Hi all,
I am simulating a flow over a cylinder using LES. Looking at several academic papers on the same topic, statistics are calculated after several vortex shedding cycles.
I would like to know if there is a way to calculate the number of shedding cycles for my case in order to get more reliable statistical information
Thank you.
No, the initial conditions are quite arbitrary and not physically correlated. You have to monitor some global quantity like the total kinetic energy in time and wait until the initial condition is disregarded. That happens when the total KE will oscillate around a mean value. After that, you can cumulate the fields to compute the statistics.
Clearly, the more the initial condition is close to a physical field, the lesser you have to wait. For example you can use as initial condition a preliminar RANS solution.
You can measure the force coefficients at each time point and then calculate the dominant frequencies from the time series data using FFT.
Plot lift or drag coefficients, and then do an FFT analysis.
Note that the frequency of the lift coefficient signal is the same as the vortex shedding frequency. On the other hand, the frequency of the drag coefficient signal doubles the vortex shedding frequency.
f: vortex shedding frequency
f_L= f: lift coefficient frequency
f_D = 2f: drag coefficient frequency
Hi Filippo Maria Denaro , can you explain why monitoring total kinetic energy, not any other flow variables?
Al Shahriar
Reaching the statistical energy equilibrium between production and dissipation of kinetic energy is a way to control the numerical solution has disregarded the arbitrary initial condition.
Filippo Maria Denaro
Hi Filippo, could you please provide us any reference ? I am really interested
Antonio Mezzacapo
references for what topic, exactly ? Sagaut is the standard textbook for LES/DNS.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Plating