How does the ADC work out wind components ?

[anengineer]

Speaking as a computer, rather than an aircraft, engineer, I've always wondered how the on-board computers manage to calculate wind components. Is it calculated from the GPS data, i.e. expected position - actual position = drift ? I've never quite got my head around how a huge lump of an airliner screaming along at mach .8 can manage to calculate / measure that there is a 5kt headwind ! How is it done ?

[kenparry]

The names of the boxes depend on whether you are talking Airbus or Boeing. My experience is on the latter.

The FMC (Flight Management Computer) is the box that calculates the wind. The inputs are:

Track, heading and groundspeed from the nav system (varies between models, but typically inertial with updates from GPS and terrestrial navaids); and

TAS from the ADC (Air Data Computer)

The FMC then does a vector triangle solution, the 3 sides of the triangle being:
Heading & TAS
Track & groundspeed
Wind direction and speed.

On the ones I flew, the displayed data was updated about once or twice a second. Brilliant stuff: going down finals into Palma on a sea-breeze day, for example, you could watch the wind changing from tail to head as you went down the slope. On a lumpy day anywhere, instant and accurate display of the wind affecting you - often much different from that reported by the tower from their distant measuring point.

[glum]

My understanding:

An inertial unit alone can calculate drift angle, since it knows the aircraft heading & the actual track across the ground. Difference between the two is your drift angle.

Similar thing for airspeed wit the addition of an Air Data Computer - you have probes / computer for calculating airspeed as you fly through the air, and an inertial unit which is calculating your speed across the ground. Difference gives you the windspeed component.

Olden days (C130H for example) they used things like doppler radar for this, but I'm not sure that's still in use now.

[SNS3Guppy]

An inertial unit doesn't really know your heading; it knows your track, referenced to whatever input keeps it updated, or once updated it before departure.

The air data computer doesn't know heading or track; it knows air pressures.

The magnetic reference system, or compass, knows heading, but not track.

The flight management computer, flight management system, or other computation systems in use collect inputs from various systems on board the airplane, and use these systems to come up with the data that's presented to the pilot.

The ADC provides data which is used for a number of systems on board the airplane, depending on which airplane is in question. Information may be provided to the navigation computers, but also to aircraft instrumentation, pressurization, and engines and engine instrumentation.

Which inputs are used to determine winds aloft really depend on the system in use, but the basic theory is always the same. The computer needs to know how fast it's moving, and in which direction. If the computer knows the heading and track, and the true airspeed, the computer can calculate the wind component...no different than calculating the wind component using a hand-held aluminum E6B computer.

ADC input is but one input to the computer to allow it to make that calculation.

[grounded27]

Woahhhh, slow down.

An ADC calculates raw air speed, TAT in addition helps with MACH/off topic but bottom line this is what is required for airspeed calculation.

The bottom line is that an ADC or DADC does not work out wind components as in reference to calculating navigational position/progress (can not think of why other you would ask this question)

[glum]

"An inertial unit doesn't really know your heading; it knows your track"

Yes it does, it knew it the moment it aligned on the ground - it's strapped to the airframe and knows which is the front!

Track is merely measurement of the movement in both the longitudinal and lateral axis at once.

[Brizeguy]

Had a Hawker through last shift, crew complained of a split between the wind indications.
Here is the answer from Collins


Quote

Wind is the "garbage collector" in the system.

The FMS computes track angle and ground speed from the successive position computations. The FMS receives Heading and TAS from the system. The FMS then solves for the final side of the vector triangle and computes wind (side 1- HDG/TAS; side 2- TRK/GS; side 3- Wind direction/magnitude).

In a perfect world, the difference between TRK/GS and HDG/TAS is actual wind (direction, magnitude). Of course, the world is far from perfect...

Any errors in position can affect the TRK/GS side of the triangle. Any sensor errors can affect the HDG/TAS side of the vector triangle. Consequently, the wind computation is the collective result of all (position and sensor) errors. A portion of the computed wind vector represents the "actual wind", and the remainder is a summation of all other (position and sensor) errors.

Previously, when a customer called in and reported errors in wind, I would ask them if the two FMCs (assuming dual equippage) had numbers for track angle and ground speed that were 'pretty close' (meaning within a degree for track angle and with a knot or two for groundspeed). If the answer was yes, then the FMCs are navigating fine. The differences in wind are a result of collective errors from either the heading and true airspeed data, the position computations, or some of both.

In those cases where the track angle and ground speed agreed within a degree and a knot, there is nothing wrong with the system, so no amount of changing LRU's will affect the result (unless the issue was sensor error, and by changing a box the sensor error is reduced). In some cases, changing a LRU did not affect the position computation performance, but actually made the wind issue worse.

I would expect that stronger winds will indicate closer to each other in direction, where lighter winds will be more likely to differ by greater angular values.

Probably not the answer you were hoping for, but this has been my experience with our systems. Due to the nature of the computation (and I'm guessing this issue is more pronounced with the AHRS based system), wind is a garbage collector. I've never correlated the differences seen between AHRS and IRS equipped aircraft, but I can believe that the heading sensor could make a huge difference in the goodness of the wind values provided.

I am aware that there is one OEM that chose to strap the system to provide IRS wind (instead of FMS wind) as the value displayed on the EFIS


Best Regards,


Franklin S. Gutierrez III
Principal FMS Product Support Manager
Customer Service
Rockwell Collins,Inc.

[WeekendFlyer]

As someone who has spent the last 7 years doing aircraft avionics systems testing I hope I can provide you with a satisfactory answer!

The ADC provides air data from pitot and static pressure sources, and also an air temperature sensor (usually total temperature). By means of look-up tables and a bit of maths the ADC then outputs various speeds and height/static pressure data. For navigation purposes the most important speed is True Air Speed (TAS), which is the speed the aircraft would be making over the ground if the airmass it was flying in was stationary. The aircraft TAS vector points in the direction of the aircraft heading (HDG), which on modern aircraft is provided by the INS / IRS / ADIRU depending on aircraft type.

Aircraft actual ground speed (GS) and track (TRK - which differs from HDG by the drift angle due to wind) are measured by GPS and/or INS / IRS depending on aircraft type, but if the aircraft is certified for area navigation (RNAV) it will almost definitely have a GPS. GPS cannot measure aircraft heading, so there will always be a sensor for this (e.g. INS or IRS), and neither can it measure TAS (or any other airspeed), hence the need for an ADC or ADIRU.

Once you have the HDG/TAS vector (from the ADC and whichever sensor provides heading data) and the TRK/GS vector (from the GPS or INS), a simple vector calculation will provide the instantaneous wind direction/speed vector. The 3 vectors always join up to form a triangle.

Hope this helps!

WF