LES of jet injection giving pulsations
 

I am doing some work with LES of jet flow injections into a relatively stagnant body of the same working incompressible fluid and then impinging about 7 diameters away. My simulated injection tubes are 15 diameters long prior to the entrance and about 3 to 4 diameters after the entrance into the stagnant fluid I am seeing a pulses with a vortical annulus around the jet that moves down towards the impingement site. These are at extremely regular intervals. I'm running the simulation with a 0.001s timestep, the jet is 1cm, the inlet velocity into the pipe entrance is 1.058m/s.

My concern is that the pulses might be artifacts of some inlet condition as opposed to the actual flow phenomenon. I have experimental PIV data at hand of the same geometry/conditions but the flow does not have such strong features.


Has anyone seen this before either in experimental or numerical? The thing that comes to mind is simple turbulent instabilities in flow past a rod, but this is a jet flow and the entrainment of the surrounding fluid I feel like would represent itself differently than this.

Try posting up some images so that we can take a look at the phenomenon. It's hard to visualise through words alone.

This is a very short animation of 20 frames each spaced 5ms(5 time steps) apart. This phenomenon has been occurring over a longer period of time, but for illustration purposes, this should do.

The inlet tube extends well above the image, and it extends about 1 diameter into the body. The missing section in the center is an adjacent cylindrical rod.


Note: display vectors were reduced for this image, the mesh is not this coarse.

http://img59.imageshack.us/img59/2596/largepulsens.gif

How high is the Re number of the jet??? If you are working with some high Re, you probably are seeing the transition to turbulent flow zone of your system, as expected in a LES simulation. Have you tried to analyze the statistical data of this simulation? Another possibility is that your time step is too big for the simulation... try to use some cfl condition as commonly used in papers of LES jet simulation...

Good luck...