On July 20, 1969 — 43 years ago today — Neil Armstrong made a nice, soft landing on the moon. It was way better than the time he almost died practicing for the event in the Lunar Landing Research Vehicle.

When President Kennedy promised America a moon landing by the end of the decade, NASA more or less knew how to get to the moon, but only in theory. The practical side was another matter, and one of the biggest unknowns was how to actually land on the surface.

What would a lunar lander look like and how would it fly? At the Flight Research Center at Edwards Air Force Base in California, Assistant director Hubert M. “Jake” Drake established a group of four engineers to look into the challenges of piloting a lunar landing. Armstrong, a civilian pilot at the time, was the only pilot-engineer in the group.

WATCH VIDEO: Enjoy our top 5 lunar moments; astronauts singing, dancing and falling over… on the moon!

ANALYSIS: NASA’s Apollo Landing Sites Will Be Protected

Nicknamed the Drake Group, the team quickly identified a vertical approach as better suited to a moon landing, something closer to a helicopter than an airplane. But the other piece of the puzzle was designing a vehicle that could mimic the flying characteristics of a vehicle landing on the moon. That was a challenge. The moon’s one-sixth and no atmosphere is impossible to replicate for an Earth-bound flight test program.

But they could mimic the weight and movement of a vertical landing on the moon. The solution was to mount a cockpit on a large engine that could rotate freely around its three axes. Properly calibrated, this engine could provide the necessary thrust to give the pilot the sensation of flying in one-sixth gravity. The free movement of the cockpit around three axes was as close to an atmosphere-free flight as they could get.

The Drake Group’s research influenced the design of the Lunar Landing Research Vehicle, the vehicle for NASA to study and analyze piloting techniques needed to fly and land the Apollo Lunar Module. The LLRV was nicknamed the flying bedstead — it looked more like a bedstead on an engine than any kind of aircraft.

ANALYSIS: Apollo 11′s Engines Discovered in the Atlantic

The training flight profile had an astronaut fly straight to the test altitude before switching the engine into lunar mode by adjusting its power to support five-sixths of the vehicle’s weight. Then in this lunar gravity simulation mode, the astronaut would make a soft landing using the same instruments he would have on board the lunar module.

In 1968, Armstrong and Pete Conrad became the first astronauts to start training in the LLRV. They were backup commanders to the first manned Apollo missions, putting them in line to command early lunar landing flights (and they did, Armstrong on 11 and Conrad on 12). By May, Conrad had made 13 flights and Armstrong 20. On May 6, Armstrong made his 21st flight in the vehicle. It was a disaster.

The test started out normally. Armstrong took off vertically off and rose to an altitude just under 500 feet where he switched the engine into lunar simulation and began the test landing. He descended as though on the moon, but about 230 feet above the desert floor the LLRV started pitching forward and picking up speed. Armstrong tried to counter this forward momentum with his attitude control thrusters but they didn’t respond. Instead, the vehicle rolled sharply to the right and started losing altitude.

Within seconds Armstrong was flying on his side and realized ejecting was his only recourse. A little less that 200 feet from the ground he separated from the LLRV. The vehicle crashed in a field, the fuel still on board starting a fire (as can be seen from the dramatic photograph, top). It was damaged beyond repair at an estimated cost of half a million dollars. Armstrong, meanwhile, landed by parachute a few seconds later. He walked away uninjured, though admitted later that the force of ejecting had caused him to bite his tongue.

HOWSTUFFWORKS: How the Apollo Spacecraft Worked

The post-accident investigation found that the cause of the crash wasn’t related to the design of the LLRV but a problem with the propellant. The helium in the propellant tanks had depleted earlier than normal leaving insufficient pressure to force the hydrogen peroxide fuel to the attitude control rockets. That’s why Armstrong’s attempts to right the vehicle were ineffective.

NASA built off what it learned from the LLRV and created the sophisticated descendant called the Lunar Lander Training Vehicle. On June 14, 1969, about a month before Apollo 11’s launch, Armstrong got back in the saddle and made his first flight in the LLTV since his crash.

Over a period of three days he made eight flights — 14 takeoffs and landings, six of which were done in lunar simulation mode. In total, he got 40 minutes 14 seconds of practice time in before heading to the moon.

He described flying the LLTV as “a great deal different than any other kind of aircraft that I’ve flown.” He also commented that he was pleased to have had the chance to fly it just before launching on Apollo 11. When he returned to Earth on July 24, 1969, Armstrong attributed his successful landing at the Sea of Tranquility to the near-perfect simulation afforded by the training vehicle.

Image: Neil Armstrong floats safely to the ground after the Lunar Landing Research Vehicle (LLRV) crashed and exploded on May 6, 1968. Credit: NASA