Pierce-Arrow Society Feature Article
Early Pierce-Arrow Cast Aluminum Body Technology
by Roger Sherman
The completed cast aluminum body on a World War I era 38-C-4 Convertible Coupe
(University of Michigan Libraries)
A Unique Solution to a Universal Problem
For sixteen of its thirty-seven years (1901-1938), Pierce-Arrow used cast aluminum for the entire structure of their automobile bodies. It was a process unique to the Pierce-Arrow Motor Car Co. The necessity for this innovation was the same as that which forced the simultaneous development of the conventional automobile body technology of the time: stresses brought on by the operation of a powered car on the roads of the early twentieth century damaged the structure of the carriage-derived bodies on early cars. Even the vibrations of the automobile’s machinery strained body joints and caused unexpected distortions in the wood, accompanied by alarming squeaks and shudders.
The Composite Body
The conventional solution to these problems was to build up a sturdy framework of hardwood, to which metal outer panels were attached. The completed body was then mounted on the chassis of the car. This so-called “composite” body was the standard method used to construct car bodies for the next thirty years. Pierce rejected this solution for more than a decade.
The Pierce Unorthodoxy
In 1904, after three years of intensive development, the newest motor cars built by the then George N. Pierce Company of Buffalo, New York had become front-engined Panhard-style chassis with channel section frames carrying powerful four-cylinder water-cooled engines, thereby joining the ranks of companies producing what became the standard layout for automobiles. These imposing Great Arrow models were aimed at the luxury car trade, and the company was determined not to market them with flimsy coachwork. Although this onetime bicycle manufacturer had no experience in carriage construction, and in contrast with most automobile manufacturers of the time, Pierce resolutely undertook to develop their own body building capability, and set out in a whole new direction.
Ten years later, while speaking to a dealer conference at the factory, Herbert M. Dawley, one of the important influences on the Pierce-Arrow cast aluminum body, remarked that “The cast aluminum body is a thing that is distinct and unique in the Pierce-Arrow factory. It is the one thing that has been developed that you will not find in other cars.” At the time he spoke these words Dawley knew the Pierce-Arrow car occupied the very summit of prestige in the luxury car market, seen in enclaves of social prominence and the loci of great financial and governmental power that included the White House itself. The company’s earlier decision seemed quite justified.
The man who developed the cast aluminum body at Pierce came to the firm in 1904 from Brewster & Co., Manhattan’s most renowned carriage and automobile body builder. His name was James R. Way, and his idea was to use cold rivets to unite cast aluminum panels together to make a complete body structure. While the concept was simple enough, it took years to actually perfect the technology.
When, faced with unprecedented success, the George N. Pierce Co. built its forward-looking automobile factory up on Elmwood Avenue in 1906, the Body Building was its tallest structure. Eventually, as enlarged, that building would be the plant’s longest as well. In it were spaces for finishing the aluminum panels; cutting and fitting the wood pieces used for tacking strips, seat bottoms and floor boards; preparing and painting the assembled bodies; installing upholstery and fittings and mounting the bodies to the chassis. Over time methods changed, but this body factory supplied the vast majority of bodies for Pierce-Arrow until the company went out of business in 1938.
What follows is a description, taken from contemporary sources of the exacting operations used to produce the remarkable cast aluminum bodies that clothed Pierce-Arrows in their golden years before the start of the Great War.
Designing the Body
To begin with, as with any motor car body, some sort of general idea about the body’s structure and lines was developed. At Pierce this was done by Mr. Way and his assistants in the Body Department. At first, difficulties casting large body sections limited the possibilities, but that was overcome by 1911. Refinement of foundry practice and development of suitable alloys allowed larger and more complex castings.
Once the general shape was determined individual body parts were modeled for casting. From these sample parts patterns were developed for making the casting molds for production. Master patterns for each part were sculpted from wood laminate, and then used by the foundrymen to make two metal surface patterns—one for the front of the casting and another for the back. These were used to form a casting of that particular panel. The master patterns were a crucial item because they had to accommodate all the requirements of the part. For example, each production casting shrank as it cooled, and the patterns had to compensate for that.
Making the Body Panels
As was true for most automobile firms in those early days, Pierce-Arrow had no foundry of any kind. They machined the cast parts received from vendors to their specifications. The vendor for body panels was the Elmwood plant of the Aluminum Castings Co. in Buffalo who then shipped them down the street to Pierce-Arrow for final finish and assembly. Factory records indicate that the time it took to “order, make and deliver alum. castings” [sic] for a body took three or four days.
