ECU estimating of torque output

Hi. When logging I notice there is an "Estimated Torque Output" in the software I am using. How would an ECU go about estimating the torque an engine is producing? Is there a known and respected method for doing this, a formula that people could use for their own 'back of the napkin' calculations?

Thanks,

Two methods I'm aware of are estimating how much potential energy is in the air the engine consumes (Brake Specific Air Consumption or BSAC) or in the fuel the engine consumes (Brake Specific Fuel Consumption or BSFC). Both estimates should get relatively close but both methods involve some assumptions about how well the engine converts fuel (or air) to power. Both methods also ignore ignition timing which can have a significant effect on power.

Another method that can incorporate ignition timing is checking the vehicle's acceleration (speed vs time) and using Newton's second law (force = mass * acceleration) to back-calculate the force (power) from the engine. This is how the Virtual Dyno software works, as well as most smartphone-based apps since they tend to use GPS and an onboard accelerometer. This isn't perfect either since the slope of the road surface and headwinds/tailwinds can affect the estimate, but it does have a chance of measuring changes from ignition timing.

It's a good idea to compare any of these calculations against a known dyno graph to get a feel for how well they are estimating your engine's actual performance. And there are definitely some advantages to using a dyno for initial engine tuning, such as the ability to repeat the same run under very similar conditions and the ability to safely stop the vehicle quickly if a problem occurs.

Yes there are clearly benefits to a dyno, the cons are there too, $$$ :)