Randomization of material testing sequence

Is there any way to randomise the order of the shapes burned in the material test? The localised heat effects of ramping power/frequency/speed in the x axis tend to give unreproducible results when these are taken into consideration. I have tried putting a 0% sublayer in to reduce this, but it makes the time to run a test significantly longer.
Stainless steel and Titanium especially so.
Essentially, I want the antithesis of optimise settings. Give me the biggest distance possible between material test squares!