Numerical and Experimental Investigation of A Compressor with Active Self-Recirculation Casing Treatment for A Wide Operation Range
Liangjun Hu,Harold Sun,James Yi,Eric Curtis,Jizhong Zhang,Ce Yang,Eric Krivitziky
DOI: https://doi.org/10.1177/0954407013486741
2013-01-01
Abstract:A turbocharger compressor with a wide flow range and a high efficiency is important to the application of advanced clean combustion technologies, such as homogeneous charge compression ignition and low-temperature combustion, in diesel engines. Self-recirculation casing treatment is one of the techniques that can extend the compressor surge margin without much efficiency penalty. The underlying physics of the self-recirculation casing treatment technology were investigated with computational fluid dyamics modeling and bench testing in this study. It is identified that, if the bleed slot of the self-recirculation casing treatment is located upstream of the impeller passage's throat area, self-recirculation casing treatment improves the surge margin but the throat still limits the maximum flow capacity of the compressor. On the other hand, if the bleed slot of the self-recirculation casing treatment is located at the impeller passage's throat area, the self-recirculation casing treatment improves the maximum flow capacity but results in a significant compressor efficiency penalty in the low-flow range. An active self-recirculation casing treatment design was proposed. The active self-recirculation casing treatment design extends the compressor flow capacity and improves the surge margin without an efficiency penalty through dual bleed slots with one upstream and the other downstream of the leading edge of the splitter blades. In the choke condition, the upstream bleed slot will be closed; near the surge condition, the downstream bleed slot will be closed. In the middle flow range, both bleed slots are closed. Both the numerical data and the bench testing results show that the maximum flow rate could be extended by about 15% and the surge margin by about 20% without an efficiency penalty. The mechanism of the performance improvement is also numerically studied.