How to Solve the IVP Using an Integration Factor?

  • Context: MHB 
  • Thread starter Thread starter hiyum
  • Start date Start date
  • Tags Tags
    Ivp
Click For Summary

Discussion Overview

The discussion revolves around solving the first order differential equation \( y' = y + 2te^{2t} \) using various methods, including integration factors and Duhamel's Principle. Participants explore different approaches to find the general solution, while clarifying concepts related to linear differential equations.

Discussion Character

  • Technical explanation
  • Mathematical reasoning
  • Debate/contested

Main Points Raised

  • Some participants note the equation is a first order differential equation and express it in standard form.
  • One participant points out that the problem is not an initial value problem (IVP) due to the absence of an initial condition.
  • Another participant suggests separating the equation into homogeneous and non-homogeneous parts, proposing a solution for each and combining them.
  • A different approach is introduced using Duhamel's Principle, with a detailed outline of the method and its equivalence to the Variation of Parameters technique.
  • Some participants derive the general solution as \( y = Ce^t + (2t - 2)e^{2t} \) through different methods, while others present variations of this solution.
  • One participant discusses using an integration factor and simplifies the left-hand side of the equation, leading to an integral expression for the solution.

Areas of Agreement / Disagreement

Participants present multiple competing views and methods for solving the differential equation, with no consensus on a single approach or solution. The discussion remains unresolved regarding the best method to apply.

Contextual Notes

Some participants express uncertainty about the classification of the problem as an IVP and the implications of using different solution methods. There are also unresolved mathematical steps in the derivations presented.

hiyum
Messages
2
Reaction score
0
\[ y'=y+2te^{2t} \]
 
Physics news on Phys.org
hiyum said:
\[ y'=y+2te^{2t} \]
This is a first order differential equation. Putting it into standard form:
[math]y' = y + 2t e^{2t}[/math]

[math]y' - y = 2t e^{2t}[/math]

Multiply both sides by [math]e^{-t}[/math] (which is the integration factor):
[math]e^{-t} y' - e^{-t} y = 2t e^t[/math].

What can you do to simplify the LHS?

-Dan
 
But this is NOT an "IVP" (initial value problem) because there is no initial value given.
 
Since this is a linear differential equation with constant coefficients we can also separate it into its 'homogeneous' and 'non-homogeneous' parts, find solutions to each, then add.

The 'homogeneous' part is y'= y which has the obvious general solution $y= Ce^t$ for any constant C.

To find a solution to the non-homogeous part try $y= (At+ B)e^{2t}$.

Then $y'= Ae^{2t}+ 2(At+ B)e^{2t}= (2At+ A+ 2B)e^{2t}$.

So we have $(2At+ A+ 2B)e^{2t}= (At+ B)e^{2t}+ 2te^{2t}$. Dividing by $e^{2t}$, $2At+ A+ 2B= At+ B+ 2t$.

This has to be true for all t. Taking t= 0, A+ 2B= B or A= -B. Taking t= 1, 2A+ A+ 2B= A+ B+ 2. 2A+ B= 2.

Since A= -B. -2B+ B= -B= 2 so B= -2 and A= 2.

$y= (2t- 2)e^{2t}$ satisfies $y'= y+ 2te^{2t}$: $y'= 2e^{2t}+ (4t- 4)e{2t}= (4t- 2)e^{2t} $ while $y+ 2te^{2t}= (2t- 2)e^{2t}+ 2te^{2t}= (4t- 2)e^{2t}$.

Therefore the general solution to $y'= y+ 2te^{2t}$ is $y= Ce^t+ (2t- 2)e^{2t}$.
 
I'm usually not one for replying to answered threads, especially in light of the expert help already provided by topsquark and County Boy. That said, I'm detailing another solution outline here because the method on which it's based, Duhamel's Principle, is one of my favorites. What I like about this method is its generality, including its application to inhomogeneous PDEs. Note that in the case of linear inhomogeneous ODEs, which this problem is, Duhamel's Principle is equivalent to a technique known as the Variation of Parameters.

We consider the problem
$$ y'-y = 2te^{2t}, \qquad y(0) = y_{0}\qquad (*)$$
Set $v = y-y_{0}$ and note that $v' = y'$. Then, upon substituting for $y$, $(*)$ becomes
$$v'-v = 2te^{2t} + y_{0}, \qquad v(0) = 0 \qquad(**)$$
This is an inhomogeneous linear evolution equation, whose solution is given by
$$v(t) = \int_{0}^{t}(P^{s}f)(t)ds,$$
where $(P^{s}f)(t)$ is the solution of the problem
$$v'-v = 0, \qquad v(s) = f(s) = 2se^{2s}+y_{0}.$$
See Duhamel's Principle - Wikipedia for a reference.

The solution to $v'-v = 0$ is given by $v(t) = Ce^{t}.$ The condition $v(s) = 2se^{2s}+y_{0}$ implies $C = 2se^{s}+y_{0}e^{-s}$. Thus,
$$(P^{s}f)(t) = \left[2se^{s} + y_{0}e^{-s}\right]e^{t}.$$
According to Duhamel's Principle, the solution to $(**)$ is
$$\begin{align*}
v(t) &= \int_{0}^{t}(P^{s}f)(t) ds\\
& = \int_{0}^{t}\left[2se^{s} + y_{0}e^{-s}\right]e^{t}ds\\
& = e^{t}\int_{0}^{t}2se^{s} + y_{0}e^{-s}ds
\end{align*}
$$
Using integration by parts on the $2se^{s}$ term we obtain
$$
\begin{align*}
v(t) &= e^{t}\left[2se^{s}-2e^{s} - y_{0}e^{-s}\Biggr|_{0}^{t} \right]\\
&= 2te^{2t}-2e^{2t}+(2+y_{0})e^{t} - y_{0}
\end{align*}
$$
Using $v = y - y_{0}$ to substitute back for $y$ and setting $C = 2+y_{0}$, we obtain
$$y(t) = Ce^{t} + (2t-2)e^{2t},$$
as was previously demonstrated.
 
That last is, I believe, equivalent to "variation of parameters". Having determined that the general solution to the associated homogeneous equation is $y= Ce^t$, we look for a solution to the entire equation of the form $y(t)= u(t)e^t$ (we allow the "parameter", C, to vary). Then $y'(t)= u'(t)e^t+ u(t)e^t$ and the equation becomes $y'- y= u'(t)e^t+ u(t)e^t- u(t)e^t= u'(t)e^t= 2te^{2t}$.

Dividing both sides by $e^t$, $u'(t)= 2te^t$.
Integrating that, $u(t)= 2te^t- e^t= (2t- 1)e^t$.
$u(t)e^t= (2t- 1)e^{2t}$ and the general solution to the equation is $y(t)= Ce^t+ (2t- 1)e^{2t}$.
 
I liked topsquark's suggestion to use the integration factor $e^{\int^t -1\cdot dt}=e^{-t}$.
Then we get:
$$y' - y = 2t e^{2t} \implies e^{-t} y' - e^{-t} y = 2t e^t \implies (e^{-t} y)' = 2t e^t \implies e^{-t}y = \int 2t e^t\,dt \implies y = e^t \int 2t e^t\,dt$$
 

Similar threads

  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 3 ·
Replies
3
Views
1K
  • · Replies 4 ·
Replies
4
Views
1K
Replies
1
Views
2K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 2 ·
Replies
2
Views
1K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 2 ·
Replies
2
Views
1K