# Need solution for IVP

1. Jul 22, 2012

### BobMarly

Not sure of solution for IVP

1. The problem statement, all variables and given/known data
(x^2)*y"+4*x*y'+4*y=0
y(1)=1 y'(1)=2

2. Relevant equations
then (r^2)+3r+4=0
get (-3+/-7i)/2

3. The attempt at a solution
More y'=..... one large formula
Am I on the correct track, is there something I missed, or is there a shortcut here
it seems that by putting in IV's, it doesn't lead to anything clean

Last edited: Jul 22, 2012
2. Jul 22, 2012

### Staff: Mentor

Re: Not sure of solution for IVP

First off, does this function satisfy your differential equation? If so, you have the right general solution.

Next evaluate your solution and y'(t) at 1 and solve for the two constants. The check there is verifying that y(1) = 1 and y'(1) = 2.

At this stage in your learning, you should get in the habit of verifying that solutions you find are actually solutions. You've already done all the hard work, so checking your work is simply a matter of 1) showing that your solution satisfies the DE, and 2) that your solution and its derivative satisfies the initial conditions.

3. Jul 22, 2012

### BobMarly

How do I know it satisfies the equation?

4. Jul 22, 2012

### LCKurtz

Re: Not sure of solution for IVP

No it doesn't.

You are confusing the characteristic equation for constant coefficient DE's with this problem. The "solution" you propose is what you would get for those roots of r if this were a constant coefficient equation. But it isn't. This is an Euler equation and you get that indicial equation by looking for a solution $y = x^r$. Look in your text for the form the solution takes for complex conjugate roots. It will have x's times cosines and sines of ln x's.

5. Jul 23, 2012

### HallsofIvy

Staff Emeritus
In fact, the reason the "Cauchy-Euler" equation is as simple as the equation with constant coefficients (and easily confused with it) is that the substitution $t= e^x$ converts the Cauchy-Euler equation $Ax^2d^2y/dx^2+ Bx dy/dx+ Cy= 0$ into the 'constant coefficients' equation [itex]Ad^2y/dt^2+ (B- A)dy/dt+ Cy= 0.
The two have the same characteristic equation. In your case the constant-coefficients solution is
$$e^{-3t/2}(C_1cos(7t/2)+ C_2sin(7t/2))$$
so that the solution to the original CauchyEuler equation is
$$e^{-3(ln(x))/2}(C_1cos(7ln(x)/3)+ C_2 sin(7ln(x)/3))= x^{-3/2}(C_1cos(7ln(x)/2)+ C_2sin(7ln(x)/2))$$