Journal of Advanced Materials, Vol. 38, No.

1, 2006

PROCESSING AND PROPERTIES OF A LIGHTWEIGHT

FIRE RESISTANT CORE MATERIAL FOR SANDWICH
STRUCTURES
K. N. Shivakumar, S. D. Argade, R. L. Sadler, M. M. Sharpe, L. Dunn, G. Swaminathan,
Center for Composite Materials Research
Department of Mechanical and Chemical Engineering
N. Carolina A&T State University, Greensboro, NC
and
U. Sorathia 1
Naval Surface Warfare Center, Carderock Division, W. Bethesda, MD

1
"The technical views expressed in this paper are the opinions of the contributing authors, and do
not represent any official position of the U. S. Navy. This paper is declared a work of the US
Government and is not subject to copyright protection in the United States."

1
Journal of Advanced Materials, Vol. 38, No. 1, 2006

ABSTRACT

A process for syntactic foam made from fly ash, a waste product of coal

combustion from thermal power plants, has been developed using phenolic resin binders

at low levels. The fly ash consists of hollow glass or ceramic microspheres and needs to

be treated to remove contaminants. The production process is easily scalable and can be

tailored to produce foams of desired properties for specific applications. Complex

shaped parts also are possible with appropriate compression mold tooling. Mechanical

properties, compression, tension, shear and fracture toughness, have been determined in

this preliminary investigation on this syntactic material and are found to be comparable

or better than commercially available core materials. Initial testing for fire resistance has

indicated very encouraging results. Further work is being continued to develop this core

material with superior mechanical and fire resistance properties.

1. INTRODUCTION

Core materials are used extensively throughout the composites industry to

fabricate stiff yet lightweight products. The typical use of core materials is in sandwich

construction consisting of top and bottom face sheets and middle core material. The core

provides an order of magnitude increased flexural stiffness compared to the laminate

without the core. Alternatively to achieve an identical flexural stiffness, the sandwich

composite laminates saves considerable amount of weight compared to the laminate. The

face sheets are comparatively thin and are made of a material of high strength and

stiffness. The core is relatively thick and provides sufficient stiffness and strength in the

direction normal to the plane of the face sheet. In principle in sandwich construction the

2
Journal of Advanced Materials, Vol. 38, No. 1, 2006

face sheets carry the bending stresses applied by a load and core carries the shear stresses

(e.g.1-3).

A variety of core materials are used in the composites industry. In aerospace

applications honeycomb cores made from aluminum, phenolic-resin impregnated

fiberglass, paper, polypropylene, and Aramid fiber are extensively utilized. Skins are

made of sheets of aluminum alloys for aircraft structures. Fabrication of extremely

lightweight flat panels is possible with honeycomb cores; however, costs and moisture

problems with honeycomb composites negate its superior stiffness advantage. (3-5)

For composite applications that are less weight critical and where cost is an

important factor, core materials of the following type are considered in sandwich

construction: End-grain balsa; closed-cell foam made from thermoplastics such as PVC

or polyimide; foamed thermoset resins such as phenol-formaldehyde; carbon foam and

syntactic foam (4). End-grain balsa and many foamed plastics are not suitable for

complex shape constructions. Another major drawback with most of the currently

commercially available core materials is that they are not fire resistant or would emit

noxious gases when exposed to a fire (6).

Syntactic foams are made by embedding preformed hollow microspheres in a

resin matrix. The lightweight hollow microspheres reduce the density of the resin and

create a thick mixture that can be applied by hand or sprayed or can be compression

molded in a suitable mold. Syntactic foams are used in many applications, such as

underwater buoyancy aids, aerospace plug manufacturing and structural components for

ship hulls and bulkheads (7-9).

