TECHNOLOGY
Aluminum Engines May Stage a Comeback New alloys, new diecasting methods, new engine designs improve aluminum's performance The modern aluminum engine as standard equipment in mass-produced U.S. automobiles has been talked about for 20 years. Now, fresh approaches to engine design and manufacturing methods may be close to bringing it out of hiding. With a potential market of some 700 million lb. per year of casting alloys in view, aluminum producers have pushed development programs to meet the auto makers' show-me attitude about production economics. Results to date are prompting a shift of emphasis by these producers. Though technical problems remain, much of the effort
now is toward demonstrating feasibility of new designs in production and operation. One question pervades Detroit's thinking about a new design or material—can it do the job more cheaply than the present choice? The first wave of aluminum engines couldn't. Introduced as options in some 1960 and 1961 models by General Motors, Chrysler, and American Motors, these engines cost car buyers more without offering a clear advantage over the usual conventional cast-iron engine. Production was phased out until only GM's Corvair (the aluminum engine was always available as standard equipment), whose rear-end engine placement and air-cooled design require high thermal conductivity and light weight, is now powered by an aluminum engine. Following this failure to score, Reynolds Metals and Aluminum Co. of
MODIFIED. Harold H. Macklin, Jr. (right), of Reynolds product development division and Benjamin T. Inge, a technician, inspect a test engine modified by installing MD-57 (a Reynolds silicon-aluminum alloy) wet sleeves 34
C&EN
DEC.
20,
1965
America put emphasis on cost-cutting technology—elimination of iron sleeves in aluminum engine blocks, for example, and changes in block design to take full advantage of die-casting techniques. The auto makers themselves, though once burned, are again showing interest: Both Reynolds and Alcoa are engaged with one or more of Detroit's big four in development projects or technology exchanges related to aluminum engine production. Reynolds says it has been engaged with every division of every auto maker for seven years in intensive development of aluminum engines. Aluminum lost its first bid for the production engine market partly because of low production volume and limitations in casting methods and design. Machining or assembly of aluminum engines probably never reached peak efficiency. Optional aluminum engines, further, had to follow the iron engines rather closely in design, an expensive approach. Simple replacement of iron by aluminum in an unchanged design cuts engine weight by almost two thirds. This savings, though, is more than offset by the light metal's greater c o s t two to two and a half times the cost of gray iron for blocks of identical design. To be competitive with an equivalent iron engine, the aluminum design must take maximum advantage of high-speed casting techniques to offset the higher materials cost. Castability. Here the castability of aluminum comes into play—many of its alloys can readily be die-cast, while iron can't. Iron engine blocks and cylinder heads are cast in sand molds. The casting's walls must be made fairly thick since the molten iron, which flows into the mold by gravity, may not properly fill narrow cavities in the mold. In die-casting, molten metal is injected under high pressure into a steel die. Cavities in the casting are formed by metal cores which are automatically inserted and withdrawn. Die-casting is geared to rapid, high-volume production and can turn out precision castings with thin walls and sections.
ASSEMBLY. A worker at Chevrolet's Tonawanda, N.Y., plant begins assembling the crankcase for an aluminum Corvair engine. One by one, auto makers have stopped producing aluminum engines, until now only Corvair uses them
Chrysler die-cast the blocks for its line of slant-6 aluminum engines. Doehler-Jarvis division of National Lead Co. used the same method to make in-line, six-cylinder blocks for AMC. Minimal redesign was allowed for these optional engines, however, and the result was little more than diecast copies of thick-walled, sand-cast blocks. GM turned to semipermanent mold casting for its aluminum engines. This method, in which molten metal is gravity-poured into a steel mold fitted with sand cores, is closer to ironfoundry experience. It's slower than die-casting, though, and requires heavier casting design and more machining of the finished part. The auto industry is now working to develop a die-casting technique that makes unusually dense (low-porosity) aluminum castings. To be economically successful, an aluminum engine must be made standard equipment for a high-volume automobile line. Full advantage of diecasting techniques can then be taken to produce a design that is not only lightweight but is also competitive in cost with iron. This seems to be the consensus, based upon experience gained from production of the firstwave engines. Another costly feature of the firstwave aluminum engines was the ves-
