ENGINEERING RESEARCH INSTITUTE THE UNIVERSITY OF MICHIGAN ANN ARBOR LITERATURE SURVEY OF POROUS AND FIBROUS LOW-TEMPERATURE INSULATION J.' T. anchero J. R. Sinek Project 2254 THE ARO EQUIPMENT CORPORATION BRYAN, OHIO August 1957

An,~IS

The University of Michigan ~ Engineering Research Institute TABLE OF CONTENTS Page ABSTRACT ii OBJECTIVE iii INTRODUCTION PART I. ANALYSIS OF LITERATURE DATA AND RECOMMENDATIONS 2 Theory of Heat Transfer Through Porous and Fibrous Insulation 2 Conduction 2 Radiation 3 Overall Heat Transfer 3 Physical Properties of Porous and Fibrous Insulation Under Vacuum 3 General 3 Properties of Santocel 5 Recommendations Resulting from Literature Survey 6 PART II. ANNOTATED BIBLIGRAPHY 8 A. Literature References Other Than Patents 8 B. Patent Literature 15 U. S. Patents 15 British Patent 16 French Patent 16 German Patents 16

The University of Michigan ~ Engineering Research institute ABSTRACT The literature on porous and fibrous insulations suitable for use in liquefied gas containers has been reviewed and an annotated bibliography is presented. OBJECTIVE The primary objectives of the literature were to determine the insulating materials that have been used or proposed, and the information available on the effect of certain variables on rates of heat transfer through layers of granular insulation.

The University of Michigan ~ Engineering Research Institute INTRODUCT ION The theoretical analysis of rates of heat transfer to the inner container of the Aro 5-liter oxygen container presented in a previous report* showed that the major portion of the heat leak occurred by radiatiban. - It was suggeste in that report that the use of granular insulation in the annular space between the inner and outer shells of the container be considered and a number of references were cited in support of this view. Since only a cursory examination of the literature had been made at that time, it was decided to prepare as complete a literature survey as possible of published information on lowtemperature insulation, with special emphasis on finely-divided solid materials for use in a space under vacuum. The primary objectives of the survey were to determine the insulating materials that have been used or proposed and the information available on the effect of the following variables on rates of heat transfer through layers of granular insulation~ 1. Pressure on the insulated space. 2 Particle size and orientation. 3. Gas initially present in the insulated space,., Thickness of the insulated space. 5. Temperature of the retaining walls. 6. Addition of powdered metals (such as silicon) to the insulation. 7. Materials and surface treatment of the containing walls. The results of the literature survey are presented in two parts. Part I includes a discussion of the theory of heat transfer through porous insulation, summarizes the pertinent information found, and presents recommendations. Part II is an annotated bibliography of the literature on which Part I is based. For convenience, all references except patents are given in Part II-A and are arranged in alphabetical order according to author. Patents are listed in Part II-B and are presented in chronological order for each country. United States patents are listed first, followed by British, French, and German patents *Banchero, J. T., Calvert, So Elzinga, Eo Ro,, and Sinek, Jo Ro, The University of Michigan Engineering Research Institute Report 2254-6-P, November, 1956.

The University of Michigan ~ Engineering Research Institute PART I. ANALYS IS OF LITERATURE DATA AND RECOMMENDATI ONS THEORY OF HEAT TRANSFER THROUGH POROUS AND FIBROUS INSULATION Heat transfer through porous and fibrous insulation is due mainly to conduction through the gas, conduction through the solid, and radiation. Convection is a negligible factor in this type of insulation0 Conduction. —From kinetic gas theory, the thermal conductivity k of a gas is proportional to the absolute pressure P of the gas. It is also proportional to the length of the mean path its molecules actually traverse, this is the mean free path L of the gas, or the distance D between the solid barriers confining the gas, whichever is smaller:o k ac PL (L/D z 1) (1) k oc PD (L/D E 1) (2) Equation (1) may be simplified, since the mean free path L is inversely proportional to P. Equation (2) may also be simplified, since D is constant for a given piece of equipment. k = const (L/D _ 1) (1A) k cx P (L/D - 1) (2A) Usually,the mean free path L is smaller than D at atmospheric pressure. Upon gradually lowering the pressure of a body of gas, thereby increasing its mean free path L, the conductivity remains constant (Eq. 1A) until L becomes equal to D. From this transition point on, a decrease in pressure brings with it a proportional decrease in conductivity (Eq. 2A)o In a Dewar vessel, D is the inter-wall distance, and generally varies from 5 to -0 cm in4commercial vessels. The transition pressure (at which L = D) is 10 to 10 mm Hg, and th pressure at which the conductivity is 10% of its normal value is therefore 10 to 10-5 mm Hg.2 Such pressures are attained in a Dewar vessel. The chief advantage of porous and fibrous insulators is the fact that here D, the distance between the solid barriers confining the gas, is no longer the inter-wall distance but the mean pore diameter, and is therefore much smaller.

