Gimbal distance and Speed Range Estimates using Lines of Bearing and/or DCS

[DavidB66]

That's because the plume is 10 times thinner when viewed from the side compared to viewing it directly from the back, see figure in post 315 above. A 10x thicker layer of hot gases is much more visible.

That's a fair point. It just doesn't seem obvious in that particular video. If the main factor in visibility were the angle from which the plume is viewed, one might expect to see a continuous change in visibility of the plume as the angle changes. But at 0:11 in the video, when the planes are viewed at roughly a half-turned position, the plumes are already hardly visible behind the nozzles.
However, that is just one video, and others, like this one, show a variety of effects:

Source: https://www.youtube.com/watch?v=PLzD1SCk__g&t=91s

For example, at 1:02 the plane (a BAe Typhoon) shows a narrow concentrated plume projecting behind the nozzles and gradually blending into the air behind, while at 1:13 the same plane shows a huge flaring plume extending much further back; then at 1:20 another plane (an F18) also shows a large plume. I don't know how far these differences are due to characteristics of the engines or fuel (e.g. the use of after-burners), or to the optics of the cameras.
So far as glare is concerned, the optics definitely come into it, as glare by definition is the appearance of light (in this case IR) spreading beyond where it should appear with a perfect lens, sensor, etc.

 

[Itsme]

That's a fair point. It just doesn't seem obvious in that particular video. If the main factor in visibility were the angle from which the plume is viewed, one might expect to see a continuous change in visibility of the plume as the angle changes. But at 0:11 in the video, when the planes are viewed at roughly a half-turned position, the plumes are already hardly visible behind the nozzles.

Isn't that exactly what you'd expect? If you look straight down the exhaust plume from the back of the jet, you're looking through a very thick layer of hot gas (red arrow in figure, note that the green 2000 Kelvin region is even longer than the figure itself). If you move a sideways just a bit, the thickness of this layer very rapidly diminishes (blue arrow in figure):

So not seeing a continuous change but a sudden change instead is exacty
what you would expect from a thermal image of the exhaust plume.

It is, however, not what you would expect from a glare. This is because the IR emission of the metal nozzle of a jet is as high or even higher than that of the exhaust gas. Your remark is a very good one, the IR emission of the jet exhaust gases is limited. In the video you posted the nozzles are as bright or even brighter that the exhaust plume in every case The same can be seen in the ATFLIR image of a jet engine, the nozzle rims are in fact brighter in IR than the hot gas in the engine:

This means the glare is not expected to suddenly decrease when no longer looking straight into the engine.

I don't know how far these differences are due to characteristics of the engines or fuel (e.g. the use of after-burners), or to the optics of the cameras.

I think this video was posted by a visitor of this show, using the same camera, and a lot of times you see the use of afterburners (which of course adds to the spectacle of a flight show).

So far as glare is concerned, the optics definitely come into it, as glare by definition is the appearance of light (in this case IR) spreading beyond where it should appear with a perfect lens, sensor, etc.

Yes you're absolutely right. No optical system has a zero point spread function. But according to Teledyne FLIR, the glare in thermal imaging is minimal, which means a very small point spread function. So a very bright IR source may become a bit thicker at the edges, that's all. It certainly does not turn a point source into a big blob.

 

[Mick West]

This means the glare is not expected to suddenly decrease when no longer looking straight into the engine.

And yet that's exactly what it does:

Looking straight into the engine. Large glare, obscuring the plane

1 second later. No longer looking straight in, smaller glare. Plane more visible as exposure adjusts for the dimmer average light

2 more seconds. Side on, no engine glare, or plumes. Plane now black as very hot heat sources are hidden.

 

[Itsme]

Agree, and my point exactly: it behaves like an exhaust plume would, and not like a glare would behave.

 

[Mick West]

Agree, and my point exactly: it behaves like an exhaust plume would, and not like a glare would behave.

But you just said

This means the glare is not expected to suddenly decrease when no longer looking straight into the engine.

So, what is your point exactly? WHAT behave like an exhaust plume and not a glare?

 

[JMartJr]

2 more seconds. Side on, no engine glare, or plumes. Plane now black as very hot heat sources are hidden.
View attachment 48433

Pardon me for "reposting" but it seem to me to be very important... there is no visible hot exhaust plume in this, the image it "should" be most visible in, nor in the other two. None at all.

 

[Itsme]

But you just said



So, what is your point exactly? WHAT behave like an exhaust plume and not a glare?

See post #322 for explanation: There is NO sudden huge increase in IR brightness when looking straight into the engine. The nozzles are as bright or even brighter in IR as the heart of the engine, so the alleged big glare blob caused by a bright IR source should be visible from an angle as well.

