TEETH SURFACE FAILURES IN AUSTEMPERED …foundrygate.com/upload/artigos/Austepered Ductile Iron Gears.pdf[270 ] paper XIII GEARS 2003 Faculdade de Engenharia da Universidade do Porto,

GEARS & TRANSMISSIONS Workshop paper XIII [ 269 ]

Faculdade de Engenharia da Universidade do Porto, Portugal, 5th June 2003

TEETH SURFACE FAILURES IN

AUSTEMPERED DUCTILE IRON (ADI) GEARS

Luís Magalhães 1; Jorge Seabra 2

1Departamento de Engenharia Mecânica Instituto Superior de Engenharia do Instituto Politécnico do Porto

2Departamento de Engenharia Mecânica e Gestão Industrial

Faculdade de Engenharia da Universidade do Porto

OVERVIEW

This work presents results from gear tests with ductile iron gears submitted to different austempering treatments. Gear tests were performed using an FZG gear test machine and FZG type C gears, which were produced from the same ductile iron and differently austempered, originating six different Austempered Ductile Irons (ADIs). Presented results allow better understanding of these ADIs’ surfaces behaviour when submitted to severe contact conditions, namely high contact pressure levels and low specific lubricant film thickness.

1 OBJECTIVES

The main objective of this work was to study how different ADI surfaces behave when

submitted to severe contact conditions. Having in mind the future perspective of producing gears

using these materials, and considering that the specific contact conditions associated with

gearing kinematics are hardly reproduced by other usual contact fatigue testing equipments,

some ADI gears were produced according to the FZG-C geometric standards and then tested

using an FZG test rig.

Although produced from the same base nodular iron, the tested gears were submitted to

different austempering treatments, producing six materials with different metallurgical structures

and mechanical properties. Beside the characterization of each material surface behaviour,

performed tests also aimed a better understanding of the relation between austempering

treatments’ parameters and contact fatigue resistance of each ADI variety.

[ 270 ] paper XIII GEARS 2003


2 AUSTEMPERED DUCTILE IRON

Austempered Ductile Iron (ADI) has been progressively adopted for the manufacturing of

mechanical parts. This material family includes varieties that match high-strength steel resistance

(tensile strength reaching 1600 MPa), along with very strong wear and scuffing resistance, and

also includes “softer” varieties, presenting high ductility (rupture elongation reaching 12%),

properties that are fundamentally dependent on the base material constitution and heat-treatment

parameters, allowing the production of well-equilibrated materials.

ADIs are suitable for moulding production, as they are easily melted, a process that allows

significant cost-savings when compared to steel conventional manufacturing processes [1].

Energetic savings are also important, as austempering heat treatments require relatively low

temperatures, when compared to conventional steel quenching and tempering or other

conventional surface hardening treatments. These features have been exploited as great

advantages when substituting conventional steel parts, along with the advantages related to the

material itself [2]: lower specific weight (10% less than steel), better vibration absorption

(contacting surfaces produce lower noise levels), the ability to work with EP and AW addictive-

free lubricants (a relevant ecological feature), the graphite self-lubrication capacity (avoiding

scuffing in extreme contact situations) and the TRIP phenomenon, that causes a significant

hardness raise when high pressure levels are imposed to the surfaces (mechanically deformed

austenite transforms into martensite), improving contact fatigue resistance.

3 TESTED ADIs

All tested ADIs were produced from the same base material, a nodular iron that contains 1%

Copper and 0.5% Manganese as main alloy elements [3, 4]. ADIs produced from this material

already revealed a very high scuffing resistance when gears were submitted to FZG scuffing tests

[5] using addictive-free lubricant oils.

For the set of tests included in this work, six ADIs were produced, based on three different

austempering temperatures and on two different austempering procedures.

Temperatures adopted for isothermal transformations were 260, 300 and 340ºC. Two

treatment forms were adopted for each austempering temperature: a conventional

“austenitization + isothermal transformation” austempering treatment and a non-conventional

“austenitization + sub-martensitic cooling stage + isothermal transformation” treatment, here



referred as “double” treatment (the letter “D” tags all materials austempered with “double”

treatments). Table 1 shows a list of all tested ADIs and respective mechanical properties.

Table 1 – Mechanical properties of the tested ADIs

ADI Tensile stress Yield stress Rupture elongation

Impact resistance Hardness

ºC MPa Mpa % KJ HV 260 1625 1325 3.2 53 480 260D 1440 1110 1.1 - 445 300 1435 1170 6.3 73 425 300D 1370* 1093* 3.0* - 433* 340 1145 845 7.6 87 350 340D 1300 1075 3.7 98 420

* extrapolated values

4 TESTED GEARS

Gears were produced according to FZG type C standard, suitable for contact fatigue tests

[6]. Table 2 shows some geometric parameters of these gears.

