TECHNICAL PAPERS

Production of Magnesium Powertrain Components via Thixomolding®

Stephen LeBeau, Raymond Decker, D. Matthew Walukas
Thixomat, Inc.

Abstract
The use of magnesium alloys in automotive applications continues to grow at an unprecedented rate. Since 1990, the rate of utilization of magnesium in automotive applications has increased some 20 % on an annual basis. This paper will review some of the key attributes that make magnesium an attractive structural material. Emphasis will be placed on powertrain applications and technical barriers that need to be overcome to increase magnesium consumption by as much as 20 to 50 times over the current levels. Thixomolding® produces net-shape parts from magnesium alloys in a single step process involving high-speed injection molding of semi-solid thixotropic alloys. A description of the process and status of commercial developments will be presented. Semi-solid processing of magnesium alloys provides a vehicle for enhancing the performance and increasing the utilization of lightweight materials in commercial applications.

Introduction
The use of magnesium in the transportation industry has been receiving tremendous interest in the past few years [1]. The growth in magnesium components is being driven by demands for reduced vehicle emissions and improved fuel efficiency. The automobile industry made a voluntary commitment to reduce fuel consumption by 25% in comparison with 1990 levels by the year 2005. At the same time, this will substantially reduce CO2 emissions and conserve finite oil reserves. For cars, there is a general expectation that U.S. government standards will increase from the current levels of 8.7 L/ 100 km (27.5 mpg) to 7.7 L/ 100 km (31 mpg) in the next 10 years. The seriousness of the commitment by the automotive industry is demonstrated by the development of mass-produced vehicles capable of fuel consumption of only 3 L/100km (80 mpg) such as Volkswagen's 3L-Lupo. [2]

There are many ways to reduce fuel consumption and emissions, such as with improved powertrain efficiencies, IC-diesel hybrids, alternate fuels, aerodynamics and mass reduction. While all of these cost more than the current systems, mass reduction is just about the least costly, but only if the reductions are large, such as in the 20 to 40 % range. Mass reduction is also being driven by market demands for added features (sound, lights, comfort, entertainment) and increased safety features which tend to increase the mass of vehicles. These demands for more comfort features are direct conflict with demands for increased fuel efficiency. According to Sweder [3], a 10% mass reduction in the average vehicle results in the following benefits:

· Fuel economy - 2 mpg improvements
· Improved performance - 0.5 sec reduction in 0 to 60 mph
· Emissions reduction - 7 % less feed gas
· Safety - 10 % less kinetic energy
· Payload and trailer towing - 300 lb. Improvement
· Braking - 10 foot reduction in barking distance
· Increased Feature availability - DVD, in-vehicle entertainment systems, etc. permitted due to mass reductions

Advantages of Magnesium

Magnesium is important in vehicle mass reduction as a result of having several important characteristics:

1. It has a lower density. At 1.8 kg/Liter, magnesium alloys are 20 % lighter than polymer composites, 30 % lighter than aluminum alloys and 75% lighter than steels.

2. It has outstanding castability. Magnesium can be cast into components, which at 1.0 - 1.5 mm wall thickness and 1-2 degree draft angles are ½ the thickness of aluminum cast components. Magnesium's higher fluidity allows many steel fabrications to be replaced by one large cast magnesium component. For example, a typical steel cross car beam instrument panel consisting of 35 different parts can be replaced by a die cast magnesium IP containing only one-tenth the number of components.

3. It has enhanced surface properties. Magnesium die castings demonstrate a "skin effect' in the sense that material/mechanical properties near the surface are much improved over the bulk. Thinner magnesium castings may then have sufficient strength per unit area to support loads better than thicker sections and thereby compete with heavier aluminum and plastics sections.

The mechanical properties of magnesium die cast components are not altogether different from aluminum. The properties of the most common alloys, AZ-91D (9 % Al, 1 % Zn) and AM-60B (6 % Al) are compared with AL 380 alloy in Table 1. In non-structural components such as covers, cases and housings, where ductility is not an important mechanical attribute, AZ-91D provides the same functional performance as 380 aluminum, but at a lower mass. There are a growing number of structural components such as instrument panels, seats and steering wheels where ductility, energy absorption and impact resistance are important material considerations. In these applications AM50/AM60 alloy castings can be used, since they have ductility exceeding 10% and better impact properties.

