Conducting Corrosion Tests in Field Applications: Standard Guide For

 

 

Designation: G4  −  01 (Reapproved 2014)

Standard Guide for

Conducting Corrosion Tests in Field Applications1 This standard is issued under the fixed designation G4; the number immediately following the designation indicates the year of original adoption adopt ion or, in the case of revis revision, ion, the year of last revision. A number in parentheses parentheses indicates indicates the year of last reapproval. reapproval. A super superscrip scriptt epsilon (´) indicates an editorial change since the last revision or reapproval.

G30   Pr Prac acti tice ce fo forr Ma Maki king ng an and d Us Usin ing g UU-Be Bend nd St Stre ress ss-Corrosion Test Specimens G36 Practice G36  Practice for Evalua Evaluating ting Stress Stress-Corr -Corrosion osion-Crack -Cracking ing Resistanc sist ancee of Met Metals als and Alloys Alloys in a Boil Boiling ing Magnesiu Magnesium m Chloride Solution G37 Practice G37  Practice for Use of Mattsson’s Solution of pH 7.2 to Evaluate the Stress-Corrosion Cracking Susceptibility of  Copper-Zinc Alloys G41  Pra G41 Practic cticee for Det Determ erminin ining g Cra Cracki cking ng Sus Suscep ceptibi tibility lity of  Metals Exposed Under Stress to a Hot Salt Environment G44 Practice G44  Practice for Exposure Exposure of Metals and Alloys by Altern Alternate ate Immersion in Neutral 3.5 % Sodium Chloride Solution G46 Guide G46  Guide for Examination and Evaluation of Pitting Cor-

1. Sco Scope pe 1.1 This guide covers procedures procedures for conducting conducting corro corrosion sion tests in plant equipment or systems under operating conditions to evaluate the corrosion resistance of engineering materials. It does not cover electrochemical methods for determining corrosion rates. 1.1.1 While intended intended primarily for immersion immersion tests, general guidel gui deline iness pro provid vided ed can be app applica licable ble for exp exposu osure re of test specimens in plant atmospheres, provided that placement and orie or ient ntati ation on of th thee tes testt sp speci ecimen menss is no nonn-re rest stri ricti ctive ve to air circulation. 1.2 The values values stated in SI uni units ts are to be reg regard arded ed as the standard. The values given in parentheses are for information only.

rosion G47  Test Method for Determining Susceptibility to StressG47 Test Corr Co rros osio ion n Cr Crack ackin ing g of 2X 2XXX XX an and d 7X 7XXX XX Al Alum umin inum um Alloy Products G58 Practice G58  Practice for Preparation of Stress-Corrosion Test Specimens for Weldments G78 Guide G78  Guide for Crevice Corrosion Testing of Iron-Base and Nicke Nic kell-Bas Basee Sta Stain inles lesss Al Allo loys ys in Se Seaw awat ater er an and d Ot Othe herr Chloride-Containing Aqueous Environments 2.2   NACE Standard:4 RP0497   Fie RP0497 Field ld Cor Corros rosion ion Eva Evaluat luation ion Usi Using ng Met Metalli allicc Test Specimens

1.3   This standar standard d doe doess not purport purport to add addre ress ss all of the safet sa fetyy co conc ncer erns ns,, if an anyy, as asso socia ciated ted wi with th it itss us use. e. It is th thee responsibility of the user of this standard to establish appro priate safety and health practices and determine the applicability of regulatory limitations prior to use. See also 10.4.2 also  10.4.2.. 2. Referenc Referenced ed Documents Documents 2.1   ASTM Standards:2 A262 Practices A262  Practices for Detectin Detecting g Suscep Susceptibility tibility to Inter Intergranu granular lar Attack in Austenitic Stainless Steels E3 Guide E3  Guide for Preparation of Metallographic Specimens G1 Practice G1  Practice for Preparing, Cleaning, and Evaluating Corrosion Test Specimens

3. Signi Significanc ficancee and Use NOTE  1—This guide is consistent with NACE Standard RP0497.

G15 Terminology Relating to Corrosion and Corrosion TestG15 Terminology ing (Withdrawn ing  (Withdrawn 2010)3 G16 Guide G16  Guide for Applying Statistics to Analysis of Corrosion Data

3.1 Obs Observ ervatio ations ns and data der derived ived fro from m cor corros rosion ion test testing ing are used to determine the average rate of corrosion or other types typ es of atta attack, ck, or bot both h (se (seee Term ermino inolog logy y   G15 G15), ), that occur during the exposure interval. The data may be used as part of  an evaluation of candidate materials of construction for use in similar service or for replacement materials in existing facilities.

1

This guide is under the jurisdiction of ASTM Committee G01 Committee  G01 on  on Corros Corrosion ion of  Metals and is the direct responsibility of Subcommittee   G01.14   on Corrosion of  Metals in Construction Materials. Current edition approved Nov. 1, 2014. Published November 2014. Originally approved in 1968. Last previous edition approved in 2008 as G4–01 (2008). DOI: 10.1520/G0004-01R14. 2 For referenced ASTM standards, visit the ASTM website, www.astm.org, or contact ASTM Customer Service at [email protected]. For  Annual Book of ASTM  Standards volume information, refer to the standard’s Document Summary page on the ASTM website. 3 The last app approve roved d vers version ion of this historica historicall sta standa ndard rd is ref refere erence nced d on

3.2 The data developed developed from from in-plant tests tests may also be used as guide lines to the behavior behavior of existin existing g plant materials for the purpose of scheduling maintenance and repairs. 3.3 Corrosi Corrosion on rat ratee dat dataa der derived ived from a sin single gle exp exposu osure re generally gener ally do not provide information on corro corrosion sion rate chang changee

4

www.astm.org.

Available from NACE International (NACE), 1440 South Creek Dr., Houston,

TX 77084-4906, http://www.nace.org.

Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States

1

 

 

G4 − 01 (20 (2014) 14)

versuss time. Corrosion rates may increa increase, se, decrease, or remain versu constant, depending on the nature of the corrosion products and the effects effects of incub incubation ation time required at the onset of pittin pitting g or crevice corrosion.

