Units to be used for various quantities in CRIMSON

121 views
Skip to first unread message

Satish Chimakurthi

unread,
Oct 20, 2020, 2:47:23 PM10/20/20
to CRIMSON Users
Hello CRIMSON users,

I am a bit confused about the units to be used in CRIMSON and would like to double check if my understanding is correct. In the slide that I attached herewith, I have shown units for various quantities (including some BC's and solver parameters) assuming two different scenarios related to length unit in a geometry that is created elsewhere and imported (as IGES, for example) into CRIMSON: 1) geometry length unit is 'mm', and 2) geometry length unit is 'm'. Could someone please take a look at it and let me know if everything looks good? If not, please correct me and also feel free to edit the PPT and upload it back if you would like, for the benefit of other users.

Thanks in advance,
Satish
Units to be used in CRIMSON.pptx

C. Alberto Figueroa

unread,
Oct 21, 2020, 9:50:12 AM10/21/20
to Satish Chimakurthi, CRIMSON Users
Hello Satish

CRIMSON makes no assumptions on units. It is up to the user to make sure that all quantities are consistent.

If you use Pa (kg*m/s^2/m^2), this is consistent with using 'm' for length.

If you use 'mm' for length, you need to scale your pressure units accordingly. You can't mix using length in mm and using pressure in Pa, which assumes meters.

Hope this helps.

- Alberto

--
You received this message because you are subscribed to the Google Groups "CRIMSON Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to crimson-user...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/crimson-users/06fa2d5f-10dc-40e0-a10d-9a09c93b26e4n%40googlegroups.com.


--

C. Alberto Figueroa, PhD

 

Edward B. Diethrich M.D. Professor of Biomedical Engineering and Vascular Surgery

Fellow, American Institute for Medical and Biological Engineering

University of Michigan

North Campus Research Complex B20-210W

http://bloodflow.engin.umich.edu/

http://www.crimson.software/

@BloodFlowSim

 




Satish Kumar Chimakurthi

unread,
Oct 21, 2020, 11:08:08 AM10/21/20
to alberto....@gmail.com, CRIMSON Users
Thank you very much Aberto. Let me then see what is causing the problem with my case whose results are not matching with the reference data! BTW, one reason I was confused about the units is because of an earlier post in the forum that suggests the use of 'Pa' with 'mm'!

So, even though the UI shows a particular set of units (e.g., g/mm^4.s for resistance - see below), I can simply enter the corresponding value in kg/m^4.s as long as I am consistent throughout. For example, in this case, I would ensure that the flow rate values in my .flow file are in m^3/s. Is this correct? Just making sure. Thanks in advance.

image.png

Best Regards,
Satish

--
Satish Chimakurthi, Ph.D
Herndon, VA, 20171

**The ultimate measure of a man is not where he stands in moments of comfort and convenience, but where he stands at times of challenge and controversy. - Martin Luther King, Jr**
 

C. Alberto Figueroa

unread,
Oct 21, 2020, 11:50:50 AM10/21/20
to Satish Kumar Chimakurthi, CRIMSON Users
Correct.

Thanks

C. Alberto Figueroa, PhD

Edward B. Diethrich M.D. Professor of Biomedical Engineering and Vascular Surgery

Chris Arthurs

unread,
Oct 22, 2020, 6:50:08 AM10/22/20
to Satish Kumar Chimakurthi, CRIMSON Users
Hi Satish,

I just want to clear up something here:
> one reason I was confused about the units is because of an earlier post in the forum that suggests the use of 'Pa' with 'mm'!
It is absolutely the case that the pressure units are Pa when the length-scale of your model is in mm. The earlier forum post is correct.

However, as Alberto says, you have to ensure that your units are dimensionally consistent - if you change the length-unit for your geometry, then you must change everything else. The CRIMSON GUI assumes mm as your length-unit, and all the hints in the GUI for units correspond to that (as in your image). However, they are only hints - so long as you are consistent throughout the program, you can type whatever quantities you want into the boxes - if you have a non-mm length scale in your geometry - and it's simply up to you to remain consistent.

I like to check my unit consistency by using a simple equation like P=QR (pressure = flow rate * resistance), replacing each symbol with its unit, then manipulating them algebraically. For example:

* When Q has units mm^3 s^-1, and R has units g mm^-4 s^-1
* P=QR implies that P has units g mm^-1 s^-2
* You can multiply by 1000 on the top and bottom: (1000 g) (1000 mm)^-1 s^-2 to get kg m^-1 s^-2; this is a Pascal.

You can do this with - for example - Q=C dP/dt for a compliance C and a pressure gradient P, too.

Out of interest, how did you generate the reference data you are trying to compare with?

Best,
Chris

On Wed, Oct 21, 2020 at 4:50 PM C. Alberto Figueroa <alberto....@gmail.com> wrote:
Correct.

Thanks

C. Alberto Figueroa, PhD

Edward B. Diethrich M.D. Professor of Biomedical Engineering and Vascular Surgery
University of Michigan
North Campus Research Complex B20-210W
http://bloodflow.engin.umich.edu/
http://www.crimson.software/
@BloodFlowSim

   
On Wed, Oct 21, 2020, 11:08 Satish Kumar Chimakurthi <satish...@gmail.com> wrote:


--
Chris Arthurs, D. Phil.
King's Prize Research Fellow
Imaging Sciences and Biomedical Engineering
King's College London
UK
==
CRIMSON Cardiovascular Modelling Environment: http://www.crimson.software/
==
Twitter: @carthu15p

Satish Kumar Chimakurthi

unread,
Oct 22, 2020, 10:49:27 AM10/22/20
to Chris Arthurs, CRIMSON Users
Hi Chris,

* You can multiply by 1000 on the top and bottom: (1000 g) (1000 mm)^-1 s^-2 to get kg m^-1 s^-2; this is a Pascal.  
You are correct. Sorry, I didn't realize that earlier.

