BASIC DESIGN OF LARGE APPARATUS FOR MEASURING

INSULATORfS THERAMAL CONDUCTIVITY

Shoji Kamiya, Kenjiro Haraguchi,and Eiji Kawagoe
Kanto Institute, Noda plant, Kawasaki Heavy Industries, LTD.
118 Futatsuzuka, Noda, Chiba, 278 JAPAN

Abstract(1)

In WE-NET project, we have been studying a large capacity liquid hydrogen (LH2)storage system for ground tanks and transportation tankers. In this study, developing thermal insulators for a large storage tank, which can provide with the excellent thermal insulation for decreasing boil off rate of LH2 and the optimized strength structure required for large scale, is indispensable. Various insulators having different thermal structures and materials are being devised. Thermal characteristics of devised insulators should be evaluated by the correct method valid for designing the full scale insulator, not by the conventional method for a small specimen. Hence we have performed the basic design of a large apparatus for measuring large insulatorfs thermal conductivity to get its thermal data needed for designing a full scale tank.
This apparatus adopts a calorimeter with a guard reservoir which evaluates the insulatorfs thermal conductivity by measuring boil off rate of LH2. High and low temperature of both insulator sides are 300K and 20K respectively. Diameter and maximum thickness of an specimen are about 1`1.2 m and less than 0.3m respectively. To decrease the edge heat loss of an insulator which would cause the experimental error of a large one, a guard reservoir is equipped with.
This paper describes the basic design of this apparatus and consideration of measurement error due to the edge heat loss.


1 Introduction

To design a large capacity LH2 storage tank (ex.capacity:50,000m3,BOR 0.1%/day), developing a high performance insulator suitable for a large tank is indispensable.Thermal conductivity of insulation material studied before are not valid for the design of a full scale storage tank, because a full scale insulator has thermal anisotropy and complicated structure. And we also can not use the conventional method described in JIS,ISO,ASTM code to evaluate the complicated insulator. Hence we must design the apparatus for measuring the practical thermal performance.


2 Insulators to be evaluated(2)

Various LH2 tanks devised are classified into the following types, concerning their insulators and structures.
Type 1: Vacuum powder insulation + Flat bottomed tank and spherical tank
Type 2: Vacuum solid insulation + Flat bottomed tank.
Type 3: Multi layer vacuum insulation and vacuum powder insulation + Membrane tank
Type 4: Non vacuum powder insulation + Flat bottomed tank and spherical tank
Typical thermal conductivity of insulation are shown in the table 1.

Table 1 Typical thermal conductivity of insulation(W/m K)

Pressure

Muliti layer

Powder

Micro-sphere

PUF

Vacuum

5 x 10-5

1 x 10-3

1`10 x10-3

7 x 10-3

Atmosphere

------------

1 x 10-1

---------------

7 x 10-2


3 Design concept of experimental apparatus

An apparatus for measuring the insulatorfs thermal conductivity is designed according to the following concepts.
(1) Measuring various large specimens
    Any specimens which can be valid for simulation of the full scale insulator
(2) Simulating the environment of the insulator as much as possible
    By considering the environment of insulatorfs usage, experimental conditions such as vacuum pressure , temperature range and etc. are decided.
(3) Decreasing measurement errors
    Measurement errors due to edge heat losses , non-symmetrical condition and etc. should be minimized.
(4) Insuring safety measures in instrumentation and piping system.


4 Basic design of experimental apparatus

The designed apparatus adopts a calorimeter with a guard reservoir in which heat flux transferred through an specimen measured by LH2 boil off rate. Fig 1 shows its schematic diagram. This apparatus consists mainly of a calorimeter reservoir, a guard reservoir, a hot boundary surface plate, a vacuum chamber, a loading mechanism, and other instrumentation. The guard reservoir, which surrounds the calorimeter, shields all heat flux besides heat flux passing the specimen and minimizes the measurement error due to specimenfs edge heat loss. Temperature of the hot boundary plate is maintained to about 300K by means of a heater controller. The load mechanism acts compressive force between the specimen and the hot boundary to decrease thermal contact resistance. The gas pressure in a calorimeter should be maintained to within the rated pressure, because latent heat of LH2 is sensitive to its pressure. In the case of a non -vacuum type specimen, helium gas is supplied to prevent the decreasing pressure.
The mean thermal conductivity of the specimen is calculated by the rate of heat input to the calorimeter. The rate of heat input(Q) and the heat flux density(q) is calculated from the following formulas.
Q = V LƒÏq =Q/A
where
V: Volumetric boil off rate, L: Latent heat of vaporization
ƒÏ: Gas density at the gas flow meter temperature and pressure
A: Surface area of calorimeter reservoir (metering area)


5 Considering edge heat loss(3)

5.1 Edge heat loss influenced by an specimenfs dimensions
n the general guarded hot plate method, the heat loss from the edge of the specimen ,which depends on dimensions of hot plates and the specimen, causes the error of the heat flow measurement. The theoretical error(Ee) can be derived from the following approximate expression.
Ee=(ƒ³T - ƒ³) /ƒ³

where ;
ƒ³T: Actual heat flow rate
ƒ³: Ideal heat flow rate in the unidirectional condition
e : Edge factor, defining specimen edge temperature as Te=T2+e(T1-T2)
d : Thickness of a specimen, b: Guard width
2l : Side length of the metering section from gap center to gap center.
Fig. 2 shows the error affected by dimensions of hot plates and an specimen, assuming the edge number 0.5. The error tends to decrease while the ratio of specimenfs length and thickness increases and the ratio of the metering sectionfs length and the guard width decreases. In our design , the calorimeter and the guard reservoir correspond to the central and the guard hot plate respectively.

5.2 Edge heat loss under vacuum condition
n our design, the specimen is measured under vacuum condition, so that edge heat losses due to non adiabatic conditions is negligible. But as for the specimen having the complicated thermal structure, it is needed to estimate the heat loss due to its unhomoginity. For checking its edge heat loss effect, the distribution of heat flux density about the estimated specimen consisting of PUP and stainless steel is computed by using the two-dimensional finite element analysis ( NASTRAN code). Fig.3 shows the distribution of heat flux density on the 20 K surface. From this distribution, we find that the heat flux increase remarkably at the 55 cm distance, which makes the measurement error. This indicates the necessity of a guard reservoir to cancel this error. The guard reservoir dimension will be decided for considering each specimenfs configuration.


Acknowledgment

This work was conducted in the project g WE-NET(World Energy NETwork)h consigned to the Engineering Advancement Association of Japan(ENAA) from the New Energy and Industrial Technology Development Organization(NEDO).
The author would like to thank the members of the working group for helpful suggestions.


Reference

1) A.Iwata, S.kamiya and E.Kawagoe , gTechnical Development for Large Storage of Liquid Hydrogen in WE-NETh, 11th World Hydrogen Energy Conference 1996
2) 1995 Annual Report on Results,NEDO-WE-NET9553,1995
3) g Thermal insulation-Determination of steady-state thermal resistance and related propertiesh ISO8302 ,1991