CFX Coronary Flow Simulation

ld1305 · February 1, 2012, 12:52

Hi, I have a coronary artery that I am running through CFX to try and find the pressure at certain points, mainly before and after a blockage, but elsewhere would be useful too.

I'm having a couple of problems, though:

1. I can't seem to make the flow go through all the outlets, there is one inlet, and 4 outlets and just by using a quick velocity streamline it's obvious the flow only goes one way.

2. I can't seem to find a way to find the pressure at given locations anyway

3. The main problem: I'm not too sure how to make the flow pulse, as would be natural in the human body, I'd need it to be every 0.75 seconds, though not too sure otherwise. I also have mass flow rate and properties of blood etc if that's any use

Thanks very much,
Lauren

ld1305 · February 1, 2012, 12:53

Some printscreens of the crude mesh (it's to be improved later) and some other things I've managed are here

ghorrocks · February 1, 2012, 18:25

Your questions:
1) Then you boundary conditions or simulation is wrong. Are you sure the pressures are correct? I suspect some of the outlets really do have back flow at some stages of the pumping cycle.

2) The easiest way is to colour the outer wall by pressure in CFD-Post. Or you can put points and get the pressure at those points. In the solver you can define monitor points and get the pressure during simulation.

3) This is simple to implement, use a CEL expression (or interpolation function) to define the mass flow rate versus time.

Also I suspect you will find the deformation of the vessels significantly affects the results.

FrankW · February 2, 2012, 03:30

Hi Lauren,

is your model scale right? I can see a ruler with a lenght of 20-30 meters on 2 pictures.

Frank

ghorrocks · February 2, 2012, 05:43

Good point. Is this a scan of the arteries from Godzilla?

ld1305 · February 2, 2012, 10:48

Yeah... I need to change the scale, I'll sort that out afterwards. For now I have it to go through all the different parts of the artery though, just need to work on the pulsatile flow a little, though this seems to be the trickier bit to be honest :/

ld1305 · February 2, 2012, 10:50

As it is I can#t comment on the deformation of the vessels as I have no way to accurately predict their movement. This has been done before using a different imaging technique, but I don't have access to more advanced images yet

ghorrocks · February 2, 2012, 17:26

You should get the scale right straight away. It will change the Re number of the flow, and separations and lots of other flow features. Therefore the resistance balance will be totally different and hence the flow split will be totally different.

It is easy to do and you do not need to remesh. Just do a mesh transform using a scaling function in CFX-Pre with your existing mesh.

ld1305 · February 4, 2012, 14:14

I've changed the scale, but the major problems I'm having now are with finding the correct values for the blood, which is obviously causing issues with getting the correct results... don't happen to know anywhere do you? I have most of the basic numbers, just a few more needed for the analysis

ghorrocks · February 5, 2012, 18:19

There is lots of papers in the literature on this, do a google search.

ld1305 · February 5, 2012, 19:14

I have done and can't fine many of the numbers I'm looking for, mostly with regards to how to set it up in CFX as a material, I can't model it as a pure substance (because it isn't) and because there is no molar mass, and the other ways don't seem much help either...

ghorrocks · February 6, 2012, 20:14

Blood rheology is a whole field of research in itself. It sounds like you better do some research on it before proceeding - as you point out blood is a very complex fluid and there is many different levels of simplification you can use to model it.

ghorrocks · February 6, 2012, 20:14

Blood rheology is a whole field of research in itself. It sounds like you better do some research on it before proceeding - as you point out blood is a very complex fluid and there is many different levels of simplification you can use to model it.

ld1305 · February 7, 2012, 04:17

I know I don't need any of the thermal properties of blood, only those which could effect pressure and mass flow rate, I don't know that the shear stresses etc are even important to be honest.

