2008 MIDAS Technical Conference

Construction Stage Analysis Reflecting Long-Term Deformation

1. Introduction

In high-rise buildings, the axial deformations of columns cannot be ignored, and special considerations are
required for design and construction. A vertical member undergoes both elastic deformation and
deformation due to creep and shrinkage. The elastic deformation takes place instantaneously due to dead
loads and live loads applied to the structure, while deformation due to creep and shrinkage occurs over
many years. Most of the vertical deformations in a high-rise building, however, take place during its
construction.

Due to the difference in axial stiffness and load distribution areas on vertical members, differential
shortening inevitably develops. If this differential shortening in vertical members, which takes place during
and after the construction, is not considered in analyses of high-rise buildings, structural safety will be
compromised due to additional stresses in the horizontal members and, subsequently, in the vertical
members. The structural safety problem is also magnified when incorporating serviceability issues such as
the curtain wall function, floor unevenness, excess stress in piping, etc. As such, total displacements of
vertical members must be calculated at the design stage. Comparatively reasonable and accurate results
can be predicted when construction stage analysis is carried out reflecting the creep and shrinkage
behavior of concrete.

Conventional structural analysis has the assumption that all structural loads are instantaneously applied to
the entire completed structure. However, since most buildings are constructed by one story or several floor
units at a time, or even if it is the same story, the construction sequence and loading sequence may be
different depending on the construction plan. Therefore, the actual structural behavior can be significantly
different from the conventional analytical behavior based on the above assumption.

Vertical members (columns and walls) in high-rise reinforced concrete buildings not only exhibit elastic
shortening, but also have shrinkage and creep effects that develop from long-term compressive loading. In
lower stories of a building, additional stresses in girders become very large due to differential shortening
and undergo significant redistribution of the member forces.

In order to analytically solve the problem described above, the construction stage analysis function of
midas Gen considers shrinkage and creep during construction stages to simulate the construction process
of a high-rise building. Also, with input variables, such as the strength of concrete, construction duration of
building components, casting condition, ambient condition, etc., the elastic shortening, shrinkage and creep
of vertical members can be estimated and are reflected in the analysis. Change in strength gain based on
the maturity of concrete members is also reflected in the calculation of modulus of elasticity at various
construction stages.

MIDAS Gen 2009 and Integrated Solutions -32-

 2008 MIDAS Technical Conference

In the following example, construction stage analysis considers the creep and shrinkage effects of a 40-
story building consisted of an exterior concrete frame and interior shear walls, as shown in figure 1. The
displacements of vertical members and the girder member forces are compared and evaluated with the
results from conventional analysis (analysis in which construction stages are not considered).

2. Model Summary

2.1 Structural System

The structural system, as shown in figure 1, is a 40-story building constructed with core walls and perimeter
RC columns & RC girders. In order to evaluate the influence that gravity has on the displacements of a
vertical member, and the member forces of horizontal members, the member sections are selected
according to the dead load and the live load. The typical plan view of a story is shown in figure 2.

Figure 1: 3D Analysis Model

MIDAS Gen 2009 and Integrated Solutions -33-

 2008 MIDAS Technical Conference

8,000 9,000 8,000

4,500 4,500

9,400

3,800
9,400

3,800 1,800
9,400

Figure 2: Typical Story Plan View

2.2 Design Loads
Table 1: Design Loads
Type of Loads Symbols Loads

Frame Self Weight SW -

Slab Self Weight Dc 360 kgf/m2
Finishing and Masonry Loads Ds 160 kgf/m2

Live Loads LL 250 kgf/m2

2.3 Construction Sequence

During the construction stages (please refer to table 2 and figure 3), after the core walls are
constructed on the 4th story (after construction of three preceding stories is completed through a 5 day
cycle), the girder, column, and slab construction is initiated at the same time. During the construction of the
21st story, the interior finishing for the floor is initiated, starting from the 1st story, in a 5 day cycle. It is
assumed that the live loads applied to the structure reach 100% loading 90 days after the completion of
construction.

