View Full Version : inertial reference system (A320)

21st Sep 2000, 00:34
IRS alignment, how is it that the aircraft knows which way that it is pointing?

How does earth rate function determine this?

There is no magnetic input to the Honeywell system. Is it a function of Lat.?

21st Sep 2000, 11:34
Ok, its been a while since I've studied this, but I'll have a shot anyway.
When you do the align you enter in the lat and long, so the system knows where on earth it is. When it alings itself the aircraft has to be stationary, so that any movement felt by the system is caused by the earths rotation. If the system is pointing to north (that would be true north) then one of the axes of the system will not detect any displacement, as all the rotaion will be at 90 degrees to it.
Right, so it knows where true north is, which brings me to the fact I have no idea how it then works out where magnetic north is. I can only imagine it either knows the magnetic variation over the earth, or knows where the magnetic north pole is and works out the variation for itself.
And that concludes a thorough and deatailed answer to your question :)

21st Sep 2000, 14:42
The simple explanation: you know when McGyver is locked in a trunk and the bad guys drive him off to someplace, but he figures out where he is by keeping track of which way the car turns, and how long it goes at what speed? That, in a nutshell, is how the IRS works. You tell the system exactly where it is on startup, by programming in the coordinates of the stand, then a very high tech gyroscope keeps track of every change in motion from then on.

The technical explanation: It uses 'laser ring gyros.' Instead of having spinning mass like mechanical gyros do, the laser gyro has two homogeneous light beams following a circuit in opposite directions. There are two kinds of ring lasers, one with fibre optics, and a better one where the light follows a triangular path with mirrors. How does this act as a gyro?

Remember that nothing can travel faster than 'c', the speed of light, including light itself. (You have possibly been subjected to a thought experiment involving the headlights on relativistic trains. No matter how fast the train goes, the light from the headlight will not go faster than c, so that when the train is going at the speed of light, the headlight does not send a beam ahead of the train.) No matter how much of a "push" the light gets from its source, it won't go any faster than if it came from a stationary source.

When the aircraft is stationary, the light beams going in each direction a given ring gyro have exactly the same path length. Where they meet after a circuit of the instrument, they have a particular phase difference. I'm not certain that the instrument is tweaked to have it be zero, but without loss of generality, let's say that it is zero. The beams are perfectly in phase.

Working from these principles, consider what happens to the gyro in the horizontal plane, when the aircraft turns to the left. The gyro has turned around so as to reduce the path for the light beam going clockwise. The light going counterclockwise has had its path increased and it doesn't get a push from the turning motion. So the light beams will meet out of phase. Just as your VOR instrument works out what radial you are on by the phase difference between the two signals received, the computers associated with the ring laser gyro works out how much you turned, by the phase difference between the two beams.

The crew of an A319, who first told a disbelieving me about this system, swore up and down that it uses only inertial data.

(Oops, this post didn't actually answer the question that was asked. See below.)

[This message has been edited by Luftwaffle (edited 24 September 2000).]

22nd Sep 2000, 09:18
Yeah, that was a good explanation of how the things works, but the question that hasn't been answered is how does the system know where magnetic north is, having no external inputs other than TAS, which aint got a lot to do with magnetic north.
Anyone out there doing an ATPL care to enlighten us??

22nd Sep 2000, 19:07
What's wrong with a compass?

23rd Sep 2000, 01:12
I fly the A320?A320 and just looked through the flight manual. This is really a good question. It appears as if I have to agree with the other Airbus pilots: Heading (and we use magnetic heading like anybody else) is said to be produced by the Inertial Reference part of the system. A little something out of the limitations section of our flight manual: Inertial reference system: Ground alignment of the IRS has been demonstrated to be satisfactory below 70 degrees North latitude. In the NAV mode(note: our normal navigation mode), the IRS provides valid magnetic heading only between 73 degrees North and 60 degrees South .......

23rd Sep 2000, 02:41
Just read about some gyro stuff. What I said before about the north and south limitations: That has very little to do with the magnetic field of the planet. When you turn on an INS, it enters the alignment phase. It is supposed to be sitting on the ramp without movement. It still detects movement though; the movement of the earth since the gyro/laser gyro measures all movements in space. It take this movement/ acceleration and calculates at what latitude the unit is. Rotation speed is higher at the equator than further north and that's how the computer figures out the current latitude. This gets more difficult the closer you get to the poles because the rotaional speed differences get smaller and it gets more difficult to exacly determine the current latitude. The INS looks at this angular acceleration and since it knows where the rotation is from, it knows that the true north must be in a 90 degree angle to that. In flight, the computers constantly correct the sensed data for the rotation of the earth at any latitude and also for the aircraft following the curved surface of the planet which, when followed at a certain altitude, is sensed as a continous change of direction downward in space. Still couldn't find any of the magnetic stuff, but one thing is for sure: Either there is magnetic variation data in the data base which the system uses to interpolate magnetic variation at any give time or there must be some sensing input. So if you enter the current position during Alignment, the unit not only checks your entered position with the one it stored upon previous shutdown (reasonableness test), it also compares the latitude you entered with the one calculates to be on (system performance test). Before i go back to looking for more info, a little note from the manual: If there is a fault in the system where I can still use the IRS for attitude and heading info, I have to enter the magnetic heading (from the old Whiskey Compass) via entry in the overhead panel and redo this about every 10 minutes, just like the good old gyro compass.

