results
indicated that the samples movement were
partly reversible and irreversible. Quiet a large
portion of the movement was irreversible. This
situation is similar to clay brick (Vekey,
2001).
The results suggest that the expansion of the
samples were of similar mechanism with
cement paste or concrete, i.e. due to moisture
adsorption. Due to adsorption of water
mole After heating, the samples were left at
room environment (21.0 0.5 C, with 502%
RH) until stabilized. The results are presented
in Figure 9. Referring to
Figure 9 (strain in vertical direction), within
the first heating cycle the samples expanded to
750 microstrain (10
-6
). Then the samples were
taken out from the oven and conditioned at
room environment for 2 days. Within the
fist day at room environment the samples
shrunk almost to its original position and then
slightly expanded on the next day due to the
moisture absorption from the environment (as
had been experienced). Similar procedures
were carried out on the next two cycles.
Thermal expansion of the blocks was found
highly reversible, similar to concrete masonry
(CST, 2007). The coefficient of thermal
expansion on the Building Blocks was around
600-700 microstrain or (10
-6
) per 70C, or
about 8.6-10x10
-6
/C. This coefficient should
have affected by the size of the samples
(100x100x65mm) and the level of curing
regime applied. The coefficient is comparable
to the coefficients of expansion of concrete
masonry units, i.e.7.2 to 9.0x10
-6
/C
(Drysdale et al., 1994).

Creep Performance
After the expansion test with results as in
Figure 7 (after the volume of the samples
stable), the samples were then tested for creep
at the same environment. The stress applied
was 1 MPa.
This stress is commonly applied in masonry.
The asphalt used was of 60/70 pen, and the
content was varied and gave optimum at
5.5%. At this stage the curing regime applied
was 200 C for 24 hours as shown in Figure
11.
Then at optimum asphalt content, the heat
curing regime applied was 200 were tested C
for 4, 8, 16, and 24 hours. The samples at dry
condition and after soaking in water for 24
hours. The results are given in Figure 12.
Some additional properties of the samples are:
1
creep strain = total strainelastic strain
shrinkage or expansion.
* the samples were tested for creep after the
expansion stabled (at zero expansion).
The creep test results indicated that the all of
0 1 2 3 4 5 6 7
200
0
-200
-400
-600
-800
sp1 sp2 avg
300
200
100
0
-100
-200
-300
-400
0 10 20 30 40 50 60 70
sp1 sp2 avg
53
0 4 8 12 16 20 24
5
4
3
2
1

Comp
Level
(MPa)
Total
Strain
()
Elastic
Strain
()
Creep
Strain
1

()
Exp. at
creep test*
()
1 173.25 54.45 118.8 *
2 143.55 49.5 94.05 *
4 113.85 34.65 79.20 *
the samples gave creep strain at least equal of
concrete block commonly used in the UK
(Sear, 2005, BS6073 1981). In order to ensure
a better deformation resistant, creep strain of
less than 100 microstrain is recommended
(Tapsir, 1985), therefore compaction level of
at least 2MPa is suggested.

Figure 10. The creep strain of the samples
compacted at different compaction level
hammer

RECENT EXPERIMENT

Table 3. Creep performance of the CDWblock
samples

The author had carried out recent experiment
on masonry block unit using locally available
construction demolition material in Bali. After
carrying out trials, the aggregate particles
composition was determined to consists of
40% coarse particles (14-2.36) mm, 50% fine
particles (2.36-0.075) mm and 10 % filler
(passing 0,075mm). Waste concrete was
used for the coarse particles. The fine particle
portion consisted of 25% waste concrete, 50
% concrete blocks, and 25 % clay brick.
Rice husk ash was used as the filler.
Compaction was carried out by applying 50
blows Marshall experiments in order to
evaluate specific creep, i.e. creep strain per
MPa unit stress. The creep test results are
shown in Figures 10 and summarized in
Table 3. Given in Table 3.
The average porosity, water absorption, and
initial rate of suction (IRS) of the samples
curedat 200C for 4-24 hours is summarized in
Table 4.

Figure 11. Compressive strength of the
samples at varied asphalt (60/70 pen) content,
cured at 200 C for 24

Figure 12. Compressive strength of the sample
at 5,5% optimum asphalt content, cured at 200
C with varied curing duration

In line with the Indonesian National Standard
the (SNI) for masonry concrete block/unit,
the minimum compressive strength is 25
kg/cm2 or 2.5 MPa (SNI 03-0348, 1989).
Referring to the results in Figures 11 and 12,
0 4 8 12 16 20 24
160
120
80
40
0
1 4 4,5 5 5,5 6 6,57
5
4
3
2
sp1 sp2 avg
sp 1
sp 2
54
the compressive strength of the block unit
well meets the standard. There is no
requirement on porosity, water absorption and
IRS. The IRS values in Table 3 suggest the
use of mortar with higher water content, as
the unit gave relatively high IRS and water
absorption.

CONCLUSIONS

Some conclusion can be withdrawn from the
investigation, i.e.:
1. The CDW materials were found very
suitable for producing asphalt bound
construction demolition waste masonry
block (CDWblock).
2. Compaction level minimum of 2 MPa
and the curing regime applied (200 C for
24 hours) were found to give satisfactory
performances (satisfy compressive
strength and creep performance).
The performances of the CDWblocks
were found at least equal to the concrete
blocks commonly used in the United
Kingdom (UK).
3. The volume stability of the CDWblock
is affected by relative humidity (RH).
Higher RH environment tends to cause
higher excuring duration (hour) pansion,
and vice versa.
4. The volumetric movement of the
CDWblocks due to environment moisture
is not fully reversible, but highly
reversible due to thermal exposure.
5. The compressive strength of the
masonry unit recently tried using locally
available construction demolition
material,well meet the minimum
Indonesian standard of 25 kg/cm2.

ACKNOWLEDGMENT

The author would like to express his gratitude
and appreciation to The School of Civil
Engineering Leeds University-United
Kingdom, the Engineering and Physical
Sciences Research Council (EPSRC)-United
Kingdom for providing funds, and to his
research supervisors Dr. S. E. Zoorob and
Dr. J. P. Forth who had given guidance
during the authors post doctoral research in
the year of 2004-2007.

REFERENCES
British Standard Institution BS6073, Part 1.

(1981). Precast Concrete Masonry Unit.
Specification for Clay Bricks, British
Standard (BS) 3921.
British Standard BS 812. (1995). Testing
Aggregates, Part 2: Methods of
Determination of Density.
British Standard (BS) 594-1. (2005). Hot
Rolled Asphalt for Roads and Other
Paved Areas-Part 1: Specification for
Constituent Materials and Asphalt
Mixtures.
CEB-FIP. (1993). CEB-FIP Model Code
1990. Comite Euro-International du
Beton, Netherlands.
Chen, W. F. (1982). Plasticity of
Reinforced Concrete. McGraw-Hill
International Company, Washington, D.C.
Craighill, A. and Powell, J. C. (2006). A
Lifecycle Assessment and Evaluation of
Construction and Demolition Waste.
Research Project Report funded by the
Engineering and Physical Sciences
Research Council (EPSRC)-UK, Centre
for Social and Economic Research on the
Global Environment University of East
Anglia and University College London,
ISSN 0967-8875.
CST/Cast Stone Institute. (2007). Design
Tips Technical Bulletin #52 Allowing
for Movement of Masonry Materials. At
http://www.caststone.org/bulletins/tb_52.p
df ; accessed on24 October 2007.
De Witte, F. C. and Kikstra, W. P. (2008).
"DIANA Finite Element Analysis, User's
Manual: Release 9.3. TNO Building and
Construction Research, Delft, The
Netherlands.
DoE/Department of the Environment. (1994).
Guidelines for Aggregates Provision in
England. Minerals Planning Guidance
Note 6, HMSO, London.
Domone, P. L. (1994). Part Three: Concrete,
in Construction Materials Their Nature
55
and Behaviour. Edited by J.M. Illston,
Second Edition, E&F Spon, p. 87-195.
Drysdale, R. G., Hamid, A. A., and Baker, L.
R. (1994). Masonry Structures: Behavior
and Design. Prentice Hall.
Forth, J. P., Zoorob, S. E., and Thanaya, I. N.
A. (2006). Development of
Bitumenbound Waste Aggregate Building
Blocks. Proceedings of the Institution of
Civil Engineers, Construction Materials
159, Issue CMI, Thomas Telford-London,
February 2006, p. 23-32.
Forth, J. P., Zoorob, S. E., and Thanaya, I. N.
A. (2008). The Time-Dependent
Performance of Bitublock Single Leaf
Masonry. Proceedings of 14th
International Brick and Block Masonry
Conference, Manly Pacific Hotel,
Sydney, Australia.
Goto, Y. and Joh, O. (2004). "Shear
Resistance of RC Interior Eccentric
Beam-Column Joints. 13th World
Conference on Earthquake Engineering
(13 WCEE), Vancouver, Canada, August
1-6, 2004, p. 13.
Indonesian National Standard. (2002).
Earthquake Resistant Design Standard
for Building Structures. SNI 03-1726-
2002, National Standards
InstituteHumphreys, H., Partners
Consulting Engineer. (1994). Managing
Demolition and Construction Wastes.
Report of the study on the Recycling of
Demolition and Construction Wastes in
the UK, Report for the Department of the
Environment, HMSO, London.
Naito, C. J., Mosalam, K. M., and Moehle, J.
P. (2000). Evaluation of Reinforced
Concrete Bridge-Joints. 12
th
World
Conference on Earthquake Engineering,
Auckland, New Zealand, p. 1-8.
Neville, A. M. (1991). Properties of
Concrete. 3
rd
Edition, Longman
Scientific and Technical, p. 373375.
New Zealand Standard. (1995). Concrete
Structures Standard and Commentary
NZS 3101, Part I and Part II. SANZ
(Standard Association of New Zealand),
Wellington.
Park, R. (2002). A Summary of Results of
Simulated Seismic Load Tests on
Reinforced Concrete Beam-Column
Joints, Beams and Columns with
Substandards Reinforcing Details.
Journal of Earthquake Engineering, Vol.
6, No. 2, 147174.
Paulay, T. (1989). Equilibrium Criteria for
Reinforced Concrete Beam-Column
Joints. ACI Structural Journal, Vol. 86
No. 6 (November-December), p. 635-643.
Sear, L. (2005). Blocks Made in UK 100 %
Utilization of Bottom Ash. ECOBA
Conference Paper, United Kingdom
Quality Ash Association (UKQAA) at
http://www.ukqaa.org.uk/Papers/ECOBA
%20Conference%20FBA%20in%20block
%20-%20Sear.pdf; accessed on 08
February 2006.
Soutsos, M. N., Jones, N., and Millard, S. G.
(2004). Precast Concrete Building
Blocks Made with Aggregates Derived
From Construction and Demolition
Waste. Proceedings on Use of Recycled
Materials in Buildings and Structures,
Barcelona.
Standar Nasional Indonesia (SNI) 030348.
(1989). Mutu dan Cara Uji Bata Beton
Pejal.
Tajima, K., Mishima, T., and Shirai, N.
(2004). 3-D Finite Element Cyclic
Analysis of RC Beam/Column Joint using
Special Bond Model. 13th World
Conference on Earthquake Engineering
(13 WCEE), Vancouver, Canada, August
1-6, 2004, p. 15.
Tapsir, S. H. (1985). Time-Dependent Lost
of Post Tensioned Diaphragm and Fin
Masonry Wall. Ph.D Thesis, Civil
Engineering Department, University of
Leeds, UK.
Teguh, M. (2007). Seismic Performance of
Prestressed Concrete Pile-to-Pile Cap
Connections. Wednesday 17
th

May
2006, ISBN CD-Rom 0-Ph.D Thesis,
Department of Civil and Environ-
9553490-0-1 , 978-0-9553490-0-3, Edited
by Dr. Lindon Sear, Paper Ref. A16.
Teguh, M., Duffield, C. F., Mendis, P. A., and
Hutchinson, G. (2006). Seismic
56
Performance of Pile-to-pile Cap
Connections: An Investigation of Design
Issues. Electronic Journal of tructural
Engineering (eJSE), Vol. 6, p. 8-18.
Thanaya, I. N. A., Forth, J. P., and Zoorob, S.
E. (2006). Incorporation of Fly Ash
and Furnace Bottom Ash in Bitublock.
Proceedings of AshTech 2006,
International Coal Ash Technology
Conference, The Birmingham
Hippodrom Theatre, Birmingham, West
Midlands, UK, Sunday 15
th
the
Environment, Arup Economics and
Planning, HMSO, London.
Vekey de, R. C. (2001). Brickwork and
Blockwork, Construction Materials, Their
nature and Beahviour. Third Edition,
Editted by J. M. Illston and P. L. J.
Domone, Spon Press, London and New
York, p. 288.
Whitbread, M., Marsay, A., and Tunnell, C.
(1991). Occurrence and Utilisation of
Mineral and Construction Wastes.
Report for the Department of Metal
Engineering, University of Melbourne,
Australia.
Whiteoak, D. (1991). The Shell Bitumen
Handbook.
57
Pengembangan Peta Hazard Kegempaan untuk Pulau Jawa dan
Perbedaannya dengan Hasil-Hasil Studi Terdahulu

Lalu Makrup, Sarwidi, dan Susilo
Staf Pengajar Program Studi Teknik Sipil, UII Yogyakarta dan Alumni Program Magister Teknik Sipil UII Yogyakarta (Pegawai
Pemkab Kulon Progo), email: lalu.makrup@ftsp.uii.ac.id, sarwidi@ftsp.uii.ac.id

Abstract: The island of Java with high population density, tectonically it is an area with a fairly
high seismicity in Indonesia. Seismic hazard analysis for Java was done using the theory of total
probability. Results of calculation and development of PSHA map show that seismic hazard with
risk exceeded 10% during 50-year building useful life or nearly 500-year return period of
earthquake show that maximum soil acceleration at bedrock is between 0.1 g to more than 0.5 g.
Range of this acceleration is similar to a study conducted by Irsyam et al., (2008) and ISO 2010
and is different from SNI 2002. However, different patterns of acceleration map exist between
calculated and others. Java earthquake map is also developed with return period of 500 years (10%
probability of exceedance in 50 years) and 2,500 years (2% probability of exceedance in 50 years)
with spectra period 0.0 seconds, 0.2 seconds and 1.0 second.

Keywords: Hazard analysis, probabilistic seismic hazard, peak ground acceleration, and Java
earthquake map.

PENDAHULUAN

Pulau Jawa dengan jumlah penduduk dan
tingkat kepadatannya yang tinggi secara
tektonik merupakan kawasan dengan tingkat
aktivitas kegempaan yang cukup tinggi di
Indonesia. Kondisi ini disebabkan karena
pulau ini berdekatan dengan zona subduksi
selatan Jawa. Disamping rawan gempabumi
akibat aktivitas tumbukan lempeng tektonik
pada zona subduksi, Pulau Jawa juga rawan
gempabumi akibat aktivitas sesar-sesar lokal
di daratan. Kondisi tektonik semacam ini
menjadikan Pulau Jawa sebagai kawasan
seismik aktif yang kompleks. Salah satu upaya
untuk mengurangi dampak akibat bencana
alam gempabumi adalah melalui pengembang
an peta bahaya akibat gempa bumi. Peta
bahaya (hazard) kegempaan atau disingkat
peta gempa dikembangkan berdasarkan
analisis probabilitas bahaya kegempaan
(probabilistic seismic hazard analysis). Peta
ini merupakan peta percepatan tanah
maksimum di batuan dasar (peak ground
acceleration) yang dapat digunakan untuk
perencanaan struktur bangunan tahan gempa.
Peta hazard kegempaan dibuat berdasar kan
data catatan kejadian gempa yang pernah
terjadi sebelumnya dan diolah dengan prinsip-
prinsip statistik dalam analisis seismic hazard
probabilistik (PSHA). Sejak diterbitkannya
peta wilayah gempa pada tahun 2002, sampai
dengan saat ini di Indonesia telah banyak
terjadi gempa dengan kekuatan signifikan
yang menimbulkan kerusakan dan korban
jiwa. Sebagai contoh adalah gempa Yogya
Mei 2006 dan Tasikmalaya September 2009.
Gempa Yogya M6.3 memakan ribuan korban
jiwa dan harta benda di Yogya dan Jawa
Tengah, sedang gempa Tasikmalaya M7.3
dengan puluhan korban jiwa dan harta benda.
Dalam studi ini dilakukan analisis hazard
gempa untuk Pulau Jawa dengan bantuan
Program PSHA Seismic Risk Model (Makrup,
2009). Model sumber gempa yang digunakan
adalah model tiga dimensi dikombinasi dengan
model sumber seismisitas background.
Sebelum ini sekitar tahun 2002 kebawah di
Indonesia, perhitungan hazard gempa
dilakukan dengan menggunakan program
komputer yang mengaplikasikan model
sumber gempa area (dua dimensi). Sebagai
58
Makrup, dkk: Pengembangan Peta Hazard Kegempaan untuk Pulau Jawa dan ... Terdahulu
contoh adalah peta gempa dalam SNI 2002.
Peta gempa ini dikembangkan oleh 4 tim yang
berbeda. Yang pertama adalah peta gempa
yang dikembangkan oleh Shah dan Boen,
(1996) dari kosultan swasta, menggunakan
program GSHAP (The Global Seismic Hazard
Assessment Program), yang kedua oleh
Engkon Kertapati, (1999) dari pusat geologi,
Bandung menggunakan program EQRisk
(McGuire, 1976), yang ketiga oleh Nayoan,
(1999) dari Direktorat Jendral Pengairan
Depatemen Pekerjaan Umum menggunakan
program GSHAP dan yang keempat
Firmansyah dan Irsyam (1999) dari ITB
Bandung menggunakan program EQRisk
(McGuire, 1976). Karena tidak ada
kesepakatan tentang peta gempa yang mana
yang akan digunakan dalam SNI 2002, maka
nilai numerik dari keempat peta tersebut lalu
dirata-rata, dan peta gempa dengan nilai rata-
rata inilah yang digunakan sebagai peta gempa
untuk SNI 2002 (Gambar 1).

