Landing

Suggest to try that yourself and keep an eye on the VSI instead of the accelerometer ;)

Btw, what has the accelerometer to do with this issue?

My apologies for the delayed response. It has not been a good weekend so far.

I misspoke. I was thinking variometer and, for some reason, typed accelerometer. Perhaps senility is finally setting in. It was also early morning and I was writing with minimal caffein in my system.

The above formula is valid for all planes. No need to test it. That's simple math.

 

...

I assume you haven't done what I asked and checked in the cockpit. What in these images has always been my experience. If you are correct, then one of these instruments is wrong. Or I am completely misinterpreting the conversation:

The same is true in the Su-27's pit.

This may be why Cap'n made his original statement, whether he was consciously aware of this or not.

Thanx for your reply :) Just tested and the problem is the wrong (way too coarse) glideslope deviation indication.

(There are a few other problems with the F-15 ILS as well but this one is the most important in this case)

The ILS GS starts indicating at (way too late) approx 4.3NM which means at 1300ft AGL.

With a minimum of approx 200ft this leaves 1100ft or just 1min from intercepting the GS until reaching the minimum.

If you fly 800ft/min or 1000ft/min in this short time span doesn't matter at all with the too coarse GS deviation scale in the F-15 as it doesn't even noticable move away from the center position.

The GS indicition in the DCS F-15 in its present state is unfortunately pretty useless.

Well, actually, you intercept the glideslope in the sim at 9.6NM (I assume the units are NM and not miles). That's the point at which the HUD bank steering index becomes the landing steering cross. From that point, even if you're a bit slow (like I am) in making the required adjustments (steering, AoA units, and power), by the time you hit the 5.5NM mark, you should be in equilibrium with the velocity vector centered on the steering cross and both superimposed over the runway landing marks because you are now flying down the center of the glideslope with 21 units, the correct power setting and 1000 fpm down on the VVI. That you are centered on the glide slope will be confirmed at 4.3NM, when the ADI ILS bars become active and are centered. (FWIW, this is roughly similar to how it works in the Su-27 cockpit as well.)

So it's not like there are last minute adjustments to be made that'll keep you bouncing around until you land. By the 5.5 mark, you're centered on the glideslope and, if you're maintaining 21 units AoA, you'll be sitting on a descent rate of 1000 fpm. You are free to argue that the instruments aren't accurate enough to make any judgements but, then, I have to question what you're original arguement with Cap'n was, since he was referencing what's available to the pilot in the cockpit.

And, if you allow that the cockpit instruments are accurate, then, in the "Light" screenshot I posted above, a descent rate of roughly 780 fpm should be indicated. So, there's a problem. Either you are incorrect or the sim is incorrect. If the latter, you should post a bug report so we can get it fixed.

If you fly level at 3000ft AGL until 10NM (GS intercept point IRL) from the runway (ILS DME) and precisely maintain a 1000ft/min ROD at 200kts and at the next try at 150kts (both at 21 units AoA) you will be surprised how far short of the runway you will hit the ground during the second attempt ;)

I might do that when I have some free time but only if you can tell me how to control for the differences (measurement errors) you'll introduce as you transition from level flight onto the glideslope. I'm gussing that you'll bounce around for awhile before settling in. I certainly do. As I noted above, if you're slow you'll be dialed in at around 5.5. That's up to 4.5NM worth of error introduced into the experiment.

bbrz is right about the glideslope geometry. Sort of.

More specifically, right if all bodies are measured in the frame of reference of the airfield talking about speed along the glidepath.

AoA and IAS are both in the frame of reference of the aircraft and compare the motion of the aircraft relative to the air, not the ground.

Most of this argument is people forgetting or not knowing precise definitions and the implications of those definitions,

The math and physics checks out,

if you do it right.

Edit: I'll admit that it took me a little while to figure out how bbrz could be doing the geometry right and everyone else could be doing the airmanship right and yet somehow getting different answers.

bbrz is right about the glideslope geometry. Sort of.

 

More specifically, right if all bodies are measured in the frame of reference of the airfield talking about speed along the glidepath.

 

AoA and IAS are both in the frame of reference of the aircraft and compare the motion of the aircraft relative to the air, not the ground.

 

 

Most of this argument is people forgetting or not knowing precise definitions and the implications of those definitions,

 

 

The math and physics checks out,

 

if you do it right.

 

Edit: I'll admit that it took me a little while to figure out how bbrz could be doing the geometry right and everyone else could be doing the airmanship right and yet somehow getting different answers.

Another way of saying it, I think, is that the energy taken out of the system has to equal the energy put into the system. His formulation is only correct, if you are comparing the same aircraft at equal weight. In that case, increasing the airspeed means there is excess energy available to put into forward speed and a resultant increased rate of descent. But when you increase the weight but want to maintain the same AoA and glide slope, the extra energy is only sufficient to add to lift, not forward speed, and the result descent rate remains the same. If it did both, we'd have free energy and whoever bottled it would be a billionaire.

IMO never trust ILS in the F-15C(At least how it’s implemented now). Use the technique from the flight manual (basically what GG said) and F-10 map if needed for distance.

Also remember that at higher gross weights your CG will be further forward. You’ll need more aft stick at touchdown to keep the nose up relative to a lighter gross weight. This is mentioned in the manual.

To the original poster, have you had any success? I think we’ve made it clear as mud :joystick:

Ok, I managed to come up with a slightly more understandable explanation than my previous one. At least I hope it's more understandable.

Ironhand got the core of the problem, obeying the laws of physics.

This all based on physics, and physics as currently understood demand conservation of mass and energy. That means that a math model of a physics problem has to obey those laws.

bbrz was using a purely positional model. You could call it algebraic, geometric, or analytic geometry, because mathematicians aren't always sticklers for consistency in naming. Whatever you choose to call it, it is inadequate as a good representation of the aspects of landing a plane brought up in this thread because it can't keep track of the conservation principles demanded by physics.

For that you need vector mathematics.

The motion of the plane as being talked about is the sum of four vectors: thrust, drag, lift and weight. It is possible to change these vectors and still arrive at the same sum. Then with that force vector you can go on to calculate acceleration, velocity, and finally, position. Trying to skip straight to position does not give accurate answers.

Here's an example of different input vectors resulting in the same sum. If thrust = (5,0,0) and drag =(-2, 0, 0) then the result in a standard Cartesian coordinate system is 3x.

Change that to thrust = (7, 0, 0) and drag = (-4, 0, 0) and the result of summing them is 3x.

Vectors behave a bit oddly compared to the math people learn before encountering them, which is why students encountering them often struggle a bit. They don't act quite like people who've never seen them before expect them to act.

Before I digress too far on mathematics, the salient point is that there is a lot of scope for decent rate to remain the same with other parameters changing as long as the net sum of force vectors acting on the plane remains the same.

bbrz performed their math correctly, it just wasn't the correct form of math for modelling the system in question. It was a good effort based on incomplete understanding of how to solve the problem.

That's not the only incomplete understanding of things from bbrz.

You have absolutely no idea what you are talking about.

According to your 'logic' you could never fly in IMC but if you would do that you would be managing the plane instead of flying it.

That's absolutely ridiculous and shows a total lack of knowledge about RW flying.

The folks at the FAA who write things like their Basic Airmanship Manual (available free as a PDF), and presumably know something about training pilots in real life, spend a great deal of effort to encourage pilots to avoid being "head down." Translating from aviation speak to English that means, "don't look at the instruments, look out the cockpit canopy." If you ever watch videos of instructor pilots with new students you'll notice that instructors do a great deal of telling student pilots to stop looking at the instruments and look out the canopy.

There's a reason for this. Humans are really bad at instrument navigation. The instruments are a backup in case looking out of the canopy doesn't work. You don't fly tight formation in bad visibility in real life. You don't land manually if there isn't decent visibility at ground level in real life. Human pilot IFR is an approximate navigation system, not a precision navigation system. If a human pilot can't land visually for the actual touchdown, they'll get diverted to somewhere with better visibility.

