David C. Arters

An investigation of long and short duration tests for evaluating engine non-starts caused by CCD flaking

Two engine dynamometer test protocols are compared for their ability to discriminate and duplicate the field phenomenon of engine non-start due to combustion chamber deposit (CCD) flaking. The first, a protocol based on a 625 hour deposit accumulation cycle, has been shown in prior work [1, 2] to reflect field experience and discriminate the effects of various fuel additive treatments. The second, a protocol based on a 60 hour deposit accumulation cycle, was developed in an attempt to significantly reduce the time, and thus cost, of testing. Results indicate the shorter protocol is repeatable and has similar discrimination with respect to fuel and fuel additive impact on the no-start phenomenon. There are, however, differences in the results that indicate there may be a severity difference between the tests. The tests both show there are clearly differences in the engine no-start impact of deposits formed by fuel and additives. It is also found that while all CCDs produce flaking, and in all cases significant cylinder pressure loss occurs in one or more cylinders during cold start, neither of these observations correlate with start or no-start results. Clearly, engine no-start due to CCD flaking is a complex phenomenon that cannot be easily accounted for via secondary measurements.