A question about ensemble_correlation.ipynb

[Ivan Nepomnyashchikh]

Hello dear OpenPIV community.

I am looking at this example of ensemble correlation averaging.

My question is where dt comes into play in that example. By dt I mean the time between the pulses in a double-pulse.

In that example, we find correlation in block [7]. In block [12], we use the correlation to find velocities. But we don't supply dt to the function that calculates velocities. How is that possible?

Thank you in advance.

Ivan

[Alex Liberzon]

openpiv eatimates displacement fields in pixels per dt. 

if the user provides the scaling (pix/mm) and dt (sec), then the output can be mm/sec 

--
You received this message because you are subscribed to the Google Groups "openpiv-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpiv-user...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/openpiv-users/e8ee3618-a695-49fc-bfd9-085df15cd168n%40googlegroups.com.

[Ivan Nepomnyashchikh]

Thank you, Dr. Liberzon.

If one were to choose to provide dt,  where exactly would one supply dt  in this example?

I've gone through the codes for the functions involved with the example: I couldn't find a function that accepts dt as an argument. But that could be due to my inexperience, that's why I'm asking for clarification.

Ivan

[Alex Liberzon]

it does not really matter where, maybe a line before saving the data ( in this example we do not save). 

i usually introduce dt only in the post processing, before plotting the final figures. see pivpy examples for instance 

To view this discussion on the web, visit https://groups.google.com/d/msgid/openpiv-users/c5c3297c-b4e8-4ad7-99a7-4c50d9ad1f17n%40googlegroups.com.

[Ivan Nepomnyashchikh]

Ah, I see, thank you.

That's what I didn't realize: I need to introduce some additional function to the code that would take dt as an argument and return velocities in mm/s.

I didn't know such function exists in OpenPIV. Before I tried ensemble averaging, I had only used

pyprocess.extended_search_area_piv  which takes dt  as an argument an does the units conversion on the background. So I didn't have to worry about the units conversion.

I'll go ahead and look for such a function in the examples.

Thank you.

Ivan