3
Journal of Advanced Materials, Vol. 38, No. 1, 2006

North Carolina A&T State University, Composite Materials Research Center has

developed (10) a production method for syntactic foam made from a waste product

generated by the utility industry. The fly ash, a byproduct of coal combustion already

exposed to high temperatures of >1000oC, is in the form of hollow glass or ceramic

bubbles and is normally collected in the filter bag houses to reduce the particulate

emissions from the thermal power plants. The fly ash contains some lime and gypsum

that are products of desulfurization used to minimize SO2 emissions. The fly ash needs to

be classified to separate the heavier components from it. By utilizing the treated fly ash

and combining it with an appropriate resin that has a high char value (contributes

minimum fuel value in case of a fire), low-cost syntactic foams can be produced with

properties tailored to the specific applications. Phenol-formaldehyde or other suitable

high-temperature thermoset resins that have a low fuel value and a high char yield can be

employed as binders. Fire resistant syntactic foams can be produced using phenol based

thermoset resins with low binder percentages. In the event of a fire this foam material

should generate little or no smoke and emit no toxic gases, a factor that would be of great

importance for naval and other military applications as well as building constructions.

The difference between the present and previous syntactic foams is the use of low binder

content (about 6% by volume) and flyash.

2. EXPERIMENTAL

2.1 Materials A class of fly ash known as Cenosphere or Recyclosphere grades CG 100

and SG 300 was obtained from Sphere Services Inc. The binder resin was a phenol-

formaldehyde resole resin, Durite SC 1008 supplied by Borden Chemical Co. The fly ash

was treated with a silane coupling agent, aminoalkyl triethoxysilane, obtained from either

4
Journal of Advanced Materials, Vol. 38, No. 1, 2006

Gelest Company or Aldrich Chemicals. An epoxy resin additive D.E.N. 431 from Dow

Chemical Co. and a silicone additive, polydimethyl siloxane diglycidyl ether, obtained

from Aldrich were used in a few syntactic foam panels.

2.2 Syntactic Foam Processing A schematic of the processing is shown in Figure 1. The

fly ash was treated to remove lime components by a dilute acid (pH ~ 4), and the heavier

than water fraction of the ash was separated and removed by settling. The lighter floating

fraction material was further washed with water 3-4 times to dilute the acid content. The

floaters were scooped out and then they were thoroughly dried at 110oC in a convection

oven. Subsequently, the treated fly ash was treated with a silane coupling agent, as per

instructions from the silane manufacturer. The fly ash after silane treatment was dried in

an oven to attain a free-flowing material.

Figure 1 Process flow diagram for producing syntactic foam from fly ash

The treated and dried fly ash was typically admixed with the resole resin diluted

with suitable solvents in a low-shear planetary motion mixer to uniformly coat the fly ash

particles. The volatile solvents from the fly ash mixture were removed while mixing in a

stream of warm air. The coated fly ash mix was subsequently placed in a compression

5
Journal of Advanced Materials, Vol. 38, No. 1, 2006

mold of 152x152x25 mm or 330x330x25 mm dimensions and was pressed in a laboratory

hot press to be finally cured at 162oC for one hour. The press pressure depends on the

required density of the panel. The minimum pressure used is about one atmosphere (100

kPa) for low density panels and 700 kPa for higher density panels. It was found that to

achieve reproducibility from sample to sample the void fraction in the foam panels had to

be controlled as low a value as possible. The foam samples were post cured at a

temperature of 177oC for three hours.

2.3 Characterization and Evaluation Mechanical testing was carried out using

compression, tension, shear and fracture toughness according to ASTM standards C-365,

D-3039/D-3039M, D-5379 and E-399, respectively. The specimen configurations used

for these tests are shown in Figure 2. The density measurements were carried out on the

cored specimens used in compression tests. Tensile, shear and fracture toughness

specimens were prepared by machining syntactic foam panels as per test requirements

using templates. Both compression and fracture toughness tests were used as quality

control tests during the process development. Panels of size 152x152x25.4 mm were

fabricated and four compression samples of diameter 29.2 mm and two 3-point bond

fracture toughness specimens were machined. Figure 3 shows the specimen layout.