For really low to zero vis conditions landings are handled by autopilots at airfields with compatible precision landing systems. The airfield portion of the system provides a reference to compensate for errors or drift in the aircraft instruments, and the autopilot prevents the human pilot from making the mistakes that are expected even from human pilots with IFR ratings and thousands of hours of flight experience. People just aren't good at monitoring many instruments at one time, and they're also not very good at turning instrument readings into an accurate and up to date idea of where the aircraft is and how the aircraft is moving.

Treating a human IFR rated pilot as a precision navigation system in very low visibility has resulted in what's called, "controlled flight into terrain," so many times in the history of aviation that aviation authorities have developed extensive rulesets designed to discourage pilots from being foolish enough to think that a set of instruments and an IFR rating will allow a safe landing when below IFR minimums.

The folks at the FAA who write things like their Basic Airmanship Manual (available free as a PDF), and presumably know something about training pilots in real life, spend a great deal of effort to encourage pilots to avoid being "head down." Translating from aviation speak to English that means, "don't look at the instruments, look out the cockpit canopy." If you ever watch videos of instructor pilots with new students you'll notice that instructors do a great deal of telling student pilots to stop looking at the instruments and look out the canopy.

This is where you lose me. Unless you turn off or have lost electrical power all the information (AoA, IAS, altitude, etc.) is right there, directly in front of you on your HUD. You don’t need to be heads down to see any of it. Just remember to be aware of your weight (fuel gauge and stores) before the approach. You’d have to be willfully ignoring it to not see this information while looking out the canopy.

This is where you lose me.....

At a guess, to combat fixation, especially where trainees are concerned.

Contact flying backed up by instrument cross-check. Air forces teach this all the time AFAIK.

This is where you lose me. Unless you turn off or have lost electrical power all the information (AoA, IAS, altitude, etc.) is right there, directly in front of you on your HUD. You don’t need to be heads down to see any of it. Just remember to be aware of your weight (fuel gauge and stores) before the approach. You’d have to be willfully ignoring it to not see this information while looking out the canopy.

Keep in mind that especially for instructional materials written aimed at civilian aviation, the assumption is that initial training aircraft will not have a HUD. They might not even be IFR rated aircraft.

The main concern is that since students start with VFR and will be flying solo after not really all that many hours, keeping eyes in the pit is an extremely dangerous habit. In VFR keeping eyes out of the pit is the only warning system for preventing collisions, mid-air or with terrain. When the pilot starts to solo, there's no Instructor Pilot to start screaming in terror because someone is so fascinated by their compass heading that they haven't noticed a looming mountain.

It's bad form for an Instructor to lose students as soon as they start soloing because the Instructor never bothered to teach them to watch where they were going.

Not to mention that blind blundering around tends not to make friends with the pilots that you near-miss. Then when they start yelling on the radio ATC may get upset, when they file reports with your aircraft ID calling you a dangerous idiot the national aviation authorities get upset, etc.

The point was that eyes out is the "official" preferred primary source of information for pilots. You check the pit as necessary. It's nice to know if instruments agree that things look ok really are ok. After checking, you're eyes out again though.

In IFR "as necessary" might be nearly continuous, but it's continuous because there's no choice, not continuous because the instruments are preferred over the view out the windscreen.

This is sort of derailing again though, so I'll stop with the general aviation stuff since this is supposed to be F-15 specific.

Quite a bit of misleading.....

Why don't you just capture a video via Shadowplay (or AMD equivalent) and post it up. That way people can see what it is you are trying to convey.

Why don't you just capture a video via Shadowplay (or AMD equivalent) and post it up. That way people can see what it is you are trying to convey.

...

I'd guess that it is a question of technique and training.

Especially in VMC it's rather easy. If the ILS FD cross is exactly centered within the flight path marker, both symbols will be sitting exactly on the runway threshold. Just tried an ILS approach flown with only the HUD FD and occasional cross checking with the vertical speed. If you follow the FD exactly the ROD will be ~750ft/min at 150kts...

