Residence Time of RBC's and Haemolysis(Help for Searching)

kastarkas · September 15, 2016, 01:51

Hai all,

This cfx forum has given me lot regarding the cfd analysis for a centrifugal blood pumps.

Wonderful to be a part of this forum.

I have a problem with the setting up of "additional variables" (Eularian Approach) in CFX, for modelling the shear induced haemolysis of blood cells(RBC's).

I have finalised the model: dHb/Hb = (3.62*10^-7)*(T^0.785)*(Tau*2.416).
(Ref : Garon and Farinas Hyperbolic Model)

I have also seen some relevant posts regarding this. Can someone help me to search with links of threads or any other relevant documents/tutorials for setting up the "Additional Variables" in CFX ?

Any help is blissful. Thank you all in advance.

ghorrocks · September 15, 2016, 02:24

If I understand you correctly the equation you quote defines the gain in red blood cell damage as a function of local shear stress and T (is that time, temperature or something else?).

To model this is would appear an additional variable with a transport equation (but no diffusion) would be suitable. The initial condition is presumably Hb=0 everywhere and probably at the boundaries too, and the equation you quote would be a source term which would generate Hb.

Read the section on additional variables to understand the different implementations, and source terms to understand total sources or flux approaches.

kastarkas · September 15, 2016, 02:38

Thank you for the reply Glenn. The "T' is time and 'Tau' is shear stress only as you said.

Great. I l look upon the section of "Additional Variables" and i l update asap.

Thanks a lot.

kastarkas · September 17, 2016, 08:12

Hai Glenn,

I went through the documentation for source terms and additional variables. I am a bit confused on arriving at inserting the expressions and source terms here.

This is my case.

Transport Eqn : (∂/ ∂t + ∇.V) D = I

(V - Velocity Vector, D - Linear Blood Damage Index)

D = d^(1/0.785)

d = ΔHb/Hb =(3.62*10^-7)*(t^0.785)*(tau^2.416)

[t = exposure time of RBC, tau = shear stress on blood, ΔHb/Hb = Gain in blood damage. ]

'I' is mentioned as the source term causing damage.

My doubt is :

1. Shear stress on blood = (2*Density*Turbulent KE), is that right ?. I saw in the forum thread in which you helped. Should i create this as an additional variable ?

2. You told me about Hb is presumably zero at boundaries and at initial conditon. So shoud i add a term for this equation at the inlet surface..?

3. While inserting an expression in CFX, how will i select the 'time of exposure' from the in built list of variables. ?

Thank you again.

kastarkas · September 19, 2016, 01:39

Hai all,

Can someone help me out ?

ghorrocks · September 19, 2016, 01:47

1. I have no idea. I am not an expert in blood damage.
2. No, these are boundary and initial conditions. So it is specified in the boundary and initial conditions.
3. Time of exposure of what? Is this simulation time, or something else?
4. Only if you need to use and area or volume integral. I have not seen anything which appears to require one. But I don't understand what you are doing so there might be something I am not aware of.
5 and 6 - I do not understand these questions.

kastarkas · September 19, 2016, 06:09

Thank you glenn for that.

The time, that I mentioned is actually the total time that a blood cell is exposed to shear, and dHb/Hb is given as a funtion of this time and the shear stresses on this cell. So how am i supposed to include that while creating the Expression ?

And also can u please tell me am i right on these steps :

1. create additional variable (dHb/Hb).
(specific and no units).

2. Activate the addtional variable in domain.

3. create source terms with at inlet, with the value of source term as that of the expression equating dHb/Hb .

ghorrocks · September 19, 2016, 08:07

I don't really understand the model you are trying to implement so cannot comment on the detailed implementation.

So I would recommend you do a simplified model to understand how this works so you can adapt it to your case. Take a simulation such as the static mixer tutorial. Add an additional variable with a transport equation. Then add a source term set to 1 [1/s^1] and set the additional variable to zero at the boundaries. When you get this working it should give you the fluid residence time in seconds. Note you will need to sort out exactly what type of additional variable and other details. I will leave that as an exercise for you

So do a simple model first and make sure you understand that before you tackle your more complex one.

kastarkas · September 19, 2016, 15:44

Great. Thank you again.
Very much of a relief on hearing that....I l try to implement the basic ones and get it working....and i l come back right here after that...Thanks a lot..

kastarkas · September 21, 2016, 08:58

Hello everyone,

I need to create a simple expression for time(t) = {V^(1/3)} / u in CEL CFX.

where V = volume of each computational cell and u = magnitude of velocity of each cell.

