Fluids

Fluids

Fluent – inconsistencies in force coefficients after modifying reference values

Tagged: 

    • davide.salvatore
      Subscriber

      Hi all I would like your help to understand how the force coefficients are actually computed inside Fluent.

      The scenario where the inconstinsetencies to which the title refers to is the following:

      I have a set of external aerodynamics computations, automated through a journal file. For each flow condition (in therms of thermodynamic values of the asimptotic flow) I have multiple runs with varying angle of attack.

      In one case (one flow condition, multiple angles of attack), due to an error in the journal, the reference values where not updated and so they were set to a wrong value from previous asimptotic flow conditions.

      In order to correct the values of the force coefficients, I ran another journal file, which consisted in:

      loading the .cas file (once, common to all the cases);
      for each case (i.e. for each angle of attack):
         loading the .set file and loading the .dat file;
         computing the reference values from the far field;
         export the force coefficients;
         save the .set file with updated reference values;
         continue to the next case.

      The workflow worked without apparent problems and the reference values were updated to the correct values.
      Since the the ratio between the previous (wrong) reference dynamic pressure and the new (corrected) one resulted to be 0.7058, I was expecting the same ratio between the corrected coefficients and the wrong ones. This happens for three out of four of the computed coefficients: cat, clm, cn (even if also here the fluctuations in the ratios should not be there in my opinion). But for one of them (caf) the ratio is completely different.
      clm, cn and caf are computed on the same surfaces, so that I can't really figure out why this is happening.

      old_ref_dyn_pressure / new_ref_dyn_pressure = 0.7058 (according to the reported reference values)

      new_c / old_c:
      AoA   cat          caf         clm         cn
      2        0.7062    0.2057    0.7065    0.7063
      6        0.7061    0.2153    0.7063    0.7062
      10      0.7062    0.2156    0.7062    0.7062

      Any idea how this could happen?
      Doesn't Fluent compute the force coefficients by simply dividing the computed force by the reference values?

    • DrAmine
      Ansys Employee

      The reference values are used to calculate force coefficients as described in the manual.  Not all reference values are populated from the boundaries you calculate the values from. So I recommend assessing the values for all your design points. What is the caf in your case? How are you defining / calulating it?

    • davide.salvatore
      Subscriber

      Hi DrAmine,

      thank you for your reply.
      It is clear that not all the reference values are populated from the boundaries, however, the ones of interest in this case that are not (the Area and the Lenght) are set in the general case file and they are not modified during the automated process. For what concerns the ones that are updated by the "compute from" command (density, velocity and pressure), they were not set to the right values in the first istance, when the computation was performed. However, after running the correcting juornal I described in the first post, they were set to the right value. From computing the ratio between the reference dynamic pressure computed with the old (wrong) values and the new (corrected and checked) values that I was expecting the same (inverted) ratio between old and nex values of the force coefficients (no other activity was performed, no additional modifications or iterations).

      Instead what I get is the situation depicted in the table, please note that for each angle of attack (AoA) we are referring to the same case, and even in this case you can see a huge difference in the ratio for caf and cn, which are force coefficients computed on the same surfaces, just different directions.

      The case in exam is approximately an ogive-cylinder body the meaning of the coefficients are:

      caf: axial (x) force coefficient (forebody)
      cn: normal (z) force coefficient
      clm: moment about y axis
      cat: axial (x) force coefficient (total)

      The first three are computed on the whole body except the base surface, cat is computed including the base. All of them are defined whith the preset force monitor option. So no expressions on my side.

    • DrAmine
      Ansys Employee

      If you just report the forces (not the coefficient): are you getting the same values or different values at the end of the run (hopefully converged run)? 

Viewing 3 reply threads
  • You must be logged in to reply to this topic.