cfd simulation of solar water heater using cfx 13.0

avinash madhavan · December 20, 2011, 17:15

hi.
i want to simulate a thermosiphon solar water heater . i have attached the picture of a model.i have only considered a fluid domain.
since the model is being tested indoors i do not want to consider the radiation effect. so there is only natural convection and conduction.
the source of energy is a 1000 watt spot light. i have entered this as the source of heat flux in set up. is that right?
the only other available data is the mass flow rate at the inlet which is 0.01 kg/s. and the temperature of water at inlet is 20 degree celicis.
can anyone suggest how to proceed..
because i run the simulation and i wanted to get the temperature of water at the outlet. which i didn't get

ghorrocks · December 20, 2011, 17:19

I have moved this post to a new thread - please do not hijack other threads, start a new thread.

Have you read the FAQ?
http://www.cfd-online.com/Wiki/Ansys..._inaccurate.3F

avinash madhavan · December 20, 2011, 17:28

i am new to this site. i couldnt find how to post a new thread.
thanks

avinash madhavan · December 20, 2011, 17:32

if any one can help me with this query i would be very thankful.
i have been struggling with this problem for a while.
thanks

ghorrocks · December 20, 2011, 18:04

I already have replied - have you read the FAQ? Your question seems to be "Why is my simulation inaccurate?" and this has been asked so much that we have done a FAQ on it. I will not assist you in further details until you have covered the basics in the FAQ.

avinash madhavan · December 21, 2011, 07:15

hi ghorrocks.
i went through the FAQ's. i am new to the software . i presume my boundary conditions are right. but my mesh is not so fine. i have inserted inflation and edge sizing. is there any wat i can improve the mesh quality..

jbritton · December 21, 2011, 09:33

Your mesh isnt good enough, it needs to improved all over. Also your geometry is different from the original picture. The flow field will be different so you'll get different results.

avinash madhavan · December 21, 2011, 10:35

hi.
i am considering the fluid domain only..i have not taken into account the copper channel in which the fluid flow..
can yu give me any suggestions on how to improve the mesh as i am new to this software.. i tried inserting inflations on the boundaries..
any help will be appreciated
regards
avinash madhavan

ghorrocks · December 21, 2011, 17:05

Do the meshing tutorials for meshing skills.

You will need a MUCH finer mesh than this to get anything close to reality. Also you will need to move your inlets and outlets further away from the active region. And I reckon your assumption that you can model the fluid domain without the solid domain is probably wrong too.

avinash madhavan · December 21, 2011, 17:18

as i told you i am new to this software, using my knowledge i have tried to improve the mesh using edge sizing and inserting inflation. will that be ok to get a proper mesh or should i go for some other options.
thanks

ghorrocks · December 21, 2011, 17:39

The FAQ I pointed you to explains what you need to do.

If you are unsure whether your mesh (or anything else for that matter) is good enough, then do a sensitivity study. Do two meshes which are significantly different (eg 1mm average edge length compared to 0.5mm average edge length, in 3D this will be a mesh about 8 times bigger) and compare the results. Keep refining the mesh until the results converge to a tolerance you are happy with.

manjunathms · December 23, 2011, 04:26

The problem set up in CFD is not correct from whatever data u have supplied. If u are simulating natural convection flow then inlet mass flow should not inlfuence the flow of waterthrough the pipes. Watershould flow by itelf(buoyancy flow). In ur case i understant u have one inlet and one outlet...and such case inletmass flow pushes the waterto outlet. This becomes a forced circulation flow...even when u have small mass flow at inlet.

You will have to redesign ur experimental set up also .

avinash madhavan · January 4, 2012, 12:06

hi. i have tested the model both indoors and outdoors. i have got an outlet water temperature of 40 degrees when inlet water temperature was 20 degree.. i think my mesh is not so accurate. i have inserted body sizing face sizing and edge sizing. the total no of elements was 1260,550.. can any body help me on how i can improve my mesh as i have only a few boundary conditions to specify. i have used a buyoncy effect in this case.
thanks
happy new year to everyone

avinash madhavan · January 9, 2012, 11:20

is it possible to obtain the value of overall heat loss coefficient in cfd post .

ghorrocks · January 9, 2012, 17:11

Yes. You can integrate expressions or variables over surfaces and you can evaluate the overall heat loss from that.

