Autopilot ILS in storms and gusty wind

[Gregg_Seipp 565]

Hi all,

 

I'm tuning the ILS on an a largish GA twin engine airplane so that it holds the Glideslope better.  I've never flown an ILS RW, with or without an autopilot.  I've got two different configurations I'm working on:

• One setting has trouble seeking to the signal in gusty winds like thunderstorms...tends to overshoot, up and down with negative dynamic stability when it's really bad.  Once it gets onto it, however, it holds onto it like death.
• The other one finds the signal better, even in the storm, but doesn't hold it as tightly...prefers to drift along with the GS close but slightly off.

From your experience, which would be more realistic?  I've been hunting for videos but I haven't found anything useful to use.

 

Thanks for any info you can provide.

Gregg

[Gregg_Seipp 565]

Maybe I should have asked the question this way....

 

For real world pilots....in your opinion, which FSX/P3D airplane autopilot simulates acquiring and holding the ILS (especially the glideslope) the best? 

[kamaz74 81]

The default Bendix 140 is far more reliable in fsx than in any of the airplanes I flew which were equipped with it (2 glass C172SP's and an Arrow). Moderate turbulence and winds >10 would disconnect the AP if in APCH mode. Was far easier to hand fly an ils using the needles. I've heard newer STEC and Garmin units like those you get in the cirrus are far more reliable.

[Gregg_Seipp 565]

 

 

The default Bendix 140 is far more reliable in fsx than in any of the airplanes I flew which were equipped with it (2 glass C172SP's and an Arrow). Moderate turbulence and winds >10 would disconnect the AP if in APCH mode. Was far easier to hand fly an ils using the needles. I've heard newer STEC and Garmin units like those you get in the cirrus are far more reliable.

 

Hi Kylan,

 

Yeah, I can imagine all sorts of things.  In these smaller airplanes, if the trim motor can't keep up with the forces caused by the rapidly changing wind, it might be similar to a 'dutch roll' effect...a bad situation getting worse.  The AP disconnecting might be a blessing.  Unfortunately, unless it's built into the aircraft logic outside of the FDE, there is no disconnect logic.  It's going to keep trying even if there's a full deflection.  On the one hand, sometimes the worse the weather is, the more you could use the autopilot...on the other hand, there are some things an AP just can't do so you better be fully caffeinated and ready.  For my hardest testing, I was using the 'Strong Thunderstorm' weather in P3D.  Sometimes the AP could handle it...sometimes it really, really couldn't.  I just had no idea what it *should* be able to handle.

[gregmorin 199]

Gregg

 

In severe weather you do not want to chase the needles or altitude. Fly attitude when caught in thunderstorms with significant up and down vertical currents. The point of an ILS is to get you pointed at the runway, at altitude where you can make a successful landing and have enough runway to roll out. Fly wind correction headings and keep coming down to your decision height. Auto pilots will chase their settings and when  you are getting bounced around the oscillations will become larger and larger and can get you into serious trouble. When you see the corrections increasing in frequency and intensity go to hand flying. This is part of the 5% pilots get paid for in the old saying 95% boredom and 5% stark terror.

 

My two cents.

 

Greg

Gregg

 

When you break out look in the direction the localizer needle is deflected for the runway.

 

Greg

[Gregg_Seipp 565]

 

 

In severe weather you do not want to chase the needles or altitude. Fly attitude when caught in thunderstorms with significant up and down vertical currents. The point of an ILS is to get you pointed at the runway, at altitude where you can make a successful landing and have enough runway to roll out. Fly wind correction headings and keep coming down to your decision height. Auto pilots will chase their settings and when you are getting bounced around the oscillations will become larger and larger and can get you into serious trouble. When you see the corrections increasing in frequency and intensity go to hand flying. This is part of the 5% pilots get paid for in the old saying 95% boredom and 5% stark terror.

 

Thanks Greg.  Any thoughts on which one of your airplanes, in your opinion does it the most realistic?

[gregmorin 199]

Gregg

 

I hand fly 99% of my approaches.

 

I occasionally let the Flight 1 King Air, PMDG 737 and Eaglesoft Citation X handle an approach to 700' AGL, in relatively stable air. I think these aircraft set the standard for high performance commercial aircraft. I encourage you to try hand flying your approaches, it will make you a better pilot.

 

Greg  

[Gregg_Seipp 565]

 

 

I encourage you to try hand flying your approaches, it will make you a better pilot.