MIDAS Gen 2009 and Integrated Solutions -34-

 2008 MIDAS Technical Conference

Table 2: Construction Sequence

Total
Stage Duration
Stage Element Group Construction Description
Load (days)
Period (days)

#CS1 1st Fl. core SW/Dc 5 5

#CS2 2nd Fl. core SW/Dc 5 10

#CS3 3rd Fl. core SW/Dc 5 15

#CS4 4th Fl. core 1st Fl. Frame SW/Dc 5 20

#CS5 5th Fl. core 2 Fl. Frame SW/Dc 5 25

#CS6 6th Fl. core 3 Fl. Frame SW/Dc 5 30

#CS7 7th Fl. core 4 Fl. Frame SW/Dc 5 35

… … … … … …

#CS21 21st Fl. core 18 Fl. Frame SW/Dc/Dd 5 105 1st Fl. Int.
Finishing
#CS22 22nd Fl.core 19 Fl. Frame SW/Dc/Dd 5 110 2nd Fl. Int.
Finishing
… … … … … … …

#CS40 40th Fl. core 37 Fl. Frame SW/Dc/Dd 5 200 20th Fl. Int.
Finishing
#CS41 38 Fl. Frame SW/Dc/Dd 5 205 21st Fl. Int.
Finishing
#CS42 39 Fl. Frame SW/Dc/Dd 5 210 22nd Fl. Int.
Finishing
#CS43 40 Fl. Frame SW/Dc/Dd 5 215 23rd Fl. Int.
Finishing

#CS44 Dd 85 300 24~40 Fl. Int.

Finishing

#CS45 90 390

#CS46 LL 260 650 Live Loads

MIDAS Gen 2009 and Integrated Solutions -35-

 2008 MIDAS Technical Conference

Stage 4 Stage 8

Stage 21 (Input Finishing Loads)

Stage 35 Final Stage

Figure 3: Sequence of Construction Stages

MIDAS Gen 2009 and Integrated Solutions -36-

 2008 MIDAS Technical Conference

2.4 Material Properties

In order to define the properties of concrete shrinkage and creep, the ACI standard is used, and the
concrete material properties are shown in table 3.

Table 3: Concrete Material Properties

Comp. V/S
Humidity Slump Aggregate Air Cement
Column Strength Ratio Code
(%) (cm) (%) (%) (kg/m3)
(kgf/cm2) (mm)
C1 500 55 12 60 4.5 450 300 ACI
(1~5th)
C1 500 55 12 60 4.5 450 275 ACI
(6~10th)
C1 400 55 12 60 4.5 380 250 ACI
(11~15th)
C1 400 55 12 60 4.5 380 225 ACI
(16~20th)
C1 350 55 12 60 4.5 350 200 ACI
(21~25th)
C1 350 55 12 60 4.5 350 175 ACI
(26~30th)
C1 300 55 12 60 4.5 320 150 ACI
(31~35th)
C1 300 55 12 60 4.5 320 125 ACI
(36~40th)
C2 500 55 12 60 4.5 450 225 ACI
(1~10th)
C2 400 55 12 60 4.5 380 200 ACI
(11~20th)
C2 350 55 12 60 4.5 350 175 ACI
(21~30th)
C2 300 55 12 60 4.5 320 150 ACI
(31~40th)
Girder G1 270 55 12 60 4.5 302 127 ACI

Girder G2 270 55 12 60 4.5 302 100 ACI

Link beam 270 55 12 60 4.5 302 127.3 ACI

Wall 500 55 12 60 4.5 450 266 ACI

(1~10th)
Wall 400 55 12 60 4.5 380 266 ACI
(11~20th)
Wall 350 55 12 60 4.5 350 226 ACI
(21~30th)
Wall 300 55 12 60 4.5 320 226 ACI
(31~40th)

MIDAS Gen 2009 and Integrated Solutions -37-

 2008 MIDAS Technical Conference

The concrete creep coefficient data and graph, as well as the shrinkage strain data and graph, as per the
ACI standard, are shown in figure 4 and figure 5, respectively.

