Automotive Casting Success Stories-2001.

These case histories from the 2001 SAE World Congress detail innovative design and production scenarios manufacturers are using to meet customer demands.

It is not an overstatement to say that as goes the automotive industry, so goes the foundry industry. With more than 31% of all castings produced in 2000 consumed by automakers, the industry's health is closely tied to the success or trials of domestic vehicle production.

The makeup of that 31% of casting shipments continues to change, however, as the automakers' engineers continue to redesign components to lighter materials in search of new opportunities for cost and performance improvement. Increased global competition and on-the-horizon government emissions regulations have spurred this redesign effort and forced the automakers and their foundry suppliers to push the limit of manufacturing and think "outside the box" for new solutions to old problems. In many cases, engineered cast metal components are providing solutions with the flexibility needed to meet these design challenges by consolidating multiple components into a single net shape, virtually eliminating the need for secondary machining and assembly processes

At the 2001 Society of Automotive Engineers (SAE) World Congress & Expo held last March, 43,000 transportation industry officials gathered to discuss some of the solutions being developed. In a session titled "Automotive Casting Processes and Materials," 19 presentations focused on engineered casting solutions. This article highlights a sampling of these solutions.

ADI Crankshaft

The TVR Tuscan Speed Six is a high-performance sports car produced by TVR Engineering, Ltd., Blackpool, England. Equipped with an in-line 6-cylinder engine, the maximum output is 420.3 Nm at 5250 rpm. Due to this stress, the engine requires a crankshaft with high fatigue strength and stiffness, Following is a look at the development of that crankshaft as described by Kristin Brandenberg and Kathy Hayrynen, Applied Process, Inc., John Ravenscroft, TVR Engineering, Ltd., and Arron Rimmer, ADI Treatments, Ltd.

In the initial design, the TVR engineers decided upon a forged steel crankshaft (Fig. 1) with a weight of 10 lb. The problem was that the forging would require the use of expensive equipment to manufacture (forging die tool, etc.) and the weight of the component in a vehicle--already looking for low weight, affordability and high performance--was unacceptable. Another option considered for the crankshaft was machined steel, but this was more time-consuming to produce and at a high cost as well.

The engineers considered a ductile iron crankshaft (Fig. 1) because the material is easy to manufacture and can be cast with more dimensional precision than steel, making it more economical. In addition, it is 10% less dense than steel, leading to a lower weight for the same given shape, and ductile iron's properties provide the added benefit of reduced noise stemming from internal damping. The problem was that the material had too low of a fatigue strength for this application. In testing at 6000 rpm, the ductile iron crankshafts failed in some cases with a fatigue crack at the fillet radius on the flywheel end of the crankshaft.

The engineers then decided to design the crankshaft for austempered ductile iron (ADI) (Fig. 1). It was a logical choice because TVR already was using it on its V8 engine, however, there were initial reservations in regard to distortion in the component's manufacture. During the heat treatment (austempering) process, the component can grow, causing it to be out of tolerance. But this growth is predictable and can be accounted for in the design stage of the crankshaft. In addition, the ADI crankshaft was rough-machined prior to heat treat and finish machined afterward to accommodate design tolerances.

The ausferrite matrix of ADI lends itself to mechanical dampening superior to that of ductile iron. Also, ADI is 10% less dense than steel, providing the same weight reduction opportunity of ductile iron. The engine designers also had to account for the thermal expansion of the ADI crankshaft in operation, which meant that their aluminum engine block for the engine was a better fit than an iron block.

In the end, TVR engineers settled on the ADI crankshaft due to its low cost of manufacture and weight reduction opportunities when compared to the steel forging. In addition, the ADI crankshaft was sufficiently strong to handle the loads of a high-performance engine (Table 1).

