rapid tempering

7/30/2019 Rapid Tempering

1/6

HEATTREATING

PROGRESSTHE BUSINESS AND TECHNOLOGY OF HEAT TREATING

An ASM International

Publication

TM

SM

www.asminternational.org MARCH/APRIL 2006

RAPID TEMPERING

OF AUTOMOTIVEAXLE SHAFTS

RAPID TEMPERING

OF AUTOMOTIVE

AXLE SHAFTS

Exclusive HEAT TREATING PROGRESSReprint

By: Thomas Neumann and Kenneth Pickett

Compliments of:

1081 Chemin Industriel St-Nicolas, Que,

bec, G7A 1B3 Canada

Phone: (418) 831-2576 Fax: (418) 831-3206www.pyromaitre.com


2/6

Rapid tempering of AISI1050M steel using anin-line tempering furnacewith an ultrahigh rate ofair convection movementreduced tempering cycletime by 80% with nodiscernable degradationof mechanical and fatigueproperties.

Thomas Neumann**Ford Motor Co.Sterling Heights, Mich.

Kenneth PickettSchoolcraft CollegeLivonia, Mich.

**Member ASM International and member,

ASM Heat Treating Society

Automotive axle shaftforging/extrusions aremachined and formed tonear net shape and then in-duction hardened and

quenched in water. Hardened shaftsgenerally are tempered in a furnaceat a temperature of 280 to 325F (140

to 165C) for a minimum of 75 min-utes to achieve a post tempered hard-ness of 60 HRC minimum at the

bearing journal surface.Tempering is carried out in a two-

zone gas-fired furnace having a size-able foot print, which was designed

to accommodate many differentshaft designs. Shafts are manu-ally loaded and unloaded ontoconveyers that hold the shaft ina vertical position as it travels

through the furnace. Total cycle time,including load, tempering, cooldown and unload is between 3 and4 hours.

The introduction of a new axleshaft design created a need to pur-chase new manufacturing equip-ment/processes that could support

lean manufacturing concepts. Leanmanufacturing is a production phi-losophy that emphasizes minimiza-tion of the amount of all the re-sources, including time, used tomanufacture a component or as-sembly. The existing temperingequipment/method was not compat-ible with lean manufacturing of axleshafts.

Rapid Tempering: An AlternativeAlternate furnace technology for

tempering interconnecting shafts andconstant velocity (CV) joints was

being used at a sister division, whichafforded an opportunity to performtempering tests on axle shafts usingits furnace technology, called forcedair convection tempering (FACT). Ituses an efficient furnace design con-

sisting of multiple zones withinthe furnace to allow for rapid,uniform component heatingand soaking, which results inrapid heat transfer to the com-

ponents. The design provides highrates of thermal radiation, convec-tion, and conduction.

Implementing the rapid tempering

method required a new mindset indefining adequate tempering. His-torically, time at a particular ambientfurnace temperature was considered

RAPIDTEMPERINGOF AUTOMOTIVE AXLE SHAFTS

Fig. 1 Thermocouples located at the spline end (a) and bearing region (b) of the axle shaftand overview of instrumented shaft with data recorder attached (c)

(a)

(c)(b)


3/6

critical regardless of actual furnaceloading or heat transfer rate. By com-parison, tempering criteria usingFACT methodology was redefinedas the uniform achievement of thedesired tempering temperature in thecase hardened component whilemaintaining the minimum surfacehardness requirements in the bearing

journal.

Validating FACTTempering tests were conducted

using the FACT tempering furnaceto validate its capability to processaxle shafts, and based on the positivetest results, a specialized horizontalFACT furnace was designed, pur-chased, installed and commissionedin 2003. Furnace performance wasvery good, and resulted in additional

testing and the purchase of a higherthroughput vertical FACT furnace in2005 for other established axle shaftproduction lines.

