This is the story of a seemingly impossible challenge that we faced, and how we overcame it using innovation and physics based optimisation.
The challenge was the inter-cooler duct, which takes compressed air, a swirling high-speed flow, expands it, and channels it into the rectangular inlet of the intercooler. This may sound simple, but airflow doesn’t like to be turned, rotated, and expanded all at the same time. We learned this lesson during our work on the serpentine intake project, where we faced a similar issue of flow separation and significant blockage.
But the inter-cooler duct posed an even greater challenge. We couldn’t use solutions like a vortex-generator or compressed air injection due to blockage issues and weight limitations of additional equipment. To make matters worse, the tightly packaged engine assembly allowed us only a tiny margin of movement, with only +/- 2-3 mm surface deviation room.
We knew that we had to find a solution quickly. We were limited by the design envelope, which meant that we couldn’t change the CAD model too much, and we didn’t have time to use optimization algorithms based on parametric CAD models.
Instead, we turned to mesh-morphing, which allowed us to create numerous design variations quickly and test them in CFD straight away, without the need to re-mesh every time. This was a game-changer, as it enabled us to experiment with different shapes and configurations and rapidly converge on a viable solution.
Our approach involved a three-fold strategy. First, we changed the shape of the duct near the separation region to reduce the adverse pressure gradient. Second, we changed the upstream rate of turning by adjusting the duct curvature. Finally, we reduced the adverse pressure gradient by changing the rate of cross-sectional area expansion.
We manufactured and tested the duct on an engine testing rig, even though there was still some separation in CFD results.
What happened next was truly remarkable. In just two weeks, through a virtual-only project, we were able to achieve an on-track improvement equivalent to four weeks of wind-tunnel testing! It was an incredible surprise because we had not fully cleared the separation.
This project demonstrated the power of mesh-morphing and the value of an iterative, experimental approach to engineering problem-solving. It also taught us a valuable lesson: small changes in complex systems can often lead to significant gains. By changing all three aspects of the duct, we were able to achieve a remarkable reduction in separation, leading to improved engine performance.
So, the next time you’re faced with a daunting challenge, remember that even small changes can yield substantial improvements.
