Discussion Forum

Hi

you have a non trivial issue there, your media are only partially absorbing. HT is set up by default for absorbing materials, so you need to integrate the heat flux per boundary and domains depending on the spectra and absorptions and distribute the fluxes and sources accordingly, not to say you have internal reflections and probably also radiation exchange.

You should start sketching out the exchange in-out relationships, and decide which contributions of radiation, versus conduction and convection you want to include, then set up your model

Nice and interesting model by the way

--
Good luck
Ivar

Hi Ivar.
Thank you for help.
Before, I'm new a comsol user.
The problem did formated two divide.
The target is to measure temperature of outgoing water on the thermal pannel.
Therefore I'm thinking transfer as picture. (Model 1 to Medel 2)
Dear Ivar. Have you different advice for me?

Hi

I see you have started well, by dividing into smaller models (indeed I have been told that the best way to digest an elefant, is to start to chop it into smaller parts, but I'm no hunter so I havent really tried, literally ...)
What stroke me first was your solar energy flux 1kW that is over a large wavelength range, and glass, as many other materials are only partially transparent to these wavelengths. COMSOL is not, today, set up to automatically integrate over the spectral wavelengths for material absorption coefficients (and I see nothing on your drawing about that chapter).
But the physics solver is compatible, so it's a question of setting up the material data base to operate that way, perhaps it will come with one of the future versions ? I hope so anyhow it would decrease the time to set up such models.
Another item, I have learned by using COMSOL, is that cooling with small water pipes can often be simulated by studying a 1m standard length of pipe, often in 2D-axi as the copper jacket will mostly diffuse the thermal flux all around rapidly, and then one apply or map the BC results to the more complex model. But if you have the same tube running over all your area, you will have quite some temperature increase and a parametric sweep of increasing inlet temperature for your 1m standard water pipe model can then give a better estimation of heat extraction withnthe increasing temperature.

In the new v4.3 out these days, I see from the release notes on the main web site, that COMSOL has added solar heat flux sources based on latitudes

Anyhow my advice, start simple.
With HT heat conduction, then split up heat fluxes and distribute the 1kW into the different material layers depending on their transmission.
Run separately your conjugated heat fluid flow => Solid exchange model, estimate the time for the copper jacket to stabilise the temperature all around the water flow, compared to flow velocities, as I suspect it is the only element that might distroys the validity of a 2D-axi pipe flow model, and 2D-axi solves so much quicker than 3D.
Also start i laminar conditions, then turbulent if applicable, depends on you flow speed, and will give a better heat exchange, but imply higher pressures and more energy to pump your water through. Later, based on the temperature you get, you can start to consider radiative exchange, that you can estimate and compare the power levels to the flux leves in your first model. In your case you have view factors close to 1, that is one calculation burden less

Nice project ;)

--
Have fun COMSOLING
Ivar

thanks Ivar.
How can I open this file ? ("model.xml")

www.ewp.rpi.edu/hartford/~fonteb/EP/Other/Models/Final%20Test%20Model%20Template/

Hi

you have one if not two variable naming conflicts. COMSOL is not reminding you about double use of variable names (internal COMSOl defined ones) as it allows you to freely change teir signigficance. This is nice , but when you hit a COMSOL varaible name and use it for another purpouse you end up with surprises.

I wish COMSOl could propose a naming conflict checker, so that we could run it over a model just to remin us with potential naming conflicts

In your case I expect "H" to be already used, therefore it flags an unit error in some of your BC settings. and "eps" is already used by COMSOL for the smallest integer different from 1 that can be resolved, its in the order of 1E-12 so the side effects of setting it to "6" can be uncontrollable ;)
Now I have redone your calculation with these names changed, without noticing major differences, but I cannot check everything

Try avoiding 1 character names, but I agree it's difficult to avoid naming collisions, there are so many COSMOL variables in there :)

A few suggestions:
- Not that it changes anything, but you could also set the initial temperature to T_amb, just to be systematic
- To make the T graphs easier to read, use rather T-T_amb so you see the temperature increase, else its quickly done to get confused in scale (and you can leave [K] as relative [K] = [degC]
- When you "Line plot along a edge COMSOL proposes the arc length, I often prefer the coordinates as you never know fully in which direction the "arc" is counted. Replace the X-axis expression in the Line graph node by "x" or "y" or whatever appropriate
- When you calculate the average temperature, you should also consider to normalise over the flow, as you have a non linear (parabolic type) flow profile and often much more water comes out "cold" in the centre. (with your low flow its not really "cold", but anyhow worth considering). This will probably change your cell efficiency value too
- Do not forget you are in 2D, so the HT part is calculating per meter (hence the [*/m] units) so this should be considered for the absolute values too

--
Good luck
Ivar

Hello,

I am Meera and I am a student, I would like to develop a laser heating of double layered materials. my materials are stationary and the layers movers in the meander shape. Can you please give me some clews regarding how to approach and develop such model.

Thanking you.

with regards;
Meera

Hello,

I am Meera and I am a student, I would like to develop a laser heating of double layered materials. my materials are stationary and the layers movers in the meander shape. Can you please give me some clews regarding how to approach and develop such model.

Thanking you.

with regards;
Meera