I am preferring to use Response Surface Methodology (RSM) in optimization of friction stir spot welding (FSSW) process, could you please tell me if there are any limitations to use this method in optimization.
In FSSW, there are four affected factors; tool rotation speed, welding speed, axial load and pin diameter, while there are two outputs will be tensile strength and hardness.
The only issue may be that DoE assumes that the distribution of response variables is normal. Although hardness can be expected to be normally distributed, I will be surprised if tensile strength is (here I assume that specimens will break without necking). I would expect TS to follow the Weibull distribution. On a separate note, I recommend that you also use elongation as a response variable. Because elongation is a measure of the energy absorbed by the specimen, it will be a good indication of the structural quality of the weld.
Good luck.
Dear Raheem Kh. Al-Sabur,
Limitations of "Response Surface Methodology (RSM)" are mentioned below.
1. The First or main limitation of the method is that Response Surface Methodology (RSM) is a ‘black box’ approach (David C. Cox & Paul Baybutt, 1981). That is, estimating the accuracy of approximation, or one can say in other words the magnitude of the approximation errors, is difficult.
2. The Second limitation of the method is that it is a local analysis. The developed response surface is invalid for regions other than the studied ranges of factors.
I hope I have answered your question.
With Best Wishes,
Samir G. Pandya
In RSM analysis the data has to be fitted to to a second order polynomial. If system under question ie curvature can be fitted to second order polynomial then RSM gives good results.Moreover RSM uses Method of Least Squares to estimate error.
Dear Murat Tiryakioglu
Thank you for your reply.
In brittle materials, the Weibull distribution widly used but there are several studies carried out using a larger class of probability models including Weibull , normal, log-normal, gamma and generalized exponential distributions.
Also there are many articles using RSM to predict the tensile strength behaviour.
Thanks again
Dear Samir G. Pandya
Which tool or technique you suggest to use for DoE , analysis of results and optimization instead of RSM?
It can predict the variation with in the limits of DOE effectively. If you predict the results outside design points, either you need many confirmation trials or it will not be beneficial there.
Dear Raheem,
RSM is a great tool if you define the design space over which to operate. Design Expert is one tool you could use
- Badges
- Science topic
- Similar topics
- Analytical Chemistry
- Extraction