Transmission Valve Body

ZFGetriebe, a leading automatic transmission manufacturer located in Sint-Truiden, Belgium, faced numerous challenges when approaching the testing phase of new design programs. Certain hydraulic components, such as valve bodies, had to have mechanical and metallurgical properties that were virtually identical to the final diecast parts. Going to diecast tooling, however, was out of the question since design changes were inevitable, and the costs and lead times for diecast tooling would place the development program in jeopardy. The diecast tooling, which took 4-6 months to build, cost ZF$100,000-$140,000.

The geometry of the porosity-free casting required precision of [+ or -] 0.006 in., a wall thickness of 0.120 in., deep and narrow pockets, a draft angle of 0-0.5[degrees], and numerous through-holes with 0.25-in. diameters. These stringent requirements were forcing ZF to resort to machining such complex components from solid blocks of aluminum. For an initial batch of 25 parts, which took 24-28 hr/part, the cost of machining totaled $26,000 with subsequent batches costing $20,000. Aside from high cost and long lead times, the mechanical properties of machined aluminum were different from cast aluminum, and the test results of the transmission were not precise enough.

In 1996, ZF brought a transmission component for a small passenger car to Nest Technologies, Inc., Studio City, California, which used the Precicast sand/ceramic casting process, a technology developed for prototyping and "bridge" production of complex parts that are normally diecast. In this process, a thin layer of ceramic material is applied by vapor deposition on the surface of the tool and then backed up by a highly compressed sand and resin mix. This combination of ceramic and sand enables production of complex geometry such as deep and narrow pockets, holes, thin walls, tall pins and fins, as well as parts with minimal draft angle.

After receiving the transmission CAD data from ZF, Nest began the project with a full thermal and mold filling simulation in order to minimize liquid shrinkage and eliminate porosity. The computer simulation predicted if a "cold-shut" would occur and allowed the gating and feeding systems to be dynamically modified until a perfect result was achieved. Once the tooling was designed in CAD, it was machined from high-density plastic board to a polished state. The tooling then was mounted on a fully automated computer controlled sand/ceramic molding machine that has a production rate of up to 30 molds/hr. Casting was done using a "stopper-rod" pouring system and a special degassing technique that eliminates porosity. The final step of the process was a three-stage heat-treatment process that uses fluidized bed technology to achieve mechanical properties similar to those achieved with diecasting.

ZF's latest transmission design required a valve body with tolerances of [+ or -] 0.006 in. to assure proper function of the valve. The numerous deep pockets were 0.120 in. wide, and the holes for screws had to be held to a tolerance of [+ or -] 0.006 in. to eliminate secondary machining operations. The first two evaluation parts (Fig. 2) were delivered to ZF within 4 weeks after receipt of CAD data, and subsequent batches of 25 parts were produced for various testing phases in I hr. A total of 300 parts were manufactured before ZF authorized the production of the final diecast tooling. At Nest, the cost of producing the first batch of 25 castings was $18,000 (a 31% cost reduction) with subsequent batches at $5000 (a 75% reduction).

"The precision and reliability of the sand/ceramic casting process helped us successfully continue our new design programs," said Richard Glare, ZF's director of research and development.

Clutch Spider Casting

New casting methods have opened the door to new casting opportunities requiring greater mechanical properties and finer microstructures. In a presentation by Alan Druschitz, Thomas Prucha, Adam Kopper and Thomas Chadwick, Intermet Corp., three innovative processes- squeeze casting, rheocasting and semi-solid casting-were compared for the production of a clutch spider casting (Fig. 3) for Arctic Cat in secondary 356 aluminum alloy.

Squeeze casting is a diecasting process that uses 100% molten metal and relatively slow shot velocities. At a designated time during casting solidification, high pressure is applied to "squeeze" the last liquid metal into any remaining porosity. As a result, a denser casting can be produced.

The semi-solid casting process uses a specially solidified solid billet as a starting material. The solid billet is heated until it is about 50% liquid and then injected into the casting cavity. The unique flow characteristics and low heat content of the semi-solid metal produce cast components with little to no porosity, and die life is improved.

