The Origins of Wagstaff Inc.: Part 1—Two Emerging Technologies Collide

Abstract

The journey from a machine shop to a position of a global leader in Direct Chill (DC) aluminum casting technology began when a distressed father with six mouths to feed purchased a new lathe for the backyard shop housed in the pump house and paid for the lathe by machining components for the lathe manufacturer. Eventually, his two sons, Bill and Frank Wagstaff, would focus their life visions at the root of two developing technologies: (1) the dawning of computer-controlled machine tools, and (2) the emergence of DC casting equipment to produce semifinished aluminum products, i.e., extrusion ingot and large format rolling slab ingot, as the dominant process after WWII. In doing so, they merged their skills in machine tool design and materials research into a business known and recognized across the world as the premier molten aluminum solidification technology.

In the Beginning

In 1946, George Brinton and his wife Laura Everton Wagstaff started a small machine shop in a 12’ × 16’ backyard building that doubled as a pump house with the hopes that it would help them make ends meet. They had no idea that their actions were laying the foundation for the future of Direct Chill casting for the aluminum industry. George wanted to purchase a lathe for their business, but in the aftermath of World War II, it was impossible for him to secure a business loan. Therefore, George made a manufacture-to-own agreement with the lathe manufacturer to machine components for them in payment for the equipment. Work at the machine shop was a family affair. At young ages, George and Laura’s two sons, Bill and Frank, learned to clean and care for the machines followed by “the boys” taking on increasingly challenging special projects that taught the fundamentals of machine operation, which gave them an opportunity to gain a valuable perspective into machine tool design. They learned early how to use the skills they were developing to problem-solve and to discover better ways of doing things when they initiated projects such as retrofitting an old automobile starter motor onto their bicycle.

At first, George worked his day job in his new machine shop and put food on the table working the swing shift at the Kaiser Aluminum Trentwood Rolling Mill. Eventually, the upstart machine shop served several road building contractors with the Roy L Bair Company as the largest account. The Bair Company maintenance superintendent had such confidence and trust in George Wagstaff that at the end of each month, he brought his wife with him to the Wagstaff’s for a business and social visit and sitting at the kitchen table he would write a work order, after the fact, for each job completed according to the timecards accumulated during the month.

George and Laura’s first son, William (Bill) George Wagstaff joined the upstart machine shop as the third employee in 1958 to work alongside George and Wendel Olsen. Bill had just graduated with a Mechanical/Manufacturing Engineer degree (BS) from Utah State University (USU) in a slow economic cycle when jobs in his field were few and far between. Bill thrived at the shop with new challenges brought to the shop daily. This constant regiment of problem and resolution brought satisfaction to the young engineer as his innovation cycle was refined day to day and week to week. These energies eventually drove Bill out of the shop into the business world in search of a more stable product for the market that would bring value to the customer and that would highlight the fledgling’s developing skills at difficult problem-solving.

Wagstaff as Aluminum Casting Mold Supplier

Eventually, Wagstaff developed a relationship with the Kaiser Corporation where a short Direct Chill (DC) casting mold used in producing round extrusion ingot (billet) was in the works. In 1962, Bill met Pat Davy, who quickly established a relationship with Bill and via “back of the envelope” sketch, outlined his scheme for a new mold. The key to the Kaiser approach was not only the short bore, but also the use of a common pressured water box to deliver water sprays to the backside of the billet mold and the emerging surface of the hot billet. The challenge was the precision and accuracy needed to drill a series of 7,000 very small (3/32″) holes in a confined space at a 30-degree angle without reference datum to start the drill (Fig. 1). This first step was seen as a reckless endeavor by George as the other high-level machine shops, Central Machine and General Machine, both well respected in the community, chose to “no quote” the project.