'V' can be obtained through a syntax for "volume of finite volumes" ie. (volcvol)
although i am not able to find the short name for it in the documentation.

i got volcvol from forum only.

How can 'u' be accessed ? what is the variable for that. is this the usual variable "velocity" of cel or is there anything else ?

Thank you all in advance..

ghorrocks · September 21, 2016, 19:53

The names for all variables are described in the CFX Reference manual, in the chapter titled "Variables in ANSYS CFX". You will find the velocity magnitude can be accessed as either "vel" or "Velocity".

kastarkas · September 22, 2016, 03:54

hai,
Thank you Glenn

I needed velocity of each cell, . that is the problem

Lance · September 22, 2016, 06:26

Quoting Glenn:

Quote:

Originally Posted by ghorrocks

The names for all variables are described in the CFX Reference manual, in the chapter titled "Variables in ANSYS CFX". You will find the velocity magnitude can be accessed as either "vel" or "Velocity".

vel or Velocity is the velocity in each finite volume. The volume is, as you already found out, Volume of Finite Volumes or volcvol. They are described in the above mentioned section of the reference manual.

kastarkas · September 22, 2016, 06:41

Wow. Thank you Lance on that.
I l try that in 'Pre', and come back with update.

Thank you.

kastarkas · September 26, 2016, 04:13

Hello everyone,

With all the help from the thread, I think I moved a bit ahead. Thanks a lot for that.

kastarkas · October 3, 2016, 01:44

Hello friends, i am doing the transient analysis of the same problem.....I am having a problem here....
setup is MRF, transient rotor stator interface.., additional variables and source terms....Time step : 0.0001 s, Total time : 0.02 s, 10 iterations per time step.

It comes around 180 time steps...and it crashed exactly at 104 previously...there was a "less storage space" message and i cleared out the space in the drive where the files are saved.

then i run again and i get the same message..

I have attached the image of the error as well...

Can somebody please fix this for me...

+--------------------------------------------------------------------+
| Writing transient file 105_full.trn |
| Name : Transient Results 1 |
| Type : Standard |
| Option : Every Timestep |
+--------------------------------------------------------------------+

+--------------------------------------------------------------------+
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| write_compressed_dataarray: (fwrite failed) syserr:: No space lef- |
| t on device |
| |
| |
| |
| |
+--------------------------------------------------------------------+

+--------------------------------------------------------------------+
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| iocnt: write data failed |
| |
| |
| |
| |
| |
+--------------------------------------------------------------------+

+--------------------------------------------------------------------+
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| WRITE_BLKS_SER: write data failed: what=G/MFLOIP_FL1 where=ZN1/ES- |
| 1 |
| |
| |
| |
| |
+--------------------------------------------------------------------+

+--------------------------------------------------------------------+
| Writing crash recovery file |
+--------------------------------------------------------------------+

+--------------------------------------------------------------------+
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| write_compressed_dataarray: (fwrite failed) syserr:: No space lef- |
| t on device |
| |
| |
| |
| |
+--------------------------------------------------------------------+

+--------------------------------------------------------------------+
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| iocnt: write data failed |
| |
| |
| |
| |
| |
+--------------------------------------------------------------------+

+--------------------------------------------------------------------+
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| WRITE_BLKS_SER: write data failed: what=G/COURANT_FL1 where=ZN1/V- |
| X |
| |
| |
| |
| |
+--------------------------------------------------------------------+

+--------------------------------------------------------------------+
| An error has occurred in cfx5solve: |
| |
| The ANSYS CFX solver exited with return code 1. No results file |
| has been created. |
+--------------------------------------------------------------------+

End of solution stage.

