IRS failed Cosine but goes NAV

[Xavier]

Hi,

Working through Bill Bulfer's FMC guide, I am currently playing with wrong manually entered coordinates for IRS alignment and corresponding consequences.

As I wanted to test the behaviour of the IRS when I twice entered a latitude that would not pass the sin/cosine test, I was pretty surprised to see them obviously going NAV mode, despite the instructor station's Analysis page confirming a failed Cosine test...

What may I have missed ? I made a .situ file in case of, but I'm not familiar with uploads so will only do it if it is of any interest.

Thanks in advance for the help,

Regards

[Avi]

When you enter the same wrong coordinates the second time, it is accepted.

Cheers,

[United744]

Boeing philosophy is generally one of "tell the pilot he is being silly, but don't stop him".

[John H Watson]

Even when the same wrong latitude is entered and accepted, an EICAS fault message should appear. The IRUs also carry out a final sine/cosine check before entering NAV mode. Wrong identical latitudes may be accepted, but gross navigational errors may result.

Rgds
JHW

[Xavier]

Ok, my bad then.
I was under the impression that there were two different cases :

- too big a difference with "last pos" (over 1°) would be overidable by a second entry ; it is more or less a mean for the system to get confirmation that it was not just a typing error

- sin/cosine test failed would prevent the IRU from entering NAV mode, as it would prove that the IRU is unreliable since its computations are wrong

[United744]

- sin/cosine test failed would prevent the IRU from entering NAV mode, as it would prove that the IRU is unreliable since its computations are wrong

The test failure is only because the difference of where it *THINKS* it is (sensed position), is different by more than a certain limit from the ENTERED position. If you "insist" that the (incorrect) entered position is correct by entering it twice, the system assumes you are correct and accepts it, because you could have moved the aircraft 3000 miles and not told it.

Physics however, will ensure high drift rates, and premature failure of navigation if you tried to actually fly it (ignoring the obvious problems of using a system that thinks it is half-way around the world from its real location).

[Avi]

because you could have moved the aircraft 3000 miles and not told it.

If I remember correctly the Bulfer's guide says you can move the IRU box (3000 miles) without it knows this (as a spare part in the cargo compartment, not as an active box).

[Hardy Heinlin]

Has someone tested this with the real IRS actually? Will it really go to NAV mode when the sine/cosine test invalidates the pilot-confirmed latitude at the end of the alignment?

As we know, it doesn't matter if the IRS has been switched off and moved 3000 miles before starting a new alignment; during the alignment it will sense its new latitude on its own within a certain  tolerance. If the difference between the sensed latitude and the pilot-confirmed latitude exceeds the allowed sine/cosine tolerance, there must be a fault in the sensor system. What error message will occur then?


Regards,

|-|ardy

[Xavier]

Thanks for your answers.

Bill Bulfer's guide gives details about the logic behind the sin/cosine test and consequences of its failure (p.50), but they are not 744 specific and mention annunciator lights that we don't have... and a pretty different behaviour.
Sadly no mention of what the 744 EICAS or s/p should read.

[kevmac86]

Hi all,  just to throw my 2pence in.

As I understand the IRU's, if an incorrect longitude is entered twice then the IRU's will accept the position, they cannot workout longitude.
As for latitude, if you enter the latitude wrong twice,  the IRU will report a failure and not enter navigation mode, This is indicated by the EICAS advisory message on the main EICAS display. The IRU must be switched off for 30 seconds until the EICAS memo message disappears. The mode selector switch should then be placed to NAV, where a new alignment begins.

This helps to find a failed IRU.

Kev Mac

[Hardy Heinlin]

Hi Kev,

I think you're right. If nobody is complaining, I'll add this feature in the next update and let the advisory IRS (L, C, R) appear.


Cheers,

|-|ardy

[kevmac86]

Thanks Hardy that is a good solution.

It would be nice if any BA drivers out there could test it next time they are in Braincrank ,  I don't think any of the engineer instructors I know are doing a 744 course in the near future, so I can't get them to test it.


Kev Mac

[United744]

Hi,

I had a look in the QRH but found nothing for IRS alignment failure as a result of bad position entry. The QRH for IRS LEFT/CENTER/RIGHT doesn't state to check entered position (or even what to do if it occurs on the ground).

It states to set the affected unit to ATT and maintain level flight for 30+ seconds, or switch it OFF if the message persists, but otherwise doesn't say anything about ground alignment phase (and I can't see you deliberately departing with one or more INS in ATT mode).

[Hardy Heinlin]

Hi,

I'm wondering if the scratchpad message "CYCLE IRS OFF-NAV" should also appear when the afore-mentioned sine/cosine test fails at the end of the alignment phase.

Or should this failure be considered a case for the maintenance engineers? I.e. don't recycle OFF, then NAV; forget it, captain, call maintenance.


Cheers,

|-|ardy