Discussion Forum

Hi

I believe you are missing something in your model: if you have a constant voltage applied, you get a given current and hence a total power entering the system, each second you add some joules that responses in a temperature increase.

This will not stop so long you are not sinking (extracting some power) someway, so I believe COMSOL is right , this time, at least (as mostly it is,by the way ;)

--
Good luck
Ivar

Thanks for your fast answer....
I'm applying voltage using "joule heating -> electric potencial" ... i think i understand what you said but i still don't know how to do it. What i'm trying to simulate, it's a heated wire that will be used for heating, like a heating blanket. In that case there's nothing sinking, the power applied in the wire is dissipated through the wire and somewhere the temperature will be constant, RIGHT?

Thanks for the help btw

---
Rafael Oliveira

Hi Rafael,
Did you take into account heat loss from your wire? If not this temperature will grow without bound, due to the heat source (Joule heating).
Regards,
Andrzej

Hi

the temperature will stabilise if the heat losses becomes equivalent to the heat generated. If you have no heat losses the temperature will continue to rise, forever.

In a light bulb, what is happening is that as the temperature rises the resisitivity of the wire changes (increases) which alreday is limiting the maximum power dissipated (for a given voltage, but not for a constant current !), then as the temperature rises the filament/wire will loose more and more energy by radiation, and some by convection/conduction from the gas used around.

If you want to simulate this, you need to
1) use a non linear resisitivity, that is dependent on the local temperature,
2) define losses (energy that leaves the system, not only entering) such as radiation losses related to the 4th power of the absolute temperature of the object, via the Stefan Boltzmann constant and the material emissivity, including some energy that is received by the surroundings, and finally some conduction and convection, often defined as an energy flux depening +/- linearly on the temperature difference, between the object and its surrounding

Tout un programme ;)

--
Good luck
Ivar

hi all...
now i understand clearly what you've said but even with that i'm a begginer with COMSOL software. i'm trying to learn by myself but i think it's more complicated than i thought, because there's too many options and too many details that i can't handle alone... i was trying to do a "simple" simulation but to do it right i have to know really what i'm doing and every COMSOL function that i have to use...
i think i'll quit :P

Thank you anyway ...

Best regards
----
Rafael Oliveira

Hi

start to think "physics": energy balance, inputs/outputs forces/flux in/out, etc the build you model, you will see you will soon become an expert, just do not give up ;)

--
Good luck
Ivar

Don't give up!

I think you need to put your problem on paper and figure out the overall energy balance of your system, then the boundary conditions. In your case, what seems missing is an appropriate boundary condition on the wire surface, such as natural or forced convection. If your wire is in a blanket, I think you need to determine a heat flux term on your wire that will depend on the wire temperature, you need to establish that on paper first.

Comsol is easy! Just a few days and you will "capture the concept", thrust me.

Salut.

Hi Everyone,

I am trying to simulate a simple MEMS heater device ( a thin wire with a diameter of 2 um attached to two big gold contact pads), but so far I have not been able to get rational results. I made this device (in reality!) and I tested it and it worked but when I use COMSOL to simulate it, the results are completely irrational.

This heater works with a 0.5 V and it is made of gold. When I apply the voltage, the temperature increases to ~4700 K, if this had happened in reality,my device would have melted, but it did not.

I define the temperature of my big pads as room temperature, so they are supposed to act as heat sinks. Also, I defined inward heat flux with a coefficient of 10 (W/m^2.K). So, heat produces and dissipates (if I am right!).

I really appreciate if you can tell me what has caused this really high temperature.



Thank you,

Sincerely,

Keivan

Hi

I'm sure you have a BC issue, as I have done many MEMS joule heating simulations and get good coherent results ;)

I'm slightly puzzeled by you heat "influx" ? you should use heat convection loss I assume (and 10[W/m/K] is rather high valid for forced convection in my view but OK) if this sign is wrong, I suppose that 4-5 tousand K is quite normal ;)

And yre you sure the heat exchange Au electrode to the support SiO2 ? or Si is correct, because even 0.5V in a micron thin Au wire will give quite high values, depends on 1) the wire length, sure but b) the resisitivity change with temperature.

so for 2) are you using constant electric conduction fr Au or a linear temperature dependent one ?

Finally, try rather to use latest version and patch, the early v4.0 was really not "ironed out", and I found it frustrating to use

--
Good luck
Ivar

Dear Ivar,

Thank you so much for your kind reply and explanation. At that time, I thought that nobody was going to answer my question, so I just stopped working on that problem. Today while I was browsing I saw your quote. I really appreciate your time and kind attention.

I was disappointed in solving that problem and I gave up. I will start working on it according to your points.

Thank you again,

Sincerely,

Keivan