Accepted Manuscript Lightweight of a cross beam for commercial vehicles: Development, testing and validation S. Cecchel, D. Ferrario, A. Panvini, G. Cornacchia PII: DOI: Reference: S0264-1275(18)30293-4 doi:10.1016/j.matdes.2018.04.021 JMADE 3834 To appear in: Materials & Design Received date: Revised date: Accepted date: 26 January 2018 27 March 2018 9 April 2018 Please cite this article as: S. Cecchel, D. Ferrario, A. Panvini, G. Cornacchia , Lightweight of a cross beam for commercial vehicles: Development, testing and validation. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Jmade(2017), doi:10.1016/j.matdes.2018.04.021 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ACCEPTED MANUSCRIPT Lightweight of a cross beam for commercial vehicles: development, testing and validation EP T ED MA NU SC RI PT ABSTRACT characterisation. AC C KEYWORDS: non-ferrous alloys; aluminium; EN AC-43500; HPDC; road simulator test bench; mechanical 1 ACCEPTED MANUSCRIPT EP T ED MA NU SC RI PT INTRODUCTION AC C 1. 2 ACCEPTED MANUSCRIPT RI PT This material guarantees high nominal mechanical properties even at an as-cast state [46] and has good corrosion resistance [44], which allows the avoidance of additional protective treatments currently used on the traditional steel structure, being therefore environmentally friendly and economically affordable. AC C EP T ED MA NU SC The advantages achieved could be summarised in weight reduction of about 50%, avoidance of painting operations, thanks to the good corrosion resistance of the alloy selected, and excellent recyclability. These benefits entail the achievement of an economically affordable solution. At the same time, the use of appropriate materials, a new concept of design and a careful function integration would allow the structural limits explained in the introduction to be overcome. [49-51]. [50]. 3 ACCEPTED MANUSCRIPT 2. METHOD 2.1 Description of structure and material NU SC RI PT In Figure 1 it is possible to observe the component, which has overall dimensions of about 1260x450x4 mm and weighs about 15 kg. MA Figure 1: HPDC aluminium suspension cross beam. ED The selected material is EN AC-43500 alloy and its chemical composition as measured by optical emission spectrometry is listed in Table 1. Table 1: Chemical composition of EN AC-43500. Si 11.04 Fe Cu Mn Mg Ti 0.11 0.003 0.56 0.17 0.06 EP T % To produce the suspension cross beam prototypes, an ITALPRESSE TF cold chamber machine was used with a maximum locking force of 3000 t equipped with Fondarex vacuum.  AC C 2.2 Experimental setups Microstructure A microstructural characterisation was conducted on samples machined out of the component (Fig. 2). Different zones were analysed in order to investigate the potential dissimilarity due to the different solidification rate of the metal into the die. In addition, the microstructure was analysed both near to the surface (skin layer) and into the bulk of the casting for all the samples. For microstructure characterisation, specimens were wet ground through successive grades of SiC abrasive papers from P120 to P4000, followed by diamond finishing to 1 μm. The samples were examined using Optical Microscopy Leica DMI 5000M and Scanning Electron Microscopy (SEM) LEO EVO 40. Semi-quantitative chemical analyses were obtained by means of an EDS (Energy Dispersive Spectroscopy–Link Analytical eXL) probe, with a spatial resolution of a few microns. 4 ACCEPTED MANUSCRIPT  PT Figure 2: Position of samples machined out of the component for the microstructural analysis. Tensile tests, hardness tests and heat treatments SC RI Tensile tests were conducted according to UNI EN ISO 6892-1:2009 [52] with an electromechanical INSTRON 3369 testing machine at a crosshead speed of 2 mm/min. The strain was measured using a 25 mm extensometer. Flat tensile samples with length, width and thickness of about 100, 12 and 4 mm were used. Experimental data were collected and elaborated to provide yield strength (Rp0,2), ultimate tensile strength (Rm) and elongation to fracture (A%). At least four measurements were taken, and the average value was considered for each condition. NU The hardness tests were performed on the shoulders of the tensile specimens prior to the tests, following ASTM E 18-03 procedures. Vickers hardness HV60 tests were performed on a hardness tester GALILEO ERGOTEST COMP 25, with a 60 kg load and a dwell time of 15 s. At least four measurements were taken, and the average value was considered for each condition. After the tensile test, fracture surface observations were conducted using Scanning Electron Microscopy LEO EVO 40 in secondary electron (SE). AC C EP T ED MA Design Of Experiment (DOE) and ANOVA methodologies were applied to assess the influence of the time and temperature of heat treatment conditions on the final alloy’s mechanical properties. These analyses were conducted on samples separately cast from the component (Fig. 3). The statistical methodology used during this work to investigate and model the relationship between the factor and the response is the factorial analysis of variance (ANOVA) 2 3 (2 levels, 3 factors). The software Minitab was used for these studies. Next, an additional analysis with an introduction of centre point was performed. The centre point is used to investigate whether the model between the response and the factors is linear. The centre point replicates are treated as an additional factor in the model. In particular, the factors selected for the analysis are solubilisation time (tsolubilization), aging temperature (T aging) and aging time (taging). The solubilisation temperature was fixed at 490°C. For T5 heat treatments, tsolubilization=0 