FOCUS: Improve two significant defects in the molding of 8mm data cartridge shells.
A supplier was providing 8mm data cartridge shells to the company. Two chronic defects existed which threatened the OEM business of this product. The first related to short molding. The second related to a gate defect.
The company proposed to assist the supplier with a project to eliminate or significantly reduce these defects. Molding process parameters were brainstormed and some initial simple tests were done for factor selection. A first experiment was designed utilizing three system pressures, P1, P2, P3. The supplier asked if a full factorial could be used, as an introduction for them into Design of Experiments. So, the experiment was 3 factors, 3 levels each. A continuous measurement scale was desired, and was thus developed for the magnitude of the gate defect and the magnitude of the short molding defect. The first experiment and subsequent optimization provided significant improvements, but an additional factor was desired for testing; whether the plastic was virgin plastic or reground. So, a second experiment was planned. The supplier agreed to a more efficient test; the Box-Behnkin design. The second experiment, and subsequent optimization, yielded very good results, which were confirmed by follow-on testing and control charting.
Neither of these defects were ever found again for the life of the production of this product.
* Due to nondisclosure agreements, the organization referenced in this example cannot be disclosed.