UNDERSTANDING and MEASURING DECARBURIZATION Understanding the Forces Behind Decarburization Is the First Step Toward Minimizing Its Detrimental Effects

Home , Decarburization

22 UNDERSTANDING AND MEASURING DECARBURIZATION Understanding the forces behind decarburization is the first step toward minimizing its detrimental effects. George F. Vander Voort, FASM*, Struers Inc. (Consultant), Cleveland

ecarburization is detrimental to crostructure of carbon contents close to the maximum depth of decarburization the wear life and fatigue life of the core may be difficult to discern. The is established. Because the carbon dif- steel heat-treated components. MAD determined by hardness traverse fusion rate increases with temperature ADVANCED MATERIALS & PROCESSES | FEBRUARY 2015 2015 FEBRUARY | & PROCESSES MATERIALS ADVANCED D This article explores some factors that may be slightly shallower than that de- when the structure is fully austenitic, cause decarburization while concentrat- termined by quantitative carbon analysis MAD also increases as temperature rises

ing on its measurement. In most produc- with the electron microprobe. This is es- above the Ac3. For temperatures in the

tion tests, light microscopes are used to pecially true when the bulk carbon con- two-phase region, between the Ac1 and

scan the surface of a polished and etched tent exceeds about 0.45 wt%, as the rela- Ac3, the process is more complex. Carbon cross-section to find what appears to be tionship between carbon in the austenite diffusion rates in ferrite and austenite the greatest depth of total carbon loss before quenching to form martensite and are different, and are influenced by both (free-ferrite depth, or FFD) and the great- the as-quenched hardness loses its linear temperature and composition. est depth of combined FFD and partial nature above this carbon level. Decarburization is a serious prob- loss of carbon to determine the maxi- lem because surface properties are infe- mum affected depth (MAD). rior to core properties, resulting in poor In some cases, there is no free Decarburization basics wear resistance and low fatigue life. To ferrite at the surface. In research stud- Decarburization occurs when car- understand the extent of the problem, ies, this may be supplemented with a bon atoms at the steel surface interact two characteristics that may be present Knoop hardness traverse to determine with the furnace atmosphere and are at a decarburized steel’s surface can be the depth where hardness becomes con- removed from the steel as a gaseous measured: Free-ferrite layer depth (FFD, stant. The Knoop-determined MAD is of- phase[1-8]. Carbon from the interior diffus- when present) and partial decarburiza- ten somewhat deeper than the visually es towards the surface, moving from high tion depth (PDD, when free-ferrite is determined MAD, as variations in the mi- to low concentration and continues until

Fig. 1 — Decarburized surface of as-rolled, eutectoid carbon steel (Fe-0.8% C-0.21% Mn-0.22% Si) at two different locations around the periphery show a substantial variation in the amount Fig. 2 — Erratic depth of free ferrite at and depth of ferrite at the surface. The matrix should be nearly all pearlite (4% picral etch, 500×). the surface of a bar of 440A martensitic stainless steel after quenching from 2000°F *Life Member of ASM International (1093°C); etched with Vilella’s reagent.

0215 AMP features.indd 22 2/10/2015 4:24:25 PM 23 ADVANCED MATERIALS & PROCESSES | FEBRUARY 2015 not present). If free ferrite is present, the free-ferrite layer’s maximum depth (often variable) plus the depth of partial decarburization to the unaffected core is measured. This total—FFD + PDD—is called maximum affected depth (MAD). These depths are not uniform and can vary substantially, leading to measurements of average FFD, PDD, and MAD, as well as maximum values for each. ASTM E1077 covers decarburization measurement. In practice, decarburization should be evaluated on a plane transverse to the Fig. 3 — Decarburization of 5160 modified spring steel defined by surface hardness and hot working axis, as depth variation is incremental turnings analyzed chemically for carbon content as a function of whether or not the greater around the bar on the transverse surface was descaled or was covered by mill scale, and austenitizing at 1600°F for 80 minutes. plane than at a specific constant position along a longitudinal plane. Decarburiza- tion depth can vary substantially around than planar surfaces. Sampling schemes with precision and depth measurement the periphery of a bar, as shown in Fig. 1. for large cross-sections are also illustrat- accuracy suffers. Good edge retention Qualitative measurements can be sub- ed in ASTM E1077. requires mounting of the specimen in a jective and biased. Free-ferrite depth To obtain good data, edge reten- resin, such as DuroFast, that does not ex- can also be erratic, even over a small sur- tion must be perfect—the surface must hibit shrinkage gaps between the mount face area, as shown in Fig. 2. Corners of be perfectly flat to the extreme edge. If and specimen after polymerization. square or rectangular sections normally edges are rounded, the exact location Grinding and polishing procedures must exhibit greater decarburization depths of the outer surface is difficult to define emphasize maintaining flatness. While