The University of Michigan ~ Engineering Research Institute Consequently, the transition pressure is. much higher than in a Dewar vessel ( and therefore easier to attain), and is also independent of the size of the vessel. Although porous or fibrous insulating materials are poor thermal conductors, solid conduction through points of contact between granules or fibers is unavoidable. It varies with the kind of material used, especially in terms of the density of the material, and in fibrous material it also varies according to the directional arrangement of the fibers. Solid conduction generally accounts for only a small fraction of the total heat transfer. Radiation.-At low pressures where gas conduction is relatively small, radiative heat transfer contributes considerably towards the total heat transfer. In a Dewar vessel, it is a function of the reflectivity of the surfaces with respect to thermal radiation. A third reflecting wall cuts down radiation considerably. A layer of porous or fibrous insulating material in the interwall space also reduces radiation. A recent investigation28 showed that this effect may be calculated by determining the absorption and scattering parameters of the material. It has also been correlated for powders by a fourth14 power absolute temperature term multiplied by a Beer's-law extinction term.14 Overall Heat Transfer.-The overall heat transfer through porous or fibrous insulation is generally expressed as an apparent conductivity, applying Fourier's law to the overall heat transfer rate. The only inconvenience of this practice is that the radiative contribution to overall heat transfer does not vary with insulation thickness according to Fourier's law. This should be kept in mind when using the apparent overall conductivity. This conductivity has been correlated in the literaturel, 21,24,42, 44, 46 as a function of the properties and pore dimensions of the material and of the gas properties. PHYSICAL PROPERTIES OF POROUS AND FIBROUS INSULATION UNDER VACUUM General.-The porous and fibrous insulations whose conductivity under vacuum has been investigated are listed in Table I, with the temperatures and pressures at which they were tested. In general, the experimental results correspond to theory in that, below a critical pressure, conductivity decreases linearly with pressure. When very low pressures are reached, overall conductivity levels out to an approximately constant value, since the only remaining heat transfer is substantially by solid conduction and by radiation. The most decisive property of porous insulating material seems to be the pore diameter. For porous material of Uniform particle size, there seems to be no appreciable effect when the particle size is varied from 4/10 mesh to 100/200 mesh.48 However, noticeable differences can be obtained by mixing small particles with larger particles.48 For fibrous insulation, orientation 5~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

The University of Michigan * Engineering Research Institute TABLE I INSULATING MATERIALS USED UNDER VACUUM Material' Temperature Range, Pressure Range Cef. Material Hot Side Cold Side Silox powder in air, C02, H2 R.T.* 1 mm-Atm 1 Santocel 14.5 -196 2 x 10-6 mm-Atm 5 Santocel R.T. -183 10-2 mm 10 Mipora R.T. -196 10-2 mm-Atm 13 Celite 150 10 3 mm —Atm 15 Santocel R.T. -183 10-4 mm 17 Santocel Fiberglas Rock wool R.T. -253 10- mm-Atm 18 Kapok Charcoal Cork Santocel Slag wool X ~~~Slag wool20 -183 10-3 mm —1000 mm 22 Basic Mg carbonate Brelite Santocel R.T. 10-4 mm-Atm 24 Santocel in air, C02, Freon-12 R.T. 2 mm-Atm 25 Glass wool in air, C02, Freon-12 65 - 60 10- mm-Atm 27 Perlite 10-5 mm Charcoal 2 x l0-3 mm-10-1 mm Glass wool RoT. -197 2 x 10-3 mm-10-1 mm Plastic microballoons 2 x l0-3 mm-10-1 mm Santocel 2 x 10-3 m Santocel AO 2 x 10-3 mm C orkboard Cotton Glass wool Hair felt 52 24 l0-4 mm-Atm 36 Kapok Rock wool Santocel Wood fiber *R.T. - room temperature

