Document dk1MZN7v9zzkkZjLBmYGQBXb

F us C ' P^-^T. M^r-^rGET^'S CELANESE CORPORATION OP AMERICA M*ir-w'r=-T^7^"-I! Chemcel Plant E ---[ Operating Technical Report ,iO. r.'OT TiH SUBJECT: Corrosion Tests of Various Gasket Materials COPIES TO: Mr. Bowen Mr. Anwyll Mr. Lederer Mr. Lindsey Mr. Mayfield Mr. Rogers Mr. Wilson Engineering Department Pile Report--SuppIemen'C"#! Date: December--?;--____ Is'ini'iK tu Hr11inl ff&j ihi ml Oct.15, 1948 -------- By: E. V. Kunkel PLAINTIFF'S R. C. Schueler EXHIBIT CEL - 3 Indexed Under: `4fI Mr. Fennebresque (2) : Mr. - Schnabel . Mr. Schneider PURPOSE: ` To determine the reaction of various gasket materials v :: ; - ^ when used withr stainless steels under specific condi- - ~v 'hi t.1 nrtrt'.v ; 1" . *' ^<L&BUBlMAEEr^og^msa^l2^^|t^Ens;Cdc:.i'each gaske^material tin-each, of tha^v%r|gg; "-t*h,__rlee^IIMSil estaWi r|Mepl orI itll ed on see .Tables IA, IV. .. A... . and VH_A_____a__t__ V-v v -Teflon' again was consistentlyhrr<a` *-tVved good, however, it cannot be recommended as ah. all purpose gasket material because of the price of approximately $14.00 per pound. Teflon should be used only in services where it can be economically Justified. Crane 888 tested best, or at least among the best, in each of the separate exposures. Data from these tests at present is Insufficient to state that Crane 888 is superior to other neoprene asbestos sheet. Garlock 7228 and Sterling 417 are definitely inferior to Crane 888 as they induced stress cracking in the stainless. Results Indicate, however, that Crane 888 aggravates corrosion to a somewhat lesser degree than the other materials tested. It is believed that no one specific gasket material will be found economically feasible for use as a standard material and that it will be necessary to prescribe several different materials for a limited number of corrosive conditions -- each one being the best for a specific service or services. DISCUSSION: Data is presented to show the results of three additional tests of gasket materials exposed in (1) S0-6-TJBaae?--f?)~30-6-T Reboiler; and (3) AR-3-T Base. The design of the test racks and procedure followed were the same as described under "Discussion" in Operating Techni cal Report #67 with the exception of the surface finish of racks 7, 8, and 9 as described below. An,PiA , 001549 2 I. S0-6-T Base Test racks #5 and #6 were exposed In the base of SO-o-T from October 14, 1948 until April 20, 1949. When the tower was opened for removal of the test specimens, the racks and holddown bars had completely corroded away leaving only the gasket materials. Thus the results of this test show only the effect of the corrosive media upon the gasket materials and no data is available concerning the reaction between the materials and the stainless in this ser vice. The materials found to be most resistant to the corro sive media in this test were, in the order named: 1. Teflon 2. Crane 999 3. Garlock 22 4. Crane 998 5- Crane 888 6. Ankorite 450. Table IA at the end of this report lists the rating of . each of the materials tested with respect to the effect of the corrosive media on the gasket materials only. Table IIA lists the test conditions and data. Material types and evaluation of the remaining gasket materials are listed in Table IIIA. v. 11. S0-6-T Reboller Test rack #7 was exposed to the media in the S0-6-T reboiler from June 22, 1949 until August 22, 1949. The test rack was modified before assembly by grinding all gasket surfaces with 120 grit paper in order to make all metal surfaces uniform. Two of the holddown studs failed completely during exposure and as a result four holddown bars and four