0215 AMP features.indd 23 2/10/2015 4:24:25 PM 24

SiC abrasive paper can be used for grinding, resin-bonded diamond discs such as MD-Piano provide excellent flatness and long life. Napless cloths are used for diamond polishing while low-nap clothes are used to polish with alumina or colloi- dal silica abrasives. In most cases, a ret- icule is used to make the measurement. Alternatively, many image-capture soft- ware programs allow operators to make point-to-point distance measurements. However, these systems must be properly calibrated. Hot working temperatures can produce ferrite at the surface with the amount and size of the ferrite grains growing as carbon loss increases at the surface. The upper critical tempera-

ture, Ac3, of these grains could be above 1600°F, as alloy composition and residu-

al elements influence the Ac3 of the steel ADVANCED MATERIALS & PROCESSES | FEBRUARY 2015 2015 FEBRUARY | & PROCESSES MATERIALS ADVANCED grade. Spring steels are used as an ex- ample. If a decarburized specimen is in- duction hardened, the heating rate to the austenitization temperature is extremely fast. To put all of the carbon in solution (assuming that the steel has a carbon content of 0.60-0.65%), it is heated to Fig. 4 — Frequency histograms of decarburization measurements made around the periphery roughly 1700°-1750°F. As the holding of 5160 modified bars after heat treatment. a) 1600°F specimen with a mill scaled surface has an time is short, perhaps no more than 10 s, average FFD of 0.08 mm and average MAD of 0.266 mm for 132 measurements. Note the narrow, there is little time for appreciable carbon peaked distribution of the FFD measurements and broad distribution of MAD measurements. diffusion and the decarburization depth b) 1600°F specimen with a descaled surface does not exhibit any free ferrite. No decarburization after heat treatment is a function of the was observed for almost 19.5% of the measurements and distribution of MAD values is broad. as-rolled mill decarburization depth.

Fig. 5 — Decarburized surfaces of 5160 Mod austenitized at 1600°F for 80 min. and oil quenched. Free ferrite on the scale covered specimen, top. No free Fig. 6 — Rockwell C hardness tests on the surface of 5160H, 5160M, and 9260M (after glass bead ferrite present on the specimen that was blasting) is a simple screening test to determine if the surface is decarburized or free of descaled before being austenitized (2% decarburization. The correlation is much better when there are no free-ferrite grains at the surface nital, 200×), bottom. (blue data points) than when free ferrite (red data points) is present.

0215 AMP features.indd 24 2/10/2015 4:24:26 PM 25 ADVANCED MATERIALS & PROCESSES | FEBRUARY 2015 ly produce more conservative estimates about 50-70% of the MAD determined by Experimental data than the incremental carbon analysis incremental carbon analysis or microin- If spring steels (typically ~0.6% procedure or microindentation hard- dentation tests. This depth, however, carbon) are heat treated in gas-fired ness traverses. This is because it is diffi- can be considered an effective depth furnaces, operating conditions can cult to detect the final minor loss in car- where carbon content is usually with- either increase or decrease the as- bon as the unaffected core is reached. in about 10-25% of the matrix carbon rolled depth of decarburization after Color etchants are likely to perform bet- content and responds reasonably well heat treatment, relative to the starting ter for this purpose than black and white to heat treatment. If the maximum ob- point. Austenitizing of these grades is etchants such as nital or picral, but com- served MAD is used as criteria for stock typically performed in the 1600°-1650°F parative tests have not been performed. removal, the surface’s carbon content range and holding times, which depend For annealed microstructures, the visual will be close to the matrix carbon con- upon bar diameter, are usually at least estimate of the average MAD is generally tent after machining. 20 minutes. In many cases, a protective atmosphere is not employed. An experiment was conducted using round bars of 5160 modified spring steel. Specimens were austenitized either with the as-rolled mill scale present or removed by sand blasting. Specimens were austenitized at 1600°F (871°C) for 80 minutes, then oil quenched. Part of each bar was incrementally machined (after scale removal by glass-bead blasting) and the carbon content was determined. Surface hardness readings were also recorded and results are shown in Fig. 3. Note that the specimen austenitized at 1600°F exhibits a large difference between surface carbon content and surface hardness, compared to the bar covered with mill scale to the descaled one. Figure 4 shows results of quantitative FFD and MAD measurements for the two specimens austenitized at 1600°F— including 132 measurements around the periphery of the scaled bar and 113 measurements for the descaled sample. The scaled bar austenitized at 1600°F exhibits a consistent free-ferrite layer around its periphery with an average depth of 0.08 ± 0.002 mm (95% confidence interval). Note that FFD measurement distribution is very narrow, or peaked. The MAD, however, shows an average depth of 0.266 ± 0.006 mm and distribution is broad. In contrast, for the descaled bar, no free ferrite was seen and 19.47% of the 113 measurements indicate no decarburization was present. The remaining measurements exhibit an average depth of 0.073 ± 0.010 mm, slightly lower than the scaled bar’s average FFD. The MAD distribution curve appears to be bimodal. Figure 5 shows typical microstructures observed at the specimens’ two surfaces. Visual estimates of the maximum affected depth of decarburization general-