Thermal Profilesand Furnace Uniformity

Thermocouples were attached torepresentative axle shafts at criticalregions to accurately measure theheat transfer rates of conventionaltempering and FACT methods. Be-cause temperature uniformity of the

core and surface of the shaft were ofprimary concern, thermocoupleswere located both at the surface andcore (midpoint) of bearing journal(largest cross sectional area) and thespline (smallest cross sectional area).The thermocouples were connectedto a Datapaq (Datapaq Inc., Wilm-ington, Mass.;www.datapaq.com)recorder to measure temperatureversus time. Figure 1 shows an in-strumented shaft at different loca-tions.

An instrumented axle shaft wassent through the horizontal FACTfurnace six furnace times togetherwith scrap axle shafts to represent afully loaded production furnace, andtemperature was recorded at specificfurnace locations and times. Meantemperature readings 1 standarddeviation are shown in Fig. 2. Tem-perature consistency at the conver-gence temperature (300F, or 150C)region of the furnace is very good.

The rate of temperature rise and

convergence in the representativeaxle shafts was also of interest. Theinitial goal was to rapidly heat theshaft so both the core and surface ofthe bearing and spline reached thesame convergence temperature.Figure 3 is a thermal profile by loca-tion on the axle shaft versus time for

the vertical FACT furnace. The bluetrace represents the ambient furnacetemperature and indicates the twospecific heating zones. The black andyellow traces represent the temper-ature of the spline at the surface andcore, respectively, and the red andgreen traces represent the tempera-ture at the bearing surface and core,respectively.

The instrumented axle shaft wasalso processed through the conven-

tional temper furnace for comparison

Automotive Axle Shaft Anatomy

Automotive axle shafts transmit torque from the drive train to the rearwheels, support the vehicle load, and function as the inner race for thewheel bearings. Four key design features of an axle shaft are:

Flange, wheel, and brake pilot on one end, which are used as attach-ment points for the wheel and brake rotor or drum.

The bearing (the largest cross sectional area of the shaft) adjacent tothe flange, which is highly machined and polished to tight tolerances be-cause it serves as the inner race for the wheel bearing

The extruded shaft barrel (beyond the bearing journal) Aformed or machined spline and retaining button at the end of the

shaft barrel opposite the flange.In an axle assembly, the spline is mated with a differential side gear, and

is the portion of the shaft used to transmit torque through the shaft to thewheel end. The button end of the shaft is used as a retention mechanismand keeps the shaft in place within the axle assembly.

Axle shafts used in this study were made of hot-rolled AISI1050M steelbars having a manganese content of 0.8-1.1%. The hot-rolled bar is upsetforged to form the flange, wheel and brake pilot details and the barrel isoften cold extruded to taper the body of the shaft to the desired diame-ters. The axle shaft forging/extrusions are machined and formed to a nearnet shape condition and then induction hardened and quenched in waterto 40 HRC with and effective case depth (ECD) of 0.120 to 0.250 in. (3 to6.3 mm) in the barrel of the shaft. After hardening, the shafts typically aretempered at an ambient furnace temperature of 280 to 325F (140 to 165C)for a minimum of 75 minutes. Apost tempered hardness of 60 HRC min-imum must be maintained at the bearing journal surface. After temperingthe shafts go through a series of finish machining, polishing and assemblyoperations.

Fig. 2 Temperature uniformity and re-peatability for horizontal FACT furnace. Solidblack = mean temperature +1 sigma, hollowsquare = mean temperature 1 sigma

400 800 1200 1300 1900 2100Time, s

350

300

250

200

150

100

50Temperature,

oF


4/6

purposes. Table 1 lists key heattransfer rates. The heating rate in theFACT furnace is six times faster andthe time at tempering temperature isreduced six fold compared with theconventional furnace. The spline and

bearing did not achieve the sameconvergence temperature in the ver-tical FACT furnace. In this applica-tion, the spline temperature was 10to 15F (6 to 8C) higher than the

bearing temperature. The differenceis due to the considerable differencein cross sectional area/mass. How-ever, this was considered a benefit in

that the spline portion of the shaftwas subject to greater tempering re-sulting in increased toughness, whilethe bearing portion still maintainedthe required post-temper hardness.

