# ASTM C740C740M-13

Designation: C740/C740M − 13Standard Guide forEvacuated Reflective Insulation In Cryogenic Service1This standard is issued under the fixed designation C740/C740M; the number immediately following the designation indicates the yearof original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval.A superscript epsilon (´) indicates an editorial change since the last revision or reapproval.1. Scope1.1 This guide covers the use of thermal insulations formedby a number of thermal radiation shields positioned perpen-dicular to the direction of heat flow. These radiation shieldsconsist of alternate layers of a low-emittance metal and aninsulating layer combined such that metal-to-metal contact inthe heat flow direction is avoided and direct heat conduction isminimized. These are commonly referred to as multilayerinsulations (MLI) or super insulations (SI) by the industry. Thetechnology of evacuated reflective insulation in cryogenicservice, or MLI, first came about in the 1950s and 1960sprimarily driven by the need to liquefy, store, and transportlarge quantities of liquid hydrogen and liquid helium. (1-6)21.2 The practice guide covers the use of these MLI systemswhere the warm boundary temperatures are below approxi-mately 400 K. Cold boundary temperatures typically rangefrom 4 K to 100 K, but any temperature below ambient isapplicable.1.3 Insulation systems of this construction are used whenheat flux values well below 10 W/m2are needed for anevacuated design. Heat flux values approaching 0.1 W/m2arealso achievable. For comparison among different systems, aswell as for space and weight considerations, the effectivethermal conductivity of the system can be calculated for aspecific total thickness. Effective thermal conductivities of lessthan 1 mW/m-K [0.007 Btu·in/h·ft2·°F or R-value 143] aretypical and values on the order of 0.01 mW/m-K have beenachieved [0.00007 Btu·in/h·ft2·°F or R-value 14 300]. (7)Thermal performance can also be described in terms of theeffective emittance of the system, or Εe.1.4 These systems are typically used in a high vacuumenvironment (evacuated), but soft vacuum or no vacuumenvironments are also applicable.(8) A welded metal vacuum-jacketed (VJ) enclosure is often used to provide the vacuumenvironment.1.5 The range of residual gas pressures is from 1.0.The mean free path (l) for the gas molecule may be determinedfrom the following equation:l 5kBT=2πξ2P(6)If the mean free path is significantly larger than the separa-tion between the hot side and cold side, then gaseous con-duction will be reduced.16 For many systems, a vacuumpressure of roughly 50 millitorr is the point below which thefree molecular range begins. However, some amount of gasconduction still remains until the 10-6torr level is reached.For example, some mean free path values for air at roomtemperature are approximately 0.1 m for 10-3torr and 100 mfor 10-6torr.4.2.2 The working definition of soft vacuum (SV) is a rangeof residual gas pressure from 10-2torr to 10 torr (1.33 Pa to1333 Pa) which represents a transition regime of the thermo-physical behavior of the gas. The gaseous component of theheat transfer through a material in the SV range is between freemolecular conduction and convection. This range is one ofsharp transitions and often associated with strong dependencieson the morphology, composition, and construction of theinsulation materials. The molecular flow condition is for 1.0 Kn 0.01. Thermal insulation systems operating in the softvacuum range often have all modes of heat transfer working insubstantial proportions.4.2.3 The working definition of no vacuum (NV) is a rangeof residual gas pressure from 100 torr to 1000 torr (13.3 kPa to133 kPa) which represents a continuum regime of the thermo-physical behavior of the gas. The continuum regime is associ-ated with viscous flow and convection heat transfer. Themolecular flow condition is for Kn 0.01. While most MLIsystems are designed to operate under high vacuum conditions,other MLI systems may be designed to operate under softvacuum or no vacuum conditions. In other cases, knowledge ofthe performance sensitivity due to degraded vacuum or loss-of-vacuum conditions can be crucial for system operation andreliability analysis. The three basic ranges of vacuum levels(high vacuum, soft vacuum, and no vacuum) are depicted in theMLI system performance curve given in Fig. 2.