4.7.4 4.7. 4 Pa Part rtin ing g or de deal allo loyi ying ng is a co cond nditi ition on in wh which ich on onee constit con stituen uentt is sele selectiv ctively ely rem remove oved d fro from m an allo alloy y, as in the dezincification of brass or the graphitic corrosion of cast iron. Close Clo se atte attenti ntion on and a mor moree sop sophis histica ticated ted eva evalua luation tion than a simp si mple le ma mass ss lo loss ss mea measu sure reme ment nt ar aree re requ quir ired ed to de detec tectt th this is phenomenon. 4.7.5 Pitting corrosion corrosion cannot cannot be evaluat evaluated ed by mass loss. It is pos possib sible le to mis misss the phe phenom nomeno enon n alto altoget gether her whe when n usi using ng small test specimens since the occurrence of pitting is often a

4. Limi Limitatio tations ns 4.1 Metal specimens specimens immersed immersed in a specifi specificc liquid may not corrode at the same rate or in the same manne corrode mannerr as in equipment equipment in which the metal acts as a heat transfer medium in heating or cooling the liquid. In certain services, the corrosion of heatexchanger tubes may be quite different from that of the shell or heads. This phenomenon also occurs on specimens exposed in gas streams from which water or other corrodents condense on cool surfaces. Such factors must be considered in both design and interpretation of plant tests.

statistical and its incidence can be directly related to the areaphenomenon of metal exposed. 4.7.6 4.7 .6 Stress-co Stress-corro rrosio sion n cra cracki cking ng (SC (SCC) C) may occ occur ur und under er conditions of tensile stress and it may or may not be visible to thee na th nake ked d ey eyee or on ca casu sual al in insp spec ectio tion. n. A me metal tallo logr grap aphi hicc examination examin ation (Prac (Practice tice   E3 E3)) wi will ll co confi nfirm rm th this is me mech chan anis ism m of  attack. SCC usually occurs with no significant loss in mass of  the test specimen, except in some refractory metals. 4.7.7 4.7 .7 A number number of rea reactiv ctivee meta metals, ls, mos mostt not notabl ably y tita titaniu nium m and zircon zirconium, ium, develo develop p stron strongly gly adhere adherent nt corro corrosion sion product films in corrosive environments. In many cases, there is no accep ac ceptab table le me meth thod od to re remo move ve th thee fil film m wit witho hout ut re remo movi ving ng signifi sig nifican cantt unc uncorr orrode oded d meta metal. l. In thes thesee cas cases, es, the ext extent ent of  corrosion can best be measured as a mass gain rather than mass loss. 4.7.8 4.7 .8 Som Somee mat materia erials ls may suf suffer fer acc acceler elerated ated cor corros rosion ion at liquid to atmospheric transition zones. The use of small test specimens may not adequately cover this region.

4.2 Ef Effects fects caused by high velocity, velocity, abras abrasive ive ingred ingredients, ients, etc. (which may be emphasized in pipe elbows, pumps, etc.) may not be easily reproduced in simple corrosion tests. 4.3 4. 3 The be beha havi vior or of ce cert rtain ain me meta tals ls an and d all alloy oyss ma may y be profoundly influenced by the presence of dissolved oxygen. It is es esse sent ntial ial th that at th thee tes testt sp spec ecim imen enss be pl place aced d in lo locat catio ions ns representative of the degree of aeration normally encountered in the process. 4.4 Corro Corrosion sion products products from the test specimens may have undesir unde sirab able le ef effe fects cts on th thee pr proc oces esss str stream eam an and d sh shou ould ld be evaluated before the test. 4.5 Co 4.5 Corr rros osio ion n pr prod oduc ucts ts fr from om th thee pl plan antt eq equi uipm pmen entt ma may y influence the corrosion of one or more of the test metals. For example, when aluminum specimens are exposed in coppercontain con taining ing sys systems tems,, cor corrod roding ing cop copper per wil willl exe exert rt an adv advers ersee effec ef fectt on the cor corros rosion ion of the alu alumin minum. um. On the con contrar trary y, stainless steel specimens may have their corrosion resistance enhanced by the presence of the oxidizing cupric ions.

5. Test Specimen Design Design 5.1 Before the size, shape, and finish of test specimens specimens are specifi spec ified ed,, th thee ob obje ject ctiv ives es of th thee te test st pr prog ogra ram m sh shou ould ld be determ det ermine ined, d, tak taking ing int into o con consid sidera eratio tion n any res restric trictio tions ns that might mig ht dic dictate tate fab fabric ricatio ation n req requir uiremen ements. ts. The dur duratio ation, n, cos cost, t, confidence level, and expected results affect the choice of the shape, finish, and cost of the specimen. 5.1.1 Test specimens are generally fabricated fabricated into disks or rectangular shapes. Other shapes such as balls, cylinders, and tubes are used, but to a much lesser extent. 5.1.2 Disks are normally normally made by one of three methods: methods: (1)

4.6 The accumulation accumulation of corrosion products products can sometimes have harmful effects. For example, copper corroding in intermediate strengths of sulfuric acid will have its corrosion rate increased as the cupric ion concentration in the acid increases. 4.7 Tests covered by this guide are predominantly predominantly designed to inv investi estigate gate gen genera erall cor corros rosion ion;; how howeve everr, oth other er for forms ms of  corrosion may be evaluated. 4.7.1 4.7 .1 Galv Galvani anicc cor corros rosion ion may be inv invest estigat igated ed by spe special cial devi de vices ces th that at co coup uple le on onee sp spec ecime imen n to an anot othe herr in el elect ectri rical cal contact. It should be observed, however, that galvanic corrosion can be greatly affected by the area ratios of the respective metals. 4.7.2 4.7 .2 Crev Crevice ice or con concen centra tration tion cell cor corros rosion ion may occ occur ur when whe n the metal sur surfac facee is par partial tially ly blo blocke cked d fro from m the bulk  liquid, as under a spacer. An accumulation of bulky corrosion products between specimens can promote localized corrosion of some alloys or affect the general corrosion rates of others. Such accumulation should be reported. 4.7.3 4.7 .3 Select Selective ive cor corros rosion ion at the gr grain ain bo bound undari aries es (fo (forr example, intergranular corrosion of sensitized austenitic stainless le ss st stee eels ls)) wi will ll no nott be re read adil ily y ob obse serv rvab able le in ma mass ss lo loss ss measurements and often requires microscopic examination of  the specimens after exposure.

