Gentlemen, when one uses a dynamometer, the braking device usually has a trunion support and at a specific radius from the center of the trunion it has a load cell that is measuring the torsional reaction force that the brake is applying to the engine crankshaft. Theoretically, a similar job could be done with a torque transducer on the input shaft of the brake (in series) except that the torque transducer may need to have slip rings to get its signal "out" and excitation to the strain gauges. Slip rings can get expensive and tricky to maintain.
So, if one assumes that one is operating with an engine/redrive/prop, the prop is the part absorbing the load and the engine mount will be where the reaction forces are. So one could surely make the mount in such a manner that one could have a torque arm and a load cell and should thereby be able to measure the reaction torque of the system and from that derive the engine power ? While it means a custom engine mount, it seems that would be considerably cheaper than buying a dynamometer ?
If it was to be as simple as possible (mechanically) then there would be a bearing near the output shaft of the re-drive which allowed the engine assembly to rotate exactly in line with the propeller shaft. One would then put a plate on the opposite side of the engine (at the accessory end) having a second pivot exactly in line with the prop shaft of the redrive. So now the entire engine assembly is capable of pivoting smoothly on an axis going through the prop shaft on both sides. Now one can add to the engine an arm which is attached to a load cell to measure the torque reaction. There could be some extra safety chains in both directions to anchor the engine if part of the load arm or load cell were to fail, but they would have some slack in them to allow the load cell to measure without interference or preload from the chains.
Such a system might have a tiny bit of hysteresis because of friction in the pivot and the torque value will oscillate based on the 2nd and 4th order frequency of a 4 cylinder engine but that is the case on all torque measurement equipment on any dynamometer, there would have to be some filtering applied to smooth out the signal somewhat. Do systems like this exist ? It seems like there is no reason to never have "real" data on the peformance of your engine, regardless of what you use as the "brake" ?
So, if one assumes that one is operating with an engine/redrive/prop, the prop is the part absorbing the load and the engine mount will be where the reaction forces are. So one could surely make the mount in such a manner that one could have a torque arm and a load cell and should thereby be able to measure the reaction torque of the system and from that derive the engine power ? While it means a custom engine mount, it seems that would be considerably cheaper than buying a dynamometer ?
If it was to be as simple as possible (mechanically) then there would be a bearing near the output shaft of the re-drive which allowed the engine assembly to rotate exactly in line with the propeller shaft. One would then put a plate on the opposite side of the engine (at the accessory end) having a second pivot exactly in line with the prop shaft of the redrive. So now the entire engine assembly is capable of pivoting smoothly on an axis going through the prop shaft on both sides. Now one can add to the engine an arm which is attached to a load cell to measure the torque reaction. There could be some extra safety chains in both directions to anchor the engine if part of the load arm or load cell were to fail, but they would have some slack in them to allow the load cell to measure without interference or preload from the chains.
Such a system might have a tiny bit of hysteresis because of friction in the pivot and the torque value will oscillate based on the 2nd and 4th order frequency of a 4 cylinder engine but that is the case on all torque measurement equipment on any dynamometer, there would have to be some filtering applied to smooth out the signal somewhat. Do systems like this exist ? It seems like there is no reason to never have "real" data on the peformance of your engine, regardless of what you use as the "brake" ?
Last edited by a moderator:
