EFIS Gyro Stability

[Ka8 Flyer]

Hi folks,

we all know the position in the IRU's drift over time as the accelerometers aren't 100% accurate. (Just as the heading drift on old DGs)
But what about the laser/ring gyros? I've never read that you need to check their accuracy but wouldn't the attitude be just as affected by gyro drift?
Granted, airliners are usually limited to small pitch / bank values (30-40°) and the rotation rate is also quite low but still, wouldn't turbulence cause errors?

Or is there a way to re-calibrate the gyros in straight and level flight?

Thanks,
Mark

[BOAC]

No 'drift' as you know it in a laser-ring gyro, but other errors. No real need to 're-calibrate' either.

[Ka8 Flyer]

Hi BOAC,

thanks for the reply. May I ask why?
I mean, the gyro measures a relative value (namely the rotation around the 3 axes) and this value is then added to the current value to get the absolute orientation in 3D space.
Or am I missing something?

[BOAC]

All done electronically - it knows where it is - usually :) (INS) and the 'earth rate' is applied appropriately. That is 'earth rate. You seem to be trying to cover all sorts of 'spinning gyro' errors (which just do not exist) in your post? Friction - take it as nil. 'Accelerometers' - pretty well perfect. Profile rate - INS driven. Add Schuler Loop tuning and you have a pretty good system. Obviously no software is perfect, but platform errors are pretty well a forgotten item.

[FlightPathOBN]

There are also the GPS update...

[BOAC]

There is no GPS update to the platform! We are discussing IRUs not FMCs.

[FlightPathOBN]

The standard ADIRU on Airbus platforms is provided by Honeywell....

Honeywell (NYSE: HON) announced today that its Air Data Inertial Reference System (ADIRS), which provides navigational data, has become the standard inertial reference system for the Airbus A320, A330/A340 and A380 family of aircraft.

"Three Air Data Inertial Reference Units (ADIRUs) each comprise three ring laser gyros and three quartz accelerometers to accurately sense aircraft position and attitude. By integrating data from the inertial sensors and Global Positioning System (GPS), ADIRS provides automatic initialization, faster align times and 100 percent availability of Required Navigation Performance approaches."
Honeywell's Inertial Navigation System Becomes Standard Equipment On Airbus (http://www51.honeywell.com/honeywell/news-events/press-releases-details/04.30.10ADIRS.html)


Evolve dude...

[Ka8 Flyer]

So if I understand you correctly, the laser gyros are practically error free and the modern accelerometers, while being very accurate, aren't 100% accurate (causing the eventual map shift if not updated by an external source via VOR/LOC/DME or GPS through the FMC).

So what about the mechanical counterparts found in GA aircraft, the ones spinning at 30k RPMs? Can the attitude drift there?

@OBN:
No, the external sources will never update the actual internal position provided by the ADIRS. That's done in the FMGC in the Airbus. And definitely no GPS update will be used for attitude data (and that's what this post is mainly about).

[BOAC]

Basically yes - all systems have errors. Any error, however small, in the achievement of a perfect alignment of a platform will induce a false acceleration/s which will integrate to position errors over time. These are corrected in the FMC by some sort of update as you list. I don't know how good your 'GA' ones are, but if they are like the 'old' gymballed platforms we began with, you would be pushed to see any attitude error - we are talking fractions of a degree normally. It will be 'map shift' you will see plus possibly a 'residual' groundspeed on stand. Map shift is becoming a rarity in airline usage as GPS updating becomes the norm. In the 'old days' a shift of 30 nm or more was not unknown and I had one of those at the end of the Black Sea after an hour without a DME update. Didn't half screw up my descent planning into Baku :)

FPOBN has got confused during his 'evolution', I think, between using GPS for alignment and updating, and is under the impression that a GPS can produce attitude information of the accuracy required for a platform re-alignment. That'll be the day! That WILL be evolution, dude.

Many people get confused between the 'gyrosopic' attitude functions of a platform and the position derived from accelerations to that platform and then displayed to the pilot.

