FMEP of engine including viscosity effect

I am trying to use engine models to get at the fuel consumption dependence on lubricant viscosity.
From what I have read, it is usual to write the FMEP as:

FMEP = c0 + c1*N + c2*N*N

where N is the engine speed (rpm)

Can I assume that the first term is related to boundary friction, so could be affected by the presence of a friction modifier ?
Could I then assume that the second term is related to hydrodynamic lubrication, so can be modified according to the viscosity grade of oil (and oil temperature) I am looking at.
The third term is related to pumping friction, so I assume that there is no viscosity dependence for this term.
Does this approach sound sensible, or am I making some fundamental errors ?

 

[ivymike]

I don't think that I'd accept the assumptions you've stated.

Your fundamental approach is a curve fit, so if you already have engine data (measurements) you should probably be able to find coefficients to make things match up pretty well. If you're planning to pull coefficients out of a hat, I think you're barking up the wrong tree. Further, I don't expect that the coefficients that you calculate based on a set of measurements for an engine with a particular lubricant viscosity grade will be applicable to an engine using a different viscosity grade.

 

[mattsooty]

Dr Bob,

Perhaps some sort of DoE approach to this problem would be in order. I am guessing that, as was said above, you have either engine data or an engine to do this research with. By using an appropriate DoE you would be able to quantify your results use regression analysis fit a nice function to describe them.

MS