To be presented at GPC Powertrain Conference, Ann Arbor , MI Sept. 2002

Alloy	% Elongation	Modulus (GPa)	Yield (MPa)	UTS (MPa)
AZ91D	3	45	160	240
AM60B	8 - 12	45	130	225
380 Al	3	71	165	320

Table 1. Mechanical Properties of Die Cast Magnesium and Aluminum Alloys

Magnesium alloys have tighter dimensional tolerances (+/- 0.001 mm/mm) that are about 50 % better than aluminum alloys. The result is that magnesium cast components can be cast closer to net shape that reduces materials usage and lowers costs. Magnesium also has reduced latent heat and less affinity for iron than aluminum that leads to 2 - 4 times the die life (from 150,000 - 200,000 shots for aluminum to 300,000 - 700,000 shots for magnesium). The lower latent heat also permits up to a 25% increase in productivity due to reductions in shot-to shot cycle times versus aluminum.

The development path that led to the use of die-casting for large-scale model production series may be traced to the early work by the Volkswagen Company with initial prototypes produced via sand casting as early as 1934. [4] When the production numbers increased in the fifties, die-casting was introduced for increased production capacities. The high point of this early consumption reached a peak in 1971 when the use of magnesium by VW reached an annual level of 42,000 tons. Much of that production was for engine block and gearbox housings for the air cooled Beetle and Transport models at a level of 20 kilograms per vehicle of magnesium. In the yeas that followed, the introduction of water-cooled front engine designs, the inadequate heat resistance of magnesium as well as problems with poor corrosion resistance led to a substantial reduction in magnesium usage. Cost pressures relative to aluminum also made magnesium less favorable eventually leading to the complete cessation of gearbox production in 1982.

Since the development of high purity magnesium alloys, i.e. with very low tolerance ranges for Ni, Cu, and Fe contents, the corrosion resistance of magnesium alloys has surpassed that of aluminum secondary alloys. Table 2 compares the corrosion resistance of the earlier versions of the magnesium alloys (AZ91A, AM60A, AS41A) with the new high purity alloys (AZ91D, AM60B, AS41B and AE42A) relative to mild carbon steel and 380-aluminum alloy. Recent studies on the corrosion resistance of Audi 5-speed manual gearboxes indicate significantly better results for surface corrosion are achieved with AZ91D alloy compared to 380 aluminum alloys. [5].

Material	Corrosion rate, MPY
Carbon steel	30
380 Al	15
AZ91B	680
AM60A	700
AS41A	680
AZ91D	4
AM60B	22
AS41B	4
AE42A	6

Table 2 Salt Spray Corrosion Performance Mg versus Fe and Al.
(10 day ASTM B-117 salt fog)

The growth of the use of magnesium die-castings that has been occurring since 1991 and is expected to continue to grow for the next 10 years is shown in the table below for the North American market. The North American market has led the way for several years now but the European market is actually growing at a greater pace [6] and is expected to match the North American numbers during the next decade. It is interesting to note that Volkswagen (Dead Sea Works), General Motors (Norsk Hydro), Ford (Queensland, Australia) and Toyota (Magnola Project, Canada) have aligned themselves with existing magnesium producers or new magnesium producers to secure long term supply sources of magnesium.

Year	Metric Tons ('000)
2000	52
2001	58
2002	65
2003	77
2004	87
2005	95

Table 3. North American Demand Forecast for Die Cast Magnesium

Current Commercial Applications

There are over 60 separate and individual components where magnesium is being used or developed for use in automobiles today. The most popular of those components are:

1. Instrument panel substrates
2. Seat Frame
3. Steering Column Components
4. Engine Valve Covers
5. Transmission housings
6. Intake Manifolds

Vehicle platforms with significant quantities of magnesium include the following:

1. GM full sized vans - Savana & Express - up to 26 kg
2. GM minivans - Safari & Astro - up to 16.7 kg
3. Ford F-150 truck - 14 kg
4. VW Passat, Audi A4 & A6 - 13.6 to 14.5 kg
5. Buick Park Avenue - 9.5 kg
6. Alfa Romeo 156 - 9.3 kg
7. Mercedes SLK roadster - 7.7 kg
8. Chrysler minivans - 5.8 kg

The following table lists components for the automotive drivetrain section that are either in production or under development [6].