h imposed in the DOE matrix was considered. Specimens were solutionised in a preheated electric oven, immediately followed by water quenching at room temperature. Next, they were immediately refrigerated at a temperature of about −18 °C. The time lapse between quenching and refrigeration did not exceed 5 minutes to avoid alteration to the mechanical properties of the alloy due to intermetallics precipitation during natural aging. Afterwards, they were artificially aged in a preheated oven. The experimental matrix composed of control factors with different levels considered is reported in Table 2. Table 2: Control factors for each experimental run. Run tsolubilization (h) Taging (°C) taging (h) P01 0 150 1 P02 0 150 8 P03 0 190 1 P04 0 190 8 P05 5 150 1 P06 5 150 8 P07 5 190 1 P08 5 190 8 P09 2.5 170 4.5 5 ACCEPTED MANUSCRIPT The response variables considered were Vickers hardness (HV60), yield strength (R p0,2), ultimate tensile strength (Rm), elongation to fracture (A%) and the Quality Index (QI) calculated according to the following equation [53]: 𝑄I= Rp0,2+210 log(A%)+13 Four repetitions were performed for each condition. The results obtained can be a useful database for the selection of the proper heat treatments for the designing and production of a specific automotive component. Optimisation of the elongation or a maximum yield strength could be required depending on the application. In particular, for the HPDC cross beam object of this study, a T5 heat treatment was selected after this activity and performed on some components. More details about this choice will be shown in the results section. AC C EP T ED MA NU SC RI PT Usually, quality tests during component production are based on tensile tests of samples separately cast, which require an easier and less expensive setup than samples machined out of the casting. In this context, during this research, the mechanical characterisation of the as-cast component was conducted both on samples machined out of the components and on samples separately cast from the die casting (Fig. 3) in order to check the reliability of the latter specimens in comparison to the real properties of the component. Samples were machined out of 5 cross beams in different zones, highlighted in the bottom side of Fig. 3, in order to evaluate potential anisotropy due to the different solidification of the metal into the die. In detail, from each component 4 “A” and “B” and 2 “M” samples were machined in symmetric positions. It can be noted that samples “B” present two ribs on the grip section, which were machined in order to obtain specimens with constant thickness. A representative picture of the pre- and post-test coupons is reported in the bottom right side of Fig. 3. Figure 3: Samples separately cast from the component (top) and samples machined out of the component (bottom). In addition, a T5 heat treatment at 170°C for 4 hours (selected after DOE activity) was conducted in an industrial furnace at the same time on 3 components and 6 samples separately cast. After that, samples were machined out of these T5 heat treated components in the same way explained before. Finally, the same T5 heat treatment was reproduced in a laboratory preheated electric cast oven, on other samples separately, in order to confirm the reliability of this laboratory scale method.  Salt spray test A salt spray corrosion test was performed according to ISO 9227 Standard [54]. In order to evaluate potential galvanic corrosion effects during the test, the cross beam was previously assembled with all the elements usually connected to this component in the vehicle. A sodium chloride solution with a concentration of 50 g/l was sprayed through a series of nozzles. The temperature inside the spray cabinet was maintained at 35 ± 2 °C while the maximum exposure time was 500 hours. During the test, the cabinet was stopped at 50 h, 100 h, 215 h, 330 h, 408 h and 500h and the assembly was given a visual inspection, in order to estimate the development of the corrosion phenomenon. At the end of the experiment (after 500 h), the assembly was extracted from the cabinet and analysed. In particular, a detailed visual inspection of the component and its interfaces with other elements and a fatigue test bench road simulator of the corroded component was 6 ACCEPTED MANUSCRIPT done. It is worthwhile to note that the resistance to corrosion is also a very important requirement in the production of safety relevant component.  Fatigue test bench road simulator SC RI PT During this research, the fatigue behaviour of the cross beam axle was investigated on a MTS329 4DOF road simulator bench test available at the Streparava Testing Centre. First of all, in order to perform the test, it was necessary to assemble a physical prototype of the cross beam with a full suspension model to carry out the functional tests (Fig. 4). Figure 4: Full suspension cross beam assembly. AC C EP T ED MA NU Then, it was necessary to fix this assembly to a segment of the chassis frame in order to replicate the boundary conditions and the constraints found on the vehicle. The link between the actuators and the wheel hub was made through a flange with the same wheel force transducer used in field data acquisition. Fig. 5 shows the full suspension assembly mounted on road simulator test bench. Figure 5: Full suspension assembly on road simulator test bench. During the test, the shock absorbers were water-cooled by means of a circuit linked with a refrigerator that kept the water at a constant temperature between 18°C and 20°C. Rubber parts (bushings and bumpers) were cooled through a compressed air system. When required by the program, an oil-pneumatic