tige of iron required in their assembly. In the usual automotive engine, aluminum pistons run against bare cylinder walls in the cast-iron block. If both piston and cylinder were bare aluminum, however, their contact would cause galling or scoring of the mating surfaces. As a result, the cylinders in aluminum engines now on the road are lined with cast-iron sleeves. Sleeves add to engine cost—$5 for an eight-cylinder block, according to one Alcoa estimate, plus the costs of labor and a slower die-casting rate. And they add to engine weight. Their elimination is now a primary goal of aluminum engine development. Both Reynolds and Alcoa see the answer as a block made of a silicon-aluminum alloy. In the block's bare bores run aluminum pistons, covered with a long-wearing material, such as iron, that is compatible with aluminum cylinder walls. Doehler-Jarvis, a major producer of die castings, disagrees. In its view, casting difficulties with the hypereutectic silicon alloys will lead to higher manufacturing cost which will defeat their use. If gray-iron sleeves are eliminated, the aluminum bores must be cast with taper and machine-stock allowance. Cutting away this surface—part of the casting's "skin"— when the bores are machined can ex-
pose porosity within the .casting. Doehler-Jarvis feels that porosity is still a severe problem that the aluminum producers apparently don't take seriously enough. The Toledo firm maintains that an aluminum engine properly designed for die-casting and containing grayiron wet sleeves can be manufactured competitively now. (Wet sleeves are inserted into the block during engine assembly instead of being placed in the die before the block is cast. ) This would permit the use of common aluminum alloys, such as 380, with which foundries are already familiar. Compromise. Choosing an alloy for a sleeveless block means striking a compromise—very hard, high-silicon alloys stand up well to piston ring wear but are more difficult to work with than are softer, low-silicon compositions. Reynolds has definitely settled on a 17% silicon alloy and is running with it, concentrating its development effort on casting and machining methods suitable for its alloy. Reynolds is trying to skirt the problems of working with the long-wearing hypereutectics. The company's automotive section chose this route when no hypoeutectic alloys passed its cold-scuff test (cold scuffing is serious scratching of the bores). An engine is brought to subzero temperature in this test, then run under controlled conditions on a dynamometer at no greater than 40° F. water-jacket temperature. Bores of the 17% silicon alloy— MD-57—Reynolds points out, show no scuffing or significant scratching from cold-scuff tests. These bores aren't coated, but they are given a special surface finish during the honing operation by a silicon-oil slurry. Silicon nodules at the finished cylinder-wall surface are in relief, rising microscopically higher than the metal surface. Road testing under normal conditions shows better performance foi* such cylinders than for iron— 0.000575 in. wear compared to 0.005 in. in one 80,000-mile test. Reynolds' cylinder-wall finish won't work with hypoeutectic alloys, which contain no primary silicon. The finishing technique is still at the laboratory stage, Reynolds says. Reynolds now favors an iron-plated piston for use with its bare aluminum bore. Developed by GM, the plating technique turns out pistons at roughly half the cost of Reynolds' previous DEC.
20, 1965
C&EN
35
choices—steel piston skirts or an ironepoxy piston coating. The Reynolds tack now is to demonstrate that MD-57 can be reliably cast and economically machined in production volume. The largest casting yet made in this alloy is a 30-cu. in. marine engine, though a 90-lb., V-8 automotive engine block will be diecast soon. Castability studies have been under way for more than a year at Southern Die Casting & Engineering's High Point, N.C., foundry. High on the list is how to control casting procedures to avoid silicon precipitation and reduce casting porosity. Metcut Research Associates, Cincinnati, has just begun a machinability study for Reynolds, aimed at prescribing the best cutting fluids and edges for use with the alloy. Both studies should be completed by the end of next year. Some commercial experience has already been gained with MD-57. S-K-S Die Casting Co. has used it for several months to cast a 12-oz. disk. The Berkeley, Calif., firm chose MD57 for its high fluidity—typical of highsilicon alloys. The casting has a particularly thin center web between rim and hub, requiring good flow to fill the die properly. S-K-S rates the alloy's castability as satisfactory—under carefully controlled conditions. Heating and holding are done in electric induction furnaces instead of the usual gas-fired furnaces. Close temperature control and constant motion prevent silicon precipitation. The disks are cast at 1350° F., somewhat higher than temperatures for common diecasting alloys such as 380. Alloys. After several years of investigation, Alcoa's Application Engineering division is still on the fence as to which approach to take. It has evaluated alloys ranging from 7% silicon to hypereutectic compositions of up to 20% (eutectic composition is 12% silicon) and has cast experimental blocks in alloys across this range. Alcoa isn't convinced that passenger-car service conditions require the wear resistance of high-silicon (16 to 18%) alloys. Hypoeutectic compositions present fewer casting and machining problems. Though its work with hypereutectics—including alloy X392 at 20% silicon-continues, the company is also giving much attention to hypoeutectic die-casting alloys containing from 9 to 1 1 % silicon. At normal operating temperatures, wear 36
