0
$\begingroup$

I recently solved (non-numerically) the differential equation of the simple pendulum problem,

$$y''=-c*\sin(y)\tag{1}.$$

I then tried to solve the equation with drag. I failed. I don't want to solve it numerically because I'm doing this for fun and that would be fun, $$y''=-c*sin(y)-a*y'\tag{2}.$$

So how would I go about adding drag to $y$ ad hoc, or without actually solving the second differential equation?

The constants of $y$ in my solution shift the equation along the $x$-axis and change the slope at $y(x)=0$ & $|y'(x)|=y'(x)$.

If you give me an equation of how the amplitude should change over time that would also work since the slope and amplitude are directly related.

$\endgroup$
12
  • 1
    $\begingroup$ Please share your non-numerical solution to the first equation. $\endgroup$
    – G. Smith
    Commented Mar 28, 2020 at 17:29
  • $\begingroup$ @G.Smith It's a standard result. The solution is an elliptic integral. $\endgroup$
    – alephzero
    Commented Mar 28, 2020 at 17:44
  • $\begingroup$ @alephzero Isn’t that for the period, not $y(x)$? $\endgroup$
    – G. Smith
    Commented Mar 28, 2020 at 17:50
  • $\begingroup$ @G. Smith You can use one to find the other. $\endgroup$ Commented Mar 28, 2020 at 17:56
  • 1
    $\begingroup$ Why can you not extend your series solution to handle the drag case? $\endgroup$
    – G. Smith
    Commented Mar 28, 2020 at 20:42

1 Answer 1

0
$\begingroup$

I recently solved (non-numerically) the differential equation of the simple pendulum problem.

As @G. wrote in the comments:

Please share your non-numerical solution to the first equation.

The ODE:

$$y''=-c \sin(y)$$

is usually solved with the small angle approximation, which is assumed valid for small angles $y$:

$$\sin(y) \approx y$$

So the ODE becomes:

$$y''=-cy$$

which does have a simple analytical solution:

$$y=C_1\sin(\sqrt{c}x)+C_2\cos(\sqrt{c}x)$$

where $C_1$ and $C_2$ are integration constants, determined from boundary/initial conditions.

Now if we add the damping term $ay'$, slight reworking gives:

$$y''+ay'+cy=0$$

This is a second order linear homogeneous ODE which also has analytical solutions. You can find the algorithm on plenty web resources, just google it.

$\endgroup$
1
  • 1
    $\begingroup$ Neither I nor, I assume, the OP are interested in the well-known small-angle approximation. But this answer is useful to others who are. $\endgroup$
    – G. Smith
    Commented Mar 28, 2020 at 20:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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