- #1
Vuldoraq
- 272
- 1
Hi, Please could someone help me with the following? Sorry for the extra long post!
a) Using the Maximum/Minimum principle, find the maximum and minimum points of the solution of the following mixed problem;
[tex]u_t - c^2u_{xx} = 0, \ \ 0 < x < l, \ \ t>0[/tex]
[tex]u(0,x) = x(x-l) \ \ 0< x< l [/tex]
[tex]u(t,0) = 0 = u(t,l) \ \ t > o[/tex]
and
b) Let u, belonging to the set of continuously differentiable once and continuously differentiable twice functions, be the solution to the following mixed problem;
[tex]u_t - u_{xx} =0, \ \ 0 < x < 2, \ \ 0 < t < T[/tex]
[tex]u(0,x) = 0[/tex]
[tex]u(t,0) = f(t)[/tex]
[tex]u(t,2) = 0[/tex]
where [tex]0 \le f(t) \le 1[/tex]
Prove that,
[tex]0 \leq u(t,x) \leq 1 - \frac{x}{2}[/tex]
hint: Write the initial boundary value problem for [tex]w(t,x) = u(t,x) - v(t,x)[/tex] where [tex]v(t,x) = 1 - x/2[/tex] and use the maximum principle.
None?
I'm really stuck on these. For a start I don't really understand the Maximum principle, so any information would be most helpful. I think it means that if you can find a maximum on the boundary of a function, then you have found the maximum for the whole space. The same is true for the minimum.
So is the boundary defined by our boundary/initial conditions? And how would you then find the maximum/minimum on that boundary?
Here is what I have for part a)
The only boundary that gives us any information is the one where [tex]u(0,x) = x(x-l) \ \ 0< x< l [/tex]. The minimum of this boundary must be zero, since x lies between zero and l. The maximum can be found by differentiating once and setting equal to zero. This gives x=l/2. Plugging this back into the original boundary equation gives me a maximum of l/4. Therefore we have,
[tex]0 \le u(x,t) \le \frac{l}{4}[/tex]
Is this the right thing to do? It doesn't seem very rigorous to me, so I think that I have done something wrong.
For part b)
We get an equation for w,
[tex]w_t - w_{xx} = 0 [/tex]
[tex]w(0,x) = \frac{x}{2} - 1 [/tex]
[tex]w(t,0) = f(t) -1 [/tex]
[tex]w(t,2) = 0 [/tex]
I find the minimum on the boundary of w to be minus one, because the lowest of f(t)-1 is when f(t)=0. I also get the maximum of w to be zero, corresponding to when x=2 or f(t)=1. Therefore we have the following,
[tex]w(t,x) \ge -1 [/tex]
[tex]w(t,x) \le 0 [/tex]
since w = u-v we have,
[tex]u-v \ge -1 [/tex]
[tex]u-v \le 0 [/tex]
which implies,
[tex]u \ge v - 1 = -x/2 [/tex]
[tex]u \le v = 1-x/2 [/tex]
Therfore I get, [tex]-x/2 \le u(t,x) \le 1-x/2[/tex] which is clearly wrong!
Please could someone help me make sense of all this? I have spent hours and hours trying to work it out and I must be missing something obvious. Thanks for your time!
Homework Statement
a) Using the Maximum/Minimum principle, find the maximum and minimum points of the solution of the following mixed problem;
[tex]u_t - c^2u_{xx} = 0, \ \ 0 < x < l, \ \ t>0[/tex]
[tex]u(0,x) = x(x-l) \ \ 0< x< l [/tex]
[tex]u(t,0) = 0 = u(t,l) \ \ t > o[/tex]
and
b) Let u, belonging to the set of continuously differentiable once and continuously differentiable twice functions, be the solution to the following mixed problem;
[tex]u_t - u_{xx} =0, \ \ 0 < x < 2, \ \ 0 < t < T[/tex]
[tex]u(0,x) = 0[/tex]
[tex]u(t,0) = f(t)[/tex]
[tex]u(t,2) = 0[/tex]
where [tex]0 \le f(t) \le 1[/tex]
Prove that,
[tex]0 \leq u(t,x) \leq 1 - \frac{x}{2}[/tex]
hint: Write the initial boundary value problem for [tex]w(t,x) = u(t,x) - v(t,x)[/tex] where [tex]v(t,x) = 1 - x/2[/tex] and use the maximum principle.
Homework Equations
None?
The Attempt at a Solution
I'm really stuck on these. For a start I don't really understand the Maximum principle, so any information would be most helpful. I think it means that if you can find a maximum on the boundary of a function, then you have found the maximum for the whole space. The same is true for the minimum.
So is the boundary defined by our boundary/initial conditions? And how would you then find the maximum/minimum on that boundary?
Here is what I have for part a)
The only boundary that gives us any information is the one where [tex]u(0,x) = x(x-l) \ \ 0< x< l [/tex]. The minimum of this boundary must be zero, since x lies between zero and l. The maximum can be found by differentiating once and setting equal to zero. This gives x=l/2. Plugging this back into the original boundary equation gives me a maximum of l/4. Therefore we have,
[tex]0 \le u(x,t) \le \frac{l}{4}[/tex]
Is this the right thing to do? It doesn't seem very rigorous to me, so I think that I have done something wrong.
For part b)
We get an equation for w,
[tex]w_t - w_{xx} = 0 [/tex]
[tex]w(0,x) = \frac{x}{2} - 1 [/tex]
[tex]w(t,0) = f(t) -1 [/tex]
[tex]w(t,2) = 0 [/tex]
I find the minimum on the boundary of w to be minus one, because the lowest of f(t)-1 is when f(t)=0. I also get the maximum of w to be zero, corresponding to when x=2 or f(t)=1. Therefore we have the following,
[tex]w(t,x) \ge -1 [/tex]
[tex]w(t,x) \le 0 [/tex]
since w = u-v we have,
[tex]u-v \ge -1 [/tex]
[tex]u-v \le 0 [/tex]
which implies,
[tex]u \ge v - 1 = -x/2 [/tex]
[tex]u \le v = 1-x/2 [/tex]
Therfore I get, [tex]-x/2 \le u(t,x) \le 1-x/2[/tex] which is clearly wrong!
Please could someone help me make sense of all this? I have spent hours and hours trying to work it out and I must be missing something obvious. Thanks for your time!
Last edited: