Greenie Board Possibilities?

[SpazSinbad]

OK - found the same symbology 'fly up / fly down through glideslope' in the 2009 LSO NATOPS - missed it on the first quick squizz.

fsxnp, it seems you use the APARTS symbology. Better to use LSO symbology? Or is the APARTS easier to use for computer I guess?

This is where the LSO knowledge can say about 'degree of deviation' to record events. As you say the approach is 'around' the actual glideslope so the LSO not only records the actual deviations but the movement to other deviations. I think the phrase implies the 'grossness' of the correction from one deviation above the glideslope to end up below the glideslope - for example.

It a pilot anticipates the necessary corrections (this is ideal) then if for example he is above the glideslope by a small amount it is possible to make a correction down then correct with an UP to stop the DOWN then again anticipate with a smaller UP then smaller down to effectively go down to the glideslope but not go below and then to remain on glideslope. In practice the aircraft deviates by a small acceptable amount which gets smaller as the aircraft/pilot sees a more accurate glideslope - all going well.

To me the phrases indicate that the corrections were not refined - therefore noteworthy - gross - too much. Not sure how that can be modelled except in comparison of 'what to what' if the aircraft is high then goes low through the glideslope by an agreed margin - then this is noteworthy. The difference between the too high to the too low by flying through the glideslope is a big change (and still in error).

[neutrino]

Serge, I think for the "up/down", if I interpret it correctly, you can simply monitor deviations from a calculated "ideal" vsi. For example, let's say from given {wind/wind direction/carrier speed/carrier course}, you calculate an ideal vsi of 720 fpm. If in the groove you register a vsi of 400 fpm - this means that the jet is going too fast UP towards or through the glideslope. Or if you see a vsi of 1000 fpm - he is flying DOWN through the glideslope quite fast. On the other hand you can have a hundred corrections within 650-800 fpm - that's a really smooth approach.

For example check my landing here - my vsi stays within 670-740 fpm for the entire approach, which is +/-5%. This is as smooth as I can get  

Even more accurate will be to divide the vsi deviation by the distance - this will give you the angular speed through the glideslope. For example, you can descend 1500 fpm at 1.5 miles from the ramp, but this will be much less dangerous than descending at 800 fpm at 0.1 miles before the ramp (provided your "ideal" vsi is 700 fpm). If you are looking at the two jets through the PLAT cam - the jet at 0.1 miles will be descending twice as fast than the jet at 1.5 miles. A third jet at 700 fpm will appear stationary on the PLAT cam at any distance.

N.B. A pilot's job is not to maintain any pre-determined vsi or even look at the vsi indication! He need only maintain AOA and glideslope (ball/lineup/AOA). The vsi is just a means to evaluate the corrections relative to the glideslope with a computer.

[SpazSinbad]

A new 'how to deck land' PDF now 1.5GB (smaller without some large embedded videos) is available:

http://www.a4ghistory.com/_HowDeckLand24mar2011pp1711.pdf

Keep in mind this information has been taken from the main 4.4GB PDF so it is skewed towards the interests of that PDF - the A4G Skyhawk and various aircraft in the RAN FAA. However you will be surprised at the variety of aircraft mentioned - including FSX aircraft talked about in this forum - as well as a bunch of material about FCLP - not just carrier landings.

[SpazSinbad]

neutrino, your idea about descent rates is good but needs some further refinement IMHO. I'm not sure how it can be achieved so I'll describe what is meant.

Carrier Pilots are not SMOOTH as such. This is not an Air Force Style landing. In a much earlier thread now I recall making a very bad approach video for illustration purposes but in the end I did get there (I think). In a roundabout way this video perfectly illustrates in a gross way how a carrier pilot flies the ball (not that anyone in the real world would be allowed to fly in such a dangerous manner - the LSO would wave them off before they got too close).

It is legitimate to make large corrections a long way from arrest so as to get 'on the ball' as quick as possible. The closer one gets when 'on the ball' corrections necessarily should be much smaller ideally but the point is to 'fly the ball' and be 'on the ball'. Smoothness is not a benchmark. Flying the ball accurately is the benchmark hence the often large elevator movement of a Hornet as it crosses the ramp - to keep the ball where it is - rather than have it move - in this situation. So at the ramp there is not a requirement to change the ball position but keep it in the mirror which apparently will guarantee an arrest. Ideally though the ball is centred all the way to touchdown.

