LOW-PRESSURE CARBURIZING PROCESS … Pressure Vacuum Carburi… · rial Specifications (AMS) pertaining to pyrometry (AMS 2750) and heat treatment of parts (AMS 2759) It was determined

With more aerospace customers looking for improved mechanicalproperties in bearing materials than offered instandard through hardened steel grades,New Hampshire Ball Bearings Inc. initiated aprogram to develop a low-pressure carburizingprocess that could providethe required properties innewer carburizablegrades including M50 NiL.

Steve Carey*New Hampshire Ball Bearings Inc.Peterborough, N.H.

Dan Herring*The Herring GroupElmhurst, Ill.

*Member of ASM International and member, ASM Heat Treating Society

iTech Div., New HampshireBall Bearings Inc. (NHBB),manufactures precisionbearings for use in aero-

space and industrial applications,including various ball bearing androller bearing configurations (Fig.1). Historically, a majority of the ma-terials used in aerospace bearing ap-plications have consisted ofthrough-hardened materials suchas AISI 52100 alloy steel and M50(intermediate high speed, molyb-denum type tool steel) alloy steels.In the past several years, bearingmanufacturers have seen an in-crease in the number of inquiriesfrom aerospace customers who arechallenged with bearing applica-tions that require superior mechan-ical properties compared with thoseprovided by standard through-hardened materials. To meet theseever-increasing requirements, car-burizable grades of steel such asM50 NiL (a nickel-low carbonvariant of M50 tool steel having thechemical composition 0.11-0.15% C,4.1% Cr, 3.4% Ni, 4.2% Mo, 1.2% V,balance Fe) are now being specifiedfor these applications. In doing so,wear resistance and fatigue strengthproperties (Table 1) are achieved onthe case-hardened surfaces compa-rable to those of through-hardenedmaterials, but with a higher level offracture toughness in the core of thepart. NHBB developed low-pres-sure carburizing technology tohandle these materials.

Furnace RequirementsThe first step in developing LPC

technology at NHBB was to specifythe requirements for the furnace de-sign and operation. It was deter-mined that one key feature of such asystem would be its flexibility to per-form both carburizing cycles andstandard vacuum heat treatment.This required a furnace that wouldproduce a minimal amount ofsooting to avoid contaminating sur-faces of non-carburized parts. Otherfeatures identified as being essentialincluded:

• Ability to perform both oil quench-ing and gas quenching

• Optimum case uniformity• Operator-friendly controls and

programming• Data acquisition capability• Must meet all Aerospace Mate-

HEAT TREATING PROGRESS • MAY/JUNE 2007 43

H

LOW-PRESSURE CARBURIZING PROCESS DEVELOPMENT OF M50 NiL

Table 1 — Typical Mechanical Property Data for Various Bearing SteelsTempering temperature, Fracture toughness Fatigue life

Alloy °F (°C) Hardness, HRC (KIc)*, ksi /in. (MPa • m0.5) (L10 dynamic life factor)

440C 350 (175) 58-62 19 (20.88) 0.8 52100 375 (190) 60-64 18-20 (19.78-21.97) 1-6 M50 1000 (540) 60-64 18-20 (19.78-21.97) 10 M50 NiL 975 (525) 47 max. 50-52 (54.94-57.14) 12-16 * ASTM E399.

Fig. 1 — Representative precision bearings.

rial Specifications (AMS) pertainingto pyrometry (AMS 2750) and heattreatment of parts (AMS 2759)It was determined that a LPC fur-

nace would be best-suited to meetthese requirements, and NHBB in-stalled in 2005 an Ipsen InternationalInc. (Cherry Valley, Ill.) AvaC (acety-lene vacuum carburizing) system(Fig. 2).

Process DevelopmentAfter an extensive literature review

on LPC and gathering informationon LPC from various other resources,NHBB created a cause and effect di-agram outlining the potential sourcesof inherent variability in the LPCprocess (Fig. 3). At the same time, car-burizing simulation programs weredeveloped to model the case profiles

of M50 NiL and other materials. Sev-eral iterations of the M50 NiL simu-lation programs were conductedbased on the results of the develop-ment cycles.

Initial carburizing cycles performedon the furnace were designed to verifythe furnace would operate properlyand produce the required carburizedsurface. These baseline cycles werecompleted using standard carburizingsteels (AISI 1018 and AISI 9310). Car-burizing parameters for these initialcycles were selected using valuescommon to industry practices forthese grades of steel.

