Fatigue life improvement on 20MnCr5 steel through surface modification for auto transmission application

Fatigue life improvement on 20MnCr5 steel through surface modification for auto transmission application

archives of civil and mechanical engineering 19 (2019) 360–374 Available online at www.sciencedirect.com ScienceDirect journal homepage: http://www...

6MB Sizes 2 Downloads 21 Views

archives of civil and mechanical engineering 19 (2019) 360–374

Available online at www.sciencedirect.com

ScienceDirect journal homepage: http://www.elsevier.com/locate/acme

Original Research Article

Fatigue life improvement on 20MnCr5 steel through surface modification for auto transmission application Ramesh Srinivasan a,*, S. Natarajan a,b, V.J. Sivakumar a,c a

Centre of Excellence in Corrosion and Surface Engineering. (CECASE), National Institute of Technology, Trichy, Tamilnadu, India b Department of Metallurgical and Materials Engineering, National Institute of Technology, Trichy, Tamilnadu, India c Department of Management studies, National Institute of Technology, Trichy, Tamilnadu, India

article info

abstract

Article history:

This research represents a unique approach in improving the fatigue life of 20MnCr5 (89–91

Received 15 August 2018

HR15N hardness) shafts which were surface treated through gas carburizing process in a

Accepted 25 November 2018

sealed quench furnace followed by double tempering. Crack initiation on 20MNCr5 trans-

Available online

mission shafts always occurs at the end of spline location (Location X), propagates longitudinally and finally ruptures at stem location. This was confirmed through optical microscope

Keywords:

and SEM (make:Jeol) images. The fatigue experiments were carried out at room temperature

Gas carburizing

in 11,500 Nm torque test machine (Model – MTS-663-144-01). The torque test was carried-out

Inter-granular oxidation

by applying a fully reversed cyclic load with the frequency of 5 Hz for the torsional load of

Non-martensitic transformation

3000 Nm. Effect of double tempering, surface roughness, carbon case depth, Microstructure

products

such as retained austenite and non-martensitic transformation products (NMTP) have been

Retained austenite

investigated in this research. The outcome of the research shows that increase in case depth

Fatigue life

(CD) (out of increased gas carburising time) resulting in increased inter-granular oxidation (IGO) and NMTP in its microstructure. The presence of retained austenite on the surface of the shaft has not yielded any remarkable improvement in fatigue life of the shaft. Presence of tempered martensite with lesser percentage of retained austenite on the surface improved the fatigue life drastically from 12,000 cycles to greater than 35,000 cycles at high torque load of 3000 Nm. © 2018 Published by Elsevier B.V. on behalf of Politechnika Wrocławska.

* Corresponding author at: Department of Metallurgical and Materials Engineering, National Institute of Technology, Tiruchirappalli, India. E-mail address: [email protected] (R. Srinivasan). https://doi.org/10.1016/j.acme.2018.11.005 1644-9665/© 2018 Published by Elsevier B.V. on behalf of Politechnika Wrocławska.

361

archives of civil and mechanical engineering 19 (2019) 360–374

1.

Introduction

Fatigue life of the shaft is an important factor to decide the performance of the transmission. The effects of combined bending and torsional loading are taken into consideration for the design of transmission shafts.

1.1.

