See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/266374952

Variation of the Thermal Conductivity of Elastomeric Foam With Pressure

Conference Paper · August 2008

DOI: 10.1115/HT2008-56227

CITATIONS READS
2 557

2 authors:

Erik Bardy Joseph Mollendorf

Grove City College University at Buffalo, The State University of New York
29 PUBLICATIONS   181 CITATIONS    84 PUBLICATIONS   1,973 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

ElectroHydroDynamic View project

Electrohydrodynamics View project

All content following this page was uploaded by Erik Bardy on 15 April 2015.

The user has requested enhancement of the downloaded file.

 2008 ASME Summer Heat Transfer Conference
HT2008
August 10-14, 2008, Jacksonville, Florida USA

HT2008-56227

VARIATION OF THE THERMAL CONDUCTIVITY OF ELASTOMERIC FOAM WITH PRESSURE

Erik R Bardy, Member ASME Joseph C Mollendorf, Fellow ASME
Department of Mechanical Engineering Center for Research and Education in Special Environments
Grove City College Department of Mechanical and Aerospace Engineering
Department of Physiology and Biophysics
The State University of New York at Buffalo

ABSTRACT INTRODUCTION
In a previous study, the authors used well-known upper Modeling the change of the effective thermal conductivity
and lower bounds to develop the form of an empirical (further referenced as k*) of elastomeric foam under
correlation for the thermal conductivity of foam insulation applied hyperbaric pressure is useful in predicting the
as a function of ambient pressure. The correlation was performance of closed cell insulations in high pressure
in terms of three constants which were determined by environments (e.g. foam neoprene used in wetsuit
performing a nonlinear regression on experimentally insulation). In order to model k* of composite foam
measured thermal conductivity values of foam neoprene using theoretical correlations, the thermal conductivity of
insulation at varying ambient pressure. In this previous the constituents and knowledge of the shape of the gas
work, the thermal conductivity of neoprene rubber was cells needs to be known [1-25]. Elastomeric composite
determined using the three constants, one of which is foam, such as foam neoprene, has an elastomeric
the reciprocal of the thermal conductivity of air. In the rubber as a constituent. When these elastomeric closed
present paper, we show that the three correlation cell composite foams are placed under increased
constants can, alternately, be determined by using hyperbaric pressure both the volume fraction of gas
values of the constituent thermal conductivities (e.g. air decreases and the shape of the gas cells change [26].
and rubber), and the effective thermal conductivity at The decrease in gas volume fraction can be modeled
one pressure only. Previously reported values of the using the ideal gas law [26], but the gas cell shape
measured effective thermal conductivity of foam change with applied hyperbaric pressure is not easily
neoprene insulation under applied pressure, up to 1.18 predictable. In addition, at atmospheric pressure, the
MPa, were found to have a maximum difference of about gas cell shapes are not homogenous [26].
14% of the values predicted using the correlation. It was In a previous study by Bardy et al. [26], an
also found that the accuracy of the correlation did not empirical correlation was developed for k* of foam
depend strongly on the reference pressure used. insulation as a function of increasing hyperbaric
pressure as shown in equation (1).
KEY WORDS: Elastomeric, empirical correlation, foam
insulation, pressure, effective thermal conductivity
1 + a⎛⎜ P ⎞⎟
NOMENCLATURE k (P ) =
* ⎝ Pa ⎠ (1)
a, b, c = constants used in equation (1)
b + c⎛⎜ P ⎞⎟
kg = thermal conductivity of gas constituent (independent ⎝ Pa ⎠
of pressure)
kr = thermal conductivity of rubber constituent Equation (1) is independent of gas cell shape and
(independent of pressure) represents the functional form of an upper and lower
k0 = reference point effective thermal conductivity bound used to estimate k* of composite foams [6, 16,
k* = effective thermal conductivity 21, 23, 27-29] when the volume fraction of gas is
P = ambient pressure assumed to change as a function of pressure (P)
Pa = atmospheric pressure according to the ideal gas law. The upper bound was
P0 = reference pressure formulated assuming the gas and rubber are arranged
r, R = constants used in equation (2) thermally in parallel and the lower bound assuming they