Fig. 1

Early mold cross section a fabricated mold, b water-slot mold, and c water-holed mold. Note that a slot mold refers to a gap in the mold that forms a continuous water curtain around the ingot periphery, while a hole-mold has an array of holes around the periphery to generate the water curtain

Before long, this unique capability of drilling a multiplicity of small, angled holes began to catch wind and move beyond the round extrusion ingot DC mold circuit. Invitations appeared for the young, passionate engineer to visit other customers, including some that produced rectangular rolling ingot. Cooling the rectangular ingot mold was much the same as for the billet but employed a picture frame spray ring of rectangular tubing with holes to direct water against the back of the mold wall and through a slot to the emerging ingot. Bill became well acquainted with the challenges of fabricating casting mold assemblies and holding the necessary tolerances to yield a quality product. After much thought, he proposed a completely machined water-jacketed mold cut from a solid aluminum block. This was such a radical departure from the current technology that the others probably thought this young engineer was a bit of a dreamer. However, he proposed building a machined test mold for the same price as a fabricated mold, and Kaiser agreed to his offer probably thinking they would never see a machined mold. Bill laboriously machined a mold from a block of aluminum that performed so well that the skeptics became believers in the success of the new technology. About the same time (1966–67), Boeing Aircraft Company was replacing a series of hydraulic tracing profiling machines manufactured by the Cincinnati Milling Company. The machines, which had been used during and after the Korean War to manufacture airframe components, were being sold for salvage. These hydraulic profiling machines were capable of machining large, two-dimensional geometries, roughly 4’ × 12’ for the aerospace industry, and this envelope seemed perfect as a mechanism to provide a more reliable mold section in contrast to the bent extrusion-based structures commonly used in the industry of the day. These profile mills employed hydraulic positioning to follow a cutting path parallel to the stylus movement on the tracing pattern. The pattern for the mold bore and jacket was a ½” aluminum plate with a series of flat, cold-rolled steel segments screwed to the plate to define the geometry.

Figure 1a shows the cross section of a fabricated mold using right angle aluminum extrusions for the sides and ends, welded in the corners, and bent to form the contour for the as-cast ingot surface. The water pipe rectangular steel tubing with water passage holes on the inside was attached to the underside of the aluminum extrusion fabrication, and then an aluminum plate cover was attached to the bottom surface forming a water-slot. This formed a very good mold in principle but it was difficult to manufacture and presented a challenge to meeting the necessary tolerance requirements.

Figure 1b shows the cross section of an integral jacketed water-slotted mold machined from a solid aluminum block and machined to a pre-defined contour giving much tighter tolerances than could be achieved with fabricated molds. The internal baffle, which can be seen in this cross section, is used to create a uniform flow condition around the mold, thereby negating any flow inlet effects which might disrupt the uniform water flow.

Figure 1c shows a cross section of an integral jacketed water-holed mold made in the same manner as the water-slotted but employing individual water jets instead of a continuous sheet of water to cool the emerging cast ingot. The big advantage of this design was the rigid attachment between the mold wall and the jacket wall to minimize mold wall deflection. Each of the different rolling slab ingot producers had their own “proprietary designs,” tailored to their products and casting practices.

As new companies entered the slab casting world, the water-slot design was dominant for customers anticipating relatively small or low-aspect ratio ingots and planning to cast low in the mold. Since water-holed mold designs employed discharge holes for cooling water impingement onto the exiting ingot, Wagstaff designed a unique machine for drilling multiple holes with accuracy and synchronized precision center spacing around the mold bore. This machine automatically followed around the mold and was nicknamed the “Creepy Crawler” since it slowly followed the mold bore drilling holes autonomously.

Computer-Controlled Machining Gives Wagstaff a Manufacturing Edge

George and Laura’s second child, Frank Everton Wagstaff, graduated from University of Utah (U of U) with a Doctor of Philosophy in Ceramic Engineering and with a minor in Metallurgy (1962). After serving an Air Force commitment (1962–1965) at AeroSpace Research Labs at Wright Patterson AFB in Ohio, he went to work at the General Electric (GE) Research Laboratory, in Schenectady, NY, which at the time was considered one of the most high-tech labs in the country. While in the Air Force, Frank had the opportunity to see some of the earliest versions of Numerical Control (NC) manufacturing used to machine aircraft engines. With his childhood experience as a backdrop, he was simply “amazed at what could be done with a computer.” Then, during Frank’s tenure at General Electric (1965–1968), he was introduced and encouraged to use computers in his work. He learned FORTRAN and developed these skills further in his work at GE. Frank joined the family enterprise in 1968, bringing some of the new developments in ceramics and metallurgy, as well as his newly developed high-tech knowledge and skills, to the aluminum industry.