Literature survey

Fatigue strength and fatigue limit of the shafts are broadly influenced by the conditions such as. (1) surface condition, (2) stress concentration, (3) residual stress [15], (4) microstructure, (5) heat treatment processes [17], (6) corrosion fatigue [16] and (7) material. Surface roughness was the most important influencing factor in short life regime, and the fatigue life was found to decrease with increase of surface roughness [2]. Also, Arola and Williams [3] found that the high-cycle fatigue life of machined specimens of AISI 4130 steel is surface-texture dependent, and that the fatigue strength decreased with an increase in surface roughness from 2 to 6 mm. Rotational bending Carbon steel specimens (weight % of Fe – 98.5, C – 0.214, Mn – 0.725, Si – 0.133, S – <0.05, P – <0.05, Cr + Mo + Ni + Sn – <0.32) under machined condition are subjected to fatigue test at 50 Hz reversed cyclic load. It was observed that more number of fatigue crack initiation sites in courser specimen as that of finer specimen. These fatigue cracks were observed to be trans-granular in nature [4]. Carburizing process involves the production of a high level of carbon at the surface layer of steel of transmission shafts made of low carbon low alloy steel. It is observed that the presence of internal oxidation, retained austenite which is formed during gas carburisation process decreases the fatigue strength [5–7,13]. Some researches revealed that moderate retained austenite increases fatigue resistance, whereas other study suggests keeping the retained austenite in minimum condition for optimum fatigue results [8]. Effective and total case depth (ECD and TCD) is one of the most important characteristics which affect fatigue performance of carburized steels. Process parameters like time and temperature during the various stages of a carburizing process, hardenability, part shape or geometry affect the case depth [5]. In transmission shafts an increase in the case depth increases fatigue performance [9]. Exceeding the limiting value cannot be recommended for machined parts which have been subjected to a carburizing treatment. In order to obtain good fatigue resistance of transmission shafts, the case depth should be kept at an optimum range depending upon the shaft geometry and processes involved. When the shafts are subjected to carburising atmosphere, the internal oxidation will be formed on the surface layer. It is obvious that fatigue crack initiation and development are confined to surface grains, but the effect of the depth of internal oxidation upon fatigue strength is not fully understood [11]. It has also been reported that the fatigue strength is dependent upon the depth of internal oxidation, while some studies suggest otherwise [18].

1.2.

Primary objective

The prime objective of this research is to explore possible avenues in optimising the process parameters in turn to maximise the fatigue life. One of the options explored in gas carburisation process of transmission shafts was to introduce double tempering operation to reduce retained austenite, followed by introduction of machining and super-finishing operations to remove NMTP and IGO from the shaft surface. With these surface improvements [12] and microstructure modifications it is clearly evidenced that the fatigue performances of the transmission shafts have experienced 3 fold improvement in its fatigue life.

2.

Materials and methodology

2.1.

Material type

Steel grade of 20MnCr5 material's chemical composition is given in Table 1. This material was considered for research study due to its widespread usage in automobile transmission applications. This material has good hardenability, strength and toughness, finding its applications for high torque requirements such as shafts and gears.

2.2.

Sample preparations

'Transmission shaft samples (Fig. 5) are taken for the research whose manufacturing process flow is given below (Fig. 1). Shafts are cold forged, normalised, soft turned, case carburised and tempered to maintain the required metallurgical as well as mechanical properties as given below. Then, shafts were subjected to double tempering process and subsequently to surface machining and super finishing operations so as to maintain surface hardness of 89–91 HR15N and surface roughness of 30–50 mm as well.

2.3.

Experimental setup

Carburization of steel involves a heat treatment of metallic surface using a source of carbon potential. Carburization can be used to increase the surface hardness of low carbon low alloy steel. Gas carburizing is normally carried out at a temperature within the range of 900–950 8C. This transmission shafts is carburised, hardened and tempered in an enclosed atmosphere called sealed quench furnace (Fig. 2). This heat treatment process (Fig. 4) involves pre-heating, carbon activation and diffusion, Hardening, oil quenching and tempering. Then same cycle was repeated in second tempering which is known as double tempering. This heat treatment is a batch type process where in the batch is washed prior to

Table 1 – 20MnCr5 nominal composition, all values in wt %. C 0.19

Si 0.30

Mn 1.25

S 0.020

P 0.015

Cr 1.18

Al 0.03

362

archives of civil and mechanical engineering 19 (2019) 360–374

Fig. 1 – Manufacturing process flow chart.

heat treatment and post heat treatment. This heat treatment cycle (Figs. 3 and 4) comprises of diffusion of Endo (Endothermic gas) and LPG with Prefixed carbon potential at specified temperature and duration. This cycle of process is fully programmed and automated. Machining in turning operation (Fig. 6) is carried out with at most care to maintain the required surface roughness of the shaft as the surface roughness play a very important role on the fatigue strength of the shaft [10]. The feed lines formed in turning operation may lead to the formation of stress raisers and hence this surface is finished through super finishing operation with 600 mm diamond emery cloth to achieve 0.5 Ra surface roughness. This also ensures the uniform surface roughness throughout the shaft to avoid stress raiser normal to circumferential direction [2].

