INS and Gyro Drift.

I had a interesting conversation with a couple of FAE (Ecuadorian Air Force) pilots, regarding INS and gyro drift. The following images are a simplification of what they told me about the issue:

As I said before, this is the result of my conversation with them.

The gyro drift is caused by the mechanical components of the INS gyros.

That the direction of the drift is totally random for each flight.

That modern laser ring gyros are more exact than old electro-mechanical ones and subject to less gyro drift but that it still exist in these devices.

That current modern aircrafts, both civilian and combat, no longer use INS as their primary navigation system, relying instead on GPS.

My mental picture of INS drift is - correct me if I am wrong - that the accellerometers and gyros are not working 100% exact due to friction, etc.

So if I fly a turn, the INS detects a certain heading change, but which is off by a small percentage. There are more than one sensor involved when maneuvering - as usually the aircraft changes it's attitude in more than just one plane. So if I fly that turn, not only the heading change sensed by the INS might be a bit off, but probably also other parameters like altitude or whatnot.

What I try to say ... or ask(?) is, the INS drift does not just depend on flying time, but on the amount of maneuvering. Correct? Now, this 1 nm drift per hour - I understand that is just an average? So if I get into a dog fight, I can assume that (if I survive :o) afterwards my INS is completely messed up and the INS drift would easily exceed 1 nm?

My mental picture of INS drift is - correct me if I am wrong - that the accellerometers and gyros are not working 100% exact due to friction, etc.

 

So if I fly a turn, the INS detects a certain heading change, but which is off by a small percentage. There are more than one sensor involved when maneuvering - as usually the aircraft changes it's attitude in more than just one plane. So if I fly that turn, not only the heading change sensed by the INS might be a bit off, but probably also other parameters like altitude or whatnot.

 

What I try to say ... or ask(?) is, the INS drift does not just depend on flying time, but on the amount of maneuvering. Correct? Now, this 1 nm drift per hour - I understand that is just an average? So if I get into a dog fight, I can assume that (if I survive :o) afterwards my INS is completely messed up and the INS drift would easily exceed 1 nm?

No. The gyros are stabilized against aircraft maneuvers. Hard maneuvers will not affect the INS at all. The problem comes not from friction but from stuff like earth's gravity, the coriolis effect on rotating bodies, micro errors in the alignment of the gyros during construction, etc. Stuff that is so small that its effects are only felt during long times.

Hrm, interesting. Seems I found a new topic worth studying a bit more in depth.

Thanks!

.

That current modern aircrafts, both civilian and combat, no longer use INS as their primary navigation system, relying instead on GPS.

Indeed the INS drift is regularly canceled by GPS update.

It's INS with embedded GPS.

Thanks for the info! So I have to fly over geographical point to reset the INS. What is this point and can I create it somehow? (or is it somekind of fixed point)

As I said before, this is the result of my conversation with them.

 

The gyro drift is caused by the mechanical components of the INS gyros.

That the direction of the drift is totally random for each flight.

That modern laser ring gyros are more exact than old electro-mechanical ones and subject to less gyro drift but that it still exist in these devices.

That current modern aircrafts, both civilian and combat, no longer use INS as their primary navigation system, relying instead on GPS.

Hi Zeus,

My 2 cents on those 4 points:

- no comment

- not exactly: there exist a feature in the INS that "learns" from errors (over several flights) and includes an "auto-correction" feature; as such a feature does exist, then it is proof the drift is not really random. OTOH, it can't be predictable precisely either, so positions update are still needed.

- yes

- this is not exactly true; GPS isn't "smooth" enough to be used as a primary "alone" for all the features an INS provides (provided); the real deal today is a hybrid INS/GPS system, with the INS as primary (still), and GPS as an aid; this has all the advantages of the INS (precision in real time, lots of inertial parameters) without the main disadvantages: 1/ initial position is obtained automatically from a GPS fix (no more typing, slow and error-prone) and 2/ intertial position is updated from GPS fixes all the time without having to manually do that from a known landmark (it's more precise, works everywhere...) => essentially reducing the drift to zero.

