Discussion Forum

Posted: 1 decade ago Nov 7, 2011, 8:56 a.m. EST

Hi, You set 573.15K as the inflow temperature but physical properties of the engine oil are determined in the range 273

Posted: 1 decade ago Nov 7, 2011, 9:08 a.m. EST

Hi, OK!!But im trying with other fluids and i cant achieve to see the heat losses in the fluid,i mean,the temperature at the entrance should not be the same that at the exit of the pipe.I included another "INITIAL VALUES"(velocity and a temperature of 573 K ) for the fluid in the pipe,an INLET,OUTLET,OUTFLOW and TEMPERATURE of 573 K at the INLET. I want the fluid enters at 573K,but then it should loose heat,transferred to the rest.I dont know if im doing right!! If you have any advice or suggestion,i will be grateful!!1 Thanks!!!

Posted: 1 decade ago Nov 7, 2011, 9:58 a.m. EST

This is what i have,but the temperature nside the pipe is the same! Please,help!!

Attachments:

• Prueba_Conjugate HT.mph

Posted: 1 decade ago Nov 7, 2011, 10:42 a.m. EST

Hello! I think you should look at the model "free_convection" that you find in the model library under Comsol Multiphysics - Multiphysics - free convection. Read the model pdf. If you are working in 3D the boundaries becomes surfaces of course. I hope it helps, Good luck! /Rasmus

Posted: 1 decade ago Nov 7, 2011, 11:15 a.m. EST

Hi Your fluid is a "initial value 2" already defined at max temp. Try to use 300K as starting temperature then you should see the fluid temperature change, provided you give a step shorter than 50 sec, i.e. start with 0 1 5 20 range(50,50,1000) You should also add a boundary layer on the fluid tube wall, or leave COMSOL define the default fluid mesh by its own (physics controlled meshing) I would also propose to add a laminar inflow instead of your flat profile, or define an initial velocity and pressure drop via Poiseille and a parabolic velocity profile with a boundary "s" parameter". It has been discussed on several threads already, as well as to be found in several of the model libbrary examples. Conjugated heat transfer is complex physics, with fluid, heat and transport, often it's easier to start step by step, one physics at the time. But now you model looks quite nice, no ? -- Good luck Ivar

Posted: 1 decade ago Nov 7, 2011, 11:31 a.m. EST

Hi, You just need to set up time-varying functions for both inlet conditions, flow and temperature; then your model works. You also need to make sure your mesh is refined enough. Cheers

Attachments:

• Prueba_Conjugate HT AF.mph

Posted: 1 decade ago Nov 7, 2011, 12:22 p.m. EST

Hi Ivar, Firstly,thanks for your response! I have defined the "INITIAL VALUE 2" at about 573 K,but it will affect just to the simulation time,wont it??I mean,it will affect the time the solar salt takes to reach the required temperature,or (at the same simulation time) at the temperature reached by it.I dont really know.I will try,you know more than me. Anyway,the temperature by the fluid will be the same in all the pipe.I just want the fluid to enter at 573K(it will always enter at this temperature) and transfer the heat to the rest,so it will have heat losses and his temperature should be different at the end of the pipe,isnt it?,but i dont succeed to put this in COMSOL. I will leave Comsol defines the default fluid mesh by its own. If i achieve this,my model will look nice.This is just a previous step to another simulation.I simulated something similar in 3D,but with no fluid,i just put a TEMPERATURE boundary condition in the inner wall of the pipe to remain at 573K,but this is not correct! Sorry,i hope you understand me and you can help me!! Thank you for your time,i really appreciate it and im very grateful!! Regards, Carlos

Posted: 1 decade ago Nov 7, 2011, 12:29 p.m. EST

Hi Amir, Thank for your response!! How can i set up the time-varying function for the inlet conditions??? Idont have problems with the flow.The temperature is my problem.I just want the fluid to enter at 573K ,see how the heat is transferred to the rest,so his temperature changes at the end. The heat transferred at the beginning should not be the same that the one transferred at the end because the temperature of the fluid should be different,so the temperature of the solar salt also. I hope you can help me! Thank you for your time,im really grateful!! Regard, Carlos

Posted: 1 decade ago Nov 7, 2011, 3:06 p.m. EST

Hi I believe I understand you. Now if you have an initial condition for your fluid of constant temperature, and an inlet temperature constant you will after some time get a uniform temperature all over, certainly the fluid will stay at the same "white" maximum temperature. What you can do is to estimate the flux inlet and outlet, take the difference and hence see the energy you are transferring, over time. Integrate the flux over these boundaries, and multiply by your simulation depth value to get the volume out. You can do it in postprocessing, but often I find it easier to define a few operators and a few specific variables to call them directly in the postprocessing section (derived values and tables) Pls read carefully the KB articlres such as (for heat flux calculations) http://www.comsol.com/support/knowledgebase/973/ http://www.comsol.com/support/knowledgebase/1031/ The heat transfer pipe to fluid passes via the boundary layer with noslip conditions, this means that you have no or little tranport only conduction over a thin layer, to resolve this correctly use the mesh boundary layers (or the deault 4.2a meshing that now adds them for you, zoom in on hte pipe boundary). You should also check your diffusivity "alpha = k/Cp/rho", thermal time step values "Dt" and mesh size "h" there are some relations to respect here too, as for RF or ACDC, and even structural (search the Forum for the comments of Nagi about mesh sizes related to time dependent solving -- Good luck Ivar

Posted: 1 decade ago Nov 7, 2011, 4:39 p.m. EST

Hi, I dont know if i have understood you!I think you are saying to me to try by putting heat flux at the enter of the pipe,but how can i determine it?You talk about postprocessing,but im a little bit lost,sorry,i have just started with this program. I have tried to open the articles but i cant. I will try tomorrow with the heat flux! Thanks you for your time!! Regards, Carlos

