Re: ntdvmm error
- To: mathgroup at smc.vnet.net
- Subject: [mg95513] Re: ntdvmm error
- From: dh <dh at metrohm.com>
- Date: Tue, 20 Jan 2009 05:49:08 -0500 (EST)
- References: <gkpps1$doi$1@smc.vnet.net>
Hi,
the problem comes from fpend[A_, B_] and fpendw[A_, B_]. E.g. look at:
fpend[0.001,0.1]. The equation to solve reads:
{-0.00001`+(w')[x]+(vz'')[x]==0,
-2 w[x]-(-6+w[x] (3+2 w[x]+2 w[x]^2))/(2 (10+2
w[x]^2+w[x]^4))-(vz')[x]==0,vz[0]==0,(vz')[0]==0.001`,w[0.5`]==0.1`}
note that we can easily solve the second eq. for (vz')[x]. If we
differentiate the resulting expression we get vz'', what allowes us to
eliminiate vz and get an equation for w:
{-0.00001+3 (w')[x]+((-6+w[x] (3+2 w[x]+2 w[x]^2)) (4 w[x] (w')[x]+4
w[x]^3 (w')[x]))/(2 (10+2 w[x]^2+w[x]^4)^2)-((3+2 w[x]+2 w[x]^2)
(w')[x]+w[x] (2 (w')[x]+4 w[x] (w')[x]))/(2 (10+2
w[x]^2+w[x]^4))==0,w[0.5]==0.1}
NDSolve will solve this easily.
Knowing w we can solve the equation for vz.
Hope this helps, Daniel
SK wrote:
> Hi
>
> Im trying to solve two coupled nonlinear differential equations using
> the shooting method (see variables A and B). The file is can be
> downloaded at
>
> http://web.mit.edu/~shahriar/Public/submission.nb
>
> When I run it I keep getting the error
> NDSolve::ntdvmm: Cannot solve to find an explicit formula for the
> derivatives. NDSolve will try solving the system using a mass matrix
> method. >>
>
> and then Mathematica hangs and doesnt do anything.
> If I change Hz->0 for T, Mathematica solves it properly but it doesnt
> seem to like it when Hz is nonzero.
>
> Any help on getting past this error will be greatly appreciated
> Thanks
> S
>