The following tips and suggestions should be used to guarantee the best possible fit for any data collected.

How Should the analysis proceed?

• The first sweep through the data should use the auto initialize option.
• The next parameter to calculate should be the Fast Arm Turning Rate (NCON7)
• If the projectile is spin stabilized the next parameter should be the Slow Arm Turning Rate (Fin stabilized projectiles have roughly the same turning rates for the fast and slow arm)
• Then finally solve for the damping rates (NCON3 and 4)

What type of fit error level is considered good?

• Depending on how accurate the data was read a good fit error would be roughly half of the error in collecting data (i.e. if the data can only be read to 0.5 deg resolution a good fit would be 0.25 deg error)
• Good fit error also depends on the magnitude of the yawing motion collected. If there is only small motion (i.e. below 2 degrees highest yaw) the analysis will have a hard time resolving accurate coefficients.

When do I know I've got the best fit to the data

• If after the second cut at the analysis, where you have calculated another of the parameters, the fit error does not substantially reduced (more than 10%) watch the error levels for the ACON coefficients themselves. If these start to increase any further analysis would begin to skew the coefficients derived.

The analysis has been run completely and the fit error is still larger than the error in reading the data.

• Look at the plotted data for bad data points. If found verify that the data was entered correctly. If entered correctly double check the Yaw cards and verify the data was read correctly.
• Verify that the initial conditions from testing such as MET and Muzzle Velocity are correct.
• Verify the physical properties of the projectile.
• Check to see if the yaw motion has enough magnitude or that there is enough data per cycle.

In general, the objective is to achieve the best fit to the data with the least number of unknowns (examine uniqueness of solution).

Previous  Top  Next

© 2002 Arrow Tech Associates