There is, however, a sudden and huge increase in thickness of the exhaust plume along the LOS of the thermal camera when looking straight into the engine.

 

[Itsme]

Pardon me for "reposting" but it seem to me to be very important... there is no visible hot exhaust plume in this, the image it "should" be most visible in, nor in the other two. None at all.

If the thickness of the plume along the LOS of the thermal camera drops below the detection limit, you won't see it. Another aspect is the amount of thrust given by the engine, which may vary while taking a curve.

 

[Mendel]

Another aspect is the amount of thrust given by the engine, which may vary while taking a curve.

Are you suggesting pilots reduce thrust when flying a curve?

 

[Mick West]

See post #322 for explanation: There is NO sudden huge increase in IR brightness when looking straight into the engine. The nozzles are as bright or even brighter in IR as the heart of the engine, so the alleged big glare blob caused by a bright IR source should be visible from an angle as well.

There is, however, a sudden and huge increase in thickness of the exhaust plume along the LOS of the thermal camera when looking straight into the engine.

Sorry, that's nonsense, which you should be able to see by watching the video of the jet banking away.

It very obviously glare. There is nothing that looks like a glare. It does not turn into an increasingly transparent plume as it rotates. It diminishes and then vanishes because the hot interior of the engine is now shielded by the exhaust nozzle.

 

[markus]

Isn't that exactly what you'd expect? If you look straight down the exhaust plume from the back of the jet, you're looking through a very thick layer of hot gas (red arrow in figure, note that the green 2000 Kelvin region is even longer than the figure itself). If you move a sideways just a bit, the thickness of this layer very rapidly diminishes (blue arrow in figure):
View attachment 48428
So not seeing a continuous change but a sudden change instead is exacty
what you would expect from a thermal image of the exhaust plume.

There are two cases (well, a continuum of cases, with these two as its extremes):

• Plume is IR-transparent: You'd expect the same total brightness throughout the frame regardless of plume orientation. If you're right on the cusp of overexposure maybe changing perspective would attenuate the glare, but you wouldn't go from "literally blinding the sensor" to "invisible" simply by decreasing the local intensity by a factor of 10 or so, especially while keeping the intensity integrated over the frame constant.
• Plume is IR-opaque: in this case only the external surface of the plume would be visible, in which case viewing it from the side would make the image brighter. If this affects the camera exposure setting, the plume would appear more distinct.

In either case there doesn't seem to be a rationale for why the black region should almost disappear as the airplane turns.

The video posted by @DavidB66 contains a smoking gun:


Source: https://youtu.be/PLzD1SCk__g?t=147


I cued it up at 2:27, right as the F-22 nears the cusp of a loop maneuver. In the following couple of seconds there's a moment where the afterburners become briefly occluded, and the afterburner glare becomes noticeably smaller. There's nothing else it can be here: it can't be a change in thrust, first because it wouldn't make any sense and second because jet engines don't respond that fast. It can only be glare.

 

[TheCholla]

Sorry, that's nonsense, which you should be able to see by watching the video of the jet banking away.
View attachment 48457



It very obviously glare. There is nothing that looks like a glare. It does not turn into an increasingly transparent plume as it rotates. It diminishes and then vanishes because the hot interior of the engine is now shielded by the exhaust nozzle.
View attachment 48458

On this video we very clearly see the glare from the hot engine disappears when the jet is seen from the side. How is this consistent with a right-to-left trajectory proposed by @markus? And Gimbal being a distant plane ? The lines of sight tell us a plane would have to be seen from the side at some point.

 

[Itsme]

@Mick in reply to post 330: You use an optical image of a jet exhaust to suggest the interior of the exhaust is much 'hotter' and therefore much brighter in IR than the exterior of the nozzles. The actual thermal images, however, tell a different story and show that the nozzles are at least as bright in IR as the interior of the engine. That is because the hot gas leaving the engines is less prone to emit IR than the metal of the nozzles, see posts 318 and 322. An optical image in that sense can be misleading.

@markus in reply to post 331: The plume is of course neither fully transparent not fully opaque. It is a gas cloud emitting IR itself.

General remark: trying to quote a reply on this site is veeeeeery slow and often fails (getting an 'oops, it seems something went wrong' message). Therefore I resorted to referring to the other posts instead. Are more people having these problems?

 

[Edward Current]

The actual thermal images, however, tell a different story and show that the nozzles are at least as bright in IR as the interior of the engine.

They do? The video Mick posted transitions almost instantly to a higher exposure as the plane banks. Watch the background noise suddenly increase.