O = Production X = Development

Table 4. Automotive Drivetrain Magnesium Components

The possibility of using magnesium for a specific component is essentially determined by its requirements profile. Primarily, that involves the effects and loads to which it is subjected under operating conditions and the resulting requirements for rigidity and strength that must take account of the effects of temperature and the surrounding environment. In addition there is a whole series of primary or associated requirements such as safety, comfort, weight, cost-efficiency, manufacturability, and environmental safety.
A good example of this is the B-80 transmission housing for the VW Passat and Audi A-4 and A-6 platforms. In this case the gearbox housings are being produced with AZ91D alloy as a replacement for the previous aluminum housing. Because of the lower elastic modulus the loss of stiffness had to be counteracted with greater use of ribbing (which also improves acoustics) and localized increases in wall thickness. On the other hand, the magnesium alloy's good casting properties allowed thinner walls to be used at points of low stress. The result was a real weight savings of 4.5 kg in the case of the B80 gearbox (equivalent to a 26% weight savings which is very close to the theoretical achievable limit). An interesting aspect of this application is that the greater tendency to creep and galvanic corrosion and the lower heat resistance of the magnesium alloy were tackled by design strategies. The new high purity AZ91D eliminated the need for secondary coatings (chromating and wax coating) used in the earlier housings used by VW back in the 1970's. Ford also has successfully used AZ91D in 4 wheel drive transfer cases, without any high temperature creep problems, because the case and methods of joining the case were designed properly.

Another example is the use of magnesium for engine valve covers. A continuing debate goes on as to what is the best material for this application and the designer has a number of choices. Magnesium, aluminum, steel and plastic are all used for valve covers, depending on which car company and which model you are looking at. The reasons this application keeps growing in magnesium are:

1. Weight reduction
2. Elimination of oil leaks
3. NVH control

Most of the current designs are not taking full advantage features that magnesium can offer for this application:

1. Magnesium recyclability.
2. Use 'as cast" or mechanical finish that eliminates the need for paint.
3. Good higher temperature creep properties offering insurance against oil leaks.
4. Casting characteristics that allow many of the features to be cast in, thus eliminating the need for additional component costs for machining, fittings, attaching screws and dis-similar metal oil baffle plates.

Restrictions to Extensive Use of Magnesium in Drivetrain Components

Despite the recent advances in high purity magnesium alloys and applications already cited, the overall average vehicle content for magnesium is only 2 - 3 kg. on average worldwide.

The following areas of concern have been identified as barriers to the increased use of magnesium in automotive applications:

1. Mechanical property database - without sufficient data, product designers cannot fully take advantage of the full potential mass reductions provided by magnesium. An important problem is the lack of realistic database from actual components and not test bars, as historical evidence has indicated substantial differences between the two.
2. Expert systems for design and manufacturing - an expert system which integrates design features (using the database described above) and the manufacturing process capabilities is essential for the production of the lowest cost components.
3. New alloys - the lack of resistance of creep and heat have restricted or proved an obstacle to the use of magnesium in engine and transmission components in the past or made design modifications necessary that have diminished the material's weight saving potential.
4. Improvements in current manufacturing methods or development of new improved methods - since the quality of the casting, i.e. inclusions, porosity and heat cracking restrict the application of magnesium or result in designing based on less than optimum properties the quality of components must be improved.

Thixomolding® of Magnesium Structural Components

It is well known that the mechanical properties of cast metals and alloys are strongly influenced by a variety of casting defects and artifacts, all of which reduce strength and ductility. These include entrapped gas porosity, shrinkage porosity, oxide, flux, and dross inclusions and planar defects such as cold shuts and folded oxidized surfaces. Die-castings are especially prone to high levels of porosity due to high velocity turbulence of molten metal entering the die and oxide and flux inclusions resulting from melting and liquid metal transfer systems. Elimination of defects and reductions of porosity will lead to improvements in the quality and performance of metal castings. In addition to elimination of defects, control of the structure of cast metals, including the size, distribution, and location of precipitates, the size and morphology of crystalline phases, the amount and location of porosity and the final grain size of the material have a direct impact on the resulting properties.

