I am solving the following differential equation and boundary conditions

dsol = Flatten[DSolve[{
 0 == s''[x] - c n'[x] , 0 == n''[x] - c s'[x], 
 n'[L] - c s[L] == 0, s'[L] - c n[L] == 0
 }, {s[x], n[x]}, x]] // FullSimplify

which yields:

 {n[x] -> C[1] Cosh[c x] + (C[2] Sinh[c x])/c,  
 s[x] -> (C[2] Cosh[c x])/c + C[1] Sinh[c x]}

My question is: are the C[1] and C[2] integration constants identical for the two solutions? Or should I treat them as 4 separate quantities to be determined?

  • 1
    $\begingroup$ What happens if you try substituting these solutions you got into the original DE? Note the result of that, and repeat the experiment where the C[1] and C[2] of either s or n have been replaced with C[3] and C[4]. $\endgroup$
    – J. M.'s torpor
    Jul 30 '17 at 2:51
  • $\begingroup$ Yep, that was easy. C[1] = C[3] and C[2] = C[4] $\endgroup$
    – BeauGeste
    Jul 30 '17 at 2:58
  • $\begingroup$ You can answer your own question now, if you want. $\endgroup$
    – J. M.'s torpor
    Jul 30 '17 at 2:59

Nope, the constants are the same.

Originally I thought that Mathematica might treat the solutions as solutions to two different differential equations and then start its numbering of integration constants back at C[1].

However it's easy (HT @J.M.) to show that's not the case. Leave the first solutions as is but change the constants of the second solution to C[3] and C[4]. Now put back in the original differential equations. You will see that they are only satisfied if C[1] = C[3] and C[2] = C[4].


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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