fdm['ic/gamma-deg'] sign switches

[fernandafenelon]

Hello everyone,

When I try to re-asign a value for fdm['ic/gamma-deg'], it stores the value with the opposite sign.

Assigning:
fdm['ic/gamma-deg'] = 4
Reading:
Print(fdm['ic/gamma-deg'])
fdm['ic/gamma-deg'] = -4

Does anyone know anything about it?

[seanmcleod]

So setting ic/gamma-deg doesn't store the actual angle as part of it's state, it actually uses the supplied angle to calculate a climb rate and stores that as part of the IC state.

Now note that the climb rate is stored in a NED velocity vector, where the Down axis is positive down, so note the minus sign in:

_vt_NED(eW) = -hdot;

 /** Sets the flight path angle initial condition in degrees.
      @param gamma Flight path angle in degrees  */
  void SetFlightPathAngleDegIC(double gamma)
  { SetClimbRateFpsIC(vt*sin(gamma*degtorad)); }

void FGInitialCondition::SetClimbRateFpsIC(double hdot)
{
  if (fabs(hdot) > vt) {
    cerr << "The climb rate cannot be higher than the true speed." << endl;
    return;
  }

  const FGMatrix33& Tb2l = orientation.GetTInv();
  FGColumnVector3 _vt_NED = Tb2l * Tw2b * FGColumnVector3(vt, 0., 0.);
  FGColumnVector3 _WIND_NED = _vt_NED - vUVW_NED;
  double hdot0 = -_vt_NED(eW);

  if (fabs(hdot0) < vt) { // Is this check really needed ?
    double scale = sqrt((vt*vt-hdot*hdot)/(vt*vt-hdot0*hdot0));
    _vt_NED(eU) *= scale;
    _vt_NED(eV) *= scale;
  }
  _vt_NED(eW) = -hdot;
  vUVW_NED = _vt_NED - _WIND_NED;

  // Updating the angles theta and beta to keep the true airspeed amplitude
  calcThetaBeta(alpha, _vt_NED);
}

Now the issue is that when retrieving gamma the sign convention with positive down in the NED frame isn't reversed, we should be passing:

-GetClimbRateFpsIC() to asin().

Or actually GetClimbRateFpsIC() should be flipping the sign.

  /** Gets the initial flight path angle.
      @return Initial flight path angle in degrees */
  double GetFlightPathAngleDegIC(void) const
  { return GetFlightPathAngleRadIC()*radtodeg; }

  /** Gets the initial flight path angle.
      If total velocity is zero, this function returns zero.
      @return Initial flight path angle in radians */
  double GetFlightPathAngleRadIC(void) const
  { return (vt == 0.0)?0.0:asin(GetClimbRateFpsIC() / vt); }

  /** Gets the initial climb rate.
      @return Initial rate of climb in feet/second */
  double GetClimbRateFpsIC(void) const
  {
    const FGMatrix33& Tb2l = orientation.GetTInv();
    FGColumnVector3 _vt_NED = Tb2l * Tw2b * FGColumnVector3(vt, 0., 0.);
    return _vt_NED(eW);
  }

[fernandafenelon]

I see.. If I use the flight-path/gamma-deg to retrieve the value in a trim scenario, it should work. Right?

[seanmcleod]

I've just submitted a pull request to fix GetClimbRateFpsIC().

But yes, if you take a look at how gamma is computed for flight-path/gamma-deg you'll notice that there is a minus sign for NED velocity component in the eDown axis.

gamma = atan2(-in.vVel(eDown), Vground);

[fernandafenelon]

Ok, thank you!