I'm just looking for some guidance on how to proceed when dealing with a heat equation that looks like a Neumann condition problem, but has nonzero boundaries. Here's the general problem I would like to solve:

$$u_t=ku_{xx}, \,\,\,\,0\leq x\leq L, \,\,\,\,t>0$$ $$\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\textbf{BC: }\,\,\,\, u_x(0,t)=a, \,\, u_x(L,t)=b, \,\,\,\,t>0$$ $$\textbf{IC: }\,\,\,\, u(x,0)=f(x),\,\,\,\, 0\leq x\leq L$$

Is it as easy as just applying a stead state argument like I would do for the boundary conditions $u(0,t)=a, \,\,\,\,u(L,t)=b$? If the boundary conditions were instead time-dependent would I also be able to use the same adaptation of the steady state as well? This seems like the obvious answer, but I simply cannot find any references that provide a solution to such a problem.


1 Answer 1


Answering my own question:

We need to reduce the problem to an equivalent one which is the sum of two easily solved differential equations.


where $w(x)$ satisfies the boundary conditions $w_x(0)=a$, and $w_x(L)=b$. The caveat of this problem is that we will need to integrate our result afterwards to determine the boundary conditions of $v(x,t)$. This is easy enough though, because $w(x)$ does not depend on $t$. We take $\frac{dw}{dx}$ to be linear: $$\frac{dw}{dx}=\frac{(b-a)}{L}x+a.$$ Then, integration gives $$w(x)=\int_0^x \left[ \frac{(b-a)}{L}t+a\right] dt=\frac{(b-a)}{2L}x^2+ax$$

Now, we need only solve for $v(x,t)$. The fact that $w(x)$ is a second degree polynomial means that $v(x,t)$ is a nonhomogeneous differential equation with homogeneous boundary conditions. I.e.


$$\textbf{BCs:}\,\,\,\,v_x(0,t)=v_x(L,t)=0,$$ $$\textbf{IC}\,\,\,\,\,\,\,\,\,v(x,0)=f(x)-w(x).$$

The solution for $v(x,t)$ can easily be found using Duhamel's Principle, which is well documented for all homogeneous boundary conditions.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .