Once the formula for the S/N ratio has been determined, we will carry out an orthogonal table experiment. There are various orthogonal tables. The two orthogonal tables that are often used in quality engineering（QE) are "L12" and "L18". There are various orthogonal tables in the "design of experiments method" that have existed for a long time. Since there are many interactions between columns (parameters) of the orthogonal table (affected by the parameters in the column), you need to be careful. QE recommends the use of L12 and L18 orthogonal arrays due to the small interaction. "L36" has a large number of rows, so it will take time to perform one orthogonal array experiment. L12 orthogonal table has level 1 and 2, but 11 items parameters can be examined. Since it has level 1 or 2, it is suitable to see trends of many parameters in screening experiments to determine which is more stable, such as large/small, long/short, high/low. For L18 orthogonal array, you can study 2 levels only for column A and 3 levels for columns B to H. After the L12 orthogonal table experiment, I think it is good to carry out the L18 orthogonal table experiment to see the appropriate parameter range. Due to the small number of parameters, you should not plan an L9 orthogonal table experiment.
Click here for the orthogonal table → orthogonal table