The University of Michigan ~ Engineering Research Institute and fiber density have an effect on heat transfer, but data have been reported mainly at atmospheric pressure. There has been relatively little research on the influence of different gases on the conductivity of porous and fibrous insulation under vacuum. Besides gaseous heat transfer, adsorption plays an important part. Carbon dioxide has a 20-30% lower conductivity than air when used with porous and fibrous insulators at atmospheric pressure and between 25 and 500C, and Freon-12 has a 30-50% lower conductivity than air under similar conditionsas reported by one author. 27 The effect of thickness of the insulation depends on whether heat transfer is mainly conductive or radiant under the given conditions. In conduction, heat transfer is inversely proportional to thickness. Where radiation is the main factor, as when the pressure is very low and the temperature drop large (low-temperature insulation), heat transfer is independent of thickness unless the insulation layer is so thin that it is no longer opaque to thermal radiation. Some porous and fibrous material may be rendered more opaque to thermal radiation by adding powdered silicon, aluminum, or graphite. 8 The addition of 10% silicon to Santocel reduces conductivity by 10-20%.35 The reflectivity of the walls does not significantly reduce heat transfer when porous or fibrous insulation is used.5 Examination of the data in the literature (see Table I) shows that Santocel has the most outstanding properties of the insulating materials f5r which data have been reported. Consequently, a more detailed account of this material and its properties are given in the following section. Properties of Santocel.-Santocel is a porous silica aerogel manufactured by the Monsanto Chemical Company. The grade used for thermal insulation (Santocel A) is a granular material containing 93-96 weight percent silica and 2-5% sodium sulfate, so that for most purposes it may be considered as chemically inert. It has excellent temperature stability since it has been reported that it will withstand temperatures as high as 8000C apparently indefinitely without breakdown of structure.24 It is a free-flowing powder with a bulk density of 4.5-5 lb/cu ft. A given mass of the aerogel contains 94 volume percent air in the form of minute air cells. It has been -estimated that it has a mean pore size of 5 5 m25 and a specific adsorption surface of 400 sq m/g.26 The thermal conductivity of Santocel at room temperature and atmospheric pressure is 10% less than that for still air, and it has the lowest thermal conductivity at atmospheric pressure of any insulator so far reported.24 At low pressures (0.1 ~), only a few types of glass wool have lower thermal conductivities at a mean temperature of 100~F.27 The thermal conductivity of Santocel at various pressures and for different

The University of Michigan. Engineering Research Institute conditions of temperature is listed in Table II. For purposes of comparison, the apparent thermal conductivity of a Dewar-type vacuum space was calculated as 0.011 Btu/(hr)(sq ft)(~F/in.), based on the data reported for the heat leak into a spherical, 8-liter, liquid oxygen container having silvered surfaces and a vacuum of 1 x 10-6 mm Hg.52 TABLE II THERMAL CONDUCTIVITY OF SANTOCEL IN AIR Btu/(hr)(sq ft)(OF/in.) Between Liquid N2 and At A roximatel Room Temperature Pressure, Room Temperatures Ref. 24 mm Hg Ref 36 Ref 5 Ref. 22 Ref. 35 20-40 mesh 100-200 mesh. 1 0.0357 0025 0.074 o0 o43 lo0-l 0 0218 0.017 0.044 0.027 10-2 0.0125 0.012 o.o36 0.025 10-3 o.00o87 o.01 o.0o35 o.024 L10-4 0.0073 0.023.o0255 Although Santocel is a gel, it has a sufficiently open structure so that moisture absorption is small even at large relative humidities,24 beirfg about 3 to 4 percent. Its behavior under mechanical pressure up to 15 psi (repeated compression and release) shows an initial reduction in volume of about 10 percent due to closer packing of the particles and no further change.24 RECOMMENDATIONS RESULTING FROM LITERATURE SURVEY Porous or fibrous insulators have certain advantages over a simple Dewartype vacuum space. These are: 1. For pressures in the vacuum space from atmospheric down to about 1 x 10-3 mm Hg absolute, such insulators have a lower heat-leak rate than a Dewar-type space. This holds for vacuum spaces as low as 1 cm in thickness. 2. The reflectivity of the confining metal walls has little, if any, effect when porous or fibrous insulators are used at temperatures below room temperature. 3. The porous or fibrous insulators provide a certain degree of mechanical support to the confining walls.