Table 2 – Characteristics of the FZG-C gear.

FZG – C gears units

centre distance 91.5 mm tooth width 14 mm pinion pitch diameter 73.2 mm wheel pitch diameter 109.8 mm pinion addendum diameter 82.46 mm wheel addendum diameter 118.36 mm module 4.5 - pinion number of teeth 16 - wheel number of teeth 24 - pinion profile shift factor .1817 - wheel profile shift factor .1715 - pressure angle 20 grades working pressure angle 22.5 grades maximum sliding rate 64 % medium surface roughness 0.3±0.1 µm

Teeth surface roughness was higher than desired. Values as high as 2.3 µm (Ra) were

measured at some pinion active surfaces, while wheels were slightly smoother. Average values

of 2 µm were considered for calculation purposes.



5 FZG GEAR TEST RIG

The FZG gear test machine is widely used equipment for lubricant oil testing. Tribology

laboratories often use these machines to test materials, as using the same type of lubricant

provide results that depend only on the characteristics of the gears themselves, namely when they

are produced with different materials. Figure 1 shows a scheme of the used FZG machine.

Figure 1 – Scheme of the FZG gear test rig.

This machine allows contact pressure to reach more than 2 GPa at teeth contact, providing

more than 530 Nm torque and a maximum wheel rotational speed of 3000 rpm. Lubricant

temperature can reach 220ºC, and is permanently controlled. Vibration level control allows the

detection of spalling or other severe damage that may occur at teeth surfaces.

6 LUBRICANT

An ISO VG 150 lubricant oil containing some EP (Extreme-Pressure) and AW (Anti-

Wear) addictives was used during all phases of these gear tests. Lubricant samples were analysed

after each test, allowing the quantification and identification of small metallic particles resulting

from teeth active surfaces’ wear. Table 3 shows some characteristics of the used lubricant.

1 – test pinion 2 – test wheel 3 – drive gear 4 – load clutch 5 – locking pin 6 – load lever and weight 7 – torque measuring clutch 8 – temperature sensor



Table 3 – Lubricant properties.

Kinematic viscosity at 40ºC 148.8x10-6 m2/s

Kinematic viscosity a 100ºC 14.68x10-6 m2/s

Thermal condutibility 0.1318 W/m2K

Density 895 Kg/m3

7 GEARS’ RUNNING IN

All gears were submitted to a running-in period in stage FZG 4, to which corresponds a

contact pressure of about 0.7 GPa. During this period, different rotational speed was imposed to

the gear wheel, according to Figure 2. Each gear wheel accomplished 270.000 cycles during this

phase.

The running-in period is important because it allows teeth profiles to adapt each other under

moderate charge conditions, smoothing surfaces by lowering roughness and favouring load

distribution.

Surfaces’ roughness resulting from manufacturing processes was relatively high, averaging

2 µm (Ra), but was significantly lowered during this running-in phase.

0

500

1000

1500

2000

2500

3000

3500

0 50000 100000 150000 200000 250000 300000

rpm

cycles

Figure 2 – Running-in sequence of tested ADI gears.

Lubricant temperature was kept at 85ºC during surfaces’ running-in. Lubricant samples

were also token after each running-in period in order to be analysed by ferrometric and

ferrographic techniques.



8 GEAR TESTS

Twelve gear tests were performed, two tests per ADI, each one at a different contact

pressure level. A set of tests was performed using stage FZG 8, and another one using FZG stage

10. To these stages correspond pitch diameter contact pressures of 1.4 GPa and 1.8 GPa,

respectively.

Gear wheel rotational speed was kept constant (3000 rpm) during the tests. Table 4 shows

some other relevant operating conditions.

Table 4 – Contact conditions during FZG-C gear tests.

Stage 10 Stage 8 units

Maximum sliding rate (A) * 63 63 % Maximum normal force (C,I,D) 11014 7080 N Maximum Hertzian pressure (C) 1.79 1.439 GPa Minimum lubricant film thickness (A) 0.23 0.35 µm Minimum specific lubricant film thickness (A) (Ra = 2 µm)

0.06 0.08 µm

* letters refer to gearing points, defined as A, C, I, D, B.

Lubricant temperature was controlled along the tests, aiming permanent 100ºC, but it was

not possible to avoid some overheating, mainly during tests at 1.8 GPa, when lubricant

temperature reached values as high as 116ºC.

