Aerodynamics Numerical Optimization and Experimental Verification of a Low Reynolds Number Mixed Flow Blower for Multiple Operational Points

Small compression systems are widely used in household appliances to move air, and their performance has a direct impact on energy consumption and acoustic comfort. The development of high-performance fans and blowers is usually based on designer experience and existing hardware. However, proper design relies on matching the component performance to the application requirements such as range of operation, system restriction and noise regulations. In addition, products usually have different operational modes which change the restriction of the system and/or the compressor rotational speed. Modern CAE tools allow for parametrizing turbomachinery baseline cases, creating automated workflows for Design of Experiments and run optimization algorithms to generate the best candidate for specific criterion. Selecting the right parameters to optimize it is in itself a challenge usually not given the right importance. For example, when optimizing for compressor efficiency it is important to make the right point of operation and performance envelope is taken into consideration. This paper describes a methodology for using commercially available tools to numerically optimize a compression system and then experimentally verify the validity of the optimization design variants and its performance for an existing system restriction curve. The baseline design case — a mixed flow compressor with low Reynolds number — is a blower currently being used in existing products. There is an opportunity to improve its performance by optimizing the design to better match the new product restriction curve while respecting the same overall shape and size — only the number of blades and blade angle distribution were changed, which allows the optimized design to fit even into the existing products. Two CFD based Design of Experiments and one surrogate based optimization have been performed and their best design candidates for different criteria were selected, prototyped, and experimentally tested in an aeroacoustics rig. The optimum design candidates were achieved in a relatively modest number of CFD runs. The new design widened the operating curve, which allowed the design to operate in a lower RPM to meet the same performance as before the optimization. The combination of new blade design and lower RPM resulted in a compressor with the same overall sizes which consumes almost 5% less power while still delivering same flow rate and pressure as before.