You are on page 1of 10


Int. Comm. HeatMass Transfer, Vol. 22, No. 5, pp, 751-760, 1995
Copyright © 1995 Elsevier Science Ltd
Printed in the USA. All rights reserved
0735-1933/95 $ 9 . 5 0 + . 0 0



N.D.Kaushika and M. Arulanantham
Centre for Energy Studies
Indian Institute of Technology,
Hauz Khas, New Delhi - 110016, India

(Communicated by P.J. Heggs and A.R. Balakrishnan)
A physical model is formulated to evaluate the radiative heat transfer across
transparent honeycomb insulation materials. The model takes into account the
refraction, reflection, absorption and scattering characteristics of cell walls and
enables the computation of elemental transmittance, integral transmittance and the
geometrical shape factor. The predicted results of heat loss across a honeycomb
array exhibit by far the best agreement with the result of an earlier experiment. Thc
model is used to investigate the effect of wall thickness, aspect ratio, and
absorptivity of the wall on radiative heat transfer. The results are presented in the
form helpful for engineering design of TIM.

Transparent Insulation Materials (TIM) represent a new class of thermal insulation which
can be used for preventing heat losses as well as achieving solar energy gain in such outdoor
systems as building walls/roofs, storage tanks as well as transport pipelines o f hot fluids. TIM
consists of a transparent cellular structure immersed in an air layer. There are many material
types; four generic types describe [1 ] them according to geometry of cellular structure. The most
documented configurations include honeycombs, capillaries and parallel slats.
For optimising TIM performance, maximisation of solar transmittance and minimisation
of heat losses are the two major parameters to be considered. Heat is transferred through TIM

No. So for modelling the radiative heat transfer a single cell can be considered to be all opaque wall. conduction. 22. 5 by one or more of the following modes: conduction and radiation through the solid cellular media. The model is validated with earlier experimental results and used to investigate the effect of wall thickness. Kaushika and M. The problem has since been investigated with increasing degree of complexities and practical realism [3. In solar energy context Hollands [5] used the shape factor of Hottel and Keller [2] in the estimation of radiative heat transfer through honeycombs. Research on radiative heat transfer through passages dates back to the work of Hottel and Keller [2] who investigated it in context with the heat losses from furnaces. In each cell the radiation striking the wall is partly reflected. Along the ray reflected from the cell wall there is another ray penetrating from the adjacent cell in such a way that the energy along the reflected ray corresponds to the perfect mirror reflection of original incident ray (Fig. They considered the radiative heat transfer. The top and bottom covers are regarded as the two black end plates.4]. the net radiative heat flux a! a . Subsequently Morris et al [6] reported a method based on Monte Carlo ray tracing algorithms. As an approximation to the work of Morris et al [6]. convection and radiation across the air cell. The problem with Monte Carlo method is that there are no explicit equations which others can use directly for their materials and configurations. Hollands [7] suggested an empirical expression for the radiation shape factor of square and hexagonal honeycombs. with half the original thickness. scattering and absorption by cell walls. Following Edwards and Tobin [8]. In thc present paper an explicit model is formulated to evaluate the radiative heat transfer across transparent honeycomb insulation materials. partly transmitted and the remaining absorbed. across tubular glass honeycombs and took into account the reflection. and the total reflectance equal to the sum of the reflectivity and transmissivity and the emittance equal to the emissivity of the wall itself. The sidewalls are assumed to be the specular reflectors and diffuse emitters because of their smooth surface and minimal scattering in the long wave range. 1). The effect of honeycomb geometry on natural convection heat transfer is quite wcll understood but radiative heat transfer needs to be investigated.752 N.D. Arulanantham Vol. aspect ratio and absorptivity of the wall on the radiative heat transfer. The Model and Analysis TIM consists of a cellular array of repetitive nature. These authors considered the case of diffuse adiabatic walled reradiating passages.

aT4Xp(L-x) ..(xi = oT4xp(x) . 1 Schematic of Radiation Propagation Through TIM distance x from the lower plate is given by x q./2 ' (2) -!/2 I cOs0sin0 d0 dr where Tbx (0. 5 HEAT TRANSFER ACROSS INSULATION 753 . No.r) for square celled honeycomb .i aT4(x9 dxp(x-xOd(x_x r) dx' L + ! t~T4(xi) dxp(x/-x)d(x i-x) dx/" (I) Zp(X) is the diffuse radiation transmittance up to a height x from bottom and includes angular variation of reflection by walls. r) cos0sin0d0dr Xp(X) :- ~ .. The transmittance Tb~ (0. p a r t of ray a* T r a n s m i t t e d p a r t of ray b R e f l e c t e d par t of ray b • T r a n s m i t t e d p a r t of ray a .Vol. 22...(T2 Cc ) Reflected I Wa I I 11 \ '~. It may be expressed as hi2 n/2 -!/2 I "rbX(0..r) is the transmittance for the direct radiation with incident angle 0 and azimuth angle r at a distance x from the bottom.... (T ( x ) .. C w ) Cell o d 2a FIG.