The University of Michigan ~ Engineering Research Institute On the other hand, if the vacuum space is maintained at a pressure of about 0.5 x 10-4 mm Hg absolute or less, the Dewar-type vacuum space has a lower rate of heat leak than porous or fibrous insulators in thicknesses of 3 cm or less. Few data are available for porous insulators on the effect of the insulation thickness on the apparent thermal conductivity. For example, for Santocel, Blat et al.5 report the same apparent thermal conductivity for thicknesses of 1 and 3 cm, while Reynolds et al.35 report an increase in the apparent thermal conductivity as the thickness increases. The latter investigators reported data for 0.5-, 1.0-, and 1.8-inch thickness. The corresponding values of the apparent thermal conductivity were in the proportion 1.0:1.14:1;50. Consequently, it is not certain at what thickness the rates of heat leak become equal for the two cases. It appears that for any given design problem there is some critical distance below which the Dewar-type high-vacuum space has a lower rate of heat leak than a space filled with porous insulation (also under vacuum, but of the order of 10-3 - 10-4 mm Hg), and that this value may vary with the geometry of the system. Of the various powders and fibers for which data are reported in the liter-'ature surveyed, silica aerogel (Santocel) has the best combination of properties, such as low apparent thermal conductivity, chemical inertness and temperature stability, ease of handlings and commercial availability. Therefore, Santocel appears to be the best choice for further investigation. Since Santocel has a high specific surface, its gas adsorptivity deserves consideration. This property is advantageous in that it aids in creating and conserving -vacuum when the desorbed powder is at low temperature. To accomplish this most effectively, the insulation should be heated under vacuum to'desorb gases from the aerogel before the vacuum space is sealed off, Since the powder may be fluidized by the desorbed gases, evacuation may have to proceed slowly at the start, and a filter should be provided to prevent entry of the powder into the main vacuum system. It has also been suggested that other gases, such as carbon dioxide and various Freons be used to displace air from the aerogel. The gases used should have lower thermal conductivities and higher liquefaction temperatures than air. It is reasoned that adsorption of the gas by the aerogel and the lower thermal conductivity of the gas will lower the apparent thermal conductivity of the powder. This effect is quite pronounced at atmospheric pressure but should. decrease as the pressure is decreased.

The University of Michigan ~ Engineering Research Institute PART II. ANNOTATED BIBLIOGRAPHY A. LITERATURE REFERENCES OTHER THAN PATENTS 1. Aberdeen, J., and Laby, T. H., Proc. Royal Soc. (London), A, 113, 459 (1926). The conductivity of Silox powder (a very pure silica with some silicon) in air, C02, and H2 was measured at room temperature and pressures from 1-760 mm Hg. The conductivity was found to depend on the nature of the gas but not on that of the solid (except for pore size). The results obtained were similar to those of Smoluchowski,42 but authors criticized his experimental procedure and theoretical derivation. An empirical correlation for their own data was formulated. Stanley's patent53 was analyzed and a thermal conductivity of 0.02 Btu/(hr)(sq ft)(OF/in.) was predicted for the insulated space. The use of Silox under vacuum is suggested as thermal insulation for liquid oxygen vessels, using a relatively low vacuum and unpolished retaining walls. The heat leak for such a vessel was calculated as 14 x 10- and 5.6 x 10-4 cal/ (sq cm)(sec) for insulating spaces of 1.0 and 2.5 cm, respectively. 2. Anon., Mech. Eng., 69, 772 (1947). Santocel is described. 3. Anon., Product, 18, 168 (1947). Santocel is described. 4. Austin, J. B., Symposium on Thermal Insulating Materials, American Society for Testing Materials, Philadelphia, 1939. Pages 3-67..Detailed review of the factors influencing the thermal conductivity of nonmetallic materials, including powders. 5. Blat, M. I., Bresler, S. E.,and Ryabinin, Y. N., J. Tech. Phys. (USSR), 15, 916 (1945). Two concentric hollow steel spheres, the inner filled with liquid 02 or N2 and the outer at 14.5~C, were used to determine the apparent thermal conductivity of a vacuum space and of Santocel. Pressures from 2 x 10-6 to 760 mm Hg were used. Without Santocel in the vacuum space, the rate of heat transfer into the inner sphere depended on the conditions of the walls, being considerably less with polished walls. With Santocel, no such difference was obtained. The apparent thermal conductivity of the Santocel was found to be the same for a l-cm and a 3-cm gap. The adsorption of air by Santocel over the range 10-760 mm Hg was determined.