1 Copyright © 2008 by ASME

are arranged thermally in series. It was noted that Bardy et al. [26] and Monji et al. [30]. It is noted that the
although the individual constants (a, b and c) in equation trends in their data are monotonic, as expected.
(1) are not the same as the constants in the upper and Norton and Chan [31] reported thermal
lower bounds, they yield the same asymptotic results as conductance and compressive strain values of foam
P approaches zero and infinity. Namely, that as P neoprene at increasing pressure stops to a pressure of
approaches 0, k* approaches 1/b, or kg, and likewise, as 1.52 MPa (140 msw, meters of sea water) which were
P approaches infinity, k* approaches a/c, or kr. Note that used to calculate k*. These values were found to be
kr and kg are typically independent of pressure variation. non-monotonic (“S” shaped) with increasing pressure,
In addition, equation (1) closely represents the functional and were therefore not used for comparison with
form of theories used to predict k* of composite foams equation (2). The reason(s) for the differences in trends
that assume geometric shapes for the gas cells [1-25]. are not clear; however, the non-monotonic trends are
In the previous work by the authors [26] equation (1) unusual.
was non-linearly regressed with experimentally West [32] reported thermal conductivity, thermal
measured k* values of foam neoprene at incrementally conductance and compressive strain values of foam
increasing pressure points to extract kr (= a/c) to permit neoprene to pressures of 0.37 MPa (27 msw). An
comparison of measured values to theory. The attempt was made to regress equation (1) with the
disadvantage of equation (1) in its present form is that it experimental k* values at increasing pressure stops
can only be used in conjunction with experimentally measured by West [32] to determine kr; however, the
measured k* values at several different pressure points. resulting kr values were anomalous when compared to
The purpose of the present paper is to recast other values found in the literature. Unfortunately it was
equation (1) in a form that can be used to predict k* of not possible to completely understand the reason for
composite elastomeric foams as a function of pressure if these anomalous results. It is noted, however, that the
kg, kr are known along with k* at only one pressure point effect of the density extrema on natural convection flows
(k0 at P0). The advantage of the new form is that k* can in cold water have not been accounted for [33].
be predicted at other pressure points independent of gas Bardy et al. [26] report k* values of 5 mm and 12
cell shape thereby avoiding multiple experimental mm thick samples of foam neoprene at incremental
measurements. The validity of equation (1) as a model pressure stops to 1.18 MPa (107 msw). Monji et al. [30]
was verified by comparison to previously published report k* values of two 5 mm thick samples (further
experimentally measured k* values of foam neoprene noted as 1st 5 mm thick sample and 2nd 5 mm thick
rubber as a function of increasing pressure [26, 30]. sample) and one 8 mm thick sample of foam neoprene
at incremental pressure stops to 0.51 MPa (40 msw).
ANALYSIS
To compare equation (2) with experimental data it
Equation (1) can be algebraically manipulated to be in
was necessary to have a value of kr, as was done by
the form shown in equation (2).
Bardy et al [26]. Accordingly, equation (1) was
r + R⎛⎜ P ⎞⎟
regressed with the k* values reported by Monji et al. [30]
k * (P ) ⎝ P0 ⎠ (assuming kg = 0.026 W/m-K). The regression yielded kr
= (2) values of 0.063, 0.070 and 0.076 W/m-K. These
kr 1 + R⎛⎜ P ⎞⎟ numbers fall slightly below the range of values found in
⎝ P0 ⎠ the literature (kr = 0.100 – 0.192 W/m-K, [30, 31, 34-37]).
The kr values determined by Bardy et al. [26] were found
kg
where r= (3)
to be 0.112 and 0.144 W/m-K for the 5 mm and 12 mm
kr thick samples, respectively.
In the results reported by Bardy et al. [26] a total of
k0 − k g 9 incremental pressure stops were made (including
and R= (4) atmospheric pressure). In this study, six of the nine
kr − k0 pressure stops are used for comparison. Monji et al.
Equation (2) amounts to an empirical correlation as a [30] reported a total of 5 pressure stops (also including
predictive model of k* as a function of increasing atmospheric pressure). All experimental k* values were
pressure when kr, kg and k0 known, and was used for normalized by kr for comparison to equation (2).
direct comparison to experimental data. The value of k0 was set equal to k* measured at
The effective thermal conductivity (k*) of foam the reference pressure. Once k0 was determined,
neoprene as a function of increasing pressure has been equation (2) was used to predict k*(P)/kr for all remaining
reported in previous studies, e.g. Bardy et al. [26], Monji pressure stops. Equation (2) was compared to
et al. [30], Norton and Chan [31], and West [32]. In the experimental data at each pressure stop for each data
present paper, comparison of equation (2) to set.
experimental data was limited to k* values reported by