Figure 4: Concrete Creep Coefficient Data & Graph by ACI Code

Figure 5: Shrinkage Strain Data & Graph by ACI Code

MIDAS Gen 2009 and Integrated Solutions -38-

 2008 MIDAS Technical Conference

3. Bending Moment Comparison for Girders and Shortening of Vertical Members

3.1 Comparison of Shortening in Vertical Members

In high-rise buildings, there are two types of vertical deformations, which take place at the nth story, the
vertical deformation that takes place immediately after casting concrete up to the nth story, and the vertical
deformation that takes place after subsequent floors are cast above the nth story. In a reinforced concrete
building, additional stresses on a girder at the nth story are not generated at the time of casting the nth floor,
since there is no relative vertical displacement between the columns supporting the girder. However,
additional loads above the nth story cause shortening in the columns, and if differential shortening exists
between the columns, additional stresses are generated in the girder. Again, shortening is caused by
elastic deformation and deformations due to the long term effects of creep and shrinkage. Therefore, the
deformation is calculated after the completion of the nth story floor structure, which is referred to as “subto”
shortening. Generally, each floor is leveled on-site when casting concrete.

The elastic deformation and deformations due to creep and shrinkage of an actual structure cannot be
physically isolated; but, for the purpose of analysis, they are separately calculated. The observation
8,000 9,000 8,000

points for the shortening of vertical members are selected at a column and a wall, highlighted in figure 6.
4,500 4,500
9,400
3,800
9,400
1,800
3,800
9,400

Figure 6: Observation Points for Shortening in Vertical Members

Deformation after 650 days, which is 260 days after applying the live loads, is calculated. Disparity may
occur depending on the magnitudes of the live loads assumed for shortening calculation. In this case,
100% of the design live loads is applied.

From the analysis results, deformations due to the creep and shrinkage effects, as shown in table 4 and
table 5, contribute to 61.7~73.6% of the total deformations of the column and 70.1~83.1% of the total
deformations of the wall.

MIDAS Gen 2009 and Integrated Solutions -39-

 2008 MIDAS Technical Conference

Table 4: The Deformation of Column (650days)

Column
Floors Elastic Deformation Deformation due to C. & S. Total Deformation
deformation ratio(%) deformation ratio (%) (mm)
40th 14.9 26.4 41.4 73.6 56.3
39th 16.5 28.1 42.1 71.9 58.6
38th 17.8 29.5 42.5 70.5 60.3
37th 18.9 30.7 42.7 69.3 61.6
36th 19.7 31.6 42.5 68.4 62.2
35th 20.2 32.4 42.1 67.6 62.3
34th 20.8 33.2 41.9 66.8 62.7
33rd 21.3 33.9 41.5 66.1 62.8
32nd 21.5 34.4 41.0 65.6 62.4
31st 21.5 34.8 40.2 65.2 61.6
30th 21.3 35.2 39.1 64.8 60.4
29th 21.4 35.7 38.6 64.3 60.0
28th 21.3 36.1 37.8 63.9 59.2
27th 21.1 36.4 36.9 63.6 58.1
26th 20.8 36.6 35.9 63.4 56.6
25th 20.3 36.9 34.7 63.1 54.9
24th 20.0 37.2 33.8 62.8 53.8
23rd 19.6 37.4 32.8 62.6 52.4
22nd 19.0 37.5 31.7 62.5 50.7
21st 18.4 37.7 30.4 62.3 48.8
20th 17.6 37.8 29.0 62.2 46.6
19th 17.1 37.9 28.0 62.1 45.1
18th 16.5 38.0 26.8 62.0 43.3
17th 15.8 38.1 25.6 61.9 41.4
16th 15.2 38.4 24.3 61.6 39.5
15th 14.1 38.2 22.8 61.8 36.8
14th 13.4 38.3 21.6 61.7 34.9
13th 12.6 38.3 20.3 61.7 32.9
12th 11.7 38.2 18.9 61.8 30.6
11th 10.7 38.1 17.4 61.9 28.1
10th 9.7 38.0 15.8 62.0 25.5
9th 8.9 38.0 14.6 62.0 23.5
8th 8.1 38.0 13.2 62.0 21.3
7th 7.2 37.9 11.8 62.1 19.0
6th 6.2 37.8 10.2 62.2 16.5
5th 5.2 37.7 8.6 62.3 13.8
4th 4.3 37.6 7.2 62.4 11.5
3rd 3.4 37.5 5.6 62.5 8.9
2nd 2.3 37.3 3.9 62.7 6.3
1st 1.2 37.0 2.1 63.0 3.4

