Hao Hu

Failure analysis of melt urea pump impeller

Abstract A melt urea pump is out of work after only 6 days in service. The failure of the melt urea pump impeller is analysed using different methods including chemical composition analysis, metallographic observation, SEM and energy dispersive spectroscopy analyses, etc. The results show that the material of the impeller is not the reported AISI 321 stainless steel but AISI 316 stainless steel. There are many honeycombed holes on its impeller, which is the typical feature of cavitation damage. Therefore, the main failure reason of the impeller is the cavitation erosion, and the electrochemical corrosion of the medium also accelerated the failure of the impeller. The failure of the impeller is not related with the as casted defects of the material because no obvious as casted defects are found in the material.

Simultaneous Reverse Time Migration of Primaries and Multiples without Multiples Prediction

Prestack reverse time migration (RTM) is usually regarded as an accurate imaging tool and has been widely used in exploration. RTM using multiples can improve imaging quality compared with conventional RTM which only uses primaries, but the multiples often need to be predicted, which is time consuming and prone to errors. We propose a new RTM approach which uses primaries and multiples simultaneously in migration process. In our approach, the source function in conventional RTM is replaced with recorded field data including primaries and multiples, together with a wavelet retrieved from direct waves; the back propagated primaries in conventional RTM are replaced with complete recorded field data. The imaging condition of proposed approach is same as the crosscorrelation imaging condition in conventional RTM. The Sigsbee 2B synthetic dataset are used for numerical experiments. The numerical results show that the proposed approach can provide better illumination of shallow scatterers compared with conventional RTM. The proposed approach is easy to implement and avoids tedious multiples prediction.