Gambar 1. Peta gempa PGA Jawa dengan
level hazard 10 % probabilitas terlampaui
dalam 50 tahun atau mendekati perioda ulang
gempa 500 tahun (SNI, 2002)

Peta gempa Jawa dengan pemodelan sumber
gempa tiga dimensi yang digabungkan dengan
model sumber gempa sismisitas background
dua dimensi, telah dikembangkan oleh Irsyam
et al., (2008). Program yang digunakan dalam
analisis adalah program EZ Frisk (McGuire,
2005) dengan tidak mengakomodasi fungsi
atenuasi terbaru NGA (The Next Generation
Attenuation) karena pada saat pengembangan
peta ini EZ Frisk belum mengakomodasi NGA
(Gambar 2).
Peta gempa Jawa dalam rencana peta gempa
SNI 2010 dikembangkan dengan pemodelan
sumber gempa tiga dimensi dan digabung
dengan model sumber gempa sismisitas
background Frankel, (1995).

gempa 500 tahun (Irsyam et al., 2008)

Peta ini dikembangkan menggunakan program
USGS (Asrurrifak, 2010) dengan
mengakomodasi fungsi atenuasi terbaru NGA
(Gambar 3).

gempa 500 tahun (SNI, 2010)

Model sumber gempa tiga dimensi yang
digabung dengan model sumber gempa
sismisitas background dua dimensi, dan fungsi
atenuasi generasi baru (NGA) diakomodasi
dalam program kegempaan SR-Model (The
Seismic Risk Model, Makrup, 2009). Program
ini digunakan dalam PSHA untuk studi
mengembangkan peta gempa Jawa dalam
paper ini. Hasil analisis akan diberikan pada
sebelum bagian akhir dari paper ini.
59
ANALISIS SEISMIK HAZARD
PROBABILISTIK

Konsep probabilistik telah memberi ruang bagi
ketidakpastian dari site, lokasi, dan laju
keberulangan gempa serta variasi karakteristik
gerakan tanah secara eksplisit dapat
dipertimbangkan dalam melakukan evaluasi
terhadap hazard kegempaan. Analisis hazard
kegempaan probabilistik (PSHA)
menyediakan kerangka yang memungkinkan
ketidakpastian dapat diidentifikasi,
kuantifikasi, dan dikombinasikan dalam cara-
cara yang rasional untuk menyediakan
gambaran yang lebih lengkap tentang hazard
kegempaan (Kramer, 1996).
Tahapan analisis dalam hazard kegempaan
dengan metode Probabilistic Seismic Hazard
Analysis (PSHA) secara tipikal diuraikan
menjadi empat tahap proses (Reiter, 1990)
yang dapat dijelaskan sebagai berikut ini.
1. Tahap pertama adalah identifikasi dan
karakterisasi sumber gempa, termasuk
didalamnya adalah karakterisasi distribusi
probabilitas dari lokasi rupture yang
berpontensi menyebabkan gempa pada
sumber. Dalam kebanyakan kasus,
diterapkan distribusi probabili tas yang
sama untuk masing-masing zona sumber.
Hal ini secara tidak langsung menyatakan
bahwa gempa mungkin sama-sama akan
terjadi pada setiap titik dalam zona sumber
gempa. Distribusi ini, dikombinasikan
dengan bentuk geometri sumber untuk
mendapatkan distribusi probabilitas yang
sesuai dengan jarak sumber ke site.
2. Langkah berikutnya adalah karakterisasi
seismisitas atau distribusi perulangan
kejadian gempa. Hubungan empiris
perulangan kejadian gempa (recurrence
relationship), yang meng ekspresikan laju
rata-rata (average rate) suatu gempa
dengan magnitude yang berbeda akan
terlampaui. Hubungan ini digunakan untuk
meng karakterisasi seismisitas masing-
masing zona sumber gempa. Hubungan
empiris ini dapat mengakomodasi
magnitude maksimum gempa.
3. Gerakan tanah yang terjadi disuatu lokasi
akibat adanya gempa dengan besar gempa
berapapun dan lokasi kejadian dimanapun
dalam masing-masing zona sumber gempa,
dapat ditentukan dengan menggunakan
predictive relationships.
4. Langkah terakhir adalah mengkom bina-
sikan ketidakpastian dari lokasi gempa,
magnitude gempa dan prediksi parameter
gerakan tanah untuk mendapatkan
probabilitas dimana parameter gerakan
tanah akan terlampaui selama perioda
waktu tertentu.
Untuk satu kejadian gempa tertentu,
probabilitas satu parameter gerakan tanah Y
akan melampaui satu nilai y* tertentu dapat
dihitung menggunakan teorema probabilitas
total (Kramer, 1996). Bentuk fungsi dari
teorema probabilitas total (McGuire, 1976)
adalah:
| | | | drdm ) r ( f ) m ( f r , m | * y Y P * y Y P
R M } }
> = >
(3)
dengan | | r , m | * y Y P > diperoleh dari
hubungan prediktif probabilitas parameter Y
yang akan melampaui nilai y* pada lokasi
yang ditinjau untuk kejadian gempa dengan
magnitude M dan jarak hiposenter R,
sedangkan f
M
(m) dan f
R
(r) adalah fungsi
kerapatan probabilitas (PDF) masing-masing
untuk magnituda dan jarak. Untuk
memperoleh nilai tahunan rata-rata ter
lampaui ( = annual rate of exceedance)
persamaan (3) dapat ditulis menjadi :
| | drdm r f m f r m a A P a
R M A
) ( ) ( , | ) (
} }
> = v
(4)
dengan v adalah rate dari kejadian gempa.
Penyelesaian persamaan (4) dalam domain
sumber gempa sangat rumit, sehingga secara
umum penyelesaian tersebut di tuangkan
dalam bentuk software.

METODE

Tempat dan Data
Data gempa yang digunakan dalam penelitian
ini adalah seluruh kejadian gempa tercatat
yang pernah terjadi di Pulau Jawa dan
sekitarnya dengan magnitude minimal Mw = 5
dan kedalaman hipocenter gempa maksimal
60
250 km. Data gempa yang dipakai dalam
penelitian ini diperoleh dengan men-download
dari Advanced National Seismic System
(ANSS) dan National Earthquake Information
CenterUnited States Geological Survey
(NEIC-USGS) dengan rentang waktu sampai
dengan kejadian gempa bulan Desember 2010,
meliputi data episenter gempa yang terjadi di
Pulau Jawa dan sekitarnya yaitu yang berada
pada rentang koordinat 100 120 bujur
timur dan 2 13 lintang selatan.

Konversi Skala Magnitude
Supaya didapatkan hasil analisis seismic
hazard yang akurat maka dalam analisis
seismic hazard diperlukan data kejadian gempa
yang memiliki skala magnitude yang sama,
dalam hal ini adalah skala magnitude moment
Mw. Konversi skala magnitude ke dalam skala
momen magnitude untuk gempa-gempa di
wilayah Indonesia dilakukan dengan
menggunakan persamaan yang diperkenalkan
oleh Hendriyawan (2001), sebagai berikut:
M
w
= 0,99M
B
+ 0,253 (5)
M
w
= 0,08M
S
2
+ 0,04M
S
+ 3,01 (6)
Dengan menggunakan persamaan (5) dan (6),
data gempa yang mempunyai skala magnitude
selain Mw kemudian dikonversi ke dalam
skala magnitude Mw sehingga diperoleh skala
magnitude yang seragam.

Pemilihan Gempa Utama
Pemisahan gempa utama dari gempa-gempa
ikutan (foreshock dan aftershock) dilakukan
agar diperoleh data gempa dengan kejadian
yang independen. Kejadian gempa dependen
atau gempa ikutan, seperti foreshock dan
aftershock yang terjadi dalam suatu rangkaian
kejadian gempa harus diidentifikasi dan di
lepaskan dari katalog sebelum katalog tersebut
digunakan dalam analisis. Dalam studi ini,
untuk mengeliminasi foreshock dan aftershock
dari katalog gempa digunakan kriteria Gardner
dan Knopoff (1974). Hasil dari proses analisis
pemilihan gempa utama (dependency analysis)
adalah data gempa utama yang kemudian
digunakan pada proses analisis selanjutnya.

Analisis Kelengkapan Data Gempa
Analisis kelengkapan data gempa dilakukan
untuk mengetahui kelengkapan data gempa
dalam rentang waktu tertentu. Kelengkapan
data diperlukan mengingat hasil proses analisis
probabilistik yang akan dilakukan sangat
tergantung dari data tersebut. Analisis seismik
hazard memerlukan data gempa dalam kurun
waktu tertentu, dimana kejadian gempa
independen dalam rentang magnitude tertentu
dapat dikatakan lengkap dalam suatu katalog
gempa. Data pencatatan kejadian gempa
historis untuk kejadian-kejadian gempa besar
lebih lengkap dibanding kejadian-kejadian
gempa kecil. Hal ini disebabkan pada masa
lalu jumlah alat pencatat gempa belum banyak
dan teknologinya belum maju seperti sekarang,
sehingga alat-alat tersebut hanya mencatat
kejadian-kejadian gempa besar. Jika data yang
tidak lengkap digunakan dalam analisis resiko
gempa, maka hasil yang didapat akan terlalu
kecil (underestimated) untuk gempa-gempa
kecil dan terlalu besar (overestimated) untuk
kejadian gempa besar (Aldiamar, 2007).

Identifikasi dan Pemodelan Sumber Gempa
Sumber gempa yang digunakan dalam
pembuatan peta hazard kegempaan Pulau Jawa
meliputi zona sumber gempa subduksi dan
sumber gempa sesar di daratan. Secara umum
digunakan tiga zona sumber gempa dalam
studi ini yaitu (1) zona megatrust yang
merupakan sumber gempa dangkal pada
daerah subduksi dengan kedalaman maksimum
50 km, (2) zona benioff yang merupakan
sumber gempa pada daerah subduksi dengan
kedalaman antara 50 km hingga 250 km, dan
(3) zona shallow crustal pada daerah
transform fault. Bentuk zona ini adalah
patahan yang biasa disebut sesar (fault) dengan
kedalaman hingga 20 km. Sumber gempa
subduksi dibagi menjadi dua yaitu zona
subduksi megatrust dan zona subdaksi benioff.
Kejadian gempa akibat thrust fault, normal
fault, reverse slip dan strike slip yang terjadi
sepanjang pertemuan lempeng dapat
diklasifikasikan sebagai zona subduksi. Zona
subduksi megatrust terletak pada pertemuan
antar lempeng tektonik yaitu pertemuan antara
61
lempeng samudera yang menunjam masuk ke
bawah lempeng benua dan memiliki sudut
penunjaman yang kecil (landai) pada
kedalaman sampai dengan 50 km, sedang kan
zona subduksi benioff adalah kelanjut an dari
zona subduksi megatrust dengan kemiringan
sudut penunjaman yang relatif lebih besar.

Tabel 1. Zona sumber gempa Pulau Jawa dan
sekitarnya
No Zona Sumber Gempa
A Megatrust
1. SouthernSumatra Megatrust
2. Jawa-1 Megatrust
3. Jawa-2 Megatrust
4. NusaTenggara Megatrust
B Benioff
1. SouthernSumatra Benioff
2. Jawa-1 Benioff
3. Jawa-2 Benioff
4. NusaTenggara Benioff
C Shallow Crustall
1. Sesar Sumatra
2. Sesar Cimandiri
3. Sesar Baribis
4. Sesar Bumiayu
5. Sesar Semarang
6. Sesar Sampanahan
7. Sesar Opak
8. Flores Back Arc

Gambar 4. Zona sumber gempa untuk Pulau
Jawa

Gempa pada fault umumnya terjadi pada zona
dangkal dengan mekanisme strike-slip,
reverse, oblique, atau normal. Oleh karena itu
zona fault hanya sampai kedalaman 20 km.
Zona sumber gempa subduksi dan fault yang
secara umum sudah dikenal sebagai sumber
gempa aktif di Jawa dapat dilihat pada Tabel-1
dan Gambar 4.

Atenuasi
Salah satu model fungsi yang harus tersedia
dalam PSHA adalah fungsi atenuasi. Fungsi
atenuasi merupakan for mula yang
menunjukkan hubungan antara parameter
gerakan tanah misalnya percepatan (a),
magnitude (M), jarak (R) dan kondisi tanah
site (f
s
) dari kejadian gempa pada suatu
wilayah. Fungsi atenuasi kebanyakan
dikembangkan di Amerika Serikat dan sangat
sedikit dari tempat lain. Oleh karena itu hingga
saat ini belum ada fungsi atenuasi yang
dihasilkan secara spesifik bagi wilayah
Indonesia karena kurangnya data-data yang
dibutuhkan untuk menghasilkan fungsi
atenuasi tersebut.
Fungsi atenuasi yang digunakan di Indonesia
diambil dari negara lain. Dalam penggunaan
fungsi atenuasi ini diasumsi kan bahwa kondisi
tektonik dan site di Indonesia mirip dengan
kondisi tektonik dan site dimana fungsi
atenuasi itu dikembangkan. Dalam studi ini
fungsi atenuasi yang digunakan untuk zona
sumber gempa subduksi adalah fungsi atenuasi
Youngs (1997) dan fungsi atenuasi Atkinson
Boore (2003). Sedangkan zona fault digunakan
fungsi atenuasi Boore, Joyner dan Fumal
(1997) dan fungsi atenuasi Boore Atkinson
(2006) NGA.

HASIL ANALISIS

Perhitungan dengan PSHA dilakukan atas
dasar level hazard 10% dan 2% probabili tas
terlampaui dalam jangka waktu masa guna
bangunan 50 tahun atau masing-masing
mendekati periode ulang gempa 500 tahun
2500 tahun. Perioda spektra yang ditetapkan
dalam perhitungan adalah 0 detik (PGA), 0.2
detik dan 1.0 detik.
Hasil perhitungan dengan periode ulang
gempa mendekati 500 tahun menunjukkan
bahwa percepatan tanah maksimum pada
62
batuan dasar untuk Pulau Jawa adalah antara
0.1g hingga lebih dari 0.5g. Sedangkan untuk
periode ulang gempa mendekati 2500 tahun
menunjukkan bahwa percepatan tanah
maksimum pada batuan dasar berada diantara
0.15g hingga lebih dari 0.8g.

perioda ulang gempa mendekati 500 tahun

Gambar 6. Peta gempa Jawa dengan perioda
spektra 0.2 detik dan perioda ulang gempa
mendekati 500 tahun

mendekati 500 tahun

perioda ulang gempa mendekati 2500 tahun
Hasil pengembangan peta gempa untuk pulau
Jawa adalah berupa peta kontur percepatan
gerakan tanah dibatuan dasar dengan periode
ulang gempabumi 500 tahun dan 2500 tahun
dengan periode spektra 0 detik (PGA), 0.2
detik dan 1.0 detik. Peta hasil analisis dapat
dilihat dalam Gambar-5, 6 dan 7 untuk perioda
ulang gempa 500 tahun serta Gambar-8, 9 dan
10 untuk perioda ulang gempa 2500 tahun.

PEMBAHASAN

Yang akan dibahas disini adalah perbedaan
peta gempa PGA Jawa dengan perioda ulang
500 tahun antara peta hasil studi dan peta hasil
studi terdahulu. Secara visual perbedaan ini
ditampilkan dalam bentuk gambar.

