Numerical modeling of microscale plasma actuators

We present the study of the dielectric barrier discharge (DBD) plasma actuator for both macro and microscale applications. Such actuators create a stable glow discharge at atmospheric pressures and generate cold plasmas and electrohydrodynamic (EHD) force that impart directed momentum to the surrounding fluid. There are phenomenological and physics based reduced order numerical models available for predicting these forces at macroscale. In microscale the physical model is not known. This research covers problems of two distinct spatial scales.At macroscale, we apply plasmas for mitigating heat transfer problem in gas turbines. Specifically, novel film cooling concepts of turbine blades are investigated using plasma discharge. A phenomenological approach is utilized for modeling the local body force generated by plasmas. An active three-dimensional plasma actuation is predicted for different cooling hole geometries. Such an approach utilizes the EHD force which attaches the cold jet to the work surface by actively altering the body force in the vicinity of an actuator. Results show above 100% improvement of film cooling effectiveness over the standard baseline design.Also at macroscale, we study the physics of plasma induced bulk flow control using first-principles based reduced order force model. We introduce two novel designs of horseshoe and serpentine actuators, and both designs have zero net mass flux (ZNMF). These actuators show active modification of the boundary layer thickness suitable for flow separation control and flow turbulization using the same actuator.We study the physics of microscale plasma actuation using the high-fidelity finite-element procedure which is anchored in a Multi-scale Ionized Gas (MIG) flow code. First, a two-dimensional volume discharge with nitrogen as working gas is investigated using a first-principles approach solving coupled system of hydrodynamic plasma equations and Poisson equation for ion density, electron density, and electric field distribution. The quasi-neutral plasma ( Ni &ap Ne) region and the sheath (Ni >> Ne) region are identified. As one approaches the sheath edge, there is an abrupt drop in the charge difference and sharp increase in electric field strength. By decreasing the gap between electrodes, the sheath becomes dominant in the plasma region. Based on the simulation results, we have deeper insight into the microscale force generation mechanism through understanding the physics at microscale. Subsequently, we investigate a novel first generation micro plasma pump using the same microscale hydrodynamic plasma model. We find the average flow rate is around 28.5 ml/min for micro plasma pump. Such micro plasma pumps may become useful in a wide range of applications from microbiology to space exploration and cooling of microelectronic devices.In order to improve the performance of micro plasma pumps in real world applications, a three-dimension plasma simulation is needed. We introduce a flow shaping mechanism using surface compliant microscale gas discharge. For the case of quiescent flow, horseshoe plasma actuator creates an inward and downward electric force to pinch and eject fluid normal to the plane of the actuator.We extend our two-dimensional hydrodynamic model into the two-species three-dimensional DC plasma simulation to study two cases of micro plasma pumps. Both plasma governing equations and Navier-Stokes equations are solved using a three-dimensional finite element based MIG flow code. The results show the highest charge separation and electric force close to the powered electrodes. We find three vortical structures inside the pump which can not be found in our two-dimensional simulation. To reduce the vortices inside the micro plasma pump, the location of the actuators and the input voltage may be key factors. The three-dimensional flow simulation at 5 Torr predicts an order of magnitude lower flow rate than that predicted earlier of two-dimensional micro plasma pump simulation for atmospheric condition. The predicted flow rate in Case...