While at working for the American Foundrymen’s Society, I was introduced to computer programming. I was enthralled with the potential of the computer to assist with control in foundries and taught myself how to program. Once I started consulting, I began to create programs for general foundry use and programs for specific foundry operations.

Some of the programs are:

Predict

This program provides information that allows improved mechanical properties of cast iron while offering the potential for reduced alloy and charge costs.

Convert

Shareware program containing over 2800 conversion formulas. It does the math, too.

Stats & Stats2

When I started getting interested in programming, I was already deeply interested in the power of statistics for tracking and controlling variables. My frustrations initially were at the cost of programs for statistical analysis. They were simply more expensive than I could afford. Once I did find some that I could afford, the programs dealt with one variable at a time. Foundry variables don’t come one at a time. In foundries, we get chemical analyses with numerous elements, sand tests with a number of variables from each test, and just about every other test comes up with numerous variables that should be tracked at one time.

My answer was to write a DOS base program that would accept as many variables as desired and would alert the user of any entries that were not in statistical control. Of course, once I went that far I started developing all sorts of files full of data from that program. Then the challenge became to be able to better quickly analyze that data. I soon added the capability to compare sets of data, perform simple regression analysis, perform linear multiple regression analyses, non-linear multiple regression analyses, and more.

A number of my clients and former clients still use Stats, the DOS version of the program, but I have been working on converting into a windows program. Stats2, the new version, does use a different file structure that allows using the files in most spreadsheet programs.

These programs are only available to clients because there is no documentation. Only with ongoing interaction with the people using the program, can its power be fully utilized.

SFP

While assisting one foundry, I was completely frustrated with the lack of information they had about their production and scrap. In order to concentrate on the most significant problems, I believe you need to know in detail about scrap and production. Knowledge of scrap and production by part, by customer, molding area, metal type, defect type is essential to building an effective program to improve quality.

That foundry kept promising me that they would get their present program rewritten to get the needed information. After about a year of waiting, I gave up and started writing a program they could use to get the information.

I no longer work with that foundry, but it is my understanding they are still using the program.

Of course, that program is designed specifically for that foundry, but the framework in that program can be used to develop a similar program for any foundry. If you don’t know which of your molding units produces the most scrap and of what type maybe you should think about a program that will give you the information you need to design an effective quality program.

Identification

My latest project is a program to assist with the training of inspectors.

One of the problems every foundry quality manager faces is making sure that he is trying to solve the right problem. Trying to correct defect problems when the defect has been improperly identified is an exercise in futility. To circumvent this problem I am writing a program that will train inspectors to correctly identify defects.

There will be three stages to the program. The first stage will be a learning stage. The user will select which defect is to be reviewed. A description of the defect will appear as well as pictures of castings that have the defect in question.

The second stage will be a self-testing mode. In this stage the user will be shown a picture of a defect and be asked to identify it. If the defect is identified incorrectly the program will tell the user to try again.

The final stage will be a true test of the users knowledge. Pictures will be shown and the user will be asked to identify the defect. This time all answers will be recorded and the results will only be available to a person with a special code to access the information. That will usually be the person who is responsible for the training.

Want to be notified when the program is ready? Let me know and I’ll let you know.  rwl@lobenhofer.com.

Special Programming

I have also done some special programming in fields outside of the foundry industry. Such programming usually has involved data manipulation of one form or another. If you’re looking for someone to write a specialized program for your needs, give me a call and we’ll see if I can help you.