While the required carburizing re-sults were achieved in the furnace,results of initial test cycles indicatedthe need for modifications to the pro-gramming format for improvedrecipe flexibility (i.e., allow for addi-tional boost/diffuse segments andprovide the ability to program inminutes/seconds instead ofhours/minutes) and modificationsto the furnace hardware, both ofwhich were handled on-site.

Following the initial tests, addi-tional baseline test cycles on the stan-dard carburizing steels were per-formed to verify the effects of severalkey process variables on resultingcase depth, microstructure, and near-surface carbon content. Variables in-cluded:

• Carburizing temperature• Boost/diffuse times and respec-

tive ratios

44 HEAT TREATING PROGRESS • MAY/JUNE 2007

Fig. 2 — Carburizing chamber of the low-pres-sure carburizing furnace.

Fig. 3 — Cause and effect diagram outlining potential sources of variation in thelow-pressure carburizing process.

Fig. 5 — Photomicrograph showing veryhigh level of retained austenite in the chamferarea of an inner roller-bearing ring; 3% nitaletch. 200×

Fig. 4 — Case hardness profiles of two M50 NiL LPC development cycles.

0.002 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 0.050 CoreDepth, in.

66

62

58

54

50

46

42

Har

dn

ess,

HR

CPart design Operator Method

FurnaceMeasurement Material

Surface finish

Geometry

Case depth req.

Non-carburized surfaces

Microindentation hardness

Retained austenite

Surface carbon

Residual stress

Microstructure

Temperature uniformity

Gas dispersion systemDesign

Quenching

Vacuum leak rate

Experience

Training

Controls

Programming

PM load transfer time

Type

Part cleanliness

Prior microstructure

Robust carburizing

process (Y)

Carburizing temperature

Boost time

Gas type

Gas flow rate

Partial pressure

Masking

No. of boost/diffuse segments and ratios

Load surface area

Quench method

Hardening/tempering

Load configuration and fixturing

• Number of boost/diffuse seg-ments

• Relationship between boost/dif-fuse cycles and surface carbon

• Partial pressures during boostand diffuse segments

• Relationship between part sur-face area and gas flow rate

• Carbon flux as a function of tem-perature

In all instances, resulting casedepths, microstructures, and surfacecarbon measurements were consis-tent with expected levels.

M50 NiL Carburizing DevelopmentNHBB used the results of the ini-

tial baseline tests to develop a robustcarburizing procedure for M50 NiL(AMS 6278). M50 NiL is a carburiz-able grade of steel used in the aero-space industry in critical applicationsrequiring higher core fracture tough-ness properties compared with thethrough-hardened (non-carburiz-able) M50 grade of steel. The rela-tively high level of chromium in M50NiL (4.0-4.25%) presents a challengewhen developing a carburizingprocess due to an increase in the ef-fective carbon absorption of the ma-terial [1].

The first few M50 NiL test cycleswere modeled to target an effectivecase depth of 0.030 to 0.035 in. (0.762-0.889 mm). Effective case depth is de-fined as the perpendicular distancefrom the case hardened surface to adepth equal to a hardness of 58 HRC.Initial carburizing temperatures inthe range of 1600 to 1900°F (870-1040°C) were selected based on tem-peratures commonly used in the in-dustry for LPC. In all cases, the partswere heated to the carburizing tem-perature, exposed to multipleboost/diffuse segments, cooled tobelow 200°F (95°C) in a protective at-mosphere, then austenitized, oilquenched, and processed throughmultiple subzero/tempering cycles.Several short boost cycles were com-bined with progressively longer dif-fuse times to overcome the effectivecarbon absorption of M50 NiL mate-rial. For example, representative casehardness profiles (Fig. 4) from two ofthe development cycles where allprocessing parameters were identicalwith the exception of the number of

boost/diffuse segments revealed thata much deeper effective case was pro-duced with five additional boost/dif-fuse segments. While the case hard-ness profiles were promising basedon targeted depths, metallurgicalevaluation of the resulting case mi-crostructures revealed high levels ofretained austenite, particularly in thecorner and chamfer areas of an innerroller-bearing ring (Fig. 5).

Surface-carbon results from testspecimens included in the loadsshowed carbon levels in excess of1.1%. It was necessary to lower thenear-surface carbon concentrationsto the 0.70% range to optimize themicrostructure of the M50 NiL testpieces. “Leaning out” the surfacecarbon level was accomplished bylowering the carburizing tempera-ture, shortening boost times, ex-tending diffuse segments, and in-creasing the number of boost/diffusesegments. Resulting microstructuresshowed no evidence of retainedaustenite or excessive carbide forma-tion in the corner and chamfer re-gions. However, the effective casedepths of the test cycles were lessthan 0.020 in. (0.508 mm).

