e > ;s Hat P o s X-822 e e T OAK RIDGE NATIONAL LABOR Operated by g@ggflmusmu AUTHDRIZED UNION CARBIDE NUCLEAR COMPANY Division of Union Carbide Corporation 0 R N l_ = CENTRAL FILES NUMBER Ouk Ridge, Tennassee 58-8- x DATE: August 13, 1958 CUPY’NO-ii;y SUBJECT: Screening Tests of Mechanical Pipe Joints for a Fused Sszlt Reactor System TO: Distribution FROM: W, B, MeDonald, E. Storte, A. 8. Olson ABSTRACT The testing and evaluation of three types of mechanical Joints in a circulating molten fluoride salt system, at temperatures up to 15009, has been accomplished. ‘ The feasibility of these jolnts for use in a large scale molten sall system 1s discussed, Design criteria and operating techniques are described. Measurements have been made of the leakage rates of helium through the joints. The effects of thermal cycling, atmospheric exidation, salt corrosion, and thermal stresses have been noted. Disassembly and reassembly procedures are described. e LEGAL NOTICE — This repart was prepored as an occount of Government sponsored work. Meither the United States, nor the Commission, nor any person acting on behalf of the Commission: A. Mckes any warranty or representation, express or implied, with respect to the sccuracy, completeness, of usefulness of the informotion contained in this report, or that the use of any information, apparatus, method, or process disclosed in this report moy not infringe privately own=d rights; or B. Assumes any liabilities with respect to the use of, or for damages resulting from the use of any information, apparatus, method, or process disclosed in this report. As used in the above, '‘person acting en behelf of the Commission' includes any employee or contractor of the Commission to the extent that such employse or contractor prepares, handles of distributes, or provides access to, any information pursuant te his employment or contract with the Commission. 1.0 208 5.0 - = & 5.0 i o Table of Contents Totroduehblon o ¢ o o o o o o o o o o o s s o 2 o 0 o o o o » BUMMBYTY o o o o o 5 o o s o 6 o o o » o o s » o o o o s o a Description of Test Joints o & o ¢ 4 v 6 o 4 o o & o o 2 s Criteria for Jolinh Performant® . v o o s o o o v ¢ & @ WM » }J N & £ Frasge Flangs Joinb . o 4 5 ¢ o 6 o 6 ¢ o 6 0 0 o 5 5 5.3 Indented Seal Jolnb o o 4 o ¢ ¢ o o ¢ o ¢ o s o o & o o fio% Gafit S@&l J@int * % B & e ©°© B » H e r e H& 4 & © ° " o ° 5 The Test LOOD 2 o o o o o o o 8 o 0 s o ¢ o o s s o o » 2.6 Tostrumenbabion o o o v o v s 0 6 s 6 o 0 e s o 5 » o Test ProcedU'® o o o o 2 o o o o o o o o« o o o o o o o « o o 4.1 Freeze Flange Joint . . o ¢ 4 o 6 o o 6 o 5 o o o o o o L2 Coast 88l JOIOE v v o v o o s 0 0 o 0 s o e b e e e e L.rofi Lnd@fitfi‘fi.sa&lggintaaoosoooo-oo-coona ol lzougslon of Besult8® . o o o o 4 4 ¢ o 6 6 o o o o o o o o ol Fresza Elaflge Jfliflfi g 8 & * 8 8 & & © B # S & 8 e & ® 0 .2 Cafifi Seal JOint & ® ¢ © & © & & % & & » O © s & © @ 0 5o§ Ind@mt@fi S@&l J@int " & ® ° o © & s & ¢ £ & o & & & B 0 7 - 0 . “Fufiflr@ leStS ¢ © © © © & 9 B8 8 & 4 & & © e & & & & e w s 0 Preliminary Mechanicel Joint Specification . . .« . + « ¢« ¢ & Gravhs of FPreeze Flange Thermal Cyeling Tests . o o o o o o o th Graphs Cast Seal Flange Thermal Cycling Tests . « o o o & rraphs of Indented Seal Flange Thermal Cycling Tests .+ o & Report oa Metallurgical Examination of the Cast Seal Flanges > Report on Metallurgical Examination of the Freeze Flange Joint Assembly Drawings of the Mechanical Joint and Loop « & o o & o 5 W 5 j-t 66 Tl Gl 100 105 107 1.0 Iotroduction To raduce down-time during maintenence operations on the molten salt pover reactor system to a minimum, the phil@sgphy has been adopted that all system components on which diract mainbtenancs is not possible shall be removable and replaceable with sparss, by reamote manipulation. Re- pair work on falled compomants shall be accomplished in suitably equipped independsnt facilities, It is preregquisite to ths removal ard replacemsnt of system com- ponzctbs by remote manipulation that a reliable msethod be developed for parting system pipipng and rejoining it to its originsl integrity. Until prasent experimsntal efforts to perform critical welding remotely have uceeedaed to the point wherse they can be applied with confidence to the maintenasce of a reactor system, the use of mechanical Joints for this porpose most be considered. The cobjective of the work reported here was to scresn the various concepts of mechanical Jolnts offered by individuals or groups within the Reactor Projects Division, and to selact for develop- ment those showing promise of successful application, Thrae small scale mechanical pipe joints of alternative types were tasted in a fused salt pump loop, The Joints werse cold leak checksed on a mass spactrometer lsak detector kefore installation in the loop, vare then lustalled and subjected to a saries of thermal cycles betwaen 1100 and 1300°F, removed from the loop and leak chackad., The Joints were then parted, remads; and lzak chacked again, An indicated lesk rate of less than 1 x .1.()@7 ce of halium per second was reguired for acceptable parformance at =ach check. Of the three joints tested, the "freeze flange" Joint was found sat- isfactory for immediate development, the "indented seal" Joint was found promising if modified, and the "cast seal” joint was found not suitable for development. In accordance with these results, further tests of the small-scale "freezs flange” and "indented seal’ joints have been initiated with sodium as the process fluld, and a pair of "freeze flange" Joints in 4" pipe size has besn fabricated for testing in a large fused salt system. Results of these tests will be reported separately. Descripbion of Test Joints 3.1 Criteria for Joint Performance For the screening tests, the following criteria were used, based upon the Prelimiusry Mechanical Joint Specification (see Appendix 1) prepared by the MSEP Group: | (a) The joints shall be fabrieated of materials compatible with fused fluoride salt fuels. (b} The joint leak rate, cold, before installation in the test loop, shall be less than 1 x 1077 cc of helium per second, as in- dicated by comparison with a standard leak used in conjunction with a mass spectromeler leak detector. (¢) The joint shall be installed in a test loop circulating fused salt and subJjected to a minimum of 50 thermal cycles between the temperatures of 1150 and 13%00°F, (4) No fluid leakage shall be acceptable, (¢) Upon completion of the thermal cycling tests, the Joint shall bs cut from the loop and cold leak checked {sse 3.1.Db). (£} It shall be demonstrated that each joint can be parted and remads accaptably leak tight, without requiring more than super- ficial cleaning, etc., such as could be psrformed Dy remote mayipulatlon, %.,2 Fresze Flange Joind The principle of the frozen ssal mechanieal jolunt, or frseze flangs, is illustrabed in Fig, 1, The frozen salt seal is formed in the gap vetween the flange faces. The ring insert provides a labyrinth re- striction to the passage of salt into the gap. The labyrioth is im- portant only when the test loop was filled with salt, before the frozen saal was formed. After the frozen seal is forwmad, 1t retains the molten salt in the process stream. The metal seal ring provides s gas-tight seal , walch retalos fission gases. For an effectlive gas seal, it was found necaesseary to manufacturs the seal ring to close toleracces., The 0. D, and 1. D, of the ring were within 0.004% inches on a diasmeter of seven inches, and concentric within 0.001 inches, The finish requirsd wvas sixteen microinches. The same Lolerancaes were reguired in making tha geal ring groove in the flange faces, An alr chnsnpel, in each balf of the Tlangs asssmbly, provides cool- ing for the seal ring and insurss the formation of the frozen salt seal. Toe flange hass a thin cross saction between the flange hub, which is welded to the loop tubing, and the seal ring, near the outer edge of the flangs, This was designed to pressat a small cross sectional area for radial beat conduction. FPour webs add strength to the flange and re- duce warping. On the test installation, sight gulck-opsuing toggle clamps wers usaed on each flanges assgembly, as shown in Fig. 2, toe provide the necessary UNCLASSIFIED ORNL-LR-DWG 27825 1-SOFT-IRON OR COPPER SEAL RING +H¢ GAP (~Yie in) " AIR CHANNEL FOR COOLING FROZEN-SALT SEAL SRR ERRD \\\\\\xx IR S S I R IR X D K _ /WELD OF FLANGE TO LOOP TUBING —s=— SALT FLOW 7 U RIE % \\ / \/é\\}\gfi:—mm INSERT TO PROVIDE LABYRINTH /§ FOR SALT LEAKAGE 3 FROZEN-SALT SEAL / X /b NARROW SECTION TO REDUCE HEAT T TRANSFER FROM THE MOLTEN SALT ; IN THE LOOP TUBING IR 7N £ -’:x. b SEAL ; 243 INDICATES REGION OF TRANSITION FROM LIQUID TO SOLID SALT Fig.1. Cross Section of Freeze Flange Joint. UNCL ASSIFIED PHOTO 30204 ¥ i ‘ Y L p 3 ' 1 ] . 