Hello everyone,
I am now employing cold rolling to quantitatively adjust macroscale properties. I always used the same sample size, the same roller and speed, but the repeatibility is still not satisfactor. Is anybody farmiliar with improving this situation?
Best regards,
Qiao
Qiao, it will help us help you if you tell us what you mean by "repeatability." For example, do you mean the macro-scale properties have too much variation? When you say, for example, "same" roller and speed, how do you quantitatively determine the variation in the process inputs? [Are you familiar with Statistical Process Control and Designed Experiments (such as Components of Variance, Factorials)?]
Statistical Process Control (SPC) will allow you to quantitatively determine the amount of variation in a given variable (inputs, in-process, outputs --> e.g., macroscopic properties). SPC analysis will also allow you to determine whether those variables are in a state of "statistical control" (varying only due to the natural, or random, variation of the variable --> Common Cause variation) or if there are Special Causes present (not part of the natural variation --> a cause outside the process). You should first identify and remove Special Causes if their effects are undesired, or incorporate them into the process if they are beneficial, until the process is in a state of statistical control.
Then, if the amount of Common Cause variation is still unacceptable, you can use a type of Designed Experiment called Components of Variation (also called Nested Designs; also known as Gage R&R if the measuring process is non-destructive) to identify which parts of the process are contributing how much variation to the overall total variation. This allows you to determine where in the process to focus your variation-reduction efforts.
Start with a Google search on "Statistical Process Control" and on "Nested Designed Experiments." For more detail, I can highly recommend "Understanding Statistical Process Control" by Donald Wheeler and "Improving Quality Through Planned Experimentation" by Ronald Moen, Thomas Nolan, and Lloyd Provost. Chapter 8 illustrates the combined use of SPC and Nested Designs.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size