+--------------------------------------------------------------------+
| An error has occurred in cfx5solve: |
| |
| Error reported by IO module: write_index: warning: new index not |
| at end of file, file may be bad |
+--------------------------------------------------------------------+

Lance · October 3, 2016, 03:42

Quote:

Originally Posted by kastarkas

there was a "less storage space" message and i cleared out the space in the drive where the files are saved.
[...]
then i run again and i get the same message..
[...]
Can somebody please fix this for me...

| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| write_compressed_dataarray: (fwrite failed) syserr:: No space lef- |
| t on device |

1) Buy a bigger hard drive
or
2) Run the simulation on another hard drive
or
3) dont save data as often.

kastarkas · October 3, 2016, 06:54

Thank you Lance on that..
So is this regarding the memory issue alone ? Does any other error messages have any other significance...?

Lance · October 4, 2016, 02:46

The errors you presented in the last post all relate to writing data to disc, and that the disc is full.

kastarkas · October 4, 2016, 04:50

OK lance.
I will do the needful regarding the memory and update right here as it is fixed.
Thanks a lot .

September 15, 2016, 01:51	Residence Time of RBC's and Haemolysis(Help for Searching)	#1
kastarkas Member kastarkas Join Date: Aug 2016 Posts: 45 Rep Power: 10	Hai all, This cfx forum has given me lot regarding the cfd analysis for a centrifugal blood pumps. Wonderful to be a part of this forum. I have a problem with the setting up of "additional variables" (Eularian Approach) in CFX, for modelling the shear induced haemolysis of blood cells(RBC's). I have finalised the model: dHb/Hb = (3.6210^-7)(T^0.785)(Tau2.416). (Ref : Garon and Farinas Hyperbolic Model) I have also seen some relevant posts regarding this. Can someone help me to search with links of threads or any other relevant documents/tutorials for setting up the "Additional Variables" in CFX ? Any help is blissful. Thank you all in advance.

September 17, 2016, 08:12		#4
kastarkas Member kastarkas Join Date: Aug 2016 Posts: 45 Rep Power: 10	Hai Glenn, I went through the documentation for source terms and additional variables. I am a bit confused on arriving at inserting the expressions and source terms here. This is my case. Transport Eqn : (∂/ ∂t + ∇.V) D = I (V - Velocity Vector, D - Linear Blood Damage Index) D = d^(1/0.785) d = ΔHb/Hb =(3.6210^-7)(t^0.785)(tau^2.416) [t = exposure time of RBC, tau = shear stress on blood, ΔHb/Hb = Gain in blood damage. ] 'I' is mentioned as the source term causing damage. My doubt is : 1. Shear stress on blood = (2DensityTurbulent KE), is that right ?. I saw in the forum thread in which you helped. Should i create this as an additional variable ? 2. You told me about Hb is presumably zero at boundaries and at initial conditon. So shoud i add a term for this equation at the inlet surface..? 3. While inserting an expression in CFX, how will i select the 'time of exposure' from the in built list of variables. ? Thank you again. Last edited by kastarkas; September 27, 2016 at 08:47.*

September 19, 2016, 01:39		#5
kastarkas Member kastarkas Join Date: Aug 2016 Posts: 45 Rep Power: 10	Hai all, Can someone help me out ? Last edited by kastarkas; September 27, 2016 at 08:48.

September 19, 2016, 06:09		#7
kastarkas Member kastarkas Join Date: Aug 2016 Posts: 45 Rep Power: 10	Thank you glenn for that. The time, that I mentioned is actually the total time that a blood cell is exposed to shear, and dHb/Hb is given as a funtion of this time and the shear stresses on this cell. So how am i supposed to include that while creating the Expression ? And also can u please tell me am i right on these steps : 1. create additional variable (dHb/Hb). (specific and no units). 2. Activate the addtional variable in domain. 3. create source terms with at inlet, with the value of source term as that of the expression equating dHb/Hb . Last edited by kastarkas; September 27, 2016 at 08:50.

September 19, 2016, 15:44		#9
kastarkas Member kastarkas Join Date: Aug 2016 Posts: 45 Rep Power: 10	Great. Thank you again. Very much of a relief on hearing that....I l try to implement the basic ones and get it working....and i l come back right here after that...Thanks a lot.. Last edited by kastarkas; September 27, 2016 at 08:51.