At the foundry, the patterns were used to form the mold for the body panel in sand, the finer the sand the more precise the shape. This casting sand was held together with a “binder” that allowed the pattern to be “drawn” (pulled away from the mold) without disturbing the shaped sand. To help the sand retain its shape the patterns were “rammed” by hand into the molding sand on the hard floor of the foundry, and the “skin” of the shape in the sand was dried with an oil torch after the pattern was drawn. Each sand mold was held in a box called a “flask”, and it took two of them, fitted together, to hold the production molds for the front and back of a casting.
The first challenge of this procedure was to achieve the thinness of metal required (about 1/8th of an inch, although thickness varied with expected strains on the panel). The second challenge arose from the large size of many of the castings. In 1915, the largest body part produced with these methods was the roof panel for a coupe body, which stretched 5 ft. 6 in. on one side by 4 ft. 6 in. on the other. Molds for parts of such size required as much as two cubic yards of sand for the mold. The flasks carrying these molds were, naturally, very heavy (the pieces that came out of them, alone, weighed as much as 30 lbs.) and required extra bracing of wooden ribs to support the sand. In addition, the cast part could vary no more than one one-hundredth of an inch from that of the pattern part, so precision was required as well.
Molding the Parts
The next step was to pour the molten aluminum into the mold. This was done by hand through holes in the top flask (known as the cope). These holes, called sprues, connect inside to gates leading the molten metal into the mold itself. So that air inside the mold can be pushed out by the incoming metal, there are additional vent holes (known as risers) as well. Special pots with tapered sides and rounded bottoms were commonly used to heat the metal and then pour it into the sprue holes of the flasks. These pots are called ladles. Some of the Pierce-Arrow body castings required several ladles to pour metal in various sprues at once in order to completely fill the mold before the metal started to cool. Pouring the molten metal properly is a complicated and exacting task. Instead of heating each ladle for this pour, a crucible might have been used to fill them with molten aluminum. The way the metal was poured depended on the shape of the piece being cast. Flat pieces would be poured with the flask at an angle of 30 degrees from the vertical using a single sprue hole and two gates. Curved surfaces were poured flat through as many as five sprue holes each provided with four to six gates. A very complex pour was for the dash/cowl, which was poured flat with the curve extending downward, using four sprues and 26 gates without risers except the sprues themselves.
Once the metal is poured it is important that castings cool evenly, so round pieces of metal called chills were placed near the surface of the casting sand of the mold near thicker parts of the casting. These absorb heat rapidly from those places allowing the entire part to cool evenly. After two to five minutes the flask was shaken out and the new casting left to cool in the open air. At this point the metal shrinks at the rate of a quarter inch per foot cast. The roof panel casting mentioned before would shrink from molten metal to room temperature about 1 3/8 in. by 1 1/8 in., so getting it away from the restraint of the sand is important because the stresses could cause cracks in the piece as it cools.
Once the casting cools the extra metal left in the gates from the pouring is cut off, as are the “fins” that form along the lines where the flasks join. At this point the castings traveled over to Pierce-Arrow to be finished and assembled.
Assembling the Body
Upon entering the Pierce body plant castings were inspected and weighed. Each panel’s weight could not vary more than 10% of the weight of the aluminum pattern for the piece. Then, the panels were inspected for shape, and defects were hammered out with rawhide hammers, wooden mallets, or even steel hammers. Wooden forms were used to help achieve the correct shape. Having now produced a standard part, the process began that would give the cast body panel the right surface for paint and varnish. Next stop in the body factory was the filing room, where the irregularities from the casting sand were removed with a file. Final finish of the cast panel was achieved with an emery cloth.
Next, the body was assembled on a large, accurately level plate, using jigs to determine the exact position of each panel. Pierce-Arrow designed beads into the castings in strategic places to allow for an overlap, aiding this assembly. The joints were finished so exactly that they could not be seen. The steel rivets were hammered cold to fasten the panels together because hot rivets would melt the aluminum. The diameter of the countersunk head of these rivets was 5/16 in. thick, the shank being approximately 3/16 in. According to company documents, this work to finish and assemble the body castings took one and a half to two days, depending on the complexity of the body.
The completed body was then moved to the wood shop where, using only high-grade white ash, the sills, seats and a bit of reinforcing were added and doors were hung properly. This work took twelve to twenty-six days to complete.
So strong were the assembled bodies that stories abounded of their performance in accidents. In one case mentioned in the Hub article a closed Pierce-Arrow body survived relatively undamaged after being struck by the moving tender of a steam locomotive. Although the car was thrown from the road and down an embankment, the four passengers were, surprisingly, unhurt.
