EAGER: CFD-based Adjoint Methods for Design Applications with Unsteady Separated Flows

一种称为伴随优化的计算方法被用于车辆设计中，以找到理想性能的最佳形状，例如提高升力或减少气动形状的阻力。从历史上看，该方法已用于简单流场中的流线型，但最近已将其应用于更复杂的形状，例如复杂湍流流场中的汽车。然而，流动模拟的准确性和设计过程的有效性尚不清楚。其中一个应用，气动优化的准确性是至关重要的是冬季奥运会的雪橇运动。历史表明，奥运会的奖牌是由只有千分之几秒的时间差决定的。空气动力学是复杂的，因为它本身的几何结构是复杂的，并且骑手/雪橇后面的流动涉及大规模的非定常湍流运动（类似于车辆）。传统上，雪橇设备是由前运动员开发的，是根据直觉、感觉和经验手工制作的。在这个项目中，针对非定常流动的伴随设计的成功设计、制造和实验验证，可能会对未来汽车/卡车、船只/船舶和先进飞行器的阻力最小化设计产生重大裨益。增加转弯/机动时的性能，或加速/减速时防止气流分离。这项工作的目的是通过实验验证，研究伴随方法在湍流、非定常、分离流动中达到最佳形状的有效性。将开发一种基于伴随的方法，并将其应用于雪橇/骑手模型，该模型的目标是修改雪橇形状以最小化阻力。实验测试将在整个过程中提供中间形状的性能数据，以确定伴随过程是否准确地预测实际性能并收敛到最优解。实验验证将通过表面压力测量、阻力测量和模型周围流场测量相结合来完成。除了这些技术目标之外，还有一个更高层次的目标，那就是将科学优化的比赛技术转让给美国无舵雪橇协会，以便在奥运会上使用。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

A computational method called adjoint based optimization is used in vehicle design to find optimal shapes for desired performance, such enhanced lift or reduced drag of aerodynamic shapes. Historically, this method has been used on streamlined shapes in simple flow fields, but more recently it has been applied to more complex shapes such as automobiles in complex, turbulent flow fields. However, the accuracy of the flow simulations and the effectiveness of the design process are unclear. An application where the accuracy of aerodynamic optimization is critical is the Winter Olympic sport of Luge. History has shown that Olympic medals have been determined by time differences of only a few thousandths of a second. The aerodynamics are complicated because the geometry itself is complex and the flow behind the rider/sled involves large scale unsteady turbulent motion (similar to a vehicle). Traditionally luge equipment has been developed by former athletes and was hand shaped based on hunches, feel, and experience. The flow conditions, and historic lack of science based design optimization, make luge sleds an ideal choice for investigation of the effectiveness of adjoint based optimization to perform aerodynamic optimization The successful design, build, and experimentally validating adjoint designs for unsteady flows in this project can potentially have a significant benefit to future designs of cars/trucks, boat/ships, and advanced air vehicles for drag minimization, increased performance during turns/maneuvering, or the prevention of flow separation while accelerating/decelerating.The goal of this work is to investigate the effectiveness of adjoint methods to reach an optimal shape in turbulent, unsteady, separated flows using experimental validation. An adjoint based method will be developed and applied to a luge sled/rider model that targets modification of the sled shape to minimize drag. Experimental testing will be performed to provide performance data of the intermediate shapes throughout the process to ascertain if the adjoint process is accurately predicting the actual performance and converging towards an optimal solution. The experimental validation will be accomplished by a combination of surface pressure measurement, drag measurement and measurement of the flow fields around the model. Beyond these technical goals is the high-level goal of transferring scientifically optimized race technology to the US Luge Association for use in Olympic competition.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.