mendekati 2500 tahun

mendekati 2500 tahun

Gambar 11, 12 dan 13 adalah tampilan
perbedaan antara peta gempa pulau Jawa hasil
studi dengan peta gempa pulau Jawa dari SNI
2002, Irsyam et al., (2008) dan SNI 2010.
Peta gempa PGA Jawa hasil studi dan SNI
2002 (Gambar 11) mempelihatkan perbe daan
yang relatif kontras terutama pada wilayah
Jawa yang ditempat tersebut dinyatakan ada
63
fault. Pada gambar peta gempa hasil studi
terlihat adanya konsentrasi percepatan pada
dari wilayah Jawa Tengah sampai wilayah
Jawa barat. Sebagai contoh pada lokasi fault
Cimandiri (Gambar 11A) terlihat bahwa
konsentrasi percepatan gerakan tanah 0.3g.
Bila dibandingkan dengan peta gempa PGA
Jawa SNI 2002 maka terlihat dalam peta
tersebut seolah-olah tidak ada konsentrasi
percepatan pada wilayah tertentu yang
dianggap sebagai pusat sumber gempa. Inilah
perbedaan nyata yang dapat diperlihatkan
Gambar 11. Perbedaan peta gempa PGA Jawa hasil studi dan SNI 2002
Gambar 12. Perbedaan peta gempa PGA Jawa hasil studi dan Irsyam et al., (2008)
64
secara visual dari peta hasil studi (Gambar
11A) dan peta SNI 2002 (Gambar 11B).
Perbedaan lain yang juga dapat disampaikan
disini adalah bahwa peta Gambar 11A
memperlihatkan percepatan gerakan tanah
berkisar antara 0.1g sampai dengan 0.5g,
sedang peta SNI2002 berkisar antara 0.03
sampai dengan 0.3g. Perbedaan ini terjadi
barangkali karena perbedaan model sumber
gempa yang digunakan yaitu pada peta hasil
studi menggunakan model sumber gempa tiga
dimensi, sedang pada SNI 2002 menguna kan
model sumber gempa dua dimensi. Begitu juga
mengenai penggunaan fungsi atenuasi. Dalam
pengembangan peta gempa hasil studi
(Gambar 11A) sudah digunakan fungsi
atenuasi mutakhir NGA 2006/2007 sedang
pada peta SNI 2002 (Gambar 11A) tentunya
menggunakan fungsi atenuasi di bawah tahun
2000. Walaupun di tempat-tempat tertentu
terlihat ada kesamaan per cepatan tanah gempa
antara Gambar 11A dan 11B namun luasan
yang dicakup berkaitan dengan kesamaan ini
tidak terlalu signifikan.
Perbedaan antara peta gempa PGA Jawa hasil
studi dengan hasil Irsyam et al. (2008) dapat
dilihat pada Gambar 12. Sebagaimana
perbedaan antara peta hasil studi dan SNI 2002
(Gambar 11), maka pada Gambar 12 masih
tetap memperlihatkan bahwa perbeda an yang
relatif kontras terjadi pada posisi keberadaan
fault yang digunakan dalam analisis. Sebagai
contoh pada posisi fault Semarang dan Yogya
yang digunakan untuk mengembangkan peta
Gambar 12C terlihat adanya konsentrasi
percepatan gerakan tanah, sementara dalam
Gambar 12D ditempat tersebut tidak terjadi
konsentrasi percepatan karena pengembangan
peta Gambar 12D tidak menggunakan fault
Semarang dan Yogya sebagai dasar
perhitungan. Di Jawa bagian barat dalam
Gambar 12C terlihat adanya konsentrasi
percepatan pada posisi fault Cimandiri dan
Baribis. Sementara dalam Gambar 12D tidak
terlihat hal yang serupa. Memang ada
diperlihatkan gambar konsentrasi percepatan
menggelembung pada Jawa Barat Bagian
Timur, yang itu mungkin karena penggunaan
fault Baribis, namun tetap saja
penggelembungan ini tidak memperlihatkan
pola yang sama dengan konsentrasi percepatan
yang ada pada Gambar 12C. Adanya
perbedaan peta gempa PGA Jawa antara peta
hasil studi (Gambar 12C) dan hasil dari Irsyam
et al. (2008) (Gambar 12D) barang kali
disebabkan terutama oleh adanya penggunaan
Gambar 13. Perbedaan peta gempa PGA Jawa hasil studi dan SNI 2010
65
fault yang berbeda dan panjang fault yang
berbeda walaupun juga ada fault yang sama.
Disamping itu juga oleh penggunaan fungsi
atenuasi yang berbeda. Peta Gambar 12C
melibatkan fungsi atenuasi NGA sebagai dasar
pengembangan sedangkan peta Gambar 12D
tidak. Secara keseluruhan peta Gambar 12C
dan Gambar 12D memberikan rentang per
cepatan gerakan tanah yang sama yaitu
berkisar antara 0.1g sampai dengan 0.5g,
namun percepatan gerakan tanah yang sama
hanya terjadi pada tempat-tempat tertentu saja,
sehingga perbedaan antara kedua peta dalam
Gambar 12 tetap masih cukup signifikan.
Perbedaan antara peta gempa PGA Jawa hasil
studi dengan hasil SNI 2010 dapat dilihat pada
Gambar 13. Sebagaimana perbedaan antara
peta hasil studi dan SNI 2002 (Gambar 11)
serta antara peta hasil studi dan Irsyam et al.,
(2008) maka perbedaan antara peta hasil studi
dan peta SNI 2010 memperlihatkan pola
perbedaan yang tidak terlalu jauh berbeda
dengan apa yang diperlihatkan sebelumnya.
Pada peta SNI 2010 Gambar 13F terlihat
konsentrasi percepatan pada fault Cimandiri
memiliki pola konsentrasi yang berbeda
dengan fault Cimandiri dalam peta hasil studi
Gambar 13E. Begitu juga dengan konsentrasi
percepatan pada fault Lembang sekitar
Bandung (Gambar 11F) berbeda dengan pola
yang terjadi pada Gambar 13E, karena dalam
Gambar 13E tidak digunakan fault Lembang.
Disekitar Semarang gabungan kontribusi fault
Semarang, Pati dan Lasem memperlihat
konsentrasi percepatan (Gambar 13F) yang
berbeda dengan yang dikontribusi oleh fault
Semarang saja pada Gambar 13E. Juga
perbedaan semacam ini terjadi juga ditempat-
tempat yang lain dalam kedua Gambar 13E
dan 13F. Kesamaan percepatan gerakan tanah
dapat dijumpai hanya pada tempat-tempat
tertentu dan dalam keseluruhannya memliki
rentang percepatan yang sama antara Gambar-
13E dan 13F yaitu dengan nilai percepatan
dari 0.1g sampai dengan 0.5g.

KESIMPULAN

Dari peta hazard kegempaan Pulau Jawa
dengan resiko terlampaui sebesar 10% selama
masa guna bangunan 50 tahun atau setara
dengan periode ulang gempa 500 tahun
memperlihatkan bahwa percepatan tanah
maksimum (PGA) pada batuan dasar untuk
Pulau Jawa adalah antara 0.1g hingga lebih
dari 0.5g. Hal sama terjadi pada peta gempa
PGA Jawa hasil dari Irsyam et al. (2008) dan
SNI 2010. Namun hal ini berbeda dengan yang
ada pada peta gempa PGA yang ada dalam
SNI 2002 dengan percepatan berkisar antara
0.03g sampai dengan 0.3g. Sehingga dapat
dinyatakan bahwa peta hasil studi, peta Irsyam
et al. (2008) dan SNI 2010 memberikan nilai
percepatan tanah PGA dibatuan dasar yang
lebih tinggi bila dibanding dengan peta gempa
pada SNI-03-1726-2002.

UCAPAN TERIMA KASIH

Terima kasih kami ucapkan kepada Biro
Perencanaan Kerjasama Luar Negeri (BPKLN)
Kemendiknas, atas beasiswa yang telah
diberikan pada mahasiswa bimbingan kami
saudara Susilo. Tesis saudara Susilo telah
diterbitkan dalam Jurnal ini yang ditulis secara
bersama-sama dengan dosen pembimbing.

DAFTAR PUSTAKA

Aldiamar, F. (2007). Analisis Risiko Gempa
dan Pembuatan Respon Spektra untuk
Jembatan Suramadu dengan Pemodelan
Sumber Gempa Tiga Dimensi. Thesis,
Institut Teknologi Bandung, 2007.
Asrurrifak, M. (2010). Peta Respon Spektra
Indonesia untuk Perencanaan Struktur
Bangunan Tahan Gempa dengan Model
Sumber Gempa Tiga Dimensi dalam
Analisis Probabilistik. Disertasi, Institut
Technologi Bandung, Indonesia.
Atkinson, G. M. dan Boore, D., M. (2003).
Empirical Ground Motion Relations for
Subduction-Zone Earthquakes and Their
Aplication to Cascadia and Other
Region. Bulletin of the Seimological
Society of America, Vol. 93, No. 4, p.
1703-1729.
Boore, D. M., Atkinson, G., M., dan
NGA/Next Generation Attenuation.
(2007). Ground Motion Relations for
66
Geometric Mean Horizontal Component
of Peak and Spectral Ground Motion
Parameters. PEER Report 2007, Pacific
Earthquake Engineering Research Center,
College of Engineering University of
California, Berkeley, California, USA.
Boore, D. M., Joyner, W. B., and Fumal, T. E.
(1997). Equation for Estimating
Horizontal Response Spectra and Peak
Acceleration from Western North America
Earthquakes: A Summary of Recent
Work. Seismological Research Letters,
Vol. 68, No. 1.
Firmansjah, J. dan Irsyam, M. (1999).
Development of Seismic Hazard Map for
Indonesia. Prosiding Konferensi
Nasional Rekayasa Kegempaan, Bandung.
Frankle, A. (1995). Mapping Seismic Hazard
in the Central and Eastern United States.
Seismological Research Letters, Vol. 6,
No. 4.
Gardner, J. K. L. And Knopoff. (1974). Is the
Sequence of Earthquakes in Southern
California, with Aftershocks Removed,
Poissonian? Bull. Seismol. Soc. Am. 64,
p. 13631367.
GSHAP. (1999). The Global Seismic hazard
Assessment Program (GSHAP) 1992
1999. Ann. Geofis., p. 9571230,
(summary volume edited by D. Giardini).
Irsyam M., Dangkua D. T., Hendriyawan,
Hoedajanto, D., Hutapea B. M., Kertapati
E. K., Boen T., dan Petersen, M. D.
(2008). The Proposed Seismic Hazard
Maps of Sumatra and Java Islands and
Microzonation Study of Jakarta City,
Indonesia. J. Earth Syst. Sci. 117, S2, p.
865878.
Kertapati, E. K. (1999). Probabilistic
Estimates of Seismic Ground-Motion
Hazard in Indonesia. Prosiding
Konferensi Nasional Rekayasa
Kegempaan, Bandung.
Kramer, S. L. (1996). Geotechnical
Earthquake Engineering., Prentice-Hall,
Inc., Upper Suddle River, New Jersey.
Makrup, L. L. (2009). Pengembangan Peta
Deagregasi Hazard Untuk Indonesia
Melalui Pembuatan Software Dengan
Pemodelan Sumber Gempa Tiga
Dimensi. Disertasi Doktor, Institut
Teknologi Bandung, Bandung.
McGuire, R. K. (1976). FORTRAN Computer
Program for Seismic Risk Calculations.
US. Geol. Surv. Open-File Rep. 76-67, p.
90.
Najoan, T. F., Suharjoyo, A., Buditomo, A.,
Wibowo, S., dan Rizaldi, Nasution, R. B.
(1999). Peta Zona Gempa Indonesia
untuk Penentuan Percepatan Gempa
Maksimum di Permukaan. Prosiding
Konferensi Nasional Rekayasa
Kegempaan, Bandung.
Reiter, L. (1990). Earthquake Hazard
Analysis-Issues and Insights. Columbiai
University Press, New York, p. 254.
Shah, H. C. dan Boen, T. (1996). Seismic
Hazard Model for Indonesia,
Unpublished.
SNI/Standar Nasional Indonesia. (2002). Tata
Cara Perencanaan Ketahanan Gempa
untuk Bangunan Gedung (SNI 03-1726-
2002). Badan Standardisasi Nasional.
Youngs, R. R., Chiou, S. J., Silva, W. J., dan
Humphrey, J. R. (1997). Strong Ground
Motion Attenuation Relationships for
Subduction Zone Earthquakes.
Seismological Research Letters, 68, p. 58-
73.
67
Shoreline Change Model Using the Epr Method and the Simulation of Coastal
Vulnerability in Sambas District-West Kalimantan

M. Meddy Danial, Rustamaji, and Eka Priadi
Engineering Faculty of Tanjungpura University, Indonesia, email: meddystmt@yahoo.com, email: maji_soenantyo@yahoo.com,
email: eka_priadi@yahoo.com

Abstract: Potential threats to the coastline increase from year to year because of global climate
change that increases the temperature of the earth and sea level rise. West Kalimantan has a long
coastline of 982 km. Potential length of this beach should be maintained in order not to decrease.
District of Jawai and Jawai Selatan is one of the coastal areas, which suffered severe erosion. This
study aims to model the pattern of coastline changes and the index of coastal vulnerability to
current conditions and future conditions using physical parameters such as waves, tides,
bathymetry, land cover and sea level rise (SLR). This research is conducted using satellite image
map data is processed with ArcGIS 9.3, Erdas and Autocad. Shoreline changes were analyzed by
End Point Rate method. From the results of the shoreline change analysis obtained information that
the length of the beach erosion from Jawai Selatan to SB Nilam is about 4 km and the eroded area
is 9557.546 m
2
. Sambas coastal areas in the district Jawai Selatan until the district Jawai are
susceptible to the existing conditions (first scenario). Vulnerability score on shoreline occupied
criteria "High", where erosion rate on average per year is about 6 m. In the second scenario with
tidal elevation is 3 m, wave height is 3 m, and SLR is 0.5 m, obtained results of a vulnerability
index is generally dominated by the conditions of very high vulnerability. In the third scenario with
tidal elevation is 4 m, wave height is 4 m, and SLR is 1 m, obtained results of a vulnerability index
is generally dominated by the conditions of very high vulnerability.

Keywords: Shoreline changes, coastal vulnerability index, recent condition, and future model
analysis.

INTRODUCTION

The coast in Jawai Selatan until Jawai is
low-lying area and the topography is quite
flat. Prior to 1989, Coastal area of Jawai
Selatan was vegetated by mangrove tree and
since 1998 had damaged. In 2009, there are
several indicators of coastal damage in
district Jawai and Jawai Selatan, such as
collapse of many coconut tree and shoreline
retreat.
Erosion problem in Jawai and Jawai Selatan is
due to the fishpond conversion, sand mining
and position of breakwater which is too
close from the shoreline that can cause
serious con- sequences of rapidly increasing
erosion behind the breakwater.
Therefore, it is important to investigate the
shoreline changes, erosion rate per year and
how these condition affect the vulnerability
of coastal area in district Jawai and Jawai
Selatan.

OBJECTIVES

This study aims to identify and model the
pattern of coastline changes between 1992
until 2006. The result can be used to predict
the shoreline changes for the future.
The second objective is to identify the index
of coastal vulnerability againts current
conditions and future conditions using
physical parameters such as waves, tides,
bathymetry, topography, land cover and sea
level rise (SLR). This study is expected to
provide valuable input to local government in
overcoming the problem of coastal
vulnerability and coastal hazards.
68
Danial, e.t al.:
STUDY A

The study
tan coast to
Pontianak.
Selatan an
topography
low-lying a
Characteris
comprised
can be fou
Jawai coas
In Jawai
damage du
and occur
location of

METHOD

The metho
sensing to
satellite im
change can
on the End
vulnerabili
: Shoreline Cha
AREA
area is loc
o SB Nilam
The coas
nd SB Nila
y in this ar
area.
stic of
of sandy b
und in Ram
st is domina
Selatan, t
ue to mism
rs erosion
f study can b
DOLOGY
od of this r
o extract s
mage using s
n be analyz
d Point Rat
ity index (
Figure 1.
ange Model Usin
cated on th
m coast of S
st length b
am is about
rea is flat a
Jawai Se
beach and m
mayadi beac
ated by sand
the fishpon
management
behind bre
be seen in F
research is
shoreline p
software Er
zed using A
e (EPR) me
(CVI) can
. Study Area
ng the Epr Met
he Jawai Se
Sambas distr
etween Jaw
t 31 km. T
and constit
elatan co
mangrove tre
ch, whereas
dy beach.
nd is alrea
t of fishpo
eakwater. T
Figure 1.
using rem
position fro
rdas. Shorel
Autocad bas
ethod. Coas
be analyz
a
thod and the Sim
ela-
rict
wai
The
tute
oast
ees
s in
ady
ond
The
mote
om
ine
sed
stal
zed
us
wa

Re
De
us
req
sh
are
19
Th
to
Th
us
(A

EP
Th
tio
EP
sh
po
im
mulation of Coa
sing ArcGis
as roposed b
emote Sens
etection of
sing remote
quire sate
horeline po
e two satel
992 and 200
he software
obtain the
he flow ch
sing Erdas
Alesheikh, e
PR Method
he next ste
on from the
PR method
horeline ch
osition that h
mages.
Figure 2. St
astal Vulnerabili
s 9.3 based
by Gornitz
sing
shoreline c
sensing me
ellite imag
osition. In
llite image
06.
used for sa
position of
hart for pro
s can be
t al., 2007).
d
ep is analys
e result of re
is used to
hange base
have been e
teps of imag
ity in ... Kaliman
d on the eq
et al. (1991
hange can b
ethod. Rem
ge data t
this resea
data, name
atellite data
f the coastlin
ocessing sa
seen in
.
se the shor
emote sensi
o calculate
ed on two
extracted fr
ge satellite p
ntan
quation that
).
be analysed
mote sensing
to identify
arch, there
ely between
a processing
ne is Erdas
atellite data
Figure 2
reline posi-
ing process
the rate of
o shoreline
rom satellite
processing

t
d
g
y
e
n
g
.
a
2
-
.
f
e
e
69
EPR method is based on the concept of
empirical equation shows that Shoreline
changes in the future can be determined from
the linear relationship between the Shoreline
positions and times in the past or in
previous years. EPR method can be written in
the following equation (Li, R., et al., 1999).
Y = MX + B (1)
where:
Y = position of the coastline changes in year
n (the desired end of the year),
m = rate of erosion/year
= (Y
n
-Y
1
)/(X
n
-X
1
),
X = year used as a bench mark (the early
years of the start of the calculation),
B = point coordinates are used as a bench
mark in the early years of the start of the
calculation.
The Eq. 1 can be modified as follows:
m
EPR
= (Yn Y1) - (X n X1) (2)
where the EPR intercept is:
B
EPR
= Y1 m
EPR
. X1 = Yn m
EPR
. X
n
(3)
because the line end point can be added at the
opposite point of recent point (Xn, Yn), hence
the equation can be rewritten into:
Y
EPR
= m
EPR
. (X X
n
) + Y
n
(4)

Coastal Vulnerability Index
Vulnerability is defined as a specific condition
that can increase the likelihood of danger or
disaster resulting damage and loss. Coastal
vulnerability index is intended to predict
vulnerability due to coastal hazard
(Szlafsztein, 2005). The CVI is calculated by
using some physi-cal parameters of
topography, coastal erosion rate per year, the
slope of the shore base, rising sea level, wave
height and tidal elevation (Abuodha and
Woodroffe, 2006).
The parameter of CVI can be modified
according to the situation and condition
(Cambers, 2001). The CVI can be obtained
using the following equation:
CII = _
u.b.c.d.c.]
6

There are three scenarios of data input for
coastal vulnerability index on the beach Sam-
bas that can be shown in Table 2.