3 i Fig. 2. Freeze Flanges Installed in Inconel Test Loop. (1) Air Cooling Lines (2 places); (2) Calrod Heaters (1 place); (3) Toggle Clamp (1 place); (4) Support for Flange (1 place); (5) Clam Shell Heater (1 place); (6) Hy-Temp Insula- tion (1 place) force to maintain the gas leakage rate below the maximum allowed. Flg. 2 shows the assembled flanges with clam shell type heaters, located on the loop tubing immediately before and after each flange as- sembly. Referring to the figure, the various parts are as follows: (1) Air cooling lines, (2) Calrod heaters leads, (3) Quick-open toggle clamps, (4) Thermocouple leads, (5) Clam shell heaters, (6) Insulation. There Were no heaters nor was there any insulation 6n"the flange assemblies. The flanges were made of Inconel. The seal rings for the first test were of stainless steel. In subsequent tests, soft iron and an- nealed copper seal rings were used. 3.5 Indented Seal Joint This mechanical joint is illustrated in Fig. 3.. A dead-soft an- nealed metallgasket is used to seal the two flange halves., A raised tooth of V-shaped cross section, machined on the face of each flange, indents the flat metal gasket to provide a tight gas seal. Actual parts are shown in Figs. 4 and 5. A thin cross section between the flange hub and gasket, near the outer edge of the flange, presents a small cross sectional area for radial heat conduction. On the test installation, four standard design "C" clamps were used on each flange assembly to provide the force necessary to indent the gasket and maintain a gas leakage rate below the maximm allowed. Fig. 6 shows an assembly with clamps. The flanges were made of Inconel. For the first test, a dead-soft annealed copper gasket was used in one flange assembly, and an annealed TC 30-36 N INDENTATION SEAL FLANGE TEST NO.1. UNCLASSIFIED ORNL-LR—-DWG 31974 THERMOCOUPLE LOCATIONS TC 29-35 / ..\ /-—— TC 28-34 TC 32-38—~ /\ a— 1C 27-33 ///////.///// //\ \\\\\’\\\\\\\ e e SALT FLOW (L L L Ll L Ll s L NN N N NN N NN N NN [ SN N AN LRAISED TOOTH ) FLANGE A-IRON GASKET TC NO. 27, 28, 29, 30, 31, 32 FLANGE B-COPPER GASKET TC NO. 33, 34, 35, 36, 37, 38 S~ WELD TO LOOP TUBING — ANNEALED METAL GASKET Fig. 3. Schematic Drawing of Indented Seal Mechanical Joint, UNCLASSIFIED PHOTO 30662 2 INCHES Fig. 4. Assembly of Indented Seal Flange Before Testing. 2 INCHES Fig. 5. Parts of Indented Seal Flange Before Testing. UNCLASSIFIED PHOTO 30663 _0 [-. Tl UNCL ASSIFIED & PHOTO 31519 P g o Sy »,w—-*w-flr - ' r - ‘ s 1 ' ty i [ » 3 L : : SR, . \ Fig. 6. Indented Seal Flange Shown Assembled with "'C'* Clamps. - 12 - Armco ircn gasket in the other assembly. For the second test; an an- nealed Nickel "A" gasket was used in one assembly and a nickel-plated Armco iron gasket in the other assembly. 3.4 Cast Seal Joint A cross section of this mechanical Joint is shown in Fig. 7. Four stainless steel bolts provide the mechanical strength for the Joint. A metal insert, shown in the figure before being fused, is used to provide a salt and gas seal. The metal insert is fused between the flange faces before the test loop is filled with salt. The seal whose melting point is above that of the fuel salt, is in the solid state during operation of the test loop. The sealing surfaces on the flange faces are flash copper plated, and then nickel plated, to promote good wetting by the sealing alloy.l Because of its geometry, this Joint must be installed in a vertical run of pipe. The test flanges were made of Inconel. A cast silver seal was used in one flange assembly, while an alloy of T2% silver and 28% copper was used in the other flange assembly. An assembly of the flange is shown in Fig. 8. 3.5 The Test Loop The test loop consisted of 25 feet of one-half inch O, D., 0.045 inch wall thickness; Inconel tubing arranged as shown in Fig. 9. The various parts are identified as follows: (10) Location of flanges under test, -(11) Calrod heaters wrapped with stainless steel strip, (12) Clam lohapter 2.1 MSRP Quarterly Progress Report, Jan, 31, 1958, ORNL 2474 UC-81 Reactors - Power. _13_. UNCLASSIFIED ORNL-LR—-DWG 31972 SEE FIG.13. FOR DIAGRAM OF INCONEL LOOP FLANGE WITH SILVER INSERT : TC NO. 19, 20, 21, 22 P FLANGE WITH COPPER-SILVER ALLOY INSERT : TCNO. 23, 24, 25 26 TC 22— 26 ~——a \ A \ NS | SR L -—TC 21~25 TC 20-24 —_—————- TC19-23 —= g SEAL MATERIAL INSERT SHOWN BEFORE BEING FUSED TO FORM SEAL AN 1 15 0 1 T —— INCHES w7 QZZZZ/W// /" . WELD OF FLANGE TO LOOP TUBING Fig. 7. Cast Seal Flange Showing a Cross Section View and Thermocouple Locations. Fig. 8. Assembly of Cast Seal Flanges. UNCLASSIFIED PHOTO 43636 _VL_ B UNCLASSIFIED HOTO 31674 Fig. 9. Inconel Test Loop. (10) Indented Seal Flange (1 place); (11) Calrod Heaters Wrapped with Stainless Steel Strip (2 places); (12) Clam Shell Heaters (1 place); (13) Unistrut Frame (1 place); (14) LFB Pump (1 place); (15) Salt Sump (1 place); (16) Metal Trough Support (1 place); (17) Hy-Temp Insulation (1 place) ’ - 16 - shell heaters, (13) Unistrut frame, (14) LFB pump, (15) "Hy-temp" in- sulation, The molten salt mixture was circulated by means of an LFB-type centrifugel pump. During tests 1 and 2, on the freeze flanges, two straight sections of the loop, each about nine feet long, were used ag resistance heaters; A heavy current was passed through the tubing sections. Other sections of the loop were heated with calrod or clam shell heaters. The entire loop was preheated, prior to filling with salt, Tests 3 and 4 on the freeze flanges were conducted in a new location where resistance heating was not available, and the entire loop and pump were heated by means of calrod and clam shell heaters. A sump tank, connected to the point of lowest elevation in the loop, was used to fill and drain the loop of molten salt., The LFB pump, installed at the highest point in the loop also served as a surge tank., The pump was supplied with an oil lubricating system. A helium supply was connected to the loop for initial purging of air before heating. It was also used fio pressurize the sump for filling the loop with molten salt. The various components.at the pump end of the loop are shown in Fig. 10 and are identified as follows: (1) LFB pump, (2) Pump motor and clutch, (3) Salt sump, (4) Water cooled oil storage tank, (5) Lubricating oil pumps, (6) Helium supply regulators, (7) Helium bubbler, (8) Helium pressure gauges, (9) Lubricating oil flowmeter. Salt flow in the loop was controlled by a variable speed magnetic- type clutch and induction motor. V-belts connected the motor and clutch to the LFB pump. -17- UNCL ASSIFIED PHOTO 31360 Fig. 10. Inconel Test Loop - Pump End. (1) LFB Pump (1 place); (2) Pump Motor and Cluteh (1 place); (3) Salt Sump (1 place); (4) Water Cooled Oil Storage Tank (1 place); (5) Lube Oil Pumps (1 place); (6) He Supply Regulator {1 place); (7) He Bubbler (1 place); (8) He Pressure Gages (1 place); (9) Lube Oil Flowmeter (1 place). - 18 - All components of the loop in contact with the molten salt were fabricated of Inconel, The loop design was similar to a "standard" design established in the Experimental Engineering Department for their corrosion testing program. The entire loop, including the pump, was mounted on a test stand, also of "standard" design, made of Unistrut. The stand was mounted on wheels for easy transport between various construction stations; such as the weld shop and X-ray room. All welds in contact with the molten salt were given a dye check and X-ray examination according to departmental specifications for critical welds. The Inconel tubing used in the loop had received a thorough dye check inspection prior to construction of the loop. Twenty thermocouples were spot welded to the loop tubing and pump, per Dwg. No., SK-CKM-1623. Additional thermocouples were attached to the flanges, Precautions were taken during construction to use clean tubing and parts, and especlally to avoid contamination of the inside of the tubing and pump. 3,6 Instrumentation The chart below lists instruments and controls used on the test loop for all of the flange tests. Measurement Primary Element Indicating Element Control (a) Temperature Chromel -alumel Temperature Re- Tests 1 and 23 (Freeze ' thermocouples corders (1) Flange) Resistance Heaters, Transformer (2), Clam shell heaters, Variacs, Tests 3 and 4: (Freeze Flange and other flange tests) Clam shell heaters, Variasce (3) (b (c) (4) - 19 - Measurament, Primary Element Indicating Element Control Salt levels In Metal probes and 110 V. lights pump and sump 110 V. powsr supply Helium pressurs -~ ~ = Bourdon Gauges - - =~ Pressure regu~ for LFB punp lators (4) and sump Iubricating oil ~ = = Flowratory - = = Centrifugal pump flow for LFB and motor (5) pamp Pump spesd - - = BStrobotac - - - Motor and clutch (6) Notes on instruments and conbtrols list: (1) (2) (3) (&) Bristol recorders 0-2000°F range for Tests 1 and 2 (Freeze Flange) Brown recorders 0-2000°F range for Tests 3 and 4 (Freeze Flange and other flange tests) Reslistance heaters consisted of two lengths of 1/2" 0. D. x 0,0hs" wall Inconel tubing each approximately 9 feet long. Hevi-duty transformer and saturable reactor control rated at 110 KVA, which allows a maximum current of 2740 amps output on the %0 volt tap; 3 KW is the minimum leaskage power; 20 XW output can be controlled, but it is difficult to provide steady control of a lower pover, Wneelco conbtroller Vicksers magnetic amplifier Open elemsnt tType cylindrical clam shell heaters. Variacs, 110 V., and 220 V. Fisher Governor bleed and non-bleed types of pressure regulators, 0-35 psig, Models 67-15 and 67-16, - 20 - (5) Eastern