Heat flow in semi inifinite wall: Dirichlet problem or separate variables?

libelec
Messages
173
Reaction score
0

Homework Statement



The plane region defined by \[A = \left\{ {(x,y)/0 < x < 1,0 < y < \infty } \right\}\] defines the profile of a semi infinite wall, where there's a stationary regime heat flow. The left and right sides are isolated at 15º, and the base is isolated at 0º. Find T(x,y) in each point of the wall.

Homework Equations



Two-dimensional Heat equation: \[\frac{{{\partial ^2}T}}{{\partial {x^2}}} + \frac{{{\partial ^2}T}}{{\partial {y^2}}} = \frac{{\partial T}}{{\partial t}}\]

Image

ico65l.jpg


The Attempt at a Solution



This problem got me thinking, because it's done either one of two ways: either I treat it as a problem of separate variables and use Fourier Series, or I treat it as a differential geometry problem (Dirichlet problem, since in stationary regime the heat equation becomes Laplace's equation). I tried thinking of it as a Dirichlet problem.

If I treat it as a differential geometry problem, I have to use conform transformations. First I have to multiply the whole thing by \[\pi \], so that then the wall has its sides at x = 0 and x = \[\pi \] (the base remains at y = 0, and it remains semi infinite). Then I have to traslate it \[ - \frac{\pi }{2}\], so that now the sides are at x = \[ - \frac{\pi }{2}\] and x = \[\frac{\pi }{2}\].

So far, I transformed \[x + iy\] into \[\left( {\frac{\pi }{2}x - \frac{\pi }{2}} \right) + i\frac{\pi }{2}y\]. Now, I've only seen this kind of problem in which there are 3 contour conditions being solved by applying the transform sin(z). If I do that, then I transform the strip into Im(z) > 0, now I have transformed \[x + iy\] into \[\sin \left( {\frac{\pi }{2}x - \frac{\pi }{2}} \right)\cosh \left( {\frac{\pi }{2}y} \right) + i\cos \left( {\frac{\pi }{2}x - \frac{\pi }{2}} \right)\sinh \left( {\frac{\pi }{2}y} \right)\].

The image is then transformed into:

242y9mp.jpg


Now, using polar coordinates, T doesn't depend of the radius, \rho \, so \[{\nabla ^2}T = \frac{1}{\rho }\frac{{{\partial ^2}T}}{{\partial {\varphi ^2}}}\]. This means that \[T(x,y) = A\theta + B\].

But I can't go pass that. I don't know how to find A and B using the countour conditions.

Any ideas?
 
Physics news on Phys.org
It compiled awfully! If you don't understand what it says, I repost it:

Homework Statement



The plane region defined by \[A = \left\{ {(x,y)/0 < x < 1,0 < y < \infty } \right\}\] defines the profile of a semi infinite wall, where there's a stationary regime heat flow. The left and right sides are isolated at 15º, and the base is isolated at 0º. Find T(x,y) in each point of the wall.

Homework Equations



Two-dimensional Heat equation: \[\frac{{{\partial ^2}T}}{{\partial {x^2}}} + \frac{{{\partial ^2}T}}{{\partial {y^2}}} = \frac{{\partial T}}{{\partial t}}\]

Image

ico65l.jpg


The Attempt at a Solution



This problem got me thinking, because it's done either one of two ways: either I treat it as a problem of separate variables and use Fourier Series, or I treat it as a differential geometry problem (Dirichlet problem, since in stationary regime the heat equation becomes Laplace's equation). I tried thinking of it as a Dirichlet problem.

If I treat it as a differential geometry problem, I have to use conform transformations. First I have to multiply the whole thing by \[\pi \], so that then the wall has its sides at x = 0 and x = \[\pi \] (the base remains at y = 0, and it remains semi infinite). Then I have to traslate it \[ - \frac{\pi }{2}\], so that now the sides are at x = \[ - \frac{\pi }{2}\] and x = \[\frac{\pi }{2}\].

So far, I transformed \[x + iy\] into \[\left( {\frac{\pi }{2}x - \frac{\pi }{2}} \right) + i\frac{\pi }{2}y\]. Now, I've only seen this kind of problem in which there are 3 contour conditions being solved by applying the transform sin(z). If I do that, then I transform the strip into Im(z) > 0, now I have transformed \[x + iy\] into \[\sin \left( {\frac{\pi }{2}x - \frac{\pi }{2}} \right)\cosh \left( {\frac{\pi }{2}y} \right) + i\cos \left( {\frac{\pi }{2}x - \frac{\pi }{2}} \right)\sinh \left( {\frac{\pi }{2}y} \right)\].

The image is then transformed into:

242y9mp.jpg


Now, using polar coordinates, T doesn't depend of \rho, so then \[{\nabla ^2}T = \frac{1}{\rho }\frac{{{\partial ^2}T}}{{\partial {\varphi ^2}}}\]. This means that \[T(x,y) = A\theta + B\].

But I can't go pass that. I don't know how to find A and B using the countour conditions.

Any ideas?
 
There are two things I don't understand about this problem. First, when finding the nth root of a number, there should in theory be n solutions. However, the formula produces n+1 roots. Here is how. The first root is simply ##\left(r\right)^{\left(\frac{1}{n}\right)}##. Then you multiply this first root by n additional expressions given by the formula, as you go through k=0,1,...n-1. So you end up with n+1 roots, which cannot be correct. Let me illustrate what I mean. For this...
Back
Top