Table 2. Three scenarios of data input for
coastal vulnerability index
Coastline Change
Figure 3 show the pattern of shoreline change
between 1992 and 2006, and the prediction of
shoreline change in the future can be known.
From the shoreline change analysis using
remote sensing can be seen that between the

Scenario
1
Scenario
2
Scenario
3
Tide level 2 m 3m 4 m
Wave
height
2 m 3 m 4 m
Sea level
rise
0.1 m 0.5 m 1 m
Very low Low Moderate High Very high
Variable/ Score
1 2 3 4 5
Topography (m) 0.1 20.1 30.0 10.1-20.0 5.1-10.0 0 5.0
SLR change (m/tahun) -1.0 -1.0 - 0.9 1.0 2.0 2.1 4.0 4.1
Shoreline change (m) 2.1 1.0 2.0
-1.0 - +1.0
-1.1 - -2.0 -2.1
Tidal range (m) 0.99 1.0 1.9 2.0 4.0 4.1 6.0 6.1
Annual max wave height (m) 0.0 2.9 3.0 4.9 5.0 5.9 6.0 6.9 7.0
Table 1. Matrix for determination of coastal vulnerability index (CVI) from Gornitz (1991)
(5)
70
Danial, e.t al.: Shoreline Change Model Using the Epr Method and the Simulation of Coastal Vulnerability in ... Kalimantan
Ramayadi beach (sub distric Jawai Selatan) to
Sentebang beach (sub district Jawai) had
experienced a severe erosion of the coastline
changes in average 6.642 m/year. Coastal
areas in the district Jawai had became a place
of sand mining, but it was banned and
dismissed by the local government since 1997.
From the results of the shoreline change
analysis are obtained information that
the length of the beach erosion from Jawai to
SB Nilam is about 4.6 km in segment 1 with
an area undergoing erosion about 9557.546
m2. In segment 2, the length of erosion is
about 4.3 km with 11310 m2 of land area.
The process of sedimentation in segment 1 has
the length of sedimentation about 20.6 m and
land area has experienced erosion about 133
m21, while in the segment 2 the length of
sedimentation is about 5630 m with erosion of
land area about 2798 m2. Location of
sedimentation process starting from
coordinates 282005, 148207 (in UTM format)
to coordinates 281865.2173, 155155.0894.
Sediment process occur in coastal area along
the coast in the sub district Jawai, from the
village Nyirih to the beach in the villages SB
Nilam is because there are mangrove trees
that grow there.
Erosion and sedimentation rates along the
coast between Jawai Selatan and Jawai are
varying depending on the interaction of waves,
currents, defense building (such as breakwater,
jetty, etc), mangrove plants, and land use (such
as fishpond, sand mining, etc). Rate of erosion
in segment 1 and segment 2 can be seen in
Table 3.

Modeling of Coastal Vulnerability Index
Coastal vulnerability index (CVI) is the most
common worldwide method to determine
coastal vulnerability. However, the CVI
should not be regarded as a constant value
because it can change with time and require a
Figure 3. shoreline change from 1992 to
2006 in Segment 1
Figure 4. Shoreline change from 1992 to
2006 in Segment 2
71
simulation of some data to get a more realistic
vulnerability.
Scenario 1 has data H = 2 m, tide level = 2 m
and sea level rise = 0.1 m. The scenario 1 is
simulated as as daily condition or existing
condition.

In Figure 5, can be shown the coastal
vulnerable index is dominated by moderate
and high value. The land area of coastal is
dominated by moderate vulnerability value
where this area has flat topography and
low-lying area, generally.
CVI has high vulnerability value on the
shoreline due to along the shoreline has
experienced a very severe erosion. Along the
coastal regions have a very low vulnerability
index which are found in high topography,
such as rocky coastal.
Figure 6 shows the scenario 2 with wave
height = 3 m, tidal elevation = 3 m, and SLR =
0.5 m, the CVI, is generally dominated by very
high for both shoreline and coastal land. Sea
level rise of 0.5 m or 50 cm is predicted for the
year 2050 (Lewsey, 2002).

In Figure 7 is for scenario 3 with wave
height = 4 m, tidal elevation = 4 m, and SLR =
1 m, where the CVI score generally dominated
by very high both shoreline and coastal land.
There are only three CVI value in scenario 3,
namely moderate, high and very high. This
scenario simulated for the year 2100 where the
sea level rise increased to 1 m or 100 cm
(Lewsey, 2002).
Scenarios 2 and 3 is very useful and impor-
tant to analyze the impact of storm surge at-
tacks that have effects such as sea level rise
on coastal areas. Storm surge is the event of
sea level rise due to high winds and constant
storms accompanied by heavy rains and could
damage coastal areas.
From the simulation results of both the
scenario 2 and scenario 3, the topographical
condition in the coastal region along the coast
from Jawai Selatan until Jawai is very flat and
Figure 5. Scenario 1
72
Danial, e.t al.: Shoreline Change Model Using the Epr Method and the Simulation of Coastal Vulnerability in ... Kalimantan
low-lying land, so very vulnerable to the threat
of large waves and tidal fluctuations.

Verification
In Figure 8, can be seen that this area is the
former location of fishpond that have been
damaged.

Figure 9 shows that condition of shoreline
between beach in Jawai Selatan and Jawai had
experienced erosion. Many coconut trees that
had collapse and indication of coastal erosion
can be seen clearly.

There are two main factors cause erosion in
Jawai Selatan. First, land conversion of coastal
area to fishponds. Fishpond can cause a large
void that susceptible from wave attack.
Second, position of breakwater is too close to
fishponds area that can cause erosion quickly
due to the wave diffraction. Breakwater
position should be located more far from
shoreline (offshore breakwater) to reduce
diffraction effect. Offshore breakwater is more
adequate and more safely to protect fishpond
from erosion because it can create salient and
tombolo behind the structure.
From the results of studies in the field, the
beach will be safe from erosion if there is a
mangrove plant that can reduce wave energy
and increase the supply of beach sediment
transport.
Mangrove trees can be found starting from the
village along the coast Nyirih (Jawai district)
to the beach in the village of SB Nilam.
Figure 8. Field verification
Figure 9. Field verification
73
CONCLUSIONS

From the results of this research, there are
several conclusions that can be given as
follows:
1. From the results of the shoreline change
analysis obtained information that the
length of the beach erosion from Jawai
Selatan to SB Nilam is 3985.04 m (4 km)
and the eroded area is 9557.546 m
2
in
segment 1. In segment 2, the length of
erosion is about 4.3 km with 11310 m
2
of
land area.
2. Shoreline erosion is very strong with 5 to 6
m/year of average erosion rate for segment
1 and 10 m/year for segment 2. The rate of
sedimentation is 0.34 m/year and 11 m/year
for segment 1 and segment 2, respectively.
3. Sambas coastal areas in the sub district
Jawai Selatan until the sub district Jawai
susceptible to the (scenario 1). Sambas
coastal region was quite vulnerable to the
existing conditions, although CVI value is
dominated by the moderate vulnerability,
while the shoreline was in danger condition
with CVI value is high vulnerability.
4. In the second scenario with tidal data = 3
m, wave = 3 m, and sea level rise = 0.5 m,
obtained results of a vulnerability index is
generally dominated by the conditions of
very high vulnerability. On along the coast,
the criteria of vulnerability index changed
to "very high vulnerability ". Sea level rise
0.5 m is predicted for the year 2050.
5. In the third scenario with tidal data = 4 m,
wave = 4 m, and an increase in sea level = 1
m, obtained results of a vulnerability index
is generally dominated by the conditions of
very high vulnerability. Sea level rise of 1
m was predicted for the year 2100. This
scenario aims to detect coastal vulnerability
in the future.
6. The shoreline retreat in Jawai Selatan due
to breakwater too close to the shoreline,
can be overcome by removing the
breakwater into the sea with a certain
distance from the shoreline to grow salient
and tombolo. Furthermore, the combination
of Breakwater and mangrove plant, are
expected to increase the supply of sediment
to the eroded beach areas.

ACKNOWLEDGEMENTS

This research was carried out with the
financial support of DP2M DIKTI.

REFERENCES

Abuodha, P. A., Woodroffe, C. D. (2006).
International Assessments of the
Vulnerability of The Coastal Zone to
Climate Change, Including an Australian
Perspective. Research Report,
Department of the Environment and
Heritage, Australian Greenhouse Office.
Alesheikh, A. A., et al. (2007). Coastline
Change Detection Using Remote
Sensing. Int. J. Environ. Sci. Tech., 4 (1),
p. 61-66.
Cambers, G. (2001). Coastal Hazards and
Vulnerability. Sea Grant College
Programme, University of Puerto Rice,
Mayaguez, Puerto Rico.
Gornitz, V., White, T. W. and Cushman, R. M.
(1991). Vulnerability of the U.S. to
Future Sea-Level Rise. Coastal Zone '91,
Proceedings of Seventh Symposium on
Coastal and Ocean Management ASCE, p.
2354-2368.
Lewsey, C. (2002). Climate Change Impacts
on Land Use Planning and Coastal
Infrastructure. National Oceanic and
Atmospheric Administrations.
Li, R., Liu, J. K., and Felus, Y. (1999).
Spatial Modeling and Analysis for
Shoreline Change Detection And Coastal
Erosion Monitoring. The Proceedings of
Geoinformatics '99, Conference, Ann
Arbor, 19-21 June, 1999.
Szlafsztein, C. F. (2005). Climate change,
Sea-level rise and Coastal Natural
Hazards: A GIS-Based Vulnerability
Assessment, State of Par, Brazil. An
International Workshop, Department of
Geology, Center of Geosciences,
University of Par, Brazil.
74
Awaludin: Time-History Response of 2-D Timber Frame Reinforced with Wooden Panel
Time-History Response of 2-D Timber Frame Reinforced
with Wooden Panel

Ali Awaludin

Civil and Environmental Engineering Department, University of Gadjah Mada, Bulaksumur Yogyakarta, Indonesia,
email: ali.awaludin@ugm.ac.id

Abstract: Most engineered timber (frame) buildings under great earthquake would undergo large
deformation with their beam and column members are still elastic. This fact is potentially due to
large flexing ability and ductile behavior of their mechanical connections. Since inelastic
deformation occurs locally at their connections, non-linear analysis of the timber frame is often
referred to a localized-nonlinearity problem. This paper demonstrated a non-linear time-history
analysis of 2-D timber frame model using SAP2000 computer program. In this numerical analysis, a
collection of non-linear link elements was used to produce similar load-deformation relation of
semi-rigid joints of the frame model. The response of the frame model also was investigated when a
lateral force-resisting element such as wooden panel was installed.

Keywords: Non-linear analysis, seismic response, timber frame, and wooden panel.

INTRODUCTION

Indonesia sits on three active tectonic plates:
Pacific, Eurasia, and Indo-Australia that often
leads to a very catastrophic earthquake disaster
such as the December 2004 great Sumatra
earthquake (Saatcioglu, 2006). After this
earthquake, USGS recorded that about four to
five strong earthquakes (Richter scale of 6.0 or
more) shook Indonesia annually. As a result,
the provision of an adequate structural system
that enables a building to absorb and distribute
the earthquake energy without collapse or
catatastrophic failure is essential. Since it is
very costly to construct strong buildings
without damages during earthquakes, modern
Figure 1. Test set-up of timber joint under reversed-cyclic load
75
timber buildings would be an important
solution. The modern timber buildings are
characterized with the provision of ductile
connections, hold-down devices that firmly
connect the upper structure to foundation, and
lateral force-resisting elements such as
wooden panels or diagonal members. Through
a combination of these structural elements, the
timber buildings are enforced to perform
monolithically under any loadings including
the earthquake loads.
Wood assemblies offer high strength-to-weight
ratio over those built with steel and concrete.
This results in low inertial or lateral force
during earthquake. Wood systems are
inherently more flexible than other building
materials such as masonry and concrete that
have to be carefully designed and detailed to
ensure good seismic performance. Due to
anisotropic properties of wood and hundreds
of possible building geometry, modeling of
timber buildings even of a simple frame
construction always requires experimental
works. Many timber engineers and researchers
are currently developing various analytical
techniques to predict both elastic and inelastic
response of timber structures with minimum
laboratory works. And the required numerical
data could be derived from relatively simple
tests.
In this presented study, a seismic response of
2-D timber frame model under specified
ground excitation was numerically evaluated
using SAP2000 (1997). Moreover, the frame
model can also accommodate wooden panel
installment as one of important lateral force-
resisting elements. The static-cyclic properties
of timber joints and wooden panel were
obtained from full-scale tests.

TIMBER JOINT TEST

In general adhesive joints should not be aplied
in timber buildings to ensure ductile behavior
under earthquake loads. Mechanical timber
connections such as nails, lag screws, dowels,
and bolts are often used in modern timber
building because of their fast joint fabrication.
A timber joint model shown in Figure 1 was
selected for this study. SPF (spruce-pine-fir)
species was used with 4-mm-thickness steel
plates and 12-mm-diameter bolts. The joint
was tested cyclically until lateral displacement
measured at the hydraulic cylinder (see Figure
1) reached 70-mm. And after that, the joint
was subjected to monotonic loading until
faiure.
A typical hysteresis loop obtained from the
static-cyclic test as shown in Figure 2
indicated that joint resistance was decreasing
when the applied moment was repeated at the
same rotation level. Less resistance at rotation
close to zero points describes pinching
mechanism due to wood fiber crushing
beneath the bolts. At this stage, the joint
resistance is solely given by the bending
strength of the bolts. Hysteretic damping or
area enclosed by the loop of the joints
decreased as the number of cycles increased
and its reduction occurred with a greater rate
when the joints were cycled at higher rotation
levels. As the plastic embedment of wood
under the bolts progresses, the pinching region
or range of joint rotation with relatively low
stiffness increases.
Improvement of initial joint stiffness and
hysteretic damping of bolted timber joints
through initial bolt pretension had been
intensively reported Awaludin et al. (2008a,
2008b). For bolted joints with wood plates as
both main and side members, the improvement
can be carried out by gluing or embedding
reinforcement materials at the joint shear
planes. Among many available reinforcements,
natural fiberbased materials such as plywood,
plybamboo, and densified veneer wood
(DVW) are probably the most widely used
Leijten (2007), Awaludin et al. (2007).

WOODEN PANELS

Besides diagonal members or bracing, wooden
panels are often found in modern timber
buildings around the world. The panels usually
consist of plywood or oriented strand board
(OSB). Wooden panels or diagonal members
are expected to increase the shear rigidity of
the timber buildings during earthquakes. In
some countries with extreme weather, the
walls are often constructed in two layers with
76
gap in between. Polystyrene-based insulators
fill the gap, insulate acoustic and improve
thermal comfort of the building.
Test results of wooden panel under static-
cyclic load described here is summarized from
the work of Honma et al. (2007). The panel
consisted of 805x2595 mm
2
oriented strand
board with 9.5-mm in thickness as can be seen
in Figure 3(a). At each edge, the board was
connected to SPF lumber of 105x105mm
2

through nails CN-50 (2.8-mm-diameter and
50-mm-length). The nail spacing was about
100-mm. Before a monotonic loading until
failure, the panel was subjected to a reversed-
cyclic load to investigate its racking behavior.
The lateral load-rotation curve derived from
the test is presented in Figure 3(b). Panels with
an opening was also fabricated and tested to
examine strength degradation related to panel
opening. The opening area was about 46
percent. The simplified test setup and
experimental load-rotation curve of the panel
with opening are shown in Figure 4.
A combination of bending and shear forces
was introduced to the wooden panel by the
applied-load. The test showed that the failure
of the panel without opening was due to the
failure of the beam-column connection at the
top left-corner. The curve shown in Figure 3
clearly shows the wooden panel has small
ductility ratio as a typical characteristic of
mortise-tenon connection system. After
reaching the maximum resistance, the curve is
decreasing immediately and with minimum
apparent-yielding zone. When there are some
openings in the OSB element, the shear
rigidity of the OSB element is decreasing. Test
results of the panel with opening indicated that
the failure mode was a partial slide-off of the
OSB element from the column or beam
member. Significant bending deformation of
the nails and bearing damage of OSB or wood
fibers around the nail fasteners were observed.
Following this failure mode, in some tests, the
nail fasteners were totally moved-out from the
OSB element. The withdrawal resistance of
nail connections was exceeded. This failure
mode produces high ductility ratio as can be
clearly seen in Figure 4.
The initial stiffness and lateral resistance of
wooden panel without opening were
significantly greater than those of the panel
with opening. Under reversed-cyclic loads,
both panels experienced strength degradation
due to bending deformation of nails and
irreversible embedment of wood fibers around
the nail fasteners.
-16
-12
-8
-4
0
4
8
12
16
-120 -80 -40 0 40 80 120 160
Average rotation (0.001 rad)
M
o
m
e
n
t

(
k
N
m
)
2
Figure 2. Test set-up of timber joint under reversed-cyclic load
77
However, the hysteretic damping of the panel
without opening was much greater than that of
the panel with opening. The opening of the
OSB element definitely leads to a decrease of
shear resistance of the panel.

NUMERICAL MODEL OF TIMBER
FRAME

To elaborate the test results of timber joint and
wooden panel described previously, a 2-D
timber frame model shown in Figure 5 was
developed. Hinge support was adopted since,
in many practice, this type of support can
perfectly model a combination of shallow
foundation behavior and common timber
jointing method. At the top of each column, a
concentrated weight of 10 kN was assumed as
the frame load gene-rated by the gravity loads.
In this numerical analysis, N-S component of
Elcentro earthquake 1940 shown in Figure 6
was used as the ground excitation input.
The equilibrium equation of motion of the
system with nonlinear change in stiffness
under general time-history load p(t) is
expressed as shown in the following equation:
Figure 4. (a) Simplified test setup and (b) Hysteresis loop of panel with opening
(Honma et al. 2008)
Figure 3. (a) Simplified test setup and (b) Hysteresis loop of panel without opening
(Honma et al. 2008)
78

| | | | | |{ } { } | |{ } ) ( ) ( ) ( ) ( ) (
0
. ..
t v k t p t v k t v c t v m
d
= +
)
`
+
)
`
... (1)
{ } | |{ } ) ( ) ( t Y t v u = ........ (2)
| | | | | |{ } { } | | ) ( ) ( ) ( ) ( ) (
. ..
t F t P t Y K t Y C t Y M
sd
= +
)
`
+
)
`
..... (3)
{ } { } { } | |{ }
=
= + +
m
p
p p d
T
n
T
n n n
n
n n
n t Y k t p t Y t Y t Y
1
2
. ..
) ( ) ( ) ( ) ( 2 ) ( | | | e e .... (4)
{ } { } { } | |{ }
= + +
m
p
k
p p d
T
n
T
n
k
n n
k
n
n n
k
n
Y k p Y Y Y
1
) 1 ( ) ( 2 ) (
.
..
) (
2 | | | e e ... (5)

In the above equation, [m], [c], and [k
0
] is
mass, damping, and stiffness matrices of the
system, respectively. {v} is the displacement
vector of the system. In the equation, a single
overdot denotes the first derivative with
respect to time, which is velocity, and
obviously double overdots denote the second
derivative with respect to time, that is,
acceleration. [k
d
] is being determined from the
force-displacement relationship of a
connection. Uncouple Equation 1 by means of
the natural modes of the system defined by the
left-hand side of the equation gives and
{Y(t)} is modal matrix and generalized modal
coordinates, respectively. For a proportionally
damped system, Equation 3 is uncoupled and
the integration is actually carried out on the
equation. The last term in the Equation 4 is
modal pseudo-forces that are functions of all
modal coordinates of the system. and are
respectively, damping ratio and natural
frequency correspond to mode shape n, .
The equilibrium in one step for mode shape n
in the k cycle of iteration is shown in Eq. 5.
In each cycle of iteration, the equations are
integrated and the displacement and velocity at
the end of a time step are determined. The
analysis was carried out using SAP2000
educational version. Only the first mode shape
with damping ratio of 5% was considered for
this study.
The connection between column and beams
was modeled by a set of four NL-link elements
and one imaginary node as shown in Figure
7(a). Detail of semi-rigid joint modeling using
SAP2000 can be found in Mesic, 2003. The
properties of the four NL-link elements were
derived by some trials so that it can produce a
hysteresis loop similar to the one given in
Figure 5. Timber frame model of this study Figure 6. N-S component of Elcentro
earthquake 1940
| |{ u
n
e
n
e
{ }
n
|
79
Figure 2. An example of moment-rotation
curve under one cyclic load is shown in Figure
7(b), where the properties of the four NL-link
elements are summarized in Table 1. A
different shape of hysteretic loop can be
derived by supplying different numerical
values of the four NL-link elements. This
technique accommodates variation of
hysteretic loops as they are essentially
influenced by joint geometry including the
number of mechanical fasteners. Since the
objective of this paper is to demonstrate the
time history non-linear analysis of frame
without and with wooden panel, therefore the
properties of the four NL-link elements given
in Table 1 and Table 2 are selected after some
trials.

