<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
 <channel>
  <atom:link href="http://www.comsol.fi/community/forums/general/rss/thread/3841.rss" rel="self" type="application/rss+xml"/>
  <title>COMSOL Forums: manually changing temperature values at specified time</title>
  <link>http://www.comsol.fi/community/forums/general/thread/3841/</link>
  <description>Most recent forum messages</description>
  <pubDate>Wed, 17 Mar 2010 09:57:37 +0000</pubDate>
  <image>
   <title>COMSOL Forums: manually changing temperature values at specified time</title>
   <url>http://www.comsol.fi/shared/images/logos/comsol_logo.gif</url>
   <link>http://www.comsol.fi/community/forums/general/thread/3841/</link>
  </image>
  <item>
   <title>Re: manually changing temperature values at specified time</title>
   <link>http://www.comsol.fi/community/forums/general/thread/3841/#p10192</link>
   <description>Where would you enter the piecewise definition?  I entered &amp;quot;80*(t&amp;lt;60) + 900*(t&amp;gt;=60)&amp;quot; (as a simple example) in the Subdomain Expression field for the relevant subdomain, and the solver produced a duplicate variable error for T.</description>
   <pubDate>Wed, 17 Mar 2010 09:57:37 +0000</pubDate>
   <guid isPermaLink="false">3841.1268819857.10192</guid>
  </item>
  <item>
   <title>Re: manually changing temperature values at specified time</title>
   <link>http://www.comsol.fi/community/forums/general/thread/3841/#p10099</link>
   <description>Thanks, I will give it a shot.  &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
The reason I don't want to continuously couple them is because I am modelling a &amp;quot;pseudo-assembly line&amp;quot;, where the &amp;quot;subjects&amp;quot; move at a set time.  Basically, the model has warm boxes placed at one point, and then these move down the line.  And I want to investigate the heat transfer around them.  I thought about a moving mesh for this, but with 9 moving boxes I thought this might be too many calculations, and I thought this way might be simpler.</description>
   <pubDate>Tue, 16 Mar 2010 11:41:44 +0000</pubDate>
   <guid isPermaLink="false">3841.1268739704.10099</guid>
  </item>
  <item>
   <title>Re: manually changing temperature values at specified time</title>
   <link>http://www.comsol.fi/community/forums/general/thread/3841/#p10051</link>
   <description>Hi&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
For me you have two questions in there:&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
1) change the temperature at given times: yes why not try &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
T = T0*(t&amp;lt;t0)+T1*(t&amp;gt;=t0)*(t&amp;lt;t1)*...*Tn*(t&amp;gt;=tn)&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
or otherwise simpler: define a piecevise function fT(t) and say T=fT(t)&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
2) couple the temperature between two items, or model at given times&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Well why at given times ? why not express a continously linked physics and solve them together ? &lt;br /&gt;&#13;
There are a few examples in the doc about the weak constraints and boundary integrations, probably this is a good example for the non-ideal weak contraits (single way link), even an example somewhere about a thermal case with the temperature of one border linking around onto the other.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
If not I would try to play with some integration variables over the domain to measure and then use the equation above for the controlled domain, but one must add some &amp;quot;*at(ti)&amp;quot; where &amp;quot;ti&amp;quot; are some of the PREVIOUS solved time steps to get it to work. But I'm not sure this (the &amp;quot;at()&amp;quot; operator) would even work during the simulation, you would have to test it out carefully, never done something like that&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Hope this helps&lt;br /&gt;&#13;
Good luck&lt;br /&gt;&#13;
Ivar&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
</description>
   <pubDate>Mon, 15 Mar 2010 14:22:00 +0000</pubDate>
   <guid isPermaLink="false">3841.1268662920.10051</guid>
  </item>
  <item>
   <title>manually changing temperature values at specified time</title>
   <link>http://www.comsol.fi/community/forums/general/thread/3841/#p10027</link>
   <description>Hi,&lt;br /&gt;&#13;
I am wondering if it is possible to change the temperature of a subdomain at a chosen time.  And in addition to this, can the specified temperature be taken from another subdomain within the model?  &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
For example, I have two blocks (A and B).  A starts at 1600 K and is being cooled and at t = 60, I want the temperature of block B to equal that of block A at t = 60.  I don't have need to have the exact same gradients (the average temperature is fine).  </description>
   <pubDate>Mon, 15 Mar 2010 11:33:25 +0000</pubDate>
   <guid isPermaLink="false">3841.1268652805.10027</guid>
  </item>
 </channel>
</rss>