The aluminum molds and bottom blocks currently being machined on the hydraulic profile tracing machines at Wagstaff were logical candidates for upgrading to NC machining if the technology could be acquired at costs much less than the expensive machining centers and mainframe computers currently used in the aerospace industry. In 1969, the economy was in a recession, and Boeing was cutting back on machining resources and selling some of its oldest NC machines. These machines had large vacuum-tube, room-sized controls but had the possibility of being retrofit by keeping the electric-powered horizontal cutting tool spindles, upgrading the hydraulic axis positioning components, and installing new, small, solid-state controls mounted at the operator’s station on the machining center. With excited anticipation, Bill and Frank purchased their first salvaged, large Kearney Trecker Machining Center (Fig. 2).

Fig. 2

Kearney Trecker machining center

Tektronix, the oscilloscope manufacturer in Portland, Oregon, was selling a Cathode Ray Tube (CRT) cutter path display unit used in visually testing NC tapes before cutting chips on the machine. To go along with this testing equipment, a new line of small, solid-state NC machine controls was being introduced. Tektronix agreed to work with Wagstaff and supplied the machine controls for all the retrofit machines, which was beneficial for both companies.

General Electric had just begun offering a basic, time-shared NC parts programming software that met Wagstaff’s requirements. The system was a teletype workstation transferring the programming and data input over telephone lines to the mainframe GE computer, which performed the calculations and then transmitted the NC instructions back to Wagstaff’s teletype station, in the form of a punched paper tape input for the NC machining center. With all the pieces in place, Wagstaff could now do very sophisticated NC machining. Within a year after the first NC milling machine became operational, a new NC programming system using a minicomputer became available. The original programming language, Automatically Programmed Tool (APT), which ran on a large mainframe computer, was modified to run on a small minicomputer. United Computing, Inc. developed its version, UNIAPT, which ran on a Digital Equipment Corporation PDP-8 Minicomputer. The PDP-8, 12-bit CPU was configured to perform calculations comparable to large computers, but at slower rates. Although these minicomputers were very slow compared to the powerful smartphone computers carried in our pockets today, the onsite minicomputer had plenty of capacity and could crank away unattended. It got the job done and fit well into the overall manufacturing process. With time, a generalized “Family of Parts” program was developed that could produce an NC machining tape for any mold configuration offered by Wagstaff by simply inputting the dimensions and choosing the desired options. This newfound freedom from dial-up gave engineers and programmers an opportunity to test complicated tool path geometries and resolve problems while enabling the group to literally own their process.

Wagstaff’s Tru-Slot and Super Tru-Slot Molds

It wasn’t long before Wagstaff had perfected and manufactured their own mold designs. Figure 3a shows the cross section of a Tru-Slot mold, which is a design that takes advantage of both the water-slot and the water-hole designs [1]. Individual holes are drilled in the mold body much like the water-hole design, but the rigidly attached cover between the jacket wall and the mold bore forms a converging slot with the mold tang to create a sheet of water that cools the exiting ingot.

Fig. 3

a Tru-Slot Mold, b Super Tru-Slot Mold. Note here that distinct holes are drilled into the mold body, but the jets converge into a single curtain creating a water pattern emulating a slotted mold

Figure 3b shows the cross section of a Super Tru-Slot mold that offers an enhanced level of mold wall cooling where the individually drilled holes running through the thick mold wall increase the cooling and decrease the likelihood of an air pocket forming in the top of the mold, which can disrupt the heat transfer through the mold wall. This mold became the quality standard for conventional sheet ingot casting [2].