This I need to see.

As I'm usually not flying the F-15...

My issue as well...

Thank you. Unfortunately it'll probably be Thursday or Friday before I can view them.

As I'm usually not flying the F-15 these 2 approaches are a bit rough, but you get the idea.

I only saw the first track, question: Why did you land so far down the runway? Look like almost half way. Where you able to stop? The track did not show.

I thought this was to help people land the F-15 in better ways, yet you say:

As I'm usually not flying the F-15 these 2 approaches are a bit rough, but you get the idea.

Oh well. thank anyway.

Unfortunately it'll probably be Thursday or Friday before I can view them.

Herewith tracks converted to videos in case of playback issues and for ease of reference.

At the risk of stating the obvious, all credit to bbrz for airmanship.

No. I've only uploaded the tracks to demonstrate that the ROD on a fixed glideslope depends on groundspeed and nothing else.

TuFr99YWbtI

pPGT08FCGRE

Disclaimer:

Excuse the aspect ratio issue - DCS/ED and Shadowplay obviously don't seem to get along for reasons unknown

its like the easier they make airplanes to land the harder people try to have trouble with it

Oh, wow, thanx for the effort and your time to create these videos! :)

No bother at all :)

Was I ever out in left field on this one! My apologies for the misinformation and not recognizing my error sooner. Even before viewing Bbrz's tracks, in thinking about it, I realized that I had to be wrong. You get to the threshold sooner when heavy than light from the same starting point. That alone meant the rates had to be different. At any rate, clearly different descent rates are registered on the VVI in the situation we've been discussing.

Just to prove it to myself, I went out and flew my original ILS landings again after modifying the hell out of my flightstick axis curves. The results follow. As with Bbrz's tracks, they clearly show a difference: roughly 850 when light and 1200 when heavy at 20 units. I used 20 units rather than 21 simply because it's a bit easier to peg on the HUD AoA tape:

I always thought that I don't need any curves and I accepted the slightly twitchy behaviour ...until I tried to fly precise ILS approaches.

It doesn't take much of a curve but just a slight adjustment in the low deflection range makes a big difference and it furthermore makes takeoff rotation and landing flare easier.

I did forget to add a curve to the rudder axis and despite the super smooth and super precise movement of the pedals the result is clearly visible.

Evidently my old X-52 is twitchier than your stick because I had to make some significant adjustments to all axes (including thrust) before I could be sufficiently precise. Perhaps using a stick with some of its wires hanging out has something to do with it. It might be time to replace it. At any rate, the imprecision had been enough to mask the difference. So my apologies.

The magnet sensors on the T16000M are really superb. I could and would never switch back to anything less precise and with a lower resolution as it removes the absolutely necessary controllability/precision in DCS. (same goes for the VKB pedals)

Hmmm...I'll have to check it out. Christmas is coming.

Did you trim for hands-off approach?

Evidently my old X-52 is twitchier than your stick because I had to make some significant adjustments to all axes (including thrust) before I could be sufficiently precise. Perhaps using a stick with some of its wires hanging out has something to do with it. It might be time to replace it. At any rate, the imprecision had been enough to mask the difference. So my apologies.

Did you trim for hands-off approach?

I find that the trim isn't quite fine enough for this. It doesn't take much variation to change the rate of descent. Normally it's not an issue. But when you want to hold an exact AoA (not X plus or minus 0.5) with the tad pole and cross hairs fixed on the landing point, it tends to matter more. That being said, every adjustment I tried to make bounced me around too much until I set up some severe curving. I even had to change my thrust curve. It ended up slightly negative.

Interesting. I do have a bit of curve for the axis but not too much. I find that trimming works well enough and only minor adjustments are needed, but you have to be very on-the-ball with your throttle and glideslope to begin with ... spiky pots would probably throw you off.

Do you see spiking when you have the control display open?

I use the “user curve” option. This is an really rough example. I’ll post the refined one I use when I’m at my PC.