Hello,
   Yes, I have found the solution with the help of Coventor. Before I
mention the cause, let me address your questions in case someone out
there finds this useful.

>user-specified heat flux, I have to guess that the heat flux off the
>surfaces may be greater than the heat generated within the solid.
This

The surprising thing was that it appeared that heat was flowing out of
the hotplate surface to the statically convecting (non-forced flow) air
even though the temperature of the air was higher than the hotplate
surface. This would be a violation of Fourier's law.

>You say that the lower temperature appears BETWEEN the metal heater
and
>the holes.  Does this occurs while the external surface temperatures
are
>above ambient?

Temperatures below ambient occurred in a region between holes that
perforate a nitride layer and a 25 micron wide Pt heater. This happened
when the heater was actually above ambient.

The problem turns out to have been caused by an incorrect material
property. Specifically, the thermal conductivity of the nitride layer
was 7 orders of magnitude too low. Of course, when I began the
simulation I looked up these properties and had entered the correct
values. At some point, perhaps while moving some of the files around, I
must have changed the MPD (Material Properties Database) file I was
using. In short the lesson is: Always check your material properties if
your getting weird results even if your sure you have correct values.


Thanks for the help Coventor!
    Mike Martin


>>> Raj Gupta  05/14/02 08:27PM >>>

Hi Michael,

Have you found an answer to your problem?  Not knowing exactly the
boundary conditions you have set, specifically the emissivity and any
user-specified heat flux, I have to guess that the heat flux off the
surfaces may be greater than the heat generated within the solid.
This
can happen in reality, such as what causes dew formation on a car left
outdoors overnight.  One would guess for your situation, the reality
is
far from your simulated results.

You say that the lower temperature appears BETWEEN the metal heater
and
the holes.  Does this occurs while the external surface temperatures
are
above ambient?

Are you using forced convection?  Keep in mind that thermal
conductivity
to ambient through normal convection will scale with geometry.
Reynold's number is much smaller on micron dimensions, and turbulent
flow which can assist in forced convection will not be as efficient on
smaller dimensions.

Let me know how you manage to solve this problem.  I am very curious.

Regards,
Raj

Michael D Martin wrote:

 > Hi, Does anyone have experience with modeling hotplates or using
the
 > ETherm module in MEM CAD? I am trying to model some hotplates with
 > meander type heaters and holes in them. The problem is that I keep
 > getting temps that are below the ambient temp that is set in a
 > convection-radiation boundary condition on the top and bottom of
the
 > plate. The problem seems to occur between the metal heater and the
holes
 > in a nitride support layer. Unfortunately, Coventor has NOT been
too
 > helpful so far.
 >
 > Thanks,
 >    Mike


--

Raj Gupta, Director
    Nanotechnology & MEMS Consulting
    33 Vandewater Street
    Suite 205
    San Francisco, CA  94133-1967
Website: http://www.mindspring.com/~rajgupta