..f l"

' Fermi National Accelerator LaboratoryV FNAL-TM--1811

O DE93 005616

PS1 Satellite Refrigerator Heat Exchanger: Failure of theLN2 Heat Exchanger to Low Pressure Helium

Bruce Squires

Fermi National Accelerator LaboratoryP.O. Box 500, Batavia, Illinois 60510

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United StatesGovernment. Neither the United States Government nor any agency thereof, nor any of their

employees, makes any warranty, express or implied, or assumes any legal liability or responsi-bility for the accuracy, completeness, or usefulness of any information, apparatus, product, orprocess disclosed, or represents that its use would not infringe privately owned rights. Refer-enc_eherein to any specific commercial product, process, or service by trade name, trademark,manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recom-mendation, or favoring by the United States Government or any agency thereof. The viewsand opinions of authors expressed herein do not necessarily state or reflect those of theUnited States Governmentor any agency thereof.

November 1992

Operatedby UniversitiesResearchAssocialJonInc.underContractNo.DE-ACO2-76CH03000withtheUnitedStatesDepartmentof Energy

-I,g _ _

Br

4

Disclaimer

This report was prepared as an account of work sponsored by an agency of the United States

Government. Neither the United States Government nor any agency thereof, nor any oftheir employees, makes any warranty, express or implied, or assumes any legal liability orresponsibility for the accuracy, completeness, or usefulness of any information, apparatus,product, or process disclosed, or represents that its use would not infringe privately ownedrights. Reference herein to any specific commercial product, process, or service by tradename, trademark, manufacturer, or otherwise, does not necessarily constitute or imply itsendorsement, recommendation, or favoring by the United States Government or anyagency thereof. The views and opinions of authors expressed herein do not necessarily stateor reflect those of the United States Government or any agency thereof.

0

OPSI SATELLITE REFRIGERATOR HEAT EXCHANGER: FAILURE OF THE LN2

HEAT EXCHANGER TO LOW PRESSURE HELIUM

Bruce Squires

Fermi National Accelerator Laboratory

Research Division Mechanical Support Department

October 9, 1992

INTRODUCTION

OThe PS1 heat exchanger is one of three prototype heat exchangers built by Atomic

Welders before Meyer was given the contract to build the Satellite Refrigerator Heat Exchanger

components. This heat exchanger was first put into operation in July 1983. In November 1991,

this heat exchanger experienced a failure in the shell of heat exchanger ! causing nitrogen to

contaminate the helium in the refrigerator. The resulting contamination plugged heat exchanger

3. The break occurred at a weld that connects a 0.25 inch thick ring to heat exchanger 1. See

figure 1 for the location of the failure. The failure appears to be a fatigue of the shell due to

temperature oscillations. The flow rate through the break was measured to be 1.0 scfm for a

pressure drop over the crack of 50 psi. An ANSYS analysis of the failure area indicates that the

stress would be 83,0_ psi if the metal did not yield. This is based on cooling down the shell to

80K from 300K with the shell side helium on the outside of the shell at 300K. This is the largest

change in temperature that occurs during operation. During normal operations, the temperature

swings are not nearly this large, however temperatures down to 80K are not unusual (LN2

overflowing pot). The highest temperatures are typically 260K. The analysis makes no attempt

to estimate the stress concentration factor at this weld but there is no doubt that it is greater than

1. No estimate as to the number of cycles to cause failure was calculated nor any estimate as to

O the actual number of cycles was made.

o

q

ANALYSIS OF FAILURE AREA

®ANSYS was used to analyze the failure area of heat exchanger 1. For this analysis, both

the maximum-shear criterion (Coulomb) and maximum-distortion-energy criterion (von Mises)

were considered in determining the maximum stress at the failure area. For the cases considered

while performing the analyses, the maximum-shear theory predicted higher stresses and was u_dto form the conclusions made in this document.