The Intalco Smelter in Ferndale, Washington, owned by Alumax, Pechiney, and Howmet, opened in 1966 as the first disconnected smelter that produced slab for a diverse market of rolling mills with very high expectations in Japan. As a new operation with no past that had to be abandoned or changed, Intalco exhibited an anxious and enthusiastic readiness to be at the forefront as a leader with the newest technology and practices. They started with the integral jacketed water-slot mold. Then as the Tru-Slot and later the Super Tru-Slot came along, they were always among the first users. With intalco smelter located only a few hundred miles from Spokane, there were frequent visits and close associations that developed a strong business relationship. Wagstaff had entered the ingot mold tooling business with a mold, block, and adjacent technologies that established a new level of operational quality and precision in the industry.

With a mature NC machine shop and time available on some of the machines, Wagstaff made the decision to add contract machining to increase business and diversify the customer base. Boeing, Peterbilt, Melrose, and Nelson Irrigation were the primary customers, and Wagstaff’s NC machining capabilities were a good fit for their needs. However, suffering through economic downturns and a recession, Wagstaff’s position as a supplier became obvious. The companies Wagstaff served had their own machining resources, which they kept busy and used contract machining to handle the extra demands when needed. When times were tough, orders were not just reduced but cancelled completely. When an official from one of the companies cancelled their orders, he suggested to Wagstaff that they should strive to find a niche where they had their own technology and equipment to sell as a package, and not just be a builder of equipment. This comment spurred ideas of possible unique Wagstaff products that could be marketed to solve common customer problems and sold throughout the industry.

Wagstaff Enters the Direct Chill Casting Technology Market (MaxiCast and AirSlip)

The Wagstaff company enjoyed a comfortable relationship with Direct Chill Machine manufacturer, Loma Machine, as well as molten metal distribution pan manufacturer, M&T Manufacturing, with Wagstaff being a key supplier of the mold-supporting, water-cooled table with its corresponding starting head base. While this technology had its advantages, the disadvantages restrained the technological growth in the aluminum sector.

At that time, each primary aluminum producer had their own casting technology, unique to the semifinished products they produced, which were destined for the markets they targeted. Under special conditions, each primary producer would market and license their technology to others in the market. While Wagstaff manufactured equipment for some, it became evident to the principals that the billet production sector of the industry had an immediate opportunity and endless potential if a unique but stable technological/equipment package could be offered to the world, separate from the status quo.

It is said the Direct Chill casting pit, from a real estate point of view, is valued at par with the Ginza of Tokyo, Japan. Just as the Ginza cannot expand economically, due to the shut-down costs of the market, the Direct Chill casting pit cannot expand economically as the costs required to purchase replacement products in the market render such an expansion cost-prohibitive. Once the partners realized this, their vision focused on optimizing or maximizing the casting area.

Most furnaces were designed in capacity to melt and process molten equivalent to two drops of the installed casting equipment; the opportunity appeared on the radar screen. Without bounds, the new technology must have the following attributes:

• Address explosion safety concerns

• Safe starting head engagement to the mold

• Uniform water distribution

• Uniform molten metal distribution

• Uniform oil distribution via graphite casting ring

• Double the as-cast weight

• Minimize the operator attention needed at the cast start

• Deliver start-up, commissioning, and ongoing customer support.

