NationStates Jolt Archive


Calculus (frowny face)

Opal Isle
05-10-2004, 22:06
I need help. I'm kind of stuck...

Here is the problem:

Prove that for any point (a,b) not equal to (0,0) on the parabola y=x^2, the normal line intersects the graph twice.


Here is what I have so far:

y = x^2, therefore y' = 2x
I also rewrote y = x^2 as b = a^2, and said that x = a

Because the derivative is 2x, the slope of the tangent line is 2x, therefore making the inverse (the normal line) -1/2x
You would find the actual slope by plugging the x value of the parabola at the point you wish to find the slope of into that mini-function.

Anyway, next, to avoid confusion in the next steps, I rewrote the slope as -1/2a, simply replacing the x with an a.
To find the equation for the line at point (a,b), you must use the following equation:
y-b = m(x-a)

We already know that m = -1/2a, and that b = a^2, and that y = x^2, so I rewrote that equation (after some algebra) as:
0 = -(x-a)/2a + (a^2) - (x^2)

This is where I am at right now. Am I on the right track and just need to keep doing algerba until I show that there are two values for a that can work in the equation? Or did I mess up somewhere and I am completely off track? Your help is appreciated, thanks.
Legless Pirates
05-10-2004, 22:11
I need help. I'm kind of stuck...

Here is the problem:

Prove that for any point (a,b) not equal to (0,0) on the parabola y=x^2, the normal line intersects the graph twice.


Here is what I have so far:

y = x^2, therefore y' = 2x
I also rewrote y = x^2 as b = a^2, and said that x = a

Because the derivative is 2x, the slope of the tangent line is 2x, therefore making the inverse (the normal line) -1/2x
You would find the actual slope by plugging the x value of the parabola at the point you wish to find the slope of into that mini-function.

Anyway, next, to avoid confusion in the next steps, I rewrote the slope as -1/2a, simply replacing the x with an a.
To find the equation for the line at point (a,b), you must use the following equation:
y-b = m(x-a)

We already know that m = -1/2a, and that b = a^2, and that y = x^2, so I rewrote that equation (after some algebra) as:
0 = -(x-a)/2a + (a^2) - (x^2)

This is where I am at right now. Am I on the right track and just need to keep doing algerba until I show that there are two values for a that can work in the equation? Or did I mess up somewhere and I am completely off track? Your help is appreciated, thanks.
since tou statet that x=a, the equasion is totally useless
0 = 0/2a + 0
Lunatic Goofballs
05-10-2004, 22:13
I need help. I'm kind of stuck...

Here is the problem:

Prove that for any point (a,b) not equal to (0,0) on the parabola y=x^2, the normal line intersects the graph twice.


Here is what I have so far:

y = x^2, therefore y' = 2x
I also rewrote y = x^2 as b = a^2, and said that x = a

Because the derivative is 2x, the slope of the tangent line is 2x, therefore making the inverse (the normal line) -1/2x
You would find the actual slope by plugging the x value of the parabola at the point you wish to find the slope of into that mini-function.

Anyway, next, to avoid confusion in the next steps, I rewrote the slope as -1/2a, simply replacing the x with an a.
To find the equation for the line at point (a,b), you must use the following equation:
y-b = m(x-a)

We already know that m = -1/2a, and that b = a^2, and that y = x^2, so I rewrote that equation (after some algebra) as:
0 = -(x-a)/2a + (a^2) - (x^2)

This is where I am at right now. Am I on the right track and just need to keep doing algerba until I show that there are two values for a that can work in the equation? Or did I mess up somewhere and I am completely off track? Your help is appreciated, thanks.

The mistake is here:

y=x^2 becomes y=+or- sqr(x) [plus or minus the square root of x.]

THe proof is right before you... + or -. For any value of X other than 0, there are TWO values for y. Therefore, two points of intersection. *nod*
Opal Isle
05-10-2004, 22:17
since tou statet that x=a, the equasion is totally useless
0 = 0/2a + 0
No. The a becomes a constant while x remains a variable. I don't think you've had any calculus or else you'd understand exactly what I'm doing.
Opal Isle
05-10-2004, 22:20
The mistake is here:

y=x^2 becomes y=+or- sqr(x) [plus or minus the square root of x.]

