Real-time simulation and prognosis of smoke propagation in compartments using a GPU / Anne Küsters. Jülich : Forschungszentrum Jülich GmbH, Zentralbibliothek, Verlag, [2018], copyright 2018
Inhalt
- Leere Seite
- Introduction
- Methodology
- Physical Processes of Smoke Propagation
- Numerical Models of Fluid Dynamics
- Uncertainties and Errors in CFD
- Simulation of Smoke Propagation on CPU
- Code Structure Using Interfaces
- Extension to Complex Geometries in 3D
- Intermediate Code Testing via Unit Tests
- Verification and Validation
- Verification of the Implementation
- Validation of the Underlying Model
- Fire Induced Flow Experiment in a Compartment
- Open Plume Experiment Using Particle Image Velocimetry
- Summary
- Towards Real-Time and Prognosis Simulation Using a GPU
- Porting to GPU Using OpenACC
- Performance Portability Analysis
- Roofline Model
- Measurement Setup
- Theoretical and Measured Arithmetic Intensity
- Performance Portability Results
- Summary
- Analysis of the Real-time and Prognosis Software
- Speedup Analysis
- Runtime Performance of Used Cases
- Fire Induced Flow Experiment in a Compartment
- Open Plume Experiment Using Particle Image Velocimetry
- Real-Time and Prognosis Analysis
- Summary
- Closing Remarks
- Appendix Compiling, Developing and Running the Software
- Appendix Detailed Input Data of Test Cases
- Advection Test Case
- Artificial Diffusion Test Case
- Diffusion Test Case
- Diffusion (Hat) Test Case
- Pressure Test Case
- McDermott Test Case
- Vortex Test Case
- Lid-Driven Cavity Flow
- Flow Around a Cube
- Fire Induced Flow Experiment in a Compartment
- Open Plume Experiment Using Particle Image Velocimetry
- Tunnel Simulation for Real-time and Prognosis Analysis
- Nomenclature
- Bibliography
- Leere Seite
- Leere Seite