In rheocasting, molten metal is cooled in a controlled manner until a special microstructure is formed. When the metal is about 50% solid, it is injected into the casting cavity. As in semi-solid casting, the material has unique flow characteristics and low heat content. The resulting castings exhibit little to no porosity, and die life is improved compared to squeeze and diecasting.

The spider clutch castings were produced using all three processes (one at an Intermet facility and two at supplier facilities) and then underwent T5, foundry-specified T6 and supplier-specified T6 heat treatments. Once the processing of the castings was complete, chemical composition, microstructure, hardness and tensile properties were compared.

Chemical Composition--The chemical compositions were representative of secondary 356 aluminum for the squeeze castings, A356 for the rheocastings and A357 for the semi-solid cast components.

Microstructure--The primary difference between squeeze and semi-solid casting is the dendritic nature of the squeeze cast material and the non-dendritic nature of the semi-solid casting. The semi-solid material showed the classic microstructure, which consisted of well-rounded primary aluminum (Al) particles surrounded by Al-silicon (Si) eutectic and entrapped AlSi eutectic within the primary aluminum. The rheocast material was similar but no entrapped Al-Si eutectic was apparent within the primary aluminum. The Si was well modified with the semi-solid material, non-uniformly modified in the rheocast material and poorly modified in the squeeze cast material.

The castings produced by the squeeze casting process showed regions of eutectic pooling, which is caused when auxiliary densification is applied and remaining eutectic liquid is forced into low-density areas that are surrounded by material that has already solidified. As a result, large defects (porosity) are eliminated and mechanical properties are significantly improved.

After a T6 heat treatment, the squeeze cast material had non-uniform Si morphology, whereas the semi-solid cast and rheocast materials had well-spheroidized Si. The elongated, but thin Si particles produced by the rheocast process broke down rapidly during solution treatment.

Hardness-The hardness of the castings reflected the effects of both chemical composition and microstructure (Table 2). The higher level of impurities and high magnesium (Mg) content in the B356 aluminum alloy resulted in higher average hardness for a fixed heat treatment. The rheocast material had the lowest hardness due to low impurity content and low Mg content.

Tensile Properties--The tensile properties correlated well with hardness and showed the effects of defects (Table 3). The data for the current T6 heat treatment clearly shows the superior consistency of the semi-solid material.

Based on these properties and other information gathered in the comparison, it was determined that:

* castings produced by the two semisolid casting processes had a better combination of strength and ductility and more consistent tensile properties than castings produced via the squeeze process;

* given the same heat treatment, castings produced by the squeeze process had higher hardness than castings produced by semi-solid processes. This was due to the differences in chemical composition and microstructure.

Mirror Bracket Prototypes

The need for automakers to slash the time-to-market for their products has placed increased emphasis on rapid prototyping technologies for automobile components. Engineers often require a component to test for form, fit and function before full-scale production. A presentation by Sherif Nasser, Tekcast Industries, Inc., explored functional automotive prototypes delivered by spin casting. This process uses silicone rubber molds to provide zinc or thermoset plastic components from any type of computerized model within hours. In production, the molds are rotated at 200-1000 rpm and molten metal or liquid plastic is poured into the center of the mold and carried to the mold cavities on the edges of the mold by centrifugal force.

One of the case studies presented was for a General Motors remote-control rear view mirrorbracket (Fig. 4). Detroit-based Pinnacle Technologies, a full-service rapid prototyping service bureau, was contracted by a GM sub-contractor to provide stereolithography models and 200 zinc spin castings of these complex components for operational testing.

First, the components were spin cast into pewtersub-masters using teflon pull cores. These sub-masters were used to create a four-cavity mold in silicone rubber in which 200 zinc components were produced. Total time for production of these prototypes was 1.5 days.

The OEM then made several design changes to the component after the initial prototype was delivered. Hundreds of additional parts were cast for further evaluation and mounting on automobiles for road testing. The total cost per prototype (including mold, casting material and labor) was $100.

