Before the first shuttle orbiter flew in space, the spaceplane’s ability to fly as a glider to a precision touchdown first had to be proven. Prototype orbiter Enterprise would be carried aloft atop its Boeing 747 shuttle carrier aircraft (SCA), first in a series of captive test flights, and then for free-flying approach and landing tests (ALT). These Enterprise test flights were flown by two test pilot teams, Fred Haise and Gordo Fullerton, alternating with Joe Engle and Dick Truly.
During research for my book, Space Shuttle Stories, I interviewed Fred Haise to learn how he and Gordo approached the test series and put Enterprise through its paces. The ALT series proved to be a much-needed success for the formative shuttle program, and proved the orbiter could return to Earth safely after its hypersonic descent from orbit.
From our Feb. 12, 2020 interview:
What was most important about the shuttle approach and landing tests?
After I finished a fourth crew assignment during Apollo – I was the Apollo 16 backup commander – I joined Aaron Cohen in the orbiter project office for four years. I was on the orbiter project management side during its design and development. Which was great, because I worked with the ops team through all the design reviews, through Critical Design Review on Enterprise, and I was very happy to get assigned as one of the two crews to go fly it. So, it was a womb-to-tomb experience. I participated in the proposal evaluations to pick who built the shuttle…from the four proposals that came in to now getting to fly it for the first time. It was an exciting experience.
You have to realize that the approach and landing tests were obviously not a space mission–more like an aircraft flight test program. It took me back to where I was with NASA before, back at Edwards in flight test.
The key things that we were to achieve were verifying the subsonic aerodynamics and the handling qualities with the primary flight controls that were highly tuned through the development to be a pilot rating of 1 or 2. We worried that if over-gained, we could end up with rate limiting. And the first time we were going to see how it flew was when we cut loose from the 747 in that flight mode: how the flight control works, and the vehicle aerodynamics. Another unknown was ground effect. We were pretty confident in the up and away aero characteristics, both longitudinal and directional. We had a strategy that would give us margin with time to flare as we neared the ground. But we could run into the balloon case which would boost you into the air while you’re running out of airspeed. The other extreme possibility, like on the Concorde, we called vacuum sweep, where it could suck you into the ground, and maybe make you have a hard landing. So, ground effect was an unknown to worry about.
Probably the biggest challenge at the time was to get the four-computer set to play together. We almost gave up on it. We were down to the last software load from IBM, where the computers would all stay together. In ground test they kept voting each other out. The multiple computers and data bus architecture were a new thing, different from previous NASA spacecraft.
With that argument, I sold Aaron Cohen and Chris Kraft on having a backup computer where the whole flight control scheme was greatly simplified—just a rate command feedback system. And completely developed by a different contractor, not developed by IBM, but Draper Labs, who had done the system for Apollo. The system was dumbed down to set the handling qualities at values of 4 or 5, vs. 1 or 2 for the primary system [on the Cooper-Harper rating scale]. It was our “get it on the ground” system. By pushing one button on the top of the hand controller, you kill the input/output to those four computers, and switch to that fifth computer. That was what we were keyed to do should Enterprise go out of control as we separated from the 747. So, a goal and large technical challenge going into ALT was to get the four computers to work together.
In a program sense, I thought ALT served to “maintain program momentum.” We had a gap from when we’d flown the Apollo-Soyuz Mission, a multi-year gap where we wanted to keep NASA in the limelight … and be able to claim, “we’re flying something.”
ALT offered media exposure showing that we were making progress. On the other hand, we had to acknowledge the fact that the orbital program was slipping, about a two-year slip in the orbital launch date. ALT served a good purpose in showing progress and maintaining program momentum.
We had one simulator, a hydraulic moving-base one, and an airborne simulator, which was probably the best, the Gulfstream II. When we got into the program, each of us separated according to what we were trying to follow. For example, Crippen and Truly worried about IBM software, probably the toughest job for any of us. I worked the vehicle tests at Palmdale, and Joe Engle readied our simulators, first the moving base, which didn’t work exactly as advertised at first, and he flew a lot of the Gulfstream II development with Grumman test pilots. We had to refine our test articles before we could start our realistic training. The G2 ended up being the most accurate simulation of flying the real orbiter. Gordo took on the tedious, tough job of procedure development, both normal and contingency.
End of Part 1. Part 2 begins here: