Fatigue behavior of magnesium alloy and application in auto steering wheel frame

Fatigue behavior of magnesium alloy and application inauto steering wheel frameM AO Ping-li(毛萍莉), L IU Zheng(刘正), W ANG Chang-yi(王长义),G UO Quan-ying(郭全英), S UN Jin(孙晶), W ANG Feng(王峰), L IN Li(林立)School of Materials Science and Engineering, Shenyang University of Technology,Shenyang 110023, ChinaReceived 12 June 2008; accepted 5 September 2008Abstract: The low-cycle fatigue behaviors of AZ91HP-F, AZ91HP-T6, AZ91HP-T4 and AM50HP-F were investigated, and the potential application of AM50HP-F in steering wheel frame was studied. The steering wheel properties were characterized by bend fatigue and tensile testing, and the fatigue fracture was analyzed by SEM. The results show that the fatigue lives of AZ91HP-F and AZ91HP-T6 have little difference by comparing the low-cycles fatigue properties of different heat treatment states. The crack propagation velocity of AZ91HP-T4 is lower than that of AZ91HP-F and AZ91HP-T6. The die casting technological parameters of the magnesium steering wheel have been optimized with the aid of flow-3D software. The tensile testing results of the different part of magnesium steering wheel show that the ultimate tensile strength and elongation in the wheel arm and wheel rim have no difference and the average value are 220 MPa and 5%, respectively. The fracture is in the brittleness mode and the fatigue crack initiates at the outside of the wheel rim.Key words: magnesium alloy; low cycle fatigue behavior; magnesium steering wheel frame1 IntroductionMagnesium alloy components equal in strength are 40% lighter than steel and 20% lighter than their aluminum counterparts, which enables car manufacturers to meet regulatory requirements for lighter weight vehicles and a corresponding reduction in emissions. Among the magnesium alloys AZ91 and AM50 (or AM60) are the most widely used die-cast magnesium alloys for producing some automobile parts such as steering wheel frame, transmission case, housing, engine cradle, and pedal[1−2]. Since these mechanically loaded automobile components are often subjected to cyclic stresses, it is necessary to investigate the cyclic deformation behavior of those alloys. Previous work concerning fatigue behavior of the die-cast AZ91 and AM50 alloys have been focused on its high-cycle fatigue properties, and some helpful research results have been well documented[3−10]. However, few low-cycle fatigue data are available for these alloys[11−13]. Related investigations have revealed that the ductility of AZ91 can be greatly improved and the fatigue crack propagation rate reduced with a treatment that produces a supersaturated solution (T4)[14]. The S-N curve of AZ91 has shown that precipitation from supersaturated solutions results in little improvement in the fatigue strength although the yield stress may significantly increase (T6)[15]. It is known that the mechanical properties of die-casting alloy AM60 can be improved following a supersaturated solution and artificial aging treatment[15]. This work is concerned with the low-cyclic deformation behavior of the alloy AZ91HP in F, T4 and T6 conditions in comparison with the alloy AM50HP (F).2 ExperimentalThe testing materials are AZ91HP and AM50HP. The ingots were melted and cast into a plate of 150 mm×100 mm×5 mm with a cold chamber die-casting machine GDK200. The specimen was cut from the die-casting plate and its geometry is shown in Fig.1. The thickness of specimen is equal to that of die-casting plate and the surface of specimen was not machined after die casting, but polished with 400 and then 600 grit emery papers. Table 1 lists the parameters of high pressure die casting for AZ91HP and AM50HPFoundation