by punching from sheet material, (2) by slicing from a bar, or (3) by trepanning by a lathe or mill. Punched disks are by far thee lea th least st ex expe pens nsiv ivee an and d sh shou ould ld be co cons nsid ider ered ed if ma mater teria iall thickness is not a limitation. Some of the positive characteristics of disks are: (1) the surface area can be minimized where there is restricted space, such as in pipeline applications, ( 2) disks can be made inexpensively if a polished or machined surface finish is not required, and ( 3) edge effects are minimized for a given total surface area. Some negative characteristics are: (1) disks are very costly to fabricate if a ground finish and machined machined edg edges es are req requir uired, ed, (2) dis disks ks fab fabrica ricated ted fro from m sheet sh eet ma mater terial ial re resu sult lt in a co cons nsid ider erab able le amo amoun untt of sc scrap rap materia mate rial, l, and ( 3) di disk skss sli sliced ced from from a ba barr pr pres esen entt a su surf rface ace orientation orien tation that can result in extens extensive ive end-grain attack. Using a bar is undesirable unless end-grain effects are to be evaluated. 5.2 Recta Rectang ngul ular ar spe specim cimen enss are fa fabr brica icated ted by eit either her punching, punchi ng, she shearin aring, g, or saw cut cuttin ting. g. Pun Punche ched d dis disk k sha shaped ped specim spe cimens ens are the mos mostt eco econom nomical ical if the qua quantit ntity y is suf suffficiently high to justify the initial die cost. Fabrication is more 2

 

 

G4 − 01 (20 (2014) 14)

cost-effective for rectangular specimens than for disks when ground finished and machined sides are required, and they can be made using very few shop tools. In some cases, rectangular specimens are more awkward to mount.

heat affected zone. For example, gas tungsten arc welding has lowe lo werr he heat at in inpu putt th than an ox oxyg ygen en fu fuel el we weld ldin ing g an and d cau causes ses a narrower heat affected zone, which is also closer to the weld bead.

5.3 Materi Material al availab availability ility and machin machinability ability also af affect fect the cost of producing all types of specimens. Before the shape and size are specified, the corrosion engineer should determine the characteristics of the proposed materials.

7. Preparat Preparation ion of Test Test Specimens 7.1 Contro Controversy versy exists exists as to whether the test specimen edges edges should be machined. The cold-worked area caused by shearing or punching operations can provide valuable information on

6. Test Specimens

alloy sus alloy suscep ceptib tibilit ility y to str stress ess cor corros rosion ion cra crackin cking. g. Als Also, o, the ability to compare information among specimens of different materia mate rials ls can be af affec fected ted by the amount amount of col cold d wor work k per per-formed on the material. Therefore, the decision to machine and to test specimens with/without the residual stresses associated with cold work should be made on a case-to-case basis. 7.1.1 The depth of cold work associated associated with punching punching and shearing operations typically extends back from the cut edge to a distance equal to the specimen thickness. Removal of the cold col d wor worked ked areas can be per perfor formed med by gri grindi nding ng or car carefu efull machining the specimen edges. 7.1.2 7.1 .2 Ideally Ideally,, the sur surfac facee fini finish sh of the spe specime cimen n sho should uld replicate that of the surface finish of the material to be used for equipment fabrication. However, this is often difficult because the finish on materi materials als varies between mills, between sheet and plate and even between heat treatments. The mill scale and the amountt of oxides on the surface can vary as well. Also, surface amoun surface finishes are difficult to apply to edges that have been distorted by pun punchi ching ng or she sheari aring. ng. Sin Since ce the pri primar mary y req requir uiremen ementt is usually to determine the corrosion resistance of the material itself, a clean metal surface is most often used. The purpose of  thee tes th testt di dict ctate atess th thee re requ quir ired ed fin finis ish h of th thee sp spec ecim imen en.. Fo Forr instance, instan ce, for water treatin treating g applic applications, ations, relative changes of  weights of specimens are usually compared to optimize inhibitor additions. The specimens are gener generally ally punched or sheare sheared d and finished by blasts with glass beads. This is one of the most economical ways of preparing corrosion test specimens. Manufactur fac turing ing var variab iables les in spe specime cimen n pre prepar paratio ation n tha thatt can be removed reasonably reasonably shou should ld be elimina eliminated. ted. A standard surface finish facilitates facilitates the comparison comparison of results among test samples samples..

6.1 The size and shape of test specimens specimens are influenced by several factors and cannot be rigidly defined. Sufficient thickness ne ss sh shou ould ld be emp emplo loye yed d to mi mini nimi mize ze th thee po poss ssib ibili ility ty of  perforation of the specimen during the test exposure. The size of the specimen should be as large as can be conveniently handled, the limitation being imposed by the capacity of the available analytical balance and by the problem of effecting entry into operating equipment. 6.2 A convenient convenient size for a standa standard rd corrosion corrosion disk shaped specimen is 38 mm (1.5 in.) in diameter and 3 mm (0.125 in.) in thickness with an 11 mm (0.438 in.) hole in the center of the round specimen. This size was arrived at as being the maximum size that could easily effect entry through a normal 38 mm noz nozzle zle.. How Howeve ever, r, it is also convenien convenientt for larger larger size nozzle entries as well as for laboratory corrosion testing. A convenient standard specimen for spool-type racks measures 25 by 50 by 3 mm (1 by 2 by 0.125 in.) or 50 by 50 by 3 mm (2 by 2 by 0.125 in.). A round specimen of 53 by 3 mm (2 by 0.125 in.) or 55 by 1.5 mm (2 by 0.062 in.) is sometimes employed. These last three measure about 0.005 dm 2 in surface area. 6.3 Other sizes, shapes, and and thicknesses of specimens can be used for special purposes or to comply with the design of a special spe cial typ typee of cor corros rosion ion rac rack. k. Spe Special cial des design ignss sho should uld be reduced to a few in number in preliminary tests; special designs should be employed to consider the effect of such factors of  equipm equ ipmen entt con constr struc uctio tion n and ass assemb embly ly as hea heatt tre treatm atment ent,, welding weld ing,, sol solder dering ing,, and col cold-w d-work orking ing or oth other er mech mechani anical cal stressing. 6.4 Si 6.4 Sinc ncee we weld ldin ing g is a pr prin incip cipal al me meth thod od of fa fabr brica icatin ting g equipment, welded specimens should be included as much as possible in the test programs. 6.4.1 Aside from the effects effects of residual stresses, stresses, the main items ite ms of in inter teres estt in a we weld lded ed sp spec ecime imen n ar aree th thee co corr rros osio ion n resis re sista tanc ncee of th thee we weld ld be bead ad an and d th thee he heat at af affe fect cted ed zo zone ne.. Galvanic effects between weld metal and base metal can also be eva evalua luated ted.. The wel weld d and heat af affec fected ted zon zonee reg region ionss are relatively small; therefore, welded specimens should be made slightly slightl y large largerr than the norma normall nonnon-welded welded specimens specimens when possible, for example, 50 by 75 mm (2 by 3 in.). The optimum method of welding corrosion test specimens is to join the two halv ha lves es us usin ing g a si sing ngle le ve veee or do doub uble le ve veee gr groo oove ve wi with th fu full ll penetration and multiple passes. Double vee joint preparation is used for very thick samples. Machining the weld flush is optional, depending on how closely the sample will be examined afterward (see Practice  Practice   G58 G58). ). 6.4.2 The welding process process and numbe numberr of passes influence influence the heat input and, consequently, the width and location of the