2
RESULTS AND DISCUSSION equation (2) to pressures of 0.51 MPa (40 msw). The
Table 1 shows the percent difference between k* values highest percent differences occur when the k0 is chosen
determined by equation (2) compared to experimentally at P0 = 0.20 and 0.30 MPa for the 8 mm thick sample,
measured values at 6 pressure stops (6 different values and when k0 is chosen at P0 = 0.20 MPa for the 1st 5 mm
for k0) by Bardy et al. [26]. As can been seen there was thick sample. Figure 3 through Figure 5 show plots of
less than a 14% difference between values of k* k*(P)/kr values measured by Monji et al. [30] compared
determined by equation (2) and experimentally to values predicted from equation (2) using k0 at a
measured values to pressures of 1.18 MPa (107 msw). pressure of 0.30 and 0.51 MPa for 8 mm and two
The largest difference in predicted and measured values samples of 5 mm thick foam neoprene, respectively.
occurs when k0 was chosen at atmospheric pressure. The deviation from predicted values shown in Table 2
Figure 1 and Figure 2 shows a plot of k*(P)/kr measured can be seen in Figure 3 (k* at 0.20 and 0.30 MPa for the
by Bardy et al. [26] compared to values predicted from 8 mm thick sample) and Figure 4 (k* at 0.2 MPa for the
equation (2) using k0 at a pressure of 0.10 and 1.18 MPa 1st 5 mm thick sample). Figure 5 shows deviations from
for 5 mm and 12 mm thick foam neoprene, respectively.. the predicted values of less than 4% for the 2nd 5 mm
As can be seen, when k0 is selected at atmospheric thick sample. Although there are slight deviations from
pressure (P0 = 0.10 MPa) all predicted k*(P)/kr values predicted k* values present for all foam neoprene
were somewhat higher than the measured values. The samples, they are not significant.
percent difference was approximately 12.6% and 13.6%
for 5 mm and 12 mm thick foam neoprene, respectively,
at 0.25 MPa. As pressure increases this difference
decreases to 3.6% and 3.9% for 5 mm and 12 mm foam
neoprene, respectively, at 1.18 MPa. Likewise when k0
is selected at P0 > 0.10 MPa all predicted values of k* for
P0 > 0.10 MPa are less than 5% different than measured
values for both the 5 mm and 12 mm thick foam
neoprene. The percent difference at atmospheric
pressure (for k0 selected at P0 > 0.10 MPa) is between
8% and 12%. The reason for these differences could be
explained in terms of the experimental apparatus and
protocol. The thermal conductivity was measured using
a thermal meter that measured and controlled heat flux
(and thickness) across a test specimen placed between Figure 1 Experimentally measured k* values of 5 mm
two flat plates controlled at different temperatures (Anter thick foam neoprene from Bardy et al. [26] compared to
Corp, model QuicklineTM 16, in accordance to ASTM predictions of k* from equation (2) (*MPa, **W/m-K,
C518 and ISO DIS 8301). When measuring k*, the ***unitless).
thermal meter needed to reach equilibrium before a
reading can be taken. To ensure that a proper value of
k* was measured, values of k* were recorded every 2
minutes until k* did not vary more than 1% within 10
minutes. It was assumed at this point that equilibrium
was reached. The first recorded value of k* (at
atmospheric pressure) was measured after 12 hours had
passed. All other values of k* were recorded within 45
minutes of reaching the designated pressure. Even
though k* did not vary more than 1% after 45 minutes,
differences in trend between k* values at atmospheric
and elevated pressures could be attributed to residual
transients not detectible in a 10 minute-period.
Table 2 shows the percent difference between k*
values determined by equation (2) compared to
experimentally measured values of one sample of 8 mm Figure 2 Experimentally measured k* values of 12 mm
thick and two samples of 5 mm thick foam neoprene at thick foam neoprene from Bardy et al. [26] compared to
various pressure stops by Monji et al. [30]. There were a predictions of k* from equation (2) (*MPa, **W/m-K,
total of 5 reference pressures (5 values of k0). There ***unitless).
was less than a 9% difference between experimentally
measured values of k* and values predicted from

