Squeeze Casting of Aluminum Alloy Safety Critical Components for Automotive Applications

Transcription

1 Squeeze Casting of Aluminum Alloy Safety Critical Components for Automotive Applications Rathindra DasGupta *, Chuck Barnes *, Peter Radcliffe ** & Paul Dodd ** * SPX Contech, USA, ** SPX Contech, UK Abstract Squeeze casting is considered a high integrity process, and has given materials and design engineers a new alternative to conventional casting techniques: gravity permanent mold (GPM) and conventional (high pressure, high velocity) die-casting. In recent years, the squeeze casting process has been applied to near net shape products requiring high impact strength, high fatigue strength, pressure tightness, or high wear resistance. This paper briefly discusses the squeeze casting technology (P2000 TM ) at SPX CONTECH and provides examples of the various aluminum alloy safety critical components manufactured. The paper also includes fluid flow simulations and solidification analyses for select components, comparison between predicted and actual microstructures, mechanical properties (tensile and fatigue) based on samples machined from actual castings, and results from strength and fatigue testing of components. Key words (squeeze casting, safety critical, aluminum). 33/1

2 Squeeze Casting Technology (P2000 TM ) The unique features of the P2000 TM technology include the die design, lubricant type and application mechanism, choice of alloy and heat treatment, melt treatment and finally the casting machine itself. Figure 1 shows the schematic of a typical casting machine used. The casting process consists of pouring degassed and filtered molten metal into a vertical chamber (shot sleeve), slowly forcing the metal into a preheated, lubricated die cavity, and continuous application of pressure during solidification of the melt. The slow injection (speed slower than that used for conventional die-casting) results in reduced turbulence during flow of metal into die cavity, thereby contributing to minimal air entrapment in the matrix. The rapid heat transfer (from the contact of the molten metal with the lubricated die surface) and the continuous application of pressure during solidification of the melt in the cavity also help reduce shrink porosity. The casting/tooling design process in the P2000 TM technology follows much the same rules as for conventional die-casting [1]. However, certain casting/tooling design features do set the P2000 TM technology apart from conventional die-casting. For example, the P2000 TM casting process requires larger runners and gates to promote non-turbulent metal flow, flat areas for gate since the gate must be sawed off, and thicker walls (>2.5 mm) to allow adequate pressure during solidification [1]. In recent years, the P2000 TM technology has been widely used with various aluminum alloys (potential mass reduction of 40 to 60% can be achieved when aluminum replaces cast iron and steel fabrications) to manufacture near-net shape and lightweight safety critical automotive components including control arms, knuckles, wheels, and cross members. Although other casting processes such as gravity permanent mold (GPM) and low-pressure permanent mold (LPPM) are capable of producing these parts, the P2000 TM process is now established as a proven, practical and competitive method for producing lightweight structural aluminum castings. The wide acceptance of the P2000 TM technology can be attributed to the following: reduced or absence of shrink porosity in the matrix resulting from high cavity pressure (70 to 100 MPa) and rapid solidification rate, dimensional control similar to that of conventional die castings, heat treatable (can be subjected to solution treatment without the fear of blistering) and improved mechanical properties [1]. Smaller dendrite arm spacing (DAS) and a fibrous silicon morphology resulting from the rapid solidification rate in squeeze casting are factors responsible for the improvement in tensile properties, particularly ductility. For example, Table 1 compares the tensile properties of components made using various casting processes for select alloys [1]. It is evident that for comparable strengths, the P2000 TM process yields higher ductility than GPM and conventional die-casting. 33/2

3 The net-shape capabilities, alloy flexibility and ease of automation for high volume production are additional advantages of the P2000 TM technology when compared with GPM and low-pressure permanent mold (LPPM) processes. The benefits achieved from the P2000 TM casting process alone, however, cannot guarantee the desired level of part quality. Thus, prior to casting development trials, computer aided process simulation and modeling are used to help optimize tooling design, proper solidification, cooling line configurations, and mold (cavity) fill. This paper provides examples of the safety critical components, namely, suspension links, front and rear knuckles manufactured at SPX Contech. Also included in this study are process simulation results, mechanical properties based on samples machined from the actual components, and results from product testing. P2000 TM Applications: Suspension Links Figure 2 shows the various suspension links converted from aluminum forgings to castings. A multi-cavity die for each part number is used to satisfy the customer requirement of 800,000 pieces per year. All parts undergo non-destructive inspection (x-ray and ultrasonic) prior to shipment. Additional information on these castings and the approach to achieving a robust and reliable manufacturing process for the suspension links are outlined below. Part weight, aluminum alloy type and heat treatment: a) Tension link = 0.41 kg b) Compression link = 0.41 kg c) Camber link = 0.68 kg d) Aluminum alloy A356.2 e) Heat treatment T6 temper (solution treated, quenched, and aged) Material properties attained from actual castings: The most common material properties measured prior to shipping the production links to customers include the yield strength, tensile strength, percent elongation and hardness. These data are obtained from tensile specimens machined from actual castings. Shown below are typical data obtained from one of the camber links. Yield strength = MPa (minimum requirement is 207 MPa) Tensile strength= MPa (minimum requirement is 275 MPa) Percent elongation = 10.27% (minimum requirement is 7%) Hardness = 100 BHN (minimum requirement is 85 BHN) Approach to manufacturing high integrity suspension links: A critical feature for the proper use of the P2000 TM technology is the application of process simulation and modeling (prior to casting 33/3