That is because the hot gas leaving the engines is less prone to emit IR than the metal of the nozzles, see posts 318 and 322.

Those posts don't establish that. They talk about oxygen and nitrogen. The oxygen in the exhaust has largely been replaced by CO2, water, and other substances. Also I would assume the gases are hundreds of degrees hotter than the nozzles, and if they are partially IR-transparent, more than just the surface of the plume will emit IR, unlike the nozzles.

 

[Itsme]

They do? The video Mick posted transitions almost instantly to a higher exposure as the plane banks. Watch the background noise suddenly increase.

I mean the ATFLIR images. There is hardly any difference looking sideways to the engines compared to looking into them. Certainly not enough distance to be causing a glare:

Those posts don't establish that. They talk about oxygen and nitrogen. The oxygen in the exhaust has largely been replaced by CO2, water, and other substances. Also I would assume the gases are hundreds of degrees hotter than the nozzles, and if they are partially IR-transparent, more than just the surface of the plume will emit IR, unlike the nozzles.

The first post only gives an explanation, true. But the thermal images of the ATFLIR establish it, because the nozzle rims are brighter than the inside of the engines in white hot mode:

 

[Itsme]

distance = difference in previous post, reloading for an edit takes ages...

 

[Mick West]

Sorry @Itsme, I think the fact that were are seeing glare has been conclusively demonstrated to you, and you are just refusing to see it. I'm removing you from this thread as you are cluttering it up. If people (not me) wish to continue they can do it via PM.

 

[Edward Current]

Have our calculations involving the ATFLIR's field of view been off by 40% or more?

While working on my Gimbal simulation in Blender (which, like everything Gimbal-related, is very difficult), I realized that I have to constrain the on-axis roll of the camera to the on-axis roll of the aircraft. This is because the ATFLIR system rotates the onscreen picture as the plane banks, in order to provide a familiar-looking horizon indicator. The rotation of the picture is not necessarily what the camera is actually seeing.

So, if the aircraft were to bank at 45°, the horizon in the image would stretch from one corner to the other. I believe that is what the FOV angular value refers to — it's not the width of the field left–right or up–down, it's across the diagonal. This means that the left–right or up–down FOV is actually smaller than we've been assuming by a factor of √2, or 1.414. In the case of Gimbal, the FOV is actually 0.25° left-to-right (.35 ÷ 1.414).

This would explain very neatly why, when I did my GoFast analysis in Blender, I had to artificially reduce the FOV from 0.7° to 0.5° in order to reproduce the speed of the water surface moving across the frame. I wasn't happy with having to add this fudge, but I attributed it to overscan (like in the old NTSC video spec), which would provide more area for the stabilization system to do its thing. It turns out, though, that the overscan is to allow for rotation. It's really a circular image, and the FOV is that circle's diameter. What's 0.7° ÷ √2? It's .495° — within about 1% of the value I was forced to use.

This produces new constraints for the size of the region spanned by the Gimbal "blob," as I originally calculated here, and therefore constraints on the size of the physical Gimbal object depending on its distance:
1 NM - 1.1 feet
5 NM - 5.2 feet
10 NM - 10.3 feet
20 NM - 21 feet
40 NM - 41 feet

 

[Mick West]

I believe that is what the FOV angular value refers to — it's not the width of the field left–right or up–down, it's across the diagonal. This means that left–right or up–down FOV is actually smaller than we've been assuming by a factor of √2, or 1.414. In the case of Gimbal, the FOV is actually 0.25° left-to-right (.35 ÷ 1.414).

Interesting (and useful) if true. The FOV has always been contentious and problematic. However the "ATFLIR Principles of operation" manual says:

select the field of view (FOV) for ATFLIR operation. Alternate pressing selects either .7° by .7° narrow FOV, 3° by 3°medium FOV or 6° by 6° wide FOV.

".7 by .7" suggests horizontal and vertical. Diagonal would only need one number. Of course this could be a simplification by the manual writer, as the exact FOV isn't super important to anything discussed (or even really to pilot usage).

 

[Mick West]

The discussion of @Edward Current's blender simulation with clouds has been split to its own thread:
https://www.metabunk.org/threads/gimbal-blended-simulation-with-clouds.12209/

 

[Mick West]

Started a new thread on creating a concise explainer video to cover the evidence for a rotating glare.

https://www.metabunk.org/threads/a-gimbal-glare-explainer.12230/

 

[Mick West]

The weak link indeed was the gimbal azimuths. I extracted them frame by frame and applied gaussian smoothing with a sigma of 20 frames.

Do you have a .CSV with the smooth values per-frame?

 

[markus]

Do you have a .CSV with the smooth values per-frame?

Here you go: