The Rules state… “When two aircraft are approaching head-on or approximately so, both should yield to the right.”
But what if they’re slightly off to the right (from each other’s perspective)? Is it still a good idea to turn right?
How to deal with a situation when you need to fly directly southbound (180° mag track) or northbound (360°)? It’s impossible to keep a constant track of 360/180° and you will be off-course by +/-1° anyway flying in that direction.
The second question is about which FL do you need to stick with when you have to fly directly south or north (not about collision possibility for that matter).
igor wrote:
How to deal with a situation when you need to fly directly southbound (180° mag track) or northbound (360°)? It’s impossible to keep a constant track of 360/180° and you will be off-course by +/-1° anyway flying in that direction.
You stick to the level that corresponds to your DESIRED magnetic track, not the actual one. If your DTK is 180, then even+500, if it is 179, then odd+500.
igor wrote:
But what if they’re slightly off to the right (from each other’s perspective)? Is it still a good idea to turn right?
If there is no danger of collision then they don’t have to alter their track.
This whole issue with the semicircular rule is why I believe that randomization of cruising levels makes so much more sense than sticking to a discrete set of cruising levels and praying for you to see the other aircraft in time to avoid a collision.
The truth is that see and avoid does not work very well in practice. But SERA doesn’t care.
But I digress and this is an endless discussion. 
Alpha_Floor wrote:
The truth is that see and avoid does not work very well in practice. But SERA doesn’t care.
Well, we all know that but still everyone relies on it – not least the UK CAA! (I know about the CAA conspicuity initiative, but such equipment is expressly not for collision avoidance but for situational awareness.) An admission that see and avoid doesn’t work means that something else must be put there instead. Somehow I have the feeling that statistical arguments won’t be accepted (although that’s what we are working with in practise).
The semicircular rule is one of these many, many things in aviation that made sense at the time and in the circumstances they were designed for, but do not work in some of the circumstances the actually apply now.
The assumption underlying this rule is that
1. you have IFR / VFR traffic in the same airspace
2. IFR is separated from IFR procedurally or radar control
3. IFR traffic is mostly flying straight lines on ATS routes
4. VFR is separated from VFR by see-and-avoid
5. VFR is separated from IFR by flying +-500 ft from IFR levels.
This works reasonably well en-route in class E and above, and primarily prevents IFR / VFR conflicts. As long as IFR traffic is constrained to the ’000s, and as long as you believe that 500ft VFR/IFR separation en-route is required, there is no real alternative (random levels are not available, and a quadrantal rule would seriously reduce the number of available levels for VFR).
BUT it does NOT work in “free for all Class G airspace” without any air traffic service provision, because only (1) applies, while (2), (3) and (5) are simply not true – here a rule that does not distinguish IFR from VFR, with random altitudes, and/or a quandrantal rule are all objectively better from a separation point of view.
It also has been overtaken by reality, where there is very little IFR en-route traffic at the typical VFR cruising levels, so reserving 50% of the airspace in class E for <1% of the traffic is detrimental to safety.
But hey, were would we be if rules that were designed 60 years ago would be changed for something modern? In the 21st Century? No way, we will need at least another 50 years…
So your only alternative would be mandatory ADSB in and out as in the USA.
Do you think all aircraft owners in Europe will go for that?
igor wrote:
The Rules state… “When two aircraft are approaching head-on or approximately so, both should yield to the right.”
But what if they’re slightly off to the right (from each other’s perspective)? Is it still a good idea to turn right?
I do the same as I do with a yacht.
If it looks like it might become too close for comfort then I make a big, obvious manoeuvre which leaves the other party in no doubt whatsoever as to how I intend to put more distance between us.
igor wrote:
“When two aircraft are approaching head-on or approximately so, both should yield to the right.”
But what if they’re slightly off to the right (from each other’s perspective)? Is it still a good idea to turn right?
Depends of the rule is to clarify right of way way before getting to danger stages? or what you actually need to do in a head-on imminent collision?
Difference between “right of way” and “imminent collision”, is both pilots are aware of the conflict in the former they should have plenty for time to turn to the right and deconflict as per the rules of the air, for imminent collisions, I don’t think the answer is in ICAO Annex 2, surely one pilot has not spot the other (or none of pilots saw it comming !)
I saw few miss in gliders near o wing span (or I have been told after I landed), I am sure most of the time for an imminent collision one of the pilots is not moving at all, I can’t see how both move the same time? it’s only untill the firs one puts the wings on 90deg where the other can notice it…I doubt the practical problem has a very “very symetric” geometry & timing to question the validity of the right turn on first sight