2.4.

Torsional fatigue setup, simulations and procedure

Torsional fatigue testing bench validation machine MTS make 11,500 Nm capacity is shown in Fig. 7. The torque testing arrangement is such that either side of splines (splines are tooth on the transmission shafts that mesh with grooves in a mating surfaces of actuator and reaction or torque sensor side and transfer torque to it) are engaged between actuator side

Fig. 3 – Fixture loading.

and reaction or torque sensor side. Normally the reactor side is kept static and torque side is designed with angle of twist of 2.08 with interlock limit of 0.18. In case of ultimate torque test, the shaft is twisted till fracture. In case of torsional fatigue cycle test the torque levels are set in the system and the actuator side is programmed for continuous cyclic indexing and cycle will continue till fracture. The torsional load versus no. of cycles is recorded in the system. According to the shaft design (finite element analysis) and end application, the shaft is subjected to fatigue testing at three levels of torque loads namely high torque, medium torque and low torque. The torque levels are used based on transmission performance requirements inline with vehicle loading conditions. Ultimate tensile strength (UTS), yield point of material and hardness (89HR15N) of transmission shafts are used for determining the test levels of variable torques which are governed by the transmission design and applications. As per the hardness of transmission shafts (89HR15N), UTS was arrived at 2145 MPa. These specific high, medium and low torque levels are arrived for the intended application and are part of transmission design requirements. Shafts will have infinite life at 50% of UTS (2145 MPa) as per S–N curve. So, the operating maximum torque of transmission shaft arrived at 1000 Nm for the shafts based on transmission rig testing.

Fig. 2 – Sealed quench furnace schematics.

archives of civil and mechanical engineering 19 (2019) 360–374

Fig. 4 – Heat treatment process cycle.

Fig. 5 – Machined and polished shafts.

Fig. 6 – Machining setup.

Fig. 7 – Torsional fatigue test bench setup.

363

364

archives of civil and mechanical engineering 19 (2019) 360–374

With reference to the yield point, different proportions of torsional loads are selected to ascertain finite fatigue life in an accelerated test conditions. And hence the torque levels for testing are very specific with high load of 3000 Nm, medium load of 2400 Nm and low load of 1700 Nm. This torque testing is carried out at controlled atmosphere to avoid any influence of external condition such as temperature, pressure etc. While testing the transmission shaft (specimen) on a test bench, it was subjected to air cooling (purging) to avoid heating and hence increase of temperature is avoided. Typically the testing is maintained at 26 8C. This torque testing machine is validated and accredited by internationally certified laboratory.

3.

Results and discussion

3.1. Surface, cross-sectional morphology, microstructural characterisation and hardness testing Microstructural characterisations of both single, double tempered T& machined shafts were observed (Fig. 8) and infered under an optical microscope (Nikon: Eclips LV 150N). Here the samples were cut at location 'X' (along the start of the spline, Fig. 5) and moulded in a hot moulding machine. 3% Nital etchant was used for etching the samples.

Fig. 8 – Microstructure characterisation of single, double tempered and machined samples.

archives of civil and mechanical engineering 19 (2019) 360–374

Microstructural characteristics were studied in bothunetched and etched conditions. Un-etched condition samples were observed to have inter granular oxidation on the case region and its IGO level was photographed and measured. The etched sample was observed to have Non martensitic transformation product on the case region and its NMTP level was also photographed and measured (Fig. 8). Its is also observed to have the presence of fine tempered martensite along with some retained austenite on the case and low carbon martensite & bainite on the core region. Here detailed metallurigical parameter such as surface harness, core hardnessand case depth were measured and recorded. Further fractography (Fig. 8) anaylsis of fractured shaft has also been done using a Scanning Electron Microscope (make: Jeol, Model: JSM-IT300, Year: 2015). It is also observed to have the growth of NMTP towards the core region. Further it is also observed that the growth of NMPT is proportionate to the heat treatment cycle time. The level of NMTP measured in the single tempering sample are in the range of 19–24 m and its range in double tempering sample are between 19 and 23 m. It is further observed that retained austenite on the case region in single tempering sample was about 10% whereas in double tempering samples was about 5–10%. The surface and core hardness of single and double tempered shafts tested in Rockwell hardness tester (Dakomaster 300) was observed to be in the range of 89.5–90.6 HR15N and 39.0–41.0 HRC respectively. Carbon case depth measured from the surface was observed to be 0.90 mm on both single and double tempering samples. All the metallurgical parameters observed from single and double tempering samples are detailed in Table 2. Tempering temperature (1508), tempering time (120 min), Precipitation kinetics and Mf temperatures (at core 250 8C and

