[swamp2]

I don't think I can fully answer your questions but I can say some relevant things.

Originally Posted by bradleyland

Sorry, I wasn't clear. I was speaking in reference to the fact that most interia dynamometers use a fixed value for drive train inertia. My remarks are meant to highlight the difference in effect of drivetrain inertia between inertial dyno measurement, real world performance, and simulation.

This isn't really my area of expertise as to different dynos, but I believe none of them use any value for drivetrain inertia. After all, fundamentally they are all only measuring wheel torque, not crank torque. What is happening at the wheels is what it is, regardless of losses or inertia. Secondarily, if they do, it is almost for sure a fixed value regardless of vehicle.

Then there are the "simpler" corrections for environmental factors and conversion from wheel to crank values. The former is just a simple formula such as:



ref: SAE J1349 Aug 2004 revision

This factor, only valid up to a 7% correction, simply multiplies the entire torque/power curve by this correction factor (cf) to account for the difference between a reference environment (air density) and the actual one at the time of measurement.

The latter, wheel to crank, is also just a fixed correction, most often dividing by 0.85 (a "15% loss" as folks commonly say). I believe the intent of the latter is to capture both drivetrain inertia and true parasitic losses. Luckily, even the most unskilled dyno operators know that this 15% value is a crude estimate.

Originally Posted by bradleyland

So my point is really: on inertial dynamometers -- which assume a fixed drivetrain mass for all measurements -- wouldn't the actual difference in drivetrain mass have a more statistically significant effect than we would see in real world performance and simulation?

To make a claim, rather than asking a rhetorical question, it would appear to me that the error in measurement on inertia dynamometers (which use fixed assumptions) would be greater than the figures stated in the paragraph below:

Be careful with the difference between mass factor and inertia. Mass factors specifically (and rigorously) convert "spin-y bits into non spin-y bits". Mass factors depend on vehicle mass and tire radius. Inertia values are inherent to a component, independent of the their installation into a vehicle. The part you quoted was simply an crude sensitivity analysis to better understand how accurate the mass factors I calculated might be due to the somewhat crude estimation of the dimensions of some components (or quoted values extrapolated like the I4 to V8 engine).

Now if you want to compare "apples to oranges" I certainly believe that the errors, be them in the dyno measurement process itself, or in the use of either correction factor mentioned above are much higher than the relatively small (single low digit %) errors I mention for my calculated mass factors. After all we have the simple empirical evidence that dynos vary about 15% on the E92 M3 regardless of dyno type.