[Ka8 Flyer]

Thanks a lot BOAC for your explanations.
I still find it fascinating that a pure mechanical device would drift as little as 30 nm after an oceanic crossing of say 3000nm, so a 1% error.

Fortunately (or unfortunately?), EFIS is becoming more and more straight-forward for GA aircraft. Heck the last prop I flew (a C172) had a G1000 cockpit - compare that to a B733/B757/B767 cockpit and you'll get pretty envious if you're into EFIS :)

[FlightPathOBN]

"The IR component of an ADIRU gives attitude, flight path vector, ground speed and positional data. The ring laser gyroscope is a core enabling technology in the system, and is used together with accelerometers, GPS and other sensors to provide raw data. "

The FMC may be supplied by other manufacturers than Honeywell. The ADIRU data is available, whether it is used or not, is dependent on the box. You can have a Smiths, or Thales box with the Honeywell ADIRU, they may integrate the raw data or not.
In designing RNP and GBAS procedures, the update and drift rates are a significant part of the calculations.

This ADIRU IS GPS updated.

[Denti]

If only "raw" data is provided it is one box that provides unfiltered and unupdated attitude, accelerometer and GPS data. The GPS does not need to update the IRS portion for that.

On Boeings there is no internal GPS updating in the ADIRUs and they come right of the box GLS and RNP 0.1 approved and equipped. Our A320/A330s come right out of the box only RNP 0.3 and not GLS approved and equipped. We could equip them for both, GLS costs are over a million $ per unit and the upgrade to RNP 0.1 on A320s cost around 300k to 500k $ per unit.

[nitpicker330]

Sorry but the ADIRU/IRS itself is NOT updated by GPS or any other source during flight.

The FMC or FMGC is, but not the ADIRU/IRS

Hence the Airbus requirement to do an "IRS Performance check" at the end of each flight. :ok:

– IRS PERFORMANCE...........................................CHECK
•
Residual ground speed check:
— If the residual ground speed indication on any IRS MONITOR page is greater than 15 kt, make a Maintenance Log entry identifying the associated IRS. Perform this check within 2 min of aircraft stop.

Position error check:
•
— If any radial distance error on the POSITION MONITOR page is greater than the limits in the table below, make a Maintenance Log entry identifying the associated IRS and stating the BLOCK/NAV time and error in nautical miles.

[FlightPathOBN]

http://operationsbasednavigation.com/wordpress/wp-content/uploads/2011/04/A320FCOM.jpg

[FlightlessParrot]

On the relative accuracy of mechanical and laser gyros, as of the 1980s, I found this book interesting: _Inventing Accuracy_, Donald MacKenzie, MIT Press, 1990. More history of technology than state of the art, though.

[aterpster]

TO: Flight Path and BOAC.

It takes a team to forumate operations procedures such as what we (in this case "we" being a team of industry experts with different backgrounds, and the FAA) who cobbled together the criteria and general obstacle clearance and performance parameters for RNP AR. Two of our experts were FMS engineers. We left the details about how IRUs perform the position calculation and where in the "black box" stream the GPS sensors updated the final position calculation; the FMS position.

Others of us had flight operations backgrounds and a few of us had TERPs experience, either formal (FAA) or in my case informal.

Throughout all of this all I needed to know is that the IRU postions were massaged by the FMS (sort of like triple mix in the L1011s I flew) and GPS, when available, indeed performed a postion update at some point. Whether it actually tweaked the IRUs or the FMS positon was immaterial to us operations types unless our FMS engineers told us it matter. They spent a career being trained, then working on that stuff.

Personally, I often see operations folks such as pilots trying to be avionics engineers when they aren't. Then again, they fly airplanes a whole lot better than FMS engineers and they, the pilots, have invaluable input as to how the airplane will behave in the performance based nav world.

I don't know which of you is right, and it is unimportant to me for my purposes.:)

[nitpicker330]

All very well but if the actual IR position is updated by the GPS in flight then why
1/ require a IRS performance check post flight
2/ why do I always see residual groudspeed and position error from the 3 IRS's

[3holelover]

Maybe you're flying something without gps updating. Just because it exists, doesn't mean you should expect all birds to have it. I'm certain the 320's I know don't have it.

