Die shop sparks plantwide process control renaissance
November 29, 2001
The author relates his experience as preventive maintenance (PM) coordinator at a Big 3 automaker. The purpose of preventative maintenance is to gain control of the processes. This begins with data collection on die sets, die failure and material handling damage that will identify weak areas in the stamping process. The author's team began with one set of dies and eventually its PM program was so successful that it was implemented throughout the entire plant.
When I was invited to serve as a preventive maintenance (PM) coordinator at a Chrysler plant in 1992, a story was circulating that was relevant to not only what we learned at Chrysler, but to any stamping line.
According to the story, a line supervisor-technician on a Toyota stamping line was asked whether his parts were within acceptable specs. The supervisor looked at the array of dials and gauges and replied, "Yes, they are within spec."
He did not look at the part, not even at every thousandth part. As far as he was concerned, if the dials indicated that the part was in spec, it was in spec. Why? He had control of his process. The board he looked at controlled his process, and each dial or gauge represented each component of the process.
What stampers have to ask themselves is whether they can control their processes as well.
It is clear that each and every part of the stamping process must be monitored and controlled to achieve reliability and consistent quality. Each aspect of stamping influences, favorably or unfavorably, every other aspect of stamping. If that is so, why aren't stamping technicians on every press line looking at gauges and dials that ensure quality and reliability?
Easy question, tough answer. How a stamping operation answers this question will define its production management. On the way to arriving at the correct answer, you will hear a lot of excuses.
Our team was allowed to take control at the Chrysler plant, but first we had to prove that PM would make a difference. We did this by concentrating on one set of dies and then on all dies assigned to a line of presses. Finally we were given all the dies in the plant, about 1,200.
After working two years to implement PM in the tooling area, the PM team at our plant was ready to believe that process control (PC) actually was possible. In fact, if the plant was to continue to improve, PC was mandatory.
PC dictates that any one component of the process must be as reliable as every other component. To our benefit, the PM program identified weak links in every aspect of stamping. All we had to do was compile data and then read it.
Most of our data was downtime-related. We tracked die sets, die failure, material handling damage, and equipment failure.
Die shut heights were standardized to facilitate quick die change. "Sandwiches" were made of every die — everything necessary for a die to operate efficiently was included between the upper and lower subplates. We achieved common shut heights for all dies that ran in a set of presses. We set dies in minutes instead of hours or entire shifts. The cost of sandwiching was recovered quickly through improved die set times.
Automation booms were made specifically for each tool, which eliminated the need to readjust at every die set. Instead of cheap, homemade weldments, more dependable, commercially available components were used to eliminate variation. Although this was more expensive, we felt that it was too costly to do otherwise.
Well-maintained tooling eliminated discussion of what department was charged with the downtime — many failures were automatically assigned to die maintenance — and the finger pointed to other culprits. Presses actually were causing problems. Blank feeding systems were not maintained properly and were incapable of doing a good job.
To correct these problems, we eliminated in-house blanking, and purchased blanks became the norm — providing a definite increase in quality control.
Bad dies damage presses, and bad presses damage dies. When rehabilitated dies still were being abused by bad presses, we came to the obvious conclusion that it was time to begin a comprehensive press PM program. When regular cleaning of dies revealed that press oil was leaking during the production cycle, press maintenance people were able to locate and repair the leaks.
With better-maintained dies, the press adjustments and repairs lasted for longer periods. Presses, under a comprehensive PM program, experienced much longer mean time between failure.
A comprehensive tonnage monitor program identified excessive die wear and degeneration. A portable Difraccto machine helped to rework surface conditions where necessary. This machine identified problem surface conditions on die posts and on the actual formed part.
By closely controlling and properly washing blanks, we minimized surface damage. We touched up panel flow with jeweler's rouge where needed, which wasn't a big deal, but it proved to be a big help for concerns about surface quality.
We would never have attempted a press PM program if we had not first brought tooling under the control of a comprehensive PM program. As stated earlier, dies can ruin presses just as easily as presses can ruin dies.
As we eliminated variables, we uncovered other problems. Leaking nitro cylinders demanded a PM program of their own. Dadco, a supplier of gas die springs, helped set up the area and supplied the parts and training. Soon failure became avoidable.
We replaced cylinders safely after 1 million hits. We used leak detectors to eliminate any connection leaks. The only problem the leak detectors caused was that maintenance people constantly were borrowing them from the die shop to help them do their job. It's an indispensable tool for both the die shop and for the maintenance trades.
After-production processes eventually received PM attention as well.
Once the PM team completed its work with the stamping process, plant management decided that one controlled process would be the catalyst to improve all plant processes. A chain-reaction explosion of quality took place.
Because of the resulting increase in quality and reliability, the plant was able to eliminate panel storage previously devoted holding panels that could be delivered as needed to cover for a die failure. The newfound space made installation of 24 new subassembly lines possible.
The plant added another profit center — die construction — because fewer diemakers were needed (less than half) to service production lines and maintain the existing dies. Chrysler began to pay our new die room to build the dies that would run in our presses and in the presses of our sister stamping plants.
We evaluated each step of the improvement process and reported our findings to staff managers, including the attendant cost-avoidance savings. This step was important. Managers get busy and preoccupied, so it doesn't hurt to remind them continually that PM is working and that it is saving them money — money that could be allocated to both the plant profit and to continuous improvement of the manufacturing process. The message of this story is that nothing was too insignificant to be considered. No data was too unimportant to be recorded. No suggestion for improvement was disregarded.
We also reaped an unexpected bonus. When the product engineers and tool design staff needed to design new dies for each body style, they were able to use our recorded repair data to help them design tools without flaws.
PM is a way of life. As such, you implement it one step at a time with patience and a recognition that changes come slowly. However, change must come, and it is up to responsible managers to guarantee its implementation. We found implementing a comprehensive PM program to be a formula for success.