9 TESTS RESULTS

Three different failure modes were observed:

1 - Fracture of one ore more teeth, an occurrence that involves bending fatigue at teeth base

and is not a contact fatigue phenomenon;

2 – Surface spalling, originating craters of significant dimensions, a typical failure mode for

high contact pressure conditions;

3 - Pitting, mainly concentrated at active surfaces near teeth pitch diameter, a typical failure

mode when contact pressure levels are moderate.

These failure modes can be related to the number of cycles accomplished by the tested

gears. Fractures and spalling occurred within a short number of cycles. Pitting usually took a



relatively high number of cycles until accumulated damage became responsible for high

vibration levels (the FZG machine allows the selection of a limit vibration level. When reached,

the test is automatically stopped).

Table 5 shows the results of these tests, indicating failure modes and other relevant

parameters. These results are also shown at Figure 3, where the numbers near the graph bars

correspond to the number of teeth where spalls were found.

Table 5 – Results of all FZG-C gear tests.

ADI p0 σr p02 /σr Tmax time failure type (cycles / 1000)

# damaged

teeth

max. spall

length ºC GPa Mpa GPa ºC min. Fracture Pitting Spalling mm 340 1.8 1145 2.83 100 40 120 14 5

340D 1.8 1300 2.49 105 120 360 5 3 300D 1.8 1370 2.36 97 90 270 29 14 300 1.8 1435 2.26 106 120 360 6 3

260D 1.8 1440 2.25 100 165 495 - 260 1.8 1625 1.99 108 150 450 - 340 1.4 1145 1.71 106 180 540 1 1.7

340D 1.4 1300 1.51 113 345 1035 24 5 300D 1.4 1370 1.43 99 510 1170 [1530] 0.4 300 1.4 1435 1.37 108 276 828 10 1.2

260D 1.4 1440 1.36 117 814 1611 [2442] 0.1 260 1.4 1625 1.21 108 100 300 -

[ ] brackets refer to the number of cycles at the end of the tests.

0 500 1000 1500 2000 2500 3000

260D

300D

340D

260

300

340

ADI

cycles

1.8 GPa1.4 GPafracturespalling

X

XX

10

141

6

29

524

X#

Figure 3 – Results of all FZG-C gear tests.



10 SURFACE INSPECTION

Teeth surfaces were inspected using a portable video-microscope system, equipment that

allowed frequent inspections without having to dismount the test gears from the FZG machine.

After the tests, some teeth were cut and metalographic samples were produced to be inspected

with higher resolution microscopes.

Surface inspection and spall characterization allowed a better understanding of the

influence of the contact pressure onto each ADI surface behaviour when submitted to contact

fatigue. Some images were selected to illustrate surface condition. In most of these images,

surfaces are seen from the top of each tooth, thus rolling direction is from top to bottom.

10.1 Teeth fracture

Teeth fractures were observed during some tests where contact pressure attained 1.8 GPa.

These fractures only affected gears austempered at 260ºC, both very resistant but brittle materials

among the tested ADIs. An exception occurred when the ADI260 wheel fractured one tooth

during a 1.4 GPa test. This occurrence may be justified by an additional reason: as the ADI260

pinion lost one tooth during the 1.8 GPa test, a steel replacement pinion was used for this 1.4

GPa test. In this way, pinion and wheel could not be correctly runned-in, and profile geometric

differences may have been a cause for poor gearing conditions, vibration and abnormal stress

concentration, leading to the unexpected fracture. It is also important to refer that this wheel had

already survived a previous 1.8 GPa test where pinion teeth fractures occurred.

Figure 4 shows one tooth fractured from the 260ADI pinion, where the surface fracture

(image at the right) reveals some roughness due to the propagation of the main bending fatigue

fissure along the base of the tooth.

Figure 4 - Fractured tooth of the ADI260 pinion.



All observed fractures were very similar, always propagating along teeth base, leaving very

typical and homogeneous cylindrical and slightly rough fracture surfaces (see Figure 5). A close

observation of the 260 pinion fractured tooth revealed some matrix heterogeneities near the

active surface (see Figure 6). The location of these defects was critical, and may have, in some

extent, promoted the occurrence.

Figure 5 - Fracture surface at ADI260 pinion.

Figure 6 – Matrix heterogeneities at ADI260 pinion fracture surface.



10.2 Spalling

The larger spalls were found at teeth belonging to the flank of ADI300D pinion that was

submitted to the 1.8 GPa test. The large spall shown at Figure 7 is formed by coalescent craters

and extends over the whole tooth width (14 mm), severely destroying the surface.

Figure 7 – Tooth wide spall at 300D pinion.

Figure 8 shows another example of severe spalling at the surface of one 300D pinion tooth,

revealing a relatively poor contact fatigue resistance of this ADI when submitted to high pressure

levels.