7.2 Some of the available available finishes are: 7.2.1 Mill finish (pickled, (pickled, bright annealed, annealed, or shot blasted blasted), ), 7.2.2 Electro Electrolytic lytic polished, polished, NOTE 2—Electrolytic polishing can produce a surface layer enriched in some alloying elements while depleted in others. For example, chromium is enriched on stainless surfaces and sulfur is depleted.

7.2.3 Blasted with sand sand or steel shot, NOTE   3—Blasting many metals with sand can cause embedded sand particles and steel shot can cause surface contamination with iron or iron oxide. Glass beads are better, but not if broken pieces are allowed to be used in the blasting.

7.2.4 San 7.2.4 Sanded ded wit with h abr abrasi asive ve clo cloth th or pap paper er (fo (forr exa exampl mple, e, SiC), 7.2.5 Machin Machinee finished, and 7.2.6 Passiv Passivation ation of stainless stainless steel with nitric acid to remove remove surface iron contamination and other chemical cleaning methods used, for example, after welding. 3

 

 

G4 − 01 (20 (2014) 14)

FIG.. 1 Fla FIG Flatt Bar Rack  Rack 

7.3 The sur surfac facee fini finish sh mos mostt wid widely ely used is pro produc duced ed by sanding with an abrasive cloth or paper. Sanding removes the mill scale and oxides as well as other defects in the material such suc h as scr scratc atches hes,, pits pits,, etc. etc.,, tha thatt cou could ld pro produc ducee misl mislead eading ing results when the data are being analyzed. 7.3.1 A 120 grit finish is generally generally acceptable acceptable and is readily produced without the need for specialized equipment. Other surface finishes may be obtained through the appropriate use of  abrasiv abr asivee pap papers ers and clo cloth. th. In ord order er to pre preven ventt meta metallu llurg rgical ical change cha ngess tha thatt cou could ld af affec fectt the cor corros rosion ion res resista istance nce,, the test sample should be cooled during fabrication. Wet sanding is one method meth od of pre preven ventin ting g spe specime cimens ns fro from m hea heatin ting g up. In man many y cases, it is necessary to begin sanding with coarse abrasives

areas. Not areas. Note, e, how howeve ever, r, tha thatt alth althoug ough h the pre presen sence ce of suc such h localized attacks is a positive indication, absence of attack is not a guarantee of immunity from attack in operating equipment. 9.1. 9. 1.1 1 A ma map p sh shee eett id iden entif tifyi ying ng th thee lo locat catio ion n of th thee te test st specimens on the test rack described below is useful.

and progressively move to finer abrasives. 7.3.2 Clean polishing polish ing belts should be used to avoid contamination of the metal surface, particularly when widely dissimilar metals are being finished. For example, a belt used to sand brass should not be used to sand aluminum. Particles of one metal met al co coul uld d be beco come me im imbe bedd dded ed in th thee ot othe herr, re resu sulti lting ng in erroneous data.

the ra the rack ck.. With th this is ac acco comp mpli lish shed ed,, th thee ty type pe of ra rack ck can be selected. 10.1.1 10.1. 1 Specime Specimens ns are usual usually ly electrically isolated from one another and the rack unless special effects, such as galvanic corrosion, are under study. Insulation is achieved by sleeving all metal parts in contact with the specimens and separating them with washers. The sleeves and washers should be made from a nonconductive material such as polytetrafluoroethylene (PTFE) fluorocarbon or ceramic material.

10. Test Rack Design and Test Location Location 10.1 The purpose purpose of the rack is to support support test specimens in the process environment at the proper location and orientation. To accomplish this, the corrosion engineer should first determine the number, size, and spacing of the specimens to be tested and then establish the proper location and orientation of 

7.4 Test specimens should should be cleaned and the initial mass determined (see Practice  Practice   G1 G1)). 7.5 A pre-exposu pre-exposure re inspection of test specim specimens ens should be conducted in order to identify any pits, mechanical scratches, or residual surface treatment artifacts that could influence the corrosion behavior of the specimen.

10.2 The rack should 10.2 should be as simple as pos possib sible, le, but it als also o should be sturdy and constructed of materials resistant to the test environment. Bolts should be spot welded or double nuts used to preve prevent nt loosen loosening ing during expos exposure. ure. Occasionally an insu in sulat lated ed bo bolt lt is al alll th that at is ne nece cess ssar ary y to su susp spen end d th thee tes testt specimens. Handling this assembly requires a few more precautions than some other mounting systems but is cost effective in ma many ny in insta stanc nces es.. An Anot othe herr me meth thod od is to su susp spen end d th thee tes testt specimens by an insulated wire. This system can be used in a storagee tank or other nonagitated storag nonagitated vessels; for examp example, le, as used in chemical cleaning operations. 10.2.1 10.2. 1 A flat bar rack is usually usually made of rigid rigid material, material, such as 6 mm (0.0250 in.) thick plate, and is approximately 25 mm (1 in.) wide by 305 mm (12 in.) long. With a few mounting holes at one end, a flat bar rack is capable of supporting several specimens. The other end is attached in the process location either by welding, bolting, or clamping. See  Fig. 1. 1. 10.2.2 10. 2.2 Typi ypical cal rac racks ks are app approx roximat imately ely 305 mm (12 in.) long with 15 mm (0.625 in.) spacing between specimens. A spool rack, with adjustable plates, can be used to mount up to 36 specimens. With the support bars on the sides, the rack can be handled without touching the specimens. The rack can be easil ea sily y mo moun unted ted by str strip ipss th that at ar aree at attac tache hed d to th thee to top p an and d bottom bot tom.. The These se str strips ips can be wel welded ded,, bol bolted ted,, or clam clamped ped in place. See Fig. See  Fig. 2. 2.