gasket specimens were lost. The remaining study had corroded at vary ing rates but were still in place. However, the effect of the gasket material on the stainless plate before loss of the speci mens remained for evaluation. Table IVA lists the overall rating of each of the materials tested. Table VA lists the test conditions and data. Material types and results are listed In Table VTA. The photographs in Figures 1A and 2A show the results of the test. No gasket materials are shown here because most of them were broken or damaged In disassembling the rack. The effect of each material on the stainless plate and holddown bars appears to be of less severity than was expected. This is probably due to early failure of the holddown studs which allowed the gasket materials to become loosened and allowed a more or less free flow of the liquid, thus upsetting the conditions con tributing to a maximum corrosion of the pitting type. 001550 -3 Severe stress cracking and light stress cracking were noted on the stainless In contact with specimens 1 and 2. This can be seen from Figure 7A (photo of specimen 1 on rack #7). Stress cracking normally occurs in a direction perpendi cular to a tensile stress. Stress was set up during prepara tion of the plate by grinding. This cold working of the sur^ face in the longitudinal direction caused the greater' tensile stress to occur in that direction. Ihe major stress cracks which occurred were perpendicular to the direction of grinding. With a residual tensile stress and probably a chloride from the gasket material, conditions were suitable to cause stress cracking. The best possible means to avoid stress cracking under these conditions would obviously be to eliminate the source of the chloride since a tensile stress alone is usually not sufficient to cause cracking of this type to occur. The best material on test rack #7 was found to be Crane 888 although it was not completely satisfactory due to the fact that the material was hard and brittle at the comple tion of the test. III. AR-3-T Base Test racks #8 and #9 were exposed in the base of AR-3-T from July 7, 1949 until September 29, 19^9. These test racks were given the same 120 grit surface ^ finish as rack #7. Teflon, Durabla, RM K67O, and Crane 888, tested pre viously in this service on test racks #3 and #4, gave essen tially the same result this time. On racks #8 and #9 the Teflon was still rated ''good" and the Durabla was slightly better than before. The RMK670 was still unsatisfactory and the Crane 888 was not quite as good. This slight variation in results is probably due to the lowering of the acid content of the corrosive media, the longer test period, and the differ ence in thickness of the Crane 888 and Durabla specimens. The longer test period is definitely the cause of the greater com parative amount of corrosion on the test plates of racks #8 and #9 Table VIIA lists the overall rating of each of the materials tested. Table VIIIA lists the test conditions and data. Material types and results are tabulated in Table TXA. Photos in Figures 3A, 4A, 5A and 6a show the plates, holddown bars and gasket materials after exposure. -% Stress cracking was noted on the stainless plate in contact with specimens 5 and 6 (Sterling 417) although they were not as severe as the stress cracks noted on rack #7. Specimens 1 and 2 (Garlock 7228). 