Analysis and Simulations
RANS simulation of a ramjet
One of the more difficult and unpredictable areas of simulation is combustion. This project simulated an entire ramjet from its inlet to the nozzle, including the combustor. The results from the simulation were compared with results from an ideal ramjet. The simulation yielded approximately 90% performance compared to ideal, and also allowed insight into other phenomena not represented in the ideal equations, like inlet unstart and propellant mixing.
|
Large Eddy Simulation of a ramjet combustor
A further step in evaluating combustion performance was to go into more detail. Large Eddy Simulation is a method of directly simulating the intricacies of turbulence on a much smaller, and transient scale. The simulation ran for 10 milliseconds past a RANS-computed initial state. The theoretical output of this burner (the same as the one in the simulation above) is 24 Megawatts. The simulation showed an output between 23 and 24.5 Megawatts through the 15 milliseconds of the simulation. The variation comes from the design of the flameholder, which encourages vortex shedding to further enhance the fuel-air mixing. This corresponded to a variable power output as the mass transfer rate in the combustion varies slightly over time.
|
Damage projections for a 152mm artillery shell
It is usually taken for granted that armored vehicles are all but immune to the dangers of indirect fire. Upon reading about the results of the Soviet Artillery Effects study (pp 8-11) in 1988, I set up some explicit dynamics simulations to get a better understanding of what's going on. A number of simulations were set up using a 152mm OF-25 HE-FRAG shell, typically fired from the 2S3 Akatsiya self-propelled artillery platform, which featured heavily in Soviet mechanized formations. The first simulations involved a direct hit on a simulated armor plate made of Rolled Homogeneous Armor (RHA).
To the left is an animation of a case with the shell impacting 40mm of RHA. The blast and some of the resultant shrapnel tear into the plate, deforming it significantly in only two milliseconds! The plate suffered heavy damage and would fail in the following milliseconds. |
|
|
While the previous case reflects an artillery shell impact to the top or sides of a tank, that's not the only location that can be hit. To the right is a similar case, except the armor plate is now 150mm thick, to simulate the frontal aspect of a tank. While the peak stress is similar to the previous case, it is much more localized and exists on a smaller timescale. This is also reflected in the lower damage recorded to the plate. Unlike the 40mm case, the tank survives this with minor damage to the armor, discounting anything the radial shrapnel may hit.
|