Hello everyone, I am learning basics of design of experiments..As resources are major constrains for me right now, I have decided to go with fractional factorial design namely Taguchi Designs..I want to use L9 orthogonal array for the chemical method.. After going through literature, I came to know that linearity is assumed in this method and as there are 4 factors at 3 levels each.. Linearity may or may not be present..so how to find linearity is present or not.. Should I consider two level design like L8 orthogonal array?
Krishna Joshi
In engineering, the Taguchi method of quality control focuses on design and development to create efficient, reliable products. Its founder, Genichi Taguchi, considers design to be more important than the manufacturing process in quality control and seeks to eliminate variances in production before they can occur.
Three level design will be better for you to elucidate the sources of variability and reduce standard error with improved design robustness
@ Antonio maffei sir thank you for your response. I will definitely go through this paper.
It depends on the expected relationship between the response variable and the factors. For linear dependence between the response variable and the factors, you can use L8, for non-linear - L9. If the relationship is unknown, use L9 (to be on the safe side)
Dear Krishna Joshi,
I suggest you to using the L9 orthogonal arrey for better results,
Regards,
Z. SHAKERI.
Dear Krishna.
Before you start planning, you need to determine the number of variables to study (factors that will influence the answer) and determine how many levels for each variable you need to study. After that, you need to choose an experiment design with the right number of variables. For example: L8 can be either Two Level Designs L8 (2^7) i.e. 7 variables with two levels (high, medium), but also Mixed Level Designs L8 (2^4 4^1) i.e. 4 variables with two levels and one variable with 4 levels. Whereas L9 (3^4) is Three Level Designs i.e. 4 variables with three levels (low, medium, high).
For example: Article Application of Taguchi method and ANOVA analysis for optimiz...
Good luck.
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering
- Powders