Compression specimens were represented by C’s, indentation specimens by I’s and the

two fracture specimens were represented by MC and TC. Out of the two bend-fracture

specimens, one had the through-the-thickness crack (TC) and the other one had the mid-

plane crack (MC). The two specimens will measure the average and mid-plane toughness

of the material. Separate samples of 102x102 mm size with two different thickness (12.7

and 25.5 mm) were fabricated for fire test. Fire resistance tests were carried out at the

6
Journal of Advanced Materials, Vol. 38, No. 1, 2006

Naval Surface Warfare Center, Carderock Division, using cone calorimetry, as per

ASTM Standard E 1354. The ASTM methods and the sample preparation for the

characterization are described in more detail in the following sections.

d (29.2)

GL (88.9)
h

W (50.8) W/2

a) Compression Test:
ASTM C 365 b) Tension Test: ASTM D 3039/D-3039M (t = 12.7)

L/2

90° P
b

W (12.0) d1 (20.0)
w
d2 (4.0) a

L (76.0) r (3.0)
S

c) Iosipescu Shear Test: d) Fracture Test: ASTM E 399

ASTM D 5379/D-5379M (t = 12.7)

Figure 2 Specimen configurations

C1 C2

I1 C3

MC TC
I2 29.2 mm

C4

51 mm 38 mm 51 mm

152 mm

7
Journal of Advanced Materials, Vol. 38, No. 1, 2006

Figure 3 Compression and fracture toughness specimen layout

3. RESULTS AND DISCUSSION

3.1 Compression Behavior Fly-ash syntactic foam panels of 25.4 mm thickness were

cored using a diamond coated hole-saw. The cylindrical core specimens were lightly

polished to remove any surface irregularities and cleaned to remove the surface debris

prior to making dimension and density measurements. Four core samples were taken

from each panel, and the averages of density and compression strength were determined.

The compression tests were performed according to ASTM C365 on an Instron

4204 electromechanical testing machine. Each cylindrical sample was compressed

between two flat platens at a constant displacement rate of 0.51 mm/min while load and

displacement were recorded every second. Compressive stress and strain were calculated

as load/area and displacement/initial height, respectively.

The variation of compression strength with binder content is shown in Figure 4

and as a function of density is shown in Figure 5. The compression strength appears to

be a linear function of density. The density also is a nearly linear function of binder

weight percentage. Figure 6 exhibits compression stress-strain behavior for four typical

fly ash foam core samples from the same panel. As the peak stress crushes the top layers

of the samples, the stress remained almost constant and crushed hollow cenosheres upon

displacement of the compression platens. This continued till about 25% of the strain. The

results of four specimens were nearly same. Constant stress-strain response shows the

high ductility and/or energy absorbing capability of the material. A similar behavior was

observed by N. Gupta, et al. with syntactic foams made with epoxy resin and glass

hollow microspheres (12).

8
Journal of Advanced Materials, Vol. 38, No. 1, 2006

Compression
strength
MPa

Binder content %
Figure 4 Compression strength as a function of binder content of panels SA20-28

Compression
strength
MPa

Density g/cc
Figure 5 Average compression strength versus density of panels SA20-28

Compression
strength
MPa

Compression strain %
Figure 6 Compression strength versus compression strain

9
Journal of Advanced Materials, Vol. 38, No. 1, 2006

Compression behavior of the fly ash foam samples as a function of binder

material were also investigated to a limited extent. The binder material was modified by

adding about 20% by weight of polydimethylsiloxane diglycidyl ether or novolac epoxy

resin D.E.N. 431. The total binder to fly ash ratio (by weight) was maintained at 1:5.