The final structure of a cast metal is a function of not only the alloy chemistry but also the processing history the material experiences during production. The benefits of semi-solid processing of cast metals has been the subject of considerable research dating back to the early work performed at MIT. [7] One of the earliest reports of casting benefits owing to the casting of slushy melts came from the UK in which molten alloys were stirred as they were cooled into a slushy stage and then cast into molds resulting in some improvements in strength and porosity levels. During the same period of time foundry studies at MIT led to the development of Rheocasting, the first well-known semi-solid metal casting process. The MIT research led by Flemings, capitalized on the discovery that a morphological alteration of the solid phase occurred on rapid stirring that resulted in a thixotropic behavior of the two-phase slurry. The viscosities of slurries thus formed exhibit inverse dependencies on shear rate (2). In addition to a stabilization of the state relative to undisturbed melts, this shear thinning phenomena imparts the ability to process alloys at reduced temperatures by forging, extrusion, coining, molding, and rolling.

Thixomolding®, a semi-solid injection molding process, is a revolutionary new process for the high volume production of magnesium alloys in one step in a single integrated machine. The Thixomolding® machine, show schematically in Figure 1, is similar to a plastic injection molding machine. The process consists of introducing Mg alloy feedstock in the form of metal granules at room temperature to a heated barrel and screw of a modified injection molding machine and then raising the temperature of the material to a semi-solid region under high shear rate mixing. The semi-solid slurry, consisting of nearly spherical solid particles suspended in a liquid matrix, is then injected into a preheated metal mold to make a net shape part. For a detailed description of the Thixomolding® process and its significant advantages, the reader is referred to a number of earlier publications [8,9].

Fig. 1. Schematic of semi-solid injection molding machine.

Thixomolding® has several advantages over traditional die-casting. In contrast to die-casting where the metal enters the die as turbulent streams of atomized sprays, the semi-solid thixotropic metal fills the mold generally in a planar flow front. This results in lower porosity than die castings and an ability to recycle scrap without secondary refining operations since the alloy is melted under an argon cover gas. The Thixomolding® process has the virtue of permitting the production of parts over a wide range of solid fractions, nominally of the order of 0.05 to 0.60. This is a unique advantage over other semi-solid metal processing (SSM) methods such as semi-solid billet die casting that are restricted to solid fractions greater than 0.60 [8].

The mechanical behavior of SSM processed parts have been widely reported to exhibit improvements over die cast counterparts, principally due to a reduction in porosity and other casting defects. Thixomolding® extends the reduction in porosity to the entire range of solid fractions employed, exhibiting reductions of 50% or more of equivalent die cast parts. There is an absence of continuous porosity using properly designed tools that permits current commercial production of gas tight hermetic enclosures without the need to employ a costly secondary resin impregnation step as is common for Al die castings [10]. The reduction in porosity and absence of continuous porosity leads directly to improvements in both ductility and tensile properties
The tensile mechanical properties are shown in Table I for AZ91D, AM60, and AM50, the three most commonly used Mg alloys. It is noted that AZ91D meets or exceeds the ASTM die casting specification at solid fractions of fs=0.3 or lower. At higher solid fractions heat treating is required to achieve maximum strength levels, a requirement discussed earlier of SSP Aluminum alloys formed using the billet / preform process. AM50 and AM60 are likewise seen to exceed the ASTM die casting specifications when Thixomolded®. The data tabulated was sourced from both independent investigations and round robin studies of common batches of standardized test samples. [9-17].
The tensile mechanical properties are shown in Table 5 for AZ91D, AM60, and AM50, the three most commonly used Mg alloys. It is noted that AZ91D meets or exceeds the ASTM die casting specification at solid fractions of fs=0.3 or lower. At higher solid fractions heat treating is required to achieve maximum strength levels, a requirement discussed earlier of SSP Aluminum alloys formed using the billet / preform process. AM50 and AM60 are likewise seen to exceed the ASTM die casting specifications when Thixomolded®. The data tabulated was sourced from both independent investigations and round robin studies of common batches of standardized test samples. [9,12-18]

