Multi-objective optimization and experimental investigation of radial gas wave refrigerator

Wave rotor blade shape is a key factor affecting the performance of radial gas wave refrigerators. This study combines sampling, surrogate model, and multi-objective optimization to enhance channel design for improved overall performance. Initial samples are generated using the Max-Min distance Latin hypercube sampling method. The surrogate model suggests that achieving both maximum temperature drop and minimum power consumption is not feasible. The FAST and Sobol sensitivity analysis quantify the influence weight. The greatest impact is attributed to the inlet angle, followed by the outlet angle, and then the convergence angle. Pareto solutions are finally obtained through the multi-objective genetic algorithm (MOGA). The chosen design parameters are α=74.69°, β=85.54°, and θc=2.48°. Power consumption is reduced by 10.6 % relative to the baseline model, while the temperature drop is increased by 4.3 %. Finally, an experimental platform is established to evaluate performance. The equipment maintains significant temperature differentials across defined operational velocity ranges, with refrigeration effect scaling positively with expansion ratio elevation. Power consumption scales positively with both rotational speed and expansion ratio. Moreover, cyclic pressure differentials intensify under elevated rotational speeds and reduced expansion ratios. Experimental results confirm a similar refrigeration effect to established axial configurations while promoting autonomous cryogenic gas circulation through limited energy input.