GSXR750 Cycle Simulation, 03-9396Printer Friendly Version
Inclusive Dates: 05/08/03 - 09/08/03
Background - High performance motorcycle engines are currently the highest specific output naturally-aspirated production reciprocating internal combustion engines in the world. These engines typically have excellent flowing intake and exhaust ports as well as intake and exhaust systems tuned to provide beneficial flow pulsation characteristics at critical speeds to increase performance.
This type of engine is a challenge to model with current "one-dimensional" cycle simulation tools. Although this code has been proven on lower speed engines, its capability to model the most highly dynamic, tuned breathing systems at engine speeds approaching 15,000 rpm has not been shown.
A 2002-model 750 cm3 four-cylinder motorcycle engine is currently undergoing testing in an Institute test cell. Performance data including high speed cylinder pressure data is available as are flow bench results of cylinder head port performance and valve lift measurements. This engine will be the subject of this modeling effort.
Approach - The approach taken was to make careful measurements of the motorcycle engine's components and geometry, valve lift profiles, and cylinder head port flow performance. This information was then used to build an accurate representation of the engine in the cycle simulation software. Limited existing data were used to calibrate the model. At the same time, efforts were made to obtain additional engine test data, specifically including high speed pressure measurements in the intake and exhaust systems of the engine to help quantify the pulsation activity in those components. The model would then be compared against this high speed intake/exhaust system pressure data and tuned to match as needed. These studies would be performed with the empirical combustion model, basically forcing the model to have a heat release profile with similar characteristics to the measured data. As a secondary objective, the predictive combustion model would be tested to determine its applicability to such high speed engine combustion.
Accomplishments - The engine model was built from careful measurements made of the intake and exhaust system, including a large airbox and 4 into 2 into 1 exhaust system. The measured valve lift and port flow performance was instituted into the model.
The model was tuned to the existing data, which consisted of low speed data (e.g. engine speed, torque, fuel flow, air flow) and cylinder pressure which was available from previously existing data. Analysis of the data and attempts to match it with cycle simulation showed some limitations in the existing data. The data was matched in a general sense of the low speed data, although the trends across the speed range could be better. (See Figure 1, below.) Also, the pumping work, which is a strong indicator of the port restriction and breathing system pressure dynamics matched well with the measured data, as shown in Figure 2.
A brief sensitivity analysis was also performed of the predictive combustion model. Although a secondary focus, it was desired to understand the capability of the built-in predictive combustion model to provide a realistic combustion rate for such a high speed, high performance engine. Sensitivity analyses indicated that the ignition delay is much shorter than is predicted at default settings, but can be matched with adjustment of parameters.