4 development trials) to help optimize tooling design, cavity fill and solidification. The steps involved in the simulation process typically consist of performing a natural solidification (with no external cooling) followed by a thorough investigation into fill profiles, cooling line configurations and cooling sequences to significantly minimize or eliminate the porosity observed during natural solidification. Figure 3 shows the solidification pattern (with appropriate cooling) for a camber link after 80.18% of the melt has solidified. Further optimization of cooling line configurations, cooling sequences and casting process parameters helped produce parts meeting customer porosity specifications. Component testing during casting development trials was another approach to developing a robust and reliable manufacturing process for the links. Thus, during product testing, attention to consistency in fracture loads (or number of cycles to failure) and failure locations, correlation between actual failure locations and those predicted through finite element analysis (FEA) helped establish the appropriate process window required for producing high integrity suspension links. For example, Table 2 shows the fracture loads and failure locations obtained from left-hand (LHD) camber links (made during one of the casting development trials) following strength testing. The fracture loads (ranging from 54.2 to 60.2 kn) are fairly consistent, and failure occurred in desired areas. Determining the worst-case scenario though evaluation of components made intentionally bad (specifications for discontinuity size not met) was another avenue for achieving a robust manufacturing process for the links. It was felt that the data from such a test would help in developing preventative measures to combat a potential problem that could occur in production. Table 3 shows the results from left-hand camber links made intentionally bad with inclusions of varying types and sizes. A comparison between Tables 2 and 3 shows that the presence of inclusions in castings has a larger effect on the range of fracture loads than on failure locations. P2000 TM Applications: Front Steering Knuckles Figure 4 shows a 3.65 kg front steering knuckle, previously made from direct squeeze casting. The customer requirement of 60,000 pairs per year is met using a two-cavity die (left-hand and right-hand), strontiummodified A356.2 alloy and a T6 temper. Material property measurement, dimensional checks, x-ray and ultrasonic inspection are required prior to shipping the production steering knuckles. The typical tensile data obtained (three bars per hand from locations shown in Figure 4) are as follows. Yield strength = MPa (minimum requirement is 207 MPa) Tensile strength = MPa (minimum requirement is 276 MPa) Percent elongation = 10.5% (minimum requirement is 6%) 33/4

5 Although not a customer requirement, it was of interest to determine the low cycle fatigue (cyclic loads are relatively high, significant amounts of plastic strain are induced during each cycle, and low number of cycles to failure) behavior for squeeze cast A356.2-T6 aluminum alloy using samples machined from these steering knuckles. For a steering knuckle, even if the loads are nominally low, the material at the root of a critical notch may experience local plasticity that is strain-controlled, and the method of low cycle fatigue becomes important in predicting its life. Figure 5 shows the strain-life curve for squeeze cast A356.2-T6 with upper/lower bounds at 95% confidence level whereas Figure 6 shows the cyclic stressstrain curve for the same alloy together with part of the monotonic curves for comparison. Approach to manufacturing high integrity front steering knuckles: Various cooling line configurations, cooling sequences and fill profiles were modeled to obtain the desired part quality. In addition to flow and solidification modeling, microstructure simulation of select locations was conducted to predict and compare dendrite arm spacing (DAS) with that obtained from production steering knuckles [2]. Despite an offset between the actual dendrite arm spacing (DAS) and the simulated DAS, a significant correlation (correlation coefficient = 0.95) was observed between the two sets of data [2]. Fatigue and strength testing of the above component were conducted during casting development trials to help achieve a robust manufacturing process for the front steering knuckles. Table 4 reveals the data obtained from one such test [2]. A close examination of the data shows both lefthand (LHD) and right-hand (RHD) knuckles to exhibit very similar fracture loads. Furthermore, failure location was the same for both knuckles during strength testing. P2000 TM Applications: Rear Steering Knuckles Figure 7 shows a kg rear steering knuckle. The customer requirement of 15,000 pairs per year is met using a two-cavity die (lefthand and right-hand), strontium-modified A356.2 alloy and a T6 temper. Material property measurement, dimensional checks, x-ray and ultrasonic inspection are again required prior to shipping the production knuckles. The typical tensile data obtained (two bars per hand from locations shown in Figure 7) are as follows. Yield strength = MPa (minimum requirement is 207 MPa) Tensile strength = MPa (minimum requirement is 276 MPa) Percent elongation = 10% (minimum requirement is 6%) Approach to manufacturing high integrity rear steering knuckles: 33/5