Alloy	Ultimate Strength MPa	Yield Strength MPa	%El
*AZ91D	299 to 230	169 - 145	10 - 6
AS41A	235		7.5
AM 60	278,2	150	18.8
AM 50	268.7	140	20.0
* Note the range of properties for AZ91D reflects the influence of controlled variation in solid fraction fs. (<.05 to ~ 0.20)
ASTM Specifications
AZ91D	230	150	3.0
AM 60	210	125	10.0
AM 50	220	130	8.0

Table 5. Tensile Properties of Thixomolded® Mg Alloys

Table 6 summarizes results from two independent studies of the mechanical properties of recycled AZ91D. In both studies the source of material was sprues, gates, runners, and scrapped Thixomolded® parts. The scrap was reground using techniques similar to that for virgin feedstock. No remelting or refining steps were used. The data clearly demonstrate the ease of recycling Thixomolding® scrap. [12,13].

%Recycle	Yield Stress MPa	Ultimate Stress MPa	%El
100% (a)12	166.2	260.5	7.0
100%(b)12	167.3	254.5	6.1
50%(A)12	169.5	261.9	6.0
50%(b)12	166.7	263	6.5
10%(a)12	164.7	258.9	6.3
10%(b)12	162.4	255.6	7.0
100%13		295.5	10.3
i. (a) and (b)12 refer to solid fractions fs of <0.05 and 0.1 respectively, whereas the second investigation, ref. 13, was carried out with fs<0.05

Table 6. Tensile Properties of Recycled AZ91D.

Improvements in the high temperature creep behavior of Mg alloys above 150°C continue to be a challenge for automotive applications. While weight reduction is an important factor, stress relaxation under compressive loads of current alloys falls well below those of Al alloys. [13] Fastener bolt load torque retention levels specified are difficult to meet.

There has been limited research on Thixomolded® Mg alloys. (16,19) However in both studies there is a marked improvement in steady state creep rates for both AZ91D and AS41 as compared to counterpart die cast structures. The improvement which is enhanced with increasing solid fraction, fs, is attributed to the duplex microstructure consisting of a primary a, Mg rich, phase distributed uniformly in a divorced eutectic continuous matrix of Mg17Al12 + a Mg.

Table 7 compares the steady state creep rates for several testing conditions. LeBeau & Decker [20] have suggested that the unique microstructures and flexibility of the process in permitting micro structural design holds promise for the development of a new class of Mg based alloys.

Alloy	Solid fraction, fs	Temp, 0K	s, MPa	d in/in/sec
AZ91D(16)	4%	398	50	11.2 x 10-9
AZ91D(16)	39%	398	50	7.5 x 10-9
AZ91D(21)	< 5%	398	50	10 x 10-9
AS41D(21)	< 5%	398	50	3 x 10-9
AS41D(21)	< 5%	398	70	1 x 10-6
AS41D(21)	< 5%	423	70	8.1 x 10-6
Die Cast
AZ91D	0	398	50	13.5 x 10-9

Table 7. Steady State Creep Rates for Thixomolded® Mg Alloys.

The rate of commercialization of the Thixomolding® of Mg alloys is the most rapidly growing segment of the SSP industry and is practiced by more than forty end user licensees in the U.S., Canada, Japan, Taiwan, Germany, and Sweden collectively operating over 230 Thixomolding® machines in 2002. The growth is geometric in installed base and new licensees spurred by original equipment manufacturers in electronics, telecom, and automotive now specifying product produced by this revolutionary process [10].

Future Development and Conclusions

Lightweight construction using magnesium is in competition with lightweight construction using aluminum, plastics, or steel. The projected increases in the use of magnesium can only be accomplished by a rigorous process of bringing together the entire spectrum of engineering skills involved in bringing new products to market. The magnesium industry does not contain the infrastructure provided to the OEM and Tier 1 suppliers by the polymer, aluminum and steel industries. The projected increases in magnesium consumption in automotive applications can only be accomplished through a cooperative development effort, on an international level, between the automotive OEM producers and their supplier base. Only by adopting an integrated approach that includes the elements of design, prototyping, modeling, improved magnesium alloy development and improved manufacturing processes can the forecasts for magnesium be achieved.

