Handling constraints in NDSolve

*To*: mathgroup at smc.vnet.net*Subject*: [mg32002] Handling constraints in NDSolve*From*: "Philip M. Howe" <pmhowe at lanl.gov>*Date*: Mon, 17 Dec 2001 06:01:33 -0500 (EST)*Sender*: owner-wri-mathgroup at wolfram.com

Dear Folks, I am not sure how to use NDSolve with an equation set that contains constraints. A very simple example is shown below. There are two unknowns (x and y) which are functions of t. I have two equations and impose two initial conditions. Seems like it should work, but I am probably doing something foolish. If I remove the constraints, the system works fine. Any advice? Clear[x, y, t]; T = 1200; Eact = 149000; kzero = 5.468*10^6; rho0 = 1.834; Rgas = 8.314; L = 1; k1 = kzero* Exp[-Eact/(Rgas*T)]; s = NDSolve[ { x'[t] == x[t](L - x[t])*(1 - y[t]), y'[t] == -k1*y[t] /; t < x[t]/x'[t], y'[t] == 0 /; t =84 x[t]/x'[t], x[0] == 0.001, y[0] == 1}, {y[t], x[t]}, {t, 0, 100.} ]; y1a = y[t] /. s[[1]][[1]]; xa = x[t] /. s[[1]][[2]] Plot[y1a, {t, 0, 20}, PlotRange -> All]; Plot[xa, {t, 0.001, 20}, PlotRange -> All]; Thanks for any help you can give me. Phil -- Philip M. Howe Program Manager, Stockpile Surety Los Alamos National Laboratory (505) 665-5332 (505) 667-9498 Fax: 505-665-5249 email pmhowe at lanl.gov Mail Stop F606