3
Table 1 Percent differences of k* values determined by equation (2) at various pressure stops compared to
measured k* values of 5 mm and 12 mm thick foam neoprene as measured by Bardy et al. [26] using various
reference pressures.
P0 (MPa) / k0 (W/m-K)
Ambient 5 mm thick foam neoprene
Pressure P0 = 0.10 0.25 0.41 0.56 0.72 1.18
(MPa) k0 = 0.0518 0.0627 0.0735 0.081 0.0857 0.0942
0.10 0.0% 11.9% 10.9% 9.6% 9.4% 8.6%
0.25 12.6% 0.0% 1.1% 2.6% 2.8% 3.7%
0.41 9.5% 1.0% 0.0% 1.3% 1.4% 2.2%
0.56 6.8% 1.9% 1.1% 0.0% 0.1% 0.8%
0.72 5.8% 1.8% 1.1% 0.1% 0.0% 0.6%
1.18 3.6% 1.7% 1.2% 0.5% 0.4% 0.0%
P0 (MPa) / k0 (W/m-K)
Ambient 12 mm thick foam neoprene
Pressure P0 = 0.10 0.25 0.41 0.56 0.72 1.18
(MPa) k0 = 0.0518 0.0710 0.0860 0.0947 0.1020 0.1160
0.10 0.0% 12.3% 10.8% 11.6% 10.9% 8.4%
0.25 13.6% 0.0% 1.8% 0.9% 1.6% 4.5%
0.41 9.9% 1.5% 0.0% 0.8% 0.2% 2.3%
0.56 9.2% 0.7% 0.7% 0.0% 0.6% 2.7%
0.72 7.5% 1.1% 0.1% 0.5% 0.0% 1.9%
1.18 3.9% 2.2% 1.3% 1.7% 1.4% 0.0%

Table 2 Percent differences of k* values determined by equation (2) at various pressure stops compared to
measured k* values of 5 mm and 8 mm thick foam neoprene as measured by Monji et al. [30] using various
reference pressures.
P0 (MPa) / k0 (W/m-K)
Ambient 8 mm thick foam neoprene
Pressure P0 = 0.10 0.20 0.30 0.41 0.51
(MPa) k0 = 0.048 0.051 0.053 0.058 0.059
0.10 0.0% 7.6% 8.8% 0.2% 2.6%
0.20 6.1% 0.0% 1.0% 6.2% 8.0%
0.30 5.5% 0.8% 0.0% 5.6% 7.0%
0.41 0.1% 3.7% 4.4% 0.0% 1.0%
0.51 0.9% 4.0% 4.5% 0.8% 0.0%
P0 (MPa) / k0 (W/m-K)
Ambient 1st 5 mm thick foam neoprene
Pressure P0 = 0.10 0.20 0.30 0.41 0.51
(MPa) k0 = 0.051 0.055 0.059 0.063 0.064
0.10 0.0% 7.7% 5.7% 0.5% 0.0%
0.20 6.1% 0.0% 1.7% 6.5% 6.1%
0.30 3.4% 1.3% 0.0% 3.6% 3.4%
0.41 0.2% 3.9% 2.9% 0.0% 0.2%
0.51 0.0% 3.1% 2.2% 0.2% 0.0%
P0 (MPa) / k0 (W/m-K)
Ambient 2nd 5 mm thick foam neoprene
Pressure P0 = 0.10 0.20 0.30 0.41 0.51
(MPa) k0 = 0.053 0.059 0.063 0.067 0.069
0.10 0.0% 4.5% 4.9% 0.2% 0.6%
0.20 3.4% 0.0% 0.4% 3.6% 3.9%
0.30 3.0% 0.3% 0.0% 3.1% 3.3%
0.41 0.1% 2.2% 2.5% 0.0% 0.2%
0.51 0.2% 2.0% 2.2% 0.1% 0.0%