++

Az'

- this is not exactly true; GPS isn't "smooth" enough to be used as a primary "alone" for all the features an INS provides (provided); the real deal today is a hybrid INS/GPS system, with the INS as primary (still), and GPS as an aid; this has all the advantages of the INS (precision in real time, lots of inertial parameters) without the main disadvantages: 1/ initial position is obtained automatically from a GPS fix (no more typing, slow and error-prone) and 2/ intertial position is updated from GPS fixes all the time without having to manually do that from a known landmark (it's more precise, works everywhere...) => essentially reducing the drift to zero.

In addition to these points, all INS/GPS equipped aircraft can operate on INS alone using the same manual INS update procedures as pre GPS systems in cases where GPS is not available (jamming etc.). While INS/GPS systems don't usually require manual INS updates, in some situations they may.

The same is true of JDAM and other "GPS" weapons.

Really looking forward to seeing this stuff implemented in a sim.

Thanks for the info! So I have to fly over geographical point to reset the INS. What is this point and can I create it somehow? (or is it somekind of fixed point)

When creating the flightplan, you set one or more waypoints as INS realignment checkpoint. You place the waypoint to precisely overfly a landmark, such as a specific mountain or hill, the confluence of two rivers, etc. A specific and easily recognizable terrain feature so you can overfly it and reset the drift to 0. You can also use a navaid such as a TACAN or VOR station as a checkpoint.

The number of INS realignments that you plan to do is based on expected flight time vs the INS drift rate. In the specific case of the M-2000C, with a 1 nm/h drift rate you can probably do that once every one or two hours.

INS drift will not affect the radar. Radar contacts position are calculated with the aircraft as the center of a circle. Aircraft geographical position is irrelevant.

INS drift will affect the Bullseye and any BE based calculation since those require to know the aircraft position in the world. Same deal with any FP waypoint representation in the VTB.

Hi Zeus,

 

My 2 cents on those 4 points:

- no comment

- not exactly: there exist a feature in the INS that "learns" from errors (over several flights) and includes an "auto-correction" feature; as such a feature does exist, then it is proof the drift is not really random. OTOH, it can't be predictable precisely either, so positions update are still needed.

- yes

- this is not exactly true; GPS isn't "smooth" enough to be used as a primary "alone" for all the features an INS provides (provided); the real deal today is a hybrid INS/GPS system, with the INS as primary (still), and GPS as an aid; this has all the advantages of the INS (precision in real time, lots of inertial parameters) without the main disadvantages: 1/ initial position is obtained automatically from a GPS fix (no more typing, slow and error-prone) and 2/ intertial position is updated from GPS fixes all the time without having to manually do that from a known landmark (it's more precise, works everywhere...) => essentially reducing the drift to zero.

 

++

Az'

Hi Az,

1. As I said, it is vastly simplified. The reasons why gyro drift exist is quite complex and there are several, highly technical, papers on the issue that you can read if you like to do integral calculus. Basically the INS is so precise that the device itself interferes with its functionality.

2. Yes, I know. But as I said, this is a simplified version of the issue. The point is that you never know or can predict where the drift will happen.

3. No comment. :smilewink:

4. Very likely, the pilot who said that has retired from the air force and is now flying cargo planes for DHL. So it is probably true in his job and is speaking from memory about military aircraft.

Thanks for the updates Zeus, it's cool to "see" this system coming to life.

Quick question: as I understand it at the moment, the synthetic runway display is based on the landing waypoint in your flightplan. Is this going to remain the same or does the real system work differently?

Or more directly, to get a useful synthetic runway display will we need to perform an (accurate) INS fix update prior to landing, or is the system actually independent of the INS?

Thanks for the updates Zeus, it's cool to "see" this system coming to life.

 

Quick question: as I understand it at the moment, the synthetic runway display is based on the landing waypoint in your flightplan. Is this going to remain the same or does the real system work differently?

 

Or more directly, to get a useful synthetic runway display will we need to perform an (accurate) INS fix update prior to landing, or is the system actually independent of the INS?

1. I have not decided yet.

2. Yes. You will need to do the fix prior to engaging the APP mode. Correction: I've been told that the INS automatically realigns as soon as the ILS station is captured, so you don't need to do the manual fix.

after this beautiful technical explanation, how could someone just disable the drift?

Great wealth of information into the system thank you for the clarifications Zeus!