365

at case 60 8C) are one and the same for both first and second tempering. But, marginal reduction in retained austenite was observed after second tempering process which is mainly due to the conversion of retained austenite into tempered martensite. This is because of the precipitation of transition carbides. It has been recorded that transformation of retained austenite into other transformation products happens during tempering/double tempering. Due to precipitation of transition carbides, increase of surface hardness occurred and subsequent slight reduction in surface hardness. The precipitation of e/h-carbide is associated with a maximum in the hardness around 100 8C followed by a decrease on rising temperature. The maximal hardness is ascribed to the precipitation of coherent carbide. The subsequent decrease of the hardness can be due to the reduction to the extent of the strain fields as a consequence of the coherent–incoherent transition of the carbides [21]. From the above metallurgical results it is clearly evidenced that double tempered shaft found to have lesser percentage of retained austenite as compared to that of single tempered shaft [19].

3.2.

SEM observation of fractured sample

3.2.1.

Fractography

Fracture surface morphology of torsional failed sample has been analysed under Scanning Electron Microscope to study the failure mode. The fracture morphology has been analysed as a whole and specific areas were captured and documented. Fracture analysis has been performed in four regions i.e. on 12 o' clock, 3 o' clock, 6 o' clock, 9 o' clock positions and in core region of the failed sample (Fig. 9).

Table 2 – Metallurgical characterisation of single, double tempered and surface modified samples. Parameters

Single tempering sample

Surface hardness (HR15N) Core hardness (HRC) Case depth (mm) – HV0.5 NMTP Retained austenite (RA) % Microstructure at case Microstructure at core

89.7–90.5 39.4–41.0 0.90 19–24 m 10 Tempered martensite, free from carbides Low carbon martensite, bainite

Double tempering sample 89.5–90.1 39.0–40.9 0.90 19–23 m 5 Fine tempered martensite, free from carbides Low carbon martensite, bainite

After surface material removal 89.5–90.6 39.0–40.9 0.90 Nil 5 Fine tempered martensite, free from carbides Low carbon martensite, bainite

Fig. 9 – (a) Fractured sample, (b) SEM image at core region and (c) magnified view of (b).

366

archives of civil and mechanical engineering 19 (2019) 360–374

The observations from scanning electron microscope images show an intergranular morphology on the outer regions which are due to a brittle mode of failure owing to the presence of case (case hardening) in the shaft. Core region shows dimple morphology due to ductile mode of failure. Further to elaborate the reason for having two kinds of fracture morphology in case-hardened shaft is due to the carbon gradient present amongst case depth and core region. The carbon gradient across the cross section of the component leads to higher martensite start temperature in the core and lower martensite start temperature in the case. Another reason behind this gradient is also due to the quenching of carburised component at critical cooling rate resulting in a microstructure distribution with plate martensite in the case and lath martensite in the core [20]. Inferences which we could potentially observe from the fracture morphology (Fig. 9) are the case region which is rich in carbon due to case carburising expressed an inter-granular mode of failure i.e. crack along the grain boundary. This mode of failure is usually brittle and the reason behind it could be the presence of case on the outer region. The core region has dimple morphology resulting due to a ductile mode of failure which is due to low carbon core region compared to the case region.

3.3.

Effect of surface roughness

The influence of surface roughness on fatigue limits of double tempered and polished specimens is summarised in Fig. 10. The fatigue limit is not proportional to surface roughness for un-machined samples where NMTP contributes for the fatigue life. Result shows fatigue result is not proportional to surface roughness value which is due to the presence of NMTP in the case. Fatigue cycle values of double tempered machined and polished data are shown in the above Fig. 10(a) and (b). It is clearly evident that fatigue cycles are far higher in NMPT removed samples.