MIDAS Gen 2009 and Integrated Solutions -40-

 2008 MIDAS Technical Conference

Table 5: The Deformation of Wall (650days)

Wall

Elastic Deformation Deformation due to C. & S. Total

Floors
Deformation
deformation ratio(%) deformation ratio (%)
(mm)
40th 4.6 16.9 22.5 83.1 27.1
39th 5.1 18.3 22.8 81.7 27.9
38th 5.6 19.7 23.0 80.3 28.6
37th 6.1 20.8 23.1 79.2 29.2
36th 6.5 21.9 23.2 78.1 29.7
35th 6.9 22.8 23.3 77.2 30.2
34th 7.2 23.6 23.3 76.4 30.5
33rd 7.5 24.4 23.2 75.6 30.7
32nd 7.7 25.0 23.2 75.0 30.9
31st 7.9 25.6 23.0 74.4 30.9
30th 8.1 26.1 22.8 73.9 30.9
29th 8.2 26.6 22.6 73.4 30.8
28th 8.3 27.1 22.3 72.9 30.6
27th 8.3 27.5 22.0 72.5 30.3
26th 8.3 27.8 21.6 72.2 30.0
25th 8.3 28.2 21.2 71.8 29.5
24th 8.2 28.5 20.7 71.5 29.0
23rd 8.1 28.7 20.2 71.3 28.3
22nd 8.0 29.0 19.6 71.0 27.6
21st 7.8 29.2 19.0 70.8 26.8
20th 7.6 29.4 18.3 70.6 25.9
19th 7.5 29.7 17.8 70.3 25.4
18th 7.3 29.7 17.3 70.3 24.7
17th 7.2 29.8 16.8 70.2 24.0
16th 6.9 29.9 16.3 70.1 23.2
15th 6.7 29.9 15.7 70.1 22.4
14th 6.4 29.9 15.0 70.1 21.4
13th 6.1 29.9 14.3 70.1 20.4
12th 5.8 29.8 13.6 70.2 19.4
11th 5.4 29.6 12.8 70.4 18.2
10th 5.0 29.3 12.0 70.7 17.0
9th 4.6 29.3 11.2 70.7 15.8
8th 4.2 29.2 10.2 70.8 14.5
7th 3.8 29.1 9.3 70.9 13.1
6th 3.3 28.9 8.2 71.1 11.6
5th 2.9 28.7 7.1 71.3 10.0
4th 2.4 28.5 5.9 71.5 8.3
3rd 1.8 28.2 4.6 71.8 6.4
2nd 1.3 28.0 3.2 72.0 4.5
1st 0.7 28.1 1.7 71.9 2.4

MIDAS Gen 2009 and Integrated Solutions -41-

 2008 MIDAS Technical Conference

Figure 7: Distribution of the Vertical Displacements of Column by Stories

Figure 8: Distribution of the Vertical Displacements of Wall by Stories

MIDAS Gen 2009 and Integrated Solutions -42-

 2008 MIDAS Technical Conference

Figure 7 and figure 8 show the distribution graphs of the vertical displacements by stories for the column
and wall, respectively. As the number of stories increases in conventional analysis in which construction
stages are not considered, the maximum vertical displacements occur at the highest story (column:
43.5mm, wall: 16.4mm). When construction stages are considered, the column exhibits the maximum
vertical displacement of 62.8mm at the 33 story, and the wall exhibits the maximum vertical displacement of
32.4mm at the 36 story. These values gradually decrease with the increase in stories.