The compression behavior of these samples is depicted in Figure 7. The epoxy additive

to the binder system at the same density does increase the compression strength. The

silicone additive also improved the compression strength of the foam samples while the

density of the sample was also increased. Relatively, the change in compression strength

with the epoxy additive is higher than that for the silicone additive. These preliminary

findings will be confirmed with a more detailed study in the future.

Compression
strength
MPa

Density g/cc
Figure 7 Effect of phenolic binder modification on compression strength of flyash foam

3.2 Tension, Shear and Fracture Toughness Tension tests were performed according to

ASTM D3039 on samples designed with a dog-bone shape to ensure failure away from

the grips. Tests were run on an MTS hydraulic system using mechanical (non-hydraulic)

wedge grips and an extensometer for axial strain measurement. A constant displacement

rate of 0.51 mm/min was used while recording load and axial strain every one half

second. Stress was calculated as load/area. The shear tests were performed according to

10
Journal of Advanced Materials, Vol. 38, No. 1, 2006

the Iosipescu, or V-notched beam method, ASTM D5379 on an MTS hydraulic system

using the standard fixture. Samples were machined to the size and shape required by the

standard. A constant displacement rate of 0.25 mm/min was used while recording load

and displacement every one half second. Some specimens were strain gaged to measure

the shear strain and calculate the modulus. Shear stress was calculated by load divided

by the cross-sectional area at the notch.

The fracture toughness tests were performed according to ASTM E399 with the

bend specimen configuration. Sample dimensions are listed in Table 2. The crack starter

notch of each sample was machined out to a width of 2.3 mm and a depth of about 10.2

mm. The crack starter notch was made using a razor blade fixture mounted in a vise.

This setup ensured that a sawing motion against the end of the starter notch resulted in a

fine crack extending from the center of the starter notch. These fatigue crack starter

notches were cut to about 1.3 mm beyond the machined notch, to a total length, a, shown

in Table 2. The tests were done using a 3-point bend fixture on an MTS hydraulic load

frame. A constant displacement rate of 0.25 mm/min was used while recording load and

center deflection every one half second.

Tension and shear data taken on fly ash foam samples are given in Table 1. The

fracture toughness data are shown in Table 2. Panels 1 and 3 had a greater variation in

tensile strength from sample to sample. This was most probably due to variation in

packing density within the panel of 300x300mm. For panels 4 and 5, the samples were

more uniform. Nevertheless overall standard deviations for tensile strength and

modulus for all samples are within 15%, which is satisfactory in this preliminary

investigation. The shear strength variations from sample to sample were about 19% and

11
Journal of Advanced Materials, Vol. 38, No. 1, 2006

the average shear strength is 4.61 MPa, which is higher for this class of materials. The

fracture toughness was determined for through crack (TC) as well as mid-plane crack

(MC) specimens. The variation in properties for the two cases is not significant. The

foam toughness is in the range of 350 to 430, which is considerably small. Therefore

further modification is being done. The compression, tension, shear and fracture

toughness data of the foam samples developed is very encouraging and they compare

favorably with other commercial core materials in the market (see Table 3).

Table 1 Tension and shear properties of foam (Fly Ash: SC-1008; 5:1)
Tension Iosipescu Shear
Panel Denstiy Specimen
Strength Modulus Poisson's Strength Modulus
No* g/cc #
MPa GPa Ratio MPa GPa
1 4.88 2.15 3.41
2 5.14 1.96 5.73
1 0.471 3 7.43 2.81 0.17 3.49
4 - - 4.80
1 6.17 3.45 4.57
2 5.09 1.97 4.38
3 0.472 3 6.76 2.83 3.79 0.80
4 6.27 2.91 5.34 1.00
5 - - 4.74
1 6.15 2.36 0.16 5.09
2 7.92 2.60 3.24
4 0.461
3 6.36 2.59 4.54
4 - - 5.99 1.10
1 6.91 2.45 4.20
2 5.98 2.17 5.80
5 0.479 3 6.39 2.33 -
4 7.90 2.76 -
5 7.56 2.80 -
Average 6.46 2.54 0.165 4.61 0.97
STD (CV%) 0.97 (15) 0.40 (15) 0.88 (19) 0.15 (15)
* panel size: 330x330x12.7 mm

