• Oil analyses can only provide the information that is contained in an oil sample in the first place. For an oil sample to be truly representative of the oil present in the gear unit, it is essential that it is taken correctly. The guidelines for this can be found under Sampling.
• The Sample Information Form must be completed in full. It is essential that all the information is correct.
• Lubricant analyses should be performed at regular intervals so that trends can be reliably identified.
   

If, as in the case of your first gear damage, a relatively large part of a single gear tooth has broken out and it is clear from the break edges that this occurred as a smooth break due to an obviously old crack formation, no particles can be found in an oil sample to indicate this breakout at all. The broken out tooth part remains in a pocket of the planetary gear due to its much higher specific weight. Or it may be flushed to a corner of the gear where there is little oil turbulence. It may also have been transported to the filter and filtered out so well there because of its size that none of it made it into the sample bottle. A particle a few millimeters in size that has broken out of a gear tooth, which could not get between other teeth and has thus been crushed, then provides just as little information as a screw or washer that has fallen into the gear oil. Only if breakout particles get between the tooth flanks are they crushed in such a way that they remain in suspension in the oil for a relatively long time and can thus be found in a reference sample.

In the second case, significant pitting and pitting-like damage to the tooth surfaces have occurred. However, you state that the pits no longer have sharp edges. This is a clear indication that the damage occurred some time ago, e.g. in the form of bedding-down pitting. The rounded edges of the pittings show that they are now already scarred because, for example, a change of oil grade or a change in the load spectrum has prevented further pitting. However, these pittings are still a typical trigger for crack formation, which can lead to a violent fracture in the form of the tooth eruption described. The associated oil sample shows a very good oil condition, indicating an oil service life of less than 10,000 operating hours. The oil in the analyzed sample was certainly not in service for the approximately 70,000 hours mentioned. There must have been several oil changes over the life of the gear unit. Unfortunately, there are no trend analyses available for the gear unit that could show that pitting components were very much present in the oil sample during earlier samples. These could have been detected at that time via optical particle analysis. However, if, as the scarring of the pits shows, the damage to the tooth surfaces occurred several oil changes ago and the information about it has already been flushed out with earlier oil changes, we obviously can no longer find anything in the current oil sample.

If oil analyses are carried out regularly - and preferably every six months, or at least once a year - as trend analyses, pitting, and gray leakage, which are usually the cause of gearbox destruction, can be reliably detected. The reason for their occurrence and the degree of damage can also be detected in this way. However, if only a one-time analysis is performed, those damages that already provided their information to the oil during previous oil fillings can of course no longer be detected because they had lost their information content with the oil change. Provided that samples are taken correctly, oil analyses are a cost-effective and incorruptible tool for your preventive damage detection in gear units. By the way, it is not without reason that all gearbox manufacturers, insurance companies and manufacturers of wind turbines consider oil analyses to be an essential tool for preventive maintenance.