unit equipped with an electric operation activated the brake. The vertical actuators were displacement-controlled while the longitudinal and transversal ones were loadcontrolled. A displacement limit of 160 mm, which is the maximum distance feasible for this type of suspension, was imposed for the vertical actuators. The road data time histories that have to be replicated for the endurance test were available thanks to a tracking activity previously made on the same class of vehicles under study. The mission performed for the component analysed was a distance of 250’000 km, divided into four typical test paths (highway, urban, hill and rough terrain) with the addition of special racetrack events (steering and brakes) and impulsive single load events. The 250’000 km of the vehicle mission were divided into: 30% highway (75’000 km, 2’679 cycles), 30% hill (75’000 km, 2’953 cycles), 25% urban (50’000 km, 4’546 cycles), 25% rough terrain (50’000 km, 7’143 cycles) and 1 special event 7 ACCEPTED MANUSCRIPT every 1’000 km (total 250 steering events and 250 brake events). These technical specifications are representative of a typical LCV lifespan and are usually employed in the testing of similar suspension assemblies. The test duration was about 650 hours. The corresponding vehicle velocity was approximately 400 km/h, with 99% of real fatigue damage applied at the cross beam. At the end of the fatigue test bench road simulator the component was analysed with liquid penetrant testing. RESULTS AND DISCUSSION  Microstructure ED MA NU SC RI PT An example of the optical micrographs analysed conducted is shown in Figure 6. Figure 6: Example of Optical and SEM micrographs of sample machined out of the cross beam.  AC C EP T The analyses on the as-cast samples show for all the samples a typical microstructure composed of α-Al matrix with AlSi globular modified eutectic in the interdendritic space and some intermetallic compounds. Precipitates in EN AC-43500 are composed of α-Al15(Mn,Fe)3Si2 polyhedral structure. Two sizes of α compound were observed; the larger are probably primary phases which have formed in the shot sleeve, while the smaller precipitated after the injection in the die [55]. Generally, a finer microstructure and a higher amount of silicon were observed in the pictures collected near the surface of the samples. This effect is pronounced in the specimens machined from high thickness zones (A, B and C samples of Fig. 2), whereas it is very slight in the thinner samples (D and E samples of Fig. 2). These observations confirm the presence of a major or minor amount of Al-Si eutectic in the surface than in the bulk, depending on specific solidification conditions of the analysed region [56]. Porosities found in the samples analysed were very small and of negligible amount. This is in accordance with a previous RX analysis conducted [37], which did not detect porosity in the sampling area examined. No other defect was observed during this analysis. Tensile tests, hardness tests and heat treatments Table 3 collects the average values and the standard deviation of the response variables for each experimental combinations set up through the DOE method and conducted on samples separately cast from the component. For comparison purposes, Figure 7 reports a tensile curve for each of the run settings selected in order to guarantee a more readable format. Table 3: Response variables results for each experimental combination and comparison with as cast condition. Run HV Rp02 [MPa] Rm [MPa] A(%) QI As-cast 87 ± 1 136 ± 7 263 ± 5 5.2 ± 0.9 298 ± 13 P01 87 ± 1 141 ± 2 249 ± 6 3.3 ± 0.4 261 ± 13 P02 97 ± 1 170 ± 3 254 ± 9 4.1 ± 0.9 309 ± 19 8 ACCEPTED MANUSCRIPT 96 ± 2 171 ± 3 264 ±10 4.0 ± 0.8 309 ± 17 P04 96 ± 1 173 ± 4 261 ± 3 4.3 ± 1.7 313 ± 44 P05 59 ± 0 94 ± 4 189 ± 7 12.4 ± 3.7 333 ± 26 P06 70 ± 5 154 ± 11 229 ± 19 8.6 ± 1.1 324 ± 8 P07 58 ±0 99 ± 9 176 ± 3 11.3 ± 1.9 338 ± 18 P08 68 ± 3 155 ± 5 205 ± 11 5.7 ± 1.6 334 ± 36 P09 82 ±4 164 ± 8 229 ± 12 7.0 ± 5.0 363 ± 57 MA NU SC RI PT P03 Figure 7: Tensile curves for each experimental combination and comparison with as cast condition. AC C EP T ED In general, it can be stated that T5 heat treatment (from P01 to P04) leads to an increase of Rp0,2 up to 29% with reduction of A% up to 50% and strength almost unchanged. On the other hand, heat treatments with an artificial solubilization (from P05 to P09) cause an improvement of A% up to 138% with variation of yield strength between -30% and 20% depending on the parameters used and slight reduction of Rm up to 28%. Regarding Quality Index, T6 leads to an improvement of this parameter for each condition analysed while T5 bears not significant effects. The central point P09 (490 ° C - 2.5 h + 170°C - 4.5 h) gives the best condition of the QI. The subsequent analysis of the data by means of ANOVA 2 3 reveals a complete and more exhaustive evaluation of the relationship between the mechanical properties of the tensile samples and the independent variables. The Pareto chart, the main effect plot and the contour plot calculated for the yield strength, reported here as example, are shown in Figure 8. The same analysis was conducted for all the response variables. 