What Is “Good” Control

I’ve said before I’ve known more parents who were willing to admit their children were brats than foundry “bosses” willing to admit their foundry’s quality was poor. I’m happy to report that myopic viewpoint doesn’t completely extend to the people charged with controlling the tensile properties. Many foundry metallurgists admitted to me the control of their iron could be better. However, most of those also indicated while it could be better, it was “good enough.”

I never have figured out what “good enough” is. I suppose if you’re not getting any tensile test failures, you’re not doing any alloying to achieve the strength, and your customers are happy with your machinability of your castings, your control is good enough. Of course, in the fifty years I’ve worked in the iron industry, I’ve never seen all three of those conditions met at the same time, but I suppose it could happen.

Defining good control isn’t much easier. The way I look at it good control could be having no tensile test failures and having the fluctuation of the tensile tests lower than your major competitor. While that’s more precise, to my knowledge, no one is collecting such information and posting it on a regular basis.

The AFS Cast Iron Quality Control committee did an initial study of this in 2000. Requests for information about class 30 iron were sent to every gray iron foundry in the AFS data base. As reported in “web only” Modern Casting article in July 2002, 40 responses were returned. Of those returned, the average standard deviation was 2379 psi with a high of 3845 and a low of 1120. In my opinion these figures most likely show a slanted view of the control of the industry. It is almost certain those who did report their data were tracking it before the questionnaire was received. It has been my experience that people who track a variable usually have it under better control than those who simply assume the variable is fine.

I believe, at one time, GIRI was tracking this type of information from their members. If that is true, to my knowledge, the information was never made available to non-members.

When thinking of control of tensile properties, one must consider two significant aspects of the total process. Obviously, control of the iron characteristics is what we are asking to measure; however, what must not be overlooked is the control of the testing process. Unfortunately, the tensile testing is a significant variable in the measurement of iron control.

An initial step in showing the significance of the testing fell to the AFS Cast Iron Division’s Quality Control Committee. They reported[i] on work done showing the variability in simply the testing phase of ductile iron. Taking great care to make sure the iron they were using was “the same” they sent 10 test bars each to 11 different laboratories where they were tested. The results from that study are shown in the table below.

Standard Deviation of 10 Tensile Tests

What does this mean? It means no one should expect better control of their iron tensile tests than the repeatability of the testing procedure. The variability in the process of controlling iron will be added to the variability in the testing.

But, what is good control? Basically, I’m not sure anyone knows for sure and until significant benchmarking is done, I don’t know that anyone can really say.

I am willing to use this site as a bench marking source for iron tensile strengths. Foundries interested in submitting data should send me via email (rwl@lobenhofer.com) the average and standard deviation of their last 25 tests of each grade they produce. I will post the data without reference to the source of the data; however, I will by return mail provide the code for which their information is posted.  

Gray Iron

Class 30

Class 35

Class 40

65-45-12 Ultimate

65-45-12 Yield

65-45-12 Elongation

80-55-06 Ultimate

80-55-06 Yield

80-55-06 Elongation

100-70-03 Ultimate

100-70-03 Yield

100-70-03 Elongation

Back to page on Controlling Tensile Properties