The Thixomolding® process is in its commercial infancy having been introduced at a level matching or exceeding the state of the art for die-casting, which is a very mature technology. New developments and applications are occurring rapidly as the knowledge base continues to grow. Numerical modeling of fluid flow and heat transfer in Thixomolding® has been demonstrated [6] and will in the future result in further improvements in tool design. Research is also progressing rapidly on hot runner die designs that will lead to a dramatic improvement in raw materials utilization. Semi-solid processing of magnesium alloys provides the opportunity to adjust the microstructure of magnesium alloys to optimize their performance for the high temperature and structural requirements demanded by increased utilization by the automotive industry, particularly in the powertrain component areas. The Thixomolding® process has proven itself capable of meeting or exceeding the early promises of SSM with a bright future in store. As a safe and environmentally sensitive process it is an appealing choice for the production of net shape parts today, offering the potential to provide an economically sound approach for manufacturing well into the 21st Century.

References

1. G. Cole, Proceedings of 56th International Magnesium Association, Rome, 1999, page 21.
2. Friedrich, et.al. "The Second Age of Magnesium", Dead Sea Magnesium Conference 2000, page 9
3. T. Sweder, " Demand for Lightweight Technology", Oral Presentation, Magnesium Expo, March 2000, Ypsilanti, Michigan
4. S. Schumann, F. Friedrich, "The Use of Magnesium in Cars - Today and in the Future", Mg. Alloys and their Applications, Wolfsburg, Germany, April 1998
5. Albright D.L., Haagensen J.O., 54th International Magnesium Association, Toronto, Canada, 1997.
6. R. Rosch, P. Wanke, S. Kluge, "High Pressure Die-Casting of Magnesium", Mg. Alloys and their Applications, Wolfsburg, Germany, April 1998. page 71
7. R. Mehrabian and M.C. Fleming, Trans.AFS, 80, (1972), p 175
8. R.Decker, R. D. Carnahan, R. Vining. E. Eldener, "Progress in Thixomolding®", 4th International SSM Conference, Sheffield, England, June 1996.
9. R. Carnahan, R. Decker, R. Vining, E. Eldener, R. Kilbert, D. Brinkley, Die Casting Engineer, May/June, 1996
10. Nikkei Mechanical, 1998.3 No. 522, March 1998
11. K. Kitamura, Sokeizai, Vol. 36, No. 5, 1995, p 1
12. R. Carnahan, R. Kilbert & L. Pasternak, 51st World Mg Congress, IMA, Berlin, Germany 1994
13. T. Tsukeda and K. Saito,Keikinzoku, V. 47, No. 5, 1997, p 298
14. R. Carnahan, R. M. Hathaway, et. al. Lt. Met. 1993, Métaux Légers, Met. Soc. C/M 1993
15. D. Ghosh, K. Kang, et. al. Lt. Met. 1995, Metaus Legers. Met Soc. C / M 1995
16. S. LeBeau, Y. Yamamoto, & K. Sakamoto, Mg. Symposium, SAE Detroit, MI Feb. 1998
17. R. Carnahan, 3rd International Conference on SSM, Tokyo, Japan, 1994
18. R. Kilbert, Unpublished Private Communication
19. T. Tsukeda, et.al. JSW Annual Technical Report, No. 54, 1998, p. 665
20. S. LeBeau, R. Decker, 5th International Conference on SSM, Boulder, Co. 1998
21. I. Nakatsugawa, Japan Magnesium Association Seminars, 1998

Biographies
Dr. Stephen LeBeau is Vice President of Sales and Marketing of Thixomat, Inc. He has degrees in Metallurgical and Materials Engineering from the University of Wisconsin, Rensselaer Polytechnic Institute and Michigan Technological University. Dr. LeBeau has spent his entire career in materials development and manufacturing and is a member of SAE and the American Society for Materials.

Dr. Raymond F. Decker is Chairman and Founder of Thixomat, Inc. Dr. Decker received his undergraduate and graduate degrees from the University of Michigan. His distinguished career includes positions at INCO Research as Vice President. Dr. Decker has numerous publications and patents to his credit including extensive experience in the design and manufacture of high strength, high temperature super alloys. Dr. Decker is a Fellow of the American Society for Metals.

D. Matthew Walukas graduated from the University of Tennessee with a B.S. in chemistry and an M.S. in Metallurgical Engineering. Upon graduation, Mr. Walukas worked for USP Holdings for 4 years and has been employed by Thixomat, Inc. since 1996.