search

terrapower / armi

An open-source nuclear reactor analysis automation framework that helps design teams increase efficiency and quality
https://terrapower.github.io/armi/
Apache License 2.0
230 stars 90 forks source link

BU vs time plot prints useless/incorrect results #1902

Open keckler opened 1 month ago

keckler commented 1 month ago

One of the plots generated by default in ARMI is burnup vs time. The same plot also includes DPA vs time. In all of my runs, they contain all zeros: image

The plotting is done by this function: https://github.com/terrapower/armi/blob/59c392c1a6aa98e19c9955c88ca1a6106d1bcc35/armi/utils/reportPlotting.py#L216-L265

That function is quite fickle and incorporates a number of assumptions that are not often met by any old reactor that somebody might build. In those cases, you end up with a useless plot like the one that I showed above. Besides being useless, the plot is just wrong/misleading.

I would advocate moving highly-specific plots like this into the application that actually might use it (if any).

john-science commented 1 month ago

@keckler Is this problem new? For debugging purposes, it would be helpful to know if this worked 2 months ago.

I would advocate moving highly-specific plots like this into the application that actually might use it (if any).

Yeah, I think this is the real solution. I am not actually a big fan of having a million hyper-specific plots in ARMI. They are too hard to test and document, and surely no one outside TerraPower uses them. (@drewj-tp Am I right about that?)

Do you know where you would want to move this plot?

keckler commented 1 month ago

I am actually not sure if this problem is new or not. I would ask @mgjarrett if he's seen this same problem with any Natrium depletion calculations.

drewj-tp commented 1 month ago

Maybe related #267 ?

drewj-tp commented 1 month ago

I would advocate moving highly-specific plots like this into the application that actually might use it (if any).

Yeah, I think this is the real solution. I am not actually a big fan of having a million hyper-specific plots in ARMI. They are too hard to test and document, and surely no one outside TerraPower uses them. (@drewj-tp Am I right about that?)

I would file that into the "ARMI does too much" category and having ARMI native do fewer plots is better

john-science commented 3 weeks ago

@keckler @mgjarrett Who actually USES this?

Can I move this into someone's project?

Maybe your internal ARMI-based Application can subclass this ReportInterface, and add plots like this downstream, where people actually want them enough to support them?