Adaptation of CFD to the Design of Centrifugal Pumps

The similarity between flow patterns predicted with the aid of CFD and those observed in practice indicates, that CFD might be capable to assist the pump -engineer in coming up with improved designs. Unfortunately, in practice, the results are rather elusive. This paper discusses some of the causes of this state of affairs and presents suggestions, how to eliminate them. It demonstrates that the main cause of these disappointing results is the existing gap between science and engineering. The present explosion of information makes it very difficult for an expert in any given field of engineering to follow all the new developments in his (her) own specialized area. The less can an engineer be distracted by attempts to master an additional field of expertise. The results arrived at with the aid of CFD demonstrate, that it is a very powerful and versatile logical tool. However, versatile also means that it can be used in many different ways. Consequently, in order to arrive at satisfactory results, it has to be applied in a manner specifically suitable for handling the problem, which is being worked on. This means, that it requires an in-depth familiarity with all its aspects. Something, a CFD expert cannot afford to study. The only solution is teamwork. However, for teamwork to be successful, each participant has to know and to understand, what the others are doing. This paper demonstrates, that a relevant interpretation of the physical meaning of mathematical equations is capable to provide logical explanations to some of the even most enigmatic events. This leads to the conclusion, that such interpretations may also be capable to establish the so badly needed mutual understanding between the pump-expert, and the specialist in CFD. The above conclusions are supported by actual case histories from practice, and/or by results of tests. In addition to the above, this paper discusses a case history, which demonstrates that even in cases, in which CFD is capable to provide the required results, its use is not always the best choice. In practice, it is often much simpler, faster and considerably less expensive to arrive at the required results by means, which can be hardly regarded as associated with fluids dynamics. Finally, this paper also presents an example of a problem, which only CFD may be capable to solve.