 

I handfly about 70% of them.  I'm just trying to make the autopilot on one more realistic.  It's pretty sloppy with the glideslope.  If your speed is 5 knots high it'll float above it...5 knots low and you'll sink below it. 

[gregmorin 199]

Does this aircraft do the same when you are hand flying it? Are you getting behind the power curve? If this speed is constant it should not matter. Think elevator to control your attitude and power to maintain the glide slope. If you are high reduce power slightly. If you are low increase power. I think you may be describing a speed and power control issue. it is important to look for trends and anticipate what the aircraft is going to do. If you add power to because you are low, you may have to reduce power once you have recaptured the glide slope. You are looking for the speed and power combination to maintain the glide slope. Think I had to add power to recapture the glide, I will probably have to reduce power once I recapture I think I will try slightly more power than what I set before but LESS the what was used for the correction 

 

What aircraft are you talking about is it a turbo prop? Remember you are not going to get instantaneous power increases in a turbo prop. In recip engines a power change of 1" manifold pressure will make a difference and give a slight correction in the glideslope deflection.  I find Real Air's Duke very realistic in this regard. My starting power setting is 22" once established on the glide slope and maintaining airspeed on the blue line. If I get slightly high I will reduce to 20" and the try 21" to maintain. It will vary with atmospheric conditions.  I am also a fan of Carenado's Navajo and Chieftain.

 

I hope this helps,

 

Greg

[FormerSF3 15]

I flew a PA31-350 with a Century III.  Your second option sounds like what it used to do.

 

 

The other one finds the signal better, even in the storm, but doesn't hold it as tightly...prefers to drift along with the GS close but slightly off.

 

It could fly the ILS, but on some days you had to "help" it.

 

It really depends how much $$$ the owner spends on the aircraft.  I flew some single engine that were rock solid.

[bonchie 2,305]

Most GA autopilots I've used don't handle turbulence well, especially in regards to altitude. Lots of chasing going on if the plane is hitting up and down drafts. This is mostly my experience with older STECs and Century stuff. 

 

If you are hand flying, it's easier to make corrections where as the AP may start to porpoise a bit chasing the altitude. 

[Gregg_Seipp 565]

 

 

It could fly the ILS, but on some days you had to "help" it.



It really depends how much $$$ the owner spends on the aircraft. I flew some single engine that were rock solid.

 

 

Most GA autopilots I've used don't handle turbulence well. Lots of chasing going on if the plane is hitting up and down drafts. This is mostly my experience with older STECs and Century stuff.

 

Interesting stuff.  This, at least, tells me how variable it can be.  In FSX/P3D, my experience with most aircraft other than heavy metal is that I end up having to help them some and I often just take them off and fly them myself.  When I find some airplanes on youtube (pretty rare), I see some look pretty well behaved.  I'll start paying attention to the autopilot models and how they do.

[FormerSF3 15]

I did notice this quote:

 

 

I'm just trying to make the autopilot on one more realistic.  It's pretty sloppy with the glideslope.  If your speed is 5 knots high it'll float above it...5 knots low and you'll sink below it.

 

Did you check what the max autopilot pitch settings in the aircraft template are?  I remember I had to adjust some of these when my freeware couldn't fly an ILS at normal Vref.

This is also out of the Century Manual:

 

 

5.

Upon interception of glide path (when deviation needle drops to center) the

Glideslope Coupler will automatically engage the indicator light will come on, and

the aircraft will assure a preset nose-down attitude for descent. Drop gear and re-

adjust power as necessary to maintain approach speed. Position flaps not to

exceed 15° for approach at this time . Do not change gear or flap setting after

initial glideslope intercept for remainder of glideslope coupler operation. Power

and airspeed changes must be made carefully to prevent excessive attitude

changes.

[Gregg_Seipp 565]

Did you check what the max autopilot pitch settings in the aircraft template are? I remember I had to adjust some of these when my freeware couldn't fly an ILS at normal Vref.

I did notice that and made an adjustment...also adjusted pitch velocity and acceleration.  Lots of little pieces go into it.  Part of the trick is not to overdo it.  If the real airplane couldn't handle it I wouldn't want the sim to.  On the other hand, from my experience, most sim aircraft don't do as well as their RW counterparts.  I see a lot of default values straight from the aircraft.cfg guide.  It's understandable though because there's little, if any, documentation as to what those variables do.

[fnav77 56]

You mention flying IN thunderstorms, is it really wise, lol?