The purpose of a hold is generally to delay in a confined airspace where you will be deconflicted with other IFR traffic and nearby terrain and obstacles. To remain in the same place, you need positive course guidance and this is provided for on the inbound leg, so in theory, you have a means of intercepting and tracking back to the holding fix on the same course. Flying a continuous circle would not provide for a means of navigating to hold the station. The two straight legs allow for wind correction, so from a practical standpoint, an oval like pattern makes sense.
LeSving wrote:
VFR holdings are usually circles
Because a circle VFR in VMC around a given point is much less of a challenge than a hand flown DME-Arc around a NDB…
Maybe it helps to look at the ‘racetrack pattern’ from a different perspective.
Don’t look for a racetrack. Look for the single best way to stay close to a specific nav aid, in bad weather, that allows for wind correction, that gives you a maximum time of level and straight flight, and takes up the least geographical space to avoid terrain or protected airspaces in the vicinity.
What you get is two straight lines, opposite of each other, connected by standard rate curves. Voila – the racetrack pattern. It’s the aviation way of walking a street up and down.
Airborne_Again wrote:
So the AI would eventually show straight and level although the aircraft is turning.
On a second thought, is this correct? I don’t think so. In a coordinated turn, you have a constant rate of pitch as well as constant rate of yaw. A mechanical gyro in an AI doesn’t “measure” anything, it’s simply a gyro on a gimbal. It will seek to stay put in relation to the earth (or universe) independent of the aircraft attitude. For it to do so in relation to the surface of the earth, the spinning axis has to be vertical, pointing to the center of the earth. If the axis is off, it will start to show that error as rates of pitch and/or roll whenever the aircraft is not flying straight and level.
There is still an unbalanced force, in addition to gravity, that causes the angular acceleration, i.e. turn and also that force moves the ball into the centre, away from exact vertical.
I think if you fly a coordinated turn with a normal AI, for ever, eventually it will indicate wings-level flight. The erection mechanism, whatever it is, must eventually overcome the gyro.
It’s not like you are caging it, and both yaw rate and pitch rate will affect the gyro continuously.
Yes; it appears I was wrong above. This explanation suggests that an AI will continue to show the roll angle during a long coordinated turn – because the downward G vector is continuously rotating (and in effect painting a circle on the earth’s surface).
Aha. Of course. The aircraft is literally turning around the gyro (important to remember). In a 180 degree turn, there will be an error of 3-5 degrees due to the erection. Continuing the turn another 180 degree, and the erection mechanism will move it back to where it started.
If you start north and turn west, then the gyro will be tilted in the direction of the (average) g forces which is west. When heading north again, the average g forces will be east, and thus pull it back in.