The FACT furnaces used in thesetests and subsequent productionwere designed to facilitate both ro-

botic loading and unloading of theaxle shafts. The total tempering(cycle) time from load to unload was31 and 36 minutes for the horizontaland vertical furnaces, respectively.

This resulted in an eight fold reduc-

tion in tempering (cycle) time com-pared with that of conventional tem-pering systems. Both FACT furnacedesigns used a cool down zone al-lowing handling parts by hand aftertempering. The thermal profiles inFigs. 2 and 3 show the cool downzones in the horizontal and verticalFACT furnaces.

Test ResultsFor tempering validation testing,

axle shafts were forged from a singleheat of steel, green machined and in-duction hardened using the same in-

duction hardening system, and ran-domly separated into two groups fortempering. One (baseline) group wastempered conventionally and theother group was tempered using apreviously optimized FACT process.After tempering, shafts from eachgroup were color coded, then ran-domly reintroduced into the manu-facturing process for the required fin-ishing operations.

Mechanical properties, includingsurface hardness, microindentation

hardness, residual stress profiling,

torsional JAELand rupture strength,rolling bending, and torsional fatiguelife, of all tempered shafts were meas-ured to determine if axle shafts ofcomparable conventionally tem-pered quality could be made usingthe FACT tempering process.

HardnessSurface hardness of several shafts

from multiple treatment groups wasmeasured. Figure 4 shows the hard-ness populations in Box-Whiskerplots. Each rectangular box repre-sents the 25% and 75% quartiles of

the population, and the horizontalline bisecting the box is the median.The whiskers emanating from a boxrepresent the 0% and 100% quartiles,and the asterisks indicate values thatexceed 3 standard deviations. FromFigure 4, the as-quenched hardnessof the axle shafts is 64-65 HRC. Aftertempering, all treatments had com-parable hardness in the range of 61-63 HRC.

Microindentation hardness pro-files in the spline cross section of axle

shafts tempered using the various

Fig. 3 Thermal profile by processing time and location on axle shaft for vertical FACT furnace

Table 1 Thermal profile of conventional tempering and vertical FACT furnaces

Location onRamp rate (F/min) Time (min) at temp. Peak convergence

instrumentedfrom 100-280F 280-325F temp.,F

axle shaft FACT (new) Conventional FACT (new) Conventional FACT (new) Conventional

Bearing core 22.8 3.5 12.7 85.0 302 298

Bearing surface 22.8 3.5 14.0 85.0

Spline core 30.5 4.3 15.3 95.0 315 298

Spline surface 32.7 5.5 15.3 117.0

Average 27.2 4.2 14.3 95.5

0 10:00 20:00 30:00 40:00

Time, min:sec

350

300

250

200

150

100

300F

280F

Tem

perature,

F

Bearing surface

Bearing core

Ambient

Spline surface

Spline core


5/6

methods are shown in Fig. 5. Hard-ness profiles are identical for the threetemper treatments.

X-ray Stress AnalysisThe transformation from austenite

to martensite in the hardened axleshafts imparts compressive stress onthe barrel surfaces, which results infavorable fatigue properties of theshaft [1]. Tempering the shafts in atemperature range of 300 to 400F(150 to 205C) relieves some of thesecompressive stresses resulting in amodest increase in toughness [2],which is adequate for the requiredwear resistance, high strength, andfatigue resistance in axle shaft appli-cations.

The compressive stresses at var-ious depths in the shaft were meas-

ured using an x-ray stress analysismethod that uses x-ray diffraction toquantify the amount of crystallo-graphic strain present, and subse-quently calculates the correspondingstress value. Figure 6 shows a plot ofresidual stress versus depth from thesurface. The as-quenched (untem-pered sample) has the greatest com-pressive stresses at 0.020 to 0.025 in.(0.5 to 0.63) depth from the surface.Both the conventional and FACTprocesses similarly relieve the com-

pressive stress at that depth. How-ever, some beneficial compressivestress is still present near the surface(


6/6

methods result in shafts having sim-

ilar rupture strength, and the as-tem-pered strength is superior to shaftsthat are untempered.