The University of Michigan ~ Engineering Research' Institute 6. Codegone, C., Atti accad. sci. Torino, Classe sci. fis. mat, e nat., 81, 75 (1945-46). A method for measurement of thermal conductivity at low temperatures and atmospheric pressure is described, in which material to be tested is shaped into a hollow sphere, with a metal sphere full of solid C02 within its cavity. Applied to "cellular cement." 7. Codegone, C., Ricerca Sci., 20, 68 (1950). The above method was applied to silica gel, vermiculite, MgCO3 powder, and rock wool, using liquid 02 in the inner sphere. 8. Dewar, J., Proc. Roy. Inst., 15, 815 (1898). First to report that the introduction of finely powdered solids into the vacuum space of a Dewar vessel increases the effectiveness of the insulation. Also studied effects of metallic foil as reflective insulation. Results are qualitative only. 9. Edberg, E. A., Refrig. Eng., 64, 38 (1956). Described expandable polystyrene insulation. 10. E ichstaedt, H. Ao, NBS Report No. 3517, 8 (1955). Report on the liquid 02 container manufactured by Ronan and KunzL'l, Inc. It is a cylindrical aluminum container, built in 50-, 150- and 500-gallon sizes. In the 500-gallon container,an 8-in. vacuum space is filled with Santocel, held at 10 microns or less. Conductivity was roughly 00015 Btu/(sq ft)(hr)(~F/in.). 11. Finck, J. L., Bur. Standards J. Research, 5, 973 (1930). Investigated conductivity of fibrous materials as a functionaf bulk density. Investigated cnductivity of midt fibers, of randomly arranged and orderly fibers, and the influence of moisture. Observed radiation decrease due to the effect of Al powder as opacifying agent when dusted on kapok, etc. 12. Finck, J. L., Ind. E. Chem., 31, 824 (1939). Expressed solid conduction and radiation in fibrous insulators as functions of bulk density, and thus determined optimum density. 13. Fradkov, A. B., Doklady Akad. Nauk S.S.S.R., 81, 549 (1951). Made conductivity measurements on liquefied gas containers insulated with a 25-mm layer of Mipora (a urea-formaldehyde resin) at pressures between 10 microns and atmospheric.

The University of Michigan ~ Engineering Research Institute 14. Fulk, M. M., Reynolds, M. M., and Park, 0. E., NBS Report No. 3517, 151 (1955). Absorptivities of many kinds of metal surfaces at 760K for 300'K blackbody radiation are listed. Radiative heat transfer through powders was found to follow the T4 law multiplied by a Beer-Lambert exponential damping factor due to the powder. 15. Glaser, P. E., Ph.D. thesis, Columbia University (1955),as reported in: Glaser, P. E., and Kayan, C. F., Refrig, Eng., 64, 31 (1956). Measured conductivity of Celite (dry, calcined diatomaceous earth) in He, NH3, and air between 3 and 760 mm Hg and between 50 and 3000F. 16. Goalwin, D. S., Refrig. Eng., 59, 251 (1951). Tested techniques and equipment for the transfer of liquid 02, found glass fiber and silica gel to be preferable as insulation. 17. Hallett, N. C., Altman, H. W., Yeager, M. L., and Newton, C. L., NBS Report No. 3517, 5 (1955). Constructional details are given for the cylindrical liquefied gas containers made by H. L. Johnstone, Inc., holding 2000 to 10,000 liters. Annular space of 7 in., filled with Santocel (vacuum pre-dried) at 0.1 micron, has conductivity of 0.015 Btu/(sq ft)(hr)(~F/in.) in practical operation. 18. Johnstone, H. L., Hood, C. B., and Bigeleisen, J., NBS Report No. 3517, 139 (1955). Made experimental comparison of different insulations between 200K and row temperature, and at pressures between 0.1 micron and atmospheric. In general, powder insulation was found to be more effective than fibrous or cellular insulation, and to vacuum alone. Of all insulations tried out, Santocel proved by far the most efficient. 19. Johnson, V. J., and Wilson, W. A., NBS Report No. 3517, 246 (1955). The NBS hydrogen liquefier plant is described. Silica gel at liquid N2 temperature was used to adsorb N2 and 02 from gas to the liquefier. 20. Jones, C. L., Paint Ind. Mag., 62, 390 (1947). Described Santocel and its preparation. 10

