Research Interests
This page summarizes my main research interests. The navigation bar on the left has links to the published papers on which I am an author. For all of the papers I am the first author and hold the copyright, there is a version of the PDF posted directly on the website. For the other papers, a link to the abstract based on the DOI is provided.
Because I don't know where else to put it, here is a link to my completed Ph.D. thesis.
Hypersonic Propulsion Systems and Flight Dynamics
The components of air-breathing hypersonic vehicles are notoriously tightly coupled. In traditional aircraft the systems of an aircraft can be treated mostly separately. You want to speed up? Just burn more fuel. Design the vehicle to get more lift? All you have to do is make the wing bigger.
In a vehicle like the one below, the engine is almost the size of the vehicle itself. Not only is the engine responsible for providing thrust, it also has to provide most of the lift. Further, the nonlinear nature of supersonic flow means that the performance can be highly sensitive to small changes in speed or angle of attack. Suppose you want this vehicle to speed up. Well, you first you burn more fuel, thinking this might be the end of the story. But then you notice that burning more fuel raises the exhaust pressure, which due to the shape of the nozzle increases the lift of the vehicle. So you have to go to a lower angle of attack to keep lift in line with weight. But now the lift on the nozzle is higher, and the lift on the inlet is lower, so the vehicle wants to pitch down dramatically, and you need to use your elevators to prevent this from happening.
Generic hypersonic vehicle with stylized shocks and flame.
This wouldn't be so bad if it weren't for the high sensitivities involved in the propulsion system. The problem is that lowering the angle of attack might cause the shock waves to move around in the inlet so that performance is dramatically reduced. Or the conditions might change so much that the engine changes its mode of operation or completely stops working altogether.
This is the kind of flight dynamics that is the focus of my research. We look for topics where previous experience does not provide much guidance and flight experiments are too costly. Another aspect of this research is provide modeling for conceptual design optimization and control science. These are all topics that require a model that can simulate the whole vehicle from tip to tail in a few seconds, and our research code called MASIV (Michigan/AFRL Scramjet in Vehicle) does exactly that.
Ram-Mode Operation and Ram-Scram Transition
The type of engine that we study is called a dual-mode ramjet-scramjet. They have this funny property that it can operate in one of two modes. In one mode (scram mode, for supersonic combustion ram), the flow stays supersonic throughout the entire engine.
Sketch of isolator and combustor in scram mode. Flow is supersonic
throughout.
Ram mode. Flow is subsonic from the shock train through the end of the
constant-area section.
There are a couple of confusing or inconvenient things about this. The first is that ram mode is rather complex (due to interactions between the shock train and the combustor) and subject to unstart (which is not a euphemism for "stop"). Secondly, the engine is known (or at least predicted) to rapidly switch between the two modes. Since operating in both modes is one of the things that make these engines so potentially useful, the transition problem is a meaningful complication.
When transitioning from ram mode to scram mode, we typically see a decrease in net acceleration of about 1.5 m/s². But the pilot and control system don't directly determine which mode the engine is in; that comes from the flight condition and throttle setting. Think of it like a car that's accelerating and the automatic transmission suddenly changes the gear the car is in.
That's a somewhat suggestive analogy because obviously cars deal with that kind of transition all the time without any problems. But in flight there's more things to worry about (six degrees of freedom instead of two); let's hope that we can find a way to deal with ram-scram transition in a way that's only as disruptive as a car switching gears.
Atmospheric Entry of Spacecraft with High Ballistic Coefficients
Do you remember when Skylab came crashing back to Earth earlier than what was planned? Sometimes I use the word “remember” in a strange way, because it reentered Earth's atmosphere 6 years before I was born. But I remember hearing (at some point after I was born) that a decent amount of debris made it all the way to the ground without burning up.
Massive heat shields are required for most spacecraft that enter an atmosphere, but the lighter the craft (per unit area), the less shielding is required. The shielding requirements can be lessened even further if the vehicle provides some lift during the entry phase. Lift allows the vehicle to remain at higher altitudes and thus lower densities longer. And since heat transfer is proportional to density, this decreases the heat loading on the vehicle.
With enough contributions from both of these effects, it's possible to have a spacecraft that requires no heat shielding. Here's a design I created based on a 3U CubeSat form factor. This type of design couldn't make it on Earth or Venus (unless it was much less than 1 kg in mass), but it was on the border when it came to surviving entry into Mars' atmosphere. For Saturn's moon Titan, entry with no heat shield seems quite feasible.
Modified 3U Cubesat with deployed surfaces for lift.