Again my point is the pilot flies the ball to the best of his ability and does not worry about how that looks to others or how smooth that process may be. Generally speaking it is NOT a smooth process but may look smooth to outside viewers. Certainly from the inside is it NOT smooth. The power goes up and down rapidly and control movements can be gross with both inputs becoming smaller hopefully the closer the aircraft gets to touchdown.

As I say I don't know how that can be modelled in the sim but want to stress that smooth is not the goal. Keeping the ball within acceptable parameters is the goal - with accuracy paramount. Any emphasis on smoothness can mean that appropriate, timely corrections are not made. For example a rampstrike can occur from a high start where enough correction is not made so that the aircraft is 'high all the way' but 'coming down at the ramp' for a rampstrike. This can be because the pilot is not making enough correction early on to get back to keeping the meatball centred, or perhaps he/she is trying to be 'smooth'.

A carrier approach can be described as an endless series of corrections that get smaller the closer the aircraft gets to arrest. The ball can remained centred with appropriate constant corrections - this is not smooth Air Force flying.

http://www.a4ghistory.com/VikingS3longCarrierApproach.wmv  (2.3Mb) TURN UP VOLUME TO ELEVEN!  

There is a classic video of an S-3 Hoover/Viking doing a long approach with engine sounds heard in the cockpit. Air Force people remark on how the engines are cycling up and down. Yep - that is the key - there is nothing smooth about a carrier approach although it may look smooth. With experience a carrier pilot can anticipate ball movement and appropriate corrections to keep the ball centred, however there is a constant input of corrections with anticipated corrections for the earlier corrections and on and on becoming ideally smaller as the aircraft approaches arrest with the ball in the middle of the datum lights.

Bear in mind that carrier approaches use the same techniques for all jet aircraft. Keep the ball in the middle of the datums using 'meatball, line up and airspeed (short for 'Optimum Angle of Attack') as your only guides (not smoothness - not deck spotting). This 93 Mb PDF with embedded videos mentioned in an earlier thread is useful:

http://www.gamefront.com/files/17349862/DeckLandingInstructionVideos_pdf  (93Mb)

Very good Clemenceau Approach Video BTW!   

[neutrino]

Spaz, you may have a point - smoothness is not a priority, staying on glideslope is. The pilot should be penalized only based on the size of the deviations from glideslope, but he should be able to correct them as fast as necessary. In this case, I don't know how to interpret the symbology that puzzled Serge

[SpazSinbad]

neutrino, IF the pilot is 'smooth' by skill, technique and anticipation (of corrections) to remain on the glideslope accurately then all well and good. But the prime aim as you say is to be on glideslope - even if it looks ugly, and there have been a lot of UGLY approaches out there which turn out OK.  

Personally I think the suggestion to 'score' the approach by pilot accuracy through the boxes indicated is a good one with the 'flying through glideslope' being secondary. Within reason - whatever it takes - to stay on the ball is good. There are provisos/caveats to these kind of catch all statements though, especially in close. I'll hunt out some of them. The earlier PDFs highlight some considerations for the Hornet for example.

The first graphic is from a USN LSO Brief about the difference between flying FLOLS and NEW IFLOLS.

[SpazSinbad]

LSOs 'need to watch' guidance from:

NOTE: There is a limit to what can be done on any approach. Always best to start on the ball and stay there.

OUTER-LOOP CONTROL FACTORS FOR CARRIER AIRCRAFT
Robert K. Heffley  1 December 1990
http://robertheffleyengineering.com/docs/CV_environ/RHE_NAV_90_TR_1.pdf

"LSO View of Outer-Loop Control
Table 2-2 presents a list of outer-loop control factors from the LSO's vantage point
(Reference 45). These are useful in evaluating aspects of the task and of the aircraft
which may be crucial to success. A number of these items are concerned with where on
the final approach corrections can be made, especially when engine response is a factor.
According to this table, LSO's exercise may more caution with corrections from a
high glideslope deviation than from low. Also, the aircraft should be stabilized on the
approach by the "in-close" position (about 1/4 nm range).

