Missing combustions phenomena have been studied for a long time, in order to achieve on board diagnosis capability, due to their effect on emissions and after treatment system life. A detailed analysis of missing combustion causes can be used beyond the On Board Diagnostic (OBD) regulations observance aim, both for modeling and diagnostic purposes.
The combustion taking place in a faulty cylinder right after the fault happens is usually different with respect to a standard one: both the residual gas fraction and the fuel film effect have strong influences on post-faulty combustions. Parameters used to describe these phenomena can then be evaluated analyzing the engine post-faulty behavior. The same analysis may lead to a better fault description, thus improving diagnosis performance.
Residual gas fraction can be evaluated by means of models or with direct measurements: on one hand models always need to be validated, on the other hand the setup needed for the experimental evaluation is costly, and tests are time-consuming. A possible alternative approach could be based on the observation of residual gas fraction consequences on measured signals during particular operating conditions: missing combustion tests can be used for this purpose. Such a methodology can be easily used to validate models results, carrying out misfiring tests by means of a programmable Engine Control Unit (ECU) or a malfunction induction system. The same philosophy can be used, for example, for the calibration of the fuel film model parameters.
Missing combustions analysis can obviously also be used for diagnostic aims: the attention has been often focused on missing combustions detection, but sometimes it would be useful to go further, determining the fault causes. It is still not mandatory, according to international OBD regulations, to distinguish between missing ignition and missing injection faults, but the distinction could be useful for fault-cause isolation and for proper recovery strategies definition.
The paper shows how observations of the engine faulty behavior can be used, in order to improve physical models describing standard running conditions. Diagnostic capabilities can as well take advantage of models improvement.
A 4 cylinders 1.2 liters spark ignition port injected engine, equipped with a programmable Electronic Control Unit (ECU) has been tested on the test bench, inducing faults whose consequences have been analyzed.