Table 1. Properties of the NL-link elements for semi-rigid joint used in SAP2000
Element Type
Opening
(rad)
k
o

(kNm/rad)
Yield Moment
(kNm)
a exp
NL1 Plastic 1 NA 700 2.0 0.01 1.0
NL2 Plastic 1 NA 400 4.0 0.05 1.0
NL3 Gap 0.01 25 NA NA NA
NL4 Hook 0.01 25 NA NA NA
k
0
: elastic (rotational) stiffness
a : ratio of the post-yield stiffness to k
0
exp : indicates the sharpness ratio of yielding curve
NA : not available

Table 2 Properties of the NL-link elements for wooden panel used in SAP2000
Element Type
Shear stiffness, k
o

(kN/m)
Shear resistance
(kN)
a exp
Panel 1 Plastic 1 230 12 0.25 10
Panel 2 Plastic 1 80 7 0.25 10
k
0
: elastic (shear) stiffness
a : ratio of the post-yield stiffness to k
0

exp : indicates the sharpness ratio of yielding curve
Figure 7. (a) Semi-rigid joint modeling and (b) hysteresis loop of the joint due to one cyclic load
80
Time-history lateral displacement of the frame
without and with wooden panel subjected to
the Elcentro earthquake 1940 is presented in
Figure 8 and Figure 9, respectively. The
installment of wooden panel was modeled with
one NL-link element at the middle of the beam
element. The properties of this NL-link
element (panel 1 or panel 2), which are shown
in Table 2, were derived based on the
hysteresis loops shown in Figure 3 and Figure
4, respectively, for the panel without and with
46% opening. Figure 9 indicates that the
presence of wooden panel significantly
decreases the lateral displacement of the
frame; the lateral displacement had been
decreased by 32 percent for the panel without
opening and by 20 percent for the panel with
46% opening. Decrease of lateral displacement
due to wooden panel installment is a clear sign
of overall improvement of the seismic
performance of the frame model. This
developed analysis can also be used to further
evaluate the effect of different joint or wooden
panel hysteresis loops on seismic performance
of the timber frame.

Figure 8. Model of timber frame without wooden panel and its time-history displacement
response due to N-S component of Elcentro earthquake 1940
Figure 9. Time-history lateral displacement of the frame model with wooden panel
81
CONCLUSIONS

The seismic performance of timber buildings
is substantially governed by their connections
behavior. Therefore, performing a full-scale
test of timber frame might not be so essential
when the cyclic properties of their connections
are available. In this study, seismic response of
2-D timber frame model was analyzed
numerically using SAP2000 computer
program assuming that nonlinearity problems
occurred only at the connections. Four non-
linear link elements and one imaginary node
were necessary to appropriately simulate load-
deformation relation of semi-rigid joints of the
frame model. The numerical results showed
that a significant decrease of lateral
displacement of the frame can be attained by
installing wooden (shear) walls.

REFERENCES

Awaludin, A., Hayashikawa, T., Hirai, T., and
Sasaki, Y. (2007). "Splitting Strength of
Plybamboo-Reinforced Timber Joints
Under Loading Perpendicular-to-Grain".
Proceedings of the International
Conference on Sustainable Infrastructure
and Built Environment, Bandung, West
Java, November 2-3. (CD ROM)
Awaludin, A., Hirai, T., Hayashikawa, T.,
Sasaki, Y., and Oikawa, A. (2008a).
"Effects of Pretension in Bolts on
Hysteretic Response of Moment Carrying
Timber Joints". Journal of Wood Science,
Vol. 54, No. 2, p. 114-120.
Awaludin, A., Hirai, T., Hayashikawa, T., and
Sasaki, Y. (2008b). "Load-Carrying
Capacity of Steel-to-Timber Joints with a
Pretensioned Bolt". Journal of Wood
Science, Vol. 54, No. 5, p. 362-368.
Honma, C., Teranishi, M., Sasaki, Y., and
Hirai, T. (2007). "Earthquake-proof
Reinforcement of Deteriorated Framing of
Existing Wooden Houses with Shear-
Resistant Components". Mokuzai
Gakkaishi, Japan Wood Research Society,
Vol. 54, No. 3, p. 123-131 (in Japanese).
Leijten, A. J. M. (2007). "Densified Veneer
Wood Reinforced Timber Joints with
Expanded Tube Fasteners". Delft
University Press, The Netherlands.
Mesic, E. (2003). "Analysis of Timber Frames
with Localized Nonlinearities". Facta
Universitatis, Series: Architecture and
Civil Engineering, Vol. 2, No. 5, p. 307-
320.
Saatcioglu, M., Ghobarah, A., and Nistor, I.
(2006). "Performance of Structures In
Indonesia during the December 2004
Great Sumatra Earthquake and Indian
Ocean Tsunami". Journal of Earthquake
Spectra, Vol. 22, No. S3, S295-S319.
SAP2000 (1997). "Manual, Analysis
Reference", Computers and Structures,
Inc., Berkeley, California, USA, Vol. 1.
82
Rosyidi: Wavelet-Spectrogram Analysis Of Surface Wave Technique: A Novel Procedure ... Layer
Wavelet-Spectrogram Analysis of Surface Wave Technique: A Novel Procedure for
Non-Destructive Measurement on Pavement Surface Layer

Sri Atmaja P. Rosyidi
Department of Civil Engineering, University of Muhammadiyah Yogyakarta, KampusTerpadu, Jl. Ring Road Selatan,
Bantul 55183, Yogyakarta, email: atmaja_sri@umy.ac.id

Abstract: Reliable assessment of in situ pavements stiffness is an important aspect in effectively
managing a pavement system. The aim of this paper is to propose the new procedure, namely the
wavelet-spectrogram of surface wave (WSSW) technique for non-destructively measurement of
elastic modulus on surface layer of a pavement system. Using two receivers, surface wave
waveform on pavement surface was recorded and transformed into in frequency domain by wavelet
analysis. For this analysis, a derivative Gaussian wavelet was selected as an appropriate mother
wavelet for seismic waveform propagating along pavement surface. Thus, an interactive 2-D plot of
time-frequency spectrogram consisting of wave-energy spectrum was simultaneously generated.
CWT-filtration method was implemented in order to reduce the effect of noisy signal recorded
during measurement. From selected wave spectrogram, the unwrapped phase different-spectrum
was generated to obtain phase velocity which was performed by least-square linear regression.
Finally, the elastic modulus of pavement surface layer was calculated from a modified relationship
between phase velocity, Poisson ratio and density of pavement surface layer. The results show that
the proposed technique is able to measure in situ elastic stiffness of the surface layer. In addition,
the change of the surface layer stiffness is also able to be monitored. However, the stiffness
produced by the WSSW technique is classified as a modulus at very low strain level.

Keywords: Elastic modulus, surface layer, pavement system, and wavelet-spectrum of surface
wave technique.

INTRODUCTION

Seismic surface wave methods are well-known
of non-destructive techniques in soil dynamics
and geotechnical investigation. Over past
decade, seismic surface wave methods have
been improved and subsequently, it has been
utilized in different applications, i.e., spectral
analysis of surface wave, SASW [site
characterization (Stokoe et al. 1994), soil
density (Kim et al. 2001), ballast density
(Zagyapan & Fairfield, 2002), sloping surface
(Kim et al. 2007), pavement characterization
(Rosyidi et al. 2007)], multichannel analysis of
surface wave, MASW [residual soil
characterization (Ariestyanti, 2009), shear
wave velocity profiling (Park et al. 1998),
near-surface anomaly detection (Miller & Xia,
1999)], continuous surface waves, CSW [soil
profiling (Rosyidi, 2006)].
The surface wave methods use the dispersion
property of surface waves (dominantly by R-
wave) to determine material stiffness of each
layer at soil profile. In order to generate the
elastic/shear modulus, an elaborated data
process should be performed. The process is
divided into three steps: collecting seismic
data, constructing an experimental dispersion
curve, and conducting an inversion process of
the dispersion curve for generating a proper
profile. An impact source on a ground surface
is used to generate seismic waves at a point.
These waves are detected at two or more
receiver locations and the signals recorded
using a spectrum analyzer for post processing.
Several sets of test with different receiver
spacings are required to sample different
depths. The experimental dispersion curve can
be developed from phase information of the
transfer function at frequency ranges satisfying
the coherence criterion.
83
In inversion process, the advanced algorithm is
required to produce the stiffness profile froma
dispersion curve. Reliable inversion procedure
using stress-wave propagation theories, i.e.,
the transfer matrix method (Thomson 1950,
Hanskell 1953); the dynamic stiffness matrix
method (Kausel & Resset, 1980); and the
finite difference method (Hossain & Drnevich,
1989) should be employed. All the methods
require an initial profile model consists of a set
of homogeneous layers extending to infinity in
the horizontal direction. The last layer is
usually considered as a homogeneous half-
space. To each layer, a thickness, a shear wave
velocity, a Poisson's ratio (or compression
wave velocity), and a mass density are
assigned. Based on the initial profile, a
theoretical dispersion curve is calculated using
one of these wave propagation theories. The
theoretical dispersion curve is then compared
with the experimental dispersion curve. If the
two dispersion curves do not match, the initial
profile (number of layers, layer thickness,
shear wave velocity, or any combination) is
adjusted, and another theoretical dispersion
curve is calculated. The trial-and-error
procedure is repeated until the two curves
match, and then the associated assumed profile
is considered the real profile. In addition,
when the surface wave method is performed
on irregular stiffness profile, i.e. pavement
structure, the trial-and-error process becomes a
difficult analysis.
Due to complexity in surface wave analysis,
application of these methods on pavement
evaluation is still relatively limited. On the
other hand, in pavement evaluation system, the
need of accurate, cost-effective and non-
destructive evaluation is becoming ever
important because the rehabilitation and
management of roads is becoming increasingly
difficult due to the increasing number of aging
roads and limited budgets. As well, for
practical purpose, pavement engineer usually
needs a quick and relatively effortless analysis
for determining the structural condition of
pavement surface layers.
In this paper, a new procedure in a surface
wave methodusing continuous wavelet
transform and simple algorithm, namely the
wavelet-spectrogram of surface wave (WSSW)
technique, for obtaining the phase velocity is
introduced. Modified algorithmon phase
different calculation in this analysis aims to
avoid the use of a complex inversion algorithm
to obtain the elastic modulus of the pavement
surface layer. In proposed analysis, the time-
frequency decomposition of CWT on seismic
signals is employed to characterise the phase
information of transfer function spectrum.
In previous surface wave method, the data
analysis in frequency domain has been carried
out by fast Fourier transforms (FFT).
However, due to Fourier transform works by
expressing any arbitrary periodic function of
time with period as sum a set of sinusoidal,
this analysis is unable to preserve the time
dependence and describe the evolutionary
spectral characteristics of non-stationary
processes (Rosyidi et al. 2009). Wavelet
analysis is becoming a common tool for
analyzing localized variations of power within
a time series. By decomposing a time series
into time-frequency spectrum (TFW), one is
able to determine both the dominant modes of
variability and how those modes vary in time.
A typical result from a case study is presented
herein for the structural assessment of an
existing asphalt concrete (AC) pavement at
Cikampek-Purwakarta road, West Java,
Indonesia.

THEORY OF CONTINUOUS WAVELET
TRANSFORM (CWT)

The continuous wavelet transform has been
used in numerous studies in geophysics,
including the El NioSouthern Oscillation
(Wang and Wang, 1996), the dispersion of
ocean waves (Meyers et al., 1993), and wave
growth and breaking (Liu 1994). A complete
description of geophysical applications can be
found in Foufoula-Georgiou and Kumar
(1995) and Rosyidi et al. (2008, 2009), while a
theoretical treatment of wavelet analysis is
given in Daubechies (1992).
The continuous wavelet transform (CWT)
technique is an alternative tool for localizing
the interested frequency of seismic signal
processing particularly in non-stationary
84
Sukarno: Teori Kendaraan-Pengikut yang Disederhanakan (Studi Kasus Persimpangan Bersinyal, ... Yogyakarta)
problems. A presentation of basic wavelet
theory may be found in several literatures such
as Daubechies (1992), Kaiser (1994), and
Farge (1992).
Wavelets dilate in such a way that the time
support changes for different frequency.
When the time supports increases or decreases,
the frequency support of the wavelet is shifted
toward high or low frequencies, respectively.
Therefore, as the frequency resolution
increases, the time resolution decreases and
vice versa (Mallat, 1989). This optimal time-
frequency resolution property makes the CWT
technique useful for non-stationary seismic
analysis.
A wavelet is defined as a function of (t)
eL
2
(9) with a zero mean, which is localized
in both time and frequency. By dilating and
translating the wavelet (t), it can be used to
produce a family of wavelets as:
( ) |
.
|
\
|
=
o
t
o
t
t o
t 1
,
(1)
Where o is the dilation parameter or scale and
t is the translation parameter (o, te9, and
o=0).
Unlike Fourier transform, the wavelet has
various wavelet shape used for signal analysis
which is called the mother of wavelet. The
choice of appropriate wavelet shape used in
signal analysis depends on the seismic
waveforms. The family of wavelets can be
divided into two groups (Soman and
Ramachandran 2005) i.e. continuous wavelet
transform (CWT) and discrete wavelets
transform (DWT). Some mother wavelets
which are commonly used in the CWT are
Gaussian, Morlet, Paul and Mexican Hat. The
CWT is defined as the inner product of the
family wavelets
o,t
(t) with the signal of f(t)
which is given as:
( ) ( ) ( ) ( ) dt
t
t f t t f F
W
|
.
|
\
|
= =
}

o
t
o
t o
t o
1
, ,
,
(2)
where

is the complex conjugate of ,
F
W
(o,t) is the time-scale map. The
convolution integral from Equation (2) can be
computed in the Fourier domain. In order to
reconstruct the function f(t) from the wavelet
transform, Calderons identify (Daubechies
1992) can be used and is obtained as:
( ) ( )
o
t
o
o
o
t
t o
d d t
F
C
t f
W
2
,
1
|
.
|
\
|
=
} }

(3)
( )
< =
}
e
e
e
t
d C
2
2 (4)
where ( ) e

is the Fourier transform of (t).
The integrand in Equation (4) has an
integrable discontinuity at e = 0 and implies
that ( ) 0 =
}
dt t .
In this study, the mother wavelet of the
Gaussian Derivative (GoD) was used. The real
component of the GoD wavelet in the time and
frequency domains is defined as follows:
( )
( )
( )
2
2
2
1
1
1
0
t
e
d
d
m
t
m
m m
|
.
|
\
|
+ I
=
+
q
(5)
( ) ( )
( )
|
.
|
\
|
|
.
|
\
|
+ I
=

2
2
2
1
0
e
e e
s
e s
m
i
s
m
m
(6)
Where m is the wave number and I is the
Gamma function. The complex wavelet is
generated by the addition of a Heaviside
function in the frequency domain. This
wavelet decays with the square root of the
gamma function. The Gaussian Derivative has
wavelet's derivative order that can be varied in
order to get the best resolution of the
waveform.

