|Ph.D Student||Salib Isaac|
|Subject||Development of Fluid Mesoscopic Modeling Using Dissipative|
|Department||Department of Mechanical Engineering||Supervisor||Professor Emeritus Shimon Haber|
|Full Thesis text|
Dissipative Particle Dynamics (DPD) is a powerful mesh-less particulate simulation method. This method bears similarities to the Smoothed Particle Hydrodynamics (SPH) and the Molecular Dynamics (MD) simulation methods. In the DPD method, fluid is modeled as a set of point Dissipative Particle (DP). Each DP represents a lump of fluid molecules. The motion of the particles is controlled through soft conservative, dissipative and random forces. The DPD scheme is simulated for mesoscopic size scale. The DPD algorithm was proved to be efficient and useful to simulate complex fluids in complex domains, e.g. polymers, blood flow in veins and colloids.
In our work, we reveal the flaws of the current DPD methodology that can occur for high Reynolds simulations in bounded domains. We suggest two modifications for the method, one to retrieve an incompressible solution and the other to obtain a compressible flow. The former was done by redefining the radius of influence of the DP’s, previously thought to be an arbitrary value, and is set to equal the mean distance between adjacent DPD particles. The latter goal was achieved by means of replacing the conservative force with a pressure dependent force. The DPD pressure itself was derived using an energy balance for non-dissipative systems. The force added is proportional to the relative pressure difference between the DP’s pressure and the minimum background pressure.
Finally, we include a Boussinesq approximation like force term to the DPD model with energy conservation (DPDE) to simulate temperature buoyancy. This force compensates for the fact that DP’s are point particles of constant masses. Using this modification we retrieve convection vortex in a box in which gravity field and temperature gradients are applied orthogonally to each other.