A couple of years after Mars Pathfinder thrilled the world by delivering a small rover to the surface of Mars, the Mars Global Surveyor satellite was orbiting the planet and mapping the entire surface. Scientists had detailed information about the local vicinity of the Pathfinder landing site, but wanted a more mobile system to gather detailed geological information, perhaps from the walls of the Valles Marineris, a "valley" that is 10 times larger than the Grand Canyon. A small aircraft was proposed, and the NASA administrator liked the idea of flying in the Martian atmosphere 100 years after the Wright brothers flew at Kitty Hawk in 1903.

Early in 1999, groups at several NASA centers began studying the feasibility of a small Mars airplane. An opportunity was available to fly a "piggyback" mission on a French rocket, but there was only a small mass allowance and very limited time. Some people, including Andy Hahn and Steve Smith, worked on the airplane itself. I worked on the "entry shell": the container that would endure the harsh environment as the vehicle slowed from a speed of 6 kilometers per second to about 0.6 kilometers per second in only a few minutes.

NASA has significant experience designing entry vehicles. The Space Technology division at Ames has designed the heatshield for several recent missions. Most of them look very similar to the Pathfinder: the front is circular and almost flat, almost like a saucer or cereal bowl. The flatness helps to slow the vehicle very quickly without making it too hot, and the round shape is strong and relatively easy to build.

Unfortunately, the space on the cruise stage (the spacecraft that would travel between Earth and Mars) was not circular, but rather elongated. Airplanes are not circular, but closer to a cross shape (with a long fuselage and wide wings). It is easier to pack a folded airplane into a rectangular box than a circular container. We thought we could make a bigger container if we changed the shape, and we wanted to see how much it would help the airplane designers.

I quickly modeled the important parts of the cruise stage using Computer-Aided Design (CAD) software. Then we tried squeezing in entry vehicles of various shapes. We chose a set of about 12 parameters to describe the vehicle. The width, length and depth of the box were obviously important. Sharp corners can cause local heating problems, so we needed to control the roundness of the edges and corners. The flatness of the forebody is also very important. We also found that tilting the entry vehicle in the cruise stage helped to increase the volume available for the airplane, so we tried several packing schemes. After a couple of days, we sent three alternative shapes to the airplane designers and asked them to decide whether the non-circular entry vehicle would help them.

Figure 1. The blue shapes represent the important features of the interplanetary spacecraft. The red shape is the entry vehicle packed to fit in between the fuel tanks.

We didn't have to wait long for their response: they excitedly reported that the packing convenience was a great benefit. Now we had to figure out if the new shape would be stable (the forebody with the thick heatshield would keep facing forward throughout the atmospheric entry) and could be strong and light without too much heating. People were chiefly concerned about the stability, so we decided to fly some models in the ballistic range to demonstrate that they would not tumble.

A model in the ballistic range is shot out of a big gun, and the launch loads are hundreds of times greater than the force of gravity (see Chuck Cornelison's journal). The model must be very small (to fit inside a gun barrel that is 1.75 inches in diameter) and very strong. The mass must be distributed so that the center of gravity is in a similar position to the full-scale vehicle.

One of the big advantages of designing with a computer is that the same model can be used for several tasks. To convert the full-scale design for the entry body to a ballistic range design, we needed only to change the scale and the material properties. We chose high-strength plastic for the model and inserted a metal disc in the middle. Careful placement of the metal, which is much heavier than the plastic, allowed us to locate the center of gravity correctly.

You may have read that errors converting units can be a big problem for engineers. That almost happened in this project. I had designed with a metric system because we needed to pack the vehicle on a French spacecraft, but the machinists here are used to working with inches. Luckily they knew that the gun in the ballistic range was only 1.75 inches in diameter, so they called to resolve the confusion. Engineers should always listen to questions from the people who actually build the hardware: they have much more practical experience than most of us. They aren't always right, but they are generally pretty sensible.

It took a few weeks for the models to be built. Sometimes people who do a lot of computer calculations forget how long it can take to buy materials, and schedule workers, and cut the hardware, and measure it to ensure that it's accurate. While all that was happening, engineers in the Aerothermal Environments group ran computational fluid dynamics (CFD) simulations so we could compare predicted forces with those that would be measured in the ballistic range. Again we could use the same geometry (from CAD) so we would be sure to make the comparison for exactly the same shape.

Finally the first model was loaded into the ballistic range. There isn't much to see in the control room, because you are protected from the range by walls of thick concrete. When they fire the gun there is a dull thud, and everyone immediately looks at a bank of instruments to check that all the timers went off in sequence. After weeks of preparation the test is over in less than a second. Immediate calculations confirm that the drag is close to predictions, and when the photographs arrive the next day we see that the model was stable, as we had hoped.

Figure 2. This image from the ballistic range shows that the entry vehicle is flying nice and straight. The curved line that wraps around the front of the body is a shockwave.

The Mars Airplane couldn't be developed in time for a 2003 mission, so the project has been cancelled for now. The information we gathered in this activity may still be useful, because it can be applied for other possible missions. Maybe when students of 2000 are working at NASA, they'll send payload to Mars in a vehicle like this.