[FlightPathOBN]

Exactly, there are many ADIRU systems, and not all have a GPS update. The Boeing variants that are certified down to 0.1RNP would have to have this, or the drift rates would push it to a 0.3 RNP...
Honeywell recently announced their ADIRU for the Airbus 300 family, so it will take a bit for the certification to catch up. The A380 is already certified.

[mickjoebill]

In the land of airborn stabilised platforms (gimbals) for use by TV police and military, in general two categories of gyros have been in use. ITAR and non ITAR restricted.

In general ITAR restricted gyros are significantly better than others, one can see the difference in picture stability when using the equipment.

(ITAR restricts the export of particular components or equipment with components installed to countries that are not friendly with the western world)


Which type of gyros are in use in Airbus and Boeing ADIRUs?



Mickjoebill

[Denti]

FPO, the 737 does not have internal IRU updating and is 0.10/GLS certified as standard equipment. The GPS receivers are part of the MMR, not part of the ADIRU. And one can see quite a bit of IRS run-off after a lengthy flight, just recently after a 6 hour sector had the IRS's around 2 to 3 NM off the FMC/GPS/radio position, and the ANP was still 0.02. There is no need for internal updating as the IRS is only used for its short term stability and the GPS provides long-term stability. Filtering based on that principle is done in the FMC.

[aterpster]

Denti:

FPO, the 737 does not have internal IRU updating and is 0.10/GLS certified as standard equipment. The GPS receivers are part of the MMR, not part of the ADIRU. And one can see quite a bit of IRS run-off after a lengthy flight, just recently after a 6 hour sector had the IRS's around 2 to 3 NM off the FMC/GPS/radio position, and the ANP was still 0.02. There is no need for internal updating as the IRS is only used for its short term stability and the GPS provides long-term stability. Filtering based on that principle is done in the FMC.

Any 737s that can qualify for RNP AR IAPs of less than RNP 0.30 have to have FMS position updating to the MAP, so that in the event of a loss of GPS (sats, not airplane receivers) the FMS position derived from the IRUs has zero drift at the MAP.

None of this is required for long-range navigation.

When I flew the early generation 767, which had no updating, and no GPS, our IRU position error was typical around 1 mile for all three IRUs at the end of a 6 hour flight.

[nitpicker330]

We do have all the latest wiz bangs in our CX a/c including GPS updating RNP etc...

You cannot update the basic raw Inertial Reference Laser Ring Gyro position in flight. Once the IR is aligned on the ground it's done and dusted.

You can however adjust that position after its left the IR unit using GPS, but the IRS still thinks it is where it is…………..


Ahhhhhhhh it's all to complicated anyway, white mans magic !! ooh can I say that?

[CJ Driver]

Unless I am very much mistaken, the original question was only about attitude and heading reference gyros, and nothing to do with position. So, notwithstanding the discussion around position that ensued, here's my 2c worth:

Many (most?) EFIS solid state gyros do NOT have GPS connections to aid them. They use conventional magnetic slaving and acceleration (gravity) based erecting.

They also have errors, some of which increase during a flight. These errors are often greater than those of an inertial navigation system, although the absolute errors are still quite small. The reason that even tiny errors in an inertial system are important is that they are integrated over a long period of time. For example, if your calculation of ground track is wrong by one quarter of a degree, then after 5 hours of oceanic flight, your position solution will be "out" by quite a bit. On the other hand, if your heading display on the EFIS is out by one quarter of a degree, you will never notice or care.

As a small digression, in the very low-end EFIS systems, such as those used in experimental aircraft, they use much simpler systems which primarily rely on accelerometers. These rely on integrating accelerations to determine attitude, and they also therefore accumulate errors quite quickly. They generally DO require a GPS input as a "hint" to continually constrain the error bound.

[BOAC]

As a small digression, in the very low-end EFIS systems, such as those used in experimental aircraft, they use much simpler systems which primarily rely on accelerometers. These rely on integrating accelerations to determine attitude- care to elaborate? How do the 'accelerometers' know which way is up/sideways/left and right? I have never heard of integration of acceleration to produce attitude.