A NOVEL PROCEDURE FOR
CALCULATING ELASTIC MODULUS
OF PAVEMENT SURFACE LAYER

A proposed procedure of surface wave
analysis using continuous wavelet transform
and elastic linear model for elastic modulus
measurement on pavement surface layer is
described as follows:
1. The surface wave data are collected from
field measurement using a configuration of
the mid point receiver spacings. In this
configuration, some instruments, i.e., an
impact source of small ball bearing, a data
acquisition unit and two acceleration
85
receivers (accelerometers) are used for
recording seismic wave propagation.
Discussion of field set up is presented in the
following section.
2. The time-frequency spectrum analysis
based on continuous wavelet transform of
Gaussian Derivative (DoG) mother wavelet
is carried out for two signal waveforms
recorded from field measurement.By
decomposing a time series of seismic
waveform into time-frequency (TF)
spectrogram, the dominant modes of
variability and how those modes vary in
time can be determined very well. In this
study, the TF spectrogram is generated by
wavelet transform (Equation 2) with a
mother wavelet of GoD (Equation 6).
3. Phase information of the transfer function
(phase spectrum) are then determined from
both TF spectrograms. The data informs the
time difference of wave propragating from
first to second receiver. A mathematical
expression for calculating phase spectrum
from TF spectrogram is defined by
(Rosyidi, 2009):
( )
( )
( )
( )
( ) dt
t
t X
dt
t
t Y
W
W
f X
f Y
f H
s u
X
f
s u
Y
f
|
.
|
\
|
|
.
|
\
|
= ~ =
}
}

o
t
o
o
t
o
*
1
*
1
) , (
) , (

................................................................. (7)
where,
X(f) = spectrum input of signal, X(t),
from first receiver,
Y(f) = spectrum output of signal, Y(t),
from second receiver,
( )
( ) u t i
s u
Y
f
e
s
u t
g
s
t Y W

|
.
|
\
|
=
}

1
) , (

(8)
( )
( ) u t i
s u
X
f
e
s
u t
g
s
t X W

|
.
|
\
|
=
}

1
) , (

(9)
From Eq.7, the phase spectrogram in time-
fequency domain can be obtained by:
( )
( )
( )
( ) ( ) ( )
) , ( ) , (
) , (
,
,
,
*
, ,
s u W s u W
e s u W
s u W
s u W
s u H
X
f
X
f
b a b a i XY
f
XX
f
XY
f
X Y
= =
u u

............................................................... (10)
Thus, phase different is obtained from the
ratio of the imaginary to real part of the
phase spectrogram which is expressed as:
|
|
.
|
\
|
9
=

) , (
) , (
tan
1
s u H
s u H
|
(11)
4. The coherence function is used to visually
inspect the quality of signals being recorded
in the field and have a real value between
zero and one in the range of frequencies
being measured. The value of one indicates
a high signal-to-noise ratio (i.e. perfect
correlation between the two signals) while
values of zero represents no correlation
between the two signals. The coherence
function is a ratio of the output power
caused by the measured input to the total
measured output which is defined as
(Rosyidi 2004):
( )
( ) ( )
( ) ( ) f G f G
f G f G
f
yy xx
yx yx
=
*
2
(12)
5. A linear relationship between the phase
different and frequency from the transfer
function spectrum is then derived. The
phase velocity is examined as a function of
distancefrom the slope value (m). This
relationship can be written as:
| =
ph
V
D 360
f = mf (13)
By Equation 13, one can easily determine
V
ph
by performing a least-square linear
regression over the frequency range in the
transfer function spectrum and the slope of
the best-fit line (m) can be obtained.The
phase velocity of surface wave propagation
is independent of the wavelength for up to a
wavelength approximately equal to
thickness of the uppermost layer (Nazarian
et al., 1996). The range of wavelength to be
used can be estimated from the phase
velocities (V
ph
) of the material anticipated at
the site:
f
V
ph
=

(14)
where f is the frequency.
86
6. If one simply generates high frequency
waves and assumes that properties of the
uppermost layer are uniform, the dynamic
elastic moduli of the pavement materials
can easily be determined as follows:
E = 2 ( ) 1
2
+
S
V
g
=
2
ph
KV
g
(15)
K = (1.13 0.16)
( )
) (

2 1
1 2
(16)
where E is the dynamic elastic modulus,
respectively, V
S
is the shear wave velocity,
V
ph
is the phase wave velocity, g is the
gravitational acceleration, is the total unit
weight of the material and is the
Poissons ratio.

EXPERIMENTAL SET UP

Impact source of 5 to 15 g in weight of ball
bearings on a pavement surface was used to
generate surface waves. These generated
waves were detected using two high frequency
accelerometers where the signals were
recorded and transferred using an analog
digital acquisition to a notebook computer for
post processing (Figure 1). Several
configurations of the receiver and the source
spacings were required in order to sample
different depths. The configuration used in
this study was the mid point receiver spacings
(Heisey et al. 1982) as described in Figure 2.
However, the receiver spacing (d
2
) is
configurated as a distance which is less than
and/or equal to the thickness (H) of pavement
surface layer. The distance between a source
and first receiver (d
1
) is set as equal to
receivers spacing (d
2
). Due to the interest
pavement layer in this study is asphaltic
surface layer, the short receiver spacings of 4
and 8 cm with a high frequency source (ball
bearing) were only used. For each receiver
spacing measurement, the testing should be
repeated at lease 3 to 5 times in order to
minimise the effect of internal phase shift
between receivers and a good average of
received signals can be generated.
The surface wave data from SASW testing
was carried out at selected sites on Cikampek-
Purwakarta road in West Java, Indonesia. The
pavement profile obtained from core drilling
consists of an asphalt concrete (AC) layer (183
mm), crushed stone of base course (71 mm),
sub-base course (190 mm) over a subgrade
layer.

Figure 1. Field equipment for surface wave
recording: (a) spectrum analyzer with an
acquisition unit; (b) accelerometers for
receiving the seismic wave displacement
(Widodo & Rosyidi, 2009)

Figure 2. WSSW experimental configuration
87
RESULTS AND DISCUSSIONS

Elastic Modulus from the WSSW
Technique
The WSSW tests were performed on existing
asphalt pavements. A typical received signals
from field measurement by spectrum analyzer
instruments (Figure 1) at selected location in
Cikampek-Purwakarta roads was shown in
Figure 3. The waves were generated from
impact source arriving in the first
accelerometer (receiver 1) and its propagations
were then received in the second one (receiver
2). The body waves (Primary and Secondary
wave) and surface wave (Rayleigh wave) are
clearly shown in both signal recordings.
Therefore, the Gaussian Derivative (GoD)
continuous wavelet transform (CWT) as
mentioned in Eq. 7 was employed on both
signals for generating a time-frequency (TF)
plot. These CWT plots may overcome on
identification problem of the spectral
characteristics of non-stationary signals
measured in two receivers. The typical CWT
spectrogram for received signals with an
improved time-frequency resolution is shown
in Figure 4. The TF of GoD-CWT provides
good resolutions at high frequencies of signals.
It is also effective in the detection of frequency
bandwidth of wave groups using various
derivation order of this mother wavelet (GoD).

Figure 3. Signal recording from measurement
on a pavement structure
From Figure 4, several energy events at
different frequency bands are clearly detected
which may result in interference at
fundamental and higher mode of seismic
signals at the interface between surface layer
and base course; and modes from reflected
body waves. Based on both TF spectrograms
(Figure 4), the phase different between two
signals for every wave frequency were then
calculated (Eq.10 to 11). As a result, a transfer
function spectrum on wrapped phase different
curve was obtained as given in Figure 5. The
phase data oscillates between -t and t radian
(-180 and 180 degrees). This is the standard
illustration of spectrum presenting phase data
because the detail variation in data can be
observed in a small space.

(a) CWT Spectrogram from receiver 1

(b) CWT Spectrogram from receiver 2
Figure 4. Time-frequency plot of received
signals from measurement (Fig.3)

Figure 5 also shows that the phase different
curve reveals a smooth trend of variation in
phase with frequency up to a frequency of 25
kHz. It indicates the high-frequency surface
88
waves were detected representing the high
stiffness of the asphalt concrete surface layer
on a pavement structure. The quality of phase
data is also controlled by the coherence
function. As shown in Figure 6, the phase data
up to frequency of 20 kHz have the value of
coherence magnitude above 0.98.

Figure 5. Comparison between raw data and
best-fit curve ofwrapped transfer function
spectrum from measurement

Figure 6. Coherent function spectrum of
received signals from measurement

In order to obtain the modulus value of the
surface layer, the smoothed fitting process of
weighing function was used. The fitted curve
between the raw and the smoothed phase
spectrum is shown in Figure 5. Thus, the
phase data is unwrapped by adding the number
cycles to each phase. The unwrapped of the
raw phase spectrum is also shown in Figure 7.
The unwrapped phase spectrum is smoothed
by the linear regression as the best fit curve to
the raw data. The slope of the line is more or
less constant with frequency. A line is fitted to
the curve in the range of the frequency
corresponding to wavelengths shorter than the
thickness of the surface layer. Slope value of
the line can be used to determine the elastic
modulus of the surface layer of the pavement
profile using Equation 13, 15 to 16.
From Figure 7, the slope (m) of best-fit curve
is found to be 0.0140. Consequently, the phase
velocity can becalculated using Eq.13 which is
found to be 1028.57 m/s. Based on the phase
velocity, field configuration data and material
parameters, such as of receiver spacing (d
2
) of
4 cm, Poissons ratio of asphaltic layer of 0.25
and unit weight for pavement material (AC) of
2,200 kg/m
3
, the elastic modulus of asphaltic
(AC) surface layer is obtained as
845,662,040.80 kg/m
2
or 8456.62 MPa.

Figure 7. Comparison between raw data and
best-fit curve ofunwrapped transfer function
spectrumfrom measurement and slope analysis
to obtain m value

The result shows that the elastic modulus of
AC layer can be easily determined using the
WSSW technique. However, the value of
elastic modulus presented is relatively high. It
is due to the seismic technique measures the
dynamic stiffness at very low strain level (less
than 10
3
%). In this level, the material modulus
behaviour can be assumed as a constant and
have only a maximum value (Nazarian &
Stokoe, 1986).
In order to illustrate the usefulness and
sensitivity of this approach in testing the
surface layer, changes in the stiffness of the
existing surface and the overlay layer of the
pavement were measured in situ. Figure 8
-180
-120
-60
0
60
120
180
0 5000 10000 15000 20000 25000
Frequency, Hz
P
h
a
s
e

D
i
f
f
e
r
e
n
t

(
w
r
a
p
p
e
d
)
Raw data
Best-fit curve
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 5000 10000 15000 20000
Frequency, Hz
C
o
h
e
r
e
n
c
e
| = mf = 0.014f
R = 0.960
0
50
100
150
200
250
300
350
400
0 5000 10000 15000 20000 25000
Frequency, Hz
P
h
a
s
e

D
i
f
f
e
r
e
n
t

(
u
n
w
r
a
p
p
e
d
)
Raw data of unwrapped phase
spectrum
Best-fit curve (linear regression)
89
shows that the different stiffness of an existing
and overlay surface layer from the measured
profile can be investigated well. Based on
these results, it can be summarized that the
stiffness changes in the surface layer were
easily, non-destructively, and fast measured by
the WSSW technique.

Figure 8. Comparison between elastic modulus
of an overlay and existing surface layer at
Cikampek-Purwakarta Roads, Indonesia

Validation with the SASW Method
In order to validate the results from the WSSW
test, the spectral-analysis-of-surface-wave
(SASW) analysis was conducted at same
locations of a road pavement. In this method, a
set of transient impact sources was used to
generate surface wave energy that propagates
horizontally near the surface layer of the
pavement. The phase differences of signal data
was obtained from the cross-power spectrum.
Thus, the phase information was then
unwrapped to produce the dispersion curve of
the pahse velocity versus wavelength. An
inversion process was then iteratively
employed to confirm the experimental
dispersion curve from the theoretical model
established. A 3-D stiffness matrix model
(Rosyidi, 2007) was employed in the SASW
inversion analysis. Final profil was obtained
after 16 times of iteration with the root-mean-
square error (rms) of 35.47 m/s or average
deviation of about 5.92%.

Figure 9. The shear wave velocity profile from
inversion of the experimental dispersion curve

Figure 10. The dynamic elastic modulus of the
pavement profile from the SASW method and
its comparison with the WSSW method.

The equivalent shear wave profile from the
result of the inversion is shown in Figure 9 and
using the dynamic material equation, its
equivalent dynamic elastic modulus profile is
given in Figure 10. A modulus profile as
shown in Figure 10 is only given to 10 cm
depth of pavement structure due to the
asphaltic layer was indicated at this level. A
good agreement of elastic modulus resulted
from the WSSW and the SASW method is also
presented in Figure 10. The result shows that
-1,39E-1
0,01
0,02
0,03
0,04
0,05
4000 6000 8000 10000
D
e
p
t
h
,

m
Elastic Modulus, MPa
New pavement
Existing
pavement
Pave
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
0 200 400 600 800 1000 1200 1400
Shear Wave Velocity, m/s
D
e
p
t
h
,

m
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0 2000 4000 6000 8000 10000
Elastic Modulus, MPa
D
e
p
t
h
,

m
the SASW method the WSSW method
From 4 cm receiver spacing
From 8 cm receiver spacing
90
the difference between both methods is
calculated at 0.01 % and 3.14 % for first and
second layer of pavement surface layer,
respectively.

CONCLUSIONS

This paper summarizes a new procedure in the
wavelet-spectrogram analysis of surface wave
(WSSW) technique in order to in situ
evaluation the elastic modulus of asphalt
concrete surface layer on an existing pavement
profile. By the simple calculation on phase
spectrum from the surface wave data, the
elastic modulus of the surface layer can be
obtained without the complex calculation of
the inversion process. The calculation is easy
and can be simply implemented. This
technique is also a very sensitive non
destructive testing to monitor the change of the
modulus of the existing surface and overlay
layers.

ACKNOWLEDGMENT

This work is part of research project funded by
Directorate of Higher Education, Ministry of
National Education, Indonesia and the
University of Muhammadiyah Yogyakarta
(UMY) through hibah bersaing grant 2009.
Their supports are gratefully acknowledged.
We would also like to thank Dr. Siegfried
(PUSLITBANGJATAN, Bandung) and
research assistants at University of
Muhammadiyah Yogyakarta for their
assistance during field works.

REFERENCES

Ariestianty, S. K., Rosyidi, S. A., Chik, Z.,
Taha, M. R., Nayan, K. A. M., and
Syamsudin, A. R. (2008). Investigasi
Profil Kekakuan Tanah 2-D
Menggunakan Metode MASW. Prosiding
Pertemuan Tahunan ke 12 (PIT-XII)
Himpunan Ahli Teknik Tanah Indonesia,
Bandung, p. (VI) 1-5.
Daubechies, I. (1992). Ten Lecturers on
Wavelets. Society of Industrial and
Applied Mathematics.
Farge, M. (1992). Wavelet Transform and
Their Application to Turbulence. Annu.
Rev. Fluid Mech., 24, p. 395-457.
Foufoula-Georgiou, E. and Kumar, P. (1995).
Wavelets in Geophysics. Academic
Press.
Haskell, N. A. (1953). The Dispersion of
Surface Waves in Multilayered Media.
Bull. Seismol. Soc. Am., Vol. 43, No. 1,
p. 17-34.
Heisey, J. S., Stokoe II, K. H., and Meyer, A.
H. (1982). Moduli of Pavement System
from Spectral Analysis of Surface
Waves. Transportation Research Record
852, p. 22-31.
Hossain, M. M., Drnevich, V. P., Bush III, A.
J., and Baladi, G. Y. (1989). Numerical
and Optimisation Techniques Applied to
Surface Waves for Back-Calculation of
Layer Moduli in Nondestructive Testing
of Pavements And Back-calculation of
Moduli. Eds, Special Technical
Publication 1026, American Society for
Testing dan Materials, p. 649-669.
Kaiser, G. (1994). A Friendly Guide to
Wavelets. Birkhuser.
Kausel, E. and Resset, J. M. (1981).
Stiffness Matrices for Layered Soils,
Bull. Seismol. Soc. Am., 72, p. 1743-
1761.
Kim, D. S., Shin, M. K., and Park H. C.
(2001). Evaluation of Density in Layer
Compaction Using SASW Method. Soil
Dynamic and Earthquake Engineering,
Vol. 21, p. 39-46.
Kim, S. W., Kim, D. S., and Bang, E. S.
(2007). Evaluation of the Applicability of
SASW Method to the Sloping Surface-
Numerical Study and Model Test.
Proceeding of the twentieth KKCNN
Symposium on Civil Engineering, p. 378-
381.
Liu, P. C. (1994). Wavelet Spectrum Analysis
And Ocean Wind Waves. Wavelets in
Geophysics, Foufoula-Georgiou, E. and
Kumar P., Eds., Academic Press, p. 151-
166.
Meyers, S. D., Kelly, B. G., and OBrien, J. J.
(1993). An Introduction to Wavelet
Analysis in Oceanography and
91
Meteorology: with Application to the
Dispersion of Yanai Waves. Mon. Wea.
Rev. 121, p. 2858-2866.
Miller, R. D. and Xia, J. (1999). Feasibility of
Seismic Techniques to Delineate
Dissolution Feature in the Upper 600 ft at
Alabama Electric Cooperatives Proposed
Damascus Site: Kansas Geological
Survey. Open-file Report, Kansas
Geological Survey, No. 99-3,
Nazarian, S. and Stokoe II, K. H. (1986). In
Situ Determination of Elastic Moduli of
Pavement Systems by Spectral-Analysis-
Of-Surface-Wave Method (Theoretical
Aspects). Research Repot 437-2, Bureau
of Engineering Research. Center of
Transportation Research, The University
of Texas at Austin.
Nazarian, S., Yuan, D., and Baker, M. R.
(1996). Quality Control of Portland
Cement Concrete Slabs with Wave
Propagation Techniques. Transportation
Research Record 1544, p. 91-98.
Park, C. B., Miller, R. D. and Xia, J. (1998).
Imaging Dispersion Curve of Surface
Waves on Multi-Channel Record:
Technical Program with Biographies.
Proceeding of SEG 68th Annual Meeting,
New Orleans, Lousiana, p. 1380-1388.
Rosyidi, S. A., Taha, M. R., Chik, Z.,
and Ismail, A. (2008). Seismic Response
Identification of Surface Waves on
Residual Soils Using Wavelets Analysis.
Proceeding of International Conference
on Geotechnical and Highway
Engineering, Geotropika, Kuala Lumpur.
Rosyidi, S. A., Taha, M. R., Ismail, A.,
and Chik, Z. (2009). Signal
Reconstruction of Surface Waves on
SASW Measurement Using Gaussian
Derivative Wavelet Transform. Acta
Geophysica, Vol. 57, No. 3, p. 616-635,
DOI: 10.2478/s11600-009-0015-8.
Rosyidi, S. A. (2004). Evaluation of The
Dynamic Stiffness of Pavement Layers
Using SASW Method. Master Thesis,
Universiti Kebangsaan Malaysia.
Rosyidi, S. A. (2006). Penggunaan Metode
Continuous Source for Surface Wave
Analysis (CSSWA) dalam Kajian Struktur
Tanah. Proseding Konferensi Regional
Teknik Jalan ke-9, Makassar.
Rosyidi, S. A. (2007). Comparison between
2-D And 3-D Stiffness Matrix Model
Simulation of Sasw Inversion for
Pavement Structure. Civil Engineering
Dimension, Journal of Civil Engineering
Science and Application, Vol. 9, No. 1, p.
42-48.
Rosyidi, S. A. (2009). Wavelet Analysis of
Surface Wave for Evaluation of Soil
Dynamic Properties. Ph.D. Thesis,
Universiti Kebangsaan Malaysia.
Rosyidi, S. A., Taha, M. R., Nayan, K. A. M,
Chik, Z., Ismail, A., and Siegfried. (2007).
Development of VS-CBR-DCP
Empirical Model for Determining
Dynamic Stiffness of Pavement Base
Layer Using SASW. Proceedings of
International Conference on Advanced
Characterisation of Pavement and Soil
Engineering Materials, p. 895-902.
Soman, K. P. and Ramachandran, K. I. (2005).
Insight into Wavelets from Theory to
Practice. Prentice-Hall of India.
Stokoe II, K. H., Wright, S. G., Bay, J. A., and
Roesset, J. M. (1994). Characterization
of Geotechnical Sites by SASW Method.
Geotechnical Characterization of Sites.
R.D. Wood, Ed., Oxford and IBH Pub.
Co., New Delhi, India, p. 15-26.
Thomson, W. T. (1950). Transmission of
Elastic Waves through a Stratified Solid
Medium. Jour. of Appl. Phys., Vol. 21, p.
89-93.
Wang, B. and Wang, Y. (1996). Temporal
Structure of the Southern Oscillation as
Revealed by Waveform and Wavelet
Analysis. J. Climate 9, p. 1586-1598.
Widodo, W. and Rosyidi, S. A. (2009).
Experimental Investigation of Seismic
Parameters and Bearing Capacity of
Pavement Subgrade Using Surface Wave
Method. Sciencetific Journal of Semesta
Teknika, Vol.1, No. 12, p. 67-79.
Zagyapan, M. and Fairfield, C. A. (2002).
Continuous Surface Wave and Impact
Methods of Measuring the Stiffness and
Density of Railway Ballast. NDT & E
International, Vol. 35, p.75-81.
92
Sukarno: Analisis Antrian Akibat Daerah Sempit di Jalan Bebas Hambatan
Analisis Antrian Akibat Daerah Sempit
di Jalan Bebas Hambatan