item: Project (2007CB613705) supported by the National Basic Research Program of China Corresponding author: MAO Ping-li; Tel: +86-24-25497131; E-mail: pinglimao@MAO Ping-li, et al/Trans. Nonferrous Met. Soc. China 18(2008)s219Table 1 Parameters of high pressure die casting and heat treatmentDie casting parametersSupersaturated solutionparametersAging parametersCasting temperature/℃ Mouldtemperature/℃ Gate speed/(m·s −1)Crystal pressure/Pa Treatmenttemperature/℃Treatment time/h CoolingmediumTreatment temperature/℃ Treatment time/h Coolingmedium670−690 22080 7.0×107 415 8 Air160 10−56AirFig.1 Dimensions of fatigue specimen (unit: mm)and the subsequent heat treatments for AZ91HP .The fully reversed total strain-controlled low-cycle fatigue tests were performed with a servo hydraulic testing system MTS810 at room temperature in air. A triangular waveform with a strain rate ranging from 7.5×10−3 to 1×10−2and five strain levels ranging from 2.5×10−3−1.5×10−2were used in all tests. Lower strain rates were used for the larger strain amplitude tests and the higher strain rates for the smaller strain amplitude tests. The fatigue failure was defined as specimen fracture or 20% drop in maximum tensile load.3 Low-cycle fatigue behaviorThe fatigue life data of plastic, elastic and total strain amplitude vs. reversals to failure (2N f ) are shown in Figs.2 and 3. The plastic and elastic strain amplitudes and the corresponding stress amplitudes are obtained at a half of the cyclic life. The fatigue life data can be described in the following conventional equation and displayed very well in terms of Basquin and Manson- Coffin laws as a result of lg-lg linear regression of the elastic and plastic strain amplitudes.c b N N E )2()2(222f f f f p e t εσεεε′+′−=Δ+Δ=Δ (1)where E is elastic modulus, which is about 45 GPa for the magnesium alloys; σ′f and ε′f are fatigue strength and ductility coefficients respectively whose values have been expressed in Figs.2 and 3. By comparison of Fig.2 and Fig.3, it can be seen that there are almost no difference in the fatigue life between AZ91HP-F and -T6, but some changes have happened for AZ91HP-T4.The fatigue life for supersaturated solution treatment will be prolonged at high strain amplitude but reduced at low strain amplitude. AM50HP exhibits the similar characteristics to AZ91HP-T4, but the prolongingFig.2 Strain amplitude vs. reversals to failure (2N f ) ofAZ91HP-F and AZ91HP-T6Fig.3 Strain amplitude vs. reversals to failure (2N f ): (a) AZ91HP-T4; (b) AM50HP-FMAO Ping-li, et al/Trans. Nonferrous Met. Soc. China 18(2008) s220fatigue life at high strain amplitude shows in a little smaller degree than that of AZ91HP-T4, as shown in Fig.3. It is obvious that the transition fatigue life (NT) is elongated with solution treatment. Thereby, it will be significant that the attempt of some potential automobile application such as steering wheel would take AZ91HP-T4 as an alternative material of AM50HP for more safety.Fig.4 shows the cyclic stress—strain curves together with the monotonic stress—strain curves of AZ91HP-F and AM50HP-F. The strain amplitude (Δεt/2) and corresponding stress amplitude (Δσ/2) are obtained at half of the cyclic life (N f/2). It was found that the cyclic strength increased in an order of AM50HP, AZ91HP-T4, AZ91HP-F and AZ91HP-T6. By comparing with the monotonic stress—strain curve, all the cyclic configurations exhibit the cyclic strain hardening in different degrees depending on the composition, heat treatment and strain amplitudes. The cyclic strain hardening is accelerated with increasing the strain amplitudes. Solid solution alloy AZ91HP-T4 shows the highest and artificial aging alloy AZ91HP-T6 exhibits the lowest cyclic strain hardening of all