Reverse Torsional FatigueThe fatigue effects of repeated for-

ward and reverse torque cycles thatan axle shaft experiences in a vehicleare reproduced in reverse torisonalfatigue testing. In the test, the flangeend of the shaft is fixed and a set tor-sional stress of 75 ksi (517 MPa) is ap-plied through the spline of the shaft

in both the forward and reverse di-rections at 0.83 Hz. The load requiredto produce a 75 ksi stress is typicallyaround 50% of the JAEL load. Thenumber of cycles until shaft failureis recorded and then plotted and an-alyzed using Weibull methods.

The x-axis of the Weibull plotrecords the number of cycles untiltorsional failure and is logarithmicin scale. They-axis can be used to pre-dict the proportion of a populationthat will fail at a given number of cy-

cles. To improve the readability of thegraph, the 90% confidence intervalswere not plotted. However, all treat-ment groups had similar reverse tor-sional fatigue life; no significant sta-tistical differences were detected.

Rolling Bending FatigueA rolling bending test is used to

simulate the fatigue characteristicsof simultaneous bending and rota-tional effects on an axle shaft in a ve-

hicle. In the test, a wheel bearing is

installed onto the axle shaft and lo-

cated on the bearing diameter at thesame position that it would be in thevehicle. The flange is then fixed to afly wheel, which is, in turn, rotatedusing a belt. Aspecial side gear is at-tached to the spline end so a verticalload perpendicular to the axis of theshaft can be applied while the shaftis rotating. The value of this appliedload is calculated based on the de-sign parameters of the axle assembly.The applied tangential load causesthe shaft to bend elastically while it

is rotated producing alternating cy-cles of compressive and tensile stressat the junction of the bearing journaland the back face of the flange andthe shaft/bearing interface. The shaftis rotated under a specified load at1,000 rpm until failure, and thenumber of cycles until failure isplotted using a Weibull plot.

The Weibull plot shown in Fig. 10demonstrates considerable vari-ability in rolling bending fatigue lifefor the different tempering treat-

ments. This may be due to the sensi-tivity of this test to set-up proceduresand to the fact that this testing wasperformed over the course of severalmonths and by different technicians.Several samples of the vertical FACTfurnace had rolling bending fatiguelife in excess of 4 106 cycles andwere suspended without failure. Onthe low end, several samples hadrolling bending fatigue lives of 1.6 to2 106 cycles. The acceptance bogey

for this test is 1 10

6

cycles, which

was easily met for all tempering treat-

ments.

ConclusionsThe results of this work show that

effective tempering of induction-hardened axle shafts is dependent onactual heat transfer rates and compo-nent temperature rather than on timeat a specified ambient furnace tem-perature. This study verifies that tem-pering furnace designs that maxi-mize heat transfer rates to axle shaftsachieve:

Equivalent post-temper hard-ness at the bearing journal surfaces

Equivalent post-temper microin-dentation hardness and residualstress profiles within the shaft

Equivalent torsional JAEL andrupture strengths

Equivalent reverse torsional androlling bending fatigue life

Reduced tempering cycle time(from 4 hours to 36 minutes,or >80% reduction)

References1. Fett, G., Induction Case Depths for Tor-sional Applications,Metal Progress, ASMIntl., p 49-52, Feb. 19852. Krauss, G., Principles of Heat Treatmentof Steel, ASM Intl., p 188, 1980

For more information: Thomas Neu-mann is product design engineer, FordSterling PTO Plant, Sterling Heights,MI 48310; tel: 586/826-5346; fax: 586/826-5837; e-mail: [email protected].

Fig. 9 Weibull plot of reverse torsional fatigue for various tem-pering treatments

Fig. 10 Weibull plot of rolling bending fatigue life by temperingtreatment

Reprinted from the March/April 2006 issue ofHEAT TREATING PROGRESS magazine

rapid tempering

Documents