A manufacturer of extruded HDPE pipes experienced high scrap rates due to non-conformant pipes caused by multiple defects (i.e. surface defects, wall thickness, length, density, pigmentation, etc.). As a result, the company decided to implement Lean to address this issue.
For the initial investigation the overall weight of pipes was selected as a comprehensive variable for wall thickness, length and density (pigmentation influencing density as per blending method). After analysing 25 samples from production it was found that the process capability was in the order of 1.5 sigma (Cpk = 0.02) with a defect rate of 59.96% and a cost of poor quality (COPQ) in the order of 37.5% of the company revenue.
Making use of a Lean tool called Pareto rule (aka rule of the 20/80) the company determined that 80% of most influential defects were caused by only 20% of the root causes identified during the root cause analysis (fishbone diagram). It was observed that that the standard operating procedure (SOP) for the test of pigments was not being followed and many pipes came off the process with color defects. It was also found that variations in HDPE viscosity were caused by variations in density originated in the blending process (resin, additives, pigments, slipping agents, etc.). Higher viscosities would lead to poor flow, surface defects, higher shrinkage rates and warps delivering additional problems with maintaining a constant diameter. Conversely, low viscosities would deliver thinner walls under specified tolerances.
After a correlation analysis between blend density and viscosity, the company was able to determine the desired MFI (melt flow index) to obtain an extrusion process within specification. Following this initial implementation the company was able to reduce the defect rate to 5% and the cost of poor quality plummeted to 15% of revenue.