The only contaminant I would add would be water. Anything else and you have a real chance of causing severe damage to your engine. You could base your study on the differences in octane ratings by varying the balance of alcohol, nitro and oil. Alcohol and nitro have a different octane rating which obviously will yield a different power output, after a few trials you could use your experiment to predict the performance of a known mix or given a performance figure, predict the mixture of the fuel.