Determining Engine Inertia

[Ratty66]

Hi All,

I need to determine the inertia of an engine I currently have on a dynamometer.

Unfortunately i am unable to get the required measurements to calculate, so I need some method of calculating from dynamometer data.
Is there a method of calculating it using data from acceleration and coast down rates? similar to those used for a vehicle.

Thanks in advance

 

[BrianPetersen]

If you can motor the engine (engine off, driven by the dyno) and measure instantaneous torque and RPM while you quickly ramp the speed up and down, you could fish it out of the difference in torque between accelerating and decelerating through a certain RPM range. There will be a lot of noise in the measurement.

 

[Ratty66]

I have an AC dyno, with a torque ring, so motoring isn't an issue.

If I ramp the engine between 2500 and 3000 and back with 2s ramps, and repeat the activity at a higher speed (say 5500-6000) would this work? would a longer ramp be needed?
Is there a simple calculation?

Thanks!

 

[BrianPetersen]

The shorter in time you can get that ramp time, the more the inertia will dominate the calculation instead of the "noise" in the data from friction and the compression and power strokes. The shorter you can get that ramp time, the better.

Your first-year physics text shows the relationship between rotational torque, acceleration, and inertia. It's F = MA in rotational units. It's all greek symbols that I don't know how to type on this keyboard. If you know your starting and ending ramp RPM and the time to ramp, you can get rotational acceleration (with a positive sign for the increasing RPM and a negative sign for the decreasing RPM). If you know the average torque over this time period for the increasing RPM and the decreasing RPM you can fish out the difference between the two as the torque required to accelerate or decelerate over and above whatever the friction is doing to you. Then you can calculate the other term. Don't forget that you are dealing with the difference in acceleration between acceleration (positive sign) and deceleration (negative sign) ... X - (-X) = 2X ...

One other thing; take the spark plugs out to get rid of some of the non-acceleration-related forces involved.

 

[LionelHutz]

Run at a continuous speed to get the torque required just to keep the engine rotating. You could do this at a few speeds between the test speeds to confirm you have a decent average torque number. Or, go very slowly and get a torque curve for between the 2 test rpm's. Then, do the fast ramping. The extra torque required to accelerate the motor fast is being used to accelerate the inertia. If all goes well, you'll get repeatable results as you vary the test acceleration speed or rpm ranges.

 

[EHudson]

Torque = Rotational inertia x angular acceleration.

When Lionel says 'run' I think "motored' is the usual term for driving the engine through the dyno current. This will let you find the torque required to overcome friction and pumping losses at say 1000 and 3000 rpm.

Spark plugs out per Brian. (Earplugs in)

With an average torque value, you can time the coast down from one rpm to the other. This time value and speed change allows the calculation of angular acceleration or de-acceleration in this case. (You'll probably need the value in radians per second squared eventually.) Plug into the equation above.

This gives you the value of the whole system, not just the engine. So you'll get the inertia of the engine plus the dynamometer equipment attached. With luck, you'll have the dynamometer inertia on a name plate or in a handbook. Subtract this value from the total to get the engine inertia.

 

[gruntguru]

If you ramp up and down at a constant throttle setting, the difference in torque is due to inertia. Turbocharged SI engines will be additionally affected by the turbocharger inertia but this will be minimal if you use a low throttle setting.

If torque is measured on the driveshaft the inertia calculated is engine only. If reaction torque of the dyno stator is measured the inertia will be all rotating components.

je suis charlie

 

[EHudson]

Convince me.

It seems to me that if the dynamometer is bolted to the engine, the rate of acceleration or de-acceleration depends upon the system inertia as a whole because the whole thing speeds up or slows down together. Any difference between driveshaft torque and dynamometer stator torque is the result of internal friction in the dynamometer.

 

[gruntguru]

The ramp-up and ramp-down functions are dynamometer-controlled acceleration rates - nothing to do with inertia.

The driveshaft torque required to hold the engine to the set rate will be instantaneous engine torque plus-or-minus (engine inertia x acceleration rate). (+- depending on accel' or decel'). The torque required from the absorption mechanism of the dyno (and measured by a typical lever arm on the stator) will be driveshaft torque plus-or-minus (dyno inertia x acceleration rate).

je suis charlie

 

[EHudson]

Okay, I understand how that works if your controls have the ability to set acceleration.