Industries Model D-11 centrifugal pump with 1/5 H. F., 110 V. motor. | (6) Louis Allis Adjusto-Spede irduction motor, 5 H., P., 3 phase, ~~3600 rpu, Dypamatic magnetic-typs clutech, 400-3400 rpm at full load. 4,0 Test Procedure bl Fresze Flange Jolnt Two sets of flanges were asgsenmbled with seal rings prior Lo in- gtallation in the loop. Each flangs assembly was then laak tested using a helium leak dstector. To leak test, the tubing from one zud of a Tlangs assambly was plugged and the other end comnected to a vacuum pump., A plastic bag was then placed arcund the flange assembly and kept {1lled with helium during the leak test, Hellum leak rate tests wers madse on both flange as- semblies prior to installation in the loop for Tests 3 and 4, This work was dore by the Instrument Department and is described in detail in thair grep@rt No, 58-1-20, Both flangs assemblies wers then walded into one and of the loop as shown in Fige. 2 avd 9. Helium was circulated through the loop bafore prseheating, to ra- move air. A helium pressure was malntalned on the loop and pump duriag heating and filling, Clean fluoride salt mixture was then travsferred to the sump tank. When the temperature of all parts of the loop reached 1300°F, helium pressure was applied to the sump tank to force the salt into the loop, The pump wasg opsrated at a low speed durdng the filling opsration, The loop was vented through the pump belium iclet, as the salt displaced the helium. Salt circulation was chtalned a few minutes labter, when the pump speed was incrsassed to 2500 rpm to produce about 2 gpm flow., A halium w 21 e blanket was maintained at 2 to 5 psig on the salt surface in the pump during testing operations. Tha heaters wera then adjusted to obtain 1300°F on all parts of the loop. During the loop filling operation, cooling air was circulaeted in the air channels built into the freeze flanges. A small alr flow of ap- proximately 15 ¢fh was necessary to keep the seal ring area cool. The loop was filled rapldly to insure establishment of frozen seals batwesn the flange faces before the molten salt could raise the temperature of the flanges. After a period of isothermal operation to make certain there was no salt leak, the thermal cycle testing waes begun. Temperatures of the salt were measured by means of thermccouples located on the loop tubing at the inlet to each flange assembly., Temperature measurements were teken at varions points on each flange assembly. The salt flow was maintained at a steady rate of 2 gpm. The pump speed was checked periodically with a Strobotac and the flow in gpm cbtained from a pump calibration chart. A schematic diagram of the Inconel lcop during Tests 1 and 2 is shown in Fig, 11. During these tests, temperatures were measured at the points indicated in Fig, 12, Fluoride salt, Fuel No. 107, was circulated in the loop durlng the first test. The salt temperature was cycled 50 times betwaen 1150°F and 1300°F, and then the locp was operated isothermally for SQveralidays at 1300°F. The average cycle time was 44 minutes. A cycle is defined as a hemperature variation from 1300 to 1150°F and back to 1300°F. During the second test, In which fluoride salt, Fuel No. 30, was used, ths temperature variation of the salt wasg from 1300 to 1500°F and back to -.22_ UNCLASSIFIED ORNL-LR-DWG 31973 T FREEZE-FLANGE JOINTS o ~ PLANGE B (AX!S HORIZONTAL) ] :J "}.,,( ....... HEATER CONNECTION LUG —— | .O0P TUBING - INCONEL, Y-in. 0.0 ,x0.045-in. WALL SALT FLOW SALT PUMP I | ./ : ) e SALT FLOW ~23- UNCLASSIFIED ORNL-~LR ~DWG 27897R oy~ FLANGE B 15 8N £4 - ‘ 24 P 23— 8N 221N — _ 2! AN NOTE: NS SEE FIG.{1.FOR DIAGRAM Nl o6 OF INCONEL LOOP. Nl o7 AN =2 T FLANGE A 28 Fig. 12, Schematic Drawing of Thermocouple Locations on the Freeze Flanges for Tests 1 and 2, ~ Py - 1300°F. The number of cycles was reduced to 30 and the average cycle time was 29 minutes. After Tests 1 and 2, the {lange assemblies were disassembled and reassembled to demeonstrate ease and gpeed of handling, A motion picture film was taken of these cperations. The time required for these operations ig listed in Table 1, Both of the flange assemblies were removed from the loop before starting on theinext test., Additional machining work was performed on the flange faces to produce a better finish on the seal ring grooves. A schematic diagram of the Inconel loop for Tests 3 and 4 is shown in Fig. 13. Flu@ride_salt, Fuel No. 30, was used in these tests., The temperaturss were measursd at the points indicated in Fig. 14%. The salt flow rate was, agaln, 2 gpm. During Test No. 3 the salt temperature was cycled 50 times between 1100°F and 1300°F. The average cycle time was 60 minutes. During Test No. 4 the temperature variation of the salt was from 1100 to 1300°F and back to 1100°F; the number of cycles was 25, and the average cycle time was 02 minutes. After the thermal cycling tests, both flangs assemblies were removed from the loop for leak testing. Both assemblies were disassembled and one flange assembly was reassembled using a new copper seal ring. This assembly was again leak tested. 4.2 Cast Seal Joint Twe sets of flanges were assembled with metal inserts prior to in- stailation in the loop. One flangs assembly contalned a pure sllver in- sert, the other assembly contalined an insert of copper-silver alloy. Both Table No, 1 TIME REQUIRED FOR ASSEMBLY AND DISASSEMBLY FOR FREEZE FLANGES . DISASSRMBLY Operation Performed Time Required 1. Dump Salt from Loop 15 min, 2. Cooling Flanges to Room Temperature b hrs, 3. Disconnect Thermocouples - Heaters - Airlines 1.5 min, 4, Remove Clamps 3.5 min, | | Total % hrs. 20 min, (See Figs. 21 and 22 for view of disassembled flanges) :; “ 2 ASSEMBLY : 1. Cleaning Flanges 5 min,. 2. Replacing Flanges and Clamps - 5.5 min, 3. Reconnection of Thermocouples - Heaters - Airlines 5 min, ' Total 13.5 min, (See Fig. 2 for view of assembled flanges) Note: The time periods listed are for handling two freege flange assemblies with twe men performing the operations. The time periods were taken from a motion picture of the operations. FREEZE-FLANGE JOINTS (AXIS HORIZONTAL) FLANGE /.\/ SALT FLOW CAST-METAL -SEALED FLANGE JOINT ADDITION TO LOOP T ~ | L _! P 1 1 P B 41 "CAST-METAL.-SEALED FLANGE JOINTS (AXiS VERTICAL) UNCLASSIFIED ORNL—LR—DWG 31974 FLLANGE B / —T HEATER SECTION CLAM SHELLS 1. ] T I , LOOP TUBING - INCONEL, ;. . :I[: /2 1. OD, 004‘5 in. WALL T SALT PUMP - ] N \ ’ “ \\ r__l__ i = —— L 7] \ T / | . S \ ( —— 1 | I ! | | N -~ SALT FLOW Fig.13. Schematic Drawing of Test Loop for Freeze Flange Tests 3 and 4 and Cast Seal Flange Tests 1and 2 -27- UNCLASSIFIED ORNL-LR-DWG 31975 5 2 S lee— FLLANGE B8 28 — [ zgml,i;v —s—— SALT FLOW fij 30 33—2; 34— g 3 IS NOTE SEE FIG.13 FOR DIAGRAM OF INCONEL LOOP. IF 7/%‘3—:% FLANGE A — 36 Fig. 14. Schematic Drawing of Thermocouple Locations on the Freeze Flanges for Tests 3 and 4. - 28 . flange assemblies were then welded into one end of the loop in the location shown in Fig, 13. A special furnace was placed around each cast seal flange for melt- ing the metel inserts and to provide heat during loop operation. The loop and flangss wers fill@d with helium prior to heating. The sssembly with the silver insert was raised to 1830 - 1975°F ard held at this temperature for cas hour to insure complete fusion of the silver., The melting point of silver is 1760°F. The other asssmbly was trested in the same fashion, excapt that the temperature was raised to 1570° - 1670°F. The melting point of the copper-silver alloy is 1435°F. The loop and flanges were then filled with helium at a pressure of ap- proximately 25 psig to check for leakage through the cast metal seals. The pressure did not drop after several nours, snd the flanges gave no indication of sany leakags when checked with a scap solution coated over the flange joint and around the bolts. The loop was then placed in operation for thermal cycle testing as de- seribed in Section 4.1, During these tests, temperatures were measursd at the points indicated in Pig. 7. Fluorids salt, Fuel No. 30, was circulated in the loop during both tests on the cast seal flange. During Test Wo. 1, the salt temperaturs was cycled 5C times between 1300 and 1100°F. The averags cycle time was 60 minutes, During Test No. 2, the salt temperaturs variation wes from 1100°F to 1300°F and back to 1100°F. The number of cycles wag 25 and the averags cycle time was 62 minutes. Tests 1 and 2 on the cast zsal flanges wers run simultanecusly with Tests 3 and 4 on the freeze flanges. After the cycling tests, both flange assemblies were sszparated using the special Turnmaces to melt the cast seals Juring the pariting operation, - 29 Stainless steel rods were tack welded to the flanges to pull them apart, while the seals were in the molten state, A diagram of this apparatus is shown in Fig. 15. | The top halves of each of the two flange assemblies were thoroughly cleanad by placing them in a hydrogen furnace at 1900°F for about two hours. The bottom halves of both assemblies were glven a superficial cleaning using a brush and a vacuum cleaner. The bottom halves of the flanges represent those parts of the mechanlical Joints which could not be removed from a reactor system for cleaning and would have to be cleaned re- motely. Both flange assemblies were then reassembled using spring-loaded tie rods to apply the necessary force to push the flange halves together. A diagram of this apparatus is shown in Fig. 16. The special furnaces were used arcund each assembly to melt the cast seals during the reassembly operation, Both flange assemblies were removed from the loop and sectioned for metallurgical examination. This work was done by the Metallurgy Division, (See Appendix 5 for thelr report.) 