Path: news.mathworks.com!newsfeed-00.mathworks.com!newsfeed2.dallas1.level3.net!news.level3.com!postnews.google.com!news2.google.com!border1.nntp.dca.giganews.com!nntp.giganews.com!nx01.iad.newshosting.com!newshosting.com!69.16.185.51.MISMATCH!tmp-post01.iad!news.highwinds-media.com!not-for-mail
Date: Wed, 05 Nov 2008 17:08:11 -0500
From: Freelance Embedded Systems Engineer <g9u5dd43@yahoo.com>
User-Agent: Thunderbird 2.0.0.17 (Windows/20080914)
MIME-Version: 1.0
Newsgroups: comp.soft-sys.matlab
Subject: Re: Stiff Problems and small step sizes
References: <get1at$jfg$1@fred.mathworks.com>
In-Reply-To: <get1at$jfg$1@fred.mathworks.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Lines: 36
Message-ID: <49121953$0$4888$9a6e19ea@unlimited.newshosting.com>
Organization: Newshosting.com - Highest quality at a great price! www.newshosting.com
NNTP-Posting-Host: 75.184.122.147
X-Complaints-To: abuse@newshosting.com
Xref: news.mathworks.com comp.soft-sys.matlab:499200


William Kong wrote:
> I'm analyzing a cargo dynamics problem where a box is tied down by
> chains on to a platform. I am trying to analyze the movement of the
> box in the event of a crash. The chains I'm using the tiedowns are
> extremely stiff and strong. Thus there are large forces imparted on
> to the box to arrest its movement.
> 
> The model consists of a system of 1st order ODEs for a 6DOF problem.
> I use ode15s to solve this stiff problem.
> 
> To simulate a crash, I have both the box and platform moving at the
> same initial velocity and then the platform suddenly stops at some
> time as if it hit a wall. The box would keep going if it wasn't for
> the tiedowns. When I try to run the simulation in Matlab, I get this
> error message before it finishes:
> 
> " Warning: Failure at t=6.500000e-03.  Unable to meet integration
> tolerances without reducing the step size below the smallest value
> allowed (1.387779e-17) at time t."
> 
> Any suggestions as to what I can do about this?
> 
> Another way of simulating this crash would be to give only the
> platform an initial velocity or an initial "kick". The error doesn't
> come up in this case for some reason. However, I would prefer to do
> it the former way where both are moving together at the same speed
> and then only the platform comes to a sudden stop.

This sounds a little like the mixed continuous/discrete bouncing ball demo.
do "help ballode" and run "ballode".
Basically it uses the event detection capabilities of ODE
to switch and restart the state equations.
I've used this capability to study "stiction".