The basic technique for dealing with each of these problems is that of completing the square.
The first one, t2−9t−36.
Take a half of the coefficient of the linear term, (the 9t term), -9/2, square it, 36/4, add it to the expression and subtract it as well.
That gets you
t2−9t+364−36−364, which is still equal to the original,
and which can be written as
(t−92)2−45 .
The minimum value for that will be -45 occurring when t = 9/2, since that makes the expression inside the bracket zero. For any other value of t, the squared term will be greater than zero, making the whole expression greater than -45.
For the method to work, the coefficient of the squared term has to be 1. If the coefficient is not equal to 1, it has to be removed as a factor.
For example, if we had, (not one of yours),
−2x2−8x+17,
we would proceed as follows.
−2(x2+4x)+17=−2(x2+4x+4−4)+17
=−2{(x+2)2−4}+17 ,
=−2(x+2)2+8+17=25−2(x+2)2
Looking at that, its maximum value will be 25, occurring when x = -2, (since that makes the expression inside the bracket equal to zero).
The two variable example further down, x2+y2−4x+2y, can be dealt with in much the same way.
Write it as
x2−4x+y2+2y ,
and complete the square on x and y separately. It's then easy to see what the minimum value is, and the values of x and y that give that value.
Tiggsy.