THe proof is right before you... + or -. For any value of X other than 0, there are TWO values for y. Therefore, two points of intersection. *nod*
Uh, that's not why there are two intersections. That's why a parabola looks like a parabola on both sides of the y-axis. There is a y value for a +x and the exact same value on the -x. It's why when y=4, x can be +2 or -2, but it doesn't prove that the normal line intersects the parabola twice.
Opal Isle
05-10-2004, 22:21
It'd be nice if someone who has had some college calculus could help me.
Erastide
05-10-2004, 22:23
Well... I took calculus but didn't really like it. But I think my one question is why you took the slope and substituted "a" in for "x"? Why doesn't it stay as (-x/2)?

If you don't substitute, you get a nice equation, 3x^2 -ax -2a^2. Hehe.. .manipulating math is fun!

I'm basically clueless here. :)
Legless Pirates
05-10-2004, 22:25
No. The a becomes a constant while x remains a variable. I don't think you've had any calculus or else you'd understand exactly what I'm doing.
Dude, I'm doing Math you've never even heard of.

Try proving that at a certain point
a*x > b*x
where a*x is the derivative (don't mention my term use, I'm dutch)
and b*x is the normal line
Lunatic Goofballs
05-10-2004, 22:25
Uh, that's not why there are two intersections. That's why a parabola looks like a parabola on both sides of the y-axis. There is a y value for a +x and the exact same value on the -x. It's why when y=4, x can be +2 or -2, but it doesn't prove that the normal line intersects the parabola twice.

I tackled it backwards. But you just explained it yourself. There are two values of x that yield the same value for y. WHich means that for any and all values of y, there are two different values of x. (2,4) and (-2,4) for instance.

The important detail in your equation is to note that the removal of the ^2(which sooner or later must come out) will create two separate equations. One with a +sqr, and one with a -sqr.
Opal Isle
05-10-2004, 22:34
Because the slope is -1/2x, but the x for the slope isn't the same x in the new equation...

For instance, the slope of the line tangent to the graph y = x^2 at x = 2...
Derive the equation:y' = 2x
x = 2, therefore y' = m = 4
The graph of the tangent line has a slope of 4, and to find the equation of that line, (y-4) = 4(x-2)

The y-4 represents the y point that is being solved for subtracted from y
The x-2 represents the x point that is being solved for subtracted from x
The y point and the x point in the above problem (the first one I placed) were replaced by (a,a^2) as opposed to (2,4) since we are solving theoritically, as a proof.
Anyway, the equation of that line comes out to be y = 4x+4. That's the tangent line.
The normal line (perpandicular to the tangent line) is y = (-1/4)x + (9/2). To find the other intersection, you replace y with x^2 (since you know y = x^2 and that there are two intersections). After a lot of algebra, you come up with (4x+9)(x-2) = 0, which gives you your two Xs, 2 and -9/4. You simple plug those numbers into y = x^2 to solve for the Ys and find the two points of intersection.

However, this is not the problem. The problem is solving this theoritically to serve as a proof, everything in this post is just an explanation so you guys are more clear on what the problem is actually asking for...
Opal Isle
05-10-2004, 22:52
Any one?
Legless Pirates
05-10-2004, 22:52
Dude, I'm doing Math you've never even heard of.

Try proving that at a certain point
a*x > b*x
where a*x is the derivative (don't mention my term use, I'm dutch)
and b*x is the normal line
Do you even read the posts?
Opal Isle
05-10-2004, 22:54
Do you even read the posts?
I thought you were trying to impress me with the type of math you are doing...because I didn't understand that post--and still don't.
Opal Isle
05-10-2004, 22:59
...so feel free to explain.
Legless Pirates
05-10-2004, 23:00
It's really hard for me to explain because I'm not into your terminology.

You calculated the normal line on a certain point on the parabola right?

Any straight line makes:
1) no touch with the parabola, no the case since the normal line always intersects the parabola
2) 1 touch, but the parabola makes a 90 degree angle with the parabola so it intersects
3) 2 intersections, because at some point the "angle" (when y=x^2 then y'=2x and 2 is the "angle", I don't know how to call it) of the parabola will be greater that the one from the normal line (because the parabola "angle" increases/decreases, because it's a parabola) so the normal line has to intersect twice or have an infinite "angle", which is not possible in calculus
Legless Pirates
05-10-2004, 23:07
is it of any help?
Opal Isle
05-10-2004, 23:10
It's really hard for me to explain because I'm not into your terminology.

