F-14 AoA Indexer

18 Mar 2020 08:23 #45549 by timi
I played with Richard's F-14B yesterday and found the AoA indexer had a mind of it's own.
Took a look at the config and I think the values of the indexer are setup with AoA units
instead of degrees.

If you want to test the config please edit f-14-common.xml with the following.
It should give you the amber donut at 130 kt at 50000 lbs gross weight with
full flaps and gear down.

<aoa-indexer>
<!-- target = 11.3 deg AoA (indicated allowing for offset in AOA sensor-->
<!-- F-14AAD-1 Figure 2-124; Angle-of-Attack Displays -->

<slow-deg-start>12.96666666667</slow-deg-start>
<optimum-deg-start>7.96666666667</optimum-deg-start>
<optimum-deg-end>14.63333333333</optimum-deg-end>
<fast-deg-end>9.63333333333</fast-deg-end>

<target-deg>11.3</target-deg>
<!-- target (AFCS)-->

</aoa-indexer>
The following user(s) said Thank You: StuartC, Avionyx

Please Log in or Create an account to join the conversation.

18 Mar 2020 15:21 - 18 Mar 2020 15:25 #45576 by Richard

timi wrote: found the AoA indexer had a mind of it's own.


I'm offended by that as the F-14 AoA indexer is written to work as per the flight manual; i.e. calibrated in AoA units because that is what the aircraft uses[1].

I'm not clear why you're changing this as I've spent many hours testing and fine tuning the handling with full flaps to meet stated figures and observed characteristics. My goal is to provide an accurate simulation based on real data and avoid just making stuff up.

At 50,000lb, 130knots, AoA should be 15degrees, and descent rate around 750ft/min.


[1] AoA units are slightly higher than alpha (fuselage reference line)
Attachments:

Please Log in or Create an account to join the conversation.

18 Mar 2020 15:26 #45577 by timi
But it's not 15 degrees. Your aircraft flies at 11 degrees AoA at that configuration and the indexer should reflect that it's safe to fly in that configuration and thus showing an amber donut. The target of 15 degrees will end up in a stall.

Please Log in or Create an account to join the conversation.

18 Mar 2020 15:37 #45578 by timi
I mean this is the relation between degrees and units.


And I just replaced the numbers to represent the same scale in degrees alpha.
And to me the indexer gives sensible guidance after that.
Attachments:

Please Log in or Create an account to join the conversation.

18 Mar 2020 16:02 - 18 Mar 2020 18:28 #45579 by Richard
Your change is wrong because the AoA indexer is using a property that is already in "alpha units" (not degrees) - orientation/alpha-indicated-deg (alias of /fdm/jsbsim/aero/alpha-indicated-deg) using the data from NASA TM-81833, page 56, Figure 17.

See NATOPS NAVAIR 01-F14AAP-1 page 2-176 "Angle of attack conversion" ( my version of this chart ) - for the conversion table for mass and airspeed - this is where the figure out 130kts at 50,000lbs comes from.

Yes it is hard to fly the F-14A/B onspeed and do a carrier landing - but to the best of my knowledge that's just how the aircraft is.

If you can provide data or references that shows my aero model is wrong then I will look into it.

Please Log in or Create an account to join the conversation.

18 Mar 2020 16:39 #45580 by timi
In that case I have made a mistake with the property and will go practice
some more to see where the issue is. Likely between my ears this time. :)

Please Log in or Create an account to join the conversation.

18 Mar 2020 22:03 #45586 by timi
Ok, I did some testing and I know what causes issues for me.

On a test flight I measured the change of /orientation/alpha-deg
versus the units on the HUD and got the following.


And based on those numbers there seems to be a one to one
correlation between degrees and units. That is a change of one
degree seems to equal a change of one unit which to me feels
too tight of an angle to produce useful feedback for flying.


If a probe rotates from -10 degrees to 40 degrees to produce a
scale of 0-30 units, then I don't understand how there could
exist a correlation like that. Even if the change isn't completely
linear, surely the degrees should change more than the units,
no?

An angle so tight kind of leads to either unnecessary chasing
of the donut or ignoring the indexer altogether in my opinion.
Also there was just a 1 kt speed difference between slightly
slow and slightly fast which again is too tight to produce a
meaningful response in my opinion.

