Clemson University Computational Center for Mobility Systems (CU-CCMS) is designed to offer unique simulation capabilities that are predictive in many diverse fields. Focusing on the automotive, aerospace and other transportation industries, we provide unique solutions for engineering research, design, development and optimization.Three unique features of CU-CCMS are our massive computing infrastructure, our unique and validated methods, and our industry-focused approach.
Massive Computing Infrastructure
Thanks to a partnership with Sun Microsystems, CU-CCMS has a world-class computing infrastructure, capable of solving the most complex large-scale problems in short order. The system includes:
Compute Grid
- 430 Sun Blades with a total of 3,440 processing cores
- Ultra-high-speed, high-capacity Voltaire Infiniband Network
- 14TB RAM
- 35 Tflops total computing power
Two Sun Fire E6900 servers
- Each with 48 processing cores
- 384 GB RAM
Sun E25k Server
- 144 processing cores
- 680 GB RAM
Unique and Validated Methods
CU-CCMS builds new and unique methods on the cutting edge of computational science to solve difficult problems. Some examples are:
Laminar-to-turbulent boundary layer transition model
- Captures transition naturally from the underlying physics, rather than having a user-specified transition location
- Accurately calculates drag, down-force or lift, aerodynamics flows and high-speed jets in quiescent or low-speed environments
- Fully implemented in an efficient RANS environment
- Validated in a wide range of flow conditions
Type-II unsteady flow model
- Resolves unsteady flow fluctuations arising from shear-layer roll-up, which cannot be predicted by standard off-the-shelf transient CFD
-
Examples of such flows include unsteady ‘flapping’ flow on an airfoil or vehicle hood leading edge, and vortex oscillation at the front edge of a film-cooling jet.
Conjugate heat transfer model with curvature capability
- Solves external flow, internal flow and conduction through solids simultaneously, without being separated
- Captures the effects of convex and concave surfaces on turbulence in an economical eddy viscosity model
Semi-deterministic stress model (SDSM)
- Models the effects of inherent unsteadiness on the mean flow field, using a steady simulation
- Captures unsteady effects without expensive and time-consuming unsteady simulations
- Analogous to modeling effects of turbulence with a RANS approach without resolving turbulence
- Allows a significant reduction in CPU usage (1-3 orders of magnitude) and data storage for cases where inherent unsteadiness is critical
Industry-Focused Solutions
CU-CCMS is dedicated to solving the most complex industry problems on an industry-compatible timeline. Our employees are full-time, experienced computational engineers who don’t have academic obligations. By building the methods for "virtual experiments," we enable our clients to have a faster time to market, better concept selection and a more competitive bottom line. CU-CCMS delivers real results — on time and within budget.