It was an emergency. Another emergency. When you're up to there in alligators you can't hardly drain the swamp, but we had to catch this alligator in a hurry.

Transmission Division came to us with a problem with spot welds. They knew we had developed an ultrasonic method to test spot welds and they called us immediately to solve their problem.

Spot welds were being used to hold certain brackets to the interior of the steel cases of torque converters for automotive transmissions. Torque converters take the power of the engine and transmit it through an impeller and a turbine combination to the gears and bands in the automatic transmission so that the power can get smoothly to the drive wheels of the automobile or truck. The torque converter case is two pieces of sheet metal which come together in the shape of a bagel, about 356 mm (14 in.) in diameter, which has been sliced and put back together. Continuing the analogy, all the interior dough has been hollowed out to permit the insertion of the impeller and the turbine.

---

No amount of adjusting current and voltage could produce good welds on this new model of torque converter.

 

---

At any rate, the spot welds on this bracket inside one side of the converter case were failing. No amount of adjusting current and voltage could produce good welds on this new model of torque converter.

Using new developments based upon previous work by other engineers in Philadelphia, Tony Mansour found that he could test these spot welds and predict future failures. Based on this success in the technical feasibility study on a few converter cases, Mansour and I were invited to visit the transmission plant, recommend a manufacturing feasibility study and then suggest automated implementation equipment — namely a big, expensive system.

After talking with the very worried engineers and their harried managers, Mansour and I were taken into the plant to observe the spot welding equipment in action. The equipment was massive and heavy duty. There were two spot welding heads mounted on a large piston running vertically which brought the heads down to touch and clamp the bracket to the section of the converter case. These two heads were placed symmetrically at three and nine o'clock with respect to the shaft hole in the center of the circular section of this half of the converter case "bagel."

The end of the piston was insulated from the rest of the machine so the current, when introduced into the spot welding heads at the bottom of the piston, would flow through the welding heads, through the two layers of sheet metal to be spot welded and into the corresponding lower spot weld heads. A region of metal was supposed to melt where the current passed from one metal sheet to the other and then refreeze into a nugget when the current was turned off.

The metal parts were placed into the jigs correctly, the piston moved up and down properly and the current flowed.

The current was so many thousands of amperes that it had to be carried by large amounts of copper. Since the current source was stationary and the piston tip which required the current moved up and down with a sizable throw, the cable for the current had to be flexible. For symmetry to the two spot welding heads, the current was brought to the piston head by two sets of conductors from the two sides. To be flexible and to have a large surface area to carry the large alternating current, the conductors were multiple layers of thin sheets of copper separated by air gaps of about ten times their thickness. These sheets of copper were clamped to the piston at the center and to two electrical buses outboard. To permit the flexibility for the vertical motion of the piston, the copper sheets were extra long and drooped down in a sort of catenary shape on the two sides of the piston. As the piston raised and lowered, one could imagine looking at the cables of a suspension bridge flex if one pier moved up and down.

Mansour and I watched this welding process intently. Soon the solution became obvious. What was happening was this: with the pistons in the lowered position, the bottom copper catenary sheet was touching the frame of the machine and grounding out. Occasionally (but not on every stroke), sparks would fly as the current came on.

The NDT solution was to do no NDT at all. First, we ordered an electrician to tape large pieces of 13 mm (0.5 in.) thick rubber pad to the frame of the machine under the copper catenary conductors. Second, we recommended that the engineers adjust the heights of the various elements so that the copper would not droop so far, leaving the copper catenaries up in the air where they belonged. Good spot welds resulted when the current was subsequently set to its design specifications.

The root problem here was the old fashioned Taylor theory of management which was in use in all of American industry for so many years (see Walton [1986] for more information). Under the Taylor regimen, the workers on the floor just moved things and had no intellectual input into a process. Even if the workers had reported sparks in the wrong place, they would not have been listened to by the foremen and would not have been believed. Indeed, they would have been reprimanded for interfering and not producing. Taylor did not want workers to have any training, so they would not have even been instructed that sparks or electrical lines grounding out were undesirable. The engineers would not have been empowered to ask the workers any questions of substance.

Taylor kicked all the responsibility upstairs to the engineers and then further upstairs to their managers. What was the responsibility of the engineers? The engineers would have drawn up the process with the welding heads touching the piece parts in the right place and the current being correct for the thickness of metal and would have assumed that their spot welder would work well in a turnkey fashion with its manufacturer being the responsible party ("upstairs" from them). It would be very likely that the engineers' analysis did not go deep enough (before the introduction of Ishikawa fishbone diagrams in the 1980s) to even discuss a possible short circuit in a failure modes and effects analysis. Who would have thought it, anyway? But even in the presence of acknowledged difficulties, the engineers did not have the time or did not take the time to go to the plant floor and look at what was really happening. And, under Taylor, the managers were absorbing blame but doing nothing intellectually creative. The alligators were taking over the factory as well as the swamp.

For further study of the proper management of NDT in the context of industry and total quality management, and for further study of the cost of quality when detrimental things occur, one would be advised to take short courses on these subjects when they are offered at ASNT national conferences.

 

References
Mansour, T.M., "Ultrasonic Inspection of Spot Welds in Thin Gage Steel," Materials Evaluation, Vol. 46, 1988, pp. 650-658.

Walton, M., The Deming Management Method, New York, Putnam, 1986.

 

* Quality Systems Concepts, Inc., 379 Diem Woods Drive, New Holland, PA 17557; (717) 355-9809; fax (717) 355-9812; e-mail <papadakis@desupernet.net>

 

Copyright © 2001 by the American Society for Nondestructive Testing, Inc. All rights reserved.

[ Materials Evaluation ]