3.4. life

Effect of carbon case depth, RA and NMTP in fatigue

3.4.1.

Carbon case depth

In gas carburizing process, improving fatigue and wear resistance characteristics of steels are done by improving carbon case depth. Hard wear resistance surface is called case. This parameter is considered to be part of shaft geometry and carburized case depth from the surface which determine the fatigue performance of the shaft [16]. Here, Shafts were subjected to different heat treatment cycles (activation time varied as in Fig. 4), to obtain varied case depth level between 0.8 mm and 1.2 mm for a batch of 10 shafts. One shaft per batch (Fig. 11(a) and (b)) was measured for case depth at location 'X' cross section. Two shafts per heat treatment batch were tested for torsional fatigue at torsional load of 3000 Nm. Shafts are studied for correlation behaviour between carbon case depth and fatigue life (Fig. 11). The case depth in the sample was measured by Hardness traverse methods using Micro Vicker's hardness tester. Case depths of the samples were in the range of 0.80–1.20 mm. The samples with varied case depth were subjected to fatigue test at high torque load of 3000 Nm. After due Validation, it is observed that there is no remarkable difference in fatigue cycles between high and low case depth shafts as shown in the graph above. It is very clearly evidenced that increase of case depth did not influence fatigue cycles beyond certain case depth level. It is purely because of disproportionality of surface hardness with that of case depth. Increase of carbon case depth requires increased gas carburising time and NMTP and RA increase proportionate to the carbon case depth. This NMTP and RA present on the shaft surface causes the crack nucleation and crack propagation from the surface, which in-turn affects the fatigue performance. Both the samples resemble similar properties in terms of fracture morphology. From the fractography (Fig. 12(a)–(d)) analysis, Periphery portion shows Inter granular fracture

Fig. 10 – (a) Fatigue results (high torque W3000 Nm) of double tempered and polished samples and (b) fatigue results of double tempered, machined and polished samples. morphology and core region shows dimple morphology. The

archives of civil and mechanical engineering 19 (2019) 360–374

367

Fig. 11 – (a) Case depth vs fatigue cycles and (b) case depth measurement.

Inter granular morphology in periphery region is due to the presence of case depth which is having higher carbon content compared to the core region.

3.4.2.

EDS analysis for Sample 1 (carbon case depth 1.2 mm)

See Tables 3 and 4 and Fig. 13.

3.4.3.

EDS analysis for Sample 2 (carbon case depth 0.8 mm)

See Tables 5 and 6 and Fig. 14.

3.4.4.

EDS analysis report

EDS (energy dispersive spectroscopy) analysis for torsional fatigue failed samples has been captured under scanning electron microscope (Fig. 11). EDS analysis in SEM (make: Jeol, Model: JSM-IT300, Year: 2015) has been done to analyse the chemistry on the fractured surface. From EDS analysis of fractured surface sample we could observe higher carbon in the periphery region (Figs. 13, 14 and 17) and carbon content of 10.43%, 10.33% and 21.07% (Tables 3, 5 and 8) respectively. This was due to the presence of case depth and this reflects an intergranular fracture morphology resulting to a brittle mode of failure. Likewise when EDS analysis of core regions (Figs. 13, 14 and 17) were studied and found to have a lower carbon content than that of the periphery region. Observations of carbon in core region were 9.06%, 7.63% and 7.59% respectively (Tables 4, 6 and 9). Hence dimple morphology is observed in the core region causing a ductile mode of failure.

3.4.5.

Retained austenite

Transmission shafts are heated above austenitic temperature and quenched in oil media to form hard martensitic structure. Austenite is an FCC phase i.e. stable above eutectoid A1 temperature (723 8C). Because of material composition, final hardening temperature and quench rate, incomplete transformation occurs during quenching process resulting in austenite retention in the shaft surface.

Increase in carbon content lowers significantly the martensite start temperature (Ms) and martensite finish temperature (Mf) resulting in the formation of retained austenite below room temperature. Higher RA content can result in lower elastic limit, reduced hardness, lower fatigue cycle (at high load) performance and dimensional instability. Lower RA can result in poor fracture toughness and reduced fatigue cycle (at low load) [14]. Retained austenite present in transmission shafts normally causes increased compressive stress during its load application. Increase in localised RA results in dimensional instability of the transmission shaft which in-turn leads to crack initiation and there by reduced fatigue performance [14].