[aterpster]

OK465:

For any RNP AR below RNP 0.3 GPS is required. Even some 0.3 procedures do not allow DME/DME only solutions.

In airspace regulated by the FAA, no RNAV IAPs are permitted to use DME/DME updating. GPS is mandatory, whether conventional RNAV, RNP AR, or emerging advanced RNAV. DME/DME updating is permitted in Level 1 SIDs and STARS provided DME masking is demonstrated to be sufficent. In the case of RNP AR most, if not all, ICAO states have elected to use FAA's RNP AR criteria, as set forth in FAA Order 8260.52.

Though highly unlikely to lose GPS due to 'loss' of sats, as long as ANP is below RNP you can continue. (You probably wouldn't notice a sat loss unless you were observing that page in the MCDU.)

Indeed, a loss of GPS is unlikely but it is nonethess the assumption made for either minimums predicated on RNP of less than 0.30 and/or a missed approach that requires RNP of less than 1.0.

And in the 738, the first indication of FMS GPS positional problems (sats or receivers) will probably be an amber TERRAIN POS message on the ND as the ANP builds followed eventually by the amber UNABLE REQD RNP. With GPS probs, the IRU's will not prevent this from occurring even if they are perfect and you can't continue to the MAP.

Agree, and thus a loss of GPS on a dual-string RNP AR IAP will likely require aborting the IAP prior to the MAP.

But, the assumptions in criteria are worst case; loss of GPS at DA.


In the '73 the IR pos is third in the hierarchy as far as FMS weighting and the IRU's themselves (att or pos) are not in anyway updated by the GPS.

Which is they way most, but not all systems, function. The point is, the output of the IRUs (or IRU) is updated so the apparent IRU position has no significant errors in the event GPS is lost.

There are three types of missed approach procedures permitted by Order 8260.52:

RNP of less than 1.0, also known as "telescoping RNP," which requires at least one IRU, and because the worst case assumption is loss of GPS at the MAP, positive course guidance is then required from the IRU or IRUs, which position is updated in the FMS continuously until/if a loss of GPS occurs. The telescoping increasing size of the missed approach procedure is calculated upon agreed upon assumptions for IRU drift once GPS is lost.

Following are the two RNP missed approach containment areas from Order 8260.52. The first one requires at least one IRU and turns are permitted either with TF of RF legs (RF legs highly preferred.) The second illustrated RNP missed approach containment area does not require updated IRU position (unless the minimums are predicated on less than RNP 0.30, in which case at least one updated IRU position is required). In this second case no turns are permitted until the splay to Level 1 RNP is competed.

If the terrain is benign the third option is a conventional TERPs (or PANS-OPS) missed approach procedure to any nav aid or even DR. And, a turn is permitted (if necessary) as low as 400 feet AFE.

(The RNP 0.10 final approach segment in Figure 4-1B is an example of maximum application of criteria. the FAS could be as high as 0.30 and, if this type of missed approach is dictated by obstacles or airspace, then at least one IRU is required.)
http://i201.photobucket.com/albums/aa214/aterpster/RNPMAS.jpg

[BOAC]

No - plus padding for the post!

[NSEU]

FlightpathOBN, your manual says "for GP/IR position calculation". As far as I can see, the original question was about "attitude".

The laser gyros are susceptible to cumulative errors. The accelerometers in the IRUs/ADIRUs, however, are sensing a gravity component (not just aircraft acceleration) and this applies a long term correction factor to attitude.

Rgds
NSEU

[Dan Winterland]

And the GPS input to the ADIRS actually comes from the MMRs, so it's a separate component.

We all use FMC position which is derived from several inputs, the IRSs included. What happens to those inputs to produce the position is obviously of some interest to us, (the Aibus FCOMs are very good on this) but of little relevance to our everyday operations. If the ANP is within the RNP, then all is well and generally we are happy. I've used four different types of FMC in my career and all derived their position in different ways. One used a component called an EGI (Embedded GPS IRS) as the main source.