I was recently presented with a unique opportunity to compare the results of full ANSYS CFD simulations with the results obtained using the new ANSYS Discovery Live product, which provides results instantly upon changing the geometry without interrupting a run. I was very pleased and surprised by the speed and accuracy of Discovery Live in this comparison test.
I work for Astec, Inc., the subsidiary of Astec Industries that builds asphalt plants. Roadtec Inc., another Astec Industries company, builds asphalt pavers, reclaimers and material transfer vehicles (MTVs). An MTV helps to accomplish non-stop, non-contact paving by offering a continuous supply of Hot Mix to the paver. By separating dump trucks from the paver this way, contractors are able to make a smoother finished road.
MTVs are delivered to the job site on a flatbed truck that can reach speeds of 70 mph. Depending on the operator of the flatbed, the MTV might be facing forward or backward, resulting in different air flow patterns around the MTV. These patterns would produce very different forces on the components of an MTV during transport, so it was necessary to ensure that the components would be able to resist these forces no matter the orientation of the MTV on the flatbed.
My colleagues at Roadtec approached me initially to analyze the forces on an experimental canopy that is being considered to offer shade and comfort to the machine operator. The canopy is a thin, rectangular, non-metallic shield. Roadtec engineers wanted to make a support for the canopy to ensure that it would not fly off during transport on the flatbed, and needed to know the maximum air forces the canopy would experience in transport whether facing forward or backward.
With my experience in running CFD simulations, they turned to me for answers. I used ANSYS Fluent to model and mesh the canopy and run it through a series of CFD simulations at various angles over the course of about five days. I was able to determine the maximum forces on the canopy at a range of canopy angles, to my colleagues’ satisfaction.
About this time ANSYS had given me access to their new Discovery Live product, and it occurred to me that this was an opportunity to do a quick comparison of this new product with ANSYS Fluent. Would Discovery Live’s ability to modify geometries during a simulation run and provide instantaneous results help me to learn more about the MTV canopy in a shorter amount of time, as advertised?
The canopy is very wide but very thin, and Discovery Live needs a fair bit of thickness in the geometry to work effectively, so that was a challenge we had to solve. Once we quickly thickened the canopy and ran the simulations on Discovery Live, I was surprised at how well the results matched up with the ones from Fluent.
But we decided to take the test one step further. Analyzing the canopy in isolation with both Fluent and Discovery Live had produced similar results, but in reality the canopy does not exist in isolation. We suspected there would be significant aerodynamic effects from the rest of the MTV and the flatbed truck as it moved on the highway at speeds up to 70 mph.
So, we ran the simulations of the complete system — MTV facing both in the forward and backward positions on a moving flatbed with the canopy at various angles in Discovery Live. In Discovery Live, we were able to import and modify the full geometry of the entire flatbed, MTV and canopy significantly faster than we could have done in other tools. When it came reversing the position of the MTV on the flatbed, in Discovery Live it was just a matter of spinning the MTV geometry around and recalculating. We completed our analysis of the full, moving system in about one and a half days full using Discovery Live, which includes the time it took to learn and become familiar with the application. We saw a significant time savings compared to traditional CFD methods
Airflow over complete flatbed–MTV system with MTV facing forward and backward using Discovery Live
The results showed that the MTV coupled with the flatbed truck geometry resulted in significantly different forces on the canopy than when we had analyzed the canopy in isolation. Looking at the complete system, including the shape of the flatbed and the MTV and where the canopy is located, we saw that other parts of the equipment were deflecting air up into the canopy, producing different forces. By including the rest of the geometry and the flatbed we were able to get a much better picture of what was happening to the canopy during transport. Our engineering colleagues at Roadtec were able to design a support system for the canopy to ensure it would not be damaged under the worst conditions during transport.
Discovery Live enabled us to perform this analysis quickly and accurately, with results that were close to the much more rigorous CFD analysis of ANSYS Fluent. Both Discovery Live and Fluent were valuable in this experiment, and each has its role to play in analyzing fluid dynamics. Fluent gives us access to a range of turbulence, multiphase, and reacting models which are invaluable in solving the many varied applications for CFD we have at Astec. Discovery Live is very complimentary to Fluent allowing for very quick turnarounds of problems where the fidelity and full capabilities of Fluent are not required. With Discovery Live used for early design exploration followed by Fluent for verification and further analysis I can see great benefit in having both these tools in my simulation toolkit.