I have a kinematics problem, which I can happily do using the various non-quadratic kinematics equations. But when I do it with a quadratic equation, $s = vt - \frac{1}{2}at^2$, I get two possible solutions for $t$. The issue is they are both positive and there is only one physically possible solution.
The problem is this. A particle is travelling from A to B along a straight horizontal road with constant acceleration $0.34 \text{ms}^{-2}$. At B its velocity is $20\text{ms}^{-1}$ and the distance between A and B is $400\text{m}$. Find the time taken for the car to travel from A to B. The original question gives an intermediate velocity, so this can be used to avoid the quadratic equation. However, I am very bugged by why it is the case that I get the right answer 25.5 and a wrong answer, 92.1, which can't be discerned as incorrect, from the quadratic version.
I get that if it was a negative acceleration, a deceleration, that there could be a point when the particle would return and the two physically possible values would make sense. But the acceleration is in the direction of travel and the particle will move on forever. Shouldn't the incorrect solution be negative? Or is the use of the quadratic kinematic equations not advised in certain circumstances? What is the physical meaning of the larger, incorrect answer?
My workings are as follows.
$$s=vt - \frac{1}{2}at^2$$
$$400 = 20t - \frac{1}{2}(0.34)t^2$$
$$0.17t^2 - 20t +400 = 0$$
$$t = \frac{-(-20)\pm \sqrt{20^2 - 4(0.17)(400)}}{0.34}$$
$$t = 25.5 \quad \text{and} \quad t = 92.1$$

Hint: Your impression that "there is only one physically possible solution" is wrong. Try working out the initial velocity $u$ for both values of $t$ and compare their signs.