Calibrating a POWERSEA Model with Tank Test Data
------------------------------------------------

If you have tank test data, and especially if you have data at more than one point, then start by building a model that matches the measured heave vs. speed and pitch vs. speed characteristics, and sometimes effective force vs. speed.

1. POWERSEA does not adjust the "transom force drop-off coefficient" automatically, so start by setting that coefficient to 0.0.

2. Relative heave vs. speed:  Run the hydrostatics command and set the initial conditions using this command. Copy the initial heave value over into the calm water or static heave box in the Initial Conditions dialog, so that you can plot "Relative Heave," which is more useful that absolute heave.

3. Run a speed vs. power analysis: Ramp the speed from 500 seconds to 3000 seconds or so, so that any transient acceleration effects die out.

4. Compare the pitch and rel. heave at a couple of speeds (Fb=2.0, Fb=4.0, for example).

   a. If the pitch is too high (and you still have the transom force reduction set to 0.0), then reduce the lift created by keel camber. You do this by reducing the "Geometric Lift Coefficient." The model for lift created by keel camber assumes that the 2D approximation is valid, but up near the bow it really isn't, so this factor adjusts the influence of keel camber. You should click the radio button labeled "Manual" in the Set Coefficients box so that POWERSEA doesn't override your "Geometric Lift Coefficient" setting. Rerun the speed vs. power simulation, and adjust this value until your results improve.

   b. If the pitch is too low you could either increase the "Geom. Lift Coefficient" as in the previous step, or you could enter a non-zero value for the "Transom Force Reduction Factor" -- this is better, as it doesn't affect the dynamic response very much (e.g., vertical accelerations).

5. If you have measured resistance and want to adjust the model for that, you can adjust the "Geom Drag Coefficient" as in step 4a.

Iterate through these steps a few times before you will settle on an engineering compromise.

The built-in values of the coefficients in POWERSEA assume that your keel and chine line aren't artificially adjusted to include extra hull at the bow to model the effects of flare. That is, one of the inputs to the regression analysis is the relative position of the LCG vs. the center of planing area, expressed as a fraction of the projected length of the planing area. If your planing area is artifically extended by having the chine wrap up to the bow, then that ratio will be a little off. It doesn't make too much difference in the empirical coeffs, but it does make some. If you want to play with this, you could create two models, one with the actual chine line, and a second with the effective chine line with flare. You can copy the automatically calculated empirical coeffs from the first model into the second model, overriding the automatic calculations.