4
 CONCLUSION
An empirical correlation was presented to theoretically
predict the effective thermal conductivity (k*) of
elastomeric foam, independent of gas cell shape, as a
function of increasing ambient pressure. The advantage
of this correlation is that it allows the prediction of k* of
elastomeric foam with increasing hyperbaric pressure,
thereby avoiding experimental measurement. Use of the
empirical correlation required values for the thermal
conductivity of the pure rubber constituent (kr), the
thermal conductivity of the gas constituent (kg) and the
k* at one reference pressure (k0 at P0). With these three
values known, k* can be predicted for pressures above
and below P0. The predicted k* values were shown to
Figure 3 Experimentally measured k* values of 8 mm thick be within 14% when compared to experimentally
foam neoprene from Monji et al. [30] compared to measured k* from Bardy et al. [26] and within 9% when
predictions of k* from equation (2) (*MPa, **W/m-K, compared to Monji et al. [30]. Therefore, it is concluded
***unitless). that if kr and kg of an elastomeric foam are known, as
well as k* at one pressure point (k0 at P0), k* can be
predicted at pressures greater and less than P0 with
reasonable accuracy.
ACKNOWLEDGEMENTS
This research was supported by the Office of Naval
Research (ONR grant N00014-01-0278).
REFERENCES
1. Ashton, J., J. Halpin, and P. Petit, Primer on
composite materials; analysis. 1969: Technomic
Pub. Co., Stamford, Conn.
2. Behrens, E., Thermal conductivities of
composite materials. Journal of Composite
Materials, 1968. 2(1): p. 2-17.
3. Boomsma, K. and D. Poulikakos, On the
Figure 4 Experimentally measured k* values of the 1st 5
effective thermal conductivity of a three-
mm thick foam neoprene from Monji et al. [30] compared
dimensionally structured fluid-saturated metal
to predictions of k* from equation (2) (*MPa, **W/m-K,
foam. International Journal of Heat & Mass
***unitless).
Transfer, 2001. 44: p. 827-836.
4. Bruggeman, D., Dielectric constant and
conductivity of mixtures of isotope materials.
Ann Phys, 1935. 24: p. 636.
5. Budiansky, B., Thermal and thermoelastic
properties of isotropic composites. J Composite
Materials, 1970. 4: p. 286-295.
6. Calmidi, R. and V. Mahajan, The effective
thermal conductivity of high porosity fibrous
metal foams. Journal of Heat Transfer, 1999.
121: p. 466-471.
7. Cheng, S. and R. Vachon, A technique for
predicting the thermal conductivity of
suspensions, emulsions and porous materials.
International Journal of Heat & Mass Transfer,
Figure 5 Experimentally measured k* values of the 2nd 5 1970. 13: p. 537-546.
mm thick foam neoprene from Monji et al. [30] compared 8. Donea, J., Thermal conductivities based on
to predictions of k* from equation (2) (*MPa, **W/m-K, variational principles. Journal of Composite
***unitless). Materials, 1972. 6: p. 263-266.

