[ODE] LCP internal error, s <= 0

Russ Smith russ at q12.org
Wed Jul 31 21:32:02 2002


> Is there still interest in an example that causes:
> 
>     ODE Message 3: LCP internal error, s <= 0

yes, i'm interested. my highest priority is to fix any dynamics bugs
that arise. it is more likely however that your system is close to
singular (i.e. subtly misconfigured so that the solver can not find a
good solution). singular systems will often fail to be solved in the LCP
code. of course in an ideal world the solver would diagnose this problem
automatically and print a more informative message.

try increasing global CFM and see what happens. increasing global CFM
usually moves singular systems away from the bad configuration.

russ.

-- 
Russell Smith
http://www.q12.org