<Jf--------

180 mm

1110

blank line
!Jlonk line
IJ/ank line

Cryogenic Vacuum Insulation for Vessels and Piping

blank line

Kogan A., Fesmire J.*, Johnson W.*, Minnick J.

hlank line

Lydall Performance Materials, 68 George Str., Green Island, NY 12183 USA

*NASA Kennedy Space Center, NE-F6, Kennedy Space Center, FL 32899 USA
IJ/onk line
hlank linc

E
E

Cryogenic vacuum insulation systems, with proper materials selection and

execution, can offer the highest levels of thermal performance. Three areas of
consideration are vital to achieve the optimum result: materials, representative test
conditions, and engineering approach for the particular application. Deficiency in
one of these three areas can prevent optimum performance and lead to severe
inefficiency. Materials of interest include micro-fiberglass, multilayer insulation,
and composite arrangements. Cylindrical liquid nitrogen boil-off calorimetery
methods were used. The need for standard thermal conducti vity data is addressed
through baseline testing. Engineering analysis and design factors such as layer
thickness, density, and practicality are also considered.
blank line
blank line
blank line

INTRODUCTION
blank line

Cryogenic thermal insulation systems that incorporate a vacuum environment can provide the lowest
possible heat transfer from the local environment to the stored cryogen. Thermal conductivities in the range
of 0.01 to 1 m W Im- K are achievable with the right combination of materials in a high vacuum environment
less than 1 mill itorr. While multilayer insulation (MLI) systems can provide the ultimate thermal insulating
capability, overall system design and operational factors prevent complete utilization of their effectiveness.
Fiberglass insulation composed of low outgassing micro-fibers can provide effective high-performance
capability in vacuum as well. Vacuum level requirements are considerably less strict for fiberglass,
providing cost advantages both in manufacturing and life cycle for cryogenic vessels and piping.
Installation around piping, structural supports, and other complex geometries can be readily accomplished
using fiberglass. Compression, seams, penetrations and edge effects are known to increase heat leak
through MLI systems by 100 percent or more if the system is improperly designed.
Combining MLI and fiberglass in the vacuum annulus of vessels and piping can be done in a number
of ways. The materials can work together to meet different thermal performance, cost, or mechanical
objectives such as space and weight. Materials, representative conditions, and engineering approach must
be considered for a particular application. Deficiency in one of these three areas can prevent optimum

2110

performance and lead to costly inefficiencies. Materials of interest include micro-fiberglass, MLI, and
composite arrangements. Thermal performance data under representative cryogenic-vacuum conditions are
needed for calculating the overall efficiency of a given design and assessing the long-term economics of the
operational system.
NOllk I illc
NOIII... Iinc

TEST EQUIPMENT AND METHOD

')1 /Il/.. Ii

Cryostats using steady-state liquid nitrogen boil-off calorimetery methods are used to determine apparent
thermal conductivity (k-value) and heat flux. The Cryogenics Test Laboratory at NASA Kennedy Space
Center has developed several cryogenic insulation test instruments for testing of materials and systems
under large temperature differential and full-range vacuum conditions [1 , 2J. Cryostat testing is performed
'using laboratory standard practices. A comparative cylindrical unit, Cryostat-2, was recently reactivated at
the Lydall cryogenics laboratory in Green Island, NY for use in this study (see Figure 1).

Seneor
T1

n , T3. H
T5
vel . VC2, VC3

toc.tion
l"sicM L..)'fIt (cold

mau'. caT

Middle Lave"
OUtMde LMyer, WBT
v.euum CI'\ambe( Enenor

Figure 1. Insulation test instrument, Cryostat-2, installed at the Lydall cryogenics laboratory (left) and simplified
schematic showing locations of temperature sensors and equipment connections (right).

