Changes That Aren't

Most of us have heard the definition of insanity: to keep doing everything the same and expect different results. It makes sense that if you want different results something is going to have to change.

We foundryman are always involved in projects to change; we make changes in order to reduce scrap, lower cost, or improve specific aspects of the operation. Sometimes the changes we make aren't really changes.

That sounds contradictory. How can changing something be keeping it the same? If the changes are the changes that are always made, it may be keeping things the same. Believe me, it happens - more than most people even realize.

As an example, there’s a foundry I’ve helped previously with some small jobs that is right on the way to one of my regular clients. The foundry is at a good breaking point in my drive, and I don’t want them to forget the things I’ve done for them, so I stop by to see them every other month or so. My visits are usually with the president, but, naturally, I also know the superintendent.

It seems every time I ask about the superintendent, he’s working with a pattern maker to re-rig a problem job, expecting the pattern maker to drop off a revised pattern in a few minutes, or is at the pattern makers going over some revisions. He certainly thinks he’s making changes, but is he really making changes? In my mind, he isn’t. Recently, I realized that he was so busy changing what he always, that he didn’t have the time to make the changes that were really needed. Knowing the operation, I believe if some of the time spent on changing patterns were devoted to controlling their processes better, they wouldn’t need to make all of the pattern changes.

As with so many things in life, it’s always easier to see someone else’s faults than our own.  It wasn’t until I started thinking about his not changing by working so hard making changes that I realized I’ve been guilty of the same thing. It also pointed out to me how difficult it can be to know what to change.

My way of solving foundry scrap is by improving the control of processes. If you’re having hard castings, I’ll tell you to improve your metal control. If you’re having misruns, improve your temperature control or rigging process. If you’re having sand inclusions, improve the control of your sand properties. I recommend process improvement because it has worked for me for years – most of the time.

One of my clients wanted to take their quality to another level by reducing scrap. I immediately set to work thinking about how to improve their processes. I looked at their metal control and devised an improved system that let them control their iron to an even greater degree. The scrap didn’t improve. Their customers felt the greater control improved their machinability, but the scrap didn’t go down.

Looking at the records of the operation, it was no surprise that improving the metal control didn’t reduce scrap. Metal control wasn’t causing scrap. My process improvements could have eliminated all variation in the metal, and the scrap wouldn’t have gone down noticeably.

Sand inclusions were a problem. So, we attacked the sand control. We tweaked this and tightened down that and our control charts showed that the control of the sand properties improved, but the scrap didn’t go down. Once again I went to the records to determine why our “improvements” hadn’t worked. The records showed there was scrap to correct this time, but statistical analysis didn’t show any correlation to where the sand properties were being held or to the amount of variation the properties showed in the ranges they were currently being held. We had improved sand control, but really didn’t have any reason to expect that the improved control would reduce scrap.

We were stumped. We were about to start another program of improving the sand control when a small fire occurred that required one of the molding machines to be rebuilt. After that, it only took a couple of months worth of data for us to notice a reduction in the inclusions. A couple more months of data proved that the decrease wasn’t a fluke. Then, we had to analyze what happened and make a change that has kept the inclusions down since that time. The change wasn’t another program to reduce sand property variability, nor was periodically setting fire to the machine. A change was made in the maintenance practice. Far more checks are made now, and no longer are the p.m.’s treated like something to do if nothing is broken. A real change had been made, and different results were achieved.

Don’t misunderstand me; this isn’t condemning improving sand control or changing rigging. Obviously, both are important. If the sand weren’t well controlled, even with the improved maintenance the inclusions would have been high. Similarly, if the gating was causing erosion, the best maintenance in the world wouldn’t have prevented inclusions. Nor am I saying that improved maintenance will solve all your problems.

The thing I am saying is that the next time you’re changing something in order to improve, ask yourself if the change you’re about to make is really a change. Have you already done this before? If so, should you really be expecting this same change to get you the improvement you’re looking to achieve? Or, do you really need to find something that is really a change?