next up previous
Next: About this document ... Up: Differential Equations Previous: First Order, Variable Separable

First Order, Linear

A differential equation of the form

\begin{displaymath}\frac{d p(T)}{dT} + g(T) p(T) = q(T)
\end{displaymath}

where g(T) and q(T) are known, i.e., given functions, and p(T) is the desired solution, a function of T, can be solved using integration factors. We note that multiplying by a factor (so far unknown) h(T), one would have

\begin{displaymath}h(T)\frac{d p(T)}{dT} + h(T)g(T) p(T) = h(T)q(T)
\end{displaymath}

which could be rewritten as

 \begin{displaymath}
\frac{d h(T)p(T)}{dT}- p(T)\frac{d h(T)}{dT} + h(T)g(T) p(T) = h(T)q(T)
\end{displaymath} (5)

If the second and third terms could be made to cancel, i.e., if

 \begin{displaymath}
p(T)\frac{d h(T)}{dT} = h(T)g(T) p(T)
\end{displaymath} (6)

the resulting differential equation (5) would read

\begin{displaymath}\frac{d [h(T)p(T)]}{dT}= h(T)q(T)
\end{displaymath}

which would be integrable. But p(T) cancels from both sides of the equation (6)

\begin{displaymath}
p(T)\frac{d h(T)}{dT} = h(T)g(T) p(T)
\end{displaymath}


\begin{displaymath}\frac{d h(T)}{dT} = h(T)g(T)
\end{displaymath}

so, dividing both sides by h(T) and multiplying both sides by dT and integrating one has

\begin{displaymath}\int \frac{1}{h(T)}\frac{d h(T)}{ dT} dT= \int g(T) dT
\end{displaymath}

Integration of the left hand side gives

\begin{displaymath}\ln h(T) = \int g(T) dT
\end{displaymath}

that is

\begin{displaymath}h(T) = e^{\int g(T) dT}
\end{displaymath}

which is, acutally, a formula for h(T).

As an example consider the first-order equation

\begin{displaymath}\frac{dp(T)}{dT} + 2 T p(T) = T \,\,;\,\, p(1)=3
\end{displaymath}

where g(T) = 2T and q(T) = T. One has

\begin{displaymath}\int 2T dT = T^2
\end{displaymath}

so

\begin{displaymath}h(T) = e^{\int 2T dT} = e^{T^2}
\end{displaymath}

so


\begin{displaymath}\frac{d[p(T)e^{T^2}]}{dT}
-2T e^{T^2} p(T)
+2T e^{T^2} p(T)
=
e^{T^2} T
\end{displaymath}

Cancelling, one has

\begin{displaymath}p(T) e^{T^2} = \int Te^{T^2} dT = \frac{e^{T^2}}{2} + C
\end{displaymath}

Finally, solving for p(T) produces

\begin{displaymath}p(T) = \frac{1}{2} + Ce^{-T^2}
\end{displaymath}

One now uses the initial condition to finish. Thus p(1) = 3

\begin{displaymath}3 = \frac{1}{2} + C e^{-1^2}
\end{displaymath}

giving an equation for C.


next up previous
Next: About this document ... Up: Differential Equations Previous: First Order, Variable Separable

1998-05-11