Cryostat-2 includes a 132-mm-diameter by 500-mm-Iong cold mass and can accept specimens up to
50 mm thick. The cold-mass assembly is easily removed and mounted on a wrapping machine. Each test
measures the steady-state heat leak (watts) through the specimen at a prescribed set of environmental
conditions, including a stable warm-boundary temperature (WBT), a stable cold-boundary temperature
(CBT), and a stable cold vacuum pressure (CVP). The liquid nitrogen maintains the cold mass CBT at
approximately 78 K and the WBT is maintained at approximately 293 K using an external heater with an
electronic controller. Vacuum levels cover the full range from high vacuum (HV) (below 10-4 torr) to soft
vacuum (SV) (-1 torr) to no vacuum (NY) (760 torr).
The rate of the heat transfer, Q, through the insulation system into the cold-mass tank is directly
proportional to the flow rate of liquid nitrogen boiloff. The k-value is determined from Fourier's law for
heat conduction through a cylindrical wall. The mean heat flux is calculated by dividing the total heat
transfer rate by the effective area of heat transfer. Further details on the heat transfer calculations as well as
uncertainty analyses for each apparatus have been previously reported [1,2].

3110

Each test requires a number of temperature, pressure, gas flow, and weight measurements and
controls. All signals are processed and recorded through National Instruments (NI) compact Field Point
(cFP) hardware using Labview 8.6 Software. Temperatures are measured using Type K thermocouples
through NI cFP TC-120 modules. Warm boundary temperature is controlled through a JKEM Model 250HP-RC616 using Omega heater blankets. Pressure is measured using two MKS Baratron 627B capacitance
manometers (0.1 and 100 torr), and a Granville-Phillips 356 Micro Ion Plus transducer (full range).
Pressure is controlled through an MKS model 250 pressure controller using an MKS model 0248
proportional control valve. Gas flow from liquid nitrogen boil off is measured with four MKS MlOMB
analog, elastomer-sealed mass flow meters. To cross check the flow meter, weight change due to liquid
nitrogen boiloff is measured with a Mettler Toledo PBA430x weight scale.
Munk /in/'

MATERIALS
hli It hlc

The MLI materials used in this study consist of layers of aluminum foil (7.2 micron thick, having
emissivity of 0.03) separated by a micro-fiberglass paper spacer (Cryotherm243, 12 g/m 2). The materials
can be applied separately but are preferably collated and applied from a single roll (CRS Wrap). The
blanket material is a 25 mm thick micro-fiberglass blanket of density 16 kg/m 3 (Cryolite). Photographs
of the materials are shown in Figure 3. The removable cold mass assembly of Cryostat-2 is placed on a
wrapping machine for precise control during installation of all materials and temperature sensors.

Figure 2. Photographs of micro-fiberglass spacer for MLI (Cryothenn) and micro-fiberglass blanket material (Cryolite)

The micro-fiberglass material Cryolite was developed as an alternative to the commonly used
perlite powder insulation for cryogenic tankers, such as for liquid oxygen or liquid nitrogen, that are not
otherwise insulated with MLI. Cryolite offers low density to minimize tanker weight and maximize
capacity of the vessel; ease of installation with no settling or compaction issues; oxygen compatibility; and
fast vacuum pumpdown rates with minimal outgassing. While many stationary storage cryogenic vessels
are still insulated by perlite (as perlite settling is not as severe an issue as in the transport vessels), Cryolite
still offers the advantages of fast vacuum pumpdown, better vacuum integrity, and therefore improved
insulation properties.

4/10

The Cryolite blanket material has been used in combination with MLI in the past, with different
orders of installation arrangements. The goal of this study is to determine the best possible combination for
the composite arrangement. The following two arrangements have been tested: 1) 40 layers of MLI on the
inner vessel (cold mass) followed by one layer of Cryolite blanket and 2) Cryolite blanket on the inner
vessel followed by 40 layers of MLI.
Because Cryolite blanket has very limited protection from radiation heat transfer due to its low
opacity, the evaluation of the effect of adding several layers of aluminum foil to the blanket is also of
interest. Several such arrangements were also tested.
I t1",k lille

RESULTS
The following are the results of the experiments carried out in this study. All the experiments were
conducted over the wide pressure spectrum from high vacuum 10-5 torr) to soft vacuum (1 torr) to near
atmospheric pressure (100 torr). The residual gas was nitrogen. Results are reported in terms of the
comparative effective thermal conductivity (comparative k-value) in mW/m-K and the total heat leakage

rate (Q) in W.
MLI and MLIICryolite composite
The thermal performance data for the following three insulation systems were obtained:

System 1 - 40 layers of MLI

System 2
1 layer of Cryolite plus 40 layers of MLI

System 3 - 40 layers of MLI plus 1 layer of Cryolite

As shown in Figure 3a, the comparative k-values were lowest for the system 1, which consisted of only
MLI.

