Subproject T1 Magnetic Bearing of Rotary Axis (finished 2016)

In order to detect defective components in aircraft engines, they currently have to be dismantled. The repair costs and thus the downtime of the engine are therefore very difficult to predict at present, as the damage only becomes visible during the ongoing regeneration process. The regeneration process is therefore not only time-consuming and costly, but also involves a great deal of uncertainty in planning.  Subproject A3 is developing a methodology that enables the condition of engines to be assessed before dismantling. The idea is that defective components in the hot gas path (HGP, area downstream of the combustion chamber) of engines have an influence on the local density distribution in the flow, which propagates as irregularities in the density distribution into the exhaust jet. With the Background-Oriented Schlieren (BOS) method, these irregularities in the exhaust jet can be detected without contact, and conclusions can be drawn about the underlying defect in the engine. This makes it possible to quickly check engines for damage so that they can be serviced as needed, and the early identification of damage allows for improved planning of subsequent engine maintenance.  

At the end of the project in the current third funding period of the SFB, the developed methodology will be validated on a real engine. For this purpose, defined defects will be introduced into an engine in two measurement campaigns and the exhaust jet will be reconstructed using tomographic BOS. The influence of the defects on the exhaust jet will be investigated by comparing it with measurements of the engine in the reference case (without defects). In addition, the influence of the defects on the exhaust jet will be predicted using a numerical model and validated with the results of the measurements. After this successful validation, it will then be possible to predict the influence of further defects on the exhaust gas jet numerically and thus build up a database of possible damage cases. An engine with an unknown condition can then be measured optically at a later stage and the results of the measurements can be compared with the damage database using pattern recognition algorithms. This allows defects in engines to be detected without contact, prior to disassembly.