Experience is a great teacher. What’s more, experience is somewhat scientific in that it is reality-based. If you don’t like the result you’re getting, you try something different until things work out better. This is why for many decades injection molders have relied on experience to determine starting points (and frequently end points) for their molding process parameters.
Molding process “tribal knowledge” garnered from years of shared trial-and-error experience allows molders to make acceptable molded parts based on settings that have usually worked in the past for similar part geometries and materials. While this approach gets you up and going, it offers no reassurance that parts will be conforming run after run. What we don’t know from this approach is how much better things could be. For high-precision production molding we really need to know that our process parameters are not just good enough, but optimal.
The only way to find out is to experiment with the different settings until we arrive at a combination that produces the best results. That sounds simple enough. However, there are approximately 20 variables that affect molding process performance. If we did experiments that covered all the possible parameter combinations and interactions, the work would take months and the data generated would fill numerous encyclopedias. Fortunately, we don’t have to do that.
For nearly a decade now, PEP Lacey has been using a procedure known as Design of Experiment (DOE) that allows us to streamline molding process experiments with advanced statistical procedures that allow for testing multiple process variables in a small number of experiments.
A typical DOE project involves a multidisciplinary team of engineers representing tooling, process, manufacturing and quality. As a group, they determine which 5-7 parameters (out of about 20) have the most probable bearing on process outcomes for this particular part. The remaining ones are locked in at specific values based on experience. Experiments are then conducted to test the 5-7 identified parameters in various combinations. Data is collected from our process during each experimental part run, and parts are collected and measured so that we know which combination of variables allows us to produce the highest quality parts at the fastest reasonable cycles.
This might seem like an awful lot of work and added expense, but that is far from the case. Here is the value that scientific process development adds to our production molding process.
Bottom line: Scientific Process Development allows PEP Lacey to offer the highest ascertainable levels of quality. It also gives us a tool for identifying and stripping out cost factors to keep pricing competitive for high-volume part manufacturing.