9 SC RI PT ACCEPTED MANUSCRIPT Figure 8: Pareto Chart of the standardised effects, main effect plot and contour plot for Rp0,2. EP T ED MA NU The Pareto chart evaluates the possible interactions among the variables and highlights the factors that have statistically significant relationships, located at the right side of the red line in the graph. As expected, for each variable the most important parameter is the solubilisation time that identifies the presence or absence of artificial solubilisation. The second important parameter is aging time and the third is the interaction between solubilisation and aging time. This last relationship is considered for each property, while for the elongation a switch of the second and the third factor is observed. Aging temperature is relevant only for yield strength. The main effects charts display the influence of solubilisation and aging time and temperature on the studied properties (Rp0,2, Rm, A%, QI, HV), while the contour plots analyse the effect of the interaction of each parameter on the same properties. These last graphs are useful to make comparisons between two different data sets by taking each set of variables one by one. From the analysis of these graphs it was confirmed that heat treatment with artificial solubilisation determines an increase of A% and a decrease of Rp0,2, Rm and HV; higher aging temperature leads to an increase of Rp0,2 and HV, decrease of A% and to a Rm and a QI almost unchanged; higher aging time leads to an increase of all the properties except for A%, which decreases with time. In the range of data under consideration, the properties analysed can be described according to the following regression models (R2=0.95): Rp0,2= (8.83+11.40 ts + 0.85 Tag + 18.73 tag -0.14 tsTag-1.64 ts tag -0.10 Tag tag+0.02 tsTagtag) Rm=(187.52+7.80 ts + 0.41 Tag + 5.48 tag -0.14 tsTag-1.16 ts tag -0.03 Tag tag+0.001 tsTagtag) AC C A%=(0.13+3.16 ts + 0.02 Tag + 0.40 tag -0.008 tsTag+0.01 ts tag -0.002 Tag tag+0.001 tsTagtag) QI=(52.19+53.03 ts + 1.35 Tag + 30.69 tag -0.25 tsTag-1.32 ts tag -0.16 Tag tag+0.005 tsTagtag) HV=(48.06+3.11 ts + 0.25 Tag + 6.12 tag -0.06 tsTag-0.91 ts tag -0.03 Tag tag+0.006 tsTagtag) In addition, the centre point was used to investigate whether the model between the response and the factors is linear. The additional tests performed with the centre point were analysed in Minitab and the surface plot of the yield strength is reported in Fig. 9 as example. 10 ACCEPTED MANUSCRIPT Figure 9: Surface plot for Rp0,2 The analysis of the graphs highlights the non-linear behaviour for each parameter analysed. The effect of the different factor on the variation of properties studied is the same already observed before. PT After this examination, a T5 heat treatment was selected and performed on 5 HPDC cross beams in order to evaluate a slight increase of the resistance for heavier applications. To achieve this aim the output requested was a small increase of the yield strength (about 30 MPa), without any significant changes on stress at break and elongation. In addition, the heat treatment selected had to guarantee a release of the residual stress that reduces the deformations. For this last purpose, the temperature selected had to be rather low and holding time not too short. Following these considerations and the regression model of Rp0,2, the T5 heat treatment selected is 170°C for 4 hours. Indeed, based on the equations previously reported, the expected properties after this heat treatment are Rp0,2 160 = MPa, Rm = 259 MPa and A% = 3.8%. RI Table 4 shows the results of the tensile tests on samples machined out of the components in different positions (as-cast and after T5 heat treatment in industrial and laboratory oven) and samples separately cast from the component (as-cast and after T5 heat treatment in industrial and laboratory oven) for comparison purposes. Table 4: Tensile test on sample machined out of the component and cast into the die. T5 (170°C for 4h) Machined out of component A(%) B 134 5 266 20 4.7 1.8 118 4 242 12 4.4 1.0 117 3 234 25 6.4 3.8 134 7 258 5 5 0.9 Machined out of component A NU M MA Rp02 [MPa] Rm [MPa] A Cast into the die SC AS-CAST 182 2 288 17 3.6 1.6 Cast into the die M B 157 4 253 5 2.8 0.5 152 3 248 21 3.5 2 industrial laboratory oven oven 172 173 2 3 267 250 13 10 2.6 2.6 2 1 AC C EP T ED The T5 heat treatment led to a 30% to 35% improvement in yield strength with consequent 25% to 55% decrease in elongation. Correspondences were observed between the test results on samples treated into the laboratory oven and the industrial furnace. The experimental properties are similar to the values expected by ANOVA regression model. Indeed, the use of a linear model introduced only slight dissimilarities. In addition, the results determined that the strength of the samples is affected by the machining position. In particular, position “A” showed the highest mechanical resistance with a difference of about 20 MPa in comparison to the samples extracted from the other zones selected. This could be due to the position of samples “B” and “M” being located near to high thickness zones, which implies a longer solidification time. On the other hand, sample “A” is distant from the thickest areas of the component and hasfaster solidification. Indeed, its properties are very similar to those of separately cast samples. These results pointed out that the samples separately cast into the die are representative of the global behaviour of the component when they are properly designed, especially as pertains to their position into the mould. Finally, SEM analyses on the fracture surfaces of the tensile samples are shown in Figures 10 and 11. The following conditions were selected for the analysis of the specimen separately cast from the component (Fig. 10): as cast, T5 with maximum QI (190°C 1 h), T6 with maximum QI (490°C 2.5 h + 170°C 4.5 h), T6 with maximum A% (490°C 5h+ 150°C 1 h). The samples machined out from the component were analysed in all 3 different positions (A, M, B) both for as-cast and T5 conditions (Fig. 11 a and b). In general, it is possible to confirm the good ductility of