However, taking advantage of available space by adding a layer of Cryolite like with System 2 or

System 3 shows overall heat leak improvement of as shown in Figure 3b (heat leak chart).

This result is

expected as the overall insulation thickness is increased due to the additional layer of Cryolite.
100

100 ,--,...----,--.,.--,-....,---,--,--.-----,

:::J
"C

510

~~
E E

~3"
~ E
CIl

..io

Q.

E
0

0.1

~10 +--~~-~--+-~~~~+--+-~

..

-.

!l

CIl

:x: 1

0.005

1.;-.t~.~~'-=--lt:1=r-1

0.1
0.01

0.1

10

100

Pressure, millitorr

1000 10000 100000

System 3

+--~~-~--+-~--+--+--+-~

0.005

0.01

0.1

1000 10000 100000

Figure 3 (a & b). Variations of comparative k-value and heat leak rate with cold vacuum pressure for fiberglassIMLI composites
and MLI.

5110

Of a particular interest is the question of where to position theCryolite: Is it more efficient to have the
Cryolite closer to the cold side or the warm side?

Considering that MLI performance is governed by a

relationship, one may assume that the MLI should be preferably located on the warm side. However, as the
work of the MLI is to impose the steepest possible temperature gradient and as emissivity values are
reduced at lower temperatures, this assumption may not be valid in all cases. But to accept the warm side
placement assumption, System 2 (Cryolite/MLI) would be expected to perform somewhat better than
System 3 (MLIICryolite).
Cryolite and Cryolite/foil.
Cryolite blanket ( just as perlite) gives minimal protection against radiation heat transfer due to its
relatively low opacity. And while it is not practical to use reflective shields with Perlite, it is a quite easy
application process when used with Cryolite.

While the

improvement from the adding reflective layers seems obvious, an

objective of this study is to quantify the effect of the reflective
layers.

Two layers of Cryolite were compared to two layers of

Cryolite with Aluminum foil (7 .2 micron thickness) on each layer

having the following configuration starting from the cold inner
vessel Cryolite 1F0il/CryolitelFoil:

(Figures 4a and 4b show the

curves of k-value and heat leak respectively).

Figure 5 Photograph of CryolitelFoil configurations.

System 4

System 5

System 6

Cryolite (1 layer)
- Cryolite (2 layers)
Cryolite/foil (2 layer-pairs)

100 . , - - - - - r - - - . , - - - - - y - - - - . , . . - -

100 . , - - - - - . - - - - - , . . - - - - - , - - - - , - -

>-

.~

..,~"

.,.-co

~, +----+---+-~~-+---r-
~ E

:t:

.~

.~

0.

I:

.9

-'"
co E

10

+----4------'I--""""*-;.. ..:....:.....:!L--f--

co

Qj

+----+---+----+---r--

0.001

1 +-~~~~~-I----1-~--+--

.9

_~=---Ir.

0.

E
o
u

co
~

.....

co

~....r__

0.1

000

prest8re, millitor1

100000

o +----4----1----1----+-0.001

0.1

prest8re, miliitofPOO

100000

Figure 4 (a & b). Variation of comparative k-value and heat leak rates with cold vacuum pressure for CryolitelFoil combinations.

At high vacuum levels, the one layer of Cryolite (System 4) gives a slightly lower k-value than the
two layers of Crylolite (System 5). While in the idealized world the thermal conductivity should stay
approximately the same with the increased thickness, the increase of k-value for two layers is explained by
the slight compression that takes place during the installation of the second layer. The compression results
in a density increase, which causes the higher solid conduction heat transfer. At the higher pressures this
difference is negligible, as gas conduction starts to dominate in the soft to no vacuum regime. The heat leak
is obviously reduced due to the second layer of Cryolite insulation. As seen in both charts, addition of the

6/10

reflective layers does indeed improve the thermal performancein the pressure range of interest, below 10
millitorr.