Figure 2 shows the dimensions used in the ANSYS analysis of the failure area. The thin

gap was approximately 1/32". Upon removing the ring, it was apparent that the weld did not

penetrate into the gap; only into the ring and the shell. Therefore, the analytical model has a gap

of zero thickness with a length of 1/4". The other analysis parameters are as follows:

- The side of the ring which faces the 8" pipe is treated asperfectly insulated and the other face has convectionwith the shell side helium flow

- The 8" pipe is at a constant temperature of 80°F (300K)

- The 6-3/16" OD shell of heat exchanger .1.,is at aconstant temperature of-316°F (80K)

The effect of changing from forced convection to no convection on the one face of the ring wasonly about 3 ksi at the point of maximum stress.

Figure 3 shows the temperature distribution along the ring. The stresses created by this

distribution are shown in figure 4. The two locations of highest stress are under the ring on the

shell of heat exchanger 1. The first is on the inside of the shell and the other is in the area where

the cracks were found near the weld. The axial principle stresses are plotted in figure 5. This

clearly shows that the high stress area on the inside of the shell is compressive and that the area

near the weld is tensive. The radial principle stresses are given in figure 6. When considering

that the stress concentration factor in the region at the weld is greater than 1, it appears that large

temperature swings at the cold end of this ring contributed to the failure of this shell. Most of tbe

heat exchangers in use do not typically see this large of temperature swings, however they have

had narrower temperature swings with the warm end of the ring closer to 273K and the cold end

changing anywhere from 8OK to 250K, depending on the demand on the refrigerator and on how

well the liquid nitrogen controller works.

In order to determine the allowable temperature swings (those that cause stresses to

remain less than yield), a few temperature ranges were calculated. The stresses at the ring weld

approach 32,000 psi when the temper_.ture is dropped from 300K to 230K at the ring during

cooldown with the helium shell flow at 3(_K. If the shell flow is 225K, a shell temperature of O

II 1

6

3i

150K will cause a stress of 30,500 psi at the weld. This indicates that a temperature difference

O over the ring of approximately 70K will not cause yielding in the shell of the heat exchanger.

Calculations performed at intermediate temperature ranges support this approximation. Based on

this information, changes in temperature on the cold end of the ring should be controlled to avoid

temperature swings greater than 70K in order to increase the lifetime of this weld.

For detailed drawings on heat exchanger 1 and the surrounding heat exchanger 2, refer todrawings 9110-MD- 129233, 9110-MD- 129234, and 9110-ME- 129240.

REASSEMBLY OF HEAT EXCHANGER

In order to prevent another failure in the same area, three fixes were considered: a)

making the ring thinner, b) not welding the ring to heat exchanger 1, and c) adding a "pantleg"

extension to the ring. The first case is over stressed and the second allows more flow to bypass

heat exchanger 2 than we determined to be acceptable. The third case, shown in figure 7,

reduces the stresses to less than the maximum allowable stress that is called out by the ASME

Boiler and Pressure Vessel Code for 304 stainless steel (18,800 psi).

Figure 8 gives the dimensions of the modification to the heat exchanger. The stresses for

this configuration were calculated for two levels of convection from the pantleg and the ring.

O The first for free convection and the other for forced convection. Each convection case led tohigh stresses in different areas. All other parameters of the analysis were performed for the same

conditions as the original analysis. Figure 9 is for the free convection case showing the region of

maximum stress (15,000 psi) on the pantleg near the narrow ring on the shell of heat exchanger

1. The axial stresses are given in figure 10 and the radial stresses are in figure 11. These figures

confirm that the stresses for this case are acceptable. Figure 12 is for the forced convection case

showing the highest stress (16,600 psi) at the warm end of the 1/4" riqg. Even though this is the

location of the highest stress by the maximum shear criterion, the highest axial stress shown in

figure 13 and highest radial stress shown in figure 14 are at the same locations as the freeconvection case.

