Strange behavior with Holding Pattern

[downscc 1,327]

I've always believed this to be correct behavior, the path prediction routine has not created a solution for the holding pattern. To support this theory, I point to how the magenta line will shift entering a enroute pattern as wind adjustments, speed, etc., are resolved. Often the patter gets larger or smaller as iterations to the prediction converge.

 

Too much fascination with the magenta line. I think it's pretty cool that the FMS actually does holding patterns.

[Alpha Floor 237]

I've always believed this to be correct behavior, the path prediction routine has not created a solution for the holding pattern. To support this theory, I point to how the magenta line will shift entering a enroute pattern as wind adjustments, speed, etc., are resolved. Often the patter gets larger or smaller as iterations to the prediction converge.

 

Too much fascination with the magenta line. I think it's pretty cool that the FMS actually does holding patterns.

 

High Dan. I also think it's pretty cool that the FMC does the patterns. It's a highly non-trivial maneuver. Just the fact that the aircraft has to choose between 3 different types of entries and then perform the entry correctly amazes me.

 

I don't agree with the "too much fascination with the magenta line" bit. This has nothing to do with the classic "Children of the Magenta" discussion, it has nothing to do with being able to fly the aircraft. It's just a bug reporting thread. If the aircraft does not perform as expected, it's good to report it so that PMDG correct.

 

Maybe the OP and me are wrong, but I want to believe that the real aircraft won't generate a "circular" holding pattern which gets infinitely big when the aircraft enters it. I did not wait long enough to see what happened, maybe LNAV did eventually fly the pattern correctly, I just couldn't stand it and reverted to HDG/TRK SEL in a display of "little fascination with the magenta line".

 

Regards!

 

How interesting. Glad you were able to reproduce it. 

 

Keep us posted and what PMDG says about this.

 

 

Thanks Mark! I will!

[777simmer 117]

No, the real airplane will not do that.

 

Or at least, it should not and I have never seen it.

But that does not mean it cant happen or will never ever happen.

Many things can be coded wrong and these bugs will then be corrected in the next database update.

 

We have had all sorts of bugs in the past......like the aircraft turning in the wrong direction (left instead of right) immediatetely after take off as the SID required a 180 degrees turn to a waypoint behind us.

 

The only thing you can do is revert to HDG mode, and trun/fly as required.

Then, usually when you put the active waypoint on top again, all problems are magically solved.

[Alpha Floor 237]

Thanks for your confirmation Rob! As I thought, this must be a bug and can't be something of the real plane.

 

 

 

The only thing you can do is revert to HDG mode, and trun/fly as required.
Then, usually when you put the active waypoint on top again, all problems are magically solved.

 

Which is exactly what I did

[kevinh 409]

Thanks for your confirmation Rob! As I thought, this must be a bug and can't be something of the real plane.

 

 

 

 

Which is exactly what I did

I tried this and it didn't matter what I chose for leg time or distance I got the large magenta circle. However the aircraft did fly the hold as  programmed, only the drawn route was wrong. I suspect the hold racetrack shown earlier is a generic representation. The circle that's drawn when the hold is the next waypoint is an attempt to predict the actual path, but it only shows a circle, not the straight legs that make it a racetrack pattern. The circle was the correct radius for the speed and bank angle in my case.

 

I'm sure this holding pattern path drawing used to work properly before so it may be a bug introduced in the last update. It will be interesting to hear what PMDG say in reply to your ticket.

[Alpha Floor 237]

 

 

However the aircraft did fly the hold as  programmed, only the drawn route was wrong.

 

I guess I wasn't patient enough! I was recording my tutorial and didn't want to experiment. On another occasion I might have waited to see what she did.

 

 

 

It will be interesting to hear what PMDG say in reply to your ticket.

 

I will keep you informed!

 

Regards,

[vr-pilot 11]

Maybe this happens because the holding entry logic switches from "direct entry with flying the first 180° turn" to "direct entry with bypassing the first 180° turn" (direct turn into outbound leg). Probably the relatively high speed of 250 kts IAS instead of 230 kts IAS (entry speed) while descending further 1000ft inbound the holding fix contributes to this conflicting behaviour...

[Alpha Floor 237]

Okay guys,

 

So I got a very polite reply from a member of the team, thanking me for my input and letting me now it will be looked after and fixed for a future update.

[nawarren 83]

 

I've always believed this to be correct behavior, the path prediction routine has not created a solution for the holding pattern. To support this theory, I point to how the magenta line will shift entering a enroute pattern as wind adjustments, speed, etc., are resolved. Often the patter gets larger or smaller as iterations to the prediction converge.

 

Too much fascination with the magenta line. I think it's pretty cool that the FMS actually does holding patterns.

 

 

Hi Dan, I appreciate your a beta tester, but I don't think this comment was very helpful. James had spotted a bug, wasn't sure and asked if this was correct behaviour, and your answer seems to imply that he should be grateful that it even does a holding pattern. PMDG pride themselves on their accuracy and attention to detail and I think LNAV and holding patterns are pretty important functions which should work correctly. I'm glad that they will look into fixing it, as if you fly online on Vatsim it helps if your A/C can track a hold.