the testing materials although AZ91HP-T6 has the highest cyclic strength and AZ91HP-T4 has much lower cyclic strength.Fig.4 Monotonic and cyclic stress—strain curve4 Fatigue crack propagationAt load ratio of 0.5, load frequency of 50 Hz and under constant load amplitude control in air, Δk was measured to be 0.774 MPa⋅m1/2 for artificial aging specimens, 0.997 MPa⋅m1/2 for die-casting specimens and 1.174 MPa⋅m1/2 for supersaturated solution treated specimens. The testing fatigue crack propagation rate of AZ91HP was decreased in the order of artificial aging, die casting and supersaturated solution treatment, as shown in Fig.6. Where the value of crack length a was measured with a microscope and the stress intensity factor range, Δk was calculated with the following formulas (2) and (3) according to ASTM standard E647:2/312.）（）（Δα−+=ΔαWBPk(0.886+4.64α-13.32α2+14.72α3−5.6α4)(2)11d/d−−−−=llllnnaana(3)where ΔP =P max−P min represents load range; B is the specimen thickness, and it is equal to 4 mm; W is the specimen width, and it is equal to 20 mm; α=a/W is the normalized crack length; n is the applied cycles and l is the number of cycles.Fig.5 Fatigue crack propagation rate at 50 Hz load frequencyFig.6 Fluid field simulation result5 Application of AM50 magnesium alloy on motor steering wheelWhen magnesium alloy was used in steering wheel frame, it could not only lower the weight, but also lower the vibration of road and control system and absorb more compact energy when there was a accident. More and more magnesium steering wheels were used in auto industry. The investigation and development of magnesium steering wheel for Jinbei coach was conducted in the present study.MAO Ping-li, et al/Trans. Nonferrous Met. Soc. China 18(2008) s221The software of Flow-3D was used to optimize the die-casting parameters such as filling velocity, mould temperature and pouring temperature. According to the simulation results of fluid field, temperature field and probability of the surface defect the three parameters of filling velocity, mould temperature and pouring temperature for the present steering wheel structure and die casting running system are 2.34 m/s, 220 ℃ and 700 ℃ respectively. Figs.6 and 7 show the examples of simulation results of fluid field and the probability of surface defect.Fig.7 Surface defect probability simulation resultAM50 steering wheel frame was manufactured using the parameters according to the simulation results. The tensile tests were conducted on the five locations of the steering wheel, one is in the wheel rim and the others are in the wheel arms. The test results show that the ultimate tensile strength and elongation of five locations have no difference, and the average values were 220 MPa and 5% respectively. It is implied that the magnesium steering wheel manufactured using the current parameters has homogenous microstructure. The fatigue testing of AM50 steering wheel frame was conducted in a universal steering wheel testing machine, when the steering wheel was fixed at centre and the wheel rim was applied with ±250 kN load repeatedly. The average fatigue life of AM50 HP-F steering wheel was 110 000 cycle, and the enterprise standard for AM50 HP-F steering wheel is 100 000 cycle. It was implied that AM50HP-F had the capability of meeting the requirement of fatigue life for steering wheel frame. The fatigue fracture morphology is shown in Fig.8. It can be seen from Fig.8(a) that the crack initiated from the outside of the wheel rim. When the outside of the wheel rim underwent the maximum tension and compression stress repeatedly, it was possible for the stress concentration at outside of the wheel rim, as indicated by arrow in Fig.8(a). The morphology of fatigue crack