A carburizing process was devel-oped to maintain the superior casemicrostructure of the previous cyclewhile extending the effective casedepth to the required 0.030-0.035 in.range. The resulting case hardnessprofile (Fig. 6) for the modified cyclerevealed the effective case depth tobe approximately 0.035 in. Also, thecorresponding case microstructurewas uniform in depth around the en-tire cross section, consisting of tem-

pered martensite with no evidenceof retained austenite or excessive car-bide formation (Fig. 7). Retainedaustenite content at a depth of 0.002in. (0.051 mm) from the case hard-ened surface was measured using anx-ray diffraction method in accor-dance with ASTM E975 and was lessthan 2%. Near surface-carbon con-tent was 0.74%.

Process Validation and Optimization

Validating process results wasaccomplished by repeating the cycleand examining load-to-load variation(Fig. 8), within-part variation (Fig. 9),and part-to-part variation within load (Fig. 10). The case-hardeningprofiles show that the LPC cycle de-veloped using this analytical ap-proach produces excellent case uni-formity and reproducibility. In eachinstance, the corresponding micro-structures, retained austenite levels,and near surface-carbon contents con-formed to specified levels. Continued

HEAT TREATING PROGRESS • MAY/JUNE 2007 45

Fig. 6 — Case hardness profile of final M50 NiL LPC production cycle.

Fig. 7 — Photomicrograph showing a tem-pered martensitic microstructure in thechamfer area of an inner roller-bearing ring.There is no evidence of retained austenite orexcessive carbide formation; 3% nital etch.200×

0.002 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 CoreDepth, in.

66

62

58

54

50

46

42

Har

dn

ess,

HR

C

Using standard industry tempera-ture ranges, an optimum combina-tion of austenitizing temperature andtempering temperature was deter-mined based on results from a fullfactorial design of experiment. Vari-ation in temperature for each of theseoperations has a significant impacton the resulting case hardness pro-file, particularly for the temperingoperation (Fig. 11).

Qualification of Masking ProcessSeveral test parts were copper

plated in accordance with AMS 2418to qualify a masking process for M50NiL and other carburizing grades ofsteel. In all instances, the platingprocess was an effective way ofmasking surfaces to prevent casehardening at those locations.

SummaryNHBB developed an LPC process

for M50 NiL material, providing agood knowledge base to enhanceand apply the process to new ma-terials and part configurations. Cur-rently, effective case depths havebeen extended to 0.045 in. (1.143mm). Future development work willbe aimed at extending the process toachieve effective case depths in ex-cess of 0.060 in. (1.524 mm) for M50NiL and on the order of 0.100 in. (2.54mm) or greater for materials such as9310. Trials on Pyrowear 675 stain-less steel (Carpenter TechnologyCorp., Reading, Pa.; www.cartech.com) are also planned.

References1. D.H. Herring, Study of Vacuum Car-burizing Process Parameters, Ind. Heating,p2, Sept. 1996.For more information: Steve Carey isManager of Materials Engineering, NewHampshire Ball Bearings Inc., HiTechDiv., 175 Jaffrey Rd., Peterborough, NH03458; tel: 603-924-3311, ext. 5522; fax:603-924-9306; e-mail: [email protected];Internet: www.nhbb.com; Daniel Her-ring, The Heat Treat Doctor, is president,The Herring Group Inc., PO Box 884,Elmhurst, IL 60126; tel: 630-834-3017; fax: 630-834-3117; e-mail: [email protected]; Internet: www.heat-treat-doctor.com.

46 HEAT TREATING PROGRESS • MAY/JUNE 2007

Fig. 8 — Case hardness profiles measuredon the roller paths of M50 NiL samplesprocessed in three identical test cycles showload-to-load variation.

Fig. 9 — Case hardness profiles measured on multiple surfaces of an M50 NiL sample showwithin-part variation.

Fig. 10 — Case hardness profiles measured on multiple parts from one M50 NiL LPC pro-duction cycle show part-to-part variation within the load.

Fig. 11 — Effect of tempering temperature on the case hardness profile of an M50 NiL casehardened component.

0.002 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 CoreDepth, in.

0.002 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 CoreDepth, in.

0.002 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 CoreDepth, in.

0.002 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 CoreDepth, in.

975°F

1000°F

1025°F

66

62

58

54

50

46

42

66

62

58

54

50

46

42

66

62

58

54

50

46

42

66

62

58

54

50

46

42

Har

dn

ess,

HR

CH

ard

nes

s, H

RC

Har

dn

ess,

HR

CH

ard

nes

s, H

RC