We Now Detail Our Process to Determine the Aerodynamic Coefficients Using the Computational

A.1.2.5 4

A.1.2.5 CMARC

We now detail our process to determine the aerodynamic coefficients using the computational fluid dynamics package CMARC. This process is used in place of Fluent due to Fluent’s exceptionally long run time and because the results obtained in CMARC are reasonably close to those in calculated in Fluent.

While trying to run a full three dimensional computational fluid dynamics simulation to obtain aerodynamic coefficients, we found that running such cases would take an extremely long time to reach a converged solution. This led to our decision to come up with an alternative method to find these coefficients.

We decided to use a panel method solver called CMARC and its post-processing program POSTMARC for this analysis. The reason we chose this program is because of the time it takes to run a full three dimensional viscous case.

Fluent, the program we were using prior to CMARC, took several hours to run only a small fraction of one case while CMARC could run a full three dimensional viscous case in approximately five minutes. There would be a slight difference between the results obtained from CMARC and Fluent. This is because Fluent is a full Navier-Stokes solver whereas CMARC is a panel method solver where the accuracy of the solution is based on the number of panels in the model. The results, however, would be close enough to the Fluent solution to be useful.

To begin our CMARC model, we enlisted the help of a doctoral student, Liaquat Iqbal, who has had much experience working with CMARC and POSTMARC. We used his method of creating CMARC input files in Excel to create our model geometry. A sample of the input prompt is shown in Figure A.1.2.5.1 below.

Figure A.1.2.5.1: Sample CMARC Parameter Input

(Chris Strauss)

From Figure A.1.2.5.1, we can see that different design parameters of the launch vehicle such as stage length and diameter, nose cone length, skirt length, and (if a wing is present for an aircraft launch case) wing parameters can be easily changed to analyze the current configuration.

The launch vehicle has wings because this method was originally used to analyze launch vehicle configurations for an aircraft launch case. The launch vehicle would have a wing attached to the first stage enabling it to pitch into a vertical trajectory. After the first stage burned out, the wing would be discarded along with the first stage. This method, however, is flexible enough so that non-winged rockets can also be analyzed. We accomplished this by setting the wing span to zero.

After the parameters are entered, the Excel sheet is saved in a format that is readable in CMARC. The input file is then run in CMARC which creates an output file for use in POSTMARC. Once this output file is entered into POSTMARC, the pressure distribution and aerodynamic coefficients are found using the program’s aerodynamic coefficient calculation routine. The pressure distribution on a winged aircraft launched vehicle can be seen below in Figure A.1.2.5.2.

Figure A.1.2.5.2: Pressure distribution on winged air drop rocket at a 0 deg. angle of attack

produced in POSTMARC

(Chris Strauss)

As seen in Figure A.1.2.5.2, the pressure is at a maximum on the nose cone and the leading edge of the wings. This is as expected and thereby supports the accuracy of using CMARC and POSTMARC for the calculation of the aerodynamic coefficients. The model’s flexibility is shown in Figure A.1.2.5.3 below.

Figure A.1.2.5.3: Pressure distribution on wingless rocket at 0 deg angle of attack

(Chris Strauss)

Figure A.1.2.5.3 shows how the model can be changed easily and quickly. This model uses the same input sheet as the winged model except in this case the wingspan was set to zero to allow a wingless rocket to be analyzed. Again, the figure shows that the pressure distributions are as expected with the highest pressure on the nose cone and lower pressures along the rest of the rocket body. This again shows that the model is reasonable for aerodynamic analysis of the vehicle.

While this model appears useful when the preliminary cases are run, a major flaw is present. This is not a flaw in the model, but rather with the limitations of the CMARC/POSTMARC package. We find that CMARC calculations are only valid up to Mach 0.9. This effectively ends the use of CMARC as a primary CFD tool because the launch vehicle quickly achieves supersonic velocities after being launched. Had these supersonic cases been run in CMARC, erroneous results would have been obtained and jeopardized the integrity of the project.

Author: Chris Strauss