4.3 Inderted Seal Joint Twe sets of flanges were assembled with metal gaskets priocr to in- stallation 1n the locop. Four "C" clamps were used on each flange assembly. The clamps were each tightened with a torque of 80 - 90 foot pounds. The load ofi each gasket was approximately 12,000 pounds per inch of gasket clrcumference, Fach flange assembly was leak tested using a helium leak detector. DBoth assemblies were then welded into one end of the loop as shown in Figs. 9 and 17. _30, UNCLASSIFIED ORNL—-LR—-DWG 31976 /TACK WELD IN SEVERAL PLACES 7(7'{ S Y4—in. DIA INCONEL RODS (TWO) | | | f e LR ca TACK WELD RODS TO FLANGE AND ADAPTER (AT LEAST 6 TACK WELDS ON EACH ROD) ,__,7 Ya~in. DIA INCONEL RODS (TWO) WELD RODS TO PLATE /4/4-m. INCONEL PLATE, 8 x 8in. ] \'" FILLOOR LEVEL Fig. 15. Sketch of Extension Rods for Disassembly of Cast Seal Flanges. UNCLASSIFIED ORNL~LR~-DWG 31977 ///—5/3—in.-DIA BAR 316 STAINLESS STEEL 24-in. LONG W) COMPRESSION SPRING, STEEL t-in.OD X 2-in. LONG — T ™ [ by —3/4~in.~SCH 80 STAINLESS STEEL PIPE ] h - I y by H - TOP OF EXISTING Lo / HEATER BOX , I Il 1|i | IR 1 ! |l ! e It (! '| | :: i ! s D |1 | f |l [ I i ]! [N | 1 ¥ CAST SEAL :I Lol | FLANGE |: :: ; ASSEMBLY 1 I Ll 5I| . " I I i ' | i ! | | | | l| ' I 8 ! | | | | ! i: | ! | | ; i | | | : I :’ | | ' | .I! l! | ' : : S C= ] ' Fig. 16. Sketch of Spring-Loaded Tie Rods for Reassembly of Cast Seal Flanges. UNCLASSIFIED ORNL—-ILR-DWG 34978 INDENTATION SEAL FLANGES (AXIS HORIZONTAL) FLAMNGE A :] ] HEATER CONNECTION LLUG = |LOOP TUBING ~ INCONEL , Y/,=in.0D, x 0.045~in WALL SALT FLLOW | el ——. SALT FLOW Fig.1 7. Schematic Drawing of Test Loop for Indented Seal Flange Tests { and 2. 5.0 -53... The loop was then placed in operation for thermal cycle testing as de- scribed in Section 4,1, During Test No. 1 temperatures were measured at the points indicated in Fig. 5. Fluoride salt, Fuel No. 30, was circulated in the loop at a flow rate of 2 gpm. The salt temperature was cycled 50 times between 1100°F and 130C°F. The average cycle time was 60 minutes, Both flange assemblles were removed .from the loop after Test No, 1 for leak testing. OSubsequently, the assemblies were separated and reassembled with pew gaskets., The flanges were then re-installed in the loop for Test No, 2. During this test, temperatures were measured at the peoints indicated in Fig. 18. The salt temperature was cycled between 1100 and 13%00°F. The number of cycles was S0 and the average cycle time was 61 minutes. Again, the flanges were removed from the loop for leak testing and dis- agsembly. ©One of the flange assemblies was reassembled with a new gasket and then leak tesbed. Discussion of Results 5.1 Freeze FPlange Jeint The maximam and minimm temperatures of various locations on the fleanges during'tharmal cycling are listed for Tests 1 and 2 in Table 2; for Tests 3 and 4 in Table 3. Examination of Tables 2 and 3 will show that the greatest temperature variation, at any one point in the flanges, occurs at the flange hub near the 1loop tubing, For example, thermocouple No., 21, Table No. 2, Test No. 1, varied from 862 to 1018°F. Thus, the greatest temperature variation, at this polnt during cycling, was 156°F. The smallest variation occurs at the periphery of the flanges as indicatea, in one case, by thermocouple No. 24 in Table No. 2. The minimm and maximm temperatures indlcated UNCLASSIFIED ORNL-LR-DWG 31979 INDENTATION SEAL FLANGE TEST NO. 2. THERMOCOUPLE LOCATIONS TC 31—-34 ™ FTC 3035 -TC 29-36 7 LTC 2837 TC 3233 N //\ - TC 27-38 77 77 N TS S S Y T 7777 S SN S S SN N S N - FLANGE A—=NICKEL PLATED IRON GASKET TC NO. 27, 28, 29, 30, 34, 32 FLANGE 8—NICKEL GASKET TC NO. 33, 34, 35, 36, 37, 38 m-psn. Fig. 18. Schematic Drawing of Thermocouple Locctions for Test 2 on Indented Seal Fiange. Table No, 2 FREEZE FLANGE TEMPERATURES MEASURED DURING TBEEMAL CYCLES Test No, 1 - Salt Temperature Cycled Test No, 2 - Salt Temperabure Gycled 50 Times Between 1150 and 1300°F 30 Times Between 1300 and 1500°F Minimum Maximum Minimum Mazimum Thermocouple Temperature Cycle Temperature Cycle Temperature - Cycle ™ Temperature Cycle Number # (°F) Number (°F) Number (°F) Number (*%) Number 21 862 16 1018 33,50 960" 3 1170 21,2k 22 Flange 625 16 760 53,50 695 1,3 950 9 23 A 450 6,14%,20 577 16 500 3 652 9 2k 245 16 280 50 275 3 305 30 25 69% 10 820 L 810 6,12 gh2 1 26 Flange 4oo 16 580 4 582 6,21 665 1 o o7 A 348 16 403 50 410 6 452 1,9,27,30 . 28 205 16 240 L 250 1,3,6 260 9,24 13 798 6 957 33 915 3,6 122 21 14 Flange 560 16 635 N} 660 1,% 905 27 15 B 375 16 k7 41 465 6 595 15,21,27 16 200 10,16 228 37 255 2l 305 27 17 Th8 16 865 2 865 12,18,21 1010 1,3 18 Flange 510 16 590 2 615 12,18 | 700 1,6 19 B 315 16 362 50 410 6 450 3,6,9,30 20 185 16 215 50 260 21 278 30 * 0. 1 - Stainless Steel Seal Rings ee Fig. 12 for location of thermocouples) Test Test No. 2 - Soft Tron Seal Rings f o w3 (See Fig. 11 for diegram of Inconél loop; == Table No, 3 FREEZE FLANGE TEMPERATURES MEASURED DURING THERMAIL (CYOLES Test No. 3 Salt Temperature Cycled Test Mo, 4 Salt Temperature Cycled 5C Times Between 1100°F and 1300°F 25 Tim=s Betwaen 1100°F and 1300°%F Thermocouple Minimum Cycle Maximum Cycle Miaimam Cycle Masxeiomim Cycle No, *# Temp., °F No. Temp, °F No. Teamp. °F No. Temp. °F No., 27 Flange 185 1k 218 4G 195 1 225 2k 28 B 340 18 Lo l-hi-h2-4g 34T 12 Loo 2-l.2k 29 815 65-7-11 973 1 788 12 57 2 30 Plange 785 6 o4 L6 768 iz 932 1-2 z3 B 277 15 40 46-48-50 285 12 335 2L-25 32 182 14-15 225 50 190 1-12 218 2h-25 B A 295 15 325 50 500 1 55e 24 25 788 6 975 15 760 21 Ghs 2 z6 Flangs 737 13 903 L6 712 12 885 1 37 A 287 15 350 L6 thru 50 292 12 245 25 38 190 13-14k-15 235 46 200 1-12 233 2425 % {See Pig. 1k for location of thermocouples) Tests Nos. 3 and 4 - Copper Seal Rings {8=e Fig. 13 for diagram of Iaconsl loop) wggw - 37 - by this thermocouple were 245 and QSO“F,'respectivaly, Thus, the greatest temperature variation, at this point during thermal cycling, was 35°F, The lov maximum temperature, and the slight varlations in tempera- ture encountered in the region of theifastenings and seal ring during thermal cycling, are desirable features of the freegze flange Jjoint. The meximum radial temperature difference between the outside and inside diameters of the flange, during the same test, was 738°F. This figure is €he differsnce between thermoccouples No. 21 and No, 24, taken at their maximum values. The corresponding salt temperature was 1300°F. The maximum such temperature difference cbtained for any of the tests was 817°F. The above examples are typlcal of the freeze flange Qpera'bione Com- plete data for all tests is presented in graphical form. The points of large and small temperature variation can be seen to follow the above descriphion. Graphs showing temperatures of the flanges throughout the periods of thermal cycling are inmcluded in Appendix 2. Gas leakage rates for Tests No. 1 and 2 were not satisfactory. This was athributed to defects introduced during the initial fabrication of the flanges. Some welding was performed on the flanges after all of the machining had been completed, which caused a slight warping. In addition, the finish on the seal ring grooves was not fine enough., These defects were remedied by additional machining. Subsequent leakage rates were sat- isfactory. Gas leakage rates of the Joints used in Tests 3 and 4 are listed in Table 4, Oune leak rate was obtained on each flange before Test 3 and after Test 4. As can be seen from the data, there was an appreciable increase in Table No, L LEAX RATES OF FREEZE FLANGES AND INDENTED SEAL FLANGES Freeze Flanges Indented Seal Flanges Lesk Rate in cc/sec Leak Rate in cc/sec Test No. Flange No., Before Test After Test Test No. Flange No. before Test After Test 3 A 3 %1077 Not Removed 1 A 5.2 x 1070 2.2 x 10°° -8 , - - 3 B 2.3 x 1¢ Not Removed 1 B 1 x 10 7 1.5 % 10 9 4 A Seme as Test 3 1.7 x 107! % 2 A 1.2 x 1070 3.9 x 1077 #xx L B Same as Test 3 5 % 1670 2 B 1.9 x 1078 1.2 x 1070 #x {See Report ORNL CF No. 58-1~203 "Notes on Helium Test No. 1: Flange A - Iron Casket Lezk Detection” by H. J. Metz) Flange B8 - Copper Gaskst I% This flange asgembly was separsted and reassembled Test No. 2: Flange A - Nickel-Plated Tron Gasket Lsing a new gopper seal ring to give a lsakage rate Flangs B - Nickal Gasket of 2.0 x 107° cc/sec) {#% This result was obbained after tightening one of the "C" clamps. The clamp had looseced waen the loop was cooled to room Lempsraturs foilowing cycling tests, ) {#%% This flange assembly was separated and reassembled with a new nickel-plated ircon gasket to glve a leakage rate of 5.0 x 107° ce¢/sec) o Qg e - 39 - the leak rate after testing. A third leak rate was