Real-Time Evaluation of IMEP and ROHR-Related Parameters

Combustion control is one of the key factors to obtain better performance and lower pollutants emissions for diesel, spark ignition and HCCI engines. This paper describes a real-time indicating system based on commercially available hardware and software, which allows the real-time evaluation of Indicated Mean Effective Pressure (IMEP) and Rate of Heat Release (ROHR)-related parameters, such as 50% MFB, cylinder by cylinder, cycle by cycle. This kind of information is crucial for engine mapping and can be very important also for rapid control prototyping purposes. The project objective is to create a system able to process in-cylinder pressure signals in the angular domain without the need for crankshaft encoder, for example, using as angular reference the signal coming from a standard equipment sensor wheel. This feature can be useful both for test bench and on-board tests. In order to gain reliable results or acceptable precision on ROHR-related parameters (ROHR peak & 50% MFB, for example) a high sampling rate is required for the in-cylinder pressure. Since the angular reference signal can have low angular resolution (6 degrees with a typical sensor wheel), the in-cylinder pressure signal sampling rate must be higher than the crankshaft signal frequency. The choice for this application has been to use a high sample rate on a time base for the cylinder pressure signal, performing the transformation from the time domain to the angular domain (necessary in order to evaluate the indicating parameters) by means of an interpolation algorithm. The system features a signal conditioning module allowing to plug directly VRS/Hall effect/encoder as reference sensors; the signals, thus converted to TTL level, are digitally sampled at high frequency for the crankshaft position recognition. In-cylinder pressure signals, instead, are sampled @ 100 kHz. The conversion of these samples from the time domain to the angular domain is triggered by the sensor wheel signal. The algorithm used for the conversion can be time and memory consuming: the paper shows that the methodology used is crucial in order to save hardware resources, i.e., to increase the number of analyzed signals. As regards the hardware choice, many requirements have been taken into account: portability, sampling rate, computational power, programming language, external devices interface, etc. The final solution is based on a portable Real-Time/FPGA-based hardware, which allows performing all the necessary I/O functions and calculations in real time, even at high engine speed and with a high number of cylinders.