Gurnett & Bhattacharjee 6.
3
Xuxa Borealis
October 21, 2016
Assuming pressure balance,
B2
20
+ P = constant, show that the following function
1 0 1
x
B = B0 erf { (
) 2 y}
2 t
is a solution of the static ( U = 0) time dependent equation
B
1
=
2 B
t
0
where
2
erf (y) =
(1)
ev dv
So when I first started working this I went about it the wrong way: attempting to evaluate the erf first
and then taking the derivative. There was the splitest of seconds where I might actually complete a plasma
physics problem without help, but I was wrong.
The trick: Leibniz integral rule.
Z b(t)
Z b(t)
d
f
(
f (x, t)dx) =
dx + f (b(t), t) b0 (t) f (a(t), t) a0 (t)
dt a(t)
a(t) t
(2)
So by taking the derivative of the integral, things become so much easier. Solve the RHS and the LHS of
equ. 1 separately, and see if they are equal.
LHS =
B
t
1 0 1/2
B
= [B0 erf [ (
) y]]
t
t
2 t
Z 12 ( 0t )1/2 y
2
2B0
=
[
ev dv]
t 0
(3)
We are going to use the Leibniz rule on the time derivative of the integral:
t
Z
=
0 1/2
)
y)2
e( 2 ( t
t
1 0 1/2
y
2( t )
ev dv
1 0 1/2 2 ( 21 ( 0t )1/2 y)
(0)
dv + exp[( (
) y) ](
) exp[0](
)
2 t
t
t
Where the first and third terms go to zero because the the functions are not dependent on t, so the derivative
goes to zero.
1 0 1/2 2 ( 21 ( 0t )1/2 y)
= exp[( (
) y) ](
)
2 t
t
= exp[
1 0 2 ( 21 ( 0t )1/2 y)
(
)y ](
)
4
t
t
1
(4)
I am going to step through the derivative with respect to t because I needed to do it myself:
( 21 ( 0t )1/2 y)
y (0 )1/2 ((t)1/2 )
=
t
2
t
y (0 )1/2 ((t)1/2 )
1 y (0 )1/2
=(
)
((t)3/2 )
2
t
2
2
( 12 ( 0t )1/2 y)
y (0 )1/2
=
((t)3/2 )
t
4
( 12 ( 0t )1/2 y)
1 0 1/2
) y)
=(
(
t
4 t2
So now that I have computed that, equation 4 becomes:
= exp[
1 0 2 1 0 1/2
) y)
(
)y ](
(
4
t
4 t2
1 0 2 1 0 1/2
) y)
(
)y ](
(
4
t
4 t2
But REMEMBER plug back into equation 3 (I have made this mistake twice while working through this
problem, including while typing it up)
exp[
B
1 0 2 1 0 1/2
2B0
= exp[
(
)y ](
(
) y)
t
4
t
4 t2
RHS =
(5)
1
2
0 B
Before starting this side of the question, it is important to point out that 2 =
field, B in only dependent on y.
2
y 2 ,
because the magnetic
1
1 2
1 0 1/2
2 B =
[B0 erf [ (
) y]]
2
0
0 y
2 t
=
1 2B0
( [
0 y y
1 0 1/2
y
2( t )
ev dv])
(6)
We are going to take it one derivative at a time:
Z 21 ( 0t )1/2 y
2
[
ev dv]
y 0
Back to our friend Leibniz!!
Z
2
1 0 1/2
e( 2 ( t ) y)
1 0 1/2 2 ( 21 ( 0t )1/2 y)
(0)
=
dv + exp[( (
) y) ](
) exp[0](
)
y
2 t
y
y
Just like the LHS, the first and third terms go to zero, leaving:
1 0 1/2 2 ( 21 ( 0t )1/2 y)
= exp[( (
) y) ](
)
2 t
y
= exp[
1 0 2 ( 21 ( 0t )1/2 y)
(
)y ](
)
4
t
y
What is good about a derivative with respect to y, is that it is easier to take than the one with respect to t:
= exp[
1 0 2 1 0 1/2
(
)y ]( (
) )
4
t
2 t
Alrighty, now to take the second derivative with respect to y:
1 0 2 1 0 1/2
(exp[
(
)y ]( (
) ))
y
4
t
2 t
1 0 1/2
1 0 2
=( (
) ) (exp[
(
)y ])
2 t
y
4
t
1 0 1/2
1 0 2 1 0
=( (
) )exp[
(
)y ](
(
)y)
2 t
4
t
2
t
1 0 2
1 0 3/2
(
) y)exp[
(
)y ]
(7)
=(
4
t
4
t
Now that we have done that, we have to plug equation 7 back into equation 6 to finish salving for the RHS
of the equation.
1 2B0 1 0 3/2
1 0 2
(
=
(
) y)exp[
(
)y ]
0 4
t
4
t
Simplify:
=
1 B0 0 3/2
1 0 2
((
) y)exp[
(
)y ]
0 2
t
4
t
1 0 2
B0 0
(
)y ]
= ( ( 2 )1/2 y)exp[
4
t
2 t
Now, LHS = RHS
2B0
1 0 2 1 0 1/2
B0 0
1 0 2
exp[
(
)y ](
( 2 ) y) = ( ( 2 )1/2 y)exp[
(
)y ]
4
t
4 t
4
t
2 t
Do a little rearranging and such:
2B0 1 0 1/2
B0 0
(
(
) y) = ( ( 2 )1/2 y)
4 t2
2 t
B0 0
B0 0
( ( 2 )1/2 y) = ( ( 2 )1/2 y)
2 t
2 t
Totally equal!!
Now the remainder of the problem, I did not do, but it is plotting and it is helpful to have the following
equations:
1
1 Bx
J =
B =
z
(8)
0
0 y
and
P (y) =
1
(B 2 B 2 )
20 0
where Bx = B, the B from above, the first equation (unnumbered).
(9)