23rd Sep 2000, 02:52
I'm pretty sure that as the system is a strap down system, and alignment is carried out mathematically, and as variation is a known quantity, with a known annual rate of change,then computing magnetic north shouldn't be too difficult. The hard bit has already been done on initial alignment.
Any other ideas out there...

24th Sep 2000, 02:14
OK, I got some expert advice on the magnetic: The sytems use the WGM - the World Magnetic Model. There is an International Geometric Reference Field that is developed every 5 years and this model is fed into the data base of the inertial reference systems. The current model is for 2000 - 2005. More magnetic info: www.ngdc.noaa.gov/seg/potfld/faqgeom.html (http://www.ngdc.noaa.gov/seg/potfld/faqgeom.html)

24th Sep 2000, 16:22

I was there when 'weewads' came up with the question, and all of us in the room could not come up with a valid answer. Yes, the a/c IR's do get their sensing from earth's movement, as I do not know of any flux valves on the plane or connections to any other magnetic source, but the meaning of the actual question was from a cold start while the IR's are aligning, how does the systems know the magnetic heading while in align. The heading will show up in "ARC" and the bottom of the PFD even while the align is completing it's cycle.

Cheers all...

24th Sep 2000, 19:19
This is how I understand it:

While the aircraft is sitting there on the ground, the east-west gyro is sensing the rotation of the earth. Obviously all the bits of the earth's surface go around the earth's axis once per day, but the bits at the equatorial latitudes have further to go, so they are going faster. The gyro senses how fast left to right movement seems to be. It then compares that to the amount of west-east movement that there should be at that latitude. (It knows the latitude because you just entered it). Now it's just like a crosswind calculation: if the crosswind ("crossearth"?) component is so much, and the total wind (earthrate) is so much, then what angle must you be at to the wind (earth's rotation)? The earth rotates about the axis that points to true north, so this process has determined the position of true north.

Atomic has provided the last piece of the jigsaw. Given true north, the computer consults the World Magnetic Model to determine magnetic north in relation to true north.

On another thread, someone whom I would credit if I could find the thread, pointed out that north of 70N (or south of 70S) the speed of rotation isn't sufficient for accuracy on the IRS setting, so the system is coupled to a GPS, which presumably can tell its orientation to the satellites it has found.

I believe the task of sensing the Earth's rotation accurately is the one that takes the ten to fifteen minute start up period.

24th Sep 2000, 23:17
To sum it up:
1. Put the machine into 'align' mode
2. Sitting motionless on the ramp, the inertial system senses angular motion (and it knows that this motion can only be in a east-west direction). This happens at a certain rate, that's because the earth turns at around 15.04 degress per hour in space (the result of the 24 hour rotation and the annual rotation around the sun). At the equator, you have travel 21600 nm divided by 24 = 900 nm in space every hour by sitting on the ramp. If you go north and follow e.g. the 75 degree latitude around the earth and measure the mileage (or do the math), it falls far short of the 21600 nm at the equator. Therefore the rotational speed is much less until it reduces to practically zero at the poles (neglecting the rotation around the sun). While it may take as little as 2 - 3 minutes to complete the alignment at the equator, it takes an easy ten minutes around the 70 degree north latitude. Further north and the system might not be able to align correctly before it starts to drift and might be unable to enter NAV mode because it doesn't come up with useful numbers to satisfy its criteria for accuracy.
It just gets more difficult to do exact calculations since the starting values are much smaller further towards the poles.
3. Angular rate is determined, the airplane knows where it is. It moves from West to East with the Earth's rotation. Since it knows where East and West are , North has to be in a 90 degree angle to that, that's all there is to finding true north. Knowing true north and where on earth it sits, the inertial system now goes into its data base, pulls up the current world geomagnetic model, checks the stored magnetic variation info for where it is and applies this and, voila, there it is on your nav display: magnetic heading.
Now we can navigate around the globe, while our computer has to constantly compensate for the Earth's rotation and also for the Earth's spherical shape (they call this 'Schuler tuning') since we fly on curved paths in space following the curvature of the surface.

25th Sep 2000, 19:22
...whiz: I see what you are talking about: Why is there a mag heading displayed even though the alignment process is not complete.
Check (and I will too) how soon after you put the system in align mode the mag heading will appear. If it's there immediately, maybe it is a remebered heading from shutdown?? Depending on where your latitude is, consider that even though your INS has the same alignment period allocated for every alignment, it provides PPOS and heading much sooner at lower latitudes than higher. The INS on my plane always counts down 10 minutes before it goes into nav, but it might take only 2 or 3 to get all the calculations done when I'm on the ramp somewhere in Central America, close to the Equator.

27th Sep 2000, 01:04

I fly out tomorrow again and I will see just exactly at what point and when the mag. heading comes up, and if it was just after the initial entry of the gate/stand position, I will try a few things before entry of the coords. even. Maybe we have a different brand of software from what's on your planes, I don't know. I do see some small differences from plane to plane in our fleet, but most are all the same generally.

28th Sep 2000, 15:33
No flux valves on an A320. It's all magic!

30th Sep 2000, 00:04
Yes, the 'Magic Bus'.....there is one answer for every problem " It's automatique !!!" I just came from from PHL and ORD and I tried to watch the thing while it aligns. Less than a minute after setting the ADIRS to 'align', the attitude was up on the PFD, it took about 4 minutes for the heading information to appear and the map only shows when the alignment process is over and the system goes to NAV. Next time I go close to the Equator, I'll check and see if it's any different then.

SEO by vBSEO 3.6.1