8. Numb Number er of Test Specimens Specimens 8.1 In general, general, at least duplicate duplicate specimens should should be tested. If possible, in cases in which confidence confidence limits are required for corrosion rate measurement, then somewhere between 5 and 10 repl re plica icates tes sh shou ould ld be ru run, n, de depe pend ndin ing g on th thee sc scop opee of th thee prog pr ogra ram. m. Th Thee co confi nfide denc ncee le leve vell can be es estab tablis lishe hed d by th thee procedures shown in Guide G16 Guide  G16.. The duplicate samples should be widely separated on the test rack rather than adjacent to one another. The results for the samples should also be reported separately. 9. Identification of Test Test Specimens 9.1 Alt Althou hough gh it may be nec necess essary ary in spe special cial instance instancess to notc no tch h th thee ed edge ge of th thee sp spec ecime imens ns fo forr id iden entifi tificat catio ion, n, it is pref pr efer erab able le th that at th they ey be st stam ampe ped d wit with h a co code de nu numb mber er.. Th Thee stamped number has an additional advantage in that, should a specime spe cimen n sho show w a pre prefer ferenti ential al atta attack ck at the stam stamped ped area, a warning is given that the material is susceptible to corrosion when cold worked. It is also possible in some instances to detect stress-corrosion cracking emanating from the stamped 4

 

 

G4 − 01 (20 (2014) 14)

FIG. 2 Typical Spool Spool Rack 

10.2.3 A pipeline 10.2.3 pipeline rack is designed designed to fit between the flanges flanges in a pipeline. It can also be used at a nozzle. Because of the

rack could do extensive damage. (Test racks should be inserted after the pumps to prevent damage to the impeller in case of 

cantilever support and pipe diameter, the number of specimens that th at can be mo moun unte ted d on th this is sy syst stem em is re restr stric icted ted.. De Desi sign gn modification can be made in order to increase the number of  specimens. A potential problem with the pipeline rack is the flow restriction in the pipeline. See  Fig. 3. 3.

rack failur failure.) e.) 10.4 Retractable specimen holde holders rs overco overcome me the greate greatest st limitation of most forms of in-plant testing, which is the need to shut down in order to remove the test rack from the process. The arrangement consists of a 50 mm (2 in.) or larger nozzle that th at is fit fitted ted with a fu fully lly openin opening g ga gate te or pl plug ug va valv lve. e. Th Thee rodro d-sh shap aped ed sp speci ecime men n ho holde lderr is co conta ntain ined ed in a re retra tracti ction on chamber, which is flanged to the valve, and is fitted additionally all y wit with h a dr drai ain n va valv lvee (s (see ee   Fig. Fig. 3). The other other en end d of the retraction chamber contains a packing gland through which the specimen holder passes. The test specimens are mounted on the rod ro d in th thee ex exten tende ded d po posi sitio tion n an and d ar aree th then en dr draw awn n in into to th thee retraction chamber. The chamber is bolted to the gate or plug valve, which is then opened up to allow the specimens to be

10.3 One of the most common common reasons for for the failure of test racks is selecting fasteners that do not resist the environment. Since Sin ce the bol boltin ting g har hardwa dware re is usu usually ally highly highly str stress essed ed and contain con tainss cre crevice vices, s, cor corros rosive ive atta attack ck on fas fastene teners rs can occ occur ur rapidly rapidl y. Another common reaso reason n for failure is defecti defective ve welding of the test rack components or of the test rack to the vessel. Full-p Ful l-pene enetra tration tion welds should should be use used, d, and the are areaa to be welded should be thoroughly cleaned. Fatigue failures caused by equipment vibration or high flow rates is another leading cause of rack failures. With proper design, a rack can be built that will eliminate these failures. 10.3.1 10. 3.1 Pro Proble blems ms cau caused sed by fai failur luree of a mou mounti nting ng sys system tem also should be considered in designing the test rack. In many cases, such as with agitated vessels, pumps, etc., a loose test

moved int moved into o the ope operat rating ing env enviro ironme nment. nt. The seq sequen uence ce is reversed to remove the specimens and the process is cleared from the ret retrac raction tion cha chambe mberr bef before ore dis discon connec nectin ting g it to acce access ss the specimens, specim ens, see see Fig.  Fig. 4. 4. 5

 

 

G4 − 01 (20 (2014) 14)

FIG. 3 Pipe Insertion Insertion Rack  Rack 

10.4.1 All components of retractable specimen holders must be suitably corrosion resistant and fabricated to standards that comply with the equipment design code. The consequences of  a pr proc ocess ess lea leak k mu must st be ca care refu fully lly co cons nsid ider ered ed.. Ret Retra racta ctabl blee specimen holders are best considered in low pressure systems, that is, 1 MPa (about 150 psi) or less. However, commercially available probes and retrieval tools are available for service in systems up to 20 MPa (3000 psi). 10.4.2   Warning—In using ng retr retracta actable ble spe specime cimen n hol holder derss Warning—In usi on-line with either hot, pressurized fluids or hazardous fluids, or both, the possibility of a serious leak (or blowout) at the packing gland must be considered and appropriate precautions taken. Provisions should be made to purge and dispose of the process pro cess fluid fro from m the cav cavity ity where the specimens specimens are hel held d before they are removed from the system. Restraining devices

must be use must used d whe when n rem removi oving ng spe specime cimens ns whi while le the inte interna rnall system is pressurized. 10.5 Selectio Selection n of the proce process ss location is critical to obtaining obtaining meaningful data. The three basic process locations are (see  Fig. 4): (1) immersed stagnant, for example, the boot of the filter where whe re dea deaera erated ted con condit dition ions, s, sol solid id sett settleme lements nts pre prevail vail,, (2) immersed flowing, for example, in piping where aeration, gas and solids entrainment, and turbulence or velocity exert effects, (3) spl splash ash,, wat waterli erline, ne, or liq liquid uid lev level el whe where re the con condit dition ionss simulate partial immersion or spray. When calculating corrosion rates, the test time is not reduced to compensate for partial immersion conditions. 10.6 In certain situations, situations, process conditions conditions (in addition to the three basic locations) must be considered. For example: 6

 

 

G4 − 01 (20 (2014) 14)