The energy equation for the infinite length of the wall with adiabatic conditions is givcn by dx (q) = 0 .[9] and Kaushika et al.Pohlhausen integral technique. 4(Ax) = ~T~-~T~ (61 The resultant energy equation is solved by employing undetermined coefficients in a manner similar to Karman .754 N. The dimensionless radiant heat flux lbr qr is represented by the geometrical shape factor F12.x) 4(L) --"CP(x) 4(0) L --[ xp(X-x~) dx/-dx/ d4 ! ~p(x/. Arulanantham Vol. (4) Integrating by parts the above equation yields × FI2 = "t'p(X)+'t'P (C .2= qr(x) [ dxp(x -x/) idXp(X/-X) oT~-oT~ = Xp(X) ~ . so by following Hottel and Keller [2] ~ T I 4 call be taken as unity and ~T2 4 as zero and ~T4(x) can be called as 4.~ . the above value is substituted in the energy equation (3) and following dimensionless variables are introduced for simplification: x L ~T4 (Ax) oT4 Ax = ~ . A = -~. 5 and parallel slats may be calculated by following the methods of Hollands et al.x) dx d4/dx'. (3) The energy equation in this case is linear in ~T 4. Kaushika and M.[10]. No. (5) In order to determine the heat flux through the cellular array.~-~ 4 (x~dx/+ . It may be expressed as F. d(x l-x) 4(x/)dx / .D. In this method a simple expression for 4 is assumed with two undetermined coefficients 4o and A4 as 4 ( A x ) =40-l---A-x-x} A 4 (7) In order to determine the two parameters the dimensionless radiant heat flux F Lzis evaluated from . 22.

043 mm placed between two copper plates which were having the emissivity of 0. Ax = A to yield three simultaneous linear algebraic equations: the variation of optical constants with temperature is neglected. The above equation for F~2 may be extended to grey bounding plates with emissivities eh and e¢ as follows : 1 F~2g - 1 FI2 + 1 -~h _ _ Eh 1 . The apparatus measured . The resultant solution is given by doo = -~1 (1 +ADO) (8) where. Thus the evaluation of Ftz involves the calculation of transmittance xp(A) and the integral transmittance F(A). Ax = A/2.88.~c + _ (12) ~:c Results and Discussion The total heat loss (ht) across the cellular honeycomb has been investigated by several researchers both theoretically and experimentally.2Xp (A/2) Ado = 1 + xp(A) . A (ll) F(A) : I xp(Ax) dAx . 5 HEAT TRANSFER ACROSS INSULATION 755 equation (5) at Ax = 0.Xp(A/2)] AdO . Ft: where.Vol. No. Hollands [11] conducted experiments to measure the total heat loss across a honeycomb with cell size of 10 mm and wall thickness 0. The plates were maintained at the temperatures of 306 and 298 K: the difference in temperature of the plates is small enough to ensure the absence of free convection effects in the cell. 22. 1 +'Cp(A) .2"~p(A/2) + (4/A)F(A/2)-(2/A) F(A) (9) So we have (Io) = Xp(A) + [(2/A) F(A/2) .

01 W/m2K.. 22.5 2 i I I I 4 6 8 10 Aspect rotio FIG. No..2.. . .756 N... . modified the expression for Fi2 as follows 1 FI2 = e A + 3 5 -~t(W/m2 K) 4. ..8]. . ..~"~ ~~..~ 13 ..5 ~ 3 Hot~aras -2 . ... Theoretically total heat loss across the above transparent honeycomb insulation may be expressed as follows [Ac Nu Ka + A WKw] (T~ ...:L-S.. Subsequently..... Arulanantham Vol. ... .T~) L (A¢ + Aw) + CrFtzg (T L T2) h = t (13) Several formulations and computations for F~2 have been reported in the literature [2. The results are plotted as a function of aspect ratio in Fig. 5 the ht by a refined version of the guarded heater plate method within an experimental error of 0.TL-2_ .2 wherein .. 2 Total Heat Loss Across Honeycomb for Various Aspect Ratios The results of above two computations along with those obtained from the formulations of the present paper are compared with experimental observations in Fig.5 ~ .... Kaushika and M.. The first computational results of Hollands [5] were based on the formulation o f Hottel [2] who assumed the reradiating cell walls. .. .~ Measured 1 Proposed 1 /4 h t (BB) [11] ( BB ) 2 Hollcmds -I 3..D.. ~ ~ - [3 1 0..5 2 1..5 . Hollands [7] incorporated the effect of wall emissivity. 2 3 ( 13B ) [ S] ~ r-I Mectsured Proposed ~ ( B B ) [ 7] h I ( 5 B ) [11] ( SB ) 2....