the samples tested. Many of both types of examined samples exhibit ductile fracture, as there is a great number of small dimples indicating this fracture mode, as shown in Figures 10 and 11. However, there are still some differences between them. The eutectic Si-particles have an active role in crack propagation, when the morphology and size of Fe bearing phases are controlled, through the heat treatment conditions, and the porosity level is low. The fractographs show that the fracture surfaces especially of the tensile samples separately cast from the component are relatively flat in the skin region, but coarse and uneven in the central region; instead there are two principal fracture planes in the same surfaces of the specimens extracted from the components. However, there is no significant difference on the fracture morphology in the regions. In the fractographs, porosities, cold flakes and cold shuts are seen as the main defects observed. When cold shuts, cold flakes and intermetallics occur, they can cause laminated fracture or cracks. In particular, shrinkage porosities were observed especially on the fracture surface of the samples separately cast from the component (Figure 10), whereas, some defects, such as cold flake and cold shut, were found mainly on the fracture surfaces of specimens extracted from the components (Figure 11 a and b). It’s worthwhile 11 ACCEPTED MANUSCRIPT AC C EP T ED MA NU SC RI PT to note that cold flakes are formed when molten metal is poured into the shot sleeve and skin of solidified aluminium is pushed by the plunger into the die cavity during the filling. These small broken parts inside the castings are generally covered by an oxide layer contaminated with lubricants [57]. Cold shuts occur when molten metal flow comes into contact with the cooler die surface and solidifies before complete filling of the mould. Finally, Al-Si-Mn intermetallic phase was observed, particularly on the fracture surface of samples in as-cast and T5 conditions, which confirmed the microstructural analysis. These particles are present on the fracture surface since they are brittle and therefore this is the preferred crack propagation site. The fracture surface of these samples displays some tearing ridges, which are typical characteristics of transgranular fracture, but a great number of small dimples is also observed indicating a more ductile fracture mode. Therefore, these specimens exhibit a combination of brittle and ductile fracture, namely mixed fracture. Figure 10: SEM on the fracture surfaces of the tensile samples separately cast from the component. In the fractographs, it is possible to observe the principal casting defect identified as A, shrinkage porosity. 12 AC C EP T ED MA NU SC RI PT ACCEPTED MANUSCRIPT Figure 11: SEM on the fracture surfaces of the tensile samples machined out from: a) as-cast components, b) T5 components (b). In the fractographs, it is possible to observe some casting defects identified as: A shrinkage porosity, B cold flake, C intermetallics, D cold shut. 13 ACCEPTED MANUSCRIPT  Salt spray test NU SC RI PT After 500 h of testing, the cross beam suspension (EN AC-43500) had only a small amount of oxide layer on the surface and showed an aspect similar to its original condition, as it can be noted in Fig. 12, which represents the after test cross beam during the fatigue test bench road simulator. Fig. 12 also shows a representative detail of the component surface after the salt spray test. It is worth noting that under the oxide there are only a few shallowed pits. Figure 12: Fatigue test bench of the suspension cross beam after salt spray test.  Fatigue test bench road simulator MA This evidence confirms the previous analyses conducted on the corrosion resistance of this alloy. The results also excluded the possibility of galvanic corrosion at the interface. AC C EP T ED A representative image of the component subjected to the fatigue test bench road simulator after the analysis with liquid penetrant testing is shown in Fig.13. Figure 13: Example of penetrant testing result. The absence of cracks after the test validates the fatigue life calculation and indicates that this modified cross beam structure of an independent front suspension can endure at 250’000 km road mission, applied without fatigue failure. The same test was successfully completed also for the cross beam that was previously subjected to the salt spray test for 500 h, which confirms the high resistance to corrosion of the component. CONCLUSIONS Lightweighting is a method applied in the automotive industry that concerns the production of less heavy vehicles in order to achieve better fuel efficiency and performances. Weight reduction currently is commonly applied to car production, while in the field of Commercial Vehicles it is still rarely employed, mainly due to structural limits, linked to the lower stiffness and resistance of light alloys, and costs restraints. This research aimed at overcoming these current limits with the development of an aluminium cross beam suspension for Commercial Vehicles as a replacement of the conventional structural steel production. The component has relevant dimensions of about 1260x450 mm and was manufactured with a high locking force (3000 t) HPDC machine equipped with vacuum in aluminium EN-AC 43500. 