DISCUSSION AND ANALYSIS

MLI and MLIICryolite composite
Results show that indeed the heat flux is lower with the CryolitelMLI (System 2) combination. While the
difference in heat leak between the two systems is evident over the whole range of pressures from high
vacuum 10-5 torr) to very soft vacuum (100 torr), it is more pronounced with increased pressures. This
result is expected as with increased pressure the radiation heat transfer becomes less significant, while gas
60% ~--------------------------------------------

...c:

50%

+----------------------------------------- vs. MLI/Cryolite

conduction

becomes

more

prevalent. The Cryolite being next

to the cold mass is providing more

OJ

~40%

>

protection in this case. Figure 5

+-----------------------------------

co.
.530%

+------------------------

~20%

+------------------

.><

summarizes

percent

of

improvement of heat flux from

oS!

:I:

the

positioning Cryolite blanket next to

the inner vessel.

10%

0%
0.005

0.1

10

100

1000

10000

100000

Pressure, millitorr

Figure 5. Percent improvement in heat leakage rate for the Cryolite/MLI composite.

It is interesting to compare these findings with those for a similar study involving MLI and aerogel blanket
combinations [3].

As reported from this work, aerogel blanket performed better when positioned on the

outside of the MLI than when positioned on the cold mass. One of the possibilities is that the aerogel
blanket density is several times that of the Cryolite, allowing for more solid conduction when put next to
the inner vessel (colder temperatures). Also, Cryolite is highly permeable to gas, where aerogel blankets are
designed to minimize heat transfer by gas conduction or convection.

Thus due to different material

properties of different materials, it can easily be seen why the optimal location changes. These results
underscore the importance of fully understanding the operating environment and requirements for a specific
thermal insulation system.
Cryolite and Cryolite/foil
Figure 6 summarizes the percent of improvement in heat flux over above mentioned pressure range.

The

pressure most commonly used in the Perlite insulated vessels is in range from 1 to 10 millitorr, and as
shown, addition of the reflective layers does offer a quite significant reduction of the heat leak.

Also, the

ease of evacuation of the annular space insulated with Cryolite could offer opportunities for even better
vacuum, approaching 0.1 millitorr.

45%
40%
~

c:

35%

Q)

E
Q)

e>a.

E
-""

co
Q)

30%
25%
20%

...J
~

co
Q)

15%

J:

'#.

10%
5%
0%
0.005

0.01

0.1

Pressure, millitorr

10

100

1000

Figure 6. Percent improvement in heat leakage rate for the Cryolite/foil compared to Cryolite only

CONCLUSIONS
Thermal performance tests were conducted to detennine optimal placement and use of MLI (foil and paper
type) and Cryolite fiberglass blanket within cryogenic storage systems. These tests indicated that it is
preferential to use Cryolite on the cold side of the MLI, and that placing radiation shields within several
blankets of Cryolite drastically improves thermal performance of the insulation system at higher vacuum
levels. These results can be used to define or optimize future systems design and construction techniques.
Evaluation of additional variations in MLI and Cryolite combinations is planned. A practical benefit
of incorporating the Cryolite as part of a MLI-based high-vacuum system is two-fold. First, the Cryolite
layers can allow for better evacuation between layers. Second, the mechanical elasticity (spring effect)
offers protection to the MLI layers to minimize compression and edge effects. The total thermal
performance of the insulation system must be considered along with the mechanical performance
advantages to determine the most effective system for a given vessel or piping application.
hit/lit-

('

/JI al/ k Ii Ill'

REFERENCES
I

1.

'(

Ie

Fesmire, J.E., Augustynowicz, S.D., Heckle, KW., and Scholtens, B.N., "Equipment and Methods for Cryogenic

Thermal Insulation Testing," Advances in Cryogenic Engineering, Vol. 49, American Institute of Physics, New
York, 2004, pp. 579-586.
2.

Fesmire, J.E., Augustynowicz, S.D., Scholtens, B.E., and Heckle, KW., "Thermal performance testing of

cryogenic insulation systems," Thermal Conductivity 29, DEStech Publications, Lancaster, PA, 2008, pp. 387-396.

3. Johnson, W.L., Fesmire, J.E., Demko, lA., "Analysis and testing of multilayer and aerogel insulation configurations,"
Advances in Cryogenic Engineering Vol. 55A, American Institute of Physics, New York, 2010, pp. 780-787.