Meta-model based optimization of a large diameter semi-radial conical hub engine cooling fan

Abstract

-Turbomachinery design is an iterative process that can be time-consuming and expensive, especially when an extensive knowledge of the performance envelope is required. The approach described in the present paper can significantly cut the turnaround times down without jeopardizing the accuracy of the final result. A parameterization technique based on radial basis functions (RBF) is used and Reynolds Averaged Navier-Stokes (RANS) simulations are subsequently performed on a set of selected morphed meshes, the goal of which is to produce an aerodynamic database containing first-order, second-order and second-order cross derivatives of objectives with respect to parameters. New solutions, corresponding to any variations of the selected parameters, can thus be extrapolated thanks to the information included in the aforementioned database. In this way, a meta-model is built and can be easily explored by a genetic algorithm. This approach has been experimented on a new concept of engine cooling fan featuring low torque and high efficiency. A reference fan design has been adapted for the particular surrounding of the vehicle underhood, where the downstream flow is radially deviated from its axis by the engine. The optimization process has resulted in an efficiency improvement of three points for one of the obtained optima.Key words: Turbomachinery design / computational fluid dynamics (CFD) / parameterization / mesh morphing / meta-model / optimization / engine cooling fan Study context and reference designFan systems are used in automotive engine cooling modules to increase air flow rate through heat exchangers. These latters are dedicated to the thermal management of the vehicle, for both the engine (radiator and charge air cooler) and the cabin climate control.The willing to decrease power consumption of electrical accessories in the vehicle incites to enhance fan system efficiency and numerous studies have been performed towards this goal. Among them, an investigation [1] aimed at reducing the losses produced by the tip inverse flow, which is found between the shroud and the rotating ring attached to the blade tip. However, modifications in the blade load distribution called for an additional study [2], and performance improvements were finally observed after a blade tip readjustment. Although the rotating ring a Corresponding author: zebin.zhang@ec-lyon.fr is frequently used to limit tip flow recirculation, its contribution to shear losses is significant and might be a disadvantage to achieve low-torque design. Therefore, investigations focused on a new type of fan design without external ring.A conical hub has also been integrated to the new geometry in order to produce a down-stream semi-radial flow. This particularity is more suitable for vehicle underhood conditions, where the flow is radially deviated by the engine. For a conventional cylindrical hub fan, it has been observed that the lower radius part of the fan blade contributes very little to the guidance of the air flow through the vein in the bl...

Sampling strategies for metamodel enrichment and automotive fan optimization

“…His method proved to be very fast because he could use a cokriging technique that drew additional information from the first and second derivatives. Of course, the great difficulty there is to have access to these values, and it implies either to derive the Navier-Stokes equations [9] [10] or to assess the derivative with a differentiation method [11].…”

“…Two other Latin Hypercube Samples (LHS) designs (which were optimized for discrepancy) of 25 points each have been established for testing the meta-models trained from the 127 runs (and the points for enrichment if any). The first one is set in [-1, 1] 11 , and the second one in [-0.5, 0.5] 11 . For this later it means that all factors are limited in the range [-0.5, 0.5], i.e.…”

“…The corners (factors equal to 1 or -1) are systematically proposed (figure 5), since they are in the extrapolation zone where the variance becomes large. Unfortunately, it would be hopeless to try any global improvement by selecting the 2 11 corners. Alternatively, adding information in the center could be of interest since it would help to improve the model in the 11 directions.…”

“…By limiting the selection in the [-0.5, 0.5] 11 domain, it is observed that the first point is very close to the center (red point in figure 6), and then the other points are in the corners of the limited domain. To compare the effect of adding point either in [-1, 1] 11 or in [-0.5, 0.5] 11 , 10000 points from a Sobol sequence were generated and used to compare the mean of their kriging variance. However the mean of the variance criterion was assessed by removing the central point of this sequence because it is the first added point of the second method.…”

“…However, each simulation run has a non-neglectable cost which is not compatible with an intense sampling campaign. In the present case, 11 design variables have been selected and the full factorial plan would yield 2 11 simulations, which is of course simply not possible. Some strategies are then investigated in order to enrich initial Design of Experiment (DoE) which are classically based on a statistical approach uncorrelated from the studied physics.…”