Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I am trying to solve the following difference equation:

$$-\frac{\epsilon}{h^2}U_{n+1}+\left(\frac{2\epsilon}{h^2}+\frac{1}{h}\right)U_{n}-\left(\frac{\epsilon}{h^2}+\frac{1}{h}\right)U_{n-1}=0,\mbox{ }\mbox{ }\mbox{ }\mbox{ }U_0=1,\mbox{ }U_1=0.$$

I try $U_{n}=Aw^n$ then I get


This seems a bit far from what I want to get. I am trying to verify that the solution is


where $0\leq n\leq N$ and $\rho=h/\epsilon$.

share|cite|improve this question
In your third term in your difference equation, did you mean $U_{n-1}$? – Cameron Williams Jul 4 '13 at 18:19
If you want to verify a solution you already have, just substitute it in the equation and check it is true, and also check the initial conditions. – Mlazhinka Shung Gronzalez LeWy Jul 4 '13 at 18:39
Note that the homogeneous equation has solution $U_n=1$, equivalently you have a solution $w_1=1$. This makes life a lot easier. – Mark Bennet Jul 4 '13 at 18:39
@CameronWilliams, yes and thanks for that. – Vaolter Jul 5 '13 at 7:40
@MarkBennet, could you elaborate. – Vaolter Jul 5 '13 at 7:42
up vote 1 down vote accepted

First of all note that, inside the square root, we have $$ \left(\frac{2\epsilon}{h^2}+\frac{1}{h}\right)^2-4\frac{\epsilon}{h^2}\left(\frac{\epsilon}{h^2}+\frac{1}{h}\right) = \frac{1}{h^2} $$ So the answers read $$w_{1,2}=\frac{\left(\frac{2\epsilon}{h^2}+\frac{1}{h}\right)\pm \frac{1}{h}}{2\frac{\epsilon}{h^2}}. \Longrightarrow w_1=1, w_2=1+\frac{h}{\epsilon} $$ This means that the solution reads $U_n=A+B(1+\rho)^n$. Now let us impose boundary conditions . Since $U_0=1$, we get that $A+B=1$. Also, since $U_1=0$, we get that $A+B(1+\rho)=0$ or equivalently, $A=-(1+\rho)B$. Plugging this into the first equation, we get that $B=\frac{-1}{\rho}$, and that $A=\frac{1+\rho}{\rho}$, leading to a solution of the form $$U_n=\frac{1+\rho}{\rho} \bigg[ 1-(1+\rho)^{n-1} \bigg]. $$ We can readily check that for $n=1$, we get $U_1=0$, and $n=0$, we get $U_0=1$.

bBut this is not what you are trying to recover. Your desired answer has $N$ in it, which is defined nowhere in the statement of the problem. My conjecture is that, $n$ is limited to the range $0,N$.

Let us now consider the following set of boundary conditions: $$U_0=1,U_N=0$$ Imposing the first condition on $U_n=A+B(1+\rho)^n$, we get $A+B=1$. Imposing the second condition, we get $A+B(1+\rho)^N=0$, which can be equivalently expressed as $A=-B(1+\rho)^N$. Plugging this into the first equation, we obtain $B=\frac{1}{1-(1+\rho)^N}$ . Also for $A$ we obtain $A=\frac{ -(1+\rho)^N}{1-(1+\rho)^N}$.

For the solution, we use the values of $A,B$ obtained above, and arrive at

$$ U_n=\frac{(1+\rho)^n-(1+\rho)^N}{1-(1+\rho)^N}$$

Dividing both the numerator and the denominator by $-(1+\rho)^N$, we arrive at

$$ U_n=\frac{1-(1+\rho)^{n-N}}{1-(1+\rho)^{-N}}$$

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.