"['he PS1 heat exchanger has been modified based on this analysis. For details on the

modifications, refer to drawing ME-194492.

liJl .......I

I

4I

m •GHeOut 0 , _ _ _ #'_f--c'#'#'_-xHeatExc__

[

GN2Out _i AreaShownforANSYSAnalysisResults

LN2In Y i _ _. HeatExchanger1GHeIn _ E.aikE_e._,_:_

_, ," _. '! I ,GHeIn _ L_JU " Vacuum

Figure I. Simplified sketch of the warm end of the SatelliteRefrigerator Heat Exchanger indicating the failure area.

O8" sch. 5s Pipe

Thin Gap Between

Ring and Shell

/0.25" Thick Ring

0.125" Fillet Weld

6.1 875" OD x 0.059" Thick Shell

OFigure 2. Dimensions used for the ANSYS analysis of the failure area.

'" ' '"II'P'r'" ' ' 'ill' rl ' li _,,rll mp ,rql,I_'

5b

O

oi

i , I

t

i ot , •

-' 300 K ., 'i "," "," J"

.i o ,i

i t" ,o -o

i i il

,o • .

i i ,

• . i ....... I ........... , ,

.'.'.,.'.'.'.'.'..-,.,.:,.,..,LIi .... [:[:['r:':"l_.,; ., .,.,--:--:-_--:--:--:--:--:--I--:--:--_-:--:-li..........I........:-:.I:..-:...:.:-_ ,'"'_""",,_.........._..........t............I............_;_:;::_:;:_:_".'

.... _ ......... '.'..'..:..'..', ',............... .,._:.j.,..,.., ,- :, ".,.,.,:x;,,.,.,.,.::,:.,, :..................... .--...--,_-!-!X:-i-i- ......!i ,!,!i i !i:]:::!:!:!:!:'"-'--'- _.... ,.t,, ,, , e>", ", "_, ",.,............................................ ,.I,','"," . /r"

-:;,., %o

o i

i a

Ii ,,_H 80Kt

o

l.

,

,

_S',i,,Hear..Exchanger - Stress Analysis -

Figure 3. Temperature distribution derived from the ANSYS analysisof the failure area.

Iii,r,. M" '11" "h'M '11" ' Ir_ ..... r_ III pllr ,111' rlH lll* q"" '1"I1'I_' rl iii , i,illl,.n rl , rl ',, , 'rlr IlIlll

i

, , ,/ , , ,

o.J,i_......io.._.._i...:oo.i...!.o.

,_o.,oo.:..o:o..:o_...:o.,,...:..o:...:o._o.

F ° " "

E '_ " ' ...... '.... ' ..... D Line Stress(psi), _' ' " A "4908

D _./, ," B "14725. . . t . . _ ° ° l _k

', , C -24541r

, , D -34357

.... r .... E -44174

C H -73623

I -83439

PSI hanger - Stress Analysis

Figure 4. Stress distribution derived from the ANSYS analysis of thefailure area (maximum-shear criterion).

J/--r

!

- - AreaShown

P

....................,,' ,,lWr,,....,,"'"rnl'mP11,,IIlll_IVII'l_l:'r'l_i"I'_'ll"1"llr1'Irl,1,"_,,p':',,pFr,,.....mm,p,m,q,_,,irpl,,,l,p,,' ',rlqqilf'r,,'rml'rl"'m_l'lllr'llm_'l'"!llllll"":r!l_!"'"H"'"""='-

J

7

E F

4, ........... , , . * - _- , ,

', , q , '

.'*.... -- -- ° ', ** ....... • ° ° ° . ...... ° . .

Do , ,- 'r - - - ' ....... _. ...... • ....................

_.Ll _. , ' , ' , ' ' '

"1"''',. "'"''':''', .... "'''''"''''.. ....... "'°

--65222

' _ . --48388, " .... C --31554

• 'E , , D =-14720• E -2114

F -18947

' • , G -35781H -52615

I -69449

PSI "ExChanger - Stress Analysis

Figure 5. Axial component stresses derived from the ANSYS analysisof the failure area.

!

/,

- "Area Shown

O

8

----'1I J

/-I

-JArea Shown

.......................................................... ,L,,_......... ,_m._,

• nGHe Out _-_£-E-x_f'__'"_H eat Exc_,

mm

GN2 Out f

lE Area Shown for ANSYS Analysis ResultsLN2 In f- _ | Heat Exchanger 1GHe In _

roll _ ' I- '"

GHe In _ _ __...__ . .