3.4.6.

XRD analysis on retained austenite

The metallurgical report shows that Gas carburised and tempered shafts have 5–10% of retained austenite which decreases from shaft surface to core. The integrated intensities of the austenite (220 and 311) and ferrite (200 and 211) diffraction peaks are measured by using the volume percentage concentration of RA in the sample in XRD (Make: Bruker Model: D8, Cr_Ka, X-ray beam at 20 keV, l = 2.291 Å and Ø  156.418) (Figs. 15 and 16). Shafts were studied in order to find relationship between percentages of retained austenite (RA) versus fatigue life (Fig. 16). Shafts which have RA from 5% and 10% are subjected to fatigue testing at high torque cyclic reverse load of 3000 Nm. It is inferred that there is no remarkable increase in fatigue cycle between 5% retained austenite and 10% retained austenite. The same is plotted in Table 4. It is clearly evident that change in retained austenite percentage (5–10%) has very limited influence on surface hardness. Further experiments carried out on double tempering operation decreases the retained austenite percentage due to the conversion of fine tempered martensite. This reduction in retained austenite on the shaft surface region improves the fatigue life marginal.

368

archives of civil and mechanical engineering 19 (2019) 360–374

Fig. 12 – SEM Images for 0.8 mm, 1.2 mm and medium torque W2400 Nm.

369

archives of civil and mechanical engineering 19 (2019) 360–374

Table 3 – Elements and their weight % (case, Sample 1).

Table 4 – Elements and their weight % (core, Sample 1).

Element

Element

Weight %

Atomic %

C O Al Cr Mn Fe

10.43 6.19 0.49 1.21 1.14 80.54

31.46 14.01 0.66 0.85 0.75 52.26

Total

100

100

3.5. Effect of load on fatigue cycles: sample 3 (medium torque W2400 Nm) From the fractography analysis (Fig. 12(e) and (f)), Periphery portion shows inter-granular fracture morphology and core region shows dimple morphology. The inter-granular morphology in periphery region is due to the presence of case depth which is having higher carbon content compared to the core region. The fatigue cycles were compared in transmission shafts at different torsional loads at room temperature with cyclic loads of 3 Hz. Shafts were subjected to fatigue test at high, medium

Weight %

Atomic %

C Al Cr Mn Fe

9.06 0.71 1.18 1.28 87.77

31.45 1.09 0.95 0.97 65.54

Total

100

100

and low torque loads 3000 Nm, 2400 Nm and 1700 Nm respectively. By virtue of design, applied torsional load is inversely proportional to fatigue performance. Normally at lower torque loads stress acting on the shaft surface will be comparatively less and hence there will be delayed crack nucleation and initiation resulting in improved fatigue performance. From the above study the average fatigue life of high, medium and low torque are 12,000 cycles, 140,000 cycles and 400,000 cycles respectively. Drastic increase in fatigue life is observed between medium and high torque loads which is mainly due to the difference in torsional loads applied on the shaft surface. The details are tabulated in Table 7.

Fig. 13 – (a) Sample 1 fractured, (b) EDS analysis in periphery region, (c) Sample 1 micro and (d) EDS analysis in core region.

370

archives of civil and mechanical engineering 19 (2019) 360–374

Table 5 – Elements and their weight % (case, Sample 2).

Table 6 – Elements and their weight % (core, Sample 2).

Element

Element

Weight %

Atomic %

C O Al Si Cr Mn Fe

10.33 6.3 0.44 0.24 1.18 1.19 80.32

31.15 14.26 0.59 0.31 0.82 0.78 52.08

Total

100

100

3.5.1.

Torsional fatigue load vs no. of cycle comparison

See Table 7.

3.5.2.

EDS Analysis for sample 3 (medium torque 2400 Nm)

See Tables 8 and 9 and Fig. 17.

