Software

I’ve been a space enthusiast for as long as I can remember. I was born in the same year as Apollo 11. I made my family interrupt a holiday in France so I could listen to the launch of the first Space Shuttle mission. And just a couple of days ago, I watched the Shuttle land from its last mission. The Space Shuttle is my own personal space programme. I’m sorry it’s over and I’m anxious and excited to see what’s next.

But in 2002, I got closer to it than I ever expected – I went to Houston and flew NASA’s full-motion Shuttle simulator. Here is an article I wrote back then. If you want to fly the Shuttle today, I recommend X-Plane, F-SIM Space Shuttle or Orbiter.

Tired of X-Plane? Had enough of Flight Simulator 2002? Want the ultimate flight simulator? Well, how about NASA’s unique Space Shuttle simulator at the Johnson Space Centre? Used to train every Shuttle commander and pilot and to rehearse missions, the sim occasionally has some unallocated time and NASA uses this to give VIPs including Tom Hanks, Bill Clinton and Charles Schultz and not-so-distinguished members of the press a chance to experience flying the Shuttle. It took six months to set up the clearances, but one Friday, I got a call from the Public Affairs Office in Houston: ‘we have a free slot with your name on it. Can you be here at 8am on Wednesday?’ Since the thing is in use 16 hours a day, every day of the week I didn’t want to miss my chance and so I was on the plane to Texas two days later. This is my review of the world’s ultimate flight simulator.

Aided by jet lag, I pitched up at the security office to be badged at the unreasonable hour of 7am. Also waiting were two Texas rangers with their distinctive ten gallon hats and creased denim uniforms and a dozen fighter pilots in full Top Gun outfits. After a short wait, my escort arrived and took me to the simulator hall. I was shown the various control rooms where the instructors and technicians run the simulations – not just the Shuttle, but also the International Space Station sim. These are not vast mission control rooms you see on films. They look more like the offices of bond dealers with lots of screens and expensive swivel chairs. The simulator group employs 135 people in total but our flight only needed five operators. A full-up integrated mission simulation might have hundreds working on it in Houston, Moscow and Florida. Inside the hall itself, there are several simulators: a couple of procedural simulators, a fixed base simulator, which is used for flight training but which doesn’t physically move and my ride for the morning, the motion base simulator. This looks like the sort of thing an airline pilot might train on. There is a platform raised up on hydraulic rods, lots of pipes, cables and tubes.

I took my place in the pilot’s right hand seat and my instructor, Bill Todd, took the commander’s left hand seat. We put on headsets so we could talk to one another and the team of five controllers who were operating the sim for us. Bill issued the order over the radio and the whole platform shook into life. “Make sure you’re strapped in tight” said Bill with more sang froid than a guards officer on a cold day. “Why?” I asked. “Because the whole thing’s going to tilt now.” As he said it, I could feel the pressure on my back increasing and after a few seconds I realised I was lying flat. This was Bill’s little surprise: unlike airline simulators, this sim has a unique ability. It can tilt back ninety degrees to simulate the launch attitude. It felt very realistic. Lying on my back, it was hard to reach the buttons overhead and in front of me. All the controls are Tonka-sized so they can be used by gloved pilots but it is hard to reach them even with one gravity bearing down on me. With a pressure suit helmet on, the overhead toggles cannot be seen, so the pilots must know which switches to throw from memory. Luckily, a third crewman, sitting between and behind the two pilots, checks that they hit the right ones. Then the screens came alive and I could see the launch tower out to the left behind Bill’s seat, the sun up ahead and, by lifting my head a little, I could see a bit of coastline. ‘Welcome to Kennedy Space Centre’, I thought.

The first sortie was to be a launch and ‘return to landing site’ abort. Bill called this a “twofer” because the spectators get a launch and landing at the same time – ‘twofer the price of one’. In other words, we’d blast off from Kennedy Space Centre, reaching the edge of space over the mid-Atlantic, at which point the operators would simulate a main engine failure and we’d have no choice but to jettison the solid rocket boosters, external fuel tank and glide back to Kennedy.

Bill pressed started the launch sequence and everyone in the cabin did the ritual countdown. 3 … 2 …1 …. Lift off. The cabin vibrated and rocked back and forwards as the main engines and then the solid rocket boosters ignited then the launch tower slide smoothly. Shortly after takeoff the Shuttle rotates and flies inverted giving an extraordinary view of the entire Florida peninsula below and soon, the curvature of the Earth. There was a jolt as the solid rocket booster separated and then it was very smooth.

A few minutes into the climb, the pressure on one of the engines visibly dropped and Bill initiated the abort. This is an extremely complex ballet in the upper atmosphere that turns the Shuttle around and uses the remaining engines to slow it down and then free falling almost vertically for about 100,000 feet until the aerodynamic surfaces become effective and the thing can begin to glide back. The computers on the simulator (which, incidentally, run the same flight control software as the real Shuttle on the same hardware) control the whole thing. Luckily, my sole job on this first flight was to deploy the air data probes once we had re-entered the atmosphere (60,000 feet and 280kts) and then lower the landing gear and release the slow-down parachute, while Bill hand flew the last few thousand feet for a perfect touchdown.