Figure 8 – Large spall at 300D pinion (aprox. 12 mm wide).

The dimensions of the contact fatigue spalls revealed to be dependant from the contact

pressure, but a relation with the resistance of each tested material also became obvious.

Figure 9 and Figure 10 show spalling phenomena at the surface of teeth belonging to

different gears tested in similar conditions, both tests performed at 1.8 GPa.



Figure 9 – Spalling at 340 pinion (aprox. 5 mm).

Figure 10 – Spalling at 340D pinion (aprox. 3 mm).

In general, observed fatigue spalls revealed to be larger and deeper when ADIs were less

resistant and more ductile.

The dimensions of the contact fatigue spalls were much smaller when surfaces operated

under 1.4 GPa. In this group of tests, no spalling phenomena were observed at the surfaces of the

harder materials. Only softer ADIs presented small craters, identical to the example shown in

Figure 11.

Figure 11 – Contact fatigue spall at 340D pinion (aprox. 3 mm).



In some cases, and in spite of the small dimensions of the fatigue craters, their coalescence

damaged large portions of the active surfaces (see Figure 12).

Figure 12 – Alignment of small fatigue craters near the pitch diameter of 340D pinion.

Such small pit alignments may either be considered a minor spalling phenomenon or a

severe pitting occurrence. Based on the results obtained from these ADI gears tests, it can be

stated that these two damage mechanisms intercept when the relation between contact pressure

and material tensile strength attain certain levels.

10.3 Pitting

Large contact fatigue craters were not found only on two tests performed at a maximum

Hertzian pressure of 1.4 GPa. These tests were performed with some of the stronger and harder

tested ADIs (260D and 300D). Active surfaces suffered some damage, mainly consisting in

fatigue micro-craters and micro-fissures (a typical form of ADI pitting), and a few small fatigue

craters of low depth were sometimes found (see Figure 13)

Figure 13 – Small contact fatigue pits at 300D pinion (aprox. 0.2mm).

The hardest among the tested materials (ADI260D) resisted for a high number of cycles

when tested at 1.4 GPa, keeping teeth surfaces in a relative good condition. The test was stopped

because the vibration level was already significant.



Figure 14 shows surface condition of one 260D wheel tooth active flank, at the end of this

test, revealing a lot of pitting along the pitch diameter.

Figure 14 – Pitting at 260D wheel.

Figure 15 shows the tip of a 260D tooth after the 1.4 GPa test (left side of the image),

where some wear traces can bee seen. A concentration of micro-cavities is located at the height

of the pitch diameter (rougher area at the center of the image), and a smoother surface can be

seen below that point (at the right side of the image).

Figure 15 – Pitting near the pitch diameter of a 260D wheel tooth (the rolling direction is from left to right, and the

image height is approximately 5 mm).



11 LUBRICANT ANALYSIS AND SURFACE ROUGHNESS

Lubricant oil analysis and surface roughness measurement were used as auxiliary

techniques to characterize the evolution of the tests

Figure 16 shows ferrometric results representing the number of metallic micro-particles

found in the lubricant samples for some of the tests performed using 1.8 GPa contact pressure

(these values are already divided by the number of cycles accomplished by each gear). They

reveal a strong dependence of the wear rate relatively to the ductility (rupture elongation) of each

tested ADI. The stronger materials were the ones that better resisted wear, proportionally to the

number of stress cycles.

00.20.40.60.8

11.21.41.61.8

340 300 340D 260

DSISUC/1000DLCPUC/10

ADI

Figure 16 – Lubricant ferrometric analysis values for some gears (p0 = 1.8 GPa).

Surface roughness was measured before and after the tests. In a general way roughness

diminished strongly, particularly on 340 gear teeth, the less resistant and more ductile among the

tested ADIs. Pinions also revealed stronger roughness attenuation than wheels, a natural

consequence of the correspondent higher number of cycles. In some particular situations, final

roughness values were higher than initial ones. This occurred mainly when pitting damaged the

active surfaces, even if respective tests ended later by spalling. In this way, roughness values are

not presented here as they result from different mechanisms and cannot be individually

correlated to each ADI contact fatigue resistance.



12 DISCUSSION OF RESULTS

The strong influence of the imposed contact pressure and the tensile strength of the tested

ADIs can be well demonstrated by plotting the results of these tests, in terms of number of

cycles, against a severity factor (p02/σr), which adapts itself very well to all the results (except

for the cases where fractures occurred). Figure 17 shows this graphical representation.

y = -0.5584Ln(x) + 9.3334R2 = 0.8591

0

0.5

1

1.5

2

2.5

3

100 1000 10000cycles x 1000

300D

300

340D340 260D

pitting

FZG-C tests (not including fracture results)p 02 / σr [GPa]

300

340D

300D340

Figure 17 – Number of cycles versus p0

2/σr for ADI FZG-C gear tests.