001551 Teflon was found to be the best overall material in this service, with Durabla, Crane 888, and Ankorite 425, all being only slightly better than unsatisfactory. CONCLUSION: Of the three tests reported on, only Teflon was consistently rated "good". However, as an-all-purpose gasket material it cannot be recommended because of the price of approximately $14.00 per pound. Teflon should be used therefore only in services where it can be economically Justified. The results obtained from this test indicate that Crane 888 tested best or at least among the best in each of the separate exposures. Thus if one material were to be recom mended for use in all of the three services tested. Crane 888 would be the logical choice. However, there is insufficient data on these test racks to definitely state that other neoprene asbestos sheets are all inferior to Crane 888. Garlock 7228 and Sterling 417 are definitely inferior as they induced stress cracking. However, even the Crane 888 is not completely satis factory in that it does not prevent, and to some extent actually aggravates the corrosion of the metallic surfaces. The conclu sion is that the Crane 888 aggravates the corrosion to a some what lesser degree in most instances than the other materials tested. It is very plausible to believe that no one specific gasket material will be found economically feasible for use as a standard material and that it will be necessary to prescribe several different materials for a limited number of corrosive conditions -- each one being the best for a specific service or services. In future testing of gasket materials an attempt will be made to expose the test racks In as many different corrosive media as possible. All neoprene asbestos sheets which we have available will be tested against Crane 888 on these racks. The only exceptions to this are Garlock 7228 and Sterling 417 which induce stress cracking. In addition to this hycar bonded sheets and wire inserted hycar and neoprene rubber sheets will be tested. 001552 -5- TABLE I A - Racks 5 and 6 - S0-6-T Base Rating of the different types of gasket materials from the standpoint of effect of corrosive media on the material. Material Teflon Crane 999 Garlock 22 Crane 998 Crane 888 Ankorite 450 R. M. 670 Durabla Ankorite 425 u.s. 1771 ' G - Good P - Pair U - Unsatisfactory Rating * G P P P F FU U U U i. TABLE II A - Test Data and Conditions - Racks 5 and 6 Location - S0-6-T Base Length of Exposure - 188 days (10-14-48 - 4-20-49) Corrosive Media: Acetic Acid - 40-45# Formic Acid - 4-6# Methyl Acetate - 1# H2O - Remainder Aeration - None Agitation - Rapid Temperature - 220 P. 001553 6- - TABLE III A - Results Racks 5 and 6 - S0-6-T Base Brand & Type of Material Teflon (Polytetrafluoroethylene) Crane 999 (Neoprene Rubber) Garlock 22 (Red rubber) Crane 998 (Neoprene, Duck reinforced) Crane 888 (Neoprene asbestos comp.sheet) Ankorite 450 (Neoprene asbestos comp.sheet) R M 670 (Asbestos composition sheet) Durabla (Asbestos composition sheet) Ankorite 425 (Rubber asbestos sheet) US 1771 (Coil resistant rubber) Condition of Material Small blisters containing liquid. Flexible. Hard on surface but soft and resilient inside. Hard on surface, soft inside. Fabric attacked slightly. Hard, but strong. Not as hard as 888, but has less strength. Hard, lost most of strength. Slightly worse than RM 670. Same as Durabla. Very bad. 001554 - 6-A - TABLE IV A - Rack #7, S0-6-T Reboiler Overall rating of the different types of gasket materials tested from the standpoint of conditions listed "under Discussion" in Report #67 for ideal gasket material: Material_______________________________ Rating Garlock 7228 RMK 670 U U KM 238 RM K68 PLost Ankorite 450 P- Crane 888 F G - Good P - Pair U - Unsatisfactory TABLE V A - Test Data and Conditions - Rack #7 Location - S0-6-T Reboiler Length of Exposure - 1370 hours (6-22-4-9 to 8-22-49) Corrosive Media: Acetic Acid - 55-60# Formic Acid - 1-3# Methyl