obtained on one of the flange assemblies following the cycling tests, after disassembly and re- agsembly with a new copper seal ring. This leak rate was 2.0 x 10~8 cc/second and successfully met the speclfication listed in Section 3.1, paragraph (f). The flange assemblies were heated or conled between room temperature and the operating temperature of 1300°F at least eight times in the prepa- ration for thermal cycling tests. The rate of cooling and heating was at least 250°F per hour. Both flanges operated successfully in that there was no indication of sal% leakage, the gas leakage rates were satisfactory, and there was no indication of any cracks due to thermal stresses. (See Section 3.1), A memorandum concerning the metallurgical examination for stress cracks is included in Appendix 6. Oxidation of the copper seal rings was negligible. The rings had the same bright appearance after testing as they did originally. The Inconel Tlanges showed no noticeable salt corrosion., The flanges and seal rings are shown after the completion of Test No. & in Pigs. 19 and 20. The formation of the salt seal is shown in Fig. 21 after thermal cyciing tests and disassembly. Parts are identified as follows: (1) Frozen salt seal, (2) Lebyrinth insert ring, (3) Groove for seal ring. Both flanges are shown after disassembly in Fig. 22 and various parts identified as follows: (1) Cooling air lines, (2) Resistance heater lugs, (3) Calrod heater leads, (4) Thermocouple leads, (5) Clam shell heaters, (6) Copper seal rings im'placao The same flange is shown after removal of the salt seal in Fig. 23. The time for cleaning both flanges was not more than 5 minutes. -40- UNCL ASSIFIED PHOTO 31394 ~ Fig. 19. Freeze Flange (1/2' Tubing Size) Open with Copper Seal Ring After Clean- ing, Following Test 4. —4]- UNCLASSIFIED AR N oy PHOTO 31393 b INCHES Fig. 20. Copper Seal Ring from Small Freeze Flange After Test 4. ~AD- Fig. 21. Freeze Flange Showing Formation of Frozen Salt Seal. -43- UNMCLASSIFIED PHOTO 30205 Fig. 22. Freeze Flanges Installed in Test Loop After Disassembly of the Flanges. Photo showing both flanges disassembled. (1) Air Cooling Inlet (4 places); (2) Resistance Heater Leads (2 places); (3) Calrod Heater (1 place); (4) Thermocouple Leads (2 places); (5) Clam Shell Heaters (2 places) (6) Seal Ring (2 places) 4 UNCL ASSIFLED PHOTO 30207 Fig. 23. Freeze Flange After Removal of Salt Seal from Face of Flange. - b5 - The time necessary for the handling operations of disassembly and reassembly is indicated in Table 1. Ease and speed of handling were de- monst?ated, indicating that the freeze flange Joint is potentially suited to remote manipulation, Other tests are in progress to demonstirate remote handling of this type of Jjoint. 5.2 Cast Seal Joint The maximum and minimm temperature of various locations on the flanges during thermal cycling are listed, for Tests 1 and 2, in Table 5. Examination of Table 5 shows that the temperature variation, during thermal cycling, is nearly the same for a thermocouple located in a well near the cast seal as it is for a thermocouple located at the flange peri- phery near the flange bolts. For example, thermocouple No. 25, near a bolt, varied from 1177°F to 1287°F during Test No. 1. This is a variation during thermal cycling of 110°F. During the same test thermocouple No. 23, in a well near the cast seal, varied from 1157°F to 1296°F, a variation of 139°F. The temperature variation, during thermal cycling, in the region of the bolts is about three tifies fhat obtained in the case of the freeze flanges (See Section 5.1). In addition, the bolts were about 1000°F higher in temperature than the clamps on the freeze flanges. - The radial temperature difference, however, between the outside and inside diameter of the fl;figes is small compared to the freeze flanges, causing less thermal stress. The above examples are typical of the cast seal flange operation. Com-~ plete data for all tests are included in graphical form. Graphs showing temperature of the flanges throughout the periods of thermal cycling are included in Appendix 3, There was no indication of any salt leakage during Table No. 5 CAST SEAIL, FLANGE TEMPERATURES MEASURED DURING THERMAL CYCLES Test No. 1 Salt Temperature Cycled | | — Test No. 2 Salt Temperature Cycled 50 Times Between 1100°F and 13%00°F 25 Times Between 1100°F and 1300°F Thermocouple Minimm Cycle Maximm Cycle Minimm Cycle Maxdirmm Cycle No, * Temp. °F No. Temp. °F No. Temp. °F No. Temp. °F No. 19 1158 13 1290 1-41 1128 20 1255 2l 20 Note 1150 13 1280 1-h1-42 1122 20 1245 15-2k 2 (1) 1212 13 130% % 1155 50 1238 15 22 1188 6-13% 1323 L2 1142 20 1285 2k ! 23 1157 6-13 1296 b1 1125 20 1258 15-24 & 2l Note 1142 13 1282 4 1110 20 1243 15-2k ' 25 (2) 1177 1-13 1287 41 1130 20 1233 15-2h 26 1137 6 1315 L2 1110 20 1293 2L Notes: Note: (1) Flange with Silver Cast Seal The actual minimum temperature of all (2) Flange with Copper-Silver Alloy Cast Seal thermocouples in Test No. 2 was during (3) The salt and flange temperatures before the 18t cycle and was from 10° to L45° cycling were 1300°F lower than the minimum temperatures listed in table above, * (See Fig. 7 for thermocouple location) The salt and flange temperatures before (See Fig. 13 for diagram of Inconel loop) cycling were 1100°F, any of the testing. There was ne indication of gae leakage with the lcop and flangss under 25 psig of helium, followlrng the initial fusion of the ring lnserts. The pressure indiecation on the loop did net decrease over a pericd of several hours. There was ne indication of gas leskage using 3 goap solution on the flanges. The steps and time periods ipvolved in disassembly of these flanges are listed in Table 6. Difficulty was encountered in step (5), the re- moval of four stainless sbesl bolts and nuts from each Tlange. Fach nut had t¢ be heated to & rad heat, and a wrench used with considerable force. In step (9), separation of the flanges while hot, it was found that the flanges could not be pulled apart with a 25 to 50 pound force on the ex- tension rods comnected tu the upper flacge half. It was necessary to open the heater box arcund each flange and use a cc¢ld chisel to separate the Tlange halves, Since all of the disassembly operations would be ac- complished remotely in a reactor system, it is clear that the cast seal Jolnts, as tested, were not satisfactory in this respect. A comparison may be made in the time necessary for disassembly of the Freeze Flange Joint and Cast Seal Joint by referring to the data in Tables 1 and 6, The steps involved in reessembly of both flaanges are shown in Table 6, Steps l; 2, 3 and 6 would be accomplished remotely in sn actual reactor ingtallation. Afier reassembly, it was found that both assemblies leaked helium badly at the interfaces of the flange halves, ILeaks were detected with a scap solubtion while the flanges were under 15 psig of helium pres- sure, Photographs taken after disassembly of the flanges show that the flanges and sealing alloys wers badly oxidized. Fig. 2% shows the flange with the copper-silver alloy seal; Fig. 25 the flange with the silver seal., UNCTLASSIFIED PHOTO 30923 Fig. 24, Cast Seal Flange After Disassembly Following Loop Tests - Copper-Silver Alloy Seal. —49- UNCL ASSIFIED PHOTO 31423 Fig. 25. Cast Seal Flange After Disassembly Following Loop Tests - Silver Seal. Table No. 6 TIME REQUIRED FOR ASSEMBLY AND DISASSEMBLY FOR CAST SEAL FLANGES Operation Performed Dumping Salt from Loop Cooling Flange to Room Temperature Removal of Insulation from Heater Box Removal of Heater Box from Flanges Removal of Four Bolts and Nuts from Flanges Reinstall Heater Box Replace Insulation Heat Flanges to Melt Point of Seals o o \O 0 1 Ot F o o+ Separation of Flanges vhile Hot - O Cooling Flanges to Room Temperature Cleaning Oxide from Flanges Assemble and Install Flangees in Position . Install Heater Box and Insulation Heat Seals to Melt Point . Heat Flanges to Compress and Adjust Install Four Bolts and Nuts o WV A8 (See Fig. 8 for view of assembled flange) DISASSEMBLY . Total Total Silver Seal 15 min. 4 hours 30 min. 10 min. 40 min. 10 min. 20 min, 6.5 hours 20 min. 3 hours 16 hours ASSEMBLY 10 min, 30 min, 15 min. 4.5 hours 3.5 hours 5 min, 9 hours Time Required Copper Silver Alloy Seal 15 min. 4 hours 30 min, 10 min. 20 min. 10 min. 20 min, 4.5 hours 10 min. 2 hours 12 hours 10 min. 30 min. 15 min. k.5 hours 3 hours 5 min. 8.5 hours "Og" - 51 - Metallurglical examination confirmed that damage was done to the metal seals during the heating cpesrations. Ses Appendix 3, S e Imi&mt@a S@&l Joink The maximm and minimun temperatures of varlous locstions on the Tlaoges during thermal cycling are listed, for Test No. 1 in Table T; for Test Mo, 2 in Table 8, Exsmination of Takle 8 sbows that the temperature variation, during thermal oyeling, for s tharmoccuple locsted in a well near the metal gasket 18 about ons-half that of a thermocouple located at the hub of the flange naasr the l@@p‘twbinga For exsmpls, thermocouple No. 30, near the gasket, varised from 960°F to 1065°F, & variation during cysling of 105°F. Thermocouple No. 27 varied from 1090°F to 13L0°F, a variation during cyeling of 220°F, Tow tewperature variation, during