One of the first considerations was the strengths and weaknesses of the current aluminum billet casting technologies. Hot top metal distribution systems with individual molds fastened underneath the table produced the highest quality billet. However, separate water hose supplying each casting position, and each individual water jacket limited the number of strands that could fit into the casting machine. A water box system could accommodate many more billets with the universal distribution of mold cooling water, but the individual molten metal distribution to each mold gave a lower quality product and was more challenging to operate. The apparent answer was a combination of the best of the two systems, a hot top metal distribution system with its ease of operation coupled with water box mold cooling, which provided the maximum number of billets. This design had never been used because of the fear of water leakage coming from a defective seal in the water box and allowing contact with the molten aluminum, causing an explosion. The solution was a double water seal where any water that leaked past the first seal would automatically drain and still leave the second seal to provide protection. Other features were already available or soon to be introduced. The preferred hot top mold had an oil-injected permeable graphite sleeve as the casting surface. Wagstaff introduced the use of individual, highly insulating ceramic, metal transfer components for the modular construction of the hot top metal distribution system. The molds were designed to automatically align the starting heads. All billet casting components would be machined on the numerically controlled machining centers and lathes. This became the MaxiCast ™ Direct Chill Extrusion Ingot Casting Technology [3]. Figure 4 is a cut-away view showing the construction of the MaxiCast components.

Fig. 4

Cut-away view of MaxiCast Table

The Reynolds Metals Company (RMC) adopted the new casting technology supplied by Wagstaff over their own preferred solution for an extrusion plant near Richmond, VA. The first table started up seamlessly without event or fanfare, and a second table was ordered soon after this. Within a very short time, RMC made the decision to install the new equipment package at a new smelter they were building in Massena, NY. This order, being the largest order of extrusion ingot casting equipment to date at the Wagstaff Corporation, consisted of six tables, each table optimized to produce one diameter with a hidden capability to increase or reduce the as-cast diameter within range. A paper was published and presented to announce the successful introduction of the innovation [4]. With the announcement, others became interested. One or two purchases were made [5], and the technology took root and began its life as a breathing, living creature. Those post-start-up days were not easy. There were always opportunities to improve the equipment and to solve the various challenges of each new customer. With the world still in a recession, RMC agreed to allow Wagstaff to rent a small, under-utilized, Vertical Direct Chill (VDC) casting pit with an operator at the Richmond Research Center. This enabled Wagstaff to independently optimize the mold design and corresponding metallurgical structure offered to the customers using the new technology.

Other eastern winds from the Showa Denko Aluminum group in Japan were blowing. Mitamura et al. [6] had tested and commercially adopted some new concepts with their extrusion ingot VDC technology. This advance introduced air or compressed gas into the meniscus or pocket with a supply system, which resembled the oil distribution system. The air was injected on the opposite side of the oil into an oil-air distribution copper ring at the top of the mold. This technology produced smoother, much improved surfaces as compared to traditional DC and Hot Top cast products.

Wagstaff contacted Showa inquiring about licensing the technology for use in the MaxiCast Billet Casting System. They were anxious to sell the technology to the aluminum industry but didn't want to license Wagstaff to build the equipment and then sell it to the aluminum companies. This was a great disappointment since the new technology was the talk of the industry. However, thinking more about the process, and reasoning that since the molten aluminum contacts the permeable graphite mold wall, air injection through the porous graphite could create an air bearing and possibly achieve superior results to the Showa technology in a very simplified manner and not infringe the Showa patent. This approach hoped to eliminate many of the surface defects common to surface segregation and tribology at the molten to graphite contact point.

The next step was to simply modify a MaxiCast graphite-lined mold to provide air injection as well as oil injection through the mold wall. That mold was then tested at the next scheduled Wagstaff casting session at the Reynolds Aluminum Casting Research Center in Richmond, Virginia. As the cast got underway, observation of the billet surface showed it to be as expected from the typical MaxiCast mold. Then, after injecting air and varying the air pressure and casting speed, a very smooth surface was observed for a short time. By changing the casting parameters, the slick surface would come and go, and the results clearly indicated that the right combination hit the “home run.” In the next cast, the results were replicated, and amazing, nearly perfect as-cast surfaces were seen on the ingot. Subsequent trials developed a reliable process and operational window to produce not only marvelous and beautiful surfaces but also a simply stellar metallurgical microstructure. Patent applications were filed [7], and the new technology was given the name AirSlip ™. The design, construction, and commissioning of a small VDC casting center at the Wagstaff plant site in Spokane began immediately.

