Page Last Updated: Tuesday, 19 January 2016 11:02 EDT, 2016

HARTLEY CONSULTING
Solving
Complex Operational and Organizational Problems

PROJECT: Stick Vinyl to Nylon

Dr. Dean S. Hartley III


Project Metadata Keywords
Label Name Other Year DurationYrs
Client Deering Milliken Research Corp Milliken & Co.
Dates 1967 0.1
Employer Deering Milliken Research Corp
Partner N/A
Experimental design
Industrial issues
Optimization
Science, Math and Medicine
Statistical inference
Textiles

DETAILS

While I was a student in college, I also worked as a part-time chemistry lab assistant for Deering Milliken Research Corp. My boss was a chemist who was trying to stick vinyl to nylon. At the time vinyl was attached to cotton or polyester fabric to produce waterproof fabrics for automobile tops and furniture. The fabric provided the structure and strength and the vinyl provided the waterproof surface, which could be embossed to look like leather. The vinyl sheet was attached to the fabric by heat and pressure, aided by a simple chemical glue. My boss thought that a nylon fabric would provide greater strength. Unfortunately, vinyl did not want to stick to nylon. My boss had tried all of the chemical formulas for glues that readily came to mind, with no success. At the point I arrived, he probably would have tried peanut butter!

Chemists were not then taught sophisticated experimental design. He was doing better than random experimenting – he was holding all variables but one constant and creating trials with variations of that one variable. This resulted in long tedious sets of experiments, each with its set of tear tests, tests of bonding of the vinyl to the nylon fabric, etc. I thought there had to be a better way. I had a friend in the company OR department, so I knew it existed. I explained the problem to the head of the OR department and asked for advice. He explained that there could be significant interactions among the variables, which would be missed by the “hold all but one variable constant” technique and that there were statistical tests (specifically analysis of variance (ANOVA)) that could be used to analyze the results. At the time ANOVA required hand calculation because there were no available computer programs to do the calculations. I took this advice back to my boss and we became more efficient in our search for a way to stick vinyl to nylon.


If you arrived here using a keyword shortcut, you may use your browser's "back" key to return to the keyword distribution page.

Return to Hartley's Projects Page