Hi Daniel,

Sorry if I am missing something, but I have three (maybe naive!!) questions to pose:

Cheers !!

-Sanjay

You can detect tdc signal by running the engine without fuel and compression TDC will be the peak pressure.. this can done done but cutting engine fuel supply and crank with starter motor while reading presure signal. But this also depend on the accuracy and sampling rate of your pressure sensor

Many thanks for your replies, Sanjay and Taib! I am using a low-end commercial indicator unit without the option to connect a CAD or TDC sensor. Which is the source of my problem since the position of the curve in CAD space as delivered by the instrument is indeed fairly arbitrary (more often than not the combustion peak rather than the compression peak is close to zero CAD).

I figured since I already have a measured approximation p_meas of the real function p(V, ...) non-linearity would not be a problem. I mean, p can be non-linear wrt V or any other variable but not wrt itself. Right? If that is the case it should be possible to find a constant IMEP such that

Sum(p_meas(V) * Delta_V) = Sum(IMEP * Delta_V) = IMEP * Sum (Delta_V) = IMEP * Vd (Vd = displaced volume).

In the attached graphs (green: measurements, blue: calculated motored pressure, red: IMEP, turquoise: difference), the usual anomalies around the valve open/close points are indeed visible, but I figured these should be included in any power calculation.

Thanks for your suggestion of motoring the engine Taib, it might be possible for us to simulate that situation when the generator is synchronised with the grid and idling rather than connected to a dummy load.

Sir, when the engine is motored, the peak pressure in NOT at TDC. This is due to the heat losses to the in cylinder wall and mass leakage. The result is that the peak pressure occurs before TDC and this phase shift is known as Loss Angle (around 1 CAD).

You will find some useful information at: 

http://emilianopipitone.altervista.org/files/pubblicazioni/TDC%20Determination%20-%20ATE3072.pdf

The IMEP can be computed once the TDC position has been evaluated (or fixed), but be careful because a 1 CAD degrees error in TDC position determination may introduce an IMEP evaluation error up to 10%. 

Good job

Emiliano

Daniel,

Now that you have described the various handicaps that your setup has, I must agree with Emiliano that given all that, your attempt is a good one indeed !

Also, Emiliano's caution about the TDC detection is a good one, though it is commonly known.

BTW, by my point #2 I did not imply pressure being nonlinear with itself (maybe I was not able to convey my point too well). What I meant was about the average of a product of two variables is equal to the product of their averages assumption.

IC engines is an "old love" for me (back from my Doctoral days), though I am an Electrical Engineer by profession and career. It was great to discuss your experiment with you.

Cheers !

-Sanjay

Thanks for your interesting input and publication link Emiliano, during rudimentary numerical experimentation I did indeed observe myself that IMEP sensitivity to assumed TDC is very strong and good accuracy would of course be a enefit. For the time being I will stick to manual/visual fitting though, and see if I may be able to find time to implement some more sophisticated routine in the future. As you point out Sanjay, my setup is indeed pretty primitive, then again results do not need to be of research quality, just convincing enough for the students to buy into, i e in the sense that indicated power should preferrably turn out a bit higher than measured electrical power. :-) Thanks again everyone for enlightening me on this issue!

Have you tried mechanically indicating TDC using a sensitive dial indicator and then referencing your shaft-encoder to TDC?  During measurement we typically try to indicate TDC to within less than 0.2 CAD.  Our shaft encoders are all either 0.1 CAD or 1/8 CAD resolution.