Sukarno
Staf Pengajar Jurusan Teknik Sipil, Universitas Islam Indonesia, Yogyakarta,
email: sukarno@ftsp.uii.ac.id

Abstract: Traffic congestion which is a common phenomenon in big cities sometime is found on a
freeway as well. Congestion on a freeway happens when arrival flow is greater than departure
flow, for example when there is a road construction or an accident at downstream or greater
capacity at upstream compared than downstream. Two models, queuing and shockwave, are
accepted as a tool to analyze narrow area caused by over capacity. However, two models give
different results. The objectives of this study are to investigate those different results and to give
some suggestions if any on the next use. Parameters of queuing, which were indicators of their
characteristics, were determined using predetermined data, then the differences of their results
were discussed through concepts of queuing, delay, and flow stage changing. The modified shock
wave model was tested using the same data and the results show that modified model is slightly
different with queuing model.

Keywords: Queuing, narrow area, freeway, queuing model, and shockwave model.

PENDAHULUAN

Kemacetan terjadi hampir di setiap jalan
perkotaan tidak terkecuali di jalan bebas
hambatan. Kemacetan yang parah akan
menyebar dengan cepat dan berlangsung
sangat lama. Tujuan diadakannya jalan bebas
hambatan adalah untuk perjalanan dengan
kecepatan tinggi maka mengatasi masalah
kemacetan di jalan bebas hambatan menjadi
sangat penting. Menurut Papacostas dan
Prevedouros (1993) salah satu diantara banyak
masalah yang memerlukan perhatian dalam
manajemen kemacetan di jalan bebas
hambatan adalah perlunya prosedur analitis
untuk mempelajari efek kemacetan akibat
penyempitan (Gambar 1).
Menurut May (1990) untuk menganalisis
kemacetan akibat penyempitan didaerah bebas
hambatan biasa digunakan dua metode yaitu
metode antrian dan metode gelombang kejut.
Metode antrian memperlakukan lalu lintas
yang akan masuk daerah sempit sebagai
persoalan antrian tertentu sederhana,sedangkan
metode gelombang kejut menganalogikan arus
lalu lintas sebagai arus fluida sehingga ketika
Gambar 1. Daerah sempit dengan arus berbeda di jalan bebas hambatan
93
fluida memasuki daerah sempit akan timbul
gelombang fluida yang datangnya mendadak
(karena itu disebut gelombang kejut).
Gelombang fluida inilah yang dipakai untuk
memodelkan arus lalu lintas yang berubah
kecepatan, volume, dan kepadatannya ketika
memasuki daerah sempit (Garber and Hoel,
2009). Morales (1986) mengatakan bahwa
meskipun perambatan gelombang utamanya
berbentuk non-linier, para ahli biasa
menyederhanakan menjadi berbentuk linier.
McShane dan Roess (1990) mengatakan
terdapat perbedaan pada hasil yang diperoleh
pada kedua model. Dengan menggunakan
contoh numeris, mereka menunjukkan bahwa
jumlah antrian dan tundaan yang dihasilkan
model antrian 60% lebih rendah dibandingkan
dengan yang dihasilkan oleh model gelombang
kejutsedangkan lamanya kemacetan oleh
model antrian diprediksi 15% lebih rendah
dibandingkan hasil prediksi model gelombang
kejut. Shandutan (2011), yag meneliti
penentuan lokasi penyempitan dengan metode
gelombang kejut berhasil secara grafis
menjelaskan perambatan gelombang di hulu
daerah sempit, namun belum membandingkan
hasilnya dengan metode antrian. Walaupun
terdapat perbedaan pada kedua metode namun
dalam penerapan keduanya selalu digunakan
untuk memecahkan persoalan penyempitan di
jalan bebas hambatan.
Para ahli lalu lintas mengakui kesahihan kedua
model walaupun terdapat perbedaan hasil yang
cukup besar. Oleh karena itu, kedua model
perlu dievaluasi lebih dalam agar hal-hal yang
menjadi perbedaan dapat diketahui. Tujuan
dari karya ilmiah ini adalah untuk
mengevaluasi model antrian dan model
gelombang kejut dalam menganalisis
kemacetan akibat penyempitan di jalan bebas
hambatan kemudian mengusulkan apabila
memungkinkan modifikasi model agar kedua
metode sebanding.

TINJAUAN MODEL

Model Antrian
Penyempitan di bagian jalan bebas hambatan
terjadi ketika arus yang melewatinya melebihi
kapasitas bagian jalan bebas hambatan
tersebut. Terjadinya penyempitan ini dapat
disebabkan oleh dua hal. Pertama adalah
meningkatnya arus di hulu bagian jalan yang
ditinjau hingga melebihi kapasitas normal
(kapasitas di bagian jalan yang ditinjau).
Penyempitan ini biasanya bersifat tetap dan
efeknya terlihat hanya bila ada lalu lintas
tinggi di hulu. Kedua, penyempitan yang
terjadi karena pengurangan kapasitas yang
sifatnya sementara di bagian jalan bebas
hambatan yang ditinjau, misalnya karena ada
kecelakaan atau pengerjaan jalan.
Andaikan bagian jalan yang ditinjau adalah X
dengan kapasitas Q
c
. Bila arus lalu lintas
sebesar Q
1
(>Q
c
) datang, jalan di bagian hulu
X akan mengalami kemacetan. Andaikan Q
1

berlangsung selama t
1
(disebut arus tahap-1)
dan kemudian berubah menjadi Q
2
(< Q
c
)
selama t
2
t
1
(disebut arus tahap-2) maka
berdasarkan teori antrian tertentu (selanjutnya
disebut metode antrian), yang mengasumsikan
kedatangan kendaraan (A) dan keberangkatan
kendaraan (D) adalah linier, bentuk kurva
kedatangan dan keberangkatan dapat dilihat
pada Gambar 2. Dari Gambar 2 diperoleh:
a. Untuk kedatangan
Arus tahap-1 (berlangsung selama t
1
) :
A
1
(t) = Q
1
t dan,
Arus tahap-2 (berlangsung selama (t
2
-t
1
):
A
2
(t) = Q
2
t
b. Untuk keberangkatan
Arus tahap-1 (berlangsung selama t
1
) :
D
1
(t) = Q
c
t dan,
Arus tahap-2 (berlangsung selama (t
2
-t
1
):
D
2
(t) = Q
c
t
Jadi keberangkatan adalah konstan baik pada
arus tahap-1 maupun tahap-2.
Kendaraan yang tidak dapat memasuki daerah
sempit selama arus tahap-1 akan tertahan di
awal daerah sempit dan membentuk antrian.
Laju tertahannya kendaraan dalam antrian
adalah Q
1
-Q
c
. Pada arus tahap-2 tidak ada lagi
kendaraan yang tertahan karena Q
2
<Q
c
dan
kendaraan yang antri berkurang dengan laju
Q
2
-Q
c
. Saat t = t
1
jumlah kendaraan yang
tertahan dalam antrian mencapai maksimum
(n
m
) dan besarnya,
n
m
= A
1
(t
1
) D(t
1
) [kendaraan] (1)
94

Bila arus pada tahap-2 berlangsung cukup
lama, antrian akan hilang pada saat t = t
2
.
Waktu t
2
ini juga merupakan lamanya
kemacetan akibat adanya daerah sempit di
jalan bebas hambatan dan dapat dicari dari
lamanya arus tahap-1 (t
1
) ditambah lamanya
arus tahap-2 (t
2
-t
1
).
) Q (Q
n
t
) Q (Q
a) - (b
t ) t t ( t t
2 p
m
1
2 p
1 1 2 1 2
+ =
+ = + =
[jam]
) Q (Q
n
t t
maka macet saat t t t t bila
2 p
m
1 j
j j 2
+ =
= =

Total kendaraan yang berangkat selama waktu
macet (=t
j
= t
2
) adalah D(t
2
)
Bila D(t
2
) = N
j
maka,
N
j
= Q
c
t
j
[kendaraan] (3)
Tundaan tiap kendaraan digambarkan oleh
jarak horizontal antara kurva kedatangan
(kurva permintaan) dan kurva keberangkatan
(kurva kapasitas) dengan asumsi tidak ada
penyalipan. Dengan demikian tundaan
maksimum (d
m
) terjadi pada kendaraan
terakhir pada arus tahap-1 yaitu,
1
1 1
p
1 1
m
Q
) t ( A
Q
) t ( A
d =
atau,

jadi,
p
m
m
Q
n
d = [jam] (4)
Tundaan total (D), yaitu tundaan semua
kendaraan yang dipengaruhi oleh
kemacetan,dapat diperoleh dengan menghitung
luas segitiga (bagian yang dihitamkan) pada
Gambar 2 yaitu,
D = n
m
t
1
+ n
m
(t
2
-t
1
)
atau,
D = n
m
{(t
1
+(t
j
-t
1
)} [kendaraan jam] (5)
Tundaan rata-rata (d) dari semua kendaraan
yang terpengaruh kemacetan adalah tundaan
total (D) dibagi semua kendaraan yang
terpengaruh kemacetan (N
j
),
j
N
D
d = [jam] (6)

Model Gelombang Kejut
Dimisalkan hubungan arus (Q), kepadatan (K),
dan kecepatan (U) dibagian hulu daerah sempit
mengikuti metode Greenshield yaitu (lihat
Gambar 3a, 3b, dan 3c) yaitu,
a. Untuk hubungan U-K: K
K
U
U U
j
f
f
=
Dari hubungan Q = U K diperoleh,
b. Untuk hubungan Q-U: U K U
U
K
Q
j
2
f
j
+ =
c. Untuk hubungan Q-K: K U K
K
U
Q
f
2
j
f
+ =
keterangan: U
f
= kecepatan bebas dan K
j
=
kepadatan maksimum
Bila suatu saat arus menjadi Q
1
(>Q
c
) dengan
kepadatan K
1,
di awal daerah sempit akan
timbul gelombang kejut dan akan merambat
kedaerah hulu dengan kecepatan s
1
{besarnya
s
1
sesuai kemiringan garis (K
1
,Q
1
) dan (K
c
,Q
c
)
pada Gambar 3c} yaitu,
Gambar 2. Kurva kedatangan
(permintaan) dan keberangkatan
(kapasitas) pada model antrian
(2)
95
p 1
p 1
1
K - K
-Q Q
s = [km/jam] (7)
Panjang antrian akan meningkat kearah hulu
hingga waktu t
1
seperti ditunjukkan Gambar 4.
Setelah arus tahap-1 (lamanya t
1
) selesai, arus
tahap-2 mulai. Pada tahap ini gelombang kejut
lain terbentuk dan akan merambat kehilir
dengan laju s
2
{besarnya s
2
sesuai kemiringan
garis (K
2
,Q
2
) dan (K
c
,Q
c
) pada Gambar 3.c}
yaitu,
p 2
p 2
2
K - K
-Q Q
s = [km/jam] (8)
dalam keadaan ini antrian memendek dengan
laju s
2
dan akhirnya antrian hilang saat waktu
mencapai t
2
(Gambar 4).
Antrian terpanjang (maksimum) terjadi ketika
t = t
1
dengan panjang,
y
1
= s
1
t
1
[km]
Bila y
m
= y
1
maka,
y
m
= s
1
t
1
[km] (9)
kepadatan kendaraan dalam antrian adalah K
c
,
maka jumlah kendaraan maksimum dalam
antrian (n
m
) adalah,
n
m
= K
c
y
m
[kendaraan] (10)
kemacetan di hulu antrian berlangsung hingga
t
2
(=t
j
) dan besarnya adalah,
t
j
= t
1
+ (t
2
-t
1
)
t
j
= t
1
+
2
m
s
y
[jam] (11)
Total kendaraan yang berangkat (lepas dari
antrian) selama waktu macet (=t
j
= t
2
) adalah
N
j
maka,
N
j
= Q
c
t
j
[kendaraan] (12)
Tundaan Total (D) dapat diperoleh dengan
menghitung luas segitiga,
D = t
2
y
1

karena t
2
= t
j
dan y
1
= y
m
maka,
D = t
j
y
m
[km jam] (13)
kalau diinginkan satuan tundaan total dalam
kendaraan jam maka,
D = t
2
y
1
K
c
[kendaraan jam] (14)
Tundaan rata-rata (d) dari semua kendaraan
yang terpengaruh kemacetan adalah tundaan
total (D) dibagi semua kendaraan yang
terpengaruh kemacetan (N
j
),

Gambar 3. Hubungan Q-U-K untuk analisis gelombang kejut
96
j
N
D
d = [jam] (15)
Tundaan maksimum (d
m
) adalah tundaan yang
dialami kendaraan pada antrian terpanjang dan
dapat dicari melalui pembagian jumlah
kendaraan maksimum dalam antrian (n
m
)
dengan Q
c
,
p
m
m
Q
n
d = [jam] (16)

Penerapan Model
Kasus 1
Pandang jalan 1 lajur yang menyempit karena
perbaikan jalan. Misalkan kapasitas daerah
sempit Q
c
= 1200 kend/jam (kendaraan per
jam), arus tahap-1 Q
1
= 1500 kend/jam, dan
arus tahap-2 Q
2
= 900 kend/jam. Misalkan pula
arus tahap-1 berlangsung 1 jam (t
1
=1),
kecepatan arus bebas (U
f
) = 100 km/jam, dan
kepadatan maksimum (K
j
) = 100 kend/km.
a. Model antrian
A
1
(t)=1500 t, A
2
(t)=900 t, dan D(t)=1200 t
Panjang antrian maksimum = n
m
= A
1
(t
1
)
D(t
1
) = 1500(1) 1200(1) = 300 kend
Lamanya kemacetan =
jam 2
) 900 - (1200
300
1
) Q (Q
n
t t
2 p
m
1 j
= + =
+ =
Kendaraan total yang berangkat ketika
kemacetan berlangsung (N
j
),
N
j
= Q
c
t
j
=1200 (2) = 2400 kend
Tundaan maksimum
jam 0,25
1200
300
Q
n
d
p
m
m
= = =
Tundaan total (D)
D = n
m
{(t
1
+(t
j
-t
1
)}
= 300{1+(2-1)}= 300 kend jam
Tundaan rata-rata (d),
jam 125 , 0
2400
300
N
D
d
j
= = =
b. Model gelombang kejut
U
f
= 100 km/jam dan K
j
= 100 kend/km
Hubungan U-K:

Hubungan Q-K:

Hubungan Q-U:

Q
c
=1200 kend/jam 1200 = - K
2
+100K
K
c
=86 kend/km
Q
c
=1200 kend/jam 1200 = - U
2
+100U
U
c
=14 km/jam
atau U = 100 K = 100 86 U
c
= 14
km/jam
Q
2
=900 kend/jam 900 = - K
2
+100K
K
2
=10 kend/km
Q
2
=900 kend/jam 900 = - U
2
+100U
U
2
=90 km/jam
atau U = 100 K = 100 10 U
c
= 90
km/jam
Q
1
=1500 kend/jam 1500 = - K
2
+100K
K
1
=18 kend/km
Q
1
=1500 kend/jam 1500 = - U
2
+100U
U
1
=82 km/jam
Gambar 4. Laju panjang antrian akibat
jalan menyempit di jalan bebas hambatan
97
Tabel 1. Karakteristik antrian akibat penyempitan jalan bebas hambatan
Parameter
Model
Perbedaan
Antrian Gelombang kejut
Panjang antrian maks (n
m
) 300 kendaraan 380 kendaraan 27%
Lamanya antrian (t
j
) 2 jam 2,12 jam 23%
Jumlah kend. Berangkat (N
j
Tundaan total (D) 300 kendaraan jam 402 kendaraan jam 34%
Tundaan rata-rata (d) 0,125 jam 0,375 jam 28%
Tundaan maksimum (d
m
) 0,25 jam 0,32 jam 28%
Sumber: Hasil perhitungan (2011)

atau U = 100 K = 100 18 U
1
= 82
km/jam
Kecepatan gelombang yang menuju ke hulu
adalah s
1

Kecepatan gelombang yang menuju ke hilir
adalah s
2

Panjang antrian maksimum (dalam km)
y
m
= s
1
t
1
= 4,41 (1) = 4,41 km
atau (dalam kendaraan)
n
m
= K
c
y
m
= 86(4,41) = 380 kendaraan
Lamanya antrian (=t
j
),
t
j
= t
1
+
2
m
s
y
=1 +
3,95
4,41
= 2,12jam
antrian) selama waktu macet (N
j
) adalah
N
j
= Q
c
t
j
= 1200 (2,12) = 2544kendaraan
Tundaan Total (D) adalah
D = t
j
y
m
= (2,12) (4,41) = 4,47 km jam
atau dalam kendaraan jam,
D = t
j
y
m
K
c
= (2,12) (4,41) (86)
= 402 kend jam
Tundaan rata-rata (d) adalah,
jam 375 , 0
2520
945
N
D
d
j
= = =
Tundaan maksimum (d
m
) adalah,
jam 0,32
1200
380
Q
n
d
p
m
m
= = =
Ringkasan perhitungan karateristik antrian
dari kedua model dapat dilihat pada Tabel
1.
Dari Tabel 1 terlihat bahwa panjang antrian
maksimum hasil model gelombang kejut
27% lebih tinggi daripada hasil model
antrian, sedangkan tundaan hasil model
gelombang kejut rata-rata 30% lebih tinggi
daripada hasil metode antrian. Jumlah
kendaraan berangkat hasil gelombang kejut
6% lebih tinggi daripada hasil metode
antrian, sedangkan lamanya antrian hasil
gelombang kejut 23% lebih tinggi daripada
hasil metode antrian.