5
9. Felske, J.D., Effective thermal conductivity of foam neoprene insulation under hydrostatic
composite spheres in a continuous medium with pressure. Journal of Physics D-Applied Physics,
contact resistance. International Journal of Heat & 2005. 38(20): p. 3832-3840.
Mass Transfer, 2004. 47(14-16): p. 3453-3461. 27. Brailsford, A. and K. Major, The thermal
10. Fricke, H., A mathematical treatment of the elastic conductivity of aggregates of several phases,
conductivity and capacity of disperse systems. including porous materials. Brit. J. Appl. Phys.,
Physical Review, 1924. 24: p. 575-587. 1964. 15: p. 313-319.
11. Hamilton, R. and O. Crosser, Thermal conductivity 28. Carson, J.K., et al., Thermal conductivity bounds
of heterogeneous two-component systems. Ind. for isotropic, porous materials. International
Eng. Chem. Fund., 1962. 1: p. 187. Journal of Heat & Mass Transfer, 2005. 48: p.
12. Hashin, Z., Assessment of the self consistent 2150-2158.
scheme approximation: conductivity of particulate 29. Maqsood, A., K. Kamran, and I.H. Gul,
composites. J. Composite Materials, 1968. 2(3): p. Prediction of thermal conductivity of granite
284-300. rocks from porosity and density data at normal
13. Jefferson, J.B., O.W. Witzell, and W.L. Sibitt, temperature and pressure: in situ thermal
Thermal conductivity of graphite-silicone oil and conductivity measurements. J. Phys. D: appl.
graphite-water suspensions. Ind. Eng. Chem., Phys., 2004. 37: p. 3396-3401.
1958. 50: p. 1589. 30. Monji, K., et al., Changes in insulation of
14. Kerner, E.H., The electrical conductivity of wetsuits during repetitive exposure to pressure.
composite media. Proc. Phys. Soc. B, 1956. 69: p. Undersea Biomedical Research, 1989. 16(4): p.
802. 313-319.
15. Landauer, R., The electrical resistance of binary 31. Norton, M.P. and C.Y.L. Chan, Insulation
metallic mixtures. J Appl. Phys., 1952. 23(7): p. properties of composite cell foamed materials
779-784. suitable for wet suits. Applied Energy, 1982. 12:
16. Leach, A., The thermal conductivity of foams. I: p. 159-176.
models for heat conduction. J. Phys. D: appl. 32. West, P.B., Empirical evaluation of diving wet
Phys., 1992. 26: p. 733-739. suit material heat transfer and thermal
17. Maxwell, J.C., A treatise on electricity and conductivity. Heat Transfer Engineering, 1993.
magnetism. 3 ed. Vol. 1. 1891: Oxford university 14(4): p. 74-80.
press. 33. Gebhart, B. and J. Mollendorf, Buoyancy -
18. Nielsen, L., Thermal conductivity of particulate- Induced Flows in a Liquid Under Conditions in
filled polymers. Journal of applied polymer Which Density Extrema May Arise. Journal of
science, 1973. 17: p. 3819-3820. Fluid Mechanics, 1978. 89: p. 673-707.
19. Oka, S. and K. Yamane, Theory of the Conduction 34. Brandrup, J., E.H. Immergut, and E.A. Grulke,
of Heat in Foamed Plastics. Japanese Journal of Polymer Handbook 4th Edition. 1999.
Applied Physics, 1967. 6(4): p. 469-474. 35. Gibson, L.J. and M.F. Ashby, Cellular Solids. 2nd
20. Peterson, J.M. and J. Hermans, The dielectric ed. 1997, Cambridge, UK: Cambridge University
constants of nonconducting suspensions. J Press.
Composite Materials, 1969. 3: p. 339-354. 36. Norton, M.P., Theoretical considerations
21. Progelhof, R.C., J.L. Throne, and R.R. Ruetsch, concerning the prediction of diver wet suit
Methods for predicting the thermal conductivity of insulation behavior with and without pressure
composite systems: a review. Polymer compensation. Applied Energy, 1981. 9: p. 85-
Engineering and Science, 1976. 16(9): p. 615-625. 105.
22. Russell, H.W., Principles of heat flow in porous 37. Ohsawa, T., et al., A study of composite foams
insulation. J. Am. Ceram. Soc., 1935. 18: p. 1. for diving suits subjected to high hydrostatic
23. Springer, G. and S. Tsai, Thermal conductivities of pressure. Journal of Applied Polymer Science,
unidirectional materials. J Composite Materials, 1979. 23: p. 1233-1245.
1967. 1: p. 167-173.
24. Topper, L., Analysis of porous thermal insulating
materials. Industrial and Engineering Chemistry,
1955. 47(7): p. 1377-1379.
25. Tsao, G., Thermal conductivity of two phase
materials. Ind. Eng. Chem., 1961. 53: p. 395.
26. Bardy, E., J. Mollendorf, and D. Pendergast,
Thermal conductivity and compressive strain of

View publication stats