Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Maximum iterations reached
Posted Mar 11, 2011, 3:27 a.m. EST Fluid & Heat, Heat Transfer & Phase Change, Studies & Solvers Version 4.0a 3 Replies
Please login with a confirmed email address before reporting spam
Hello,
I've searched everywhere online and in the helpfunctions of comsol, but i can't solve my problems with Comsol.
I have a very simple model of a 20*20*20cm concrete block with a tube of 40mm and 5mm walls through the center of the block. And I've created a laminair water flow through the pipe and heat exchange of the pipe and the block. On top of the block there is a Heatflux of 1000W/m2.
I need to know how much heat is absorped by the waterflow. There are 2 problems.
- The surface integration of the inlet and outlet (multiplyed with surface for getting K) delivers value's beneath the origin temperature?
- The model only solves when;
> Iterative > Use below error level > Is set higher then 2, when i draw a plate (lamel) on top of the tube it can only solve with a number higer then 50. Maybe that is the reason that the temperature doesnt seem right? ( to high error level?)
What am i doing wrong? Because it is such a simple problem, comsol should be able to solve this right?
Thanks in advance!
Frank
PS. the error seems to apply on stationairy solver 1 ( only solves laminair flow ), the stationairy solver 2 does the heat transfer..
I've searched everywhere online and in the helpfunctions of comsol, but i can't solve my problems with Comsol.
I have a very simple model of a 20*20*20cm concrete block with a tube of 40mm and 5mm walls through the center of the block. And I've created a laminair water flow through the pipe and heat exchange of the pipe and the block. On top of the block there is a Heatflux of 1000W/m2.
I need to know how much heat is absorped by the waterflow. There are 2 problems.
- The surface integration of the inlet and outlet (multiplyed with surface for getting K) delivers value's beneath the origin temperature?
- The model only solves when;
> Iterative > Use below error level > Is set higher then 2, when i draw a plate (lamel) on top of the tube it can only solve with a number higer then 50. Maybe that is the reason that the temperature doesnt seem right? ( to high error level?)
What am i doing wrong? Because it is such a simple problem, comsol should be able to solve this right?
Thanks in advance!
Frank
PS. the error seems to apply on stationairy solver 1 ( only solves laminair flow ), the stationairy solver 2 does the heat transfer..
3 Replies Last Post Mar 11, 2011, 9:39 a.m. EST
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
are you sure it's so "simple" ?
Well you say you use two separate stationary solvers, but do you not have a coupling temperature to fluid properties too ?
My way is always to separate the problems, first solve the HT for its side with some heat loss on the interior tubes, then solve the laminar flow alone, ensuring you have some velocity profile and pressure drop (i.e. Poiseuille) as initial conditions to get the conversion faster. then couple both.
Check the materials characteristics if these are not also temperature dependent (start without always safer), you might use one stationary solver but se manually the segregation solver to perform in sequence often better
Are you sure you have all fluid properties, conduction, transport etc ? Check by performing arrow plots for flow and surface integration for energy fluxes
--
Good luck
Ivar
are you sure it's so "simple" ?
Well you say you use two separate stationary solvers, but do you not have a coupling temperature to fluid properties too ?
My way is always to separate the problems, first solve the HT for its side with some heat loss on the interior tubes, then solve the laminar flow alone, ensuring you have some velocity profile and pressure drop (i.e. Poiseuille) as initial conditions to get the conversion faster. then couple both.
Check the materials characteristics if these are not also temperature dependent (start without always safer), you might use one stationary solver but se manually the segregation solver to perform in sequence often better
Are you sure you have all fluid properties, conduction, transport etc ? Check by performing arrow plots for flow and surface integration for energy fluxes
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Thanks for the fast and good reply! :)
I tried ur way, and solved them sepperatly, both solves worked.. But when i combine them the solve gets the iterations error.
Here my settings, maybe u can detect the error;
Laminair flow ( incompressible flow)
Fluidproperties
Wall (slip)
Initial values (velocity y=1m/s, pressure 2e5Pa) That's what u recommended right?
