Discussion Forum

Hi all,

I am new to Comsol, and I think I have pieced together a somewhat functional model. However, I have a few points that I am stuck on and looking for some general feedback on my approach.

I am simulating heat exchange in 3 concentric pipes across 500m of vertical flow. Cold water enters on the surface and flows down an outer channel, which is formed by the outer surface of the middle pipe and the inner surface of the outer pipe - representing perfect insulation for now. Hot water enters at maximum depth and flows up an inner channel, which is formed by the inner surface of the inner pipe. A 3rd "heat exchange" channel is formed by the outer surface of the inner pipe, and inner surface of the middle pipe, and is a sealed volume that contains various fluids - air and water for now. Attached is an image that shows what I am trying to model.

Basic setup

• 2D array with rectangle geometry to represent the inner channel, inner pipe, heat exchange channel, and outer channel.

• The 0,0 point represents the axial symmetry at maximum depth (lower boundaries)

• Stationary laminar conjugative heat transfer

• Added water, air, and "structural steel" from materials library

Heat Transfer in solids and fluids is across all domains (ht)

• Solid domain 1 selected to be the inner and middle pipe domains, material is “structural steel”

• Fluid domain 1 was selected to be the inner and outer channel domains, material is water

• Inflow 1 - top boundary of outer channel with a predefined "cold temperature" (5 C)

• Outflow 1 - bottom boundary of outer channel, and calculating this temperature is the goal

• Inflow 2 - bottom boundary of inner channel with predefined "hot temperature" (90 C)

• Outflow 2 - top boundary of inner channel, and calculating this temperature is the goal

• Fluid domain 2 added for the material in the heat exchange channel, and temperature input is assigned to radiation heat transfer in medium, and velocity and pressure assigned to 0 (more on that later)

• Insulation 2 assigned to the top and bottom boundary of heat exchange channel

• Laminar flow (spf)

• Assigned to inner and outer channel domains to model water flow

• Incompressible flow with gravity

• Assigned reference point to be r = 0 and height “L” representing the top boundaries

• Channel inlets with the same boundaries as heat transfer, with inputs being fully developed flow at 0.1 m/s and outputs being 0 pressure with hydrostatic pressure compensate, normal flow and suppress backflow checked

Added the “heat transfer with radiation in participating media” (rpm) physics.

• Only the HX channel domain was selected

• In participating medium, used the htrpm1 multi-physics as the temperature input.

• Assigned a 0.0001 absorption coeff. And a 0.00001 scattering coeff. for air

• Opaque surface is the boundary of heat exchange channel where I assumed the “structural steel” has a grey wall, with a surface emissivity of 0.8

• In multiphysics, coupled the heat transfer in solids/fluids (ht) and radiation in participating media (rpm)

Running that simulation gives around 15 C of heat transfer between the respective channels, which seems reasonable. Given this narrow channel, and high degree of temperature differences between the inner and middle pipe, how do I model the internal natural convection in the heat exchange channel? Looking at all the different tutorials and example models I attempted this several ways with no success. Can you all give me guidance on this?

1. Has the internal natural convection already been modeled? Some reading suggested it was, but the velocity in that channel had to be manually set to 0, otherwise it wouldn’t solve…
2. Adding a surface-to-ambient radiation boundary in the solids/fluids (ht) physics.

• There is nothing under the model input and no ability to edit it.

• Under Radiation Settings, what the upside and downside radiation direction? The heat is flowing from “left to right” on the axial cross section…

• How do you handle ambient temperature? Shouldn’t that be an input? Inputting different temperatures shifts the simulation significantly.

• Adding an internal natural convection heat flux

3. I can only select the outer most boundaries, rather than the inner/middle pipe surface

• Do I delete the secondary fluid in for the heat exchange channel in the solids/fluids (ht) physics and set up a separate heat transfer in fluids physics for the heat exchange channel?

• Should I include a surface-to-ambient radiation boundary as well?

• All of them have limited/no model inputs and require external and/or ambient temperature constants. But shouldn’t those be model inputs, rather than user defined constants?

4. Adding a separate laminar flow physics for the heat exchange channel

• Include gravity and defined at reference of r=D2/2 and z=L

• What flow should be used? Since it is air, selected compressible, but weakly and non didn’t work either

• Assigned a pressure point at r=D2/2 and Z=L

• Reassigned the model inputs for heat exchange channel in the solids/fluids (ht) to be: temperature = htrpm1, velocity = spf2, pressure = spf2.

• Never got to solve, and always returned several errors about undefined variables, or void equations right after clicking solve.

• Once this is perfected for air, what changes need to be made to model the heat exchange channel for water? I am guessing I disable the radiation interfaces, but I imagine there are more complexities…

Sorry for the long post and if you made it this far thanks for taking the effort. Hopefully, I have described my problem well enough to for valuable feedback. Any suggestions on how to get convective heat in the heat exchange channel working? Any thoughts on my approach or other options that I should investigate for the system in general?