So I originally thought the indexer is messed up because I
could fly "on speed" without the donut.

So the wrong angle I setup still gave me more useful info
because it was wider and not some blip within a single
degree. And these are my thoughts and experiences not
meant to offend anyone.
Attachments:

Please Log in or Create an account to join the conversation.

26 Mar 2020 19:06 - 26 Mar 2020 19:17 #45740 by Richard

timi wrote: Ok, I did some testing and I know what causes issues for me.

...based on those numbers there seems to be a one to one correlation between degrees and units.

Sorry for the delay in replying I've spent most of the last week working on this.

The first thing is to ensure that you're flying the approach correctly; full flaps, trimmed to about 8 degrees of pitch, rate of descent around 300 to 500 fpm, using throttle to adjust descent rate and pitch to control speed. Since HeatBlur released their F-14 there's been a lot of information that has come to light - videos and I found a nice handy Case1 guide for the F-14 docs.google.com/spreadsheets/d/1cEkJW9bP...jCa9wet8s/edit#gid=0

Firstly I've looked at my data sources for the alpha probe; and I cannot find any other data apart from the data from the report that I have; so I have to assume that the figures are correct.

After some flight testing I've come to the conclusion that the handling is worse that it could be; even with lots of practice and care it is very difficult to land, more difficult that it could reasonably have been as the margins between flying and a loss of control are just too tight.

So I've revisited all of the relevant literature and come to the following conclusion;
  1. Indicated alpha is probably about right
  2. Approach speeds too high because available lift is possibly a little low
  3. Too much pitch because of (2)
  4. I've not been testing using the speedbrake; this helps to keep the engines turning faster and therefore more responsive.

For comparison I used the HB F-14B which has been tested by ex-Navy F-14 pilots and thus is probably closer to the real thing than what I have. The handling of this F-14 has a lot more lift; it has about the same alpha stability and pitching moment, however required elevator is lower and for a similar configuration speed is also lower. My view is that confirming that (2) is probably the root cause; i.e. not enough lift with full flaps and that the best approach is to tune the lift and then balance out the rest of the coefficients.

HB approach figures as measured[1]
  • Pitch angle ~8.5 degrees
  • Elevator ~9.5 degrees elevator
  • Alpha ~15 units indicated
  • Speed ~124kts
  • Fuel ~6000lbs
  • Config External tanks only
The changes I've made based on the investigation both static calculations and using OpenVSP derived data gives the following results
  1. C.G moved to AERORP; this is consistent with aerodynamic theory
  2. Airspeed / lift showed that a factor of ~1.5 (different values for main and auxiliary flaps) gives better speeds.
  3. Drag adjusted to give reasonable power settings; factor of ~1.3
  4. Pitching moment too much nose down, conclusion is that the main flaps must reverse when full flaps are deployed otherwise far too much elevator is required (~15 degrees)

I've pushed a development version of the F-14 if you'd like to test it to: github.com/Zaretto/fg-aircraft/tree/2020-flaps-tuning
--
[1] HB F-14 Case 1 approach; trimmed and stable
The following user(s) said Thank You: timi

Please Log in or Create an account to join the conversation.

26 Mar 2020 20:43 #45741 by timi
Well, wow! Thanks for your efforts Richard. I'll try that version. :)

Please Log in or Create an account to join the conversation.

27 Mar 2020 00:30 #45742 by timi
Flies nicely. Although I still have issues with the indexer but knowing my weight
I fly by airspeed and the velocity vector which brings me close to on-speed for
landing.

Please Log in or Create an account to join the conversation.

27 Mar 2020 15:51 - 27 Mar 2020 15:51 #45743 by Richard
Thanks for testing; I'm currently working on the AoA indexer again - trying to work out if what I have got is right or not.

Please Log in or Create an account to join the conversation.

27 Mar 2020 16:26 #45745 by timi
I saw some short glimpse of the donut in DCS as well while looking at some YouTube video. Didn’t even see the slow/fast lights. I don’t know... But other aircraft don’t seem to have an angle that tight.

For example the Hornet has 1.4 degrees between the slightly slow/fast and I have no problem to keep the donut visible with the trim buttons with that.

Although the F-14 does seem to trim less per push of the button in comparison.

Please Log in or Create an account to join the conversation.