The University of Michigan ~ Engineering Research Institute 21. Kannulick, W. G., and Martin, L. H., Proc. Roy, Soc. (London) A, 141, 144 (1933). Made conductivity measurements on powdered magnesia, glass, and carborundum at 1-80 mm, using air, H2, and C02. Found correlation of Aberdeen and Labyl to be unsatisfactory in this range. Derived a second-degree algebraic expression. 22. Katan, L. L., Vacuum, 1, No. 3, 191 (1951). Made conductivity measurements on liquid 02 vacuum-jacketed container, using Santocel, slag wool, basic magnesium carbonate, and brelite, an expanded volcanic mineral. Pressure range, 1 micron to 1000 mm; temperature range, 70 to -360~F. Detailed experimental procedure is described. 23. Kistler, S. S., Nature, 127, 741 (1931). First prepared silica aerogel that retains its porous structure by supercritical evaporation of the alcogel. 24. Kistler, S. S., and Caldwell, A. G., Ind. Eng. Chem., 26, 658 (1934). Measured conductivity of silica aerogel, between 0.1 micron and atmospheric pressure and at room temperature, for different particle. sizes. Although they confirmed the data of Aberdeen and Laby,1 they did not feel justified in deriving the latters' correlation. Derived theoretical correlation based on solid and gas conduction within the granules, and on solid-solid and solid-*-gas-solid conduction between granules. Radiation not included in derivation, since temperature drop in tests was low. 25. Kistler, S. S., J. Phys. Chem., 39, 79 (1935). Based on his theoretical correlation,24 he calculated pore size of silica aerogel; conductivity measurementswere made between 2 and 740 mm and at room temperature, using air, C02, and Freon. 26. Kistler, S. S., J. Phys. Chem., 46, 19 (1942). Based on his theoretical correlation,24 he calculated the specific surface area of silica aerogel. Found it to agree with results from adsorpt ion measurements. 27. Lander, R. M., Trans. A. S.H.A.E., 61, 151 (1955). The tests of Rowley et al.36 were continued, substituting air partially or totally by C02 and Freon-12. Glass wool was investigated further by varying the mean temperature between -50~ and +100~F, and by using specimens of different densities. 11

The University of. Michigan * Engineering Research Institute 28. Larkin, B. K., A Study of the Rate of Thermal Radiation Through Porous Insulating Materials, Ph.D. thesis, The University of Michigan (1957). Studied radiant heat transfer through Styrofoam, Foamglass and Fiberglas boards at atmospheric pressure and at temperatures between 700F and 400~F, using thicknesses from 1/8 to 1 in. Analyzed scattering and absorption parameters. Found that the optimum fiber diameter is 2-5 microns, and the optimum foam-bubble diameter is of the same order of magnitude, for minimum radiation. Found that increasing foam density decreases radiation if absorption is more important, but increases radiation if scattering is more important; with fibers, increasing density decreases radiation. Surface emissivity was found to have influence on radiant heat transfer. 29. Macormack, J. G., Refrig. E., 41, 311 (1941). Described "Ferrotherm," a lead-tin-alloy coated steel, having a dull surface but having a higher reflectivity for infrared radiation than polished stainless steel, and hence used as reflective insulation. 30. Mason, R. B., Ind. Eng. Chem., 25, 245 (1933). Described experiments on the use of reflective insulation, especially Al foil, at normal temperatures. Heat transfer may be made to approximate that of still air by appropriate use of Al foil. 31. McIntire, 0. R., and Kennedy, R. N., Chem. E _g. Progr., 44, 727 (1948). A list of plastics used for the production of foam insulators is given, both U.S. and foreign. Styrofoam, manufactured by Dow Chemical Company, is described. 32. McIntire, 0. R., and McCuaig, D. W., Refrig. Eng., 52, 217 (1946). Described Styrofoam, a foam insulator made by the Dow Chemical Company. Cut down radiation considerably by adding 1 percent of Al powder as opacifying agent. 33. Moench, G. C., Glas u. Hochvakuum-Tech., 1, 1, 9 (1952). Discussed the use of carbon black as insulation between the glass walls of a Dewar vessel, and the installation of a third wall in the vacuum space. 34. Ogden, F. F., and White, J. F., Refri E_., 52, 411 (1946). Described Santocel. Discussed use of opacifying agents with Santocel and concluded that they were only effective above 100~F. 12