I'm going to go and see a tutor about it later and see if he can help, but thanks

February 1, 2012, 12:52	CFX Coronary Flow Simulation	#1
ld1305 Member Lauren Dransfield Join Date: Jan 2012 Posts: 50 Rep Power: 14	Hi, I have a coronary artery that I am running through CFX to try and find the pressure at certain points, mainly before and after a blockage, but elsewhere would be useful too. I'm having a couple of problems, though: 1. I can't seem to make the flow go through all the outlets, there is one inlet, and 4 outlets and just by using a quick velocity streamline it's obvious the flow only goes one way. 2. I can't seem to find a way to find the pressure at given locations anyway 3. The main problem: I'm not too sure how to make the flow pulse, as would be natural in the human body, I'd need it to be every 0.75 seconds, though not too sure otherwise. I also have mass flow rate and properties of blood etc if that's any use Thanks very much, Lauren

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Domain format problem on airfoil flow simulation	andrenonaka	CFX	14	December 7, 2015 01:42
Flow simulation in reverse drum mixer	Patel Chirag	CFX	0	July 19, 2010 16:05
CFX steady simulation	gharek	CFX	1	April 7, 2010 19:41
how to set up high temperature gas turbine flow simulation?	adam2008	CFX	1	July 22, 2009 19:33
Simulation of Flow through Complex 3D Geometry	EmersonKB	CFX	5	July 2, 2009 09:17

February 1, 2012, 12:53		#2
ld1305 Member Lauren Dransfield Join Date: Jan 2012 Posts: 50 Rep Power: 14	Some printscreens of the crude mesh (it's to be improved later) and some other things I've managed are here

February 1, 2012, 18:25		#3
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	Your questions: 1) Then you boundary conditions or simulation is wrong. Are you sure the pressures are correct? I suspect some of the outlets really do have back flow at some stages of the pumping cycle. 2) The easiest way is to colour the outer wall by pressure in CFD-Post. Or you can put points and get the pressure at those points. In the solver you can define monitor points and get the pressure during simulation. 3) This is simple to implement, use a CEL expression (or interpolation function) to define the mass flow rate versus time. Also I suspect you will find the deformation of the vessels significantly affects the results.

February 2, 2012, 03:30		#4
FrankW Member Frank Weise Join Date: Mar 2009 Location: Germany Posts: 55 Rep Power: 17	Hi Lauren, is your model scale right? I can see a ruler with a lenght of 20-30 meters on 2 pictures. Frank

February 2, 2012, 05:43		#5
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	Good point. Is this a scan of the arteries from Godzilla?

February 2, 2012, 10:48		#6
ld1305 Member Lauren Dransfield Join Date: Jan 2012 Posts: 50 Rep Power: 14	Yeah... I need to change the scale, I'll sort that out afterwards. For now I have it to go through all the different parts of the artery though, just need to work on the pulsatile flow a little, though this seems to be the trickier bit to be honest :/

February 2, 2012, 10:50		#7
ld1305 Member Lauren Dransfield Join Date: Jan 2012 Posts: 50 Rep Power: 14	As it is I can#t comment on the deformation of the vessels as I have no way to accurately predict their movement. This has been done before using a different imaging technique, but I don't have access to more advanced images yet

February 2, 2012, 17:26		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	You should get the scale right straight away. It will change the Re number of the flow, and separations and lots of other flow features. Therefore the resistance balance will be totally different and hence the flow split will be totally different. It is easy to do and you do not need to remesh. Just do a mesh transform using a scaling function in CFX-Pre with your existing mesh.

February 4, 2012, 14:14		#9
ld1305 Member Lauren Dransfield Join Date: Jan 2012 Posts: 50 Rep Power: 14	I've changed the scale, but the major problems I'm having now are with finding the correct values for the blood, which is obviously causing issues with getting the correct results... don't happen to know anywhere do you? I have most of the basic numbers, just a few more needed for the analysis

February 5, 2012, 18:19		#10
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	There is lots of papers in the literature on this, do a google search.

February 5, 2012, 19:14		#11
ld1305 Member Lauren Dransfield Join Date: Jan 2012 Posts: 50 Rep Power: 14	I have done and can't fine many of the numbers I'm looking for, mostly with regards to how to set it up in CFX as a material, I can't model it as a pure substance (because it isn't) and because there is no molar mass, and the other ways don't seem much help either...

February 6, 2012, 20:14		#12
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	Blood rheology is a whole field of research in itself. It sounds like you better do some research on it before proceeding - as you point out blood is a very complex fluid and there is many different levels of simplification you can use to model it.

February 7, 2012, 04:17		#14
ld1305 Member Lauren Dransfield Join Date: Jan 2012 Posts: 50 Rep Power: 14	I know I don't need any of the thermal properties of blood, only those which could effect pressure and mass flow rate, I don't know that the shear stresses etc are even important to be honest. I'm going to go and see a tutor about it later and see if he can help, but thanks