Vacuum

Figure 7. Simplified sketch of the warm end of the SatelliteRefrigerator Heat Exchanger showing the modificationanalyzed with ANSYS.

8" sch. 5s Pipe

O 0.2 5" Thick Ringf

CI.}! "_--¢--

(j v

I-- r-F-

=,.

_ 3"O u')• O

o cSxX

o __u'b u'3t',.- D,-OO Orb,-- '.::t"

u5 u5_' Ali Fillet Welds are 1/16"

T

r except on the 8" Pipewhich is 1/8"

O Figure 8. Dimensions of the modification to the heat exchanger.

ut, I]lp, " II 'rl rl, , ,l,_rlllll illlI , , , II, IIpIH ,,Irl lit ,, , ,,,llr_,,Ir "'l,l_ltl lP ' mm' lr,,, m , ,11 'lr " riP' al " _1'"' II irl, tl',, li ii , , iiii .... , ,,, ..... rUl,ll_r

10

E D CCD

iiii!ii "iI ii6

ii llli'_" C

. . °

c c Une Stress (psi)

A "899.503 O

........ C "4497D "6296E "8095

- F "9894

H "13492

PS 'II_a_" _]_'changer - Stress Analysis I -15291

Figure 9. Stress distribution derived from the ANSYS analysis for themodification of the heat exchanger with free convection(maximum-shear criterion).

'-- q 0: Area Showni, It4 .... .I

II

E F G F E D C

A

i ,/

.,_ .... _.,B

c - ;- - -' ...... _.

z', : i ' 1 : : , A --12480

.'" C "-5999F D --2758

E "482 . 80 7

F "3724

G "6964

H "10205

I -13446

PSI '_ _1" _changer - Stress Analysls J

Figure 10. Axial component stresses derived from the ANSYS analysisof the modification to the heat exchanger with freeconvection.

12r

i i

|

• . . • ...... . . ° .

. ° ° • . ° °

I

io

i_2 E _ .....

,/',/. G H H I _.__: X'

,_!__,_'- . -, - . ,- -..._.j_, - .:..- .,__/i_._'"..,<.._..>. . . . . ,

B --3111_----_-_ D" ....'.....' - " '...... ---- ........ C --I 704' /L_ _ , ,

' ' ' ' ' ii •..... .. __ , z -ii11Y ' ' ' ":_.... ,.... F -2518

.... . . . H -5332

, C I -6740

'-- ' D

0i

|

PSl 'l_6a'C"_:_changer - Stress Analys_s l

Figure 11. Radial component stresses derived from the ANSYSanalysis of the modification to the heat exchanger with freeconvection.

|-' - OI 'i ,Area Shown

!l

13

!

14r

-E E E F

--.v.-i_-\B

Line Stress(psi)

,", D "-1391

, ; , , qp. . . i .. ° + _

__ i i ........: ...."...... "'-" "_ -++_'_°'_"i +:-!_-PSI hanger - Stress Analysis -

Figure 13. Axial component stresses derived from the ANSYS analysisof the modification to the heat exchanger wit.b,forcedconvection.

I-

lull AreaShown 0L-l- _ ....

lt +

154

, ...... , w. ...... ,,

. .., . . .

' b ....... .

'\..:... , ,

, D D ..... '-1 * * t t_

', H H Z _: Line Stress(psi), --- -,_-_...>,_ .... !-. . ,

''_ i B --1551

" " " - - - C --825. 63S

E -624 . 608F -1350G -2075

.....""" -'. - •'.... H -2800

, :- ._, . , z -as2s

- c.]....

PSl 'Hea_-Exhhanger - Stress Analysis -

Figure 14. Radial component stresses derived from the ANSYSanalysis of the modification to the heat exchanger wlthforced convection.

Ui

°- -]

' Area ShownI

I1,, ,J

1