28 Mar 2020 21:26 #45750 by Richard
I've just pushed some changes.

* indicated alpha (units) tuned based on handling and observed handling of HB F-14
* Completely new yaw damper that is more effective at controlling the dutch roll. This is based on description of the F-14B DFCS/SAS
* Tuned lift and pitch moment due to flaps=35 based on handling characteristics.

changes pushed to github; please test if you can.
The following user(s) said Thank You: timi

Please Log in or Create an account to join the conversation.

28 Mar 2020 21:37 #45751 by timi
Awesome! I'll try. Hopefully the DCS model is decent enough
to compare against.

Please Log in or Create an account to join the conversation.

29 Mar 2020 08:50 #45752 by timi
I think the indexer works as I would expect now. Thanks for looking into it.
Now I just need to learn how to fly a proper glide slope to a carrier. I
suppose a carrier with a meatball would help but I wanted to try on the
HMS Queen Elizabeth first. :)

There are some slight visual issues with the HUD and the vertical display
indicator when using FGCamera. Not sure why the VDI has stuff missing
from the display when you move your head. But I do understand that the
HUD is a 2D instrument.

The following user(s) said Thank You: Richard

Please Log in or Create an account to join the conversation.

05 Apr 2020 15:27 #45855 by Richard
The issues with the HUD and VDI are because these aren't using Canvas; the HUD is an XML HUD positioned dynamically based on the view and this isn't 100% accurate. The VDI can suffer from Z buffer problems because of the transparency.

Also I've just pushed a rework of the Yaw SAS - as you are chief test pilot for these changes I'd appreciate it if you could do some testing and report back.
The following user(s) said Thank You: timi

Please Log in or Create an account to join the conversation.

05 Apr 2020 19:03 #45859 by timi
Awesome! Ok, I can try. Although I wouldn't know how exactly it's
supposed to behave so I can only make generic comments.

I've spent some time this weekend with some sounds for it.

I have sounds for:
- In-cockpit jet engine whine.
- The same with the canopy open.
- Gear wind.
- Gear rumble.
- Wind.
- Catapult launch.

Not all of them make an appearance but here's a video anyway.


I think the cockpit might be slightly too loud compared to
external sounds with these sounds though but that can be
adjusted.

I can share the sounds if you find them worthy for inclusion. If not,
I'll just keep them for myself. :)

Please Log in or Create an account to join the conversation.

05 Apr 2020 19:30 #45860 by timi
I analysed this guy's videos for the sounds.

Please Log in or Create an account to join the conversation.

06 Apr 2020 13:28 #45875 by timi
Richard,

the latest update I see is this:

2020-emesary-armaments-MP-dev
Updated 4 days ago by Zaretto

Please Log in or Create an account to join the conversation.

07 Apr 2020 12:57 #45880 by Richard
Firstly I forgot to push changes. sorry.

timi wrote: I've spent some time this weekend with some sounds for it.

I have sounds for:
- In-cockpit jet engine whine.
- The same with the canopy open.
- Gear wind.
- Gear rumble.
- Wind.
- Catapult launch.


I think in the video you link what you are hearing is the air conditioning as I've read that the aircon was very loud. Also it's hard to be certain what is right from a video because of the way that audio recording devices work. It sounds too loud - I tried to adapt based on wearing a helmet based on my experience but it is my opinion.

I spent a good amount of time on the engine sounds - I tried to use samples adjusted but to my ears that only ever sounds right when at a constant rate - the accel/decel just sounds wrong. So instead I adopted the same technique as used on certified simulators - which is to split the sound down into components and that's basically N1, N2 whine, Air flow at the inlet, and exhaust.

I also use the sounds to provide cues that you could get by 'feel' in the aircraft - so I thine the N1/N2 turbine whine being audible is quite important.

Comparing with the HB-F-14 cockpit sounds what I have for engine is fairly similar - possibly a bit high pitched. Again because they had feedback from ex-navy pilots I tend to think that what they have is probably close to what it should be. The external sounds are definitely close to what I've heard in videos.

The rest of the sounds - it's hard to tell because of the cockpit background noise - if you think it's better than I will take a look at any files/pull requests.

Please Log in or Create an account to join the conversation.

Time to create page: 0.496 seconds
Powered by Kunena Forum

Latest Forum Posts

PM Notifications

You are not logged in.

PM Mailbox

You are not logged in.