Table 1. Mechanical Property Comparison
of Steel, Ductile Iron and ADI
Crankshafts, and the ASTM 897-90
Standard ADI Grade I Specifications
                         Steel   Ductile Iron  ADI   ASTM ADI Grade 1
Yield Strength (Ksi)      107         78       120          80
Tensile Strength (Ksi)    132        131       157         125
Fatigue Strength (Ksi)    58          47        62         n/a
Impact Energy (ft-lb)     240         55       104          75
Elongation (%)           23.2        10.8      13.7         10
Hardness (BHN)          226-266    262-277     300       269-321
Table 2. Hardness Comparison of Spider Clutch Castings
    Heat treatment       Casting process  Hardness (BHN)
          T5              Squeeze cast         91.8
                         Semi-solid cast       87.5
                            Rheocast           80.6
     T5 (pre-age)           Rheocast           87.1
T6 (foundry-specified)    Squeeeze cast       103.8
                         Semi-solid cast       92.6
                            Rheocast           89.9
T6 (supplier-specified)  Semi-solid cast       96.4
                            Rheocast           87.6
Table 3. Tensile Property Comparison of Spider Clutch
Castings
         Heat                Casting      Ultimate tensile
       treatment             process       strength (psi)
          T5             Semi-solid cast         42
                            Rheocast             38
     T5 (pre-age)           Rheocast             41
T6 (foundry-specified)    Squeeze cast           43
                         Semi-solid cast         44
                            Rheocast             42
T6 (supplier-specified)  Semi-solid cast         44
                            Rheocast             44
         Heat                Yield       Elongation
       treatment         Strength (psi)     (%)
          T5                   31           7.9
                               27           6.7
     T5 (pre-age)              30           7.7
T6 (foundry-specified)         38           4.2
                               37           9.6
                               34          12.3
T6 (supplier-specified)        38           8.9
                               31          15.6

The Changing Makeup of Automotive Casting Use

The car and light truck market consumes nearly one-third of all castings produced by the U.S. foundry industry. Although car and light truck production for 2001 is projected to contract to 11.3 million units (down 10% from last year), long term projections show an annual growth rate of 2.9%/year through 2010 according to Stratecast, Inc., Fort Myers, Florida.

A breakdown of casting shipments by metal to the automotive and light truck industry by Stratecast shows the following potential growth opportunities for the present and long term:

Ductile iron-The demand for ductile iron is forecasted to increase 0.4%/year until 2010. With shipments forecast at 1.06 million tons in 2001 the shipment totals are forecast to 1.178 million tons in 2010. Ductile iron is losing applications to aluminum due to weight reduction demands. It is, however, expected to continue to replace malleable iron components as well as serve as the material of choice for some brake system components. Austempered ductile iron is expected to replace some applications currently produced by forging. The weight of ductile iron castings per car and light truck is forecasted to decrease from a high of 190 lb this year to 170 lb in 2010. Growth applications for ductile iron will include crankshafts, gears and brake system components.

Compacted Graphite Iron (CGI)--CGI is expected to grown from 32,000. tons in 2001 to 54,000 tons in 2010. Chrysler has specified CGI for bedplates. for its 4.7 liter V8 truck engine.

Aluminum--The demand for aluminum castings in motor vehicles is forecasted to increase from 1.2 million tons in 2000 to 1.898 million tons in 2010. This growth translates to an increase of 85 lb per light vehicle (to 270 lb) in 2010. It is forecasted that 49% of the weight is to be in die castings. This high growth is due to an increase in engine cylinder block and head applications for aluminum.

Magnesium-Demand for magnesium die castings is expected to increase from 48,000 tons in 2000 to 263,000 tons in 2010, an annual growth rate of 18.5%. Magnesium consumption per car is expected to reach 15 lb in 2002 and 30 lb in 2008. The growth in magnesium is forecast to come in several applications, including: clutch housings, wheels, seat assemblies, motor housings and transmission cases.