14 ACCEPTED MANUSCRIPT The analyses showed:   The salt spray corrosion test on the cross beam assembled with all the elements usually connected to this component into the vehicle confirmed the following:    NU  T5 heat treatment leads to an increase of Rp0,2 up to 29% with a reduction of A% up to 50% and almost unchanged strength. Heat treatments with an artificial solubilisation cause an improvement of A% up to 138% with a variation of yield strength between -30% and 20% depending on the parameters used and a slight reduction of Rm up to 28%. The central point (490 ° C - 2.5 h + 170°C - 4.5 h) gives the best condition of the QI. The output of these experiments constitutes a useful database of the properties of EN AC-43500 alloy in the range of data considered. It can be used for the designing and production of many different automotive components. Either an optimisation of the elongation or a maximum yield strength could be required in this field depending on the specific application. A T5 heat treatment at 170°C for 4 h was selected for the specific application under study. MA  SC The following were the outputs of Design Of Experiment (DOE) and ANOVA applied to assess the influence of the duration and temperature of heat treatment conditions on the final alloy’s mechanical properties: ED  RI  Typical microstructure composed of α-Al matrix with Al-Si globular modified eutectic in the interdendritic space and α-Al15(Mn,Fe)3Si2 polyhedral structure intermetallics. A finer microstructure and a higher amount of silicon were near the surface of the samples, mainly in the specimens machined from high thickness zones. PT  A good corrosion resistance of the EN AC-43500 cross beam suspension Absence of galvanic corrosion at the interface. EP T  The results of the fatigue behaviour of the cross beam axle investigated on a 4DOF road simulator bench test were as follows: Absence of cracks after the test Demonstration that the cross beam structure of the independent front suspension can endure 250’000 km road application without fatigue failure both in the un-corroded state and after a 500 h salt spray test. AC C   In conclusion, this study and its detailed validation and test activities demonstrate the feasibility of light alloy use for this particular heavy application, achieved through careful material selection, appropriate design and accurate integration of functions. It is important to highlight that the use of advanced technologies (vacuum, high tonnage machine, primary alloys) linked with the removal of some features found in traditional steel production (painting, additional weight, etc.) led to reaching the cost and performance targets essential for project success. Moreover, the weight reduction obtained (47%) in this range of vehicles corresponds to a concrete value (from 1 to 5 €/kg) which guarantees the project’s competitiveness not only technically but also economically. ACKNOWLEDGEMENTS The Authors are grateful to Streparava SpA for their support in the fulfilment of salt spray test and fatigue test bench road simulator. 15 ACCEPTED MANUSCRIPT REFERENCES 1. F. Arena, L. Mezzana, The automotive CO2 emissions challenge, Automotive manufacturing group, (Arthur D. Little, 2014), http://www.adlittle.com/downloads/tx_adlreports/ADL_AMG_2014_Automotive_CO2_Emissions_ Challenge.pdf. 2. European commission, Proposal for a regulation of the European Parliament and of the Council PT amending Regulation (EC) No 443/2009 to define the modalities for reaching the 2020 target to reduce CO2 emissions from new passenger cars and Proposal for a regulation of the European Parliament and of the Council amending Regulation (EU) No 510/2011 to define the modalities for reaching the 2020 RI target to reduce CO2 emissions from new light commercial vehicles, (European Commission, 2012), 77319c424cbd.0001.02/DOC_1&format=PDF SC http://eur-lex.europa.eu/resource.html?uri=cellar:70f46993-3c49-4b61-ba2f- 3. R. Heuss et al., Lightweight, heavy impact. (McKinsey & Company, 2012). NU 4. H. Helms, U. Lambrecht, The potential contribution of light-weighting to reduce transport energy consumption. Int J Life Cycle Assess, 12, p.58–64, (2007). MA 5. R. Modaresi, S. Pauliuk, A.N. Lovik, D.B. Muller, Global carbon benefits of material substitution in passenger cars until 2050 and the impact on the steel and aluminum industries. Environ Sci Technol, 48, p.10776–10784, (2014). ED 6. H.C. Kim, T.J. Wallington, Life-Cycle Energy and Greenhouse Gas Emission Bene fits of Lightweighting in Automobiles: Review and Harmonization. Environ Sci Technol, 47, p. 6089–6097, (2013). EP T 7. H.J. Kim, G. Keoleian, S.J. Skerlos, Very Good-Economic Assessment of Greenhouse Gas Emissions Reduction by Vehicle Lightweighting Using Aluminum and High-Strength Steel. J Ind Ecol, 15, p. 64– 80, (2011). AC C 8. H.J. Kim, C. McMillan, G.A. Keoleian, et al., Greenhouse gas emissions payback for lightweighted vehicles using aluminum and high-strength steel. J Ind Ecol, 14, p. 929–946, (2010). 9. G.A. Keoleian, J.L. Sullivan, Materials challenges and opportunities for enhancing the sustainability of automobiles. MRS Bull, 37, p. 365–373, (2012). 10. P. Puri, P. Compston, V Pantano, Life cycle assessment of Australian automotive door skins. Int J Life Cycle Assess, 14, p. 420–428 (2009). 11. S. Das, Life Cycle Energy and Environmental Assessment of Aluminum-Intensive Vehicle Design. SAE Int J Mater Manf, 7, p. 588–595, (2014). 12. M. Bertram, K. Buxmann, P. Furrer, Analysis of greenhouse gas emissions related to aluminium transport applications. Int J Life Cycle Assess, 14, p. 62–69, (2009). 13. H.J. Kim, C. McMillan, G.A. Keoleian, et al., Greenhouse gas emissions payback for lightweighted vehicles using aluminum and high-strength steel. J Ind Ecol, 14, p. 929–946, (2010). 16 ACCEPTED MANUSCRIPT 14. M. Hakamada, T. Furuta, Y. Chino, et al. Life cycle inventory study on magnesium alloy substitution in vehicles. Energy, 32, p. 1352–1360, (2007). 