Induction Heating
Hoover-Ugine Company
After the failure of the main vacuum valve in Kramatorsk we had some time to make sales calls for the Rancocas operation. One of those calls was to a company named Hoover-Ugine located near Ann Arbor Michigan. The project manager, Bud Wahl, was not impressed with the experience that Cheston had with induction heating but was familiar with the Inductotherm group of companies. He agreed to meet us with the stipulation that Cheston would come to his office as Santa Claus.
Hoover-Ugine was a joint venture of Hoover Ball Bearing Company and a French firm Ugine. The French had experimented with compacting pieces of scrap metal into a nine inch billet that was heated to extrusion temperature using an Ajax induction heating system. It was their plan to convert this heated billet into a 3/4-inch bar that could be rolled into smaller wire without re-heating. (They were attempting a major breakthrough by increasing the extrusion ratio 4 to 5 times anything ever done before.)
Hoover's products included the metal framework for truck and automobile seats and wire used in grocery market shopping carts of the time. Their sales survey included the growing market for metal clothes hangers for the dry cleaning market.
The offer from Ajax was a six-line system with three induction coils per line. Each line was rated at 3000 KW with the first coil in the line consuming 1500 KW because the material was magnetic.
Ajax's offer for this job was about $1.4 million. My telephone offer was that the sleigh that brought me to Detroit was a 707 and that Cheston's bid was $707,000. He asked for a written offer.
We really needed the business because our order book was dry and the induction heating market was the main reason Cragmet merged into Cheston five years earlier. During those five years we sold only four jobs that included: a small gear hardening machine, an induction heater for forging large pipe fittings, a 180-cycle heating system for Brazil and the large heater to Watervliet Arsenal.
Robert Klingerman was assigned the task of figuring out the most economical material handling. His task was to pick up a bar from a feed hopper and deliver on heated part every 30-seconds to the extrusion press. The total output from the heater was 40-ton per hour.
My first selection was use 180-cycles but Bud Wahl was firm that 60-cycles would be the only system he would buy. Ted Kennedy was my source for ideas on all induction matters. He was well past retirement age and the new electronic induction power supplies Inductotherm was building were not his strong suit. His latest patent was a 60-cycle "Unicoil" reactor power supply aimed at the cast iron foundry market appeared to be the ticket that would win the sale to Hoover. He knew we were fighting costs so he recommended a tapped coil construction that he used in his early years with Northrup.
We learned that Hoover had negotiated a very low power cost per KW but were stuck with a penalty for KVA demand on any 15-minute period during the month. Kennedy's Unicoil reactor control was just fine for power companies that measure only kW usage but in this case normal line voltage swings were going to add about $18,000 per month to Hoover's power bill. General Electric had a solution for this problem that they introduced in 1964. It was an internal motorized auto-transformer that regulated voltage plus or minus 10% that they called boost-buck regulator. We had already selected a standard 20,000 KVA, 480 volt, three phase system with two lines on each phase and GE wanted $30,000 for the extra feature.
In less than a week we were armed with a good set of drawings and a cost estimate. It looked like we could win this order on technical superiority so I called Wahl for an appointment with the hint that my price offer would be increased. He refused to make an appointment because Ajax Magnethermic had reduced their price to under $1 million.
We were determined to wrench this order away from Ajax and made a last ditch effort to win so I agreed to hold the price with one extra feature for automatic voltage control at $39,000. Wahl agreed to an early morning appointment at the Hoover Ball Bearing Offices at Saline Michigan just outside Ann Arbor.
In 1973 we sold our twin engine Apache to Herb Fanders who operated out of Red Lion Airport for cash and 200 hours of use of the airplane and we still had time remaining. Bob Klingerman agreed to the four-hour flight and on the way out weather was excellent. The only problem was that we couldn't raise anyone to get the landing lights turned on at Ann Arbor and landed in semi-darkness. We found out later that pilots who knew the code could turn on the lights with a radio signal.
Bud Wahl bought our sales pitch hook line and sinker but would not agree to the extra $39,000 until he inspected out facility in New Jersey. He declined to ride with us for this inspection but agreed to that we could pick him up at the airport in Philadelphia the next morning.
