I used to work with a gentleman named Jim Arnold. He was telling me about one of the automotive plants he worked at where the bean-counters came up with a way to save several dollars per car they produced. They were going to make the panel in front of the hood and which spanned between the two headlights out of a new kind of plastic.
From the beginning, the new material was difficult to paint with lots of "poppers".
Picture in your head a pan full of bacon that is being fried over high-heat. Can you see the bubbles popping and throwing an aerosol of grease into the air? Can you see the hundreds of tiny bubbles and craters? Now "freeze" those surface imperfections (bubbles and craters) and you have a decent mental picture of "poppers".
The only fix was to use sandpaper and grind ALL of the paint off of the panel and do a total repaint.
The engineers said they had a handle on it and "it won't be like that in production".
But it was.
It was maddening. Between 5% and 80% of the cars had the defect and the failure rate wandered around in both location and percent of defective parts.
Everybody in production whined. It was super labor intensive to fix and some of the defective panels were just as defective after a total repaint as it was before.
The opinion on the floor was they should just shut-down production until the engineers figured it out.
The Plant Manager put his foot down. "We just have to process our way out of this problem" he said.
According to Jim, the Plant Manager's commands were not well received but in retrospect he was absolutely right. The engineers could not replicate the failure in the lab and the plant needed to run so the corporation could make money.
However much manpower was required to repair the panels, the Plant Manager would find it. And he did.
It took six months to run all of the issues to earth and kill them, but they did.
Jim made a mountain of money in overtime that year.
And so it is with our current tough patch. It is hard but we have no viable options other than for each of us to "...process our way through it" using whatever tools we have available.
Technical details on the paint issue and the fix below the fold
The panels were a type of Sheet Molding Compound. SMC is a blend of chopped glass roving, filler and a plastic resin that cures when cooked.
The materials are put into a mixer and mixed until homogeneous and then it is rolled out in slabs (often about 2" thick) which are cut into "books" and books were placed into the molds by hand. The two halves of the mold came together and squished the pliable material into the shape of the panel. The molds were hot and the cookie was cooked.
The problem was that the filler (ground limestone and clay) still had voids in it that were not filled with resin. The surface absorbed solvents out of the paint which "flashed" when heated in the ovens and caused the imperfections.
The lack of repeatability was due to variation in the slab across the width and from top-to-bottom. Sometimes the side of the books with the most porosity were placed face-down and the poppers were not visible. Sometimes the book was face-up and the panel was a nightmare. Sometimes the book-from-hell was placed into the left side of the mold, sometimes the middle, sometimes the right side.
Increasing the resin content would change the material properties and increase the cost. Directing the operators to keep track of the "bad" books and sides was considered unmanageable. The "fix" was to find a blend of solvents that flashed off slowly enough that it could bleed through the paint film without causing a popper.
So the 'easy to mold' material that saved a few dollars per unit cost tens thousands of dollars in overtime. Did the company ever do a cost-benefit analysis of the entire event or did the bean counters declare victory?
ReplyDeleteSome problems only appear in production that cannot be replicated in the "R" part of R&D. That is why there should be a "D" portion to develop a process from small to large scale. Process development is expensive and time consuming but is cost effective in the long run.
ReplyDeleteThat said, in our personal life most of us do not have the luxury of a development department. Preparing resin mixing for consistency...
There are chevy vans and pickups running around with sheets of white paint peeled off the gray primer. Almost always off the hood nose.
ReplyDeleteThere was an issue years ago when bas coat/clear coat came out that the paint application was so thin that uv penetrated and broke the bond with the sealer.
DeleteWhen lacquer was used before that, often the repair guys would spray entire panels over the ecoat without sealer/primer. I remember seeing those cars on the road with sheets of paint flying in the breeze.
Rick beat me to it... And the military has had its share of 'horror' stories... Ed, there are STILL vans losing their paint... sigh...
ReplyDeleteNice. Those sorts of problems are fun.
ReplyDeletePlenty of things that work great on paper or in development just don't pan out in the real world. And plenty of engineers, designers and architects have never worked in the real world so the things they create fail when exposed to actual life. Just a fact of reality.
ReplyDeleteAt one company I worked at warranty issues from a poor design were rarely fixed. When I asked why, the head of engineering told me warranty cost fell on the marketing department, Marketing would have to come up with the money for engineering to fix the issue. Our production folks ran an end around and upgraded the product, then submitted the proper documentation to QC.
ReplyDeleteReminds me of the Crown Vics I see around here, with large sections of paint missing, and what appears to be a galvanized steel body underneath.
ReplyDeleteSome not corroding, but some rusting as if they weren't galvanized at all.