15. J. Du, W. Han, Y. Peng, Life cycle greenhouse gases, energy and cost assessment of automobiles using magnesium from Chinese Pidgeon process. J Clean Prod, 18, p. 112–119, (2010). 16. C.S. Boland, R. De Kleine, G.A. Keoleian, et al. Life Cycle Impacts of Natural Fiber Composites for Automotive Applications: Effects of Renewable Energy Content and Lightweighting. J Ind Ecol, 20, p. 179–189, (2015). PT 17. J.C. Kelly, J.L. Sullivan, A. Burnham, et al. Impacts of Vehicle Weight Reduction via Material Substitution on Life-Cycle Greenhouse Gas Emissions. Environ Sci Technol, 49, p. 12535–12542, RI (2015). 18. M. Raugei, D. Morrey, A. Hutchinson, et al., A coherent life cycle assessment of a range of SC lightweighting strategies for compact vehicles. J Clean Prod, 108, p. 1168–1176, (2015). 19. S. Das, Life cycle energy impacts of automotive liftgate inner. Resour Conserv Recycl, 43, p.375–390, (2005) NU 20. S. Das, Life cycle assessment of carbon fiber-reinforced polymer composites. Int J Life Cycle Assess, 16, p. 268–282, (2011). MA 21. G.A. Keoleian, K. Kar, Elucidating complex design and management tradeoffs through life cycle design: Air intake manifold demonstration project. J Clean Prod, 11, p. 61–77, (2002). 22. V. Khanna, B.R. Bakshi, Carbon nanofiber polymer composites: evaluation of life cycle energy use. ED Environ Sci Technol, 43, p. 2078–84, (2009). 23. A.D. La Rosa, G. Cozzo, A. Latteri, et al. A comparative life cycle assessment of a composite EP T component for automotive. Chem Eng Trans, 32, p. 1723–1728, (2013). 24. I. Muñoz, I. Rieradevall, X. Domènech, et al. Using LCA to Assess Eco-design in the Automotive Sector: Case Study of a Polyolefinic Door Panel. Int J Life Cycle Assess, 11, p. 323–334, (2006) 25. A.T. Mayyas, A. Qattawi, A.R. Mayyas, et al., Life cycle assessment-based selection for a sustainable AC C lightweight body-in-white design. Energy, 39, p. 412–425, (2012). 26. C. Ribeiro, J.V. Ferreira, P. Partidário, Life cycle assessment of a multi-material car component. Int J Life Cycle Assess, 12, p. 336–345, (2007). 27. I. Ribeiro, P. Peças, A. Silva, et al., Life cycle engineering methodology applied to material selection, a fender case study, J Clean Prod, 16, p. 1887–1899, (2008). 28. A. Tharumarajah, P. Koltun, Is there an environmental advantage of using magnesium components for light-weighting cars?, J Clean Prod, 15, p. 1007–1013, (2007). 29. A. Tharumarajah, P. Koltun, Improving environmental performance of magnesium instrument panels. Resour Conserv Recycl, 54, p. 1189–1195, (2010). 30. H.L. Maclean, L.B. Lave, Life Cycle Assessment of Automobile / Fuel Options. Environ Sci Technol 37, p. 5445–5452, (2003). 17 ACCEPTED MANUSCRIPT 31. J.R. Duflou, J. De Moor, I. Verpoest, W. Dewulf, Environmental impact analysis of composite use in car manufacturing. Manuf Technol, 58, p. 9–12, (2009). 32. http://www.ricardo.com/en-GB/What-we-do/Technical-Consulting/VehicleSystems/Lightweighting1/ 33. D. Ferrario, Lightweight Design - Every Gram Counts. ATA - Ingegneria dell'Autoveicolo, 67, p. 611, (2014) 34. P. Martino, Automotive Lightweighting Materials: Current Trends and Future Scenarios, 4th PT International Conference ‘Advanced Materials and Technologies for Transport’, 27-28 February 2014, Turin. RI 35. W. Lynette Cheah, Cars on a Diet: The Material and Energy Impacts of Passenger Vehicle Weight Reduction in the U.S., PhD thesis MIT, (2010). SC 36. E. N. Weccard, Future scenarios for the derivation of material requirements – the automobile interior 2030, University of Twente thesis, (2012). 37. S. Cecchel, D. Ferrario, Numerical and experimental analysis of a high pressure die casting Aluminum NU suspension cross beam for light commercial vehicles, La metallurgia italiana, vol. 108, pp.41-44 (2016). MA 38. W. J. Joost, Reducing Vehicle Weight and Improving U.S. Energy Efficiency Using Integrated Computational Materials Engineering, JOM, 64, 1031-1038, (2012), DOI: 10.1007/s11837-0120424-z ED 39. J. Galos, M. Sutcliffe, D. Cebon, M. Piecyk, P. Greening, Reducing the energy consumption of heavy goods vehicles through the application of lightweight trailers: fleet case studies. Transp. Res. D 41, EP T 40–49 (2015). 40. M. Ravi Chandra, S. Sreenivasulu, S. Altaf Hussain, Modeling and Structural analysis of heavy vehicle chassis made of polymeric composite material by three different cross sections, IJMER, 2, 2594-2600, (2012). AC C 41. S. Cecchel, D. Chindamo, E. Turrini, C. Carnevale, G. Cornacchia, M. Gadola, A. Panvini, M. Volta, D. Ferrario, R. Golimbioschi.; Impact of reduced mass of light commercial vehicles on fuel consumption, CO2 emissions, air quality, and socio-economic costs, Science of the Total Environment, vol. 613-14, p. 409-417, (2018). 42. N. Sakundarini, Z, Taha, S. H. Abdul-Rashid, R. A. R. Ghazila, Optimal multi-material selection for lightweight design of automotive body assembly incorporating recyclability, Materials and Design, 50, 846–857, (2013) 43. S. Cecchel, D. Ferrario, R. Golimbioschi; A cross member of an axle assembly for a light commercial vehicle, WO2017/158535, (2017). 44. S. Cecchel, G. Cornacchia, M. Gelfi; Corrosion behavior of primary and secondary AlSi High Pressure Die Casting alloys, Materials and corrosion, doi:10.1002/maco.201709526, (2017). 18 ACCEPTED MANUSCRIPT 45. G. Cornacchia, S. Cecchel, A. Panvini; A comparative study of mechanical properties of metal inert gas (MIG)-cold metal transfer (CMT) and fiber laser-MIG hybrid welds for 6005A T6 extruded sheet, The International Journal of Advanced Manufacturing Technology, https://doi.org/10.1007/s00170017-0914-9 (2017) 46. Rheinfelden, http://rheinfelden-alloys.eu/wp-content/uploads/2016/01/05-HB DG_Ci_Sf_Cm_Td_Ma_RHEINFELDEN-ALLOYS_2015_EN.pdf, (2015). 