# Parabola and normal lines

• Apr 22nd 2010, 05:03 AM
TsAmE
Parabola and normal lines
If c > 1 / 2, how many lines through the point (0, c) are normal lines to the parabola y = x ^ 2? What if c <= 1/2? Could someone please help?
• Apr 22nd 2010, 06:31 AM
tom@ballooncalculus
Let a be the x-value of where the normal intersects x^2. Express m, the gradient of the normal line, in terms of a. Then (at least if you do a sketch) you can see that c, the height of where the normal intersects the y-axis, will be equal to a^2 minus ma, which you should find is equal to a^2 + 1/2. (Which is at least 1/2)
• Apr 22nd 2010, 02:52 PM
TsAmE
I see but the answer was that there were 3 normal lines when c > 1/2 and 1 normal line when c <= 1/2.
• Apr 23rd 2010, 12:13 AM
tom@ballooncalculus
Quote:

Originally Posted by TsAmE
I see but the answer was that there were 3 normal lines when c > 1/2 and 1 normal line when c <= 1/2.

Yes, because the y-axis goes through any c (or rather, through any point (0,c).
• Apr 23rd 2010, 03:23 AM
TsAmE
How could it be only 3 normal lines for c > 1/2? Aren't there infinite normal lines?
• Apr 23rd 2010, 04:38 AM
tom@ballooncalculus
Quote:

Originally Posted by TsAmE
How could it be only 3 normal lines for c > 1/2? Aren't there infinite normal lines?

No, there are infinite points (0,c), but any particular point (0,c) where c > 1/2 lies on two non-vertical normals plus the vertical. Any point where c < or = 1/2 lies only on one normal, the vertical. Draw a sketch. Or a few (for different a).
• Apr 23rd 2010, 07:34 AM
TsAmE
Quote:

Originally Posted by tom@ballooncalculus
No, there are infinite points (0,c), but any particular point (0,c) where c > 1/2 lies on two non-vertical normals plus the vertical. Any point where c < or = 1/2 lies only on one normal, the vertical. Draw a sketch. Or a few (for different a).

But for each point dont you get many normals as they can have many gradients?

I have inserted my sketch
• Apr 23rd 2010, 10:48 AM
tom@ballooncalculus
You can have many lines (with different gradients) through a point (0,c) but only 3 of them will hit the curve at right-angles to it (or at right-angles to, rather, the curve's tangent at each of the points where they hit). Which is what it means for a line to be 'normal' to the curve.
• Apr 23rd 2010, 02:54 PM
TsAmE
But couldnt the curve have many gradients as you arent given the point thats tangent to the curve? meaning the normal wont only cut 1 tangent at 90 degrees but will also cut the other tangents of different slopes? (y=2x gradient seems as if it could cut any where on the parabola)
• Apr 23rd 2010, 03:52 PM
tom@ballooncalculus
Quote:

Originally Posted by TsAmE
But couldnt the curve have many gradients as you arent given the point thats tangent to the curve? meaning the normal wont only cut 1 tangent at 90 degrees but will also cut the other tangents of different slopes? (y=2x gradient seems as if it could cut any where on the parabola)

A line is only a normal if it cuts the curve at 90 degrees, so any cuts it makes with tangents other than at 90 degrees and on the curve are beside the point.

Good that you did the sketch, it will help you - but now make sure you only show lines that are truly (or at least roughly) normal to the curve.

Also, show a, and the geometry and/or algebra that enable you to determine c from it. E.g. the altitude of the point on the curve, a^2, and the height of a triangle with base a and hypotenuse with slope 1/(2a).