SUVAT is five equations relating five quantities: displacement s, initial velocity u, final velocity v, acceleration a and time t. Each equation leaves one of them out, so with any three known you can always reach the other two — usually by picking the equation that omits the quantity you neither have nor want. All of it rests on a single assumption: the acceleration does not change over the interval. Everything else follows from that.
A car brakes from 30 m/s to a stop in 45 m. You know s, u and v and want a, so use the equation without t: v² = u² + 2as gives a = (0 − 30²) ⁄ (2 × 45) = −10 m/s², almost exactly 1 g. Double the entry speed to 60 m/s and the distance becomes 180 m — four times as far, because the speed is squared.
Because there genuinely are two. Throw a ball upward at 20 m/s and ask when it is 15 m high: it passes that height at 0.99 s on the way up and again at 3.09 s on the way down, moving at the same speed in opposite directions. Solving for time uses a quadratic, and solving for a velocity takes a square root — both have two roots, and both roots are real motions. A calculator that silently returns one of them has picked for you without saying so.
That you finished on the negative side of where you started. s is displacement, not distance travelled. A ball thrown up that lands back in your hand has s = 0 even though it has covered several metres — and if it lands in a well below you, s is negative.
Whatever your chosen positive direction says. Pick one — up, or forwards — and every quantity takes its sign from that choice. If up is positive then gravity is a = −9.81 m/s², a ball thrown up starts with positive u, and braking gives a negative a while the car still moves forwards. Mixing conventions halfway through is the single most common way these problems go wrong.
Almost always air resistance, which SUVAT does not contain. A dropped feather, a long fall, a cyclist coasting — all of them have a force that grows with speed, so the acceleration is not constant and no SUVAT equation applies. The rule of thumb is that it holds well for dense objects over short falls, and starts to mislead beyond a few seconds of free fall. For a fall with real drag, see the Galileo drop simulator.
No changes to this tool’s own behaviour since the earliest archived release (v1.98). The full history for the site is in the changelog.