The exhaust mass flow measurement for motorcycles poses a unique challenge due to presence of pulsations arising from an unfavorable combination of the engine displacement-to-exhaust system volume ratio and the long or even unequal ignition intervals. This pulsation phenomenon significantly impacts the accuracy of the differential pressure-based measurement method commonly employed in on-board measurement systems for passenger cars. This paper introduces an alternative approach calculating exhaust mass flow in motorcycles, focusing on statistical modelling based on engine parameters.
The problem at hand is rooted in the adverse effects of pulsations on the differential pressure-based measurement method used in the EFM. The unfavorable combination of engine characteristics specific to motorcycles necessitates a novel approach. Our proposed alternative involves utilizing readily available OBD parameters, namely engine speed and calculated engine load as there is mostly no data for intake mass flow. The intake mass flow is then calculated using the SAE-based equation, offering a suitable and robust method for exhaust mass flow estimation in motorcycles.
Acknowledging certain simplifications in the model, such as the absence of lambda information and variations in engine calibration quality, deviations were systematically measured for various motorcycles on a chassis dynamometer and subsequently verified. The database-driven calculation may not always align precisely with the reference, but it consistently demonstrates a strong regression on the two parameters, validating its overall reliability.
Furthermore, the paper explores the integration of the collected comparison data with the chassis dynamometer into statistical modelling software. A model specific to each motorcycle was developed, and its quality was assessed through renewed measurements on a chassis dynamometer. Notably, the driving cycle selected for these measurements deviated significantly from the previously used cycle to create the model. The values recorded from the OBD during this drive serve as an input for the modelling process to further compute the exhaust mass flow derived from the CVS. This approach significantly improves accuracy and offers a viable alternative to direct measurements, particularly for small two-wheelers.
The relevance of this statistical modelling-based approach is particularly pronounced for small two-wheeler vehicles, as they face challenges associated with the additional weight of an EFM and the accuracy limitations of such devices. The proposed methodology provides an alternative that maintains high determination accuracies while addressing the specific intricacies of motorcycle engine dynamics. This research contributes to advancing exhaust mass flow calculations for motorcycles, offering a valuable tool for on-board exhaust mass flow estimations on two-wheelers.