Ok here I go :-)What exactly drifts in the IRS?Is it only the position, ie. latitude and longitude or also groundspeed?I've been thinking of this, and what I think is that the groundspeed is basically a "drift factor". If IRS1 has a much higher groundspeed than IRS2 and IRS3, you can assume IRS1 will drift much faster/further than IRS2 and IRS3, as when you integrate the groundspeed over time, you get distance you flew in that time and thus the "position offset". Right?Thus the groundspeed of a specific IRS will not drift over time, it will only display a different value than the other IRS's throughout the flight?Hope this makes sense :-)Thanks,Mark

IRSes start with gyros to measure attitude/heading and accelerometers to measure, well, acceleration. Thus, they integrate acceleration once to get velocity and again to get position. I think that after they are properly calibrated, the position error ends up being a random walk, so I don't think you can say that a specific IRS groundspeed will not drift over time. The noise in the system is basically noise in measuring acceleration and its direction, and thus the noise affects velocity and position. The goal of calibration is to remove biases (i.e., make the errors zero centered).I believe that aviation IRSes are accurate to a few miles after hours of use, so the velocity errors must be quite low, probably on the order of a few kts. I seem to recall Ian or someone else here saying that the IRS errors are logged after every flight to make sure they are within spec. (I actually used to work a bit next door at the Charles Stark Draper Labs that was big into designing IRSes for aircraft, spacecraft, and ballistic missiles, although I wasn't that involved in IRSes. I think Draper himself basically invented the IRS while part of the MIT Instrumentation Laboratory.Anyone remember the Trident missile that did a real tight 360 right after breaking out of the water and had to be destroyed during a test? We had a supplementary IRS onboard for evaluating the primary IRS, and we got back pieces from the ocean floor. The destruction charges did quite a number on parts of our system, but some of our circuit boards worked fine after cleaning out the salt water.)Lee Hetherington (KBED)

"IRSes start with gyros to measure attitude/heading and accelerometers to measure, well, acceleration"I thought the gyros were the accelerometers? Having three laser gyros all in a 90 degree angle to one another and using the optical doppler effect, you can measure the difference between the normal laser light's frequency and the "accelerated" light's frequency, and knowing the speed of light, you can calculate the acceleration.At least that's how I always thought the IRS's work.(light that is accelerated doesn't spread out faster, it's frequency is changed)

I thought the ring laser gyros were just used to measure rotation. My understanding is they set up an optical standing wave pattern in a circle (or square or triangle), and as you rotate you'll pick off and count peaks that you pass.The Doppler effect has to do with velocity, not acceleration. (Think of a train whistle going by at constant velocity.)Lee Hetherington (KBED)

Clear as Ponskey to me.

But if you know the time over which this velocity was accelerated, you can compute the total acceleration, right?

Would the real IRS pros please help us out? I'm not a real IRS pro. I just play one in the forums. ;-)Lee Hetherington (KBED)

Ahhh yes ,and speaking of ponskey have you met his brother Ponderosa?

Dito :-)

Hi guysIt's the IRS guru speaking... ;-)I'm doing a flight back from BIKF to EKCH and thought I'll take the time to catch up on IRS operations.According to the FMC Guide the inertial guidance system consists of three laser ring gyros, and three accelerometers. The accelerometers meters the acceleration in each axis and the laser ring gyros (also strapped down in the three axises) are giving information on the changes between these three accelerometers. Basically... :-)This information is based on the 737 equipment but I think the main idea applies to the 767 as well.Regards

My head just exploded ;-)I will now vanish in a puff of logic.

Well guysI thought I'd give you an explanation in depth as well ;-)The gyros are measuring the aircrafts speed as well as the speed of earth's rotation which is approximately 15 degrees/hour. That is one revolution around its own axis every 24th hour and one revolution around the sun every 365,25th day. This latter speed is substracted from the aircrafts speed to give a speed in inertial space.The 15 degree/hour rate is a sum of two rates on the ground. The earth rate and the azimuth. At the notrh pole. The earth rate is zero and the azimuth is 15degr/hr. At the equator the earth rate is 15 degr/nr and the azimuth is zero. Between these locations the earth rate vary with the cosine of the latitude. The computers are substracting this value from the signal recieved from the gyro pointing eastwards. Without this compensation any IRS could mistakenly believe that it's upside down after 12 hours of operation or similar errors in pitch, roll and heading. If you would enter the wrong latitude the IRS would presume the wrong earth rate. This in turn means that the IRS can never find true north. In that case the ALIGN would not stop flashing.Wanna know some more?.... lolTGIF :-beerchug

Hi MatsOk, so the IRS's want to know an irrelative speed, movement (assuming our solar system, our galaxy and our universe are not moving, which is not the case but doesn't matter as we can set our point of relativity anywhere, right?) (same as saying the potential energy is 0 when a second mass is put on the surface of the earth, everything's relative :-) )Ok, so the IRS can measure the earth rate and the azimuth. So this would explain how the IRS can calculate the latitude, while longitude has to be input, as the azimuth / earth rate are constant at a given latitude on any longitudinal position, correct?To measure true north, does the IRS need the longitude or is it able to measure it by its own using the azimuth / earth rate's movement?To get back to my original question :-), what will eventually drift?Position? Groundspeed indication?Thanks for any info,Regards,Mark

Follow these Linkshttp://www.pprune.org/forums/showthread.ph...0&highlight=IRShttp://www.pprune.org/forums/showthread.ph...3&highlight=IRShttp://www.pprune.org/forums/showthread.ph...6&highlight=IRS

Thanks for the links but, at the risk of sounding arrogant, they are a tad too basic ;-)They all say the same - the position drifts over time. But does groundspeed? ;-)Regards,Mark