Acetate - 0.5# H20 - Remainder Aeration - None Agitation - Rapid Temperature - 220-240 P. 001555 TABLE V I A - R e s u lts - Rack #7, S 0 -6 -T R e b o ile r Sheet) 001556 TABLE V I A - R e s u lts - Rack #7 S 0 -6 -T R e b o ile r p1 O Cp H P CO o VO CO cq bO 03 Ch C'- CO on IP. O P 03 1 on on rH rH P aCO Z 1 11 C r-t O CO *H H Pp P Oi 03 0 0+3 c ca os a 4J CQ O P 03 rH p -p * *H o p P X3 CO x 03 rH P P * a p pP CO X 03 rH p p * H O f-t PP ca z 03 rH p p *H T3 P PP CO X - OP a co- x: co cq c PP P ft*3 03 < O O i 1 03 C rH IQ CO O 13 X O i-3 P 03 C rH 03 CO O 03 P O P 1 <m CD OX C3 CO C P CO rH P P P CU--- a-p 03 < C oo I 1 03 C P OQ CO O 13 X O pp 1 CQ C rH 03 CO O 03 P O pp O ii (C3 CO 03 ca CQ a-p >>P c E* < o 03 -> p o ^ CO O r-t 03 CO 0a-p >>p c E-1 < O cn n rH 03 C -P ---c- X C0 P C3 S'-- '1 pa E- O *0 C <M CO O rH ca os -p a ca CgQ E o 0a3 zO CO -U 03 O P 1c 03 CO C3 -P PP P CO o C 03 rH O X2 Z CO m VO o 03 O -p Z -H a o ca 0) 4-> OP *H flj 4J O 03 C rH X3 zO CO OJ VO o rH * cca 0 03 3p) rl Prl >ea 03 CO rH 03 rH P (0 H ea a cn cn VO 0 Pa Op C H pa 2 JO 0 ca 4-> p CO z0 OJ VO 0 x 0 O c0 ZP 4-? C0 03 PO P3 rH ap 0 ca 03 Pi P bO bO 1) O CO > C O 03 Drlrt CQ E CO CM VO O j' co QJ 00 C 03 -rH VO 03 O CQ P -p 0 a oj a I s o03 x03i so CZ 01U --< p 03 03 03 o) cn p p C O CO 0) 4-) p o p 03 oinao a osi-a* OPE < <c z03 ca oo 4-> 03 03 03 C 03 03 P 03 O X P p P CQ p0 am JO E0 in 0 03 a J- 03 P <C Z 03 O O p CO 03 03 co C cn 03 00 03 O P CP w 03 a 03 c 0 o> a CO 03 P P Z 03 O O'--< 0 P CO 03 0) CO C 03 03 CO 03 O P P P GO 03 a cn - P Ec 0 03 a CO 03 <P Z 03 O O'-- 0- 00 OV O rH rH CM rH rH 001557 l i i t -9- TABLE VII A - Racks 8 and 9 - AR-3-T Base Overall rating of the different types of gasket materials tested from the standpoint of conditions listed under "discussion" in Report #67 for ideal gasket materials. Material_______________________ Rating Teflon G Durabla F- Crane 888 F- Oarlock 900 U.S. F.O.P. 899 Ankorite 450 Ankorite 425 Sterling 417 Garlock 7228 RM K 670 KM 238 RM K 68 U U U FU U U U U G - Good F - Fair O' - Unsatisfactory TABLE VIII A - Test Data and Conditions - Racks 8 and 9 Location - AR-3-T Base Length of Exposure - 1912 hours (80 days) (7.7-49 to 9-29-49) Corrosive Media: Acetic Acid -0.2# Formic Acid - 1.0# H2O - Remainder Aeration - None Agitation - Rapid Temperature - 230 F. 001558 TABLE IX A - R e s u lts _ Racks 8 and 9 , A R -3-T Base e h a rd e n e d , s o ft a t edge H ardened in c e n te r. S o ft & p lia b le Somewhat fle x ib le . C o n s i d e ra b ly h a rd e i th a n o rig in a l S o ft around edges. C o n d itio n of M a te ria l Pi oX O Pi---I x: co co c -P -P CQ -h a-p ^ A< C Qo on o o E 3 on O iH OX E a ro u n d edges H ardened in c e n te r. S o ft and p lia b le E 3 ^ Op OX A E a ro u n d edges S lig h tly g d , e d e t n a e S lig h tly i x A >>-0 P P A Pi PC x a) a b0*O A A p Pi bOp P A 'O X CO .C A P >*o 1-t Pc X .C A a p bO'O a x iH <H CO Pi CO X *000) a>A1 t*>TJ a P Q> i-l Pc X xb:O'aa 003) PX p u bO CD r1 A "O GO .C A Pi i i fdt r ao hs E 3 OE rH OX <d E E 3 OV E Op OX A E E 3 b- E PP OX eA E 3 C"- E OP OX <0 E VO O O TABLE IX A - R e s u lts - Racks 8 and 9 - A R -3-T Base Pi A O X -P A rs B OH C P -P Pi iH a-p --A< C oo E 3 CVJ E CU p OX A E Pi X OOP. AA A -P (23 a-p >< c Eh O X <3 A -P A PP <A rH Pi Cu o c AO a > E-> 3 bO rH O C A r--t -H Pi rH P A AP C X iH a o 3 A $-4 td c p p3 PO H rH Pi p A rH C AA Pi a > a CCA AAA osu CQ rH AA C P^-- 4 Pi P. C O O A -H pP XA E- S in VO o 00 ^ cu c O eu o C Pi rH t- A P AOA CAP X A Op O A P. C3 Pi P A c aa o aa o A >iP i-t o a a P. E-> A P AX E CO s AZ A O i a---< o a AO i az eo 00 E 3 t"-- OP OX E 3 C-- E Op OX * CO