I use Fluent to simulate the drag of bluff bodies. why the results is larger than that of EXP results? Who can give me the reasons. Cheers

Most turbulence models don't handle separation that well.

Did you check your reference values (Report->Reference values)? In 2-D, the "depth" reference value is what's used as your "third-dimension" to calculate forces, mass flow rate, etc.

If it's not that, then a common reason is that your mesh in the stagnation region is too coarse, so it's over predicting your local stagnation pressure. If it's a smooth body, you could be predicting the wrong separation point due to grid choice and turbulence model. If it's a common body, then you'll be able to find experimental data on the pressure coefficients along the body, or if its something you've come up with, then hopefully you've used pressure taps in your experiment to get an understanding of what the pressure looks like along the body. Either way, looking at the pressure coef along the body should help you understand where the drag is coming from and then you can try and figure out if it's realistic or not.

Hope this helps, and good luck, Jason