This representation shows that small number of cycles correspond to high p02/σr values

with relatively good correlation. There is also a tendency relating failure modes with this severity

factor, as it’s lower values correspond to the tests where only pitting occurred. With the

exception of ADI3001.4 GPa test, values of p02/σr higher than 1.5 always leaded to spalling, and

lower values leaded to pitting.

Another important aspect is that to softer materials (where large spalling occurred)

correspond relatively high p02/σr values among the tests performed under the same contact

pressure. Although this is a natural mathematic consequence resulting from the form of the used

factor, it allows the conclusion that fatigue failure modes and associated damage can be well

related to this severity factor. This relation appears to be valid not only to the dimension of

fatigue craters but also to the amount of accumulated damage when pitting is the predominant

failure mode.

A note must be made about the small number of tests, meaning that this correlation cannot

be generalized and is only valid for this set of contact fatigue results, i.e., these results are not

statistically representative of the contact fatigue resistance of the tested ADIs. On the other hand,

when globally considered these results point to some tendencies and establish a contact pressure



range where failure modes and damaging mechanisms could be clearly identified at the active

surfaces of the tested ADIs.

13 CONCLUSIONS

Results from FZG-C contact fatigue tests performed with gears made from different ADI

varieties revealed that the nodular iron used to produce these gears can be successfully

austempered, originating ADIs resistant to high-level contact fatigue stressing.

Among these materials, the harder and more resistant ADIs (with higher tensile stress)

were the ones that better resisted to contact fatigue solicitations. Less resistant and ductile

materials leaded to the worse results among the performed tests.

A particular note to the varieties austempered at 260ºC, which kept the working surfaces in

very good condition, revealing a notorious contact fatigue resistance. In the other hand, only

these ADI gears suffered teeth fractures, a condition that is mainly dependent on the bending

fatigue resistance of the teeth base (not studied in the scope of this work). Although some

already referred factors may have had some influence on the observed fractures (namely the

presence of foundry defects at the sub-surfaces, which can be of significant influence) the low

ductility of these ADIs may be the ultimate responsible for such occurrences.

The double austempering heat treatment was apparently benefic, having in consideration

the comparison between results obtained from simple and double ADIs treated at the same

temperature, but the small number of performed tests does not allow a secure conclusion on this

subject.

It can be concluded that austempering heat treatments should maximize the tensile strength

of the ADIs but should not be a cause for noticeable embrittlement, as a significant loss of

ductility can severely diminish the bending fatigue resistance of gear teeth. A compromise

situation seems to be required, in order to produce materials that can be effectively used for the

manufacturing of high-load transmission gears.



REFERENCES

[1] - R.A.Harding, The use of austempered ductile iron for gears, 2º Congresso mundial de engrenagens, Paris, 1986.

[2] - L. Magalhães, Resistência ao Desgaste e à Gripagem de Engrenagens em ferro Fundido

Nodular Austemperado (ADI), Dissertação de Mestrado, Faculdade de Engenharia da Universidade do Porto, Dezembro de 1995.

[3] - H. Santos, A. Pinto, V. Torres, Cu-Mn ADI: A low cost high performance material, 58th world

foundry congress, Polónia, 1991. [4] - H. Santos, A. Duarte, J. Seabra, Austempered ductile iron with tempered martensite,

International Journal Cast metals, 15, 2002. [5] - L. Magalhães, J. Seabra, Wear and scuffing of austempered ductile iron gears, WEAR 215

(237-246), 1997. [6] - Winter, H. and Michaelis, K., FZG Gear Test Rig - Description and Test Possibilities, Co-

ordinate European Council Second International Symposium on “The Performance Evaluation of Automotive Fuels and Lubricants”, Wolfsburg, West Germany, June 5-7, 1985.

ACKNOWLEDGMENTS

To the Fundação para a Ciência e Tecnologia for the financial support given to this work under

contract PRAXIS XXI – 3/3.1/2666/95.

To Professor H. Santos from the Department of Metallurgy and Materials of the Faculdade de

Engenharia da Universidade do Porto, for the ADI heat treatment.



TEETH SURFACE FAILURES IN AUSTEMPERED …foundrygate.com/upload/artigos/Austepered Ductile Iron Gears.pdf[270 ] paper XIII GEARS 2003 Faculdade de Engenharia da Universidade do Porto,

Documents