Residual Distribution Methods
Accurate modeling and simulation of fluid transport in multi-dimensions require a numerical method that can mimic multi-dimensional flow physics. Most of the current commercial and research codes in CFD are based on Finite Volume (FV) methods which are known to be based on one dimensional physics and are very sensitive to grid changes. The Residual Distribution (RD) methods are known to be less sensitive to grid irregularities compared to Finite Volume (FV) methods and can be designed to incorporate multi-dimensional upwinding. In this research, a new signals distribution is developed for the RD method. The first is an isotropic signals distribution which would ensure conservation of primary variables and entropy-conservation. The second is an artificial signals distribution which would offset the isotropic signals to ensure entropy-stability and multi-dimensional upwinding. This new RD approach is compact, preserves second order accuracy on unsteady and advection-diffusion problems and can reach fourth order accurate for steady-state cases using any consistent time-integration method.
1. Hossain Chizari and Farzad Ismail. Accuracy Variations in Residual Distribution and Finite Volume Methods on Triangular Grids. Bulletin of Malaysian Mathematical Sciences Society, DOI: 10.1007/s40840-015-0292-0 (IF=0.80, Q1)
2. Farzad Ismail and Hossain Chizari. Developments of Entropy-Stable Residual Distribution Methods for Conservation Laws I: Scalar Problems. Journal of Computational Physics (IN PRESS)http://dx.doi.org/10.1016/j.jcp.2016.10.065(IF=2.56, Q1)
Vorticity Capturing
Vortical flows are a common occurrence in many aerodynamics problems. However, prediction of vortical flows remains a challenge in computational fluid dynamics (CFD). Most numerical methods in CFD will produce predicted solutions which are excessively dissipated or attenuated compared to the true vortical solutions, and in some cases will artificially create unphysical vortices. This research focuses on the development of numerical methods that can faithfully ‘capture’ vortical flows.
1. Farzad Ismail and Philip L. Roe, 2005, Toward a Vorticity Preserving Second Order Finite Volume Scheme, 17th American Institute of Aeronautics & Astronautics (AIAA) CFD Conference, Toronto, Canada.
2. Farzad Ismail, A Finite Volume Method That Controls Vorticity Based on the Limited Rotated Richtmyer Scheme, International Proceedings of AUN/SEED-Net Aerospace and Mechanical Engineering 2012, Ho Chi Minh, Vietnam.
3. Farzad Ismail, On Vorticity Prediction Using Limited Rotated Richtmyer Scheme Solving the Linearized Euler Equations, Bulletin of Malaysian Mathematical Sciences Society, 2013, Vol 36(1), pp. 7-22 (IF=0.80, Q
Entropy Consistent Flux Function
In CFD, there are lots of physical attributes being lost due the discretization process. Numerical methods are specifically designed to include certain physics of the fluid flow. For example, discretely enforcing conservation of mass, momentum and energy are the pinnacle of successful finite volume methods in computing compressible flow problems. Overall, this part of method development in CFD is a mature field but there are missing physics which currently are neither included nor properly enforced in the state-of-art numerical methods. These include entropy control, multi-dimensional flows, combustion, etc. Without compromising the discrete conservation laws, the work herein is develop a method that discretely enforces the Second Law of Thermodynamics (entropy), known as the entropy-consistent flux function both for the Euler and Navier Stokes equations.
1. Farzad Ismail and Philip L. Roe, Affordable Entropy-Consistent Euler Flux Functions II: Entropy Production Across Shocks, Journal of Computational Physics (IF=2.37, Q1), 2009, 228, pp.5410-5436
2. Akmal Nizam Mohammed and Farzad Ismail, 2011, Entropy-Consistent Flux Function in CFD: Integrating Physical Viscous Effects, International Conference on Mechanical and Manufacturing Engineering 2011 (ICME2011) Putrajaya, KL.
3. Akmal Nizam Mohammed and Farzad Ismail, Study of an Entropy Consistent Navier Stokes Flux, International Journal of Computational Fluid Dynamics (IF=0.8), 2013, Vol 28 , pp. 1-14
Shockwave Anomalies
The carbuncle phenomenon usually occurs when a blunt body is subjected to a strong shockwave. This can happen for instance, when a supersonic or hypersonic flow passes a cylinder. The physical solution would have a bow shock ahead of the body but the predicted carbuncle solution would produce a pair of weaker oblique shocks in front of the geometry. As a result, the prediction of pressure and velocities around the geometric are severely compromised since the post shock conditions are incorrect. This carbuncle problem was first reported in 1988 but until now there is no satisfactory explanation of the phenomenon, let alone a satisfactory cure to it. There are other shock anomalies such as the slowly moving shock problem, the Noh’s problem (heat transfer), etc.
1. Philip L. Roe, Hiroaki Nishikawa, Farzad Ismail and Leonardo C. Scalabrin, 2005, On Carbuncle and Other Excrescences, 17th American Institute of Aeronautics & Astronautics (AIAA) CFD Conference, Toronto, Canada.
2. Farzad Ismail, Philip L. Roe and Hiroaki Nishikawa, A Proposed Cure to the Carbuncle Phenomenon, Computational Fluid Dynamics 2006, Springer, 2009, 149-154, ISBN 978-3-540-92778-5
3. Keichi Kitamura, Philip L. Roe, Farzad Ismail, An Evaluation of Euler Fluxes for Hypersonic Flow Computation, American Institute of Aeronautics & Astronautics (AIAA) Journal (IF=1.00, Q1), 2009, 47, 1, 44-53
CFD Code Development and HPC
This is an in-house code development project for solving compressible fluid dynamics. The code is based on a finite volume method and is still in preliminary stages. Currently, it is a 2D code but 3D code development will start in the near future. The group is currently looking for experts in computational turbulence and fluid-structure interaction, overlapping and moving grids to continue the development of the code.
1. Nur Khairunnisa Hanisah Roslan and Farzad Ismail. Evaluation of the entropy-consistent Euler flux on 1D and 2D test probloems, (to appear in American Insititute of Physics Journal).
2. Farzad Ismail, Pablo M. Carrica, Tao Xing, Frederick Stern, Evaluation of Linear and Nonlinear Convection Schemes on Multidimensional Non-Orthogonal Grids with Applications to KVLCC2 Tanker, International Journal for Numerical Methods in Fluids (IF=0.97), 2010, 64, , 550-586.
3. Frederick Stern, Farzad Ismail, Tao Xing and Pablo M.Carrica, 2008, Vortical and Turbulent Structures Using Various Convection Schemes with Algebraic Reynolds Stress-DES Model for the KVLCC2 at Large Drift Angles , 27th Symposioum on Naval Hydrodynamics, Seoul, S. Korea.
4. Pablo M. Carrica, Farzad Ismail, M. Hyman. S. Bhushan, F. Stern, Turn and Zigzag Manuevers of a Surface Combatant using a URANS Approach with Dynamic Overset Grids, Journal of Marine Science & Technology (IF=0,7 Q1), 2013, 18(2), pp. 166-181.
Modelling and Simulation of Fluid Transport on Ocean and River Flows