Design of experiments to measure residual stress in machining

[sydneyjongleur]

Hi,
I am looking to carry out turning machine trials on steel, using carbide and ceramic inserts to determine residual stress profiles, using a design of experiments framework.

I am not overly familiar with the complexities of design of experiments and was hoping that someone could point me in the right direction of what method of DOE to choose ie. Taguchi, ANOVA, full factorial etc. and a brief synopsis of how to conduct it.

The factors involved would be feed, speed and depth of cut, and the carbide/ceramic insert. I would be looking to vary the speed, feed and depth of cut and test on both the carbide and ceramic (although the speeds on the carbide and ceramic would be different). Would this be a 2 level 4 factor DOE (2⁴), or are there more levels involved as I want to vary the parameters of the factors (speed, feed, depth of cut).

I am a little confused!

I hope that this makes sense and any help would be appreciated.

 

[sydneyjongleur]

I should note that the residual stress will be getting measured using X ray diffraction or hole strain gauge method

 

[rb1957]

is this a student project ?

 

[CoryPad]

A two level four factor DOE would allow you to have low and high values for material, speed, feed and depth. ANOVA is an excellent tool for determining differences amongst the factors.

 

[sydneyjongleur]

So, I would be carrying out 16 experiments? Before the experiments would I determine a low and high value to test at?

ie.
feed 0.05mm/rev low 0.25mm/rev high,
doc 0.1mm low 1.0mm high,
speed 50m/min low 300m/min high,
carbide low ceramic high

Also, what I am not quite sure of is if all factors have to be compatible.

What I mean by this is, for instance, the feed and doc specified can run on both carbide and ceramic, however, carbide inserts can run at 50m/min but cant run at 300m/min and vice versa ceramic can run at 300m/min and not 50m/min.

If this is the case and they cant run together would I need to set up a 2 level 3 factor DOE (2³) for carbide and 2 level 3 factor DOE (2³) for ceramic.

 

[rb1957]

maybe define feed rates qualitatively ... fast and slow ?