The effect of combustion chamber design on the combustion rate in an SI engine

The effect of combustion chamber design on combustion rate has been investigated experimentally and theoretically. The experimental work concentrated on the measurement of cylinder pressure and flame speed using a piezo-electric pressure transducer and multiple ionisation probes together with a data acquisition/processing system. A total of twenty one chamber designs of varying shape, compression ratio and spark plug arrangement were tested over a range of operating conditions on a single cylinder S.I. engine. The pressure data were analysed to obtain values of pressure rise rate, cyclic dispersion and combustion (mass burn) rate whilst the ionisation data were processed to yield flame travel angles and flame dispersion. The results obtained show that for a given compression ratio, the flame speed is not significantly affected by chamber design. In contrast, the combustion rate and pressure parameters are highly dependent on the chamber design; more compact arrangements giving higher combustion rates and reduced cyclic dispersion. A computer simulation model of the compression, combustion and expansion phases of the engine cycle was developed to predict the effects of the combustion chamber design parameters. Based on the experimental results, the model assumes that the ratio of laminar to turbulent burning velocity is independent of chamber design. The influence of chamber shape on the burnt volume, flame front surface area and heat transfer surface areas is modelled using a simple but effective geometric integration technique. This technique allows an infinite variation of the design parameters to be specified for a large range of chamber shapes with a minimum of input data being required. The model predicts that chamber design does have a major effect on combustion rate and cylinder pressure but shows that the influence of individual design is highly dependent on the setting of all other parameters. The effect of squish area is shown to be due to it changing the compactness of the chamber, optimum squish area being about 50% for conventional engines with higher areas being suited to higher compression ratio designs. Spark plug arrangement is predicted to be the most effective way of controlling the combustion rate with a single centrally located spark plug or alternatively, dual spark plugs, giving large increases in combustion rate. Computer model predictions have been compared directly with experimental results obtained in this study and with experimental results reported by two other independent workers. Good agreement was obtained thereby giving support to the assumption of the flame speed being unaffected 'by chamber design. The model was also used to predict squish velocities in fired engines. The results show that the velocities and, in particular the reverse squish, can be significantly modified by the combustion process with a strong dependence on ignition timing being evident. The predictive model has been modified to yield a heat release program capable of analysing experimental pressure time data to predict combustion rate, flame speed, turbulent burning velocity and many other variables. The predicted flame speeds were in good agreement with corresponding experimental values obtained from ionisation probes. In conclusion, the study has confirmed the importance of combustion chamber design as a means of improving the combustion rate but has shown that the flame speed is not affected by chamber shape (i.e. squish). The semi-empirical simulation model has been shown to predict the effects of the chamber design parameters to an acceptable degree of accuracy.