On the flight back we encountered a front over Pennsylvania that shook our bird like a bowl full of jelly. As we crossed the Appalachians from western Pennsylvania toward the Susquehanna River Bob remembers that I made a self serving and soothing announcement. "Things are going to smooth out right after we cross the mountains." I never liked to fly in that dangerous Apache after a hair-raising experience in the same airplane with Rowan near Asheville, NC just after he purchased it in 1960.
Bud was impressed with our brown building at 65 Indel Avenue and even thought it might be the headquarters for the whole operation on Indel Avenue. We wined and dined him that evening at Pirates Inn on the banks of Rancocas creek. I told him that I was on the first 747 flight that landed in Frankfort because it could not land in the London fog. In the spirit of use of airframe numbers he agreed that the contract would be $747,000.
I was fresh from Russia and Bud was a very strong anti-Soviet thinker. You should never argue with a customer, but his attitude needed some instruction. He trapped me in a question that led me into a corner of the end justifying the means. I was not an intellectual thinker at the time and would have to work on that. It was the time of "detente," but I realized that night and later when recalling conversations in Moscow, that people were not ready to end the cold war.
I met with Rowan to discuss the technical details of this difficult heating job. The conversation moved on to the Russians and he agreed that the cold war would end someday and that trade was the way to end it. I left his office with encouragement on the pending Russian business.
Klingerman tackled the mechanical portion of the job using Jack Murray as the design draftsman and I tackled the coil and induction portion using Frank O'Brien as the design draftsman. Ted Kennedy guided us through the design like students. He always went back to Ohm's law and reminded us that reinventing something was usually cheaper than finding and old design.
The coil selected was about 10-foot long with 750-volt capacitors over the full length. The 480-volt main transformer was connected to taps that converted the coil into an auto-transformer.
I was not an expert on coil design but by that time had more than twenty years of experience in a wide range of induction coil use and designs. During the years as a melter I had several discussions with Rowan, Inductotherm, about cold design parameters and correct frequencies. During the years as design engineer at Beryllium I met Sid Sedgeworth, Ajax Magnethermic, and Rowan and began to understand the different methods of coil construction used by those two companies. During the years as a salesman for Ajax I met Bruce Goodrich, founder of induction heating of aluminum, and the Tama's who first brought 60-cycle melting to America.
The big argument between Ajax and Inductotherm centered on the copper factor used in a coil design. During my years with Ajax/Cragmet/Cheston I carried a worn copy of Tudbury's, Basics of Induction Heating.
I sweated over just how small we could make the coil diameter since the amount and cost of the capacitors is a square function of the relationship of the coil and load diameter. Special Metals in Utica, NY worked with one of the refractory companies to make slip castings and we selected a one-inch refractory liner using this method. We left another inch for high temperature water-cooled skids on which the billets would be pushed through the coils while being heated.
Lona was assigned the task of winding the special copper shape we purchased through Rancocas Metals. During the winding process a thin wall copper tube was soldered a thin to it for water cooling. After a few tries Lona whipped the coil construction problem and with considerable difficulty solved the problem of the refractory lining.
The capacitor/reactor bank was designed as cost effective as possible while not worrying about a few kilowatts being lost in the transmission bus bar. The power system used two vertical two-inch copper pipes to carry the current to the coils that were to be mounted just above the cabinet. These pipes had tabs on each side to connect capacitors with the smallest possible bus connections. They also served as the closed water system we connected to a cheap open water-cooling tower.
I took on the task of putting the components in the cabinets. Ajax and Inductotherm used low bay buildings with overhead cranes to assemble their power supplies. As reactor and trippler components got heavier they worked out methods to use fork trucks with special handling attachments. I decided to use the overhead crane on this job. Lona built a cabinet "up ender" so we could lay the cabinets down. To prove a point I asked one man to join me on a Saturday morning and we picked the capacitors from pallets while removing them from cardboard boxes. Before the long day was over the capacitors and reactors were in place for one of the lines. The next day I departed for the Ukraine for another long stay.
The cost when the job was complete was well below budget and ahead of the promised shipping date. Klingerman was the project engineer in the field during the installation phase and all the mechanical handling worked just fine. One induction line was started and the heating worked like a charm.
