
Re: How to use FloWorks to Compute the force of drag?
Bill McEachern Aug 11, 2010 10:53 AM (in response to Nicholas Reynolds)use a surface goal and pick the force component goals you want as surface goals. You need a relatively fine mesh to get sensible results. You cna check the verification manual for flow over a 2D cylinder if you want to see how it does  pretty well really up to R'numbers of about ~6 million.

Re: How to use FloWorks to Compute the force of drag?
Nicholas Reynolds Aug 11, 2010 2:12 PM (in response to Bill McEachern)I tried a global goal and a surface goal for a sphere. I selected the force and the normal force in the direction of flow and the results being very similar were both about half the amount of the results I calculated analytically. Is it because I need a higher mesh.

Re: How to use FloWorks to Compute the force of drag?
Bill McEachern Aug 11, 2010 2:59 PM (in response to Nicholas Reynolds)maybe. depends on the flow regime. What r'number did you do your hand calc at? the Cd varies with R'number  goes from like .5 to ~.2 if I recall at super critical.

Re: How to use FloWorks to Compute the force of drag?
Nicholas Reynolds Aug 11, 2010 3:16 PM (in response to Bill McEachern)The velocity I imput into the general settings was 50m/s so that gives a reynolds number of 6,780,898 (Re=[v*l*ρ]/μ : ρ=1.207kg/m^3 μ=0.0000178 kg/m*s) The only term im not positive on the is l (i treated it as the depth the compoment so for a sphere it would be the dia).

Re: How to use FloWorks to Compute the force of drag?
Bill McEachern Aug 11, 2010 3:29 PM (in response to Nicholas Reynolds)The Kepsilon turbulence model that Flow sim uses starts to diverge from experimental in the flow regime you are considering. Check the validation manual and pick a super cirical R'number that Flow sim does a good job of and figure out what it takes to get a good number and then you can see how you do in the 6 million range. You can always put on the autoadaptive meshing and through it to result resolution to 8 and give it a rip. That should produce as good a number as it is likely to get with out much mucking about. Though it might take a bit to compute.