12
Journal of Advanced Materials, Vol. 38, No. 1, 2006

Table 2 Fracture toughness of fly ash foam (Fly Ash: SC1008; 5:1)
Specimen Span S
Panel # Cut w mm b mm a mm Pmax N KIC kN/m3/2
# mm
3 25.6 25.2 11.5 102.5 154.7 354
1
4 25.7 25.3 11.5 103.0 164.5 373
1 25.7 25.5 11.5 103.0 165.1 370
TC
2 4 25.7 25.6 11.8 102.7 167.0 387
5 25.9 25.6 11.7 103.6 184.1 416
3 F1 25.0 25.1 12.1 102.2 152.7 410
2 25.8 25.3 11.9 103.3 151.7 358
1 3 25.6 25.3 12.1 102.6 161.4 391
4 25.8 25.4 12.1 103.4 176.2 420
MC
4 25.7 25.6 11.9 103.0 176.3 408
2
5 26.1 25.5 11.8 104.2 179.3 405
3 F1 25.4 25.1 12.2 102.2 167.9 430

Table 3 Properties of core materials

Alba Touchstone's Fly Ash
Property Balsa CL-PVC
Core Cfoam Foam
Density kg/m3 160 200 200 400 400
Compression strength MPa 12.7 4.4 3.6 13.8 15.6
Tensile strength MPa 13.5 6.4 2.1 3.4 6.5
Shear strength MPa 3.0 3.3 2.8 2.1 4.6
454
Maximum temp. °C 163 90 232 449 (Not
tested)

3.3 Evaluation of Fire Resistance Fire resistance of the fly ash foam was determined at

the NSWC Carderock Division laboratories using a cone calorimeter (ASTM E1354). A

cone calorimeter is used to determine the ignitability, heat release rate and combustion

product generation rate of a material exposed to a specified irradiance level. A 100 mm

by 100 mm sample was placed beneath the conical shaped heater that provides a uniform

irradiance on the sample surface (see Figure 8). The sample mass was constantly

monitored using a load cell, and the effluent from the sample was collected in the exhaust

hood above the heater. In the duct downstream of the hood, the flow rate, smoke

obscuration, and O2, CO2 and CO concentrations were continuously measured. A spark

igniter 12.5 mm from the sample surface was used to initiate the burning of any

13
Journal of Advanced Materials, Vol. 38, No. 1, 2006

combustible gas mixture produced by the sample. Once the sample ignites, the burning

of the sample causes a reduction in the oxygen concentration within the effluent collected

by the hood. This reduction in oxygen concentration was shown to correlate with the

heat release rate of the material, 13.1 MJ/kg of O2 consumed. This was known as the

oxygen consumption principle. Using this principle, the heat release rate per unit area of

the sample was determined with time using measurements made within the duct. The fly

ash foam samples were tested at irradiance levels of 50 and 75 kW/m2 to evaluate their

performance when exposed to different heat loads.

Figure 8 ASTM E1354 Cone calorimeter test apparatus

As a preliminary evaluation, three samples A, B, and C were tested for fire

resistance properties. Samples A and B were 25.4 mm thick whereas sample C wass 12.7

mm thick. The test data are summarized in Table 4. Sample B was heaviest with a

14
Journal of Advanced Materials, Vol. 38, No. 1, 2006

density of 530 kg/m3 while A and C had densities of 340 and 390 kg/m3, respectively.

All three samples at an irradiating heat flux of 50 kW/m2 exhibited no ignition, which is

very encouraging. These samples did not emit any of the known toxic gas except for

carbon monoxide (less than 200 ppm), which is about 1/18th of the acceptable limit. At

75 kW/m2 heat flux, the samples did ignite after exposure for 26, 35 and 59 seconds for

samples A, B and C, respectively. Additional work is in progress to further improve the

fire resistance of the fly ash foam.