C&EN
DEC. 20, 1965
rates of these alloys compare well with that of cast iron, says Alcoa. In early tests, for example, chrome-plated pistons were run in bare bores of a 7% alloy; standard pistons (a 10% silicon alloy) were run in bores in the same engine. Average wear per 100 hours was 0.00008 in. for iron, 0.00005 in. for the alloy. But problems can arise when an engine with low-silicon bare bores is started cold ( as in zero or subzero weather). Lubrication is marginal, and aluminum doesn't retain oil from previous engine operation as do the graphite particles in iron cylinder walls. Cold scuffing can occur. Whether this problem actually occurs under actual operating conditions, Alcoa says, isn't yet established. Alcoa is still working to improve the wearing characteristics of the more easily worked hypoeutectics. Cylinder Heads. Though aluminum engine blocks can be die-cast with relative ease, cylinder heads are another matter. In overhead-valve engines, which are used in U.S. cars, the heads contain valve cavities and ports as well as cooling channels. It's not possible to form a single casting with such a complex internal shape by the usual die-casting methods. An alternative is to die-cast the head in two or more sections which are then bonded by epoxy adhesives. This approach doesn't look economically promising. Epoxy-bonded heads tested by Reynolds haven't shown satisfactory reliability, and Doehler-Jarvis has found little enthusiasm for its own trial design. Reynolds says it is still working on aluminum heads. Alcoa is moving toward a one-piece head made by a modified die-casting operation involving disposable cores. This approach capitalizes on aluminum's primary advantage in head design—its high thermal conductivity. In a comparison of a production V-8 iron engine from AMC and an essentially identical aluminum engine, for example, the latter ran an average of 30° F. cooler than did the iron engine. Temperatures near the exhaust valve seats ran 100° cooler in the aluminum heads. Thus a production aluminum head could perform adequately with smaller and more simply routed cooling channels than an iron head requires. These channels are formed in Alcoa's concept by placing disposable cores within the die cavity. So far, lack of a suitable coring material bars production of such a head at reason-
able cost. Although the design gets away from the fragile and intricate cores required in casting iron heads, cores must still be removed somehow from the finished castings. Coring materials that can be dissolved or otherwise quickly removed after withstanding a die-casting cycle could solve the production cost problem. They're still being sought at the company's experimental casting facility in Chicago. Alcoa's choice of materials for casting heads is aluminum alloy 13 (12% silicon) or 380 ( 7 y 2 % ) . Heads in either of these hypoeutectic alloys would require iron or steel valve-seat inserts as well as iron valve guides. (The latter are also used in iron heads.) Successful use of the harder, hypereutectic alloys in die-casting might mean the elimination of inserts and guides from aluminum head design. Neither Alcoa nor Reynolds, however, can yet say this for certain. Weight. In the end, the case for aluminum engines rests largely upon weight. The promoters insist that a lighter engine should only be the beginning of weight saving in car design. Lighter frame and suspension, smaller wheels, lighter steering gear, and other parts can follow as a consequence of cutting engine weight—saving, according to Alcoa, at least a pound elsewhere for every pound of engine weight saved. Other factors, harder to assess, could favor aluminum. Die-casting engines, for example, require less finishing than do sand-cast iron engines; a fully automated machining line^ which mass production would justify, could show a saving over present production. Operating efficiency gains from the cooler running of an aluminum engine. Against all this is the massive investment held by Detroit in its ironcasting facilities—and that required to replace them with die-casting foundries. To carry out the weight reduction made possible by an aluminum engine would mean changes in design and production of auto parts that go far beyond the usual year-to-year design changes. Hence the auto makers' show-me attitude. A demonstrable savings per car, however, then multiplied by such figures as 21/2 million Chevrolets a year, can amortize investment at a rapid clip. Detroit plays its cards close to the vest. Changing light-metal technology may cause it to deal a big hand to aluminum.