You calculated the normal line on a certain point on the parabola right?

Any straight line makes:
1) no touch with the parabola, no the case since the normal line always intersects the parabola
2) 1 touch, but the parabola makes a 90 degree angle with the parabola so it intersects
3) 2 intersections, because at some point the "angle" (when y=x^2 then y'=2x and 2 is the "angle", I don't know how to call it) of the parabola will be greater that the one from the normal line (because the parabola "angle" increases/decreases, because it's a parabola) so the normal line has to intersect twice or have an infinite "angle", which is not possible in calculus

If I want to get credit, I need an analytical proof. Besides, a line can touch a parabola only once and not be tangent. Try y = x^2 and x = 2. The line x = 2 crosses the parabola at (2,4) and no other place, yet it is not tangent to the parabola.
Legless Pirates
05-10-2004, 23:17
If I want to get credit, I need an analytical proof. Besides, a line can touch a parabola only once and not be tangent. Try y = x^2 and x = 2. The line x = 2 crosses the parabola at (2,4) and no other place, yet it is not tangent to the parabola.
it's on a random point (a,b) on the parabola and the normal line is ALWAYS on a 90 degree angle with the parabola
Suicidal Librarians
06-10-2004, 00:36
Hey, Opal Isle my 8th grade daughter (Suicidal Librarians) printed out your problem for me because I teach Advanced Placement Calculus at my high school. You have actually done a nice job getting this problem set up.

Here's what I did. I used the slope of the normal (-1/2a) and the point on the curve (a, a^2) and the point-slope form to get the equation of the normal line below (note that I have solved it for y):

y = (-1/2a)x + 1/2 + a^2

Now, you want to find where this line intersects y = x^2 so set the equations equal to one another:

x^2 = (-1/2a)x + 1/2 + a^2 then move everything to one side to get

it equal to zero.

x^2 + (1/2a)x -1/2 -a^2 = 0

Next, I multiplied both sides by 2a to clear the fraction.

2a*x^2 + x - a - 2a^3 = 0

Now, use the quadratic formula. This can get confusing since
a was used in the original problem, so I am going to use capital letters
for the A, B, and C in the quadratic formula.

In this case, A = 2a, B = 1, and C = -a - 2a^3.

After substituting these values and doing some simplification you
will have x = (-1 plus or minus the sqr. root of (1 + 8a^2 + 16a^4))/(4a).

This will become x = (-1 plus or minus the sqr. root of ((1 + 4a^2)^2))/(4a).

This becomes x = (-1 plus or minus (1 + 4a^2))/(4a).

From here, the two solutions are x = a and x = (-1 - 2a^2)/(2a).

Therefore, two points of intersection exist and the x-coordinates of these points of intersection are listed above.

I hope you can follow this, it is tough to explain these things without being
able to use all of the proper math symbols.
Suicidal Librarians
06-10-2004, 02:56
Yeah, that really was my dad. ^

I'm not that smart.
HadesRulesMuch
06-10-2004, 03:01
The mistake is here:

y=x^2 becomes y=+or- sqr(x) [plus or minus the square root of x.]

THe proof is right before you... + or -. For any value of X other than 0, there are TWO values for y. Therefore, two points of intersection. *nod*
Yep, he is right.
That's what happens when you take precalc twice. You still don't understand how it works, you just remember the process.
Opal Isle
06-10-2004, 03:55
Hey, Opal Isle my 8th grade daughter (Suicidal Librarians) printed out your problem for me because I teach Advanced Placement Calculus at my high school. You have actually done a nice job getting this problem set up.

Here's what I did. I used the slope of the normal (-1/2a) and the point on the curve (a, a^2) and the point-slope form to get the equation of the normal line below (note that I have solved it for y):

y = (-1/2a)x + 1/2 + a^2

Now, you want to find where this line intersects y = x^2 so set the equations equal to one another:

x^2 = (-1/2a)x + 1/2 + a^2 then move everything to one side to get

it equal to zero.

x^2 + (1/2a)x -1/2 -a^2 = 0

Next, I multiplied both sides by 2a to clear the fraction.