The motion base really comes into its own on approach and landing. After Bill’s demonstration, the sim was reset to put us at the top of drop for the approach into Kennedy after an abort. “You have control,” said Bill. Earlier he had commented that “people who play video games tend to do better at this than real pilots.” Since I’d spent most of my adult life designing video games and I have a PPL, I figured I had the best of both worlds and took this as a challenge. The approach bears a surprising resemblance to a standard overhead join, which you might do at any small airfield in England, in that you fly west to east over the top of the north-facing landing runway then turn and descend on the ‘deadside’ before looping back over the runway, turning downwind and then onto finals. The differences are: you start at 50,000 feet (not 2,000), the crosswind leg is over 25 nautical miles long (not half a mile or so) and the rate of descent is DRAMATICALLY faster. Put another way, I don’t normally make sonic booms on my approaches into Denham or turn base over Cambridge.

Needless to say, the Shuttle is a glider and the trick is to land first time, every time. Consequently, the Shuttle flies a very steep glide slope – not the 3-4 degrees of a commercial airline. It lands at 195-200 knots, significantly faster than a jumbo jet, in part because of its fast approach and in part because of its delta wing. Concorde lands fast for the same reason.

For the approach, Bill explained that we needed a steep nose-down attitude only pulling up on finals. The ‘picture’ of the runway (the aspect ratio of the runway which shows whether you are too high or too low) is dramatically different from what I am used too. In fact, it is more like the view a dive bomber might get. Because of the delta wing, Bill explained “you can’t fly it to the ground.” Instead, once over the threshold, you pull back on the stick and flare the aircraft, pop the chute and let it settle on its own cushion of air. Once in the flare, it’s not too dissimilar from landing a normal plane: “look at the end of the runway and hold the nose relative to that.”

I hand flew two full approaches. On the first, I managed to get it onto the runway (just) but I wasn’t terribly straight and this prompted Bill’s comment, which he delivered with a laconic Texan accent: “steer onto the centreline for the cameras – it doesn’t matter where you touch down, it’s where you end up that impresses the journalists.” My second approach was done in cloud down to about four hundred feet and I got onto the runway and held the centreline until I came to a complete stop. Bill’s comment: “pretty good, but the real astronauts have to do it in a crosswind, in pressure suits, under stress and after two weeks of weightlessness in orbit.” He told us afterwards that on the first shuttle flight, just before launch, the commander, veteran John Young, had a heart rate of 80 while his pilot, first timer Bob Crippen, was up to 160. I was definitely in the Bob Crippen camp! However, the NASA minder who was escorting me had a go and landed in the swamp. According to Bill “that was as far from the runway as I have ever landed.” This made me feel better about my performance.

As a flying machine, the Shuttle feels very stable. The joystick controller drives a fly by wire control system which triggers reaction control rockets in space and aerodynamic surfaces in the atmosphere. The controller feels very solid, like it was built by Mercedes rather than Microsoft, and it resists hasty or aggressive manoeuvres. The whole vehicle has considerable inertia and the two combined make it fly very smoothly and precisely. I’ve flown 767s and 747s on British Airways simulators and the Shuttle feels similar to them, but perhaps a little sprightlier. The simulator is fitted with the latest in Shuttle avionics including new multifunction displays. The flight control software has to cope with a number of different modes: launch, abort, orbital, docking, re-entry, approach and landing and the screen displays match the mode. The approach display is broadly similar to an airliner display but the gauges are subtly different: angle of attack assumes a higher importance and the airspeed is displayed in KEAS, which stands for knot equivalent airspeed. In the upper atmosphere, there isn’t enough air pressure to get a meaningful airspeed and this is an analogue used to give a sense of the vehicle’s speed. Where you might expect something like a direction indicator or HSI, there is a strange display which shows three small circles moving inside a constantly changing geometric shape. The three circles are predictors of the vehicle’s position at different points in the future and the shape is the HAC or heading alignment cone. This is like a funnel down which the Shuttle must fly. Luckily, the head up display (HUD) simplifies everything by giving you an incredibly simple flight director display. There is a square which represents the actual heading of the Shuttle and a diamond which indicates the desired heading as calculated by the computers. If you keep the diamond inside the square on the HUD, you’re doing fine. I suspect this is why Bill thought that people who play video games were better than trained pilots at landing the sim. The head up display also shows rate of descent, speed, angle of attack and – on finals – a projection of the runway itself to give the pilot a ‘picture’ even in cloud. For instrument approaches, this is an easier presentation than a traditional ILS.

“It is the most sophisticated simulator in the world, by far,” Bill explained after the sortie. Besides simulating all the different ways a Shuttle flies, every subcomponent is modelled so that different kinds of failures can be thrown at trainees. Finally, the whole thing is part of a worldwide simulator network and it can link to mission control in Houston, launch control in Florida, Russian mission control and mission or ISS-specific simulators. It would be like a virtual airline that plugged into a virtual operations system, other aircraft and three different air traffic control simulators.

So how does it rate as a simulator? If I were reviewing it as a consumer product I would have to say that the visuals are not as good as Flight Simulator 2002, although they are reasonable for a system that was first built in 1980 and they compare well to commercial flight sims. The accuracy of the flight models and flight control software is, of course, as good as is humanly possible. The panels, displays, seats and controllers are flight-rated and again as accurate as possible. The motion base, the piles of checklists and manuals everywhere and radio chatter from the controllers through the headsets add to the sense of ‘being there’. A particularly human touch was the half dozen kitchen timers velcro’d to every empty surface and the pens and pencils that are attached at various points. These touches reminded me that this is a simulation of a real flying vehicle and these are the sorts of things that real pilots have in their aircraft. When you consider that a pilot in training for a mission would fly several four hour profiles a week prior to launch and that they are graded by computer on the accuracy of their landings and even the amount of tyre and brake wear they cause, it is easy to imagine that for most pilots the Shuttle simulator in Houston IS a real aircraft.