3.6. Non-martensitic transformation products (NMTP) microstructure NMTP (Fig. 18) is formation of combination of bainite and pearlite at the surface of the carburised parts out of reduced hardenability during carburisation process. It is due to the depletion of alloying elements like Cr, Si and Mn which are lost due to oxidation during the carburising process resulting in

Weight %

Atomic %

C Al Cr Mn Fe

7.63 0.49 1.40 1.26 89.22

27.61 0.78 1.17 1.00 69.43

Total

100

100

reduced hardenability. During martensitic transformation, structures like bainite, pearlite, etc. forms NMTP structure. Typically, a carburising atmosphere consists of hydrogen, carbon monoxide, carbon dioxide, nitrogen, small quantities of moisture (H2O) and oxygen in ppm level. Depletion of alloying elements is caused due to reaction of oxygen with the metal surface. The reduction in hardenability of the surface layer means that the TTT curve of the surface layer shifts to the left and hence formation of non-martensitic transformation products instead of martensite. During gas carburising quenching process NMTPs are first to form on the shafts prior to martensitic transformation. During martensitic layer formation, there will be a volume growth and this volume growth induces residual tensile stress

Fig. 14 – (a) EDS analysis in periphery region, (b) Sample 2 fractured, (c) Sample 2 micro and (d) EDS analysis in core region.

371

archives of civil and mechanical engineering 19 (2019) 360–374

Table 8 – Elements and their weight % (case, Sample 3). Element

Weight %

Atomic %

C O Na Al Si K Cr Mn Fe

21.07 9.90 1.61 0.39 0.32 0.23 0.92 0.85 64.69

47.84 16.88 1.91 0.40 0.31 0.16 0.48 0.42 31.59

Total

100

100

Table 9 – Elements and their weight % (core, Sample 3).

Fig. 15 – XRD residual stress results.

Element

Fig. 16 – Retained austenite (RA) % vs fatigue cycles.

on the top layer of the NMTPs. This tensile stress drastically affects the fatigue performance of the shafts. Microstructure (Fig. 19) at surface (observed under optical microscope) containing non-martensitic transformation products (NMTP) which are prone to crack initiation and crack Table 7 – Load vs cycle results. Sample No.

Load (Nm)

Frequency (Hz)

Fatigue cycles

Fatigue tested at high torque 3000 1 2 3000 3000 3

3 3 3

10,876 12,492 11,189

Fatigue tested at medium torque 2400 1 2400 2 2400 3

3 3 3

119,126 147,266 189,663

Fatigue tested at low torque 1700 1 1700 2 1700 3

3 3 3

408,517 387,959 425,512

Weight %

Atomic %

C O Al Si Cr Mn Fe

7.59 2.62 0.39 0.27 1.28 1.16 86.70

26.14 6.77 0.59 0.39 1.01 0.87 64.21

Total

100

100

propagation. These NMTP are usually present on the surface and peculates to the core of the material. Increase in gas carburizing time affects compressive residual stresses on or near the surface of shafts, and this phenomenon can be explained by the formation of thicker case depth and non-martensitic transformation products (NMTP). X-ray diffraction result (Fig. 20) also shows that double tempered and surface modified shaft has remarkable difference in surface residual stress values. When compared to double tempered shaft, surface machined and polished shaft has more compressive residual stress at the surface. The surface residual stresses for the double tempered and machined shaft is more compressive in nature whereas double tempered shaft has tensile stress due to the formation of NMTP (Fig. 19). Compressive stresses are generated by the transformation of austenite to martensite during quenching, and therefore the highest compressive stresses observed on shaft surface [1,2]. This has the lowest retained austenite formed out of double tempering along with removal of NMTP out of machining and polishing operations. Hence removal of NMPT is carried out in machining operation with utmost care to the extent of 30–50 mm on the double tempered shaft surface.

4.