Table 2-2. Outer-Loop Control Factors
Profile:
• More ramp strikes occur when the pilot is correcting for a high deviation in-close
than for a low deviation. - Living Proof :-)
• For significant multiple deviations in close, a waveoff should be used by the LSO.
As a rule of thumb, if 2 major deviations (from among GS, LU, AOA or power)
are AFU approaching the waveoff point, use waveoff. This is especially critical
with a CQ pilot.
• For unsettled dynamics (speed, power, wing position, flight vector, pitch) in close,
the LSO should consider giving a waveoff.
• High at the ramp with less than optimum rate of descent can lead to a dangerous
long bolter. Do not hesitate to use waveoff.
• High at the ramp with excessive rate of descent can easily result in a hard landing.
• LSO should never accept a low trend on an approach.
• Be prepared for sink rate increases during late lineup corrections.
• LSO should not accept a high trend on an approach.
• Poor trends leading to the start and at the start are good indicators that the pass is
going to be a problem due to pilot disorientation or poor pilot scan.
• A poor start frequently leads to overcontrol tendencies in the remainder of the pass.
• Be alert for the "moth effect" (drift left in-close or at-the-ramp) due to pilot fixation
on the meatball at the expense of lineup control.
• During day recoveries, beware of pilot tendency to try to salvage an extremely poor
start (i. e., OSX, NESA HFX, HFX, etc.). If not stable approaching in-close position,
use waveoff.
• A major glideslope deviation at-the-start to in-the-middle is difficult for the pilot to
salvage. Extra LSO assistance may be needed to help pilot get aboard.
• If calls are necessary for aircraft with slow engine response (A-7, S-3, F-14),
they must be given well prior to glideslope interception when correction is
being made for a high deviation.
• For aircraft with excellent engine response( A-6, EA-6, F-4), be alert for pilot
overcontrol of power. This also includes excessive power reductions following
too much power."
____________________

From same source above - characteristics of aircraft for LSO guideance:

Table 2-3b. Carrier Landing Features of Existing Aircraft—LSO View.
F/A-18:
Excellent power and waveoff response..
Flat attitude when on AOA.
If back on power and cocked-up, SIC-AR is probable. [Settle In Close - At Ramp]
Easy to over-rotate on waveoff; in-flight engagement potential.
Nose adjustments must be coordinated with power changes to get glideslope correction results.

[Sludge]

Serge...

Been working on some really good Sludge mods, including a seriously RE-VAMPED .air file and updated effects.  Also, working with JIMI and Shylock, w/some of their mods.  But, I read up on your posts and you have a really good start on this "AI LSO" mission add-on.

From what I've been reading, my opinion would be to use your own IFLOLS gauge for measuring "coming up/down thru the glideslope" criteria.  Using the IFLOLS gauge as a basis of "on glideslope" (2 middle/centered balls=baseline); +/-1 ball deviation (little high/low), +/-2 balls deviation (high/low), 3+ balls deviation (very high/low).

Take a measurement (or meas. of deviation) for each phase in "the groove" (start [X]; IM; IC; AR) til touchdown, and have it stored for assessment after the trap.  An example would be: X = +2 (high); IM = 0 (on glide); IC = -2 (low); AR = -1 (little low); IW [trap gauge] = 2 (wire), 145 (fast), 6.8 (fast, little overpowered), -550 (Rate of Descent flattening).  For glideslope errors ONLY, I would grade this as: "High at the start... coming down thru glideslope in close... little low at the ramp... flat into the wires... fair 2."

I think this is what you're looking for?  Some type of basis to start an "AI grading criteria" that you can use for your mission?  Hope this helps.

Later
Sludge

[Paddles]

Thanks, guys

Sludge, I want to make a program that not only collects measurements for later assessment and grading, but also (and mostly) monitors and gives correcting advices 'live' , as a real LSO does. The data collected may be then accessed for trend analysis in an APARTS-like form or something. That's my purpose.  
Besides, to use my IFLOLS one needs to tune to an ILS frequency...

And here's why I asked what was the purpose of those / and \  symbols. In this example my aircraft starts low, then overcorrects and flies up through the GS, being high in the middle. Could I describe this situation either as 'LOX HIM' or 'LOX /' ? My guess is that the first is more informative, it tells WHERE the aircraft is in respect to the GS. The latter tells only WHAT happened. Are those interchangeable or not? Should I stay with the first and not spending my efforts on implementing more complex programming logic of the latter?