(17) In thisexample, the MLI system is 40 layers of aluminum foil andmicro-fiberglass paper under the following conditions: coldboundary temperature of 78 K, warm boundary temperature of293 K, and gaseous nitrogen as the residual gas.4.2.4 Cryopumping effects through the innermost layersgreatly aid in producing the desired high vacuum levelsbetween the layers by freezing, condensing, and adsorbing thesome of the residual gases. The assumption here is that thevacuum environment can be approximately the same as thevacuum between the layers for a properly designed andexecuted MLI system.4.2.5 Also important are the type of spacer material usedand the layer density. A spacer material that is readily evacu-ated and very low outgassing is more conducive for obtainingand maintaining the desired high vacuum condition. A lowerlayer density typically promotes better evacuation and higherultimate vacuum levels, but an exceptionally low layer densitycan make maintenance of the high vacuum condition evenmore critical.4.2.6 An acceptable CVP is achieved with a well-ventedreflector-spacer system that provides communication betweenthe interstitial spaces and the vacuum environment. Failure toprovide proper venting can result in serious degradation ofthermal performance.4.3 Mechanical Loading Pressure: .4.3.1 In practice, the reflector layers are not free-floating.Compression between the layers due to the weight of theinsulation or to pressures induced at the boundaries, or both,can cause physical contact between the reflectors producing amore direct conduction heat transfer path between the layers,thereby increasing the total heat flux of the system. The goal indesigning any MLI system for high vacuum operation is tominimize the thermal contact as much as possible.4.3.2 The effects of compression on the heat flux can beobtained experimentally using a flat plate calorimeter.(18)Experimental correlations have been obtained for a variety ofreflector-spacer combinations that indicate that the heat flux isproportional to Pbwhere b varies between 0.5 and 0.66.Typicaldata for a number of MLI systems are presented in Fig. 3 thatillustrate this effect. The typical MLI systems listed hereprovide no significant mechanical strength as the compressiveforces should be kept near zero, or less than about 10 Pa (0.001psi) for optimum performance. The overall configuration of theinstalled system, whether horizontal or vertical, as well as theunit weight of the MLI must therefore be considered for anaccurate estimation of actual system thermal performance. (19,20)C740/C740M − 1344.4 Performance Factors:4.4.1 There are three complementary ways of expressing thethermal performance of an MLI system. One way is to expressthe performance in terms of radiation transfer since theseinsulations are predominantly radiation controlling. A secondway is to calculate the steady-state heat flux. A third way is touse the classical thermal conductivity term in spite of the factthat the thermal profile across these insulations is not linear.Elaboration and a discussion of these approaches follow:4.4.2 Effective Emittance:4.4.2.1 The effective emittance of an MLI has the samemeaning as the emittance factor, E1or E2, when it is applied tothe theoretical performance of the system. The effectiveemittance of an actual system is given by the ratio of themeasured heat flux per unit area to the differences in the blackbody emissions (per unit area) of the boundaries at their actualtemperatures as given by Eq 7. The effective heat transfer areasfor both warm and cold surfaces must be applied.Ee5 q/~σTh42 σTc4!(7)4.4.2.2 The measured average total effective emittance of agiven insulation will have different values depending upon thenumber of reflectors, the total hemispherical emittance of thereflector materials, the degree of mechanical compressionpresent between layers of the reflectors, and the boundarytemperatures of the system. This effective emittance factor canbe used to compare the thermal performance of different MLIsystems under similar boundary temperature conditions.4.4.2.3 Installation Factor—The installation factor, I,istheratio of the actual system heat flux to the theoretical systemheat flux, that is,I 5 qactual⁄qtheoretical(8)The installation factor can only have values larger than 1.0.At a value of 1.0 the amount of degradation is zero and theactual performance corresponds to the theoretical performance.Degradation factors can range from 1.5 to 10 for high vacuumconditions and can be much higher for even moderatelydegraded vacuum conditions as indicated in Fig. 4. Thetheoretical system heat flux is not necessarily known, but isgenerally taken to be the idealized blanket tested underlaboratory conditions.4.4.3 Heat Flux:4.4.3.1 The heat flux, q, of a thermal insulation system canbe defined by the total heat flow rate divided by the effectivearea of heat transfer in comparable units as follows:q 5 Q⁄Ae(9)The effective heat transfer area, Ae, is the mean area throughwhich heat moves from the hot boundary to the cold bound-ary and is further defined as follows:For flat disk geometries: Ae5π4de2(10)where deis taken as the inner diameter of vessel or pipeplus one wall thickness of that same vessel or pipe.For cylindrical geometries: Ae5 2π~Le!x⁄1nSdodiD(11)where Leis the effective heat transfer length of the cylinderand do and di are the outer and inner diameters, respectively,of the insulation system.For spherical geometries: Ae5 πdodi(12)where doand diare the outer and inner diameters,respectively, of the insulation system. The heat flux can becomputed based on the MLI or the total system. Forexample, the outer diameter of the MLI is chosen for theMLI heat flux while the inner diameter of the vacuum jacketFIG. 