The first AirSlip production efforts were at the Martin Marietta Aluminum plant in The Dalles, Oregon. A 60 drop 6″ MaxiCast billet station was converted to include 20 AirSlip billets along with the 40 standard MaxiCast billets. The test casts were amazing, and the principals involved wondered which of the attributes of planning, luck, or providence was the greatest contributor to the success. That day, Martin Marietta management requested the conversion of the complete table to AirSlip as soon as possible [8]. Figure 5 shows the start of one of the first full table casts as well as a completed cast of the Martin Marietta 6″ AirSlip Billet.

Fig. 5

Start of cast of one of the first full table casts and a completed cast at Martin Marietta, The Dalles, OR

The second company willing to risk their future on the infant AirSlip process was a small 135,000 tonne smelter in the Middle East for which Wagstaff had produced a billet table for in the mid 1970s. The smelter represented itself in the market as a major producer of water for the Dubai Emirates while reducing aluminum as a byproduct. Wagstaff at that time was very cautious in the selection of their second partner with the AirSlip technology. The leadership really wanted a cautious, domestic producer with which to work, but eventually gave in to the multiple advances by the leadership of Dubai to be the second in line for the technology [9]. Figure 6 shows a cast from one of the first 6″ Billet Caster at Dubai. Today, the smelter produces 2.34 million tonnes of aluminum primary and semifinished products.

Fig. 6

Dubai 6″ AirSlip Billet Caster

The integrated MaxiCast and AirSlip technology offered existing plants the capability to double their capacity as most furnaces were designed to cast twice. The concern of temperature loss from the metal distribution entrance to the furthest casting position was usually less than 25° C and not a problem because of the high metal flow rates and high-grade insulating refractories. Many of the water systems in the industry were built using pipes in the ground with pipe diameters balanced to the original heat load or water flow. Doubling the casting capacity doubled the heat load in most systems, and while additional pumps and cooling towers could be installed, the diameter of the supply pipes emerged as a real engineering hurdle in some casting pits.

With this challenge, the development team began working on what is known today as Turbo-Enhanced Cooling Technology [10, 11], a technique involving the introduction of small, compressed gas bubbles into a planar water jet pattern to increase the localized velocity of the exiting water that impinged on the billet surface. This localized increase in velocity minimized film boiling and increased the heat transfer to the point where most of the plants with constrained water supply systems could operate year around.

New plants saw the opportunity to build mega machines casting upwards of 120 six-inch diameter billets in one cast. Figure 7 shows a typical configuration for casting 108 billets [12]. Large MaxiCast tables could be manufactured with multiple setups on the typical Kearney Trecker horizontal mill but presented a manufacturing hurdle. The original machining envelope required that each table be positioned four times against the vertically oriented machining base, which presented a potential bottleneck for manufacturing time on the K&T mill for large table production. To build these mega machines capable of handling the dreams and aspirations of many customers, the design team went back to work, engineering and manufacturing the key pieces and parts required to grow the existing K&T manufacturing envelop to four times the original size, so they could make these precision tables in one setup. Figure 2 is a picture of the jumbo Kearney Trecker mill modified for the mega casting machines.

Fig. 7

108 Drop 6″ AirSlip Billet Caster

Conclusion

It is easy to look back in the rear-view of life and clearly see the many advances and lives which have been positively affected. This technology resolved many individual casting problems and helped make the DC casting machine and cast house a much safer place to work. Worldwide, AirSlip quality billet became the standard and has been cast for almost 40 years on more than 1300 MaxiCast Billet Casting Stations.

On February 14, 1995, The Minerals, Metals & Materials Society honored and presented to Frank E. Wagstaff the 1995 Application to Practice Award for demonstrating outstanding achievement in transferring research results or findings in the fields of metallurgy and materials. Dr. Wagstaff was cited for research, development, and commercialization of the AirSlip process for casting aluminum ingots with improved surfaces and uniform fine-grain structures.

Frank Wagstaff retired in 1996, but Bill Wagstaff continued working through 2015. With his passing on January 3, 2016, that generation’s legacy was complete.