In an SAE paper written by Chrysler, "Expediting Engine Development Through Near Time Oil Analysis", a real time oil analysis program was instituted for engine dyno testing. Their key criteria for terminating engine testing was rising wear metal rates. What is interesting is looking at their conclusions and the curves that they show. For the racing engine, they terminated the test with changes in ppm Al wear that would be argued on BITOG (7 ppm to 23 ppm). However, they do have good trending to support this, by sampling oil at a regular basis without changing.
Quote:
2.0L Touring Car Racing Engine Example - A sharp increase in aluminum concentration was noticed in the oil from two race prepared engines. There was no other indication of problems. Testing of the engines was terminated. Tear down identified a cam bearing failure in motor A and a cracked rocker arm post in motor C as the sources of the aluminum. Analysis of the oil avoided catastrophic failures. The motor B test was terminated as planned.
Quote:
Early in the evaluation phase of the oil analysis testing at Chrysler Jeep and Truck Engineering, a specific test stands out both as a failure analysis example and a procedural example. Oil samples were gathered on a regular basis from a V8 engine. The engine test continued to run into the weekend. Samples taken late Friday afternoon were set-aside until the beginning of the next week when a technician was available to run the analysis. In the interim the engine failed. At both the eight and ten hour test samples, the indicators suggested that a close watch or additional inspection was necessary since both iron and aluminum were rising sharply. At the twelve hour point, test termination was definitely indicated. This incident lead to procedures being developed which maintained timely evaluation of oil analysis data.
The engine tear dwon revealed scuffed pistons in all cylinders, worn connecting rod bearings and a failed connecting rod big end with bearing. The initial evaluation suggested that piston scuffing generated increased bearing load, which resulted in the bearing and rod failures. The OSA data was then added to the investigation of the root cause.
Examination of wear metal concentration in the oil showed beginning at six hours there was a steady increase in the concentration of iron in the oil. Between ten and twelve hours the concentration of aluminum doubled. These trends continued until fourteen hours when the test was terminated catastrophically. The conclusion after examining both the oil data and the engine components was the two failure modes were initially unrelated. The bearings failed due to loss of oil supply, not because of the scuffed pistons. Debris from the failed bearings aggravated the piston scuffing but was not the initial source. The steady growth in iron and aluminum levels from six to ten hours is credited to the piston scuffing. The large jump in aluminum concentration after ten hours was caused by debris from the failing connecting rod bearing aggravating the scuffing in two cylinders adjacent the failed connecting rod. It was subsequently found that the piston fit was out of tolerance and that machining errors had lead to the oil supply problems. Compounding the difficulty in evaluating the data was that the bearings were of high aluminum content.