A discrete genetic algorithm-based efficient low-pulsating vane pump optimization method

[0108]Example 1:

[0109]With a design speed n = 2910r / min, traffic q = 50m3/ H, the rigs H = 20m, the number of single-stage monodarapel pumps of the blade of z = 6 is an example, and the shaft surface projection isfigure 2 Indicated. The method of the present invention maintains an impeller inlet width B1, Impeller import diameter D1The number of blades is z, and the blade thickness δ is unchanged.

[0110]The chromosome information was initialized using a Latin ultra-cubic sampling method.

[0111]Call the current chromosome information to call ANSYSBLADEGEN through the ANSYS WORKBENCH.image 3 As shown, each chromosome corresponds to an impeller.

[0112]The resulting impeller model is automatically divided into the ANSYSTURBOGRID through ANSYSWORKBENCH.Figure 4 Indicated.

[0113]The resulting mesh and the hydrotavername mesh are set to the ANSYSCFX through ANSYSWORKBENCH to make the initial result of the next non-constant numerical simulation with the current numerical simulation result. By the ANSYSWORKBENCH, the non-constant numerical simulation is performed in ANSYSCFX. The main frequency frequency value of the pressure pulsation in the last six calculation cycle corresponding to the impeller is obtained by MATLAB, and automatically input into the MATLAB. MATLAB feeds the results feed back the IDGA algorithm, as a chromosome evaluation value to calculate the appropriateness of the chromosome on the calculation domain, and determine whether the above results satisfy the iterative stop condition. If so, stop iteration, output optimization results; if it is, the chromosome is operated, and the new chromosome is obtained, and the above automatic modeling, automatic dividing mesh, calculation, and evaluation.

[0114]The impeller in this embodiment was optimized under a rated condition, and the pump efficiency was increased to 79.8% (78.1% of the original model), which was relatively increased by 2.18%. Such asFigure 5 (a) - Figure 5 (d)As shown, the main frequency amplitude of the pressure pulsation frequency domain main frequency of the two monitoring points is optimized, and the main frequency of 18.97 kPa and 21.24 kPa (the original model tongue is mainly transmitted by the pressure pulse frequency domain The frequency amplitudes at the point of 19.14 kPa and 23.38kPa respectively), respectively, with respect to 0.89% and 9.15%. The design parameters of the optimized model are: impeller outlet diameter D2= 132mm, impeller outlet width B2= 17mm, blade outlet chamfered radius r2= 1.5, blade import placement angle β1= 40 °, blade exit placed angle β2= 28 °, blade packet angle θ = 105 °.