propagation zone is shown in Fig.8(b), and the instantaneous fracture zone shows the feature of transgranular and cleavage (Fig.8(c)).6 Conclusions1) AM50HP, AZ91HP-T4, AZ91HP-F and AZ91HP- T6 exhibit the cyclic strain hardening in different degrees depending on the heat treatment and strain amplitudes. AZ91HP-T4 shows the highest，AZ91HP-T6 the lowest cyclic strain hardening of all the testing materials.2) No difference of the fatigue life between die casting alloy AZ91HP-F and artificial aging alloy AZ91HP-T6 is observed. Solid solution alloy AZ91HP-T4 and die casting alloy AM50HP-F have longer fatigue life only at very high strain amplitude butFig.8 Fatigue fracture morphologiesof AM50 HP-F steering wheel frame:(a) Full view of fracture; (b) Crackpropagation zone; (c) Instantaneousfracture zoneMAO Ping-li, et al/Trans. Nonferrous Met. Soc. China 18(2008) s222much shorter at low strain amplitude than AZ91HP-F and -T6. Fatigue crack propagation rate is reduced by solution treatment.3) The mechanical and fatigue properties of AM50HP-F were tested and it was found that AM50HP-F was suit for manufacturing magnesium steering wheel frame.References[1] FRIEDRICH H, SCHUMANN S. Research for a “new age ofmagnesium” in the automotive industry [J]. J Mater Proc Technol,2001, 117: 276−281[2] AGHION E, BRONFIN B, ELILEZER D. The role of themagnesium industry in protecting the environment [J]. J Mater ProcTechnol, 2001, 117: 381−385.[3] ZENNER H, RENNER F. Cyclic material behaviour of magnesiumdie castings and extrusions [J]. Int J Fatigue, 2002, 24: 1255−1260. [4] MAYER H, PAPAKYRIACOU M, ZETTL B. Influence of porosityon the fatigue limit of die cast magnesium and aluminium alloys [J].Int J Fatigue, 2003, 25: 245−256.[5] HORSTEMEYER M F, YANG N, GALL K, MCDOWELL D L.High cycle fatigue of a die cast AZ91E-T4 magnesium alloy [J]. ActaMater, 2004, 52: 1327−1336.[6] EISENMEIER G, HOLZWARTH B, HÖPPEL H W, MUGHRABI H.Cyclic deformation and fatigue behaviour of the magnesium alloyAZ91 [J]. Mater Sci Eng, 2001, A319-321: 578−582.[7] YANG Y, LIU Y B, YANG X H. High cycle fatigue properties ofAZ91D die cast magnesium alloy [J]. Special Casting & NonferrousAlloys, 2006, 26(2): 105−108 [8] HORSTERMEYER M F, YANG N, GALL K, MCDOWELL D, FANJ, GULLETT P. High cycle fatigue mechanisms in a cast AM60Bmagnesium alloy [J]. Fatigue Fract Eng Mater Struct, 2002, 25:1045−1052.[9] EISENMEIER G, HOLZWARTH B. Cyclic deformation and fatiguebehavior of the magnesium alloy AZ91 [J]. Materials Science andEngineering, 2001, A319: 578−582.[10] WANG X S, LU X, WANG D H. Investigation of surface fatiguemicrocrack growth behavior of cast Mg-Al alloy [J]. MaterialsScience and Engineering, 2004, A364: 11−16.[11] LAMARK T T, CHEMLIK F, YURI ESTRIN Y. Cyclic deformationof a magnesium alloy investigated by the acoustic emissiontechnique [J]. Journal of Alloys and Compounds, 2004, 378:202−206.[12] LIU Z, WANG Z G, WANG Y. Cyclic deformation behavior andfatigue crack propagation in AZ91HP and AM50HP[J]. MaterialsScience and Technology, 2001, 17(3): 264−268.[13] GALL K, BIALLAS G, MAIER H J, GULLETT I,HORSTEMEYER M F, MCDOWELL D L, FAN J H. In-situobservations of high cycle fatigue mechanism in cast AM60Bmagnesium in vacuum and water vapor environmental [J].International Journal of Fatigue, 2004, 26: 59−70.[14] WOLF B, FLECK C, EIFLER D. Characterization of the fatiguebehavior of the magnesium alloy AZ91D by means of mechanicalhysteresis and temperature measurements [J]. International Journal ofFatigue, 2004, 26: 1357−1363.[15] LEE S G, PATEL G R, GOKHALE A M. Inverse surfacemacro-segregation in high-pressure die-cast AM60 magnesium alloyand its effects on fatigue behavior [J]. Scipta Materialia, 2005, 52:1063−1068.(Edited by YANG You-ping)。