The attachment (once it's shown... I think it needs to be approved before it is accessible) shows a function, its derivative and its second derivative.
The bottomline:
At each x, the function has a "height".
Its derivative says if at each x the function is going up or down (for instance, if the function is going down, the derivative is negative).
The second derivative tells you if the first is going up or down. Maybe you can gain some intuition with the interest on a bank account: say you have $1000 saved. If you get $1 per week due to interests:
1. you have a positive slope.
2. such slope is constant. You get $1 more every week. No more, no less.
3. i.e., the "slope of the slope" is zero.
One way you could have a second derivative different from zero is if your bank were in trouble and reducing the interest every week, so that instead of $1 every week, you got 95 cents the first week, then 90, the 85, etc.
Then:
1. Your still have savings (f(t)>0),
2. They are still increasing every week (f'(t)>0), but
3. The weekly increase is getting smaller as time passes (f''(t)<0).
Of course, it would also be possible that the weekly increase (f'(t)) was getting bigger (not only you are getting more money every week, but the addition is every time bigger than the previous one). A single value of a function does not imply that its derivatives will be positive or negative.
Now, if you think about it, a positive second derivative means that the function is increasing more rapidly as time (say) passes. In a graph, this means that the graph is "curving up", not just increasing.
(I'll coment on the attachment once it becomes visible).