Because of the difference in axial stiffness, columns generally have greater deformations compared to
walls. After performing construction stage analysis with the creep and shrinkage effects, the magnitude of
differential shortening is greater than that of the conventional analysis, which does not consider the creep
and shrinkage effects. The 28th story shows the greatest difference (29.6%) in differential shortening
between the conventional analysis results and the construction stage analysis results, as shown in table 6.
In addition, the differential shortening due to creep and shrinkage from the construction stage analysis
amounts to 54.3%. Differential shortening in vertical members creates serious problems in serviceability of
non-structural components and induces additional stresses on the horizontal members of the structural
frame. Therefore, an exact evaluation becomes very important. Differential shortening is underestimated if
the creep and shrinkage effects are omitted in analysis. Differential shortening due to creep and shrinkage
cannot be ignored. Such difference becomes magnified with the increase in the height of buildings or the
construction time.
For reference, when midas Gen calculates the differential shortening for 10,000 days (approximately 27
years), the results show the differential shortening at the 28th story of 32.8mm, a 15% increase from the
results for 650 days.

As mentioned previously, the displacements due to creep and shrinkage contribute to the displacements of
all columns and walls by 1.5~4 times more than the displacements with no creep and shrinkage effects.
Observing the differential shortening, displacements due to creep and shrinkage are less influential at the
stories 1 to 16, and more pronounced at the stories above the 16th story, also shown in figure 9. The higher
the story, therefore, the more likely that differential shortening will be dominated by displacements due to
creep and shrinkage.

Table 6: Comparison between Differential Shortening Amounts at the 28th Story

Conventional Construction Stage Analysis
Analysis Elastic S&C Total

Differential Shortening
20.1 mm 13.1 mm 15.5 mm 28.6 mm
Amount

Ratio (%) 70.4 45.7 54.3 100

Comparison Column>Walll Column>Walll Column>Walll Column>Walll

MIDAS Gen 2009 and Integrated Solutions -43-

 2008 MIDAS Technical Conference

Figure 9: Distribution of the Differential Shortening between Column & Wall

3.2 Comparing Bending Moments of Girders

In the following four cases, analysis results are compared for the bending moments of girders connected by
the column and wall:

 Case 1: Conventional analysis without construction stages

 Case 2: Analysis assuming infinite axial stiffness of vertical members
 Case 3: Construction stage analysis without creep and shrinkage effects
 Case 4: Construction stage analysis with creep and shrinkage effects

Case 2 is a method that is typically used for the horizontal member design of a (high-rise) building, which
ignores additional stresses due to differential shortening. The maximum end moment of a girder occurs at
the wall connection, and the difference in moments along the girder becomes greater at the higher stories.

For the negative moment at the wall connection of the girder at the 40th story, as shown in table 7, the result
from the conventional analysis without the influence of construction stages is 66.76tf-m, which is a 22.6%
overestimation when comparing to the result from the construction stage analysis with the influence of
creep and shrinkage effects (54.44tf-m). For the conventional analysis, we note that the differential

MIDAS Gen 2009 and Integrated Solutions -44-

 2008 MIDAS Technical Conference

shortening between the vertical members is overestimated.

The maximum end moment is 36.92tf, when the differential shortening of the vertical members is ignored.
In this case, it is underestimated by 32.2% compared to the proper stage analysis with creep and shrinkage
effects. Therefore, when differential shortening is ignored, significant errors will exist in the horizontal
member design forces.

In addition, the construction stage analysis without the creep and shrinkage effects results in the negative
moment of 49.53tf-m; whereas, the construction stage analysis that considers deformations due to creep
and shrinkage results in 54.44tf-m. A closer result is attained, but it still underestimates it by about 9%.

The additional moment due to the displacements caused by creep and shrinkage is 4.91tf-m. It can be
inferred that the difference between Case 2 and Case 3 is 12.61tf-m, or approximately 40%. As the
building’s height increases and girder length becomes shorter, the additional moment due to the
displacements caused by creep and shrinkage becomes greater. This confirms that the creep and
shrinkage effects cannot be ignored when designing horizontal members.