December 21, 2011, 10:35	solar water heater simulation	#8
avinash madhavan New Member avinash madhavan Join Date: Oct 2011 Posts: 8 Rep Power: 14	hi. i am considering the fluid domain only..i have not taken into account the copper channel in which the fluid flow.. can yu give me any suggestions on how to improve the mesh as i am new to this software.. i tried inserting inflations on the boundaries.. any help will be appreciated regards avinash madhavan

December 21, 2011, 17:18	cfd simulation of solar water heater using cfx 13.0	#10
avinash madhavan New Member avinash madhavan Join Date: Oct 2011 Posts: 8 Rep Power: 14	as i told you i am new to this software, using my knowledge i have tried to improve the mesh using edge sizing and inserting inflation. will that be ok to get a proper mesh or should i go for some other options. thanks

January 4, 2012, 12:06	solar water heater	#13
avinash madhavan New Member avinash madhavan Join Date: Oct 2011 Posts: 8 Rep Power: 14	hi. i have tested the model both indoors and outdoors. i have got an outlet water temperature of 40 degrees when inlet water temperature was 20 degree.. i think my mesh is not so accurate. i have inserted body sizing face sizing and edge sizing. the total no of elements was 1260,550.. can any body help me on how i can improve my mesh as i have only a few boundary conditions to specify. i have used a buyoncy effect in this case. thanks happy new year to everyone

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Fan Blade Simulation. 2D CFD with FLUENT? CFX?	aerospain	ANSYS	0	December 3, 2010 06:08
[ICEM] Solar Water Heater Mesh Problem.	OzMantle	ANSYS Meshing & Geometry	17	July 27, 2010 19:14
CFX vs CFD for multiphase+particle	zeitoun	Main CFD Forum	1	June 5, 2010 21:35
Ansys CFX vs CFD for multiphase+particle	zeitoun	ANSYS	1	June 4, 2010 03:58
CFX 4 geometry file from CAD2Mesh or ICEM CFD 4	Hu	CFX	5	April 13, 2005 15:16

December 20, 2011, 17:19		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	I have moved this post to a new thread - please do not hijack other threads, start a new thread. Have you read the FAQ? http://www.cfd-online.com/Wiki/Ansys..._inaccurate.3F

December 20, 2011, 17:28		#3
avinash madhavan New Member avinash madhavan Join Date: Oct 2011 Posts: 8 Rep Power: 14	i am new to this site. i couldnt find how to post a new thread. thanks

December 20, 2011, 17:32		#4
avinash madhavan New Member avinash madhavan Join Date: Oct 2011 Posts: 8 Rep Power: 14	if any one can help me with this query i would be very thankful. i have been struggling with this problem for a while. thanks

December 20, 2011, 18:04		#5
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	I already have replied - have you read the FAQ? Your question seems to be "Why is my simulation inaccurate?" and this has been asked so much that we have done a FAQ on it. I will not assist you in further details until you have covered the basics in the FAQ.

December 21, 2011, 09:33		#7
jbritton Member james britton Join Date: Jan 2010 Posts: 38 Rep Power: 16	Your mesh isnt good enough, it needs to improved all over. Also your geometry is different from the original picture. The flow field will be different so you'll get different results.

December 21, 2011, 17:05		#9
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	Do the meshing tutorials for meshing skills. You will need a MUCH finer mesh than this to get anything close to reality. Also you will need to move your inlets and outlets further away from the active region. And I reckon your assumption that you can model the fluid domain without the solid domain is probably wrong too.

December 21, 2011, 17:39		#11
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	The FAQ I pointed you to explains what you need to do. If you are unsure whether your mesh (or anything else for that matter) is good enough, then do a sensitivity study. Do two meshes which are significantly different (eg 1mm average edge length compared to 0.5mm average edge length, in 3D this will be a mesh about 8 times bigger) and compare the results. Keep refining the mesh until the results converge to a tolerance you are happy with.

December 23, 2011, 04:26		#12
manjunathms New Member manjunath Join Date: Sep 2011 Posts: 1 Rep Power: 0	The problem set up in CFD is not correct from whatever data u have supplied. If u are simulating natural convection flow then inlet mass flow should not inlfuence the flow of waterthrough the pipes. Watershould flow by itelf(buoyancy flow). In ur case i understant u have one inlet and one outlet...and such case inletmass flow pushes the waterto outlet. This becomes a forced circulation flow...even when u have small mass flow at inlet. You will have to redesign ur experimental set up also .

January 9, 2012, 11:20		#14
avinash madhavan New Member avinash madhavan Join Date: Oct 2011 Posts: 8 Rep Power: 14	is it possible to obtain the value of overall heat loss coefficient in cfd post .

January 9, 2012, 17:11		#15
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	Yes. You can integrate expressions or variables over surfaces and you can evaluate the overall heat loss from that.