Skip to content

Compute Uint from time-averaged energy

Lukas Bentkamp requested to merge bugfix/average-Uint into develop

In DNS.py, both time-dependent integral velocity Uint(t) and averaged integral velocity Uint are computed. So far Uint=<Uint(t)>, where <> is a time-average. It might be more common to instead compute energy=<energy(t)> and then use Uint=sqrt(2*energy/3).

I tested this for a set of stationary simulations. The difference is very slight: Previously:

Rlambda = [ 59.82527534  72.59297539  80.16412996  85.47906881  90.26781386
  92.3196589   96.58668253  99.14463083 100.85668077 102.93845416
 103.41286239 106.38760229 107.45911385 109.19369953 109.60041695
 111.48464369 113.4161075  113.44344872 116.70936876 116.4221249
 118.10234642 119.87978142 120.18360381 134.019079   154.17557626
 172.88801052]

Now:

Rlambda = [ 59.83924367  72.60373282  80.18052247  85.50423767  90.28926146
  92.33381751  96.60806008  99.15791885 100.87778654 102.95101777
 103.42714578 106.40808579 107.48830375 109.20978956 109.62526639
 111.501744   113.43822281 113.46402264 116.72748148 116.44870981
 118.13622209 119.91338466 120.20334813 134.03655029 154.21000369
 172.92618076]

Please check whether this is the most common way to do this!

Edited by Lukas Bentkamp

Merge request reports