Unexpected results for 2D cylinder flow

ddvdc · March 6, 2019, 18:32

I am trying to learn how to use SU2. As a test, I am trying to solve the 2D, incompressible, inviscid flow around a cylinder using the Euler equations. I have created a simple triangular mesh using Gmsh (see attachment.), and have changed the CFG file from the Inviscid Hydrofoil example on the SU2 website to work with my mesh instead (see attachment). I am then running the case in parallel using the command

Code:

parallel_computation.py -f cylinder.cfg -n 16

. The simulation exits after the convergence criteria are met. The inlet flow conditions are 10 m/s in the positive x-direction, so I expect a symetric flow field and no residual force in the y-direction. However, the flow field is not symmetric and there is a residual force in the negative y-direction. Note also that the velocity of the free-stream is not 10 m/s. I have attached two images depicting a set of streamlines and the pressure coefficient and velocity magnitude respectively. Can anyone tell me what I'm doing wrong here?

AlexF · March 10, 2019, 22:12

Glancing over the .cfg file you posted, here are a few things that I would change/adjust:

-It looks like you are specifying a CFL of 1e6 on the first iteration, then switching to an adaptive one in the interval [1, 100] for all other iterations (you can monitor this value in the history file).

-You are specifying a convergence criterion of 10 orders of magnitude reduction, but you also have the minimum value of the residual set to -6 (log10). Perhaps your simulation is exiting prematurely; you can try decreasing RESIDUAL_MINVAL to see if the same setup will still converge.

-I would also consider turning off multi-grid until you are achieving appropriate results.

I’m not sure these modifications will solve your problem, but it may help point out other issues.

As for the velocity magnitude not equaling 10 m/s: The values are non-dimensionalized with respect to the initial values by default. You can alter this by using the “INC_NONDIM= ” option. Reference values will be output at the beginning of the simulation, or you can view them in the forces_breakdown.dat file.

Hope this helps,
Alex

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problem with grid convergence for turbulent flow around cylinder	aakie	OpenFOAM Running, Solving & CFD	3	November 13, 2018 05:39
Drag force coefficient too high for a flow past a cylinder using komega sst	Scabbard	OpenFOAM Running, Solving & CFD	37	March 21, 2016 17:16
Flow Simulation : air around an rotating cylinder using Solidworks Flow Simulation	Wyrold	Main CFD Forum	0	October 22, 2015 09:48
Flow past rotating cylinder: Problem with ForeCoeffs	raf1111	OpenFOAM	1	December 16, 2013 10:45
Incorrect Drag and Drag Coefficient for flow over a cylinder	ozzythewise	Main CFD Forum	8	June 13, 2012 07:24

March 10, 2019, 22:12		#2
AlexF New Member Alex Join Date: Jul 2018 Posts: 8 Rep Power: 8	Glancing over the .cfg file you posted, here are a few things that I would change/adjust: -It looks like you are specifying a CFL of 1e6 on the first iteration, then switching to an adaptive one in the interval [1, 100] for all other iterations (you can monitor this value in the history file). -You are specifying a convergence criterion of 10 orders of magnitude reduction, but you also have the minimum value of the residual set to -6 (log10). Perhaps your simulation is exiting prematurely; you can try decreasing RESIDUAL_MINVAL to see if the same setup will still converge. -I would also consider turning off multi-grid until you are achieving appropriate results. I’m not sure these modifications will solve your problem, but it may help point out other issues. As for the velocity magnitude not equaling 10 m/s: The values are non-dimensionalized with respect to the initial values by default. You can alter this by using the “INC_NONDIM= ” option. Reference values will be output at the beginning of the simulation, or you can view them in the forces_breakdown.dat file. Hope this helps, Alex