 

Cheers

 

Neil Waren

[kevinh 409]

 

 

 

 

 

 

Hi Dan, I appreciate your a beta tester, but I don't think this comment was very helpful. James had spotted a bug, wasn't sure and asked if this was correct behaviour, and your answer seems to imply that he should be grateful that it even does a holding pattern. PMDG pride themselves on their accuracy and attention to detail and I think LNAV and holding patterns are pretty important functions which should work correctly. I'm glad that they will look into fixing it, as if you fly online on Vatsim it helps if your A/C can track a hold.

 

Cheers

 

Neil Waren

Neil,

 

The sim tracks the hold well enough. It's just the magenta line that is wrong. More or less as Dan said.

[downscc 1,327]

 

 

 

 

Hi Dan, I appreciate your a beta tester, but I don't think this comment was very helpful. James had spotted a bug, wasn't sure and asked if this was correct behaviour, and your answer seems to imply that he should be grateful that it even does a holding pattern. PMDG pride themselves on their accuracy and attention to detail and I think LNAV and holding patterns are pretty important functions which should work correctly. I'm glad that they will look into fixing it, as if you fly online on Vatsim it helps if your A/C can track a hold.

 

Cheers

 

Neil Waren

 

As a beta tester? You must have the wrong impression of what that is, we do not represent PMDG but rather are just hobbyists like you that for a multitude of reasons are lucky to volunteer our time testing products we love.  If there is a bug to report then please report it to PMDG Product Support in the form of a trouble ticket.That is the only way a bug gets validated and put into a bug tracking system to ensure that the next time PMDG allocates resources to the product the bug may get fixed.  I interpret the OP intention was to see if others were seeing the same thing, and my comment was not intended to discourage that but simply to explain partly what was happening. If it turns out that the circles are rendered incorrectly then that is good item to report to PMDG.  I've already reported numerous other odd rendering anomalies and encourage others to do so as well.

 

As far as VATSIM is concerned, or even RW ATC, you are not expecting to see racetracks on the radar scope if there is any wind.  Simple reason is that a standard rate turn in any wind does not have a constant arc.  In practice, you'll see anything from circles to ovals to racetracks that are squished at one end and expanded at the other.  For example, this is a holding point at KLGA taken from Flightaware just a few minutes ago:

[sebihepp 1]

I just asked myself something: You specified a 5nm legs. The duration of a 180 standard turn is 1 minute. If you fly faster than 5nm in one minute the fmc might ran into problems, because it assumes the 5nm not only for inbound track, but also for the turns? (which would be impossible at that speed)

 

If we think further, then the FMC thinks the turn is finished, because of the 5nm flown, but you only flew, let's say 160 degrees. Now the FMC will process the outbound leg, but instead of 5nm it will be only 4nm for example, the other 1nm is for the remaining 20 degrees because every turn have to be 180 degrees. The result will be the end of the inbound leg way before the waypoint. 

 

And now I have no idea why this results in getting bigger circles... 

[Alpha Floor 237]

I just asked myself something: You specified a 5nm legs. The duration of a 180 standard turn is 1 minute. If you fly faster than 5nm in one minute the fmc might ran into problems, because it assumes the 5nm not only for inbound track, but also for the turns?

 

I'm not sure if I understand you correctly, but the 5 nm are only for the inbound and outbound legs. It doesn't apply to the turns.

 

I think you're making it too complicated (no offence intended . Say you're inbound on a direct entry heading over a VOR, this is the sequence of events:

- Once overhead the VOR, the aircraft will turn 180º to the outbound leg.

- The outbound leg starts when you are ABEAM of the VOR. At this point you will have completed the turn, at least ideally.

- It will count 5 nm from the point in which you're ABEAM the holding navaid.

- Once you've flown 5 nm, it will start the turn inbound.

- You should level out of the turn being on the inbound radial and 5 nm out of the station.

 

Also once you override the standard-holding pattern "constraints", it's not a standard pattern anymore.

[Alpha Floor 237]

 

 

you are not expecting to see racetracks on the radar scope if there is any wind.

 

Very good point.

[vr-pilot 11]

 

 

The duration of a 180 standard turn is 1 minute.

Exactly that is true but the term "2 minutes standard turn" (for a full turn of 360 degrees = 3 degrees of heading change per second)  is more common. So if you fly holding patterns adjusting the leg time only applies to the inbound and outbound legs and not to the turns. Otherwise the turns wouldn't be standard turns. The "good old" turn coordinator (turn rate indicator) instrument was/is fixed to that duration.

In the end a standard holding pattern means being 4 minutes on the way before being above the fix again: 1 min. turn, 1 min.outbound leg, 1 min. turn, 1 min. inbound leg...

The rule of thumb for calculating the bank angle for a coordinated (with rudder input...) 3 degrees per second standard turn says: speed (TAS actually) devided by ten plus 7. E.g. (230kt/10)+7=30 (degrees bank angle). That is one reason why holdings are never flown faster than 230 kts, because it would result in uncomfortable banks for the passengers with high g-forces. An other reason is that it doesn't make much sense do fly a holding (waiting) pattern at high speeds.

Sometimes pilots request longer leg times (2 minutes or 3 minutes) to reduce the amount/frequency of turns in order to increase the passengers' comfort when holding for a longer time period.

The shape (ground path) of the 4 minute pattern (radius of turns and length of legs) differs a lot between that of a Cessna 172 and a Boeing 777, especially with high winds, where slow flying A/Cs are much more affected...