0215 AMP features.indd 25 2/10/2015 4:24:26 PM 26 0215 AMP features.indd 26 ADVANCED MATERIALS & PROCESSES | FEBRUARY 2015 ted orspringlife would bereduced. and almost no MAD could be permit throughout the1970s,no free ferrite sign loads on thesespringsincreased drive automobile springs.Asthede spring steel lotsusedfor front wheel specimens of5160H,5160M,and9260M test was usedto detect decarburized Additionally LOM estimates yieldMADsof0.406and 0.433 mm,respectively. d)LOM image c) Qualitative andquantitative visual b) A200gfKnoophardness traverse turnings reveal aMADof0.64mm. a) Carbonanalysisofincremental annealed bar ofW1carbon steel. measurements onaspheroidize reveals a MAD of 0.51 mm. reveals aMADof0.51mm. , asimplescreening Fig. 7—Decarburization of atypical surface area. - - (d) surfaces atanumber of locations and Rockwell C tests were made on the OD bead blasting after hardening andbulk surface scan was removed by glass- for anumberofspecimens where the and burnishing.Figure 6shows data in thefinalprocessing step ofturning MAD to lessthantheamountremoved Mill processing helpedminimize the ized core—is harder andlessductile. lower volume fraction thanthespheroid- lamellar carbide structure—even witha steel metallurgists are aware, coarse counterintuitive. However, assome tool than inthecore, aresult thatmayseem the decarburized zone isactually greater Note thatthehardness atthesurface of compared to thebulkcarbon content. mentite inthelower carbon surface area annealing cannot cycle spheroidize ce tends to belamellar. Thisisbecause the rized zone is not well spheroidized but shows thatthecementite inthedecarbu- rite Examination isnotpresent. at1000× 0.7%, roughly a30%loss.So, free fer lowest carbon content isonlyto about tribution atthesurface inFig.7a.The Note theseeminglyunusualcarbon dis surface, individualcarbides can beseen. fraction thantheinterior. At theextreme exhibits asignificantly lower volume carbide inthedecarburized surface zone steel, is shown in Fig. 7d at 100×.The metallurgists inplantsthatmake tool C),atypical specimenrated by mill (~1% crostructure ofW1carbon tool steel in Fig.7.Thespheroidize annealed mi- timates by lightmicroscopy—are shown and visualqualitative orquantitative es microindentation hardness traverses, carbon analysis ofincremental turnings, burization ratings by three methods— ent versus whenitwas present. and MADwhenfree ferrite was notpres much better correlation between HRC of decarburization. Theplotshows a present) andmaximumaffected depth for maximumfree-ferrite depth(when tallographically prepared, andrated averaged. Bars were sectioned,me