Preparing the Surface for Paint
When assembly ended, any holes on the surface of the castings were filled with solder, and the body was moved to the “First Paint” shop. From now on the efforts were to finish the body inside and out for mounting on the chassis.
First step was to coat the metal and wood surfaces with a layer of primer. (The ends of wood pieces had already been primed to prevent moisture from being absorbed by the fibers of the wood.) Then, the surfaces of the body to be painted were covered with a coating of “rough stuff.” This heavy material, described by Herbert Dawley as “educated mud,” was the foundation for the carefully-manicured finish of the car. After drying, the coating was sanded and rubbed down to make “an extremely fine surface upon which to apply paint and varnish” [Dawley]. The process took from six to eight days to complete. The body was now ready for the attentions of the Second Paint Department where the first coats of color were applied. Like all the other applications of finish, the varnish, with desired color mixed in, was brushed on by hand, using specialized brushes suited to the material applied. For lakes, or transparent colors, such as English purple lake or wine, four coats were applied. Darker, or opaque, colors required only three coats of color. To reduce the visible brush marks successive coats were applied at a 90 degree angle to each other. The color coats, like all other applications to the finish, were carefully dried in ovens at 125 degrees F for an hour and a half, while clean, properly humid air circulated constantly. This allowed some certainty about the time it would take to finish drying the varnish.
The first coat of “rubbing varnish” was then applied, and oven dried for six hours. Nitrocellulose lacquer paint was a decade or more away from development, and all automobile factories had to allow for the days needed to apply a durable finish. The wider the selection of finish colors, the more time the drying took, which is one reason why Henry Ford came to prefer painting all his cars black.
Installing the Interior
When the first coat of varnish had completely dried, it was carefully “rubbed out” with “rotten stone” (a kind of pumice) for a smooth finish. The body’s next stop was at the Trim Shop where orders had already been carried out for cutting the material and men assigned to do the work. Glass and any needed wire was installed, as were disappearing seats where ordered. Upholstery was now built up. Springs of particular design were tacked in appropriate places for the given seat. Over them cushions were draped, previously made up on a frame with canvas stretched over it. Cushions were next inserted in a special machine where curled horse hair was stuffed into them and the machine put in the tufts, leaving the ends loose for the trimmers to finish. Placed in the body, the seats were packed with the hair, working from the bottom up into the cushions themselves. Over time, the trimmers developed a “feel” for the correct shape. The company commonly built particular seats to order and had a device that could furnish the required dimensions for anyone who sat in it for measurement. Dawley called it his “electric chair.”
Applying Final Paint and Varnish
Now trimmed, the body went back to the Second Paint for striping and application of the second coat of rubbing varnish, which was dried as before and rubbed out. The finish was completed with the application of flowing, or “bright” varnish, the clear coat of its day. The painters who brushed this on had to make a perfect “flowing” application, leaving no brush marks. When the body was sent for its final six hours in the drying ovens let us hope that no insect strayed onto the drying finish. Should that happen, the surface has to be sanded down and started over again.
Up to the time when the first application of rough stuff was applied, the particular body, let us say a closed limousine (a “Suburban” to Pierce), had taken thirty days to complete. The final finish and trim took another twenty-two days. Final assembly, final fitting, inspection and testing would take two days more. In other words, producing each car took nearly two months. Although Pierce-Arrow always insisted that their cast aluminum bodies were superior in strength, quiet, longevity and lightness to comparable conventional bodies, this labor intensive process compelled them to turn in 1921 to a simpler composite body using sheet aluminum panels. In 1929 the body panels were changed to steel.
The unconventional cast aluminum bodies built by the Pierce-Arrow Motor Car Company in the first two decades of the twentieth century pushed the aluminum casting technology of the time to unforeseen achievements in size and fineness. As motor car bodies they were exceptionally strong, silent and durable. Their finish was superb, as was their riding comfort. However, the tremendous cost of the hand labor used to create, assemble and finish them proved to be too much for the company to bear in the postwar market of 1920 even for a maker of cost-is-no-object automobiles. They were, therefore, successful in a structural sense while failing as a lasting technology in the automobile industry.
The operations described in this account were detailed in “The Body and Its Refinements” from The Pierce-Arrow Salesman for October 1914 and “Making Aluminum Automobile Bodies” from the July 1914 issue of The Hub, a carriage and body industry magazine. Further details of automobile body construction and finish appeared in the other sources listed. The author also wishes to thank Fay Butler, auto restorer and metalworker, for explaining the specialized terminology used and checking the manuscript for errors.