47. S. Cecchel, G. Cornacchia, A. Panvini; Cradle-to-Gate Impact Assessment of a High-Pressure Die- PT Casting Safety-Relevant Automotive Component, JOM, vol. 68, pp. 2443-2448, (2016). 48. S. Cecchel, D. Chindamo, M. Collotta, G. Cornacchia, A. Panvini, G. Tomasoni, M. Gadola; RI Lightweighting in light commercial vehicles: cradle-to-grave life cycle assessment of a safety relevant component, IJLCA, p. 1-12 (2018), https://doi.org/10.1007/s11367-017-1433-5 SC 49. F. Grosselle et al., Doe applied to microstructural and mechanical properties of Al-Si-Cu-Mg casting alloys for automotive applications, Materials science and engineering A, 527, p. 3536-3545, (2010). 50. G.O. Verrana, R.P.K. Mendes, L.V.O. Dalla Valentina, DOE applied to optimization of aluminum NU alloy die castings, Journal of materials processing technology, 200, p. 120–125, (2008). 51. D.C. Montgomery, Introduction to Statistical Quality Control, Wiley, p.128, (2005). MA 52. International Organization for Standardization (ISO), ISO UNI EN 6892−1:2009 Metallic materials−tensile testing−Part 1: Method of test at room temperature, (Geneva: ISO,2009). 53. S.P. Midson, C.A. Dallmann, J.A. Brennan, T6 Heat Treating of Strontium Modified, Low-Iron ED Conventional Die Castings. NADCA Die Casting congress & exposition 2015, (2015). 54. International Standard Organisation (ISO), ISO 9227:2012 Corrosion tests in artificial atmospheres - EP T Salt spray tests, (Geneva: ISO, 2012). 55. A. Niklasa, A. Bakedanoa, S. Ordenb, M. da Silvac,d, E. Noguésc, A.I. Fernández-Calvoa, Effect of microstructure and casting defects on the mechanical properties of secondary AlSi10MnMg(Fe) test parts manufactured by vacuum assisted high pressure die casting technology, Materials Today: AC C Proceedings, 10A, p. 4931 – 4938, (2015). 56. Z.W. Chen, Skin solidification during high pressure die casting of Al-11Si-2Cu-1Fe alloy, Mater Sci Tech A, 348, p. 145-153, (2003). 57. A. Niklas et al., Effect of microstructure and casting defects on the mechanical properties of secondary AlSi10MnMg(Fe) test parts manufactured by vacuum assisted high pressure die casting technology. Materials today: proceedings 2, p.4931-4938 (2015). 19 ACCEPTED MANUSCRIPT FIGURE CAPTIONS Figure 1: HPDC aluminium suspension cross beam. Figure 2: Position of samples machined out of the component for the microstructural analysis. Figure 3: Samples separately cast from the component (top) and samples machined out of the component (bottom). Figure 4: Full suspension cross beam assembly. PT Figure 5: Full suspension assembly on road simulator test bench. Figure 6: Example of Optical and SEM micrographs of sample machined out of the cross beam. Figure 7: Tensile curves for each experimental combination and comparison with as cast condition. RI Figure 8: Pareto Chart of the standardised effects, main effect plot and contour plot for Rp0,2. SC Figure 9: Surface plot for Rp0,2 Figure 10: SEM on the fracture surfaces of the tensile samples separately cast from the component. In the fractographs, it is possible to observe the principal casting defect identified as A, shrinkage porosity. NU Figure 11: SEM on the fracture surfaces of the tensile samples machined out from: a) as-cast components, b) T5 components (b). In the fractographs, it is possible to observe some casting defects identified as: A shrinkage porosity, B cold flake, C intermetallics, D cold shut. MA Figure 12: Fatigue test bench of the suspension cross beam after salt spray test. AC C EP T ED Figure 13: Example of penetrant testing result. 20 ACCEPTED MANUSCRIPT TABLES Table 1: Chemical composition of EN AC-43500. % Si Fe Cu Mn Mg Ti 11.04 0.11 0.003 0.56 0.17 0.06 Table 2: Control factors for each experimental run. tsolubilisation (h) Taging (°C) taging (h) P01 0 150 1 P02 0 150 P03 0 190 P04 0 190 P05 5 150 P06 5 150 8 P07 5 190 1 P08 5 190 8 P09 2.5 170 4.5 PT Run 8 RI 1 MA NU SC 8 1 Table 3: Response variable results for each experimental combination and comparison with as cast condition. HV Rp02 [MPa] Rm [MPa] A(%) QI As-cast 87 ± 1 136 ± 7 263 ± 5 5.2 ± 0.9 298 ± 13 P01 87 ± 1 141 ± 2 249 ± 6 3.3 ± 0.4 261 ± 13 P02 97 ± 1 170 ± 3 254 ± 9 4.1 ± 0.9 309 ± 19 P03 96 ± 2 171 ± 3 264 ±10 4.0 ± 0.8 309 ± 17 96 ± 1 173 ± 4 261 ± 3 4.3 ± 1.7 313 ± 44 59 ± 0 94 ± 4 189 ± 7 12.4 ± 3.7 333 ± 26 70 ± 5 154 ± 11 229 ± 19 8.6 ± 1.1 324 ± 8 58 ±0 99 ± 9 176 ± 3 11.3 ± 1.9 338 ± 18 P08 68 ± 3 155 ± 5 205 ± 11 5.7 ± 1.6 334 ± 36 P09 82 ±4 164 ± 8 229 ± 12 7.0 ± 5.0 363 ± 57 P05 P06 AC C P07 EP T P04 ED Run Table 4: Tensile test on sample machined out of the component and cast into the die. AS-CAST T5 (170°C for 4h) Machined out of component Rp02 [MPa] Rm [MPa] A(%) A M B Cast into the die 134 5 266 20 4.7 1.8 118 4 242 12 4.4 1.0 117 3 234 25 6.4 3.8 134 7 258 5 5 0.9 21 Machined out of component Cast into the die A M B 182 2 288 17 3.6 1.6 157 4 253 5 2.8 0.5 152 3 248 21 3.5 2 industrial laboratory oven oven 172 173 2 3 267 250 13 10 2.6 2.6 2 1 ACCEPTED MANUSCRIPT AC C EP T ED MA NU SC RI PT Graphical Abstract 22 ACCEPTED MANUSCRIPT HIGHLIGHTS AC C EP T ED MA NU SC RI PT -50% weight aluminium EN AC-43500 cross beam suspension for CV was developed, overcoming costs and structural limits Proof of the good resistance to corrosion and of the desired mechanical properties of the cross beam suspension Confirmation that the cross beam structure of the independent front suspension can endure 250’000 km road application without fatigue failure Final demonstration of the feasibility of the light alloy use for this particular heavy application Database of EN AC-43500 mechanical properties under different heat treatments, useful for the for the designing of different automotive components 23