Table 4 Cone calorimetry data on fly ash foam samples

5 MIN
Material/ Peak Heat Average Ave rage
Heat Time To Average Heat Total Heat
Thickness Release Heat CO
Flux Ignition of Combustion Released
(mm) Rate Release Yield
Rate
kW/m 2 sec kW/m 2 kW/m 2 MJ/kg kg/kg MJ/m 2
Panel-A/
NI 53 6 13 0.28 4
25.4
Panel-B/
50 NI 31 13 14 0.18 10
25.4
Panel-C/
NI 8 14 5 0.43 8
25.4
Panel-A/
26 47 6 5 0.35 3
25.4
Panel-B/
75 35 33 18 10 0.24 6
25.4
Panel-C/
59 59 36 26 0.12 16
25.4
NI- No Ignition

4. CONCLUSIONS

1. A process for low-cost syntactic foam made from fly ash, a waste product of coal

combustion from thermal power plants, has been developed using a resole

phenolic resin binder at a low volume percentage of about 6%.

15
Journal of Advanced Materials, Vol. 38, No. 1, 2006

2. The compression, tension, shear and fracture toughness data of the foam samples

developed were reasonably uniform, the process was repeatable, and the data

compared favorably with other commercial core material.

3. The preliminary fire resistance evaluation of the foam samples is very

encouraging, and further work will be carried out to develop foam core materials

of improved fracture toughness and fire resistance characteristics.

5. ACKNOWLEDGEMENTS

The authors wish to thank the Office of Naval Research for the financial support through

grant N00014-01-1-1033 and Dr. Yapa Rajapakse, program manager for ship structures,

for his continuing interest in this work.

6. REFERENCES

1. R. A. Malloy and J. A. Hudson, “Foamed Composites, in S. M. Lee, ed,

International Encyclopedia of Composites”, VCH Publishers, New York, 2, 257

(1990)

2. C. Hiel, D. Dietman, and O. Ishai, “Composite sandwich construction with

syntactic foam core”, Composites, 24, 447 (1993)

3. F. A. Shutov, “Syntactic polymer foams”, Adv. Polym. Sci. 73, 63 (1985)

4. R. D. Deanin, in J. C. Salamone, ed., Polymeric Materials Encyclopedia, vol. 4, p.

2554, CRC Press, New York, NY (1996)

5. F.C. Campbell, “The Case Against Honeycomb Core”, SAMPE 2004, May 16-20,

Long Beach, CA

6. U. Sorathia, “Improving the Fire Safety of Composite Materials for Naval

Applications”, SAMPE 2004, May 16-20, Long Beach, CA

16
Journal of Advanced Materials, Vol. 38, No. 1, 2006

7. N. Gupta, C. S. Karthikeyan, S. Sankaran and Kishore, Materials

Characterization, 43, 271 (1999)

8. R. A. Ruhno and B. W. Sands, in H. S. Katz and J. V. Mileski, eds., “Handbook

for Fillers in Plastics”, Kluwer Publishers, Chapter 22 , 437 (1987)

9. L. Bardella and F. Genna, International Journal of Solids and Structures, 38,

7235 (2001)

10. K. Shivakumar, R. Sadler, M. Sharpe, and S. Argade, “Fire Resistant Structural

Member”, US Patent Filed, Application No: 10/702,063 November, 2003.

11. Harry S. Katz and John V. Milewski, “Handbook of Fillers and Plastics”, Van

Nostrand Reinhold Company Inc., 1987

12. N. Gupta, E. Woldesenbet, and P. Mensah, “Compression Properties of Syntactic

Foams: Effect of Radius and Specimen Aspect Ratio”, Composites Part A:

Applied Science and Manufacturing, 35, 1, p.103-111 (2004)

17