thermal cycling in the reglon of the clamps ;is about thrze times thai obiained in the case of the freeze flange (S2e Section 5.1). This fact@rg togethear with the high temperature in the reglon of the fastenings, means that the clamps oo the Indented Seal Flange would have & greater tendency to looses than those on the freeze flange. Ths radial tamperature difference, howsver, betwsen the oubslde and inzide diamater of the flsrges is about 1/5 that of the freeze flauges cauging less thermal sirsss, The abuve esxamples are typlcal of the indented seal flange operation, Complete data for all téatg ars pressnted in graphical form. Graphs show- ing tsmperaturss of the flanges througbout the periods of thermal cycling are included in Appendix 4. Gas leakasge rates are listed in Table 4. Leak retes wers obtained Tavls No. 7 LSTENTATION SFAL FLANGE TEMPERATURES MEASUKED DURING TEEFMAL CYCLES - TEST N, 1 Test ¥o. 1 - Balt Temperaturs Cycled 50 Times Betwesn 11007 and :3007F | Mirvimun Maximum Praraocouple Lemperaturs C Terparature {ycle umhar "% Numben {°F) Numbsy o3 5 Yot M M = o7 1073 10 1320 17 2G Flange 620 10 713 : . : i 30 £32 10 735 1 33 1082 10 - 36 1%17 17 34 847 36 1015 5 Flangs 695 :6 - 45 803 36 B 673 45 780 O S -1 =3 o Flange A - Iron Gasket Flangs & - Copper Gasket * Thermocouplas damaged - count not repair during operation {Sue Pig, % fov thermocoupls location) (S2e Pig, 17 for disgram of Inconsl loop) 1 y K mggw Table Ro. 8 INDENTATION SEAL FLANGE TEMPERATURES.MEASURED DURING THERMAL CYCLES - TEST NO, 2 Test No. 2 - Salt Temperature Cycled 50 Times Between 1100° angd 1300°F Minimum Maximuam Thermocouple Temperature Cycle Temperature Cycle Numbeyr ___(*P} Number (°F) Number 27 1090 31 1310 45 28 9355 28 - L4 1040 13 29 Flfg@ 971 12 1075 39 30 960 48 1065 39 31 957 b ' 1050 s 39 - 45 A2 1077 A 1295 ks 33 1097 Ly 1310 37 Zh 538 o~k 1031 b5 35 Flange 912 28 - hh 1008 3G 36 B 927 ho - iy 1025 G 37 958 28 1065 45 38 1076 31 1295 45 Flange A - Wickel Platad Iron Gasket Flange B - Nickel Casket {See Fig. 18 for thermocouple location) {See Fig, 17 for dlagram of Inconel loop) mgga— bafore and after each thermsl cycling test. A new gaskel was then installed in each flange agsembly, followed by another leak test. None of the four gaskets, each of a different material, leaked salt. The flangs with the copper gasket leaked helium following the thermal cycling in Test No, 1 however, the leak sealed itself before the flangs ssgenbly was removed from the loop for checking with a helium leak detector, The copper gasketb was badly oxidized as well, as shown in Fig. 26. The flange with the annealed Armco iron gasket passed gas laakage tests satisfactorily. This flange and gaskét 1s shown after test in Fig. 27. The iron gasket, after removal from the flangs, is shown in Fig. 28, The outer edge of the gasket was oxidized and therafore congldersed un- suitable, As can be seen from photographs Nos. 26 and 27, a ring of frozen salt nad formed between the flange faces., This frozen seal sxtended from the loop tubing diameter of 0.410 inches to an cutside diamstsr of approxi- mately 1-1/4 inches. For subsequent teste a heater was placed arcund the outside of each flange assembly to insure that the salt bebwsen the flangs faces would be in the molten state during testing. Tha flange with the annealed nickel "A" gasket leaked gas arfter thermal cycling testing; however, this was caused by loosening of ove of the "Q" clamps, which held the flange halves togethero This flavg: and gasket is gshown after testing in Fig. 29. The flange with the nickel plated Armca iron gasket passed gas leakage tests satisfactorlly. It 1s shown after testing in Fig. 30. No oxidization of the gaskel was apparent. With the exception of the annealed Armco ircn gasket , removal of the gaskets from the flanges was difficult, It was necessary to cut the gaskets r,l’l,t'l,l]tll YT Y11 ° o INCHES Fig. 26. Indented Seal Flange with Copper Gasket After Testing. UNCLASSIFIED PHOTO 31133 UNCLASSIFIED PHOTO 31132 ’l'l,f'l’,'l"]lrl I]flfl' o 1 2 3 INCHES Fig. 27. Indented Seal Flange with Armco Iron Gasket After Testing. —57~ UNCL ASSIFIED PHOTO 31244 Biise, O ¥ Tt \ Fig. 28. Armco Iron Gasket After Testing in Indented Seal Flange. Fig. 2%, Indented Seal Flange with Nickel Gasket After Testing. UNCL ASSIFIED PHOTO 31518 ...89.. Fig. 30. Indented Seal Flange with Nickel-Plated Armco Iron Gasket After Testing. UNCLASSIFIED PHOTO 31520 _69_ 6.0 free of the flanges in a lathe. Of the four gaskets tested, the nickel plated and annealed Armco iron gasket proved most satisfactory. The Inconel flanges showed no sign of damage from ecorrosion or mechanical defects. A more suitable method of clamping is in order, to insure a tight seal during the various tempera- ture chenges. Future Tests The same freeze flange Jjoint and indented seal joint, described in Section 3 of this report, are now under test with molten sodium in an Inconel loop. An annealed copper seal ring is being used in the freeze flange. A nickel plated, annealed Armco iron gasket is being used in in- dented seal flange. The tests are similar to those described above. Both flange assemblies are shown installed at one end of the Inconel loop in Fig. 31. A larger size freeze flange jolnt has been fabricated. It is shown assembled in Fig. 32. The flange is shown open with the copper seal ring in place in Fig. 33. The principle of operation is identical to the small size flange tested and described sbove, Provision has been made on the ring seal for monitoring the gas leak rate during operation. Two such flanges have been fabricated for testing with molten salt in a loop, shown schematically in Fig. 34. The welding necks on the flanges will match a Y-inch, schedule 4O pipe. An unusual feature of the large flange is the use of a special clamp, made in two sections, which-holds the flanges around their entire circumference. The clamp is shown in the asaembly’photo- graph, Fig. 32. Holes were placed in the webs to reducerheat conduction from the salt to the seal ring area. A metal screen will be inserted in the space between the flange faces to reduce the hold-up volume of salt; UNCL ASSIFIED PHOTQO 31622 Fig. 31. Freeze Flange and Indented Seal Flange Installed in Test Loop for Cycling Tests with Molten Sodium. (1) Indented Seal Flange; (2) Calrod Heaters; (3) Toggle Clamp; (4) Air Cooling Inlet; (5) Freeze Flange - 19_ Fig. 32. Assembly of Large Freeze Flange. UNCL ASSIFIED PHOTO 31405 _39.. UNCLASSIFIED PHOTO 31406 Fig. 33. Large Freeze Flange Shown Open with Copper Seal Ring in Place. _.89.. —64 - UNCLASSIFIED ORNL-LR-DWG 31980 SALT PUMP P Yo-in. PIPE 4-in. PIPE FREEZE FLANGES FLOOR LEVEL Fig. 34. Schematic Drawing of Test Loop for Large Freeze Flange. aad to hold the salt cake together for easy snd guick removal whan the flanges are disasssmbled, This 4-inck jolat is made of Incomel, and the seal ring of dead-soft arnsaled copper with a one mll nickel plate. The clamps are of carbon stael, Two T/8-inch heat trested bolts are used to hold the clamps and flangs halves together, It is believed that Yesb results on the large flanges can be gsafely sxtrapolated to determine the sultability of the freezs fiange Joint for 2 molten salt reactor system, Approved by - 66 - APPENDIX 1 Preliminary Mechanical Joint Specificaticn - 67 - INTRA-LABORATORY CORRESPONDENCE CAK RIDGE NATTONAL LABORATORY Kovenber 14, 1957 Tos Lisgted Distribution Froms G. D. Whitman Subject: Preliminary Mechanical Joint Specification The attached specification has been prepared for your review and comnert., This document is intended for internal use and should serve primarily as a design and development gulde., Some of the values listed are more or less arbltrary; however, the criteria presented must be dealt with, and more congidered limits may be assigned after a thorough review, GDW/ds Att, Distributlion W. B. McDonald H. W, Savage H. 3., MacPherson E. J. Bresding B. W. Kinyon L. A. Mann Jo Zasler IT. III. PRELIMINARY MECHANTCAL JOINT SPECIFICATION ocope This specification selsz forth the requirements for a mechanical joint which may he used 1o a radioactive high temperature molten salt system, Service Regulraments This unit shall contaln radisactive molten salts at tempersturss up to 1300°F, and shall contaln gaseous fission products bstween normal ambisnt and 1300°F, In addition, the unit sball be adaptable to remots assembly, disassembly and leak checking. Materials of Construction The materials of construction ghall be compatible with the process fluids and in no way centribute contaminants which might reduce the life of the system or in any way raduce the efficlency of the nuclear or thermodynawmic processes. Furthermors, it would be highly desirabls to eliminete materials of construction which would result in high levels of induced activity. Mechanical Design A. Bize The unit shall be adaptsble to plpe sizee rangiog from 1/2" IPS +o 12" TIPS, B. Orientation It is highly desirable that the unit be adaptable to piping runs of random orientation; however, a single application to vertical or horizontal ruus shall be considered acceptable. C. Holdup Volume When the procegs fluid is drained from the