Posted: 1 decade ago Nov 7, 2011, 5:19 p.m. EST

Hi, If you go right-click on the Global Definition section a menu will appear, if you move the cursor over function a list of functions will appear, you may use the step function in order to have funtion that varies with time from 0 to 1, you can multiply that function by the value of your inlet velocity and temperature velocity and then solve. You may want to have a look at the model I posted, you will find the step function and you will see how I changed the boundary conditions for inlet and temperature. Cheers

Posted: 1 decade ago Nov 8, 2011, 1:56 a.m. EST

Hi bu using a constant temperature on the inlet, you are giving a heat source input, its such that the temperature is constant, not define in W, thats all. but knowing the flux (velocity) and heat capacity + density you can calculate the total heat entering, the same on the output. As the rest is isolated the heat extracted is the difference thanks to the energy conservation theoreme. But flux calculations in FREM are slightly tricky and one should take some care, to ensure correct results, thats all -- Good luck Ivar

Posted: 1 decade ago Nov 8, 2011, 4:20 a.m. EST

Hi Amir, I have tried to open your model but you are working with a previous version and i cant open it. I have found the step function,but i dont know what to do next!You say me to multiply it by the inlet velocity and by the temperature.What am i supposed to do: in boundary conditions of INLET and TEMPERATURE put " velocity *step1" and " temperature*step1 "??? Sorry,im dont know too much about this program,please give some hints!! Thank you for your time!!!!! Regards, Carlos

Posted: 1 decade ago Nov 8, 2011, 5:24 a.m. EST

Hi, I am using the latest version, 4.2a, you may want to upgrade to it. First you need to define the function 'step1' in the Global Definition section (please see the manual) then you need to write in the inlet conditions something like: v_in*step1(t) T_in*step1(t) Cheers

Posted: 1 decade ago Nov 8, 2011, 6:21 a.m. EST

Hi Amir, I dont succeed.Anyway,by doing this,i will impose the temperature of the fluid at each part of the pipe,wont i?but,what i want is to see what temperature does the fluid has a each part of the pipe,not imposed by me,due to the heat losses it is having along the pipe! Do you think that if i finally do what you say,i will achieve that goal??? Thank you very much for your time,really!! Best regards, Carlos

Posted: 1 decade ago Nov 8, 2011, 7:57 a.m. EST

Hi, To impose a Temperature at the inlet is not such a bad idea in the beginning: if you have a heated flow coming in, then you can assume that the temperature AT THE INLET is UNIFORMLY and HOMOGENOUSLY distributed, that is constant across the pipe section. Of course as the fluid enters your device the software will calculate how much heat is transferred and therefore how the temperature is re-distributed. Please note that the liquid is flowing FROM the inlet INTO the device, therefore the chances of heat being transferred backwards FROM the device into the pipe BEFORE the inlet is quite small. Cheers

Posted: 1 decade ago Nov 8, 2011, 8:24 a.m. EST

Hi, Yes,i though i was doing that,imposing a temperature at the entrance and supposing that the sofware would calculate the heat losses in the fluid and the heat transfer in the device. So i understand that if i do what you are telling me the sofware will calculate how much heat is transferred and therefore how the temperature is re-distributed (ALSO IN THE FLUID) I have being trying to define the step function,but i dont really understand what to write,and when i put in the TEMPERATURE BOUNDARY,it doesnt allow me!Could you send me an image of your model to see how you modified it???i want to achieve this1 Amir,sorry for not understanding,i dont want to be a charge,i need this thing to my project and im new with this program Thanks for your time,im really grateful!!! Best regards, Carlos

Posted: 1 decade ago Nov 8, 2011, 9:40 a.m. EST

Hi, You really need to have a look at the manual to understand the functions section. In the meanwhile you can try to have a look at the screenshots of the file I uploaded before. Cheers

Attachments:

• 1.png
• 2.png

Posted: 1 decade ago Nov 8, 2011, 11:11 a.m. EST

Hi Amir, Thank you very much for the images!!!!!!!!!! Of course im still doing,trying to understand this function.Im also looking in this forum about that,but the more similar examples,they have 2 temperatures,i mean,they want to start at 300K and reach 500K after 20 seconds,for example,or something like that. I just have 1 temperature,i dont know the other one at the final(i want the program to calculate it),so....that is why you are telling me to apply step function also in velocity???

Posted: 1 decade ago Nov 8, 2011, 11:30 a.m. EST

Hi, Let us put it this way: have you ever heard of a car able to speed up from 0 to 100 km/hr in zero seconds? No? Well that is because it is not possible, it is not physical. Exactly in the same way it is not possible to have a fluid going from zero to a finite velocity in zero seconds, therefore you need to describe the inlet velocity going up from zero to the desired value. Cheers

Posted: 1 decade ago Nov 8, 2011, 12:30 p.m. EST

Hi, So,with the step function the temperature and the velocity will vary from zero to the desirable value in a period of time(the value i put in the Smoothing section: Size of the Transition zone) and after that they will change but not affected by this function,loosing temperature due to the heat losses.¿? I will tell you tomorrow if its works! Thank so much Amir,you helping me a lot!! Best regards

Posted: 1 decade ago Nov 13, 2011, 12:01 p.m. EST

attached find ur model..i just spent couple of mins..i have redefined the entire module..instead of conjugate heat transfer module you can just use non-isothermal flow module.. cheers

Posted: 1 decade ago Nov 13, 2011, 12:09 p.m. EST

I mean non-isothermal module