FIG. 4 Retra Retractable ctable “Slip-In” “Slip-In” Specimen Specimen Holder

10.6.1 10.6. 1 Velocity effects effects should be consi considered dered if the specimens are laid out flat and parallel to the flow. If the specimens are arranged any other way, they tend to shield one another from the turbulence. The location of the specimens is critical in simulating the turbulence experienced, for example, at a pipe

taining the records in this log book throughout the test should be specifically assigned to one individual. 12.3 For specimens specimens of materials that cannot be stamped (for (for example, too hard or brittle), a system of notches can be used to identify individual specimens. Notches may be formed by filing or grinding.

wall. 10.6.2 10. 6.2 Conden Condensat sation ion (de (dew w poi point nt or   cold cold fing finger  er ) ef effec fects ts should shoul d be consi considered. dered. The test specim specimens ens represent a dif differen ferentt mass effect from a pipe, vessel wall, tube sheet, tube, etc. It may be necessary to expose several sets of specimens in a line to determine the optimum condition that duplicates condensation in the equipment. 10.6.3 10.6. 3 The effects effects of heat transfer (for example, example, when the tube wall is heated) are impossible to duplicate with conventional specimens.

13. Insta Installat llation ion of Speci Specimen men Holder 13.1 13. 1 The location location of the test spe specime cimens ns in the operatin operating g equipment will be governed by the information that is desired. This may require tests at more than one location in the same piece of equipment, such as below the level of the test liquid, at the level of the liquid, or in the vapor phase. 13.2 It is des 13.2 desira irable ble to hav havee the specimen specimen hol holder der securely securely fixed in place. The preferred position of the holder is with the long lon g axi axiss hor horizo izontal ntal so as to pre preven ventt dri drippa ppage ge of cor corros rosion ion produc pro ducts ts fro from m one spe specime cimen n to the oth other er.. Pre Prefer ferabl ably y, the specimen should be so placed that any flow of liquid will be agains aga instt the edges of the specimens specimens.. The same con condit dition ion of 

11. Selection of Materials for Evaluation Evaluation 11.1 The following 11.1 following materials, at least, should be consid considered ered for inclusion as controls: 11.1.1 11 .1.1 The material currently currently used in the proce process ss equipment equipment in which the test is being run or in the equipment of interest. 11.1.2 11 .1.2 A materia materiall that would be expected expected to incur incur the type of  corrosion of immediate concern, for example, stress corrosion, cracking, pitting, crevice corrosion, and 11.1 1.1.3 .3 One or mor moree mate materia rials ls lik likely ely to be res resista istant nt to the environment.

agitatio agita tion n of th thee liq liqui uid d sh shou ould ld th then en be en enco coun unter tered ed by al alll specimens. 14. Dura Duration tion of Exposure Exposure 14.1 The duration duration of exposure may be based on known rates of de deter terio iora ratio tion n of th thee ma mater terial ialss in us use. e. Mo More re of ofte ten, n, it is governed by the convenience with which plant operations may be interrupted to introduce and remove test specimens. In many tests, some materials may show little or no attack while other materials may be completely destroyed. In general, the duration of the test should be as long as possible, commensurate with the resistance of the materials under test. In special cases, the dur duratio ation n may be est establi ablishe shed d in reg regard ard to som somee spe specific cific phase of the operation, for example, to study corrosion in one step st ep of a ba batc tch h pr proc oces ess. s. Po Poss ssib ible le ch chan ange gess in th thee ra rate te of  corrosion may be studied either by successive exposures or by the installation of several sets of specimens at the same time, which can be removed one set at a time at different intervals. The minimum duration of the test in hours is approximately 50,

12. Initial Specimen Specimen Measurements Measurements 12.1 Aft 12.1 After er the spe specime cimen n has been cut to size and the final surface finish applied (if other than mill finish), it should be cleaned in an organic solvent and the mass determined to the nearest 0.1 mg on an analytical balance. The total surface area is also determined to an accuracy of   61 %. These measurements are filed for later use in the corrosion rate calculations. 12.2 Durin During g fabrication, fabrication, each specime specimen n should be stamped with a code number for identification. identification. The record of the details of the test exposure (dimensions, weight, location, method of  moun mo unti ting ng,, lo loca cati tion on on ra rack ck,, et etc. c.)) sh shou ould ld be ke kept pt in a permanent, bound log book. Responsibility for properly main7

 

 

G4 − 01 (20 (2014) 14)

divided by the expected corrosion rate expressed in millimetres per year (or 2000 divided by the corrosion rate in mils per year). It is desirable to run the test with various time intervals so that the changes in corrosion rate with exposure time can be evaluated.

tion of the pits will be of much more importance. Sometimes pitting pit ting is ini initiat tiated ed but is sel self-h f-heali ealing ng and sto stops. ps. (Ad (Additi ditiona onall information is provided in Guide  Guide   G46 G46..)

15.1 The condition condition and appearance appearance of the holder and specimenss af men afte terr re remo mova vall fr from om eq equi uipm pmen entt sh shou ould ld be no noted ted an and d

17.5 17 .5 If an all alloy oy is kn know own n to be su susce scept ptib ible le to lo loca caliz lized ed corrosion on a microscale, such as the phenomenon of intergranular corrosion in stainless steel, dezincification in brass, or stress-corrosion cracking of any kind, the specimen should be bent after the previously outlined examination is completed, and any cracks that develop on the surface noted. Use caution

recorded. In removing the specimens the to holder, care to keep them in proper sequencefrom relative each exercise other so that any specimen may be identified from the original record of  its position on the holder. That is important if corrosion has been so severe that identification marks have been removed.

when ben when bendin ding g mat materia erials ls sus suscep ceptib tible le to hyd hydrog rogen en emb embrit rittletlement. The results should be compared with those obtained on similar bend tests on unexposed specimens from the same lot of material. Metallographic examination (Practice E3 (Practice  E3)) is also a useful means of characterizing these phenomena.

15.2 A re 15.2 reco cord rd sh shou ould ld be ma made de of th thee ap appe pear aran ance ce an and d adhe ad hesi sion on of an any y co coat atin ings gs or fil films ms on th thee su surf rfac acee of th thee specimens after washing. It may be desirable to photograph the specimens. Color photographs may be of value. Samples of any products or films resulting from corrosion may be preserved for future study.