September 15, 2016, 02:24		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	If I understand you correctly the equation you quote defines the gain in red blood cell damage as a function of local shear stress and T (is that time, temperature or something else?). To model this is would appear an additional variable with a transport equation (but no diffusion) would be suitable. The initial condition is presumably Hb=0 everywhere and probably at the boundaries too, and the equation you quote would be a source term which would generate Hb. Read the section on additional variables to understand the different implementations, and source terms to understand total sources or flux approaches.

September 15, 2016, 02:38		#3
kastarkas Member kastarkas Join Date: Aug 2016 Posts: 45 Rep Power: 10	Thank you for the reply Glenn. The "T' is time and 'Tau' is shear stress only as you said. Great. I l look upon the section of "Additional Variables" and i l update asap. Thanks a lot.

September 19, 2016, 01:47		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	1. I have no idea. I am not an expert in blood damage. 2. No, these are boundary and initial conditions. So it is specified in the boundary and initial conditions. 3. Time of exposure of what? Is this simulation time, or something else? 4. Only if you need to use and area or volume integral. I have not seen anything which appears to require one. But I don't understand what you are doing so there might be something I am not aware of. 5 and 6 - I do not understand these questions.

September 19, 2016, 08:07		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	I don't really understand the model you are trying to implement so cannot comment on the detailed implementation. So I would recommend you do a simplified model to understand how this works so you can adapt it to your case. Take a simulation such as the static mixer tutorial. Add an additional variable with a transport equation. Then add a source term set to 1 [1/s^1] and set the additional variable to zero at the boundaries. When you get this working it should give you the fluid residence time in seconds. Note you will need to sort out exactly what type of additional variable and other details. I will leave that as an exercise for you So do a simple model first and make sure you understand that before you tackle your more complex one.

September 21, 2016, 08:58		#10
kastarkas Member kastarkas Join Date: Aug 2016 Posts: 45 Rep Power: 10	Hello everyone, I need to create a simple expression for time(t) = {V^(1/3)} / u in CEL CFX. where V = volume of each computational cell and u = magnitude of velocity of each cell. 'V' can be obtained through a syntax for "volume of finite volumes" ie. (volcvol) although i am not able to find the short name for it in the documentation. i got volcvol from forum only. How can 'u' be accessed ? what is the variable for that. is this the usual variable "velocity" of cel or is there anything else ? Thank you all in advance..

September 21, 2016, 19:53		#11
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	The names for all variables are described in the CFX Reference manual, in the chapter titled "Variables in ANSYS CFX". You will find the velocity magnitude can be accessed as either "vel" or "Velocity".

September 22, 2016, 03:54		#12
kastarkas Member kastarkas Join Date: Aug 2016 Posts: 45 Rep Power: 10	hai, Thank you Glenn I needed velocity of each cell, . that is the problem

September 22, 2016, 06:41		#14
kastarkas Member kastarkas Join Date: Aug 2016 Posts: 45 Rep Power: 10	Wow. Thank you Lance on that. I l try that in 'Pre', and come back with update. Thank you.

September 26, 2016, 04:13		#15
kastarkas Member kastarkas Join Date: Aug 2016 Posts: 45 Rep Power: 10	Hello everyone, With all the help from the thread, I think I moved a bit ahead. Thanks a lot for that. Last edited by kastarkas; September 27, 2016 at 08:52.

October 3, 2016, 06:54		#18
kastarkas Member kastarkas Join Date: Aug 2016 Posts: 45 Rep Power: 10	Thank you Lance on that.. So is this regarding the memory issue alone ? Does any other error messages have any other significance...?

October 4, 2016, 02:46		#19
Lance Senior Member Lance Join Date: Mar 2009 Posts: 669 Rep Power: 22	The errors you presented in the last post all relate to writing data to disc, and that the disc is full.

October 4, 2016, 04:50		#20
kastarkas Member kastarkas Join Date: Aug 2016 Posts: 45 Rep Power: 10	OK lance. I will do the needful regarding the memory and update right here as it is fixed. Thanks a lot .