system, the residual msterial if any, shall not be more than 2 ce., Any material held up must bs re- tained for inventory purposes and shall bs handled in a mannsr that rse- - 69 - duces the spread of contamination. Alignment Tolerance The unit shall be capable of functioning to meet the requirements of this spacification with a piping centerline mismatch of # i/32" snd/or angularity misalignment between piping runs of * 1/2°, These specifications are to be met without stressing the joining piping runs beyond thelr design imit. Flow Resistance The run of piping containing the unlt shall not have any appreciable in- crease in flow resistance as compared to & similar section of regular pipe. Strength The unit shall behave the same ag the parent plping under time, tempera- turs, and stress and shall not require preferential application, Remote Servicing All special tools, fixtures, and leak checklng equipment shall be con- eidered as part of the unit package, and practical x@m@te demonstration of such apparatus will be required. General Because of the remoteness of the appllcation, it is desirable that auxiliary service requirements be kept to an absolute minimum. Leak Tightness A, Gas Leskage 1., Gas leaks shall be determined by means of a mass spectrometer leak 7 detector having a sensitivity of 1 x 10 ' cc/sec. Ths system sensitlivity shall be established before each test by calibration of thes detector with a standard leak. < - TO = No indication of gas leaksge sball be permitted. This leakages specifi- cation shall apply during and after SO temperature cycles betwean ambient and 1200°F. The rate of temperature changs shall not be less than 100°F per hour. Section V-A-2 of this specificaticn is to bhe repsated after the unit has been remotely disassembled and assemblad., Fluid Leakage 1. 2. No fluid leakage shall be permitted. The wnit shell be cycled 50 times betwsen 1100°F and 1300°F with no indication of fluid leakags. Section V-B-2 of this specification is to be repsated aftsr the unit has been remotely disassembled and agsembled, Special Leak Testing Provision shell be made to establish the leak tightness of the unit during kigh temperature operation., Means shall be provided to collact the at- mosphers surrounding the unit during high tempsraturs fluid service to establish that gas contained in the fluid doss not egress through the seal, Preparsd by: G. D. Woltmsn 111157 - TL ~ APPENDIX 2 Graphs of Freseze Flange Thermal Cycling Tests ; -] 5 PO ' 460 30 o o S i & O o N ® /G i 20 2z 24 240 28 30 372 34 CYCLE Time (HoURS) Qb LeLii=Dyg 18 Po 22 74 26 28 30 CYCLE TYME (HOUKS) o 14 o AL L (1 a 3 9 RS) U ) p °F & o ir e e LR R AR L T Lafot PR T Rt 3 U M A Lt s s g O e P VR T, i md e s e e e prer i - ; ST e et b e /200 TRl IR T T LoLniInp DOTITITTE TR LI T s T o L A T Php g e - ....................................................................................... sea s e oM - - pheeecdea o poosep e b b i e 00 .................... .......... STl b it e e e et deae et .... ....................................... ........ ........................... ...... I LT T T R ol " z3 . : LTl S — S : L —_ “T: rc_ . TETTRIUTTA o b err T - SR S : -:)’C)O O = TIphe : i . L o o - . it i TLITUT T L ST s - M ...... ..... oz & & & o J2 M e 1E 2o cvelE Time (HOURS) OF. J{." /2 1200 7 =7 50 475 0 2//00 75 50 Z 25 / /OO0 7 75 5 50 2 = G0 7 75 co Z 25 &0 800 65 (250 Z z GO GO0 7 75 50 2 HZ25 500 560 5 75 T o 25 25 4 400 300 7 475 {50 2z 225 o z 4 & 3 i i7 i+ e & Z2oi o ; CYCLE TimeE (HOURS) CYCLE TireE (WOURS) o 8 FO ’Z /4 fo 48 20 AN 5 ‘ UNCLASSIFTELD CoNLwLilaDwg 4= 31987 /OZ. / SO0 ,,,,, 5C - ‘. Cco Zi 34 300 | —r I 43 45 47 49 51 55 5 18 2o 235 37 39 4 CYCLE TIME (HoUuRs) W b i i4 /6 !8\ o 77 35 37 29 4 43 45 47 49 51 5 CYCLE TIME (o0RS) - ’ Cy P I g O 5 S = 6, - /80 n 27 39 (# oufi?S} M te 18 20 22235 37 29 41 43 a5 47 49 5l 53 CYCLE TIME (HOURS) O Y B fi'fi o G M - + by e /02 i S /& 20 22 74 2 (YCLE TimE ('NQ-U’QS) A TATSSYIOND 2EEEo © FHTerT="THHO v \H CYCLE TIME 7 N g\/‘/fie.’h?fsjl (MHTATCIYTON 14 — OriLeeLEieDirza =3 1994 UNCIASSIF LD (HouAs) TImME CYCLE 20 22 [ 214 e /0 /oL - 86 - APPENDIX 3 Graphs of Cast Seal Flange Thermal Cycling Tests 2 /<3 L /8 20 22 %% CYCLE TIME 29 4 (HOURS) 45 4’7 UNCLASSIFIED ONLelR=Dwg, 31995 49 Sl 55 Q - !8 - UNCLASSIFIED berfaoe s S deae e ...... ---------- ..... vvvvvvvvvv ...... .......... ........... “““““ s %5 x7 39 a4l 43 45 47 49 50 53 TIPE HOURS) CYCLE TymeE (Hovrs) Mégw € * Bt T T 0 LE6T /O /2 /4 /6 /& 2z 24 Po CYCLE TIME (HouRS) ,.06.. a3 AT SSVIONN BOOT L= SH~T=TINEO tata - Gl - APPENDIX 4 Graphs of Indented Seal Flange Thermal Cycling Tests ) & CYC/S £ 20 22 TINE 24 Z6 (HovRrs) 7E 30 (s N 34 36 38 40 QN T AT SSTTINS TE= B T~INTO o tay ™~ e e ..... K e 12 28 N & 20 z2Z L 26 28 30 CYCLE TimE CHOURSE) (o B3 13004 75 251 1260 74 20 255 /10 = (o e 7S s e Spls B ciiiniddTe 25 /000 75 59 25 GO0 S e ..... 2 75k : 204 25 o - o 3 . - + — - - e T T T e b s e e b T e e o et T, - + . - : — tr - pos 7 z O z 4+ G 3 % /2 /4 /o /8 Z0 Z22 24 fe 28 30 32 34 3& 38 40 1300 -~ < 200 25 /D0 s -3 Lo EOO 75 25 700 75 25 & o Z 4 G 8 O R M4 e MR P 2 24 2L PR30 32 F4 3L 38 40 CYCLE TymE (HOURS) 1200 75 50 /10¢ 75 25 SO0 x 20 TIE 22 24 (HouRS} “F {300 F 200 /00 / OO = ‘ - : 1 =1 3 20 OO0 Z00 /OS50 40 I~ fi N 5 o & /8 2Z 24 6 28 30 . ¢ py CYOLE TIME (HOURS) — — N f & o 3 P N 075 50 /oco 75 50 z5 ¢ 250 z25 ;OO0 75 25 S0 6 i ‘.) 0 iz e Rt e by s T £ - 2 /4 ) 16 zo 2z 24 cyeLs Time {(Houes) Z6 28 O > QO 3 e Ty o > W o . -3 5\ B 0o /G - Y 76 zZ CLE TIME 2Z 24 (vouis) Ll X 40 Ufl} 54 - 100 - APPENDIX 5 Report on Metallurgical Exsmination of the Cast Seal Flanges ~ 10l - SCLIDIFIED METAL SEAL DEVELOPMENT G. M. Slaughter In order to provide information to the dssigners of the solidified metal geal, several 0.252-1n,-dia tensile bars wars prepered from cast silver, These ware tested at room temperature, 1200°F, and 1M00°F. The test re- sults indicate that the room-temperature mechanical propsrtiss are aboutb the sams as the publishked data; i.e,, 20,000 psi tensile, 8,000 psi yield, and 50% elongation inm l-in, At 1200°F, the teunsile strength dropped to 2,400 psi and the yield strength to 1,900 psi, with a corresponding elongation of 6 - 10%. At 1400°F, the tensile strength was 1,500 psi, the yield strength wag 1,00 psi, and the slongation was 10%. The two flanges that were operated in test at ¥-12 were received for metallographic examination, A half-section of the silver flange is shown in Flg. 35 {¥-26029), while a half-section of the silver-copper flange is shown in Fig. 36 (¥-26103). The exsmination indicated that moderate oxi- dation of the components during opening and closing the flange had occurred, thereby impeding wetting of the base metal. The flange utilizing the silver- copper alloy appesred to be less subject to this condition than that con- taining the pure silver, probably because of the lower tempsraturss re gqulred to remelt it during opening. A pbotomicrograph of a typleal inter- face between the silver-copper and the Inconel base metal is shown in Fig, 57 (v-26241), -102- UNCL ASSIFIED Y-26029 Fig. 35. Cast Seal Flange with Silver Seal, Sectioned. -103- UNCL ASSIFIED Y-26103 Fig. 36. Cast Seal Flange with Copper-Silver Alloy Seal, Sectioned. T = - $o0r 03 - X N7 e 2 e 2 & * 4 i > o .;A - P i ¢ & ‘ ~ - g B v e BTN,y e ey - | g e W el - - f‘zr-(%~ i o = % < % 2y ) - & 1. - . . i - > N : ° - . ‘ ' . ¥ . ® 6 % : ° : . \ q . 9 - . . . . ° a ‘ “ e 1 % ° o . -104- UNCLASSIFIED Y-26241 * a 5 i . T e » ° s * . * . . . : . ° L . INCHES "{0.02 Fig. 37. Cast Seal Flange with Copper-Silver Alloy Seal; Photomicrograph of Interface of Alloy and Inconel Wall. - 105 - APPENDIX 6 Report on Metallurgical Exemination of the Freeze Flange Joint - 106 - INTER-COMPANY CORRESPONDENCE UNION CARBIDE NUCLEAR COMPANY A Division of Union Carbide and Carbon Corporation To: A. Taboada Plant: 9201-3, ¥-12 Copies to: W. B. McDonald Date: May 26, 1958 A. S, Olson RSC Files (¥C) Subject: Examinaticn of Freeze Flange This freeze flange had operated through several thermal cycles in test. Information regarding possible cracking was desired. Three sections were taken through the center hole to explore all the areas likely to exhibit cracks. One section included a welded-on web and s longitudinal section through the center hole. Another section was taken roughly 45° away longitudinally through the hole and a third in trarsverse cross sectior through the counter-sunk center of the flange. It was in this area that cracking was felt to most likely occur. None of the sections examined showed any cracks. R. S. Crouse RSC:fl - 107 - APPENDIX 77 Asssnbly Drawings of the Mechanicsl Joints and Loop Freeze Flange Mechanlcal Jolnt - 1/2" tublng size. Dwg, No. D-2-02-054-6777 Cast Seal Joint Dwg. No. C-2-02-054-6837 Indented Seal Jolnt Dwg, o, D-2-02-054-7249 Freeze Flange Joint - U-inch pipe size Dwg. No. E-2-02-054~TOL4 Towconel Test Loop Dwg. No, B-2-02-054-7126 (The drawing shows the loop configuration for fresze flenge testy Nos. 5 and 4 and cast seal flange tests Nog, 1 and 2. For the indentaed seal flange tests, the cast seal Tlaunges were cut out of tha loop and ra- placed by lengbhs of 1/2 inch Inconel tubing., The freeze flanges were cub out of the loop and the indented seal flanges inpstalled in their placs,) ~109- PARTS L/S57 PARTS LIST VTEM OWG. 0. | @Ty NANIE DWG. MO, | QTY. HAME ¥ %%Q AL TO ALY S & |0-6777-6 | /1 | FLANGE - IVCONEL [ \D8777 1) & |FLAVGE SE-ASST —__-C BORE &.a DEER 7 lpoesrrr-zi i FEINVG & DEPP7E S TLBE c" MHOLES 180° AFQR T 5 061778 4 |BracE ERE I, a;;,i_zrfi/f\.;aaifi.sflr:r A I g \O-E777-4 & EZZA//’:,;-FF-/ P At VA } ot } | B oy ay O + SN | 5 Ve 1 73 |2 44, ' 3 e 52 ' 40 Ao ' | — S | A i SECTION A-A | TS0z | | S0 506 . S £a = § ! / Los o= --62 fa/2} | : HB75 | /4. I i ! S | V=305 DR BALl ASSEMELY : DRAWING NO, __ e 2 - TTLE FEE T8 P ANGES j N% A MATERIAL ___ T o ‘ | T B R ‘1 &4 seaLs A THEL 51_{\\15;_"}'{\:‘5‘5 277-5 5 ? K’“‘fi 2 A END A & PrACES T2 4—*— - OL : EQUBLLY SPACED e -2 A a— — S NCTE: I~/ & :jl [ — 4QLl VWELDING TO 5L COARELLTELD BEFORE e /* /Zx/\é/fi// MAVC///N,«%G BOTH FACES g GROOVE o LA, 5(9/?;5 Ld. & .50:‘?5 = END F0 306 SECT/ON 4-A - LSS - o /Vf/[if.ffl Péfit_,ffi AT '—,ag'a L)/fi :;J\h\l}z(}wo O-5777-5 TITLE _FLAANVGE SHUB-A33) warenial AV YVES ‘ sonLs FULL -G 777~ 7 ! D-2-02-054-65 777 4:/-‘\/:5- _ /Tg ! 