Regarding the reference data, I got it from a thesis that has both an analytical solution as well as CFD solution for a seemingly simple case (cylindrical geometry with an inlet clinical wave form and a RCR outlet) that is summarized in the snapshot below. As far as I can see, the units in my project are pretty consistent and nothing seems to be glaringly wrong. However, the results (outlet pressure in this case) that I generated from my CRIMSON run are way OFF the data you see below. Given that I am new to CRIMSON, I am sure there are errors somewhere in my setup. I will continue to look. By the way, is there a tutorial case or something that is similar to this simple example that I can use as reference?

Thanks in advance,
Satish

image.png



Chris Arthurs

unread,
Oct 22, 2020, 11:05:08 AM10/22/20
to Satish Kumar Chimakurthi, CRIMSON Users
Hi Satish,

There's nothing this simple, but there's the idealised geometry case from the 2018 MICCAI workshop, which you can download the files for here http://www.crimson.software/workshop-granada-2018---2.html

The lecture slides in there refer to the data files in the same package. They're very similar to those explicitly placed on our Documentation page.

Best,
Chris

On Thu, Oct 22, 2020 at 3:49 PM Satish Kumar Chimakurthi <satish...@gmail.com> wrote:
Hi Chris,

* You can multiply by 1000 on the top and bottom: (1000 g) (1000 mm)^-1 s^-2 to get kg m^-1 s^-2; this is a Pascal.  
You are correct. Sorry, I didn't realize that earlier.

Regarding the reference data, I got it from a thesis that has both an analytical solution as well as CFD solution for a seemingly simple case (cylindrical geometry with an inlet clinical wave form and a RCR outlet) that is summarized in the snapshot below. As far as I can see, the units in my project are pretty consistent and nothing seems to be glaringly wrong. However, the results (outlet pressure in this case) that I generated from my CRIMSON run are way OFF the data you see below. Given that I am new to CRIMSON, I am sure there are errors somewhere in my setup. I will continue to look. By the way, is there a tutorial case or something that is similar to this simple example that I can use as reference?

Thanks in advance,
Satish

image.png


On Thu, Oct 22, 2020 at 6:50 AM 'Chris Arthurs' via CRIMSON Users <crimso...@googlegroups.com> wrote:


--
Satish Chimakurthi, Ph.D
Herndon, VA, 20171

**The ultimate measure of a man is not where he stands in moments of comfort and convenience, but where he stands at times of challenge and controversy. - Martin Luther King, Jr**
 


--

Satish Kumar Chimakurthi

unread,
Oct 27, 2020, 10:22:13 AM10/27/20
to Chris Arthurs, CRIMSON Users
Hi Chris,

The length of my cylinder along Z is 0.2m.  I loaded an IGES file wherein the length unit is 'm'. After import, when I RMB on the geometry name and click on 'Details', I see that the bounding box' Z dimension is 200. See snapshot below. I believe the display of bounding box dimensions in 'mm' is only for display purposes? In reality, the code should know that the length is actually 0.2m, correct?

TIA,
Satish

image.png

Chris Arthurs

unread,
Oct 27, 2020, 2:00:50 PM10/27/20
to Satish Kumar Chimakurthi, CRIMSON Users
Hi Satish,

I don't know the details of how your data was created and saved to file, but that sounds suspiciously to me like the model is in mm, not m. The bounding box doesn't need to conform to the image dimensions, but the fact that 200 mm = 0.2 m is highly suspicious.

CRIMSON doesn't "know" anything about length units. It just gets a number, and expects you to manage the units yourself. If you're using mm, then the hints in the GUI regarding other units will be consistent.

Best,
Chris

On Tue, Oct 27, 2020 at 2:22 PM Satish Kumar Chimakurthi <satish...@gmail.com> wrote:
Hi Chris,

The length of my cylinder along Z is 0.2m.  I loaded an IGES file wherein the length unit is 'm'. After import, when I RMB on the geometry name and click on 'Details', I see that the bounding box' Z dimension is 200. See snapshot below. I believe the display of bounding box dimensions in 'mm' is only for display purposes? In reality, the code should know that the length is actually 0.2m, correct?

TIA,
Satish

image.png

On Thu, Oct 22, 2020 at 11:05 AM Chris Arthurs <christoph...@kcl.ac.uk> wrote:

Satish Kumar Chimakurthi

unread,
Oct 27, 2020, 2:09:11 PM10/27/20
to Chris Arthurs, CRIMSON Users
Chris,

OK, thanks. It's just a regular IGES file that I exported from a CAD tool that I use and then imported into CRIMSON. I know that the length unit is 'm' in that file. I wonder why it's just not working out for me :-).

Regards,
Satish
**Ego = 1/Knowledge: "More the knowledge, lesser the Ego, Lesser the knowledge More the Ego" - Albert Einstein**
 
Reply all
Reply to author
Forward
0 new messages