17.6 The behavior behavior of the individual individual specimens in galvanic couples can be compared with that of corresponding insulated specimens exposed at the same time, and any galvanic effects can be observed. In a galvanic couple, the corrosion on one specimen will be accelerated while the other will be decelerated. As mentioned earlier, such tests are only qualitative, as the extent of the galvanic corrosion is influenced by the area ratio between the anodic and cathodic members of the galvanic couple, the relative potential difference between the dissimilar metals, and the solution conductivity. The results will apply directly only to assemblies in which the ratio of areas used in making the tests is similar to the ratio of areas anticipated in the fabricated assembly.

15. Remo Removal val of Spec Specimen imenss fro from m Test Test

16. Cleaning and Weighing Weighing of Test Test Specimens 16.1 Specim Specimens ens should be cleaned as soon as possible after removal from test. 16.2 The procedures procedures for cleanin cleaning g and weighi weighing ng specimens are described in Practice  Practice   G1 G1.. 17. Exam Examinati ination on of Spec Specimen imen Surface Surface 17.1 The spe 17.1 specime cimen n sho should uld be car carefu efully lly exa examin mined ed usi using ng low-power low-p ower magnification magnification as needed for type and uniformity of  surface attack such as etching, pitting, dealloying or parting, tarnish tarn ishing ing,, filmi filming, ng, sca scaling ling,, etc. If pit pitting ting is obs observ erved, ed, the numb nu mber er,, si size ze an and d di dist stri ribu butio tion, n, an and d th thee ge gene nera rall sh shap apee an and d uniformity of the pits should be noted (see Guide  Guide   G46 G46). ). The maximum and minimum depth of the pits can be measured with a calibrated microscope or by the use of the depth gage. Photographs of the cleaned specimens will serve as an excellent record of the surface appearance.

18. Localized Corrosion

17.2 17. 2 Det Detecti ection on of cer certain tain ef effect fects, s, suc such h aswill stress stre ss cor corros rosion ion cracking, dealloying, or intergranular attack, require lowpower pow er micr microsc oscopi opicc exa examin minatio ation. n. How Howeve ever, r, in som somee case cases, s, higher hig her res resolu olution tion and mag magnifi nificati cation on exa examin minatio ations ns may be necessary. This could include, but is not limited to, scanning electro elec tron n micr microsc oscopy opy or hig high-p h-powe owerr opt optical ical mic micros roscop copy y, or both, of metallographically prepared specimens. Mass loss is often used to evalua evaluate te inter intergranu granular lar corro corrosion sion (see Practic Practices es A262)). A262

may both propagation. process ment me ntimpact cont co ntai aini ning ng initiation crev cr evic ices es,, and such su ch as un unde derr In gask ga sket etss orequipscale sc ale deposits, depos its, variab variable le corro corrosion sion behavi behavior or may occur occur..

18.1 Met 18.1 Metals als oft often en per perfor form m dif differ ferentl ently y in aer aerated ated ver versus sus nonaerated environments, depending on how strongly oxygen reduction (cathodic depolarization) controls the cathodic reaction. The presence of other oxidizers, such as ferric or cupric ions, ion s, also can hav havee an ef effec fect. t. Oth Other er fac factor torss tha thatt can affect affect crevice corrosion behavior include, for example, crevice formerr ma me mate teri rial al an and d si size ze,, th thee re resu sult ltin ing g ga gap p pr prod oduc uced ed by tightening, and the area ratio of the shield to exposed surfaces. Some variables influence the initiation of attack while others

18.2 18 .2 Se Seve vera rall ty type pess of cr crev evic icee co corr rros osio ion n sp spac acer erss ca can n be substituted for the normal flat washer to study crevice corrosion in more detail (see Guide G78 Guide  G78)). Test specimens should be photograph photo graphed ed to docum document ent the locatio location n and overall af affected fected area ar ea of cr crev evice ice att attack ack.. Al Alth thou ough gh th thee pr pres esen ence ce of cr crev evic icee corrosion on test specimens is a positive indication, its absence does not guarantee the immunity of equipment to failure.

17.3 A distinction distinction should be made between localized corrocorrosion occurring under the insulating spacers and occurring on the boldly exposed surface. As previously noted, corrosion at or under the insulating spacers is an indication of susceptibility of the material to crevice corrosion (see Guide   G78 G78)) in the specific environment. environment. Pitting on the surfa surface ce is indica indicative tive of the

18.3 Pitting can occur on an unshielded unshielded metal surface and can lead to fai failur luree of equ equipm ipment ent displayin displaying g a low general general corrosion rate. Pitting can occur in passive type materials such as some grades of aluminum and stainless steel; it may also affect some copper base and nickel base alloys. The environmentt usu men usually ally contains contains an agg aggres ressiv sivee ion ion,, suc such h as chl chlori oride, de, which is made more aggressive if the condi conditions tions are oxidizing. Pitting can occur on usually nonpassive metals, such as steel, if, for example, a filming inhibitor breaks down locally.

pitting ten pitting tenden dency cy of the env enviro ironme nment nt on the bol boldly dly exp expose osed d surfaces of the specific alloy and specimens to be evaluated. 17.4 In the case of pitting of of the specimen, specimen, the mass loss is of little value and the study of the number, size, and distribu8

 

 

G4 − 01 (20 (2014) 14)

18.3.1 1 Pitting test data should should include a measurement measurement of the 18.3. maximu max imum m pi pitt de dept pth h du duri ring ng th thee te test st pe peri riod od an and d it sh shou ould ld encompass encomp ass a descrip description tion of the following characteristics characteristics of the pit: pi t: (1) sha shape—j pe—jagg agged, ed, circ circular ular,, elo elonga ngated, ted, (2) sec sectio tion— n— shallow, deep, rounded, conical, undercut, and (3) amount— superficial, super ficial, scatter scattered, ed, profu profuse, se, isolate isolated. d.

