How did the space shuttle land?

The shuttle orbiter had no usable engines for landing: its three main engines only worked with the external tank during ascent, and the OMS orbital engines were useless in the atmosphere. From the deorbit burn to the runway, the vehicle was a heavy glider with a double-delta wing and a very poor glide ratio. That's why the landing was planned with precision: there was only one chance, with no option to go around and try again.

About an hour before landing, on the far side of the planet, the orbiter turned tail-first and fired the OMS engines for about three minutes to slow down by just 300 km/h — just enough for the orbit to graze the atmosphere. It then turned again, nose-first, at an angle of attack of about 40 degrees, so that the belly tiles and the reinforced carbon leading edge absorbed the reentry heat, close to 1,650 °C. At hypersonic speeds, the flight computer executed a series of S-turns — wide banking maneuvers that lengthened the trajectory and shed energy without overheating the structure.

In the final approach phase, beginning at about 25 km altitude, the commander took manual control and lined the vehicle up with an imaginary heading alignment cylinder. The descent path, about 20 degrees, was roughly seven times steeper than that of a commercial airliner. At about 600 m altitude the flare began, the landing gear deployed in the final seconds, and the main wheels touched down at about 350 km/h (190 kn). A drag chute, added in 1992, and the wheel brakes stopped the orbiter within 3 km of runway, at Kennedy or Edwards.