The University of Michigan ~ Engineering Research Institute 35. Reynolds, M. M., Brown, J. D., Fulk, M. M., Park, 0. E., and Curtis, G. W., NBS Report No. 3517, 142 (1955). Thermal conductivity of' perlite at 0.01 micron and of Santocel at 2 microns or less was measured. Cylindrical jacketed-wall N2 containers were used. Found that with p erlite, finer particles are better insulators. The addition of Al powder or the use of black surfaces decreases conductivity.. However, vacuum and no powder at 0.01 micron is superior to any of the above uses of perlite~ Santocel proved superior to pure vacuum. Annular spaces of 0.5, 1.0, and 1.8 in. were used. Santocel AO, containing 10 percent Si, has roughly 80 percent conductivity of Santocel. Influence of outer shell-temperature was also measured. 36. Rowley, F. B., Jordan, R. C., Lund, C. E., and Lander, R. M., Trans. A.S.H.V.E., 58, 155 (1952). Tested conductivity of sever.l fibrous and porous insulators at a mean temperature of 1000F, at 10 and at 760 mm Hg. Readings of intermediate pressures were not considered accurate enough. Substances tested were corkboard,cotton, glass wool,- hair felt, kapok, rock wool, Santocel, and wood fiber. Apparatus used was guarded hot-plate, with a temperature difference of 50~F and a mean temperature of 1000F. 37. Sanders, V., Chem. Eng. Progr, 44, 804 (1948). Described Foamglass, a cellular glass insulator made by Pittsburgh Corning Corp. 38. Scott, R. B., NBS Report No. 5027 (1956). Empty vacuum space, with clean annealed Cu, Ag, or A1 as highly reflective surfaces, competes with powdered insulation, since the latter needs enough width to decrease solid conduction. 39. Shand, L. D., Offic. Dig. Federation Paint and Varnish Production Clubs, 277, 185 (1948). Described Santocel. 40. Simonova, L. K., J. Applied Chem. (USSR), 16, 87 (1943)(French summary). Inves.tigated conductivity and heat capacity of porous insulators as function of particle size. Worked with activated carbon and silica gel at atmospheric pressure and room temperature. Contrary to Kistler and Caldwe1124 and to Smith and Wilkes,41 found conductivity to increase on powdering the sample. 13

The University of Michigan ~ Engineering Research Institute 41. Smith, W. R., and Wilkes, J. B., Ind. Eng. Chem., 36, 1111 (1944). Investigated conductivity of carbon black as function of particle size, at atmospheric pressure. Found conductivity to be independent of particle size, similar to Kistler and Caldwell24 and contrary to Simonova,40 except when carbon black is pelletized, where conductivity increases. 42. Smoluchowsk4 M., Bull. Intern. acad. sci. de Cracovie, 5B, 129 (1910),and ibid., 8A, 548 (1911). First to conduct quantitative thermal conductivity measurements for different powders in air, at room temperature, between 1 and 760 mm. Plotted conductivity versus log P. Found solids conduction to be negligible, and that conductivity depended on the nature of the gas but not on that of the solid, except for pore size. Derived theoretical expression of conductivity based on analogy to slip-flow under vacuum. Recognized the importance of radiation. 43. Taylor, C. S., and Edwards, J. D., Heating, Piping, and Air Conditioning, 11, 59 (1939). Studied the effect of an oxide layer on the reflectivity of Al. 44. Topper, L., Ind. Eng. Chem., 47, 1377 (1955). Derived a mathematical expression for the conductivity of a porous insulation as function of pore size, pore form, and number of pores per unit volume. 45. Van Buskirk, E. C., and Surland, C. C., Chem. Ens. Progr., 44, 803 (1948). Described Flotofoam, a urea-formaldehyde foam insulator made by the U. S. Rubber Company. 46. Verschoor, J. D., and Greebler, P., Trans. Amer. Soc. Mech. Engrs., 74, 961 (1952). Derived a theoretical correlation of conductivity vs pressure for fibrous insulation as function of fiber density and diameter. Made conductivity measurements on glass fiber insulation between 1 micron and atmospheric pressure, at room temperature. The mathematical model considers solid conduction, gas conduction, and radiation. 47. Waite, H. J., NBS Report No. 3517, 158 (1955). Lists mechanical and thermal properties of Styrofoam, a polystyrene cellular insulation made by the Dow Chemical Company, and its uses. 14