2 Variation of Heat Flux with Cold Vacuum Pressure: example MLI system of 40 layers foil and paper with boundary temperaturesof 78 K and 293 K and nitrogen as the residual gas. [Note: 1 millitorr = 0.133 Pa]C740/C740M − 135is chosen to compute the total system heat flux. Accordingly,the heat flux should be stated as for the MLI only or for thetotal system. The basic form using the Fourier rate equationfor heat conduction is given as:q 5 ke~∆ T ⁄ x! (13)The Lockheed Equation gives an empirical form as follows:q 5Cs*n¯2.63~Th2 Tc!*~Th1 Tc!2*~n 1 1!1CR*e*~Th4.672 Tc4.67!n1CG*P*~Th0.522 Tc0.52!n(14)All three modes of heat transfer are accounted for by theleading coefficients: solid conduction (CS), radiation (CR),and gaseous conduction (CG). Even at high vacuum levels,some gas molecules do exist between the layers of radiationshields and spacers necessitating a term for gaseous conduc-tion. The Lockheed Equation (21) is based primarily on datafrom MLI systems comprised of double-aluminized mylarradiation shields with silk net spacers and tested using a flatplate boiloff calorimeter.) Alternatively, the general form forthe physics-based equation developed by McIntosh (22) isgiven as follows:q 5σ~Th42 Tc4!S1εh11εc2 1D1CGPα~Th2 Tc!1Csfk~Th2 Tc!x(15)The McIntosh Equation, as well as the Lockheed Equation,has three terms: one for radiation between shields, one forsolid conduction through the spacers, and one for gaseousconduction due to any residual gas molecules among thelayers. The term f is the relative density of the spacer com-pared to the solid form of the material. The use of these orCurveNo.Numbers ofLayersReflector Spacer1 10 1145—H19 Tempered Alu-minumNylon Netting (11 layers)2 10 Aluminized (both sides)PolyesterGlass Fabric (22 layers)3 10 Aluminized (both sides)PolyesterSilk Netting (33 layers)4 10 Aluminized (both sides)Polyester32 kg/m3Polyurethane Foam (11 layers)5 10 Aluminized (both sides)PolyesterSilk Netting with 0.1-mm by 12.7-mm Strips of Glass Mat(11 layers)6 10 Aluminized (both sides)PolyesterSilk Netting with 0.2-mm by 6.4-mm Strips of Glass Mat(11 layers)FIG. 3 Effect of Mechanical Compression on Heat FluxC740/C740M − 136other equations available in the literature requires adequateunderstanding of all three heat transfer modes as well as thetesting methodologies used and the influences of installationfor a given application.4.4.4 Effective Thermal Conductivity:4.4.4.1 The effective thermal conductivity (ke)ofanMLIsystem can be defined by the ratio of the heat flow per unit areato the average temperature gradient of the system in compa-rable units as follows:ke5 ~Q ⁄ Ae! ⁄ ~∆ T ⁄ x! (16)For highly-evacuated MLI systems, the effective thermalconductivity can be expressed as follows (23) :ke5 ~N ⁄ x!21⁄ @hc1 σ e ~Th22 Tc2!~Th2 Tc! ⁄ ~2 2 e!#(17)The effective thermal conductivity is determined from Fouri-er’s law for heat conduction through a flat plate as given byequation (Eq 18), between concentric cylinders as given byequation (Eq 19), and between concentric spheres as givenby equation (Eq 20):Flat Plate: ke54Qxπde2∆T(18)Cylindrical: ke5Q1nSdodiD2πLe∆T(19)Spherical: ke5Qxπdodi∆T(20)4.4.4.2 Because radiation heat transfer within an MLI sys-tem produces a nonlinear temperature gradient, kewill varyapproximately as the third power of the mean temperature.Thus, kecan be properly used for comparison of performanceof different MLI systems only when the boundary temperaturesare the similar.4.4.4.3 The total insulation thickness must be carefullydefined. Whenever keis used to describe the thermal perfor-mance of an MLI system, a statement indicating the methodused in making the thickness measurement and the accuracy ofsuch measurement is needed. In some cases, an estimate of arange of thicknesses for a given installation may be appropri-ate. Alternatively, the appropriate diameter of the vacuum canor jacket can be used to establish a thickness for determining anoverall system thermal conductivity (ks).4.5 Typical Thermal Performance of MLI—The thermalperformance of MLI systems can vary over a wide rangedepending largely upon the fabrication techniques, but alsoupon the materials used for the reflectors and spacers. (24)Performance will vary in accordance with different boundarytemperatures. Performance can also vary widely for tanks, rigidpiping, and flexible piping applications. In all cases, under-standing the total system performance, including MLI,supports, attachments, penetrations, getters, etc., is the mainpoint. Testing methods and equipment include a wide range ofboth boiloff calorimetric and electrical-based techniques.(25-31)5. Practical Performance and Applications5.1 Insulations of the type described above are generallyused when lower conductivities are req