| Multi-stage centrifugal pumps are widely used in various industries.Considering the practical needs of energy saving,centrifugal pumps have much room for improvement in hydraulic performance.The performance of the inter-stage guide vanes plays a key role in hydraulic performance.The passage guide vane as one type of inter-stage guide vanes is the research objects of this paper.In this paper,numerical simulation is used to compare the performance difference between radial guide vane and passage guide vane.The design and structural optimization analysis of passage guide vane are carried out by using artificial neural network,and genetic algorithm.The main contents of this paper are as follows:(1)Firstly,an existing multistage centrifugal pump and its inter-stage guide vanes are taken as research objects.The hydraulic analysis reveals that the loss occurring in the radial guide vanes is caused by the occurrence of vortices in the transition zone of the forward and reverse guide vanes;(2)Two different passage guide vanes are designed.The difference between the two models is the distribution of the vane angle.Then a good design is selected.The difference in efficiency and head of those models verifies the advantages of passage guide vane over radial guide vane(efficiency increased by 5.9% while head increased by 8.0% in design point);The analysis of flow path around bend area suggest that increase the flow area of fluid at the bend and modify the increasing tendency of flow area in bend,which improves the flow under the design point and high flow rate conditions;(3)The influence of the clearance between the vane and the impeller outlet and inlet on the performance of the centrifugal pump is discussed,and the optimal values are determined.The influence of guide vane outlet angle on impeller is discussed.It is good to have a relatively large outlet vane angle in order to obtain positive incident angle for impeller as it turns out.The analysis of the inlet flow path of guide vane reveals that decreasing the guide vane inlet width can improve the performance under the design point and the low flow rate condition.As for high flow rate conditions,such modification reduce the efficiency of pump.The special cut of the guide vane inlet effectively improves the efficiency of the pump,and makes up the disadvantage of reducing the inlet width of the flow passage.The manual optimal model's efficiency is increased by 1.5% while head is increased by 2.2% in design point;(4)The manual optimal design model is parameterized to determine which parameters to optimize.A database with sufficient samples is generated.After comparing multiple linear regression and BP neural network for building approximate model,neural network is chosen to build the approximate model.Genetic algorithm is performed to get the best efficiency while withholding the required head,thus the best design is obtained.The influence of the number of the vane on the hydraulic performance was determined which is that the number of guide vane should be small.The optimized design provides qualitative guidance for designing passage guide vane angle.The optimized model's efficiency is increased by 0.4% in design point while head is sufficient to design requirement.The results of this research have certain guidance for the passage guide vane design and structural optimization. |
PDF Full Text Request