3 0.7 0.6 0. 4 Effect of Wall Thickness on Shape Factor for Slat Shape factor 1.! 0 6 I I 10 15 I I I I 20 25 30 Aspect ratio 35 40 FIG.012 m m t=-023 m m ~ t=.5 0.O~Noneyco m . No.023 mm t =.012 mm t =..3i 0.~ t=.1 6 10 16 20 26 A s p e c t ro t'lo 30 36 40 FIG.0S0mm -u-.t=.5 ~ 0&~ \ ~ . 22.8 0. L b I'~\\~ ' ~\ ~ "4" A= S A: O "~ A=20 -~ A=30 o.i 0..075mm 0.2 0.0 Emissivity FIG. 5 Effect of Wall Emissivity on Shape Factor for Honeycomb 757 ..075mm 0. 3 Effect of Wall Thickness on Shape Factor for Honeycomb (Shape factor) 0.0S0 mm t =.Vol.2 0 0 0.-P -.6 ----.6 \Slot ~ ~ 0. 5 HEAT TRANSFER ACROSS INSULATION factor Honeycomb Shape 0.x*--O-- 0./.9 1.1 02 03 04 G5 0.2 0." \ ~ t =.

3 and 4 for honeycomb and slat which are made of polyester. which indicates that the amount of radiation damping by slat is always lower than the honeycomb. The geometrical shape factor decreases with both increase in wall thickness as well as increase in aspect ratio. It can be seen that the earlier two methods o f Hollands underestimate the heat losses by a factor o f about two. A detailed account of such a coupling is beyond the scope present paper and is to be investigated in future work.065) and black (~c=0. Furthermore. The geometrical shape factor as a function of emissivity of the wall material is shown in Figs.88. The shape factor for slat is always higher than the honeycomb for the same aspect ratio and thickness of wall. However. Nomenclature A aspect ratio Ac area of cell Aw area of wall D width of the TIM F(A) integral transmittance F~2 geometrical shape factor F~2~ geometrical shape factor with bounding plates h. The geometrical shape factor as a function of wall thickness for various aspect ratios is shown in Figs. From this we can observe that for IR damping one could choose the material which is very good absorber of the IR radiation or the cellular TIM with high aspect ratio since maximum of the 2rt steradian field of view is seen only by the side walls. 5 BB corresponds to black black bounding plates with emissivity of 0. the results of calculation of the proposed model exhibit deviations from the experimental values for the case o f selective and black bounding plates. 5 for honeycomb in the aspect ratio range of 5 to 40. SB corresponds to selective (~h=0. Kaushika and M. Arulanantham Vol. 22.88) bounding plates. convenient means of determining F~2 as a function of absorptivity of the material and aspect ratio of cell. No. whereas the proposed model exhibits an excellent agreement in case of black black bounding plates.758 N. Figure 5 provides -. This could be due to the appreciable coupling of conductive and radiative modes of heat transfer in SB situation. total heat loss coefficient .D. which in turn makes them a useful tool for engineering design of TIM.

Mech. Solar Energy. James. Horigomo et al. J.D. 4. K. 22. Louis. 76-HT-48 (August 1976).). Missouri.K.K. Hollands. Presented at the Nat. 1973. Progress in Astronautics and Aeronautics. 31. 5 HEAT TRANSFER ACROSS INSULATION Ka thermal conductivity of air Kw thermal conductivity of wall L height of the TIM Nu Nusselt number q radiant heat flux r azimuth angle T~ direct radiation transmittance Ti lower plate temperature T2 upper plate temperature X position coordinate (vertically upwards) 759 Greek Symbols 0 angle of incidence diffuse radiation transmittance 'rp dimensionless emissive power ~0.R. Trans.G. . 3.475. ASME paper No. Rommel. 1990. Wittwer. Platzer and M.A. Soc. AIAA. Radiative transfer through thin walled glass honeycombs. Heat Transf. 5. R. D.C. H. 3.C. 1970. Edwards and H. Amar and D. Felland.a¢ undetermined coefficients (Karman-Pohlhausen integral) ~h emissivity of lower plate Ec emissivity of upper plate emissivity of wall Stefan-Boltzmann constant References V. Am. 1933. W.C. 55. 143. AIAA. Conf. 1965. 2247.Vol. J. No. Edwards and I. Morris.K. P. Progress in Astronautics and Aeronautics. Pergamon Press. Engrs.. McMurrin. 159. Bertak. 2. St.K. 6.T. Buchberg. D. 9. Edwards. Keller. 39. R. Hottel and J. Clean and safe Energy Forever (Edited by T. 24.

1991. Edwards and R. Energy Convers. No. Pergamon Press.T. Advanced nonconcentrating solar collectors.231. Russell and R. 2119. ASME J.G.N. 8. G. 11. Heat Transfer. (Edited by A. Kaushika and M.Wilkinsun. Hollands. N. 1978. Marshall and R. p. 345. 1995 . Oxford. 32. An Introductory course. Int. 89.K. K.760 7. Leslie). Hollands.D. 162. F. D. N. 1984.G.D.E. Heat Mass Transfer 27.Wedel. K.T. 1979. Kaushika and R. 132. J.D.K. Tobin. Dixon and J. 21. Received March 29.B.Padmapriya. Solar Energy. Raithby.D.T. 22. 1967. K. 9. 10. 5 K. Mgmt.G. Arulanantham Vol.D.G. Trans.Hollands. Solar Energy Conversion.