Optimizing mesh quality and Y+ layer formation on an FSAE car

I tried both and the 32 bit one had an error, the 64 said it cant the path

It could be the language issue. Or maybe that im running windows 7? i have no idea

Change directory in the terminal window using cd c:… to the dir where the ocx is, then type the command without any of the % items…

Is this what you mean? the OCX in the same place as the program? I have no idea :sweat_smile:
image

Relax, I have duplicated your error on a new computer, I will debug this and get back to you here :wink:

ok haha, i was about to call an ambulance. Its late here and my head was about to explode. Thanks a lot for troubleshooting!

Darn, I think all it was, was that I forgot to take it out of debug mode when I compiled the project…

I had to make it version 205 now and you can get the setup file for 205 here

Run that setup program and the replace all yPlusHistogram.exe files that you have put anywhere on your computer (like in the browser downloads directory) with the new yPlusHistogram.exe…

Sorry for confusion, brain fart I hope, let me know if it works now?

Success!! This is the plot from the full car simulation

  • 1 layer BL at 3.4mm
  • 1000 iterations - it seems i should have it run longer, it did not reach 1% stability

The Y+ histogram log is still running but so far it looks good. Ill update the final histogram when all faces are accounted for.

Edit here is the final result - 30mb file, had to wait a bit

also on and off i am getting this error. When i re-click on the yplus_area.csv in the histogram program to re-calculate the Y+ it shows up. Not all the time though. When i click ok on the error box it closes the program but when i reopen it, it works agian. Just thought id let you know.

Only way I can debug that is for me to have the exact yplus_area file that has just been clicked on when the error happens…
Do you have a way of uploading the 30mb file to the web somehow and then put a link for access to it here?

yes i can make a dropbox file with the link in the file and share it with you, does that work?

Here is the dropbox link to the folder. I think the run with 88,725 had the error but im not sure.

Problem is, now that full solution dataset when you download it will always contain time 0 and 1000, so even if you continue that run, the useless 1000 iteration data will always be there…
Personally, I would start a NEW sim run with 0 start iteration and 5000 end iteration and watch it until 1% ORSI with Moving Average iterations of about 150, which is about 5x the 30 iteration period of your oscillations (even try 250 when it is running)… As the value that it is oscillating around gets more stable, you can see how increasing the the MA iterations, filters the oscillations out, giving you a better guess at the value the points are oscillating around…
Sorry, when we talked about the way to figure out how many sim iterations to run to for ‘quick future runs’, I was assuming you were talking about my sim run that already had been run to 3000… ONLY AFTER A 1% FULL RUN CAN YOU DECIDE WHAT ITERATION YOU WANT TO USE FOR A FUTURE ‘QUICK RUN’…
If I actually realized you were asking about the full car sim run I would have said to use 5000 end iteration and stop it manually when ORSI 1% is reached …
PS: To save space after solution set down loads, I always delete the useless Time 0 folder which saves a LOT of hard disk …space

Those are just the 1kb .txt summary files… I need the 30mb .csv output file from ParaView…
I see that you have not setup your windows file explorer to ‘show extensions’, it might help you if you had file extensions visible, rather that relying on Windows icon guess for which program you want to use for when you click on a file, I hate it when Windows gets more Macish :wink:

Dont worry i had run the full car simulation before we talked about when to stop / what write interval to use. I had done that sim mainly to see Y+ values so i could know if a full car with 1 layer was landing in the 30-300 range. I now know it is .

So i understand the Moving average input correctly. Is this used to calculate the average Cl or Cd value as its moving along the iterations? So a MA of my current value of 50 would show one full cycle of raw data in yellow? Changing this to 150 or 250 will flatten out the wave values so i can make a better guess of the average value between the cycles correct?

This is very useful thanks!!

I included the 30mb file, it exported as an excel icon for me

Yes, yes, yes…

My latest chart here, has column titles that explain what is happening with ORSI better than the program output screen… The ORSI value is in fact a Moving Average value at the ‘Iteration of interest’ and the % is the range of MA values in the previous number of iterations in the ORSI window :wink:

Is it possible that you had that 30mb file in a folder where yPlusHistogram was in the ‘Monitoring is ON’ green mode?

I think the error comes when my monitoring code opens a yPlusArea.csv file instead of a 2D plot csv file, I can easily plug that error but I did not forsee a person doing that…

mmm that i cannot say for sure. I have one copy of the histogram.exe in the folder with the case-foam data and another in the downloads directory. Im pretty sure i didnt have it in monitor mode but i can try to replicate this again next time i download a csv file.

OK, I will plug that error anyway in v206 :slight_smile:

PS: I can not seem to duplicate your error on that file when I try to get a yPlusHistogram with it, I get no errors… So that bug will float around for a while until it happens to me while I am in debug mode :frowning:

OK so … Here is the full simulation run. You have taken the last 500 iterations, which encapsulates one full cycle of the raw yellow data, varying 20% over the 500 iterations, which makes it difficult to estimate the “true” average of the cycle.

This next picture is showing the 500 iteration section from the entire simulation shown above. It is still fluctuating at 20% but by changing the MA to 2, it takes the average of 2 iterations within 10% fluctuation (because there are no other areas less then 10% fluctuation available) over this 500 iteration range, giving you the closest average of the data set at 0.424 Cd and -0.2367 Cl.

Correct?