2a*x^2 + x - a - 2a^3 = 0

Now, use the quadratic formula. This can get confusing since
a was used in the original problem, so I am going to use capital letters
for the A, B, and C in the quadratic formula.

In this case, A = 2a, B = 1, and C = -a - 2a^3.

After substituting these values and doing some simplification you
will have x = (-1 plus or minus the sqr. root of (1 + 8a^2 + 16a^4))/(4a).

This will become x = (-1 plus or minus the sqr. root of ((1 + 4a^2)^2))/(4a).

This becomes x = (-1 plus or minus (1 + 4a^2))/(4a).

From here, the two solutions are x = a and x = (-1 - 2a^2)/(2a).

Therefore, two points of intersection exist and the x-coordinates of these points of intersection are listed above.

I hope you can follow this, it is tough to explain these things without being
able to use all of the proper math symbols.

Tell your dad I said thanks. Heh, when I was working it I saw the a^3 and was kinda iffy about doing it because it's been like two years since I learned the quadratic thing...time to refresh..
Opal Isle
06-10-2004, 04:00
Actually...I don't understand this step:
This becomes x = (-1 plus or minus (1 + 4a^2))/(4a).

From here, the two solutions are x = a and x = (-1 - 2a^2)/(2a).

So if it's possible, I'd appreciate a further explanation of the last step by anyone who understands it. (I understand the x =a)
Csrjjsmp
06-10-2004, 04:03
Quadratic formula.
If ax^2+bx+c=0, then
x= (-b +/- sqrt (b^2-4ac))/2a
http://planetmath.org/encyclopedia/DerivationOfQuadraticFormula.html
Opal Isle
06-10-2004, 04:09
Quadratic formula.
If ax^2+bx+c=0, then
x= (-b +/- sqrt (b^2-4ac))/2a
http://planetmath.org/encyclopedia/DerivationOfQuadraticFormula.html
Sorry, I don't really consider that an explanation. It doesn't make sense to me. How do you get from
x = (-1 plus or minus (1 + 4a^2))/(4a)
to
x = (-1 - 2a^2)/(2a)

What happens to the plus or minus? Plus, I don't see how 2a^2 goes to 4a^2 or how 4a goes to 2a
Suicidal Librarians
06-10-2004, 04:17
The second solution comes from the "minus" part of the quadratic formula:
x = (-1 - (1 +4a^2))/(4a). If you subtract the stuff in parentheses it becomes (-2 - 4a^2)/(4a). Now just factor a 2 out of the numerator.
(2(-1 - 2a^2))/(4a). Then cancel it with a 2 in the denominator and you get the 2nd solution that I provided earlier:

x = (-1 - 2a^2)/(2a)
Opal Isle
06-10-2004, 04:18
Ah...I see. And had you done the plus part of it all, it'd all reduce down to a/1 or a. Okay. I'm still refreshing the quadratic formula stuff, but I think I'm getting it. Thanks a lot.
Suicidal Librarians
06-10-2004, 04:26
Glad I could help--sorry about the leap to the final solutions earlier.

I got a kick out of reading some of the early responses you were getting to your problem. Intuitively, it is obvious that a normal line will intersect twice (except the one through the origin), but that just doesn't cut it as a proof!!
Opal Isle
06-10-2004, 04:28
Glad I could help--sorry about the leap to the final solutions earlier.

I got a kick out of reading some of the early responses you were getting to your problem. Intuitively, it is obvious that a normal line will intersect twice (except the one through the origin), but that just doesn't cut it as a proof!!
Exactly...and with a D in my honors calc class, I don't think that just using intuition (especially on group projects) is going to cut it. Anyway...I get this stuff, I just don't get it all the way through. Mid term is a week from Thursday...

Oh well, my mom is about to complete her masters in Civil Engineering and said she always had Cs and Ds in her math classes, but hey, I have an A in my Honors Comp class.
Opal Isle
06-10-2004, 04:29
Actually, this stuff isn't really hard math when I think about it. The hard part is pooling all the math skills all together for one problem. I know all these skills, just can't remember them and can't intuitively select the specific formulae needed for a particular problem at the particular moment.