Search Results

It is well understood that the consumption of lube oil should be reduced to minimum due to stringent emission norms and the need to reduce visual smoke. We have used statistical methods for optimising engine performance parameters due to limited time availability for product development. In this paper, techniques that have been used to develop and optimise a small 4-stroke gasoline engine have been described. This work is focussed mainly on oil consumption due to piston assembly and cylinder block. The piston ring assembly has been modeled using simulation software (RINGPAK). Design of Experiments (DoE) tools have been used to optimise the variables / critical factors. The significant factors were identified through ANOVA (Analysis of Variance) table on the above responses. A test procedure has been judiciously finalised to evaluate the optimised parameters. The best combination of factors/levels was arrived at by using the above results and was evaluated through test rig trials.

The knock sensor is provided on an engine cylinder block to detect abnormal engine combustion (knocking) and to provide feedback to engine control unit (ECU). The ECU then modifies the engine input and avoids knocking. A commonly used knock sensor is an accelerometer that detects cylinder wall vibration and estimates knocking of the engine. Selecting the location of a knock sensor in many cases involves a challenging trial and error approach that depends upon the measurement of the knock signal at many locations on engine structure. However, a cylinder block exhibits many structural resonances. Thus, a large vibration signal at the surface of cylinder block can be either due to knocking of the engine or due to the resonances of the cylinder block structure because of normal excitation forces. Hence, this conventional method does not always yield reliable results.