Solve the first order hyperbolic equation

Click For Summary

Homework Help Overview

The discussion revolves around solving a first order hyperbolic equation given by 3 du/dx + 2x du/dt = 2u, with an initial condition of u(x,0) = 2x + 4. Participants explore various methods for approaching the problem, particularly focusing on the method of characteristics.

Discussion Character

  • Mixed

Approaches and Questions Raised

  • Participants discuss the method of characteristics, with one participant presenting their characteristic equations and initial conditions. Others express uncertainty about the correctness of their methods and seek clarification on specific steps, such as integrating and solving for constants.

Discussion Status

The discussion is active, with participants providing guidance and feedback on each other's approaches. There is a collaborative effort to clarify methods and resolve confusion regarding the integration and application of initial conditions. Multiple interpretations of the problem-solving process are being explored.

Contextual Notes

Participants are navigating through the complexities of the method of characteristics and the implications of initial conditions. There is an acknowledgment of potential confusion regarding the integration steps and the evaluation of constants.

andrey21
Messages
475
Reaction score
0
Solve the first order hyperbolic equation

3 du/dx + 2x du/dt =2u

With initial condition: u(x,0) = 2x+4



My attempt at a solution

I usually adopt the method of characteristics:

dx/a = dt/b = du/c


So from the above:

a=3, b=2x and c=2u

am I on the right track here?
 
Physics news on Phys.org
I do this in a slightly different way, my characteristic equations are:
[tex] \dot{x}=3,\quad\dot{t}=2x,\quad \dot{u}=2u[/tex]
Where the dot denotes differentiation to the characteristic variable, [itex]s[/itex]. I write the cauchy data as:
[tex] t(s=0)=0,\quad x(s=0)=r,\quad u(s=0)=2r+4[/tex]
Then just re-arrange to get rid of s and r and that will be your solution. I posted some notes on this in https://www.physicsforums.com/showthread.php?t=467445
 
Thank you hunt_mat. I shall read through ur notes. Is the method I was using incorrect for this question then?
 
Not an incorrect method, you can get the solution that way, I just think that the solution method that I presented is far easier to implement and you know if you're making a mistake or not.
 
So from what u have said, will the quotient of the characteristics be

dt/dx = 2x/3 du/dx= 2u/3
 
I think so, yes. To be perfectly honest, I was never really happy doing things this way.
 
Are u referring to the method I have adopted? Sorry for the confusion, I'm just anxious to answer this question.
 
So far, I think you're doing fine, but I haven't much experience with the method you're adopting. I have a book with it in though, so I will keep helping you with this method.
 
Ok hunt_mat I appreciate your help, now here's where I get stuck:

Integrating du/dx=2u/3?
 
  • #10
So you have computed the characteristics which are given as [itex]3t-x^{2}=k[/itex] where [itex]k[/itex] is a constant, you know that [itex]u[/itex] is a constant on these characteristics, and you know that characteristics curve passes through the point [itex](t,x)=(0,r)[/itex], what will [itex]u[/itex] be at this point?
 
  • #11
I'm a little confused how u got 3t-x2=k?

But will u(r,0) =2r+4
 
  • #12
Because I solved the equation:
[tex] \frac{dt}{dx}=\frac{2x}{3}[/tex]
You are right with the second part.
 
  • #13
Yes sorry its because I had it in the form:

t = x2 /3 +K

Ok so given u, where do I go next?

I still need to solve du/dx correct?
 
  • #14
Now you need to compute K, you know that the characteristic passes through the point [itex](t,x)=(0,r)[/itex], so you can use that to compute K. Then you need to integrate your equation for u and use that fact that at the point [itex](t,x)=(0,r)[/itex], [itex]u=2r+4[/itex], so all you have to do is substitute for r.
 
  • #15
So establishing K:

t = x2 /3 +K

0= r2 /3 +K

K=-r2 /3

Therefore:

t = x2 /3 - r2 /3

Correct??
 
  • #16
Good so far. Now do the same thing for [itex]u[/itex] and you should end up with a function with x's and r's in, and now you know an equation for r.
 
  • #17
So the next step is to do:

du/dx where u = 2r+4

so u = 2rx +4x

Is this correct??
 
  • #18
No, you have an equation for u of the form:
[tex] \frac{du}{dx}=\frac{2u}{3}[/tex]
Solve that. You will have a constant to compute, that is when you use the initial condition you were talking about.
 
  • #19
Ok this is the part I can't get past:

du/dx= 2u/3

I'm nt sure how to solve the above:
 
  • #20
You integrate it like any other equation:
[tex] \log u-\log u_{0}=\frac{x^{2}}{3}[/tex]
Now we know that at the point [itex](t,x)=(0,r)[/itex], the solution is [itex]u_{0}=2r+4[/itex], so insert this into the equation and you will be left with an equation with lots of x's and r's, well you know an equation for r...
 
  • #21
So substituting u0=2r+4 into:

logu-logu0 = x2/3

logu-log(2r+4)=x2/3
 
  • #22
Not quite, You need to evaluate everything at the point [itex](t,x)=(0,r)[/itex], I note that you have done nothing about te x...
 
  • #23
It would have been better to have written the solution as:
[tex] \log u=\frac{x^{2}}{3}+c[/tex]
Then to find c, just evaluate the equation at the point [itex](t,x)=(0,r)[/itex]
 
  • #24
Oh yes I see:

logu=r2/3 +c
Is this correct now?
 
  • #25
Yes, you also know what value of u at this point as well [itex]u=2r+4[/itex] and therefore
[tex] \log (2r+4)=\frac{r^{2}}{3}+c[/tex]So now you know what c is in terms of r, and you know what r is in terms of t and x, you can write down the solution.
 
  • #26
So

c= log (2r+4) - r2/3

So

log u = x2/3 + log (2r+4) - r2/3
 
  • #27
Yey, and now you know what r in terms of t and x are, substitute these in and you will have your solution.
 
  • #28
So:

t = x2 /3 - r2 /3

r2 /3 = x2 /3-t

r2 = x2 -3t

r = SQRT (x2 -3t)
 
  • #29
Good so far, remember there is an associated [itex]\pm[/itex] along with the answer. You have to decide which it is, but this shouldn't be too hard.
 
  • #30
So :

r = [itex]\pm[/itex]SQRT (x2 -3t)

What is the next step?
 

Similar threads

  • · Replies 18 ·
Replies
18
Views
3K
  • · Replies 12 ·
Replies
12
Views
3K
Replies
5
Views
2K
  • · Replies 8 ·
Replies
8
Views
2K
  • · Replies 5 ·
Replies
5
Views
2K
Replies
4
Views
2K
  • · Replies 105 ·
4
Replies
105
Views
11K
  • · Replies 10 ·
Replies
10
Views
2K
  • · Replies 16 ·
Replies
16
Views
2K
  • · Replies 12 ·
Replies
12
Views
3K