Conclusion

Steel grade 20MnCr5 materials used for transmission shaft application are attempted with various surface treatment processes which involved both heat treatment and machining operations. The findings of various experiments towards

372

archives of civil and mechanical engineering 19 (2019) 360–374

Fig. 17 – (a) EDS analysis in periphery region, (b) Sample 3 fractured, (c) Sample 3 micro and (d) EDS analysis in core region.

improving the fatigue life of the shaft are summarised as follows Experiment on various iterations on carbon case depth (0.8– 1.2 mm) and surface roughness (0.5–3.2 Ra) yielded no significant improvement on fatigue life(<15,000 cycles) of the shafts. Multiple attempts on tempering process were narrowed down to two stage tempering which means 'double tempering' where RA % has been reduced by 50% (from 10% to 5%). Through machining operation NMTP to the extent of about 30 mm on the

shaft surface were completely removed. This has further enhanced the fatigue life substantially (From 15000cycles to 25,000 cycles). Super finishing polishing operation has resulted in better surface finish on the double tempered shafts improves compressive residual stress up to 259 MPa (observed through XRD). Thus, it is concluded that best possible improvements in fatigue lives (from 12,000 cycles to 35,000 cycles 300%) of auto

Fig. 18 – NMTP.

Fig. 19 – Residual stress values.

archives of civil and mechanical engineering 19 (2019) 360–374

373

Fig. 20 – (a) XRD – residual stress value and (b) XRD residual stress analysis.

transmission shafts can be obtained by performing double tempering to reduce the retained austenite (<5%) percentage, followed by machining to remove the NMTP (Nil) and by improving the surface finish (<0.5 Ra) further by superfinishing polishing operation.

Acknowledgment The author would like to extend his appreciation to Sundram Fasteners Limited who has given approval to sponsor this research. Not forgotten, to the laboratory personnel and other Faculties of CECASE, National Institute of technology, Trichy and IIT-Madras for always being well-prepared to make the equipments and instrumentations ready to use for the research as and when needed.

references

[1] J. Schijve, Fatigue of Structures and Materials, Kluwer Academic Publisher, 2001. [2] H. Itoga, K. Tokaji, M. Nakajima, H. Ko, Int J Fatigue 25 (2003) 379–385. [3] D. Arola, C.L. Williams, Int J Fatigue 24 (2002) 923–930. [4] N.A. Alang, N.A. Razak, A.K. Miskam, Int J Eng Technol 11 (2011) 160–163. [5] G. Parrish, Carburizing: microstructures and properties, ASM International, Materials Park, OH, 1999. [6] S. Oda, T. Koide, M. Matsui, Y. Yamamoto, JSME Int J 32 (1989) 455. [7] T. Naito, H. Ueda, M. Kikuchi, Metall Trans A: Phys Metall Mater Sci 15A (1984) 1431. [8] Z.Z. Hu, M.L. Ma, Y.Q. Liu, J.H. Liu, Int J Fatigue 19 (1997) 641. [9] K. Genel, M. Demirkol, Int J Fatigue 21 (1999) 207–212. [10] A. Baji Dand Belai, International scientific conference on production engineering, Lumbarda, (2006) 109–115.

374

archives of civil and mechanical engineering 19 (2019) 360–374

[11] S. Cruchley, H.Y. Li, H.E. Evans, P. Bowen, D.J. Child, M.C. Hardy, Int J Fatigue 81 (2015) 265–274. [12] A. Javidi, U. Rieger, W. Eichlseder, Int J Fatigue 30 (2008) 2050– 2055. [13] A. Caetano Melado, A. Seiji Nishikawa, H.A. Goldenstein, A. Michael, A.S. Giles, P. Reed, Int J Fatigue 11 (2017). [14] P.I. Christodoulou, A.T. Kermanidis, D. Krizan, Int J Fatigue 91 (2016) 220–231. [15] R.M.N. Fleury, D. Nowell, Int J Fatigue 105 (2017) 27–33.

[16] L. Weng, J. Zhang, S. Kalnaus, M. Feng, Y. Jiang, Int J Fatigue 48 (2013) 156–164. [17] M. Widmark, A. Melander, Int J Fatigue 21 (1999) 309–327. [18] B. Jiang, S. Everitt, N. Gao, K. Soady, J.W. Brooks, P.A.S. Reed, Int J Fatigue 75 (2015) 89–99. [19] W. Gearv, J.E. King, Int J Fatigue 1 (1987) 11–16. [20] M.T. Yu, T.H. Topper, L. Wang, Int J Fatigue 4 (1988) 249–255. [21] L. Cheng, C.M. Brakman, B.M. Korevaar, E.J. Mittemeijer, Metall Trans A 19 (1988) 2415–2426.