6 Numerous fill and solidification analyses were performed to optimize tooling design, proper solidification and cavity fill. Figure 8 shows the solidification pattern for the rear knuckle with no external cooling (natural solidification) after 98.56% of the melt has solidified. The hot spots (prone to porosity) are clearly evident. Various cooling line configurations, cooling sequences and fill profiles were, therefore, investigated to eliminate the porosity. Fatigue and strength testing of the above component were also carried out during the casting development trials to ensure a robust and reliable manufacturing process. For example, Figure 9 shows the typical set-up for fatigue testing (of the upper control arm) and the desired failure location. Results from the various tests conducted are summarized in Table 5. The toe link arms exhibit approximately 50% higher fracture load than the upper control arm when subjected to strength tests. The toe link arms were also observed to have undergone a greater number of cycles prior to failure during fatigue test. Conclusions The P2000 TM technology is considered a high integrity process, and has given materials and design engineers a new alternative to conventional casting techniques: GPM and conventional die-casting. The P2000 TM technology is currently used for manufacturing a variety of safety critical components including suspension links, front and rear steering knuckles. The benefits achieved from the P2000 TM casting process alone, however, cannot guarantee the desired level of part quality. Thus, prior to casting development trials, computer aided process simulation and modeling are necesssary to help optimize tooling design, proper solidification, cooling line configurations, and mold (cavity) fill. In addition, component testing during casting trials is another mechanism to assess integrity (consistency and reliability) of parts. References 1. Corbit S.A. and DasGupta R., Squeeze cast automotive applications and squeeze cast aluminum alloy properties, SAE International Congress and Exposition, paper # DasGupta R., Xia Y. and Szymanowski B., High volume squeeze cast applications, Proceedings of the 2 nd International Light Metals Technology Conference, St. Wolfgang, Austria, 8-10 June 2005, pp /6

7 Tables Table 1: Comparison of tensile propeties [1] Alloy Process Yield strength, MPa Tensile strength, MPa Percent elongation A356.2-T6 P2000 TM A356.2-T6 GPM T6 P2000 TM T6 GPM Table 2: Left-hand (LHD) camber links fracture loads Sample identification Failure location Fracture load, kn Camber 1 Top end hoop 57.1 Camber 2 Top end hoop 54.2 Camber 3 Top end hoop 57.5 Camber 4 Top end hoop 59.5 Camber 5 Top end hoop 55.7 Camber 6 Top end hoop 60.2 Camber 7 Midsection 54.7 Camber 8 Midsection 54.8 Camber 9 Midsection 56.2 Camber 10 Top end hoop 58.8 Table 3: Effect of inclusions on fracture loads Type of inclusion Range of fracture load, Failure location kn Graphite Top end hoop and midsection Oxide (dross) Top end hoop Furnace filter Top end hoop and midsection 33/7

8 Table 4: Strength testing of front knuckles [2] Sample Left-hand Right-hand Failure location identification fracture load, kn fracture load, kn Pad Pad Pad Pad Pad Pad Pad Pad Pad Pad Average / /- 1.9 Pad Table 5: Rear knuckle testing Test type Location tested Fracture load Typical casting (kn) or number failure? (Yes/No) of cycles to failure Strength test Toe link / Yes Control arm / Yes Fatigue test Toe link 460,137 to 1,246,014 cycles Yes Control arm 43,033 to Yes 691,659 cycles Figures Moving Platen Eject Half Cavity Cover Half Stationary Platen Camber Compression Camber Molten Metal Tension Shot Cylinder Figure 1: Schematic of P2000 TM Figure 2: Suspension links [2] casting machine [2] 33/8

9 Figure 3: Solidification modeling for a camber link 3 bars per part Minimum tensile = 276 MPa Minimum yield = 207 MPa Minimum elongation = 6% Figure 4: Front steering knuckle Strain Amplitude y = x y = x Elastic Plastic Total [95% upper] Total [50%] Total [95% lower] E=72262 MPa σ f'=614.2 MPa b= ε f'= c= Nf [Rev] Figure 5: A356.2-T6 strain-life curve 33/9

10 400 σ'0.2 [95% upper]=320mpa σ'0.2 [50%]=290MPa 300 Stress [MPa] σ0.2=225mpa E= MPa K'=538 MPa n'= σ'0.2 [95% lower]=258mpa σ0.2=235mpa Cyclic 95% upper Cyclic 50% Cyclic 95% lower Test data Monotonic 1 Monotonic Strain Figure 6: A356.2-T6 cyclic stress-strain curve Two bars per hand Figure 7: Rear steering knuckle Figure 8: Natural solidification pattern for rear steering knuckle Figure 9: Set-up for fatigue testing and desired failure locartion 33/10