Optimization

Optimization is a subset of control system design where the control is frozen in time. Moshman Research recently collaborated on shape optimization of an airfoil for the Air Force shown in the figure below. Other aspects of a flow can be optimized and Moshman Research uses a combination of commercial and research simulation tools and custom software developed in-house to be able to perform flow optimization in a broader set of contexts than is commercially available.

An example of shape optimization on a NACA64A010 airfoil (blue) pitching +/- 1 degree at 17 Hz with a flow speed of Mach 0.796. The optimal airfoil shape which minimizes drag while keeping lift above a set level is shown in red.

The initial conditions of a system also fit into the optimization framework since there is no change to the “control” over time. Examples include the initial distribution of water droplets used to attenuate a passing blast wave, the initial fraction of disruption-mitigating pellet mass in a magnetic-confinement fusion reactor or the design for the thermal protection layer in a hypersonic re-entry vehicle. Moshman Research has optimization experience in all of these applications. Below is a figure representing the density of a plasma in a fusion reactor which was calculated using the NIMROD code.

Cross-section contours of the plasma density in a toroidal fusion reactor. Adding a high-Z material to the burning plasma is a technique being studied for avoiding reactor disruptions.

Determining what is the optimal initial condition for a system is mathematically very similar to determining the initial condition a system must have that would lead to observed measurements. Examples of problems Moshman Research has addressed include determining the direction, energy and location of a laser from measurements of scattered light in the atmosphere and identifying the nature of an acoustic source such as a rotorcraft.