The first thirteen attempts to extrude this small bar were a disaster. Klingerman observed that the problem was akin to attempting to push a rope as the hot material snaked from floor to ceiling and wall to wall. Hoover had sophisticated control device for the extruder and was using a method of starting slow with ever increasing speeds during the extrusion. Klingerman suggested starting fast, like casting a fishing line, and then slowing down to prevent the heat build up from melting the core of the bar.
This was a perfect solution to the problem but like all continuous processes the down stream problems must be solved on the fly. In line induction heating systems are not good holding furnaces and a very serious problem was observed when the compacted billets were re-heated.
When they tried to reheat a billet it would not heat the second time. Klingerman determined that the pieces of scrap were not fused during heating and upon cooling actually separated. Oxidation of the pieces near the surface made matters worse. This was not a problem caused by Cheston but as in all contracts the supplier has to make the system work. Klingerman recognized that a higher frequency in the first coil would put more heat on the surface and fuse the outer layer of particles.
I remembered a doubler that Sid Sedgeworth built for a job at Parma Laboratories to solve a similar problem for graphite heating in 1963. Kennedy studied the problem and agreed we could use the DC side of the reactor control with filters and produce 120 cycles.
A new coil was required but this time the number of turns was less so we were able to use heavy wall round copper and the size was increased to allow for more thermal insulation.
It took only one day for Klingerman and Solderstrom to reconnect the first system and it ran successfully and the next day the cooled billets re-heated properly. After all six lines were revised the system ran flawlessly on a two shift basis extruding a billet every three minutes using a single line.
Hoover had a secure and local source of steel scrap from punchings at the Whirlpool plant in Benton Harbor. They ran this through a crusher to shine up and homogenize the particles and then in line compacted, heated, extruded, rolled and coiled. The continuous nature of the process also greatly augmented the efficiencies. But on the other hand the continuous nature meant that one hiccup in the system idled the whole process. This has always been the bug-a-boo of any approach to continuous steel making. At Hoover they recognized this and did everything they could to maximize the reliability of the critical components. Klingerman observed they were getting on top of the thing and were producing the targeted wire with ever increasing reliability and efficiency. The process was outrageously economical and feasible.
The battle between English speaking and English goes back in history and probably will continue. The partnership between Ugine-Kulman and Hoover was always tenuous at best. The French considered Bud Wahl a cowboy riding rampant on horseback through a technical playground which only the French could comprehend and control if enough data was tabulated. Periodic visits by the French aristocracy were worth the price of admission. It was truly Loony Tunes with shouting matches, rank pulling and general pandemonium prevailing. The French would leave about 3:00 PM and be chauffeured around the tip of Lake Michigan into Chicago to dine and spend the night in a proper 4-star hotel. They would arrive back the next day about 10:00 AM to resume their harangue. This was bad enough but then Pechiney bought Ugine-Kulman and formed the new company PUK. Pechiney found in the Hoover facility the grounds for denigrating the previous management and show just how fool hardy they were. Then came disaster:
The process of extracting the die stack from the extrusion press required that a back pushing ram be brought into position and a dummy block be inserted in front of this ram. Then the pressure was applied and the dummy block pressed the stack backward out of the taper where it was set. The process was controlled from an elevated control room overlooking the extrusion press and billet heaters. On one particular day the die stack was being extracted and for some reason known only to God the dummy block apple cored. This is a form of failure which is little know or understood where a cylindrical piece of alloy steel being compressed from both ends explodes with great force and flying shrapnel with the shrapnel coming from the outer shell and the residual piece resembles an apple core or hour glass. On that fateful day Mike Ruvinetz, Cheston's serviceman and Russian translator, was standing next to a Hoover employee who was running the process. The block exploded and sent one piece about the size of an orange through double safety glazing and square into the chest of the operator exposing his heart pumping gushes of blood as he breathed his last in Mike's arms. The French on that day also lost their heart and withdrew their support from any further operations.
The French demanded that the plant shut down and placed under 24/7 security guards for an indefinite period while they reconnoitered their position and decided whether to proceed or try to sell their interest to another party. Some negotiations were in progress with US Steel of all people when the final bell tolled. On one winter night a security guard got extremely bored and bought ten cans of gasoline which he poured in ten places on the laid up tar roof and ignited them. He then stood back as the night sky was filled with fire and tar and water irreparably damaged every stick of equipment in the building. A day later a foot of snow fell on the unprotected remains. A week later Klingerman visited the site to observe utter failure.