40 Reputation

4 Badges

4 years, 352 days

MaplePrimes Activity

These are questions asked by Jjjones98

I am attempting to use the pdsolve function in Maple to explicitly solve a PDE which is basically a perturbation of the Laplace equation.  Nothing happens when I enter pdsolve, however, is this because a boundary condition is also needed to produce a solution?  The BC which I have is that f(x, y, z) goes to zero as sqrt(x^2 + y^2 + z^2) goes to infinity but I am not sure how to enter such a BC in Maple.

PDE := diff(diff(f(x, y, z), x), x) + diff(diff(f(x, y, z), y), y) + diff(diff(f(x, y, z), z), z) - exp(-t*exp(sqrt(x^2 + y^2 + z^2)))*(diff(diff(f(x, y, z), x), x) + diff(diff(f(x, y, z), y), y) + diff(diff(f(x, y, z), z), z))/(1 + m/(2*sqrt(x^2 + y^2 + z^2)))^4 = 3/2*exp(sqrt(x^2 + y^2 + z^2) - t*exp(sqrt(x^2 + y^2 + z^2)))*(diff(f(x, y, z), x)*tx/sqrt(x^2 + y^2 + z^2) + diff(f(x, y, z), y)*ty/sqrt(x^2 + y^2 + z^2) + diff(f(x, y, z), z)*tz/sqrt(x^2 + y^2 + z^2))/(1 + m/(2*sqrt(x^2 + y^2 + z^2)))^4;


Is it possible to determine an analytic solution to the following system of two differential equations for $A$ and $B$ using Maple.  My suspicion is that trial and error would find an analytic solution in theory and so that Maple could find the solution.  M is a constant and \sigma is some arbitrary function of t and the spatial coordinates. 

\[ \Bigg( \frac{\partial^2}{\partial x^2} + \frac{\partial^2}{\partial y^2} + \frac{\partial^2}{\partial z^2} + \frac{1}{2} \Bigg( 1 + \frac{M}{2 \sqrt{x^2 + y^2 + z^2}} \Bigg) \Bigg( \frac{\partial \sigma}{\partial x }\frac{\partial}{\partial x} +\frac{\partial \sigma}{\partial y}\frac{\partial}{\partial y} +\frac{\partial \sigma}{\partial z}\frac{\partial}{\partial z} \Bigg) \Bigg)B=0, \]

\[\frac{d A}{dt} = AB.\]

Furthermore, the boundary conditions are 

\[B \rightarrow -1  \: \text{as}  \: \sqrt{x^2 + y^2 + z^2} \rightarrow \infty,\]

\[A \rightarrow e^{-t} \: \text{as} \: \sqrt{x^2 + y^2 + z^2} \rightarrow \infty \]


I have written a bit of code which solves a linear system for some quantities which have been Laplace and then Fourier transformed.  

e1 := -2*D*i*k*pi + A*s = 0

e2 := 2*A*i*k*pi + 2*C*i*k*pi + B*s = a

e3 := s*C + 4/5*P*w3*(2*pi*i*k*D - 2*1/3*pi*i*k*D)/w2 = -2*(C + 2*K*(2*pi*i*k*B - 2*1/3*pi*i*k*B))/(w2*K)

e4 := s*D + 2*5/4*P*t4*pi*i*k*C/(t2*K) = -5/(2*P)*D/(t2*K)

sys := {s*C + 4/5*P*w3*(2*pi*i*k*D - 2*1/3*pi*i*k*D)/w2 = -2*(C + 2*K*(2*pi*i*k*B - 2*1/3*pi*i*k*B))/(w2*K), s*D + 2*5/4*P*t4*pi*i*k*C/(t2*K) = -5/(2*P)*D/(t2*K), -2*D*i*k*pi + A*s = 0, 2*A*i*k*pi + 2*C*i*k*pi + B*s = a}

solve(sys, [A, B, C, D])


I get at the end some fractions where everything in the fractions is a constant with some physical meaning except for k which is the only frequency as I am working in one dimension so just need one-dimensional Fourier and Laplace transforms.  s is the corresponding variable from the Laplace transform. 

I was wondering if Maple had some functionality which would enable me to inverse Laplace and then inverse Fourier transform these quantities A, B, C and D from the linear system such that I obtain an algebraic expression at the end and not a numerical result.

I have solved the following linear system for 6 variables on Maple using the following code:

sys := {w = -2*Pi*i*k_2*v + 2*Pi*i*k_2*(4*K^2)/(5*Pi)*u, z = -2*Pi*i*k_1*v + 2*Pi*i*k_1*(4*K^2)/(5*Pi)*u, p*s*x = -2*Pi*i*k_1*u - 4*Pi^2*(k_1^2 + k_2^2)*x + a_1, p*s*y = -2*Pi*i*k_2*u - 4*Pi^2*(k_1^2 + k_2^2)*y + a_2, k_1*z + k_2*w = 0, k_1*x + k_2*y = 0}

solve(sys, [x, y, z, w, v, u])

However, the solution yields z = w = 0, I was wondering if this is correct as I feel that these quantities should not be 0 in the problem which I am studying, is there a way to find a solution which does not involve setting these two to 0?


I have a function e^(-\lambda z \sqrt(x^2 + y^2)), is it possible to use Maple to find some sequence of derivatives wrt to x and y which could be applied to this function to get

(z/(1+2*sqrt(x^2 + y^2)*lambda))*e^(-\lambda z \sqrt(x^2 + y^2))



1 2 3 4 5 Page 1 of 5