19.2 The form of cor 19.2 corros rosion ion that is doc docume umente nted d sho should uld be reported, together with any observations on corrosion products or sc scal ales es.. Th Thee ex exte tent nt of ea each ch co corr rros osiv ivee fo form rm sh shou ould ld be quantified as described in Section   18 18.. 19.2.1 19.2. 1 Penetr Penetration ation damage should should be expre expressed ssed in millimetress and corrosio tre corrosion n rate ratess in mill millimet imetres res per year (mm (mm/y) /y) for uniform unifo rm or gener general al corro corrosion sion (see Practic Practicee  G1  G1). ). An evalua evaluation tion based on mass loss is also sometimes used when corrosion has been substantially uniform in distribution over the surface of 

18.4 Guideli 18.4 Guideline ness fo forr ev evalu aluat atin ing g pi pitti tting ng ar aree co cont ntain ained ed in Guide   G46 Guide G46..  The statistical nature of pitting indicates that it is more likely to occur within large specimens and is dependent on the surface finish of the test specimen. Therefore, evaluation of pitting must use the largest practical size specimen and a standardized standa rdized surface finish and prepa preparation ration technique.

specimens; it isThe expressed as mass square meter per day (g/m2 /day). use of mass lossloss dataper t o estimate to corrosion pene pe netr trat atio ion n wi will ll be su subj bject ect to er erro rorr to th thee ex exten tentt to wh whic ich h nonuniform distribution of corrosion and changes of corrosion rate with time occur. 19.2.2 19.2. 2 The depth of pitting or crevice corrosion should should be reported to the nearest 0.01 mm (0.0005 in.) for the test period and not interpolated or extrapolated to thousandths of an inch per year or any arbitrary period. The size, shape, and distribution tio n of th thee pi pits ts sh shou ould ld be no noted ted.. Th Thee su surf rfac acee ar area ea of th thee specimen and the area of the crevices should be recorded if  crevic cre vicee cor corros rosion ion occ occurs urs.. The max maximu imum m dep depth th of cre crevice vice corros cor rosion ion tha thatt exi exists sts ben beneath eath the spe specime cimen n spa spacer cer mus mustt be reported.

18.5 Prope Properly rly conceived laboratory laboratory tests (see, for examp example, le, Practices   G30 G30,,   G36 G36,,   G37 G37,,   G41 G41,,   G44 G44,,   and Test Method   G47 G47,, etc.) are valuable tools for investigating factors affecting stress corrosion corro sion cracki cracking ng of engine engineering ering alloys. Howev However, er, in-plant corrosion tests for stress corrosion cracking susceptibility come closer to representing the environmental variables that could affe af fect ct all alloy oy be beha havi vior or in ser servi vice. ce. Th Thee lim limita itatio tions ns of bo both th laboratory tests and in-plant corrosion should be recognized. For exa exampl mple, e, simp simple le exp exposu osure re of str stresse essed d spe specime cimens ns in an operating flow stream may not take heat transfer, if present in service, into account.

20. Accu Accuracy racy of Results Results 20.1 The reproducibility reproducibility of plant corrosion tests is depen depen-dent on a number of factors, including the alloys tested, the variability of the environment, and the nature of the corrosion proces pro cess. s. Acc Accord ording ingly ly,, it is imp imposs ossibl iblee to pro provid videe a gen general eral stateme stat ement nt tha thatt will apply to all circ circums umstanc tances. es. In gen genera eral, l, howeve how everr, a var variati iation on of   620 % fr from om th thee me mean an wo woul uld d be consid con sidere ered d nor normal, mal, whi while le a var variati iation on of   650 % mi migh ghtt be expected in some circumstances.

19. Repor Reportt 19.1 In reporting reporting results of corrosion corrosion tests, the conditions conditions of  the test should be described in complete detail with special attention being given to the following: 19.1.1 19.1. 1 Corro Corrosive sive medium and concen concentration tration,, 19.1.2 19.1. 2 Type of equipment equipment in which test was made, 19.1.3 19.1. 3 Proce Process ss carried out in the operating operating equipment, 19.1 19 .1.4 .4 Locat Locatio ion n an and d co confi nfigu gura ratio tion n of sp spec ecime imens ns in th thee operating equipment, 19.1 19 .1.5 .5 Temp Temper eratu ature re of cor corro rosiv sivee med media ia (m (maxi aximu mum, m, minimum, or average), 19.1.6 19.1. 6 Oxidiz Oxidizing ing or reduc reducing ing nature of corro corrosive sive media, 19.1 19 .1.7 .7 Amoun Amountt an and d na natu ture re of ae aera ratio tion n an and d ag agita itatio tion n of  corrosive media, 19.1.8 19.1. 8 Durat Duration ion and type of test (if equipment equipment was operated intermittently intermi ttently during the tests, the actual hours of opera operation tion should be stated as well as the total time of the test), 19.1. 19 .1.9 9 Surface Surface con condit dition ion of spe specim cimen en (mi (mill ll fini finishe shed, d, polished, machined, pickled, 120 grit, etc.)

20.2 The ability of corrosion corrosion test specimens to simulate the performance of the materials of construction of a process plant is lar largel gely y dep depend endent ent on the design design of the program program and the unders und erstand tanding ing of the cor corros rosive ive pro process cess inv involv olved. ed. A well well-designed test program should give results that correlate to the existing materials of construction within the limits mentioned in 20.1 in  20.1.. 21. Keyw Keywords ords 21.1 corro corrosive sive test specime specimens; ns; forms of corro corrosion; sion; general corrosion corro sion rate; in-pla in-plant nt expos exposures; ures; localized attack; specific size and surface conditions; test duration; test racks

ASTM International International takes no positi position on respecting the validi validity ty of any patent rights assert asserted ed in connec connection tion with any item mentio mentioned  ned  in this standard. Users of this standard are expressly advised that determination of the validity of any such patent rights, and the risk  of infringement of such rights, are entirely their own responsibility. This standard is subject to revision at any time by the responsible technical committee and must be reviewed every five years and  if not revised, either reapproved or withdrawn. Your comments are invited either for revision of this standard or for additional standards  and should be addressed to ASTM International Headquarters. Your comments will receive careful consideration at a meeting of the  responsible respon sible technical technical committee, which you may attend. If you feel that your comments have not receiv received ed a fair hearing you should  make your views known to the ASTM Committee on Standards, at the address shown below. This standard is copyrighted by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States. Individual reprints (single or multiple copies) of this standard may be obtained by contacting ASTM at the above  address addr ess or at 610610-832832-9585 9585 (pho (phone), ne), 610610-832832-9555 9555 (fax (fax), ), or serv service@ ice@astm astm.org .org (e-m (e-mail) ail);; or thro through ugh the ASTM webs website  ite  (www.astm. (www .astm.org). org). Permission Permission rights to photocopy the standa standard rd may also be secure secured d from the Copyri Copyright ght Clearance Center, Center, 222  Rosewood Drive, Danvers, MA 01923, Tel: (978) 646-2600; http://www.copyright.com/ 

9