# Find energy of electrostatic field

1. Aug 5, 2010

### Petar Mali

1. The problem statement, all variables and given/known data
The space between the two concentric spheres is charged by spatial density of charge $$\rho=\frac{\alpha}{r^2}$$. The radius of spheres are $$R_1,R_2$$. Integral charge is $$Q$$. Find energy of electrostatic field.

2. Relevant equations
Gauss law

$$\oint_S\vec{E} \cdot d{\vec{S}}=\frac{q}{\epsilon_0}$$

$$W_E=\frac{1}{2}\epsilon_0\int_VE^24\pi r^2dr$$

3. The attempt at a solution

Using Gauss law I get

$$E^{(1)}=0$$, for $$r<R_1$$

$$E^{(2)}(r)=\frac{1}{\epsilon_0}\cdot \frac{Q}{4\pi(R_2-R_1)}\frac{r-R_1}{r^2}$$
for $$R_1\leq r \leq R_2$$

$$E^{(3)}(r)=\frac{Q}{4\pi\epsilon_0r^2}$$, for $$r>R_2$$

And get $$W_E^{(1)}=0$$

$$W_E^{(2)}=\frac{Q^2}{8\pi\epsilon_0(R_2-R_1)}(1+\frac{2R_1}{R_2-R_1}ln\frac{R_2}{R_1}+\frac{R_1}{R_2})$$

$$W_E^{(3)}=\frac{Q^2}{8\pi\epsilon_0R_2}$$

This is my solution.

Final solution from book is

$$W_E^{(1)}=W_E^{(3)}=0$$

$$W_E^{(2)}=\frac{Q^2}{4\pi\epsilon_0(R_2-R_1)}(1+\frac{2R_1}{R_2-R_1}ln\frac{R_2}{R_1})$$

Where I make a mistake?
1. The problem statement, all variables and given/known data

2. Relevant equations

3. The attempt at a solution

2. Aug 5, 2010

### ApexOfDE

Re: electrodynamics

I think you have no mistake with your calculation and there is perhaps typo in the book.

3. Aug 6, 2010

### Petar Mali

Re: electrodynamics

Is there some other idea, other way, to solve this and check the result? Maybe with potentials?

4. Aug 6, 2010

### ApexOfDE

Re: electrodynamics

I dont understand. When we calculate the electrostatic energy, we must integrate in all space (from zero to infinity) and the formula depends only in electric field and doesnt mention the material. In region (3), there is a field here. Therefore, I think W_3 must not be zero.

5. Aug 6, 2010

### nrqed

Re: electrodynamics

Sorry, my mistake. I had misread the question. They do ask for the energy of the electrostatic field, and you are right that it is then not zero in region 3. I removed my post.

6. Aug 6, 2010

### nrqed

Re: electrodynamics

Al your steps look right (except for your sign of the ln term). So the book is incorrect, indeed, as ApexofDE pointed out

7. Aug 6, 2010

### stevenb

Re: electrodynamics

In case the consensus is not enough, I'll add that I'm in agreement that the book is wrong, and the OPs approach is correct, and the OP's math is either correct or has only a slight error. I did not check the energy in the central region, but all other calculations look correct.

8. Aug 6, 2010

### Petar Mali

Re: electrodynamics

Yes, this is mistake in writing :(

$$W_E^{(2)}=\frac{Q^2}{8\pi\epsilon_0(R_2-R_1)}(1+\frac{2R_1}{R_2-R_1}ln\frac{R_1}{R_2}+\frac{R_1}{R_2})$$

Now is OK! Do you know how to solve this using vector and scalar potential $$A(\vec{r},t)$$, $$\varphi(\vec{r},t)$$?

9. Aug 6, 2010

### stevenb

Re: electrodynamics

This is an electrostatics problem, so you can ignore the vector potential and just use the scalar potential. This problem is basically solvable in terms of voltage potential using Poisson's equation. Actually, Poisson's equation applies in the region of charge, while Laplace's equation applies both inside and outside the charged region. Boundary conditions at zero, infinity, R1 and R2 must be met.

http://hyperphysics.phy-astr.gsu.edu/hbase/electric/laplace.html

I haven't tried to work out this particular problem, but my gut feeling is that the way you have already done it is relatively easy and probably easier than using potentials. It would be good practice to verify your answer however.

Last edited: Aug 6, 2010
10. Aug 6, 2010

### Petar Mali

Re: electrodynamics

Ok! I get that! :) Thanks!