After just a few hands-on tutorial sessions, engineering seniors and graduate students at West Virginia University (WVU) were able to generate a variety of grids using Pointwise, learning how mesh topology, dimensionality, and resolution affect solution accuracy for a variety of internal and external flow problems. WVU participated in the Pointwise Teaching Partnership program in which qualified institutions receive free licenses for use in their classes. Graduate Assistant Chris Menchini said the introductory course provided most students with their initial exposure to computational fluid dynamics (CFD) software and none of them had used Pointwise previously. He said that students with past experience with Gambit and ICEM CFD agreed Pointwise was “by comparison generally much more user-friendly and more capable.” Menchini said because of its ease of use, students were able to learn cradle-to-grave how to construct a robust CFD model representative of a “real problem” in just a couple of weeks.
In the class, students learned various techniques to demonstrate meshing versatility including traditional structured and unstructured mesh generation, as well as unstructured viscous meshing using T-Rex. Using what they learned, the students were challenged to develop a CFD model to predict the drag force on a 1:4 scale fairing, shown in Figure 1.
The fairing was a preliminary concept originally developed and tested by students in a human-powered vehicle design course. The model was chosen for its readily available drag data in addition to its simple geometric contours to generate high-quality 3D mesh topologies. With the availability of the experimental data, students could compare their computational results with those recorded in the WVU closed-loop subsonic wind tunnel.
The project required students to generate the wind tunnel model and test section CAD geometry using SolidWorks 2012, import the geometry into Pointwise V17 to develop a 3D hybrid mesh, then set-up and solve the quasi-steady flow using ANSYS Fluent V14 with conditions extrapolated from measurements recorded during testing. All the while, students had to remain aware of cell count limitations imposed by the academic version of ANSYS Fluent.
With gridding guidelines and limitations in mind, students chose to use a hybrid strategy in order to manage the total mesh cell count efficiently. The mesh was generated using a dense 3D structured block extruded from the fairing surface to capture the near-field viscous/wake region. The inviscid far-field regime was handled using an unstructured block rapidly expanding in cell size away from the fairing body to economize the total number of cells.
The flow environment fell within the transitionally turbulent regime. Therefore, part of the solution process involved utilizing both viscous laminar and turbulent RANS models with wall functions and then comparing results. This allowed students to investigate the solution sensitivity to aerodynamic force calculations with respect to near wall mesh spacing as well as the meshing technique employed for transition to the far-field. For extra credit, students also could investigate how ground effect affected the lift and drag force on the fairing body. A steady laminar flow solution over a preliminary fairing design can be seen in Figure 2.
You too can use the same software championed by leading engineering companies around the world to prepare your undergraduate and graduate students for their professional careers.