4 AN AN Et & 5 5 = E v |PLy e o &3 nrLe AAVG 5.740 HA, S uATeRIAL I NEDNEL scavr AL N wonn L6777 7 i ey — / . = 2 _L— By B I a5t ) ' FER ol c—?’- — 83 REFERENCE DRAWINGS { ows. wo. n puas au sy e AK RIDGE NATIONAL LABORATORY D ] D Ll [T ° Y 0 D=5 777~/ i =877 7-5 fraceresn + 2L M oLk LFRUCLE ritLe 2 LBE e = woais ¥ oo 4 ! e COVEL swzsiin. WL TUBING_ w2yl % ok 5205 VO i < RELELE FLANGES THAWIRG scars AL muvipn-57 77 -5 scare £34 nouper S E277- 2 B W bt | T kLY JSENE VOTED | D352-4 1 [ D-2-02-054-6 777 -110- SILVER ALLOQY PARTS LIST ITEM DWW, W0, | QJTY, NAME ! |B-6839 | / |UFPPER FLANGE £ 18-€6838 | / |[OwWFERL FLANGE CAF SCL, SOC. A2 S [6e8373\ 4 Iz ¥ 3 14k A €837 5 (W7 B 5 B-¢840 |/ | 5F£4L A/VE RING IS MELTED AFTER JOINT IS ASSEMBLED. JOINT IS SURROUNDED BY ELECTRIC CLAMSHELL HEATERS OF & - QUICK REMOVAL TYPE : l / . FOR JOINT REPLACEMENT. | 7 02 P / r/ %/—{7——1@;—. C-2-02-054-6837 S T AT T A FA | "/ /é 5 £ REF o A < £ \ | N T 3y | -+ REFERENCE DRAWINGS | ows. wo. Tg j,_(;“ ”féfi A - Vits Setmwer 000, OAK RIDGE NATIONAL LABORATORY = Tpimiset| | e | EALEEITIENTAL Sc/fz:;r SEAL /f-’fwf /‘( {/4%7 :;; 1;006' FME /4 5 / A VK NoTE @:ff‘”" LCH sim feiefr — o %% SCALE , S BAAE AL e T S L CEEE, DTl ST RS DD S (\—a o e et G NGGEET 2 | PFOBE ASHEMBLY \_'.,__.__*‘ L L : o2 1 \ A T i Pd ! .54' : ____;;n__l e e e T T T e - — - e e T T — s S e — — = ——F ot LB IR AT O DAY ATUIAR B BN S A RIS R LD A TP IS TS S St R SR D O AT LD AT T AT A C T LE IS T T ATINS N & E‘\ __.v—‘lj::——’—:‘f——:—:'_’_r' — = ] LA v w oo e) . e — o e — e it [£202054 7/ = o e T E T T D Ay e AL DI A T A R L SIS ® Al / = For All weiding to be done in accordapse AT B Pt FRASENT LG 7 e BT INCONE: with P§__ = ezongt, 58 aterwiss | e motad, ' T A A TRL e T s SN L A T F—— & — DL AR TN AYASD S o B st T TR WS L AP SRSV L L A AT TEEASIE A - L LT T VD D R D b FOR Al welding to be done in accordance s A HASTEY ; M S 3 )/ ""\6/") Tl AL DT TP P A (EOTELLDY :;::d--r_tfii..é_ except as otherwise 7/ Wy . S ‘51 < — AD Dva, Ry T I ¢ '_-"T"l—L b M EO LD REFERENCE DRAMINGS 1 DWG. MO, v r TR ST - _ - i o 475 i i’ e OAX RIDGE NATIONAL LABORATORY - ETED T S ; T i HEY TR | e e Toulaandh HRAH S Vb L e - e TP -. - = + G e G seintn At ST A KT ETE TTET A AACTE TES e . i i B IL T ; L'v/}jf L‘,T('}.‘,“'—‘ ot & et ot e Dt 1 iy -A_'zt 5 jheeeld £ oS i ' - A fys bl - ¥ L e A :fsj vernveons| BT 2 —— ~ %,-V gl Fac dscat 7 - E-2-02-054- 7/ 7% xrrmos 3t | wan | am b gl 74 p F iy et ‘—'{.4_ e ' - ¥y Figure No, 1 Cross Section of Freeze Flange Joint . o o o ¢ ¢ « & 2 Freeze Flanges Installed in Inconel Test ILoop . . . % Schematic Drawing of Indented Seal Mechanical Joint 4 Assembly of Indented Seal Flangs Before Testing . . 5 Parts of Indented Seal Flange Before Testing . . . . 6 Indernted Seal Flange Shown Assenbled with "C" Clamps T Cast Seal Flapge Showing a Cross Ssction View and Thermocouple Location o o o « ¢ « 2 o « » » o 8 Assenwbly of Cast 828l Flanges . + + o » o « o & o 9]':13.(:0%1@1@@%‘&1‘099...... e & @ @ & © & ® o e @ lO m@@n@lTefitL@@P”mpmdaooqoooa»-oo-a 11 Schematic Drawing of Test Loop for Freeze Flange %@Stfilandgfloil‘@.fl.l.fi.'.. 12 Schematic Drawing of Thermocouple Locations on the Freezs Flanges for Tests 1 and 2 . . o o « & 13 Schematic Drawing of the Test Loop for Freeze Flange Tests 3 and L and Cast Seal Flange Tests 1 and 2 1% Schematic Drawing of Thermocouple Locations on the Presze Flanges for Tests 3and b . . . . . . 15 8ketch of Exteunsion Rods for Disassembly of C&St S&al Fl&flg@fi & . “ » [ * * ° - . s . . 16 8Sketch of Spring-loaded Tie Rods for Reassembly of Ca&tS@alFlaflgefi s 2 3 & & ©® @ & & B © s @ 17 BSchematlie Drawing of Test Loop for Indented Seal Flapge Testes L and 2 . + ¢ & ¢ o ¢ o0 ¢ & & .18 Schematic Drawinglof Thermocouple Locations for Test 2 on Ipdented Seal Flange., « o o o« o » ¢ o + 19 Freeze Flange (1/2" Tubing size) Open with Copper Seal ' Ring After Cleaning Following Test 4 . . . . . . 20 Copper Seal Ring from Small Freeze Flange After Test 4 21 Freeze Flange Showing Formation of Frozen Jalt Seal . . - 115 - List of Figures * L] Page HNo., a0 10 11 15 1k 15 17 22 e3 26 27 31 o 724 5k Lo ka L2 ~ 116 - Figure No, ne 26 27 29 30 A N N 55 2 57 Freezs Flanges Installed in Test Loop After Disasgembly of the Planges . .+ ¢ ¢ ¢ o +» ¢ o o o o Freezs Flange After Removal of Salt Seal from Face of Flangs . ¢ o o o o o o s o o o o s o ¢ ¢ o Cast Seal Flangs after Disassembly Following Loop Tests - COPP@-T"SilVE? Allfly Sfi&l ° ° o o * s o . a s o » . Cast Seal ¥lange After Disassembly Following LODP Testfi o Silvf;“l" S@&l - . . . o o ° o o o » 2 o Tndented S=al Flange with Copper Gasket after Testing . . Indented Sssl Flange with Armco Iron Gasket after Testing Armco Iron Gasket Affter Tesgting in Indented Seal Flange . Indented Seal Flange with Nickel Gagket after Testing . Indented Seal Flange with Nickel-Plated Armco Tron Gasket after Testdng + o« o ¢ o« o o 0o 2 s o o o o o Fresze Flange and Indented Seal Flangs Tunstalled in Tegt Loop for Cycling Tests with Molten Sodium . Azgembly of Large Freeze Flange . o ¢« o o o o o s o o o o Large Freezs Flange Shown Open with Copper Seal Ring InPlace . 4 o 4 o o s o o s o s 0 0 5 o 0 5 e o o Schematlc Drawing of Test Loop for large Fresze Flaoge Cash Seal Flange with Silver Seal, Sectlonad . . . - o » Cast Seal Fleange with Copper-Silver Alloy Seal, Ssotiloned Cast S=al Flangs with Copper-Silver Alloy Sszalj; Pholto- micrograpk of Interface of Alloy and Inconel Wall . Pags No. k9 93 56 > 58 29 104 - 247 - List of Tables Table No. Page No. 1 Time Required for Assenbly and Digassembly for Freeze FlangeS « o o o o o 2 ¢ ¢ 2 2 s o o o s a o o o ¢ 25 2 TFreeze Flange Temperatures: Measured During Thermsl Cycles . , . 35 Freeze Flange Temperatures Measured During Thermal Cycles . , ., 36 £ lLeak Rates of Freeze Flanges and Indented Seal Flanges . . . » 38 1 Cast Seal Flange Temperatures Measured During Thermal Cycles. . 46 N Time Requlred for Assembly and Disassembly for Cast Seal Fl ange 8 - . . » . - » > L o * e * ® - Q L 4 * . » - ¢ * * 50 T Indentation Seal Flange Temperatures Measured During Thermal Cycles - Test NO. L & 4 4 6 o o o s o ¢ 2 s o » 52 8 Indentation Seal Flange Temperatures Measured During Thermal Cycles ~ Test No. 2 . ¢ ¢ ¢« ¢ o ¢ o o« ¢« ¢« o« &« + 53 Distribution 1. G. M. Adamson 2. S, E. Beall 3. A. Benson, AEC L. D, §. Billington 5. K. E. Blanco 6. F. F. Blankenship T E. P. Blizard 8, A. L. Boch 9. E. G. Bohlmann 10. B. §. Bomar 11, ¢. J. Borkowski 12, W, F. Boudreau 13. G. E. Boyd 1%, E. J. Breeding 15, J. . Bresee 16. R. B. Briggs 17. ¥. B. Browm 18, ®. R. Bruce 19, D. W, Cardwell 20, W, R. Casto 21, . BE. Denter 22, R, E. Chapman 23. R. A. Charple 24, B, B, Clausing 25, PF. L. fuller 26, J. 5, Culver 27. Jo E. DeVan 28, 8, E., Dismke 29. H. ¢, Duggan 0. W. K. Bister 3. L. B. BEmlet 32. J. Y. Estabrock 3. D. E. Ferzuson 34, A, F. Frass 5. E. &, Franco-Ferreira 3., J. 5. Frye, Jr. 37 . W, R, Call B2, H, E, Goeller 3. W. R. Grimes Y. E. Guth Yi. C. 8. BEarrill A. W. Hoffman hz, A, Hollsender 4h, W. 8. Eoronbaker 45. A, 8. Householder Y6, W, H., Jordsn T P. R. Kasten Y3, G. W. Keilholtz Yo, M. T. Kelley 50. W, H. Kelley, Jr. - 119 - 51, 52, 25 54, 55. 58, 57 =5, 590 60. &1, £2. 63, 6L, 5. 66, 57. 68, 9. T70. T1. 72. 5. T4, 5. 76. TP 78. 19 80. 81, 62. 83, 81??, 85, 86, 87. 88, 8g, 90, 9l. 9z, 93, ga, 95. 96, 97& 98106, 107-109. 110-112. 113-127. v & & L o u s » 2 ° 2 * PLEHPrPQEOHURPPOPEEPPYEE RO 2 o= e ¢ L. W, A. C. G. g B. Korsmeyer . Lamb . Lane A ‘H. Levis . 3. Lindager S ‘m? 7 S, Livingston . N. Lyon G. MacPherson . DL Msnly A. Mamm B. McDeonald K. McGlothlan R. Mclally J. Miller Z. Morgsn J. Murghy P. Marray I,. Nelson R. Osborn W, Parrish Patriarca M. Perry Ramsey Ring ¥, Rupp . W, Savage W. Savolainen F. Schaffer M. Sbepkerd D. Shipley Sisman J. Skinner M, Slaughter E. Snall Syiewak Storto A. Swartout Tabhoada H. Tayler S. Toomb B. Trauger E., Unger M. Weinberg E. Winters D. Whitman Zagslier Iab Records, ORKL {RC) Leb Records Department Document Reference Section Central Reeearch Library TISE, AEC