The University of Michigan ~ Engineering Research Institute 48. White, J. F., Ind. En. Chem., 31, 827 (1939). Tested Si, Al, and graphite as opacifying agents with Santocel, at atmospheric pressure and between 100 and 6000F. 49. White, J. F., Refrig. Eng., 41, 171 (1941). Described Santocel. 50. White, J. F., Chem. Eng. Progr,, 44, 647 (1948). Described Santocel, and commented on work of Blat et al.5 51. Wilkes, J. B., Ind. Eng. Chem., 31, 832 (1939). Described use of reflective insulation. 52. Wilson, W. A., NBS Report 5041 (1957). A complete report on the design and performance of an 8-liter spherical liquid oxygen tank with a 1-in. Dewar-type vacuum space. B. PATENT LITERATURE U. S. Patents 53. U. S. Patent 1,071,817, Sept. 2, 1913 (Stanley, W.). Double-wall metallic vacuum flank, in which gas-absorbing powder at reduced pressure is contained in the annular space. Insulating power thus obtained claimed to be the same as that of a conventional Dewar flask having an absolute pressure 100 times lower, when used for storing liquefied gases. 54. U. S. Patent 1,694,967, Dec. 11, 1929 (Coolidge, W. D., to General Electric Co.). Use of lamp black and P205 as adsorbent in vacuum space of vessels used for storing liquefied gases. 55. U. S. Patent 2,093,454, Sept. 21, 1937 (Kistler, S. S.). Preparation of silica aerogel by supercritical evaporation of silica aquagel directly, without passing through the alcogel stage. 15

The University of Michigan ~ Engineering Research Institute 56. U. S. Patent 2,164,143, June 27, 1939 (Munters, C. G., to Aktiobolaget Termisk Isolation). Flexible double-jacketed container, with powder under vacuum in jacket. 57. U. S. Patent 2,396,459, March 12, 1946 (Dana, L. J., to The Linde Air Products Co.). Large, rigid, double-walled tank, not less than 2 ft ID, annular space not less than 3 in., at pressures between 1 micron and 10 mm, filled with finely divided solid material. When used as liquid oxygen tank, a heat flux of less than 5 Btu/(sq ft)(hr) is specified. British Patent 58. Brit. Patent 252,385, May 20, 1925 (Union Generale Cooperative). Use of charcoal as adsorbent in vacuum space of vessels used for storing liquefied gases. French Patent 59. Fr. Patent 657,440, July 12, 1928 (I.G. Farbenindustrie A.-G.). Use of silica as adsorbent in vacuum space of vessels used for storing liquefied gases. German Patents *60. Ger. Patent 510,577, June 20, 1927 (I.G. Farbenindustrie A.-G.). Use of silica as adsorbent in vacuum space of vessels used for storing liquefied gases. 61. Ger. Patent 528,497, July 4, 1930 (Societe l'air liquide). Heat-insulating means for apparatus at very low temperatures. 62. Ger. Patent 806,786, June 18, 1951 (Perlick, A., to Gesellschaft fur Linde s E ismachinen A.-G. ). Use of finely divided silica as filler for liquefied gas vessels. The silica forms a solid gel, which, since it will not flow, provides greater safety in te event of damage to the vessels. 16

The University of Michigan. Engineering Research Institute 63. Ger. Patent 809,561, July 30, 1951 (Kahle, H., to Gesellschaft fur Linde's Eismachinen A.-G.). Jacketed-wall insulating case for storage of liquefied gases, with an adsorbent contained in the vacuum jacket. 17

UNIVERSITY OF MICHIGAN 11111111111 1111111111111111111111111111111111111111111 3 9015 02493 9145 THE UNIVERSITY OF MICHIGAN DATE DUE