Deterministic Polishing from Theory to Practice


Abigail Hooper, Nathan Hoffmann and Harry W. Sarkas (Nanophase Technologies Corporation)

John Escolas, Zachary Hobbs (Sydor Optics)

Improving predictability in optical fabrication can go a long way towards increasing profit margins and maintaining a competitive edge in an economic environment where pressure is mounting for optical manufacturers to cut costs. A major source of hidden cost is rework – the share of production that does not meet specification in the first pass through the polishing equipment. Rework substantially adds to the part’s processing and labor costs as well as bottlenecks in production lines and frustration for managers, operators and customers. The polishing process consists of several interacting variables including: glass type, polishing pads, machine type, RPM, downforce, slurry type, baume level and even the operators themselves. Adjusting the process to get every variable under control while operating in a robust space can not only provide a deterministic polishing process which improves profitability but also produces a higher quality optic. 

In this paper, Sydor Optics and Nanophase Technologies examined how collaboration between lab experiments and plant trials using design of experiments led to improved throughput and cost reduction in an operational filter glass polishing line. Experiments not only used standard metrics like removal rate, surface quality and surface figure to determine a robust operating range, but also examined 1-pass yield and total operational time to down select a robust, deterministic and cost saving solution.