Table 7: Comparison of Bending Moment of Girder among Different Analytical Methods

Moment(tf-m)
Case
Column Ratio Mid Ratio Wall Ratio
Conventional Analysis 2.18 -27.9% 26.24 92.0% -66.76 122.6%

Deformation Ignored -22.94 293.4% 27.6 96.8% -36.92 67.8%

40th
C.S. w/ no C. & S. -12.5 159.8% 26.52 93.0% -49.53 91.0%

C.S. Total -7.82 100.0% 28.52 100.0% -54.44 100.0%

Conventional Analysis -13.86 89.4% 23.29 83.4% -48.51 105.2%

Deformation Ignored -22.94 148.0% 27.6 98.9% -36.92 80.1%

20th
C.S. w/ no C. & S. -16.8 108.4% 23.36 83.7% -45.44 98.5%

C.S. Total -15.5 100.0% 27.91 100.0% -46.12 100.0%

Conventional Analysis -30.95 101.9% 21.88 75.9% -29.68 96.8%

Deformation Ignored -22.94 75.5% 27.6 95.7% -36.92 120.4%

1st
C.S. w/ no C. & S. -30.85 101.5% 21.9 75.9% -29.74 97.0%

C.S. Total -30.38 100.0% 28.84 100.0% -30.66 100.0%

MIDAS Gen 2009 and Integrated Solutions -45-

 2008 MIDAS Technical Conference

Figure 10 shows the moment of each analysis case for the girder at the 40th story.

a) Conventional Analysis

b) Vertical Deformation Ignored

c) Construction Stage Analysis

Figure 10: Bending Moment of Girder among Different Analytical Methods at the 40 th Story

MIDAS Gen 2009 and Integrated Solutions -46-

 2008 MIDAS Technical Conference

Figure 11 and figure 12 show the end moment of the girder at each story. The end moment at the column
and wall is relatively close at the lower floors. The higher floors, however, experience larger end moments
at the wall connections with smaller end moments at the column connections. In the lower floors, the
moment due to the gravity loads, is predominant. In the higher levels, the moment due to the differential
shortening of vertical members become more significant, which cannot be ignored.

In addition, the end moment from the construction stage analysis, which reflects the creep and shrinkage
effects, is greater than the end moment from the construction stage analysis, which did not consider these
effects, but the moment is smaller when compared to the end moment from the conventional analysis.

Figure 11: End Moment of Girder by Stories (Column Connection)

MIDAS Gen 2009 and Integrated Solutions -47-

 2008 MIDAS Technical Conference

Figure 12: End Moment of Girder by Stories (Wall Connection)

Figure 13: Maximum Positive Moment of a Girder by Stories

MIDAS Gen 2009 and Integrated Solutions -48-

 2008 MIDAS Technical Conference

4. Conclusion

The construction stage analysis reflecting deformations due to creep and shrinkage of the 40-story
reinforced concrete structure consisting of core walls and exterior frame shows the following results:

 The proportion of the deformations due to creep and shrinkage that contribute to the total amount of
deformations is 61.7~73.6% for the column, and 70.1~83.1% for the wall. For concrete buildings,

therefore, deformations due to the creep and shrinkage effects must be considered.

 The amount of differential shortening due to deformations caused by creep and shrinkage is 54.3% of
the total differential shortening. Since there are considerable amounts of deformations due to creep
and shrinkage, their effects must be considered in analysis. This fact becomes more significant for
high-rise construction or for structures with longer construction periods.

 The maximum negative moment of the girder at the highest story shows a significant difference

between the construction stage analysis with the creep and shrinkage effects and the conventional
analysis, which does not consider construction stages.

 The moment in the girder due to deformations caused by creep and shrinkage has a large influence

on the girder’s member forces. About a 40% difference in values is observed between the construction
stage analysis with the creep and shrinkage effects and the analysis results ignoring the creep and
shrinkage effects. As the building’s height increases and the girder length decreases, the difference
in values becomes greater. When this influence is ignored, serious design problems will occur.

MIDAS Gen 2009 and Integrated Solutions -49-