periphery, are slightlylower thanthe visual estimate going around thebar’s qualitative estimates, based onasimple by theincremental carbon analysis. The stant atashallower depththan shown the hardness became essentiallycon- ever, thisisnotamajorproblem because MAD from actual carbon analysis.How- still rather conservative compared to the traverses thanLOM measurements, butis mate ismore accurate usingtheKnoop maximum affected depth.TheMADesti- turnings provides thebestestimate ofthe Ex Carbon analysisoftheincr amples ofthevariation indecar emental 2/10/2015 4:24:27 PM ------27 ADVANCED MATERIALS & PROCESSES | FEBRUARY 2015 - - 2/12/2015 10:32:14 AM e to ctors

uction. ings with tru r rod b y that will s ins y applicabl tails. se p is completed al.org s time clas aining atel te de our facilit SS! y

e course Glass whenGlass o t ation and increa and your staff’ e compl immedi r s -Site Tr On-Site ter th nt. Vol 151, p 31-39 and 49-55, 1943. 151, p 31-39 and Vol rmance riodic Beer ve use of sme ness af on that i ude: incl - (Lon Engineering Gases, in Decarburizing 229-231, p 92-94, 123-126, 133, Vol don), 1932. and 305-306, and De Carburization Steel Stanley, 2. J.K. Analysis, Iron – A Theoretical carburization Age, - of Decarburiza Prevention Purkert, 3. F.E. Steel, of High Carbon tion in Annealing 1982. 2, p 225-231, Vol Heat Treating, J. of Different al., Effect et Grasshoff, 4. H.W. Atmo Treating of the Heat Dew Points of on the Skin Decarburization sphere ünd Eisen, Stahl Steels, Heat-Treatable p 119-128, 1969. 89(3), Vol of Rudzki, Effects 5. G.E. Wieland and E.M. Parameters Design and Operating Furnace Metal of Steel, on the Decarburization p 40-46, February 1979. Progress, Forge: in the Drop Heating 6. R. Rolls, on of Scales and Properties Formation 34, Vol Forming, Metal Alloys, Iron-Base p 69-74, 1967. Oxida - Surface Tuck, 7. K. Sachs and C.W. ISI SR in Industrial Furnaces, tion of Steel p 1-17, 1968. 111, London, al., Decarburization 8. E. Schuermann, et 7, Vol Journal, Wire Formation, and Scale p 155-164, 1974. ustomized ati o perf s Ravi Dodeja fo C oducti ravi.dodeja@asminternation our facility orm a FREE Pe ask for your no-oblig tact onvenient access to world- Programs customized to ﬁt your needs your busi at y Inf c A pr zed courses ersonali levate - - - - • • Feature • e p you you g asses trainin Con Phone: 440.338.5414 Email: ASM’s FREE PERIODIC GLA BEER Get COMES TO YOU CUSTOMIZED TRAINING ~AM&P George F. Vander Vander F. George information: more For Inc., Struers for is a consultant Voort Cleveland, OH 44145, Rd., 24766 Detroit 847.623.7648, georgevandervoort@yahoo. www.georgevandervoort.com. com, References of Allen, The Loss and K.F. 1. A. Bramley When Heated and Steel Iron Carbon from prepared specimens with good edge edge good with specimens prepared achieved be easily can This retention. - is reason equipment and with modern of measurements Qualitative ably fast. present) depth (when the free-ferrite depth of de affected and the maximum But usually adequate. are carburization bias to subject are such measurements as is not as good and the reproducibility are measurements when quantitative select 25 randomly made using at least periphery. bar the around locations ed traverses hardness Microindentation MAD. The defining the for excellent are observed micros light FFD is easily by the inspection of and adequate copy the deep periphery detect is needed to est amount present. - - - - - T ASMINTERNATIONAL.ORG/LEARNING/COST A OW N lusions REGISTER c Dec Con surface weaker the as problem serious enabling resistance, wear reduces layer isa parts ofsteel arburization A easily. more occur to failures fatigue discussed, was test simple screening shapes certain for be used which can runs. If the sur and high production some predeter is below hardness face grade, with limit,mined varies which is examination then a microstructural analysis of carbon Chemical required. (or millings) turnings on incremental is this although performed, be can than pro research to applicable more de of rating duction. Metallographic properly depth requires carburization quantitative average, which was based was which average, quantitative the around measurements random on 25 the vi- assumed that it was If periphery. MAD around of the greatest sual estimate peripherythe bar deeper than be would chosen lo MAD of 25 randomly the mean be would result actual the then cations, surprising. rather 0215 AMP features.indd 27