Inlet (Velocity>Normal inflow velocity>1m/s)
Outlet(same as inlet)
Symmetry (defined as the walls of the innertubing)
Heat transfer (heat transfer in solids)
Heat transfer in solids
Thermal insulation
Initiial values
Inflow Heat flux (1000W/m2 on the surface)
Heat transfer in fluids (defined the inner volume of the tube, Absolute pressure = fp1, velocity =fp1)
Temperature (inlet tubing defined 290K)
That's all i have..
I also do not completly understand the part of the "material characteristics " and u say one solver? sow both laminair flow and heattransfer in 1 solver? and the segregation solver were can i change that?
I tried surface integrating for flow, and that also seems of.. 1m/s flow through a 40mm pipe should give 0.001256637m3/s and the surfaceintegration gives 0.001176m3/s? It's a bit same story as calculating the temperature..
sorry for the lot of question, im now 3 weeks working with comsol and matlab non seems to de the trick haha almost wanted to go for ansys, but hope u can help me..
I tried ur way, and solved them sepperatly, both solves worked.. But when i combine them the solve gets the iterations error.
Here my settings, maybe u can detect the error;
Laminair flow ( incompressible flow)
Fluidproperties
Wall (slip)
Initial values (velocity y=1m/s, pressure 2e5Pa) That's what u recommended right?
Inlet (Velocity>Normal inflow velocity>1m/s)
Outlet(same as inlet)
Symmetry (defined as the walls of the innertubing)
Heat transfer (heat transfer in solids)
Heat transfer in solids
Thermal insulation
Initiial values
Inflow Heat flux (1000W/m2 on the surface)
Heat transfer in fluids (defined the inner volume of the tube, Absolute pressure = fp1, velocity =fp1)
Temperature (inlet tubing defined 290K)
That's all i have..
I also do not completly understand the part of the "material characteristics " and u say one solver? sow both laminair flow and heattransfer in 1 solver? and the segregation solver were can i change that?
I tried surface integrating for flow, and that also seems of.. 1m/s flow through a 40mm pipe should give 0.001256637m3/s and the surfaceintegration gives 0.001176m3/s? It's a bit same story as calculating the temperature..
sorry for the lot of question, im now 3 weeks working with comsol and matlab non seems to de the trick haha almost wanted to go for ansys, but hope u can help me..
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
Not obvious to reply simply.
1) I cannot recommend any particular absolute initial values as these are model dependent, but my experience makes me ALWAYS add a parabolic velocity profile for the inlet and all along the tube. As well as I add some pressure drop, i.e.e a straight simple Poiseuille calculation, I gain quite some time for the solver like that. Then once solved I sometimes adapt the values to get closer to my first calculation, so that on next try I loose less time
For the pressure one should decide on gauge pressure i.e. 0[Pa] + pa=1[atm] or an absolute pressure, depends on the models.
If you inlet is pressure driven, then leave the outlet free (w.r.t velocity) with an outlet pressure
But if your flow solves nicely it should be more or less OK
Then your if solid solve alone, not much to add
Then when you couple, check that you have some conduction from the solid wall temperature to the fluid, as well as some transport of the temperature in the fluid via the velocity
For the material properties: check how they look, you can have constants for the fluid parameters, but often the one from the COMSOL DB are temperature and / or pressure dependent, this loops the solution and makes it more non-linear.
Check on the Forum, there are several nice examples
--
Good luck
Ivar
Not obvious to reply simply.
1) I cannot recommend any particular absolute initial values as these are model dependent, but my experience makes me ALWAYS add a parabolic velocity profile for the inlet and all along the tube. As well as I add some pressure drop, i.e.e a straight simple Poiseuille calculation, I gain quite some time for the solver like that. Then once solved I sometimes adapt the values to get closer to my first calculation, so that on next try I loose less time
For the pressure one should decide on gauge pressure i.e. 0[Pa] + pa=1[atm] or an absolute pressure, depends on the models.
If you inlet is pressure driven, then leave the outlet free (w.r.t velocity) with an outlet pressure
But if your flow solves nicely it should be more or less OK
Then your if solid solve alone, not much to add
Then when you couple, check that you have some conduction from the solid wall temperature to the fluid, as well as some transport of the temperature in the fluid via the velocity
For the material properties: check how they look, you can have constants for the fluid parameters, but often the one from the COMSOL DB are temperature and / or pressure dependent, this loops the solution and makes it more non-linear.
Check on the Forum, there are several nice examples
--
Good luck
Ivar