And thanks again, I appreciate everyone's help and this ain't my only question (I guess)  

[SpazSinbad]

Perhaps what fsxnp wants to achieve is at odds with what the LSO wants to achieve. I'm not sure about that issue myself. On the previous page there is this quote:

"1. Grading Passes and 2. Shorthand and 4. What is the start, in close...
101 Remember the main reason we grade passes is so a pilot can learn from every pass and the Pilot and LSO can spot trends which need correction. My philosophy as a CAG LSO was not "lets see how much detail we can go into" it was more "hit the high points and tell him what he saw."

Tell him what he saw. Not what you saw. Important distinction. If you saw him low, but he saw a centered ball because he was slow and cocked up, then you would say slocu "slow cocked up". Then in the debrief when he doesn't see the low you saw, you can explain the airspeed deviation maksked the glideslope deviation...."

If the purpose is to be like the trend analysis then that seems to be more work than necessary. So what is the aim again? Is it to grade passes and leave it that or go to the nth degree as fsxnp seems to suggest. I'm asking for clarification of aims - no criticism implied.

[GOONIE]

Spaz, I agree, the real world LSO debrief is designed to help the pilot learn.

The original intent of this thread was to see if creating a virtual "Greenie Board" was possible, to hopefully create some fun competition on this forum (or others) amongst the virtual F/A-18 pilots. Similar to the "greenie board" competition in the real world.

In order to do this correctly, a scoring/grading mechanism needs to be created based on the specs provided by NATOPS and the other details Serge is trying to understand and nail down. It sounds like Serge is trying to create a very capable virtual LSO that not only scores your approach (including storing and collecting multiple traps; using real world shorthand and grades), but also gives you guidance during the approach (similar to the TGS mission LSO calls). By creating a new mission and AI LSO, several things can be corrected or made better (when compared to TGS), correct aircraft carrier speed, better approach scoring zones, and ability to create a greenie board (real world short hand and grades), etc.  This would be a huge improvement in my opinion.

[neutrino]

Right, I think even if the program just gave you a score, without exporting it to a greenie board, will be a great achievement and important milestone of the whole system. Like, I am playing the mission, and after the trap - I get a composite score based on a couple of criteria (in order of importance): glideslope deviations, lineup deviations, AOA deviations in the groove and one final score based on wire, lineup, speed and vsi on touchdown. Serge has the responsibility to find the proper weights of each score in the composite score 

[fael097]

i understand nothing of technical details, but i hope you guys succeed.
btw, fsx has a mission where the dude says too high, too low, too fast, etc...
maybe that uses something like that but much more simple

[Sludge]

And here's why I asked what was the purpose of those / and \  symbols. In this example my aircraft starts low, then overcorrects and flies up through the GS, being high in the middle. Could I describe this situation either as 'LOX HIM' or 'LOX /' ? My guess is that the first is more informative, it tells WHERE the aircraft is in respect to the GS. The latter tells only WHAT happened. Are those interchangeable or not? Should I stay with the first and not spending my efforts on implementing more complex programming logic of the latter?



And thanks again, I appreciate everyone's help and this ain't my only question (I guess)  

Serge...

In a straightforward answer, I would say, USE the 'LOX HIM' for the initial project.  After you BETA test, more people fly the mission, you'll get FEEDBACK on what works and what doesn't. Then you'll be able to tweak/tailor the "Virtual LSO AI" to the point of making some FSX-RealWorld decisions that will work for a full launch of this mission.  And I mean decisions such as coming up with a calculation that can make a realistic 'LOX /' debrief annotation that the FSX PILOT would remember during the pass.  Spaz can give his FSX feedback and what an LSO might say different? Then, from critiques like that, you can model the vLSO to a good compromise of what will work in FSX and what Spaz and others say.

And I PROMISE, I will provide feedback on this one.  I wasn't a big help on the FCLP and Im sorry, it just never hit my "I have to fly this" bone, so I did install and fly it, but that was just cursory interest.  On this one, Im telling ya, when you get this done, I will be ONE OF THE FIRST to fly it!!

What other questions do you have?

Later
Sludge

[Paddles]

Okie dokie  
I'll start with that 'LOX HIM', as well as 'LURX LULIM' instead of 'RTL', and other similar GS/lineup gross deviations...

Some other questions:
What does B (flat glideslope) mean? Is this the case when the aircraft travels along the more shallower glideslope? If yes, ain't this the same as NERD (not enough rate of descent)?

What does RUF (rough) mean? We do know navy pilots shouldn't be smooth...  

When or why the approach is AFU (All “fouled” up)?