5 Replies Latest reply on Dec 3, 2014 5:24 AM by Romano Capocci

# Differing results on drag force for simulation and calculation

Hello,

I am carrying out a flow analysis on my model in which I wish to find out the drag on my vehicle when it is subjected to water at different flow rates. I am getting different results for my simulation and basic hand calculations.

I carried out a simulation on my model. The front surface area is 0.1064 m^2 and the flow conditions were as follows:

• water density 1035 kg/m^3
• velocity in the X direction - 3 knots (-1.54 m/s)
• laminar and turbulent flow

I set up a Global Goal of Force in the X direction. This was to check the drag on my model. The result I got was -6.8905 kgf. I tried it with different mesh settings and the results were similar. I then calculated the drag using the drag equation 1/2*(Density of water/gravity)*A*(V^2)*Cd. I used 0.9 for the Cd. When I carried out this equation I calculated the drag force to be -11.98 kgf.

To test the simulator I then tested a regular shape - a sphere. The sphere had a radius of 0.2 m giving it a front surface area of 0.125 m^2. I used the same flow conditions as before and the drag force in the X direction was 1.6 kgf at mesh setting 4 and 3.24 kgf at mesh setting 3. (mesh setting 8 gives similar results to mesh setting 4). I then calculated the drag using the equation above with a Cd 0.47. The result I got was 7.38 kgf.

These are large differences between simulation and calculation. It is baffling me because I also tried it with a cube with 0.4m*0.4m*0.4m dimensions and a Cd of 1.05 and the results of the simulations and calculations are similar.

Is there something simple I am missing? If anyone could help me with this problem I would be very grateful as I have spent weeks messing around with it.

• ###### Re: Differing results on drag force for simulation and calculation

Looking at the sphere:  the Reynolds number for a sphere of 0.2m diameter in a uniform water flow of 1.54 m/s is on the order of 10^5, well within the turbulent region. You can ignore laminar flow here, the turbulence model will take care of the boundary layer. Let's do a simple analysis using the drag equation:

Fd = .5 * (density) * (speed)^2 * (drag coefficient) * (projected area)

Fd = .5 * (1035 kg/m^3) * (1.54 m/s)^2 * .47 * (pi * [0.1m]^2) = 18.1 N

The mass required to produce 18.1 N under Earth gravity  = 18.1 N / 9.81 m/s^2 = 1.85 kg, pretty darn close to what Flow Simulation spat back at you.

p.s. why are you using this strange "kg-f" unit? I never heard of that before reading your post. Sounds like something they use in the imperial unit system to deal with the confusion between pound-force and pound-mass.

• ###### Re: Differing results on drag force for simulation and calculation

Kgf is a pretty standard engineering unit. It's as bad as the lbf lbm, and gets used for the same reasons.

• ###### Re: Differing results on drag force for simulation and calculation

Thanks for breaking that down Amit. It was a great help

• ###### Re: Differing results on drag force for simulation and calculation

this comes up a lot

have you already looked at the tutorials and validation examples about aerodynamic drag? flow over a sphere is one of the examples and i think you'll find the answers to all your questions there

generally when running these simulations the issue is either the method of output (global goal is good but surface goal is also a good double check in case you ever want to add any other geometry)

next folks generally don't have a big enough computational domain or sufficient mesh, adaptive meshing and control planes really help

and finally sometimes transient is required

check out the documentation and let us know if you still have questions

• ###### Re: Differing results on drag force for simulation and calculation

I'll have a look at the documentation Jared. Thanks again.