CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Modeling blood in CFX (http://www.cfd-online.com/Forums/cfx/80152-modeling-blood-cfx.html)

davek September 16, 2010 15:54

Modeling blood in CFX
 
I'm trying to model blood in CFX but I'm not sure what to do.

I tried defining it as a pure substance, but it wants a molar mass (blood of course is a mixture and does not have a molar mass). I don't know the proper amounts of components for blood to make a fixed composition mixture, and if I did I'm not sure it would end up acting as a non-newtonian fluid correctly anyway.

Has anyone else modeled blood before and have any suggestions? I'm also relatively new to CFX and the rest of Ansys so I could be missing something big. Thanks!

ghorrocks September 16, 2010 18:04

Looks like you need to do LOTS of reading. Blood is not even close to a pure substance - it is a mixture of plasma, red and white blood cells and lots of other stuff. Does not sound pure at all to me. There is extensive literature on the fluid properties of blood, go to your uni library and start reading.

davek September 17, 2010 09:04

Yes I know it's a mixture, as I stated. And yes that's a ton of reading.

I was asking if anyone who may have already done all that reading could offer me some shortcuts on how to model it, and if I make a mixture material will it even act correctly or am I better off making an estimation with a pure substance?

andyj September 17, 2010 18:22

Hello!
Here is a link to a Journal Article , the references may shed some light:www.cfd.com.au/cfd_conf09/PDFs/144BUI.pdf
Also check out the Univ of Utah Scientific Computing Center, they are doing a lot of work on biological (anatomical) visualization and modeling :http://www.sci.utah.edu/software.html
Here is a 212 pg thesis, color illustrated about cfd modeling of bloodflow. You would have to download and translate it from spanish to english, or use google translate , then save it as an english document:http://repositorium.sdum.uminho.pt/b...neiro_2009.pdf
good luck!
andy

andyj September 17, 2010 19:05

Here is a 41 pg article from Stanford on modeling Bloodflow:
http://med.stanford.edu/school/vascular/Pdf/FEM-Ath.pdf
And a 68pg color thesis on bloodflow modeling:
http://www.stanford.edu/~anaiman/Pub...nal_Report.pdf
Here is a color illustrated dissertation article on CFD modeling of blood flow. 40-50 pgs:
http://www2.cs.uh.edu/~hadri/papers/...tation_phd.pdf

These should give you a start and some insight on the problem and what general direction to go and what is involved.
good luck!
andy

andyj September 17, 2010 19:12

Here is a 300 page article on the physical properties of blood and modeling it with respect to fluid dynamics
http://iopscience.iop.org/0034-4885/...C5041061DC3.c1

ghorrocks September 17, 2010 21:33

Looks like Andy has been very nice to you and given you a good start to your literature review.

Quote:

I was asking if anyone who may have already done all that reading could offer me some shortcuts on how to model it, and if I make a mixture material will it even act correctly or am I better off making an estimation with a pure substance?
These questions have to be asked in context. What are you trying to model and what do you want to get out of it? For some models blood can be assumed Newtonian and pure, for some non-newtonian, for some the particles will need to be accounted for. What is important depends on what you are doing.

zmaxcomputer September 17, 2010 21:35

to model blood, just make a new substance/fluid and enter these values:
3.5cp = viscosity
density = 1.056grams/cm3

the toshiba stroke group at SUNY Buffalo uses these values for all of their simulations. Thats it!

davek September 20, 2010 08:56

Quote:

Originally Posted by ghorrocks (Post 275659)
Looks like Andy has been very nice to you and given you a good start to your literature review.

These questions have to be asked in context. What are you trying to model and what do you want to get out of it? For some models blood can be assumed Newtonian and pure, for some non-newtonian, for some the particles will need to be accounted for. What is important depends on what you are doing.

It's for a flowcell in which blood will be sandwiched between two layers of saline water. My boss stressed modeling blood as non-Newtonion, but particles like red and white blood cells don't need to be accounted for.

Thank you andyj.

zmaxcomputer, I would still would need a value for molar mass. Maybe that's all you need in fluent, I'm not sure, but I'm using cfx. Thank you though.

Lance September 20, 2010 10:30

Have you considered the Carreau-Yasuda model? It's strain rate-dependent and have model constants that include the effects of fluid rheology (in some sense).

I don't have the expression or constants at the moment, but do a web search and you'll find it.

davek September 20, 2010 12:21

Quote:

Originally Posted by Lance (Post 275833)
Have you considered the Carreau-Yasuda model? It's strain rate-dependent and have model constants that include the effects of fluid rheology (in some sense).

I don't have the expression or constants at the moment, but do a web search and you'll find it.


I hadn't, I had tried Ostwalt de Waele before. I found a paper with parameters for Carreau-Yasuda, as well as Casson so I'm trying those out. Carreaux-Yasuda didn't come out very pretty, in fact the outlets are flowing inward but I haven't done Casson yet.

This might belong in a different thread and is probably too broad or vague, but does anyone have any recommendations for fluid coming in through outlets? This has been happening frequently, whether I use Ostwalt de Waele or Carreaux-Yasuda or non-newtonian. The solver says something like "try changing boundary to Opening type" but that won't solve the problem, it would just allow fluid to come in where it's supposed to be exiting instead of making it actually exit.

My boundary condition for the inlets are defining normal velocity (0.434 mm/s) and the outlets are average static pressure (0 atm, and 0.05 pres. profile blend)

Thank you everybody for the help.

ld1305 February 4, 2012 14:26

I realise this is old, but just wondering if you got anywhere with it, as I'm trying to do a CFX analysis of blood flow through the coronary artery

Any tips?

hesamking April 14, 2012 16:22

modeling blood in artery
 
I have set up everything in CFX 13 for simulation of blood in a nasilar artery
I run the case as a transient case following the tutorial of ANSYS-CFX for a circumvent transient analysis.
I run it but after the run is done and I open the results, I cannot see any specific results. Like I expect to see pressure variation on artery walls but there is just one blue color and some variation in a little part around my inlet.
Any idea why I don't get any results?
Thanks,
Hesam

ld1305 April 14, 2012 16:26

If you double click on the wall on the left hand panel and go through the settings at the bottom of the left panel, you should be able to change the colour from a solid to change with respect to velocity or pressure etc

If it's still blue then it's because it's using a global colour scale, right slick on your artery and in that menu or a submenu you should be able to change it to local

Hope that helps :)


All times are GMT -4. The time now is 13:02.