AEM faculty spotlight:
Bringing humans to Mars is one thing. Landing once we are there is something completely different. Vehicles are going to need to be able to launch parachutes higher and be able to deal with extreme speed and turbulence. Virtual simulation will be a key to success, according to AEM Professor Graham Candler. He is helping create the parachute for the upcoming Mars Science Laboratory mission. Beyond plotting air flow, Candler sees his method as holding the potential for simulating an entire flight control system in a realistic yet virtual environment, under dynamic conditions. This would allow testing of all mission aspects – from launch to flight control software to, finally, parachute deployment and landing.
In what follows, Professor Candler discusses the current simulation technology and what he hopes and expects the future to hold.
How is what you are doing different from standard simulations done in industry today?
We are the only group doing this sort of modeling in a complex, three-dimensional way. It has been done for subsonic aircraft at cruise conditions in industry on a regular basis. But that’s a relatively simple objective function they’re trying to maximize or minimize. We are at the point where we can optimize geometries in vehicle design at different flight stages, and we are getting to a point where we can do unsteady calculations – that is, we could take the Mars Science Laboratory (MSL) capsule and fly it down its trajectory, maybe coupling with the control system so we could test the system. We are aiming to do high fidelity calculations in a virtual flight environment.
How close are you to this sort of testing, and what are the implications of such tests?
It’s beyond what we can do, but it’s definitely coming. We will aim to do, for instance, a complete, virtual simulation of part of a re-entry, or to put a scramjet-powered vehicle through a maneuver. Whether it’s accelerating or trying to start the inlet, we’re getting to a point where we can do meaningful calculations to capture a lot of those physics that have been relatively unknowable mathematically up until this point.
Briefly, how do you simulate the air flow with complex shapes, like that of an aircraft?
Essentially, we are solving partial differential equations that describe mass, momentum, and energy conservation at some infinitesimal point in space. Fundamentally it’s point-wise description of flow field. We capture mass, momentum, and energy in a certain volume and use “grids” to wrap that information around a particular geometry. Our new code is an unstructured grid code, which allows more sophistication and flexibility in the way grid is designed. We can take time steps that are 1000 times larger than conventional explicit time integration methods. Factors of 1000 make a big difference.
What was it like when you realized you had improved the capability of the method by a factor of 1000?
It’s an incremental thing. It’s interesting because sometimes grad students do not have the perspective to realize what they’ve done in some small area heralds huge breakthroughs. In the last two to three years, we have made huge improvements in what we can do. The capability of my research group – if I compare this year to a year ago – it’s a factor of 10 in the complexity of problems we can do. I have wanted to get to this point for many years. It’s a long, slow process of learning and investment. But that’s the fun of research; you try to keep advancing state-of-the-art and have a vision of where you want to get to all while planning it out and making it happen.
What could come out of this new technology?
Some of these things could enable new designs. If you could do a high fidelity calculation and predict performance of a scramjet engine-powered aircraft and optimize that in software, you can take something that might be marginally feasible and make it feasible.
We’re also learning about flow physics – it helps us understand why, for example, under certain conditions a parachute wants to collapse. In terms of the latest application – MSL - ultimately, what we’re trying to do is understand physics of these flows and make the parachute work better and reduce likelihood that the mission fails because the parachute didn’t work properly.
What is the excitement in this for you?
To me, the excitement is taking a numerical model and writing the code. You sit down in front of the computer, build some code, and have direct impact on real missions. That’s kind of fun.
Last Modified: Tuesday, 09-Oct-2007 08:34:32 CDT -- this is in International Standard Date and Time Notation