Kasus 2
Bila dicoba (untuk jalan 3 lajur) Q
c
= 4000
kend/jam, arus tahap-1 Q
1
= 5000 kend/jam,
dan arus tahap-2 Q
2
= 3000 kend/jam.
Misalkan pula arus tahap-1 berlangsung 1 jam
(t
1
=1), kecepatan arus bebas (U
f
) = 61 km/jam,
dan kepadatan maksimum (K
j
) = 393 kend/km.
Semua perhitungan dilakukan per lajur
a. Model antrian
A
1
(t)= (5000) t, A
2
(t)=(3000) t, dan
D(t)=(4000) t
98
Panjang antrian maksimum (n
m
)
n
m
= [A
1
(t
1
) D(t
1
)]
= [5000(1) 4000(1)] = (1000) kend
= 334 kend
Lamanya kemacetan (t
j
)

Kendaraan total yang berangkat ketika
kemacetan berlangsung (N
j
),
N
j
= Q
c
t
j
=(4000) (2) = (8000) kend
= 2667 kend
Tundaan maksimum (d
m
)
jam 0,25
4000
1000
3 1
3 1
Q
n
d
p
m
m
= = =
Tundaan total (D)
D = n
m
{(t
1
+(t
j
-t
1
)}
= 1000{1+(2-1)}=(1000) kend jam
= 334 kend
Tundaan rata-rata (d),
jam 125 , 0
8000
1000
3 1
3 1
N
D
d
j
= = =
a. Model gelombang kejut
U
f
= 61 km/jam dan K
j
= (393) = 131
kend/km
Hubungan U-K:

Hubungan Q-K:

Q
c
= 4000 kend/jam 4000 =
- 0,5K
2
+61 K K
c
= 102,5 kend/km
U = 61 0,47 K = 61 0,47(102,5)
U
c
= 12,82 km/jam
Q
2
=3000 kend/jam 3000 =
- 0,47K
2
+61 K K
2
=19 kend/km
U = 61 0,47K = 61 0,47(19)
U
2
=52,07 km/jam
Q
1
=5000 kend/jam 5000 =
- 0,47K
2
+61 K K
1
=38,5 kend/km
U = 61 0,47 K = 61 0,47(38,5)
U
1
= 42,90 km/jam
Kecepatan gelombang yang menuju ke hulu
adalah s
1

Kecepatan gelombang yang menuju ke hilir
adalah s
2

Panjang antrian maksimum (dalam km)
y
m
= s
1
t
1
= 5,2 (1) = 5,2 km

Tabel 2. Karakteristik antrian akibat penyempitan jalan bebas hambatan
Parameter
Model
Perbedaan
Antrian Gelombang kejut
m
Lamanya antrian (t
j
) 2 jam 2,3 jam 15%
Jumlah kend. Berangkat (N
j
Tundaan total (D) 334 kendaraan jam 613 kendaraan jam 84%
Tundaan rata-rata (d) 0,125 jam 0,2 jam 60%
Tundaan maksimum (d
m
) 0,25 jam 0,4 jam 60%
99
atau (dalam kendaraan)
n
m
= K
c
y
m
= 307,5(5,2) = 533
kendaraan
Lamanya antrian (=t
j
),
t
j
= t
1
+
2
m
s
y
=1 +
4
5,2
= 2,3 jam
j
) adalah
N
j
= Q
c
t
j
= 4000 (2,3) = 3067
kendaraan
Tundaan Total (D) adalah
D = t
j
y
m
= (2,3 ) (5,2) = 5,98 km
jam
atau dalam kendaraan jam,
D = t
j
y
m
K
c
= (2,3) (5,2) (307,5) =
613 kend jam
Tundaan rata-rata (d) adalah,
jam 2 , 0
3067
613
N
D
d
j
= = =
Tundaan maksimum (d
m
) adalah,
jam 0,4
(4000) 3 1
533
Q
n
d
p
m
m
= = =
Ringkasan perhitungan karateristik antrian
dari kedua model dapat dilihat pada Tabel
2.

PEMBAHASAN

Dari Tabel 1 dan 2 terlihat bahwa terdapat
perbedaan nilai untuk parameter yang sama
hasil prediksi dari kedua model baik untuk
kasus yang sama maupun berbeda.
Panjang antrian: Model gelombang kejut 27%
lebih tinggi daripada hasil model antrian
(Tabel 1), sedangkan pada Tabel 2 hasil
gelombang kejut 60% lebih tinggi daripada
hasil model antrian.
Tundaan: Hasil model gelombang kejut rata-
rata 30% lebih tinggi dari pada hasil model
antrian sedangkan (Tabel 1), sedangkan pada
Tabel 2 hasil gelombang kejut rata-rata 68%
lebih tinggi daripada hasil model antrian.
Lamanya antrian: Hasil model gelombang
kejut rata-rata 30% lebih tinggi dari pada hasil
model antrian sedangkan (Tabel 1), sedangkan
pada Tabel 2 hasil gelombang kejut rata-rata
68% lebih tinggi daripada hasil model antrian.
Perbedaan hasil ini dapat dijelaskan karena
adanya perbedaan konsep antrian, tundaan,
dan tahap arus yang dipunyai model antrian
dan model gelombang kejut.

Konsep Antrian
Model antrian menggambarkan kendaraan
yang akan memasuki daerah sempit sebagai
antrian berhenti yang hanya bergerak ketika
dilayani oleh daerah sempit dengan laju sesuai
kapasitas daerah sempit. Jadi model ini tidak
melihat bagaimana kendaraan bergerak dalam
antrian, mereka hanya menunggu dalam
antrian dan bergerak ketika dilayani. Panjang
antrian maksimum (N
m
) diperoleh dengan
dengan mengurangkan arus berangkat dari
arus datang, apa yang terjadi dalam antrian
tidak dilihat hingga hasilnya pasti berbeda
dengan model yang mempertimbangkan
bahwa antrian sebenarnya bergerak. Model
gelombang kejut menggambarkan antrian yang
mendekati daerah sempit sebagai antrian
bergerak dan pergerakan arus antrian ini
memenuhi hubungan arus (Q), kecepatan (U),
dan kepadatan (K) di daerah bebas hambatan
yaitu Q = U K. Antrian bergerak muncul
ketika arus datang melampaui kapasitas. Saat
itu timbul gelombang dengan laju tertentu dan
berarah ke hulu. Panjang antrian maksimum
diperoleh ketika arus datang yang lebih kecil
daripada kapasitas daerah sempit tiba. Ini
ditandai dengan bergeraknya gelombang yang
tadinya kehulu sekarang berbalik kehilir.
Kedua model akan menghasilkan panjang
antrian yang sama ketika kapasitas daerah
sempit (Qc) = 0 atau terjadi penutupan
sempurna didaerah sempit.

Konsep Tundaan
Karena berbeda dalam konsep antrian, tundaan
yang dihasilkan dari kedua model tentu saja
juga berbeda. Model antrian mengasumsikan
kendaraan dalam keadaan diam saat mengantri
sehingga,tundaan dapat dihitung dengan benar
100
karena mewakili waktu hilang ketika
kendaraan berhenti. Tidak demikian dengan
model gelombang kejut yang menggambarkan
antrian dalam keadaan bergerak, tundaan yang
diperoleh sebenarnya adalah tambahan waktu
perjalanan karena melewati antrian. Oleh
karena itu, untuk mencari tundaan waktu
perjalanan tanpa hambatan harus diketahui
terlebih dahulu.

Konsep Tahap Arus
Pada model antrian maupun model gelombang
kejut, arus tahap-1 berlangsung selama t
1
.
Asumsi ini cocok untuk model antrian karena
model ini tidak mempermasalahkan
bergeraknya antrian atau dengan kata lain
hanya mengacu pada dimensi waktu. Tidak
demikian halnya untuk model gelombang
kejut, model kedua ini mengasumsikan
kendaraan bergerak dan mempunyai hubungan
yang ditetapkan sebelumnya yaitu Q=UK,
artinya selain mengacu pada waktu model
kedua ini juga mengacu pada ruang. Oleh
karena itu, sangat penting untuk menyatakan
dimana letak perubahan arus tersebut di dalam
ruang.
Misalkan kendaraan pertama pada arus Q
1
tiba
di awal daerah sempit atau disebut titik 0 (titik
ketika waktu antrian dimulai) setelah
menempuh waktu -
1
dari suatu tempat sejauh
z
1
di hulu daerah sempit. Kecepatan arus Q
1
ini
adalah U
1
(=z
1
/
1
). Kendaraan terakhir pada
arus Q1 ini seharusnya tiba di titik 0 setelah t
1

namun kenyataannya tidak, hal ini dikarenakan
ada rambatan gelombang berarah kehulu yang
berawal dari titik 0. Kendaraan terakhir pada
arus Q
1
misalnya bertemu gelombang yang
merambat ke hulu di titik y
p
pada waktu t
p

(tentu saja y
p
<y
1
dan t
p
<t
1
).Waktu t
p
dapat
dicari dari perjalanan kendaraan terakhir
maupun dari perjalanan gelombang.
Dari kendaraan terakhir:

Dari perjalanan gelombang:

Diperoleh:

y
p
adalah panjang antrian maksimum. Bila
p
m
p
y y = maka,

Tundaan maksimum (
p
m
d ) diperoleh dengan
mengurangkan waktu perjalanan kendaraan
terakhir arus tahap-1 (kecepatan U
1
)(bila tak
ada penyempitan) dari waktu perjalanan
kendaraan terakhir ketika ada penyempitan
(kecepatan U
c
) yaitu,

Tundaan total (D
p
) ditentukan dengan
mengurangkan waktu perjalanan semua
kendaraan yang tidak terpengaruh kemacetan
(kecepatan U
1
) dari waktu perjalanan semua
kendaraan yang terpengaruh kemacetan
(kepadatan K
p
) yaitu,

.................................................................. (22)
Tundaan rata-rata (d
p
)

keterangan:

Penerapan Model Gelombang Kejut
Modifikasi
a. Dari data Kasus 1
Q
1
= 1500 kend/jam, K
1
=18 kend/km,
U
1
=82 km/jam, Q
c
= 1200 kend/jam, K
c
=86
101
kend/km, U
c
=14 km/jam, Q
2
=900 kend/jam,
K
2
=10 kend/km, U
2
=90 km/jam, t
1
=1,
U
f
=100 km/jam, K
j
=100 kend/km, s
1
=-4,41
km/jam, dan s
2
= 3,95 km/jam.
U
1
=82 km/jam dan t
1
= 1 jam, maka z
1
= 82
km

Panjang antrian maksimum
km 18 , 4
41 , 4 82
(4,41) 82
s U
s z
y
1 1
1 1 p
m
=
+
=
+
=
atau dalam satuan kendaran
kend 360 ) 18 , 4 ( 86 y K n
p
m c
p
m
= = =
Lama antrian

j
p
) adalah,
kend 2408 ) 006 , 2 ( 1200 t Q N
p
j c
p
j
= = =
Tundaan maksimum
jam 25 , 0
82
4,18
14
4,18
U
y
U
y
d
1
p
m
c
p
m p
m
= = =
Tundaan total = D
p

p
)
jam 126 , 0
2408
302,61
N
D
d
p
j
p
p
= = =

b. Dari data Kasus 2
Data ditampilkan per lajur: Q
1
=(5000)
kend/jam, K
1
=38,5 kend/km, U
1
=41,75
km/jam, Q
c
=(4000) kend/jam, K
c
=102,5
kend/km, U
c
=9,75 km/jam, Q
2
=(3000)
kend/jam, K
2
=19 kend/km, U
2
=51,5
km/jam, t
1
=1 jam, U
f
=61 km/jam,
K
j
=(393) kend/km, s
1
=-5,2 km/jam, dan
s
2
=4 km/jam
U
1
=42,90 km/jam dan t
1
= 1 jam, maka z
1
=
42,90 km

Panjang antrian maksimum
km 64 , 4
2 , 5 42,90
(5,2) 42,90
s U
s z
y
1 1
1 1 p
m
=
+
=
+
=
atau dalam satuan kendaran
kend 476 ) 64 , 4 ( 5 , 102 y K n
p
m c
p
m
= = =
Lama antrian

j
p
) adalah,
kend 2736 ) 052 , 2 ( 4000
3
1
t Q N
p
j c
p
j
= = =
Tundaan maksimum
jam 25 , 0
42,90
4,64
12,82
4,64
U
y
U
y
d
1
p
m
c
p
m p
m
= = =
Tundaan total = D
p

p
)

jam 130 , 0
2736
356
N
D
d
p
j
p
p
= = =

Ringkasan hasil gelombang kejut modifikasi
dengan hasil model antrian baik untuk kasus 1
maupun kasus 2 dapat dilihat pada Tabel 3.
102
Tabel 3. Perbandingan karakteristik antrian dari dua metode (antrian dengan gelombang kejut)
Parameter
Model
Perbedaan
Antrian Gel. Kejut Modifikasi
Kasus-1 Kasus-2 Kasus-1 Kasus-2 Kasus-1 Kasus-2
m
) 300 334 360 476 20% 42%
Lamanya antrian (t
j
) 2 2 2 2,05 0% 2,5%
Jumlah kend. berangkat (N
j
) 2400 2667 2408 2736 0,3% 2,6%
Tundaan total (D) 300 334 303 356 1% 6,6%
Tundaan rata-rata (d) 0,125 0,125 0,126 0,13 0,8% 4%
Tundaan maksimum (d
m
) 0,25 0,25 0,25 0,25 0% 0%

Dari hasil perhitungan ulang dengan metode
gelombang kejut modifikasi terlihat (Tabel 4.)
bahwa perbedaan dengan metode antrian turun
namun belum hilang sama sekali (terutama
parameter panjang antrian yang masih berbeda
cukup banyak yaitu rata-rata 30%). Perbedaan
ini disebabkan pengabaian interaksi antara
akhir arus tahap 1 dan awal arus tahap 2. Bila
interaksi diperhitungkan maka bentuk
rambatan gelombang kehilir tidak lagi linier
namun menjadi non linier yang membawa
akibat panjang antrian memendek. Pengabaian
dilakukan karena bentuk rambatan gelombang,
baik ke hulu maupun ke hilir, sudah
diasumsikan linier.

SIMPULAN DAN SARAN

1. Model antrian, yang modelnya berbasiskan
waktu, memodelkan kemacetan akibat
penyempitan di jalan bebas hambatan lebih
sederhana dibandingkan dengan model
gelombang kejut, yang modelnya
berbasiskan ruang dan waktu. Walaupun
begitu model gelombang kejut lebih
realistis karena modelnya dapat mendekati
keadaan sebenarnya di lapangan.
2. Hasil prediksi terhadap parameter
kemacetan dari kedua model berbeda.
Walaupun begitu perbedaan dapat
diperkecil dan dapat dikatakan sebanding
apabila dalam perhitungannya
dipergunakan metode gelombang kejut
modifikasi.
3. Untuk studi selanjutnya disarankan untuk
mempelajari bentuk interaksi antara akhir
arus tahap 1 dan awal arus tahap 2.
4. Karena hasil prediksi kedua model berbeda
(walaupun tidak signifikan), penggunaan
praktis dilapangan (model mana yang
dipakai) harus diserahkan pada ahlinya.
Konfirmasi dengan keadaan sebenarnya
perlu dilakukan agar revisi model (apabila
diperlukan) dapat dilakukan hingga model
yang betul-betul mewakili dapat ditemukan.

DAFTAR PUSTAKA

Garber, N. J. and Hoel, L. A. (2009). Traffic
and Highway Engineering. 4
th
Ed.,
Toronto, Canada.
May, A. D. (1990). Traffic Flow
Fundamentals. Englewood Cliffs:
Prentice-Hall Inc., New Jersey.
McShane, W. and Roess, R. P. (1990). Traffic
Engineering. Englewood Cliffs: Prentice-
Hall Inc., New Jersey.
Morales, J. M. (1986). Analytic Procedures
for Estimating Freeway Traffic
Congestion. Public Roads, Vol. 50, No.2,
Washington, D.C.
Papacostas, C. S. and Prevedouros, P. D.
(1993). Transportation Engineering and
Planning. Englewood: Prentice-Hall Inc.,
New Jersey.
Shandutan, D. A. S. (2009). A Queuing
Analysis of Bottleneck Formation and
Shockwave Propagation. Master Thesis,
(Unpublished), University of Minnesota,
Minnesota.
103

Journal UII Yogya

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Journal UII Yogya

Uploaded by

Copyright:

Available Formats

i

Vol.1, No. 1, Mei 2011 ISSN 1979-7303

British Standard Institution BS6073, Part 1.

RESULTS AND DISCUSSION

You might also like