E E 3 E Op OX <0 E E 3 in e OP OX cd E E 3 it- E CU 1-t OX CO E E 3 CO E OP OX <0 E E 3 COO E p OX cO E O C3 AOA PP CUrH H aaa a x: ao o td c CA P3 PO ap cfl PX AA Pi ax a CPA AOA o ca r) T3 C3 AO rH ap A AP AXp oa C A P AA bOP CP oa rH u . -o b0 c CA P3 PO H rH CU rH A PX A P. a a CPA AOA a m T3 4U> rH a o A 4-3 A bO |C o rH w T3 A P AA CP oa rH u A 44 H a a A 44 0 b0 C o rH Pd Pi r-t CO CO r-t 3 trObOOrH C oh n Pi A p O U. P i-i co a -U--------------- 4) > o a) 03 A P CO i-t Pi 3 a p. o 3 rH P. A rH co co rH 3 Pi n bO iH p C c E 43 Pi 4) O 43 03 a u p. r-t CO CO bO CO r-l 3 rH C Pi 3 O 3 iH 0) taO r--l hOP C A r-t 03 a hh 1) p tfl tJ Pi -H o p. x -h Pi a i-i co a. 3 I rO--t Pi rH A 3 CO h-> a x H a 43 03 Pi 43 Pi *H E a 03 P O 43 rH C 03 3 43 a bO P. 43 Pi 43 O E P. > CO o P O Pi a 43 a iH 43 C o c o A c-- VO d- in o in o VO o VO o VoO * m voo voo U\ CM cx --- e o Ao p tH 1^ Pi A P O ^3 A XX A C3 < OS co ov If 3 1 cm a E o AO P H f* Pi A P OX XXV C3X < K CO 1--' O rH a E O o A rH X AP AX A Pi X3 a 3 3 O K CO a E O o A rH X Ap AX A PX o 3 3X Q C CO A 3 C n^i >O . t Pi -p a a 4,e > O 03.-0 i XO :Z or o in 3- 43 CO A C m CO c o 43 O CO Pi P- a O 43 a Ac Pi-- Pai ra a a O X> E __ X E CZ A o Ac zA <AcoO o O'--- rH CU on rH rH m 001560 if lne xc iebntl.i S lig h tly har t dened Edges Hardened. F le xib le Hardened. In center. ges.S ligh hardened .F le x ib le on in center. S ligh ened leyd gheas .r d F le x ib le on on edges. fhlaerxd iebnleed, on edges. S lig h tly hardened, F le xib le S lig h tly fEldegxeisb le . hardened. S lig h tly C ondition of M aterial HCg. fw leyd O X L jPr Paa cq -cH aa-xp co>-- Q E3 O -EH OX CO CiHU 3E *EH oX C0 iOn *E3EH oX * <E0 E3 OOON -EXH * CEO E VOO -3EH OX CEO Ot>- *3EH OX cEO in 3EE Oo -XH cEd O *3EEH oX cEd m orH TABLE IX A - R e s u lts - Racks 8 and 9 A R -3-T Base Ooh xCC38P0HC)O---C* Paa-xPp C- -H c -- Qo e 3 OON O E -XH * CO E <Vh Jsi S-t O O CO fCfO) 0a>3>--pP< co f? JoO -CPpO a} < -p a) Ooh raH a>0) co s o CbOrCHO 03 -P jdii-nHft oc -0p3 -CpO aa cLoi aa; a> rH <U T3 3 'D -H bO 4) SZ 03 (h-P 01 MSO E 3 OOON E -XH E L> <d rH $ 3O bO rH 0) rH 03 LLii m JcCZoO -taPH bcO th aa> -Pw <4a) H O.Xc3 Li CO CO iH 3O bO rH 0) rH Li CO l. x: M CO E3 OOON -EXH CO E Li <0 r3MaH> os L4i -"PH ma a -p a Li CO rH 3 bO 0uu3 M E3 CCOU -EH OX CO E T3 03 P CO bCO m O -P wlH aH -Su3 a o 43 P (0 bcoO wrH 00c9} CO lH* *--ch ida; -h o o -P -- cu VoO m mo VoO VoO VoO * h cfl JZ 2 -i oia i--( C<0 Li O Orl o at aL)l cucd >a-zpco A CO 03 03 CcoO ac> -op--- aC; oco C)4Jr O 03 E C03 43 O03 -P r-s ON Jbtf CO Li O C3 a> aL> a03) a CO L O m a 03 * a CnO aL. n03 ft O 03 03 03 O cCo aO; 3<n oE i OC--i2- <0 OJ O 3 E z s43 <0s3 oO isi'-' CU O 3 E rl -P S'5--053 <CC OO 0oLi --c3 30c3u3 aoa> so CO VO rH itH-- CO rH ON rH E3 Oin O E -hX * to E o 03 P 03 bCO 03 O -P fH H wa a -p a cLoi lH 3 & 03 u L, IH o VoO oO cio Ea ON C0 0 O X L, C3 03 03 03 O rl3-P Li 3 S3 00 --CC 3033 o CU E3 C'- E OO -XH * cO E m E3 e opH XH CO E CU 3E E rH H oX * cd E e 3 OO PEX 2 E | 1 : 03 03 -P CO CO cu raH E CCOO co 03 Li c 03 43' C3 Li CO 43 > Li O3 CO bO C -P H 03 PX -P CO ^H cd Lu bO in VoO 03 03 P CO CO 03 raH s 6C0O OC 0L3i H c 03 0C33 Li CO 03 Lh > Li O3 09 b0 CP H 03 PX P 03 H cd CU bd in VoO co uc eo 03 03 *PCHU cO03 S0CO3 Li Jp0oZ3 rl cU0d) oC03 5C0 -**H-> *aH 4C-3O aH bc0 4*HH^ Li aH cd rH rH 3 cc bO 03 4^3 Li C Li 03 IH a in VoO viono * i ' i r> m Li O i a 1o Op CU 03 O 33 O. 333 0o3 fcCP COaDNx03 oEo * Cu 03 03 03 03 --Z'3033 00 <CC . --- as 4) W ON O 3o3. E3C0 < O fit, 33 333 O P S3 VO I3O3C'--S 313) S CC fl3r^ l- A 03 E Lai oO crHu CU CU mcu ii O43 03 03---- i S 33 03 a ao co oE VsO 0c30 03 03 S Li O cc ap cu 001561