Sent from my SM-G900W8 using Tapatalk

You shouldn't use the term "realign or align" in flight to remove or reset the INS position. This procedure is "Updating". you are telling the INS ... look I know you are drifting and your position is slightly off let me show you exactly where you are right now, now start computing again using this new updated position.

Aligning and realigning is a completely different procedure usually done on the ground. An INS full alignment takes quite a while ... like 6 or more minutes. A quick alignment can usually be done as well its a lot less accurate but may only take 30secs. Same goes for a stored Heading alignment ... great for Scramble type situations but again inherently less accurate than a full alignment. Think of an Alignment as a complete INS reboot.

Some systems can in an emergency perform a very basic alignment in flight but require wings level flight and typically a velocity input

In addition the use of external navaids like an ILS to "update" the position is always subject to a degree of quality control (done automatically). In the case of an ILS update it typically only uses the localiser info so will only correct in one axis. In the case of the Mir2K I would assume this would only affect the lateral position of the Synthetic rwy and its Course line.

You shouldn't use the term "realign or align" in flight to remove or reset the INS position. This procedure is "Updating". you are telling the INS ... look I know you are drifting and your position is slightly off let me show you exactly where you are right now, now start computing again using this new updated position.

 

Aligning and realigning is a completely different procedure usually done on the ground. An INS full alignment takes quite a while ... like 6 or more minutes. A quick alignment can usually be done as well its a lot less accurate but may only take 30secs. Same goes for a stored Heading alignment ... great for Scramble type situations but again inherently less accurate than a full alignment. Think of an Alignment as a complete INS reboot.

 

Some systems can in an emergency perform a very basic alignment in flight but require wings level flight and typically a velocity input

 

In addition the use of external navaids like an ILS to "update" the position is always subject to a degree of quality control (done automatically). In the case of an ILS update it typically only uses the localiser info so will only correct in one axis. In the case of the Mir2K I would assume this would only affect the lateral position of the Synthetic rwy and its Course line.

Thanks for the clarification in concepts.

Cant the INS use ILS / TACAN + Radar Altimeter / Baro + Compass all at once?

that should give a pretty accurate picture (at least relative to 1nm drift)

The gyros used in relative modern INS systems are not mechanical, but so called ring laser gyros, where a laser beam is send through mirrors and the time it takes for the light to reach the detector after having been send throught the mirrors are measured..

This time is affected by movement a bit like we know from the Doppler effect.

https://en.wikipedia.org/wiki/Ring_laser_gyroscope

FinnJ

As a side note, about the INS/GPS modern systems:

This is what happens when implementation got it wrong (and is "too much" confident about GPS signals): [ame]http://www.faa.gov/documentLibrary/media/Notice/GENOT_7110_711_EMB-300.pdf[/ame]

System should have detected GPS problem and reverted to INS source "only". Didn't. Woops.

= = = = =

The gyros used in relative modern INS systems are not mechanical, but so called ring laser gyros

I'm not sure M-2000C's INS is "modern" enough, though.

Cant the INS use ILS / TACAN + Radar Altimeter / Baro + Compass all at once?

that should give a pretty accurate picture (at least relative to 1nm drift)

The concept of the INS is that it is (or can be) completely self-contained and never need to rely on an external source. It is MIJI (Meaconing, Intrusion, Jamming and Interference) proof, and 1-2nm/hr is a small price to pay to have independence from external navaids.

It's not as accurate as an EGI system (which uses encrypted GPS to update the inertial), but it was state of the art back then, and it will get you into the target area and back.

It's also the responsibility of the pilot (and RIO/WSO in a 2-seater) to keep a nav picture, of which the INS is a single datapoint for developing SA.

1. As I said, it is vastly simplified. The reasons why gyro drift exist is quite complex and there are several, highly technical, papers on the issue that you can read if you like to do integral calculus. Basically the INS is so precise that the device itself interferes with its functionality.

If you guys want to learn a bit about inertial error correction, just google "Kalman Filter". Long story short, it's an algorithm that dampens inertial drift and limits it to a more predictable linear drift rate.

do I need my engine running for INS alignment process ?

do I need my engine running for INS alignment process ?

No.

No.

thanks :thumbup: