compass acceleration error
John Baron
magnetic compass in a plane, but neither my texts, nor my CFI can explain
the mechanism behind this effect.
Can anyone enlighten me on this? I always prefer to know WHY something
happens, rather than just that "IT DOES"....
Garner R. Miller
>I understand the EFFECTS of acceleration on the heading indicated by the
>magnetic compass in a plane, but neither my texts, nor my CFI can explain
>the mechanism behind this effect.
Let me take a stab at it. This may take a screenful or two.
Are you familiar with Magnetic Dip? It's key to explaining the
acceleration error, so I'll attack that first in case you're not
comfortable with it.
Some of this is difficult to picture without a diagram to point at, so
I'll try and put it into words the best I can.
If you've got the Instrument Flying Handbook handy, the picture I'm
looking at is on page 47. The Flight Training Handbook really should
include some of this stuff too, but it doesn't. I wish I had a scanner;
I'd send the picture along with this.
The magnetic field is usually drawn showing "lines of force" encircling
the earth. Visualize several heavy lines going from the North Pole,
following the curvature of the earth several miles from the surface, and
then coming back down and touching the South pole.
At the equator, the lines are a good distance from the surface of the
planet, and are parallel to the surface. As you move closer to each of
the poles, you'll see that that lines bend closer and closer to the
surface of the planet, until they almost point straight down when they
reach the Pole. If you were to follow them from the equator all the way
to the North Pole, you'd be in a gradual descent the whole way down.
What does this have to do with the compass? Well, the compass magnets
align themselves with those magnetic field lines. That's how it "finds"
North.
Have you seen the little trick where you stick two forks in either side of
a cork, and you can balance that whole contraption on the head of a pin?
The compass is mounted basically the same way. (The round compass card
and the magnets take the place of the forks, but it's still mounted on a
pinhead, with the center of gravity well below the pivot point.)
Because of the way the compass is mounted and balanced on that pin, it's
very easy for it to get tilted off of horizontal. When you accelerate the
airplane, for example, the aft end of the compass card (the side closest
to you) is tilted upward; the far side, then, is pushed downward. When
you slow down, it tilts the other way.
Again, a picture would help, but try this if you're having trouble
visualizing the tilt: Take something like a metal salad bowl with a round
bottom, and balance it upside-down on the tip of your finger. Now push
forward.
You'll see that since the center of gravity is below the pivot point,
you'll have more of an effect on the higher side, and it'll tilt away from
you. That's exactly what happens when you accelerate the aircraft's
compass, and the card starts tilting away from you.
When the card is tilted like this, the plane of rotation of the magnet has
been tilted along with it. The vertical component of these magnetic field
lines makes the north-seeking side of the compass dip towards the North
pole. The magnet is being attracted by those strong, bent magnetic
fields, and the magnet is trying to stay aligned with them.
When the card is in level flight, it can't really align itself with the
curvature of these bent magnetic fields. With the card at an angle like
that, though, the magnet *CAN* rotate to match the angled magnetic
fields. So it does.
Let's put this into practice: Say you're in in level flight, heading
East. Facing east, the north pole is to your left, so these magnetic
field lines are closer to the surface on your left (toward the pole) than
they are on your right (toward the equator).
You accelerate. The card, as before, is going to tilt away from you. The
north-seeking end of the magnet (the "S" on the compass card), then, is
going to want to point at North, just as it always does. But remember
that North isn't just to your left--It's DOWN and to your left.
Tilted, the compass card is free to truly point at the pole, which means
the north-seeking end (the "S") will move down towards the earth a few
degrees. This brings the "N" over towards you on the other side,
indicating a turn to the North. And that's the acceleration error.
You don't really care that the pole isn't exactly off your left wing, but
the compass magnet doesn't know any better. When it's tilted, the card is
free to align itself with the vertical component of those magnetic field
lines, and that's exactly what it does.
Does that help?
>Can anyone enlighten me on this? I always prefer to know WHY something
>happens, rather than just that "IT DOES"....
My mind works much the same way. When somebody tells me something works,
I have to know what exactly what makes it tick. (For example, even after
flying with them for years, I'm still amazed at how clever an invention
the constant-speed propeller is, with that ingenious use of centrifugal
force and oil pressure. The more I study it, the more intrigued I am with
it.)
I hope that helps clear things up. If not, let me know.
In the meantime, I think it's time I added a metal salad bowl to my
collection of props! It seemed much more concrete to me than a black and
white picture in the Instrument Flying Handbook.
Take care,
- Garner
--
Garner R. Miller, CFI-A
>> Pentiums & Deodorants: When being Close is all that counts
James Strickland
Your instructor really should know this... both turning and acceleration
errors are due to magnetic dip, which is simply the compass wanting
to point down at the ground because the magnetic field lines point
down to the north pole. (Assuming you're in the northern hemisphere - if
you're near the equator there will be very little dip because the field
lines are roughly parallel to the surface of the earth).
To counteract this a weight is put on the south pointing end of the
compass (you look at the back of the compass, so the south pointing end
is actually marked "North", the east side is marked "West", and so on -
that's why the compass is "backwards" wrt turns) so that the compass
will generally balance and be level. Note the word "balance" - the
compass is balancing on a knife edge, just like a teeter-totter.
I'll explain acceleration error and you can figure out turning error:
suppose you're flying east or west and you accelerate. The south
pointing side (marked "North", remember!) of the compass will lag
because of the weight on it. Thus the compass indicates a turn towards
north even though you haven't turned.
If this isn't clear, try emailing me.
--
James Strickland It's much healthier to drive than to walk or cycle
ja...@hoshi.cic.sfu.ca - after all, who wants to breathe all that exhaust?
Josef C. Frisch
|> jba...@gr.hp.com (John Baron) writes:
|>
|> >I understand the EFFECTS of acceleration on the heading indicated by the
|> >magnetic compass in a plane, but neither my texts, nor my CFI can explain
|> >the mechanism behind this effect.
|>
|> >Can anyone enlighten me on this? I always prefer to know WHY something
|> >happens, rather than just that "IT DOES"....
|>
|> Your instructor really should know this... both turning and acceleration
|> errors are due to magnetic dip, which is simply the compass wanting
|> to point down at the ground because the magnetic field lines point
|> down to the north pole. (Assuming you're in the northern hemisphere - if
|> you're near the equator there will be very little dip because the field
|> lines are roughly parallel to the surface of the earth).
|>
|> To counteract this a weight is put on the south pointing end of the
|> compass (you look at the back of the compass,
I have not heard this. This would imply that you need to get your compas
re-balanced when you signifcantly change latitude. Are you sure of this.
so the south pointing end
|> is actually marked "North", the east side is marked "West", and so on -
|> that's why the compass is "backwards" wrt turns) so that the compass
|> will generally balance and be level. Note the word "balance" - the
|> compass is balancing on a knife edge, just like a teeter-totter.
|>
|> I'll explain acceleration error and you can figure out turning error:
|> suppose you're flying east or west and you accelerate. The south
|> pointing side (marked "North", remember!) of the compass will lag
|> because of the weight on it. Thus the compass indicates a turn towards
|> north even though you haven't turned.
|>
|> If this isn't clear, try emailing me.
|>
|> --
|> James Strickland It's much healthier to drive than to walk or cycle
|> ja...@hoshi.cic.sfu.ca - after all, who wants to breathe all that exhaust?
There is an explanation which does not require the existance of a weight. (Note,
I don't know if there is or is not such a weight).
Imagine you are flying north, compass says north.
Bank right (to begin a right turn).
The north end of the compas is attracted somewhat down (by the magnetic dip).
Down in this case is clockwise. (hold your right hand flat, fingers pointing away
- north. Now rotate you hand so that the thumb is is higher than the little
finger. Now the tip of your middle finger - north compas end - tries to go down -
this is to the right. ) The compas turning right reads as the aircraft turning
left. The result is that when a right turn is initiated, the compas first shows a
left turn - this is compas lag. Apply the above rules and you can get all of the
other effects.
--- Joe Frisch ---
Not a CFI, but I am a physicist.
Roy Smith
> I have not heard this. This would imply that you need to get your compas
> re-balanced when you signifcantly change latitude.
It's certainly true of marine compasses. When you order a good marine
compass, the manufacturer wants to know what latitude range you will be
using it in.
--
Roy Smith <r...@nyu.edu>
Hippocrates Project, Department of Microbiology, Coles 202
NYU School of Medicine, 550 First Avenue, New York, NY 10016
"This never happened to Bart Simpson."
John Stephens
Josef C. Frisch <fri...@hebe.SLAC.Stanford.EDU> wrote:
|> There is an explanation which does not require the existance of a weight. (Note,
|> I don't know if there is or is not such a weight).
|>
|> Imagine you are flying north, compass says north.
|> Bank right (to begin a right turn).
|> The north end of the compas is attracted somewhat down (by the magnetic dip).
|> Down in this case is clockwise. (hold your right hand flat, fingers pointing away
|> - north. Now rotate you hand so that the thumb is is higher than the little
|> finger. Now the tip of your middle finger - north compas end - tries to go down -
|> this is to the right. ) The compas turning right reads as the aircraft turning
|> left. The result is that when a right turn is initiated, the compas first shows a
|> left turn - this is compas lag. Apply the above rules and you can get all of the
|> other effects.
|>
|> --- Joe Frisch ---
|>
|> Not a CFI, but I am a physicist.
Your explanation does fit for turning error, but it doesn't
address the acceleration error that you see when flying on an
easterly or westerly heading. The counterweight theory does.
I haven't actually taken my aircraft compass apart, but I have
dismantled some cheap handheld compasses, and they all did have a
counterweight on the south-seeking pole of the needle. And no, it
would only be necessary to modify the weight if you were to move
more than about 45 degrees of lattitude. And yes, this does tend
to shoot down the counterweight argument for aircraft compasses,
as aircraft can be expected to travel all over the world, in both
hemispheres.
******************************************************************
* . *
* John Stephens ._______|_______. Montgomery County Airpark *
* COMM-ASEL \(*)/ ( GAI ) *
* C-172P N51078 o/ \o Gaithersburg, Maryland *
* *
******************************************************************
Julian Scarfe
Stephens) wrote:
> Your explanation does fit for turning error, but it doesn't
> address the acceleration error that you see when flying on an
> easterly or westerly heading. The counterweight theory does.
>
> I haven't actually taken my aircraft compass apart, but I have
> dismantled some cheap handheld compasses, and they all did have a
> counterweight on the south-seeking pole of the needle. And no, it
> would only be necessary to modify the weight if you were to move
> more than about 45 degrees of lattitude. And yes, this does tend
> to shoot down the counterweight argument for aircraft compasses,
> as aircraft can be expected to travel all over the world, in both
> hemispheres.
If you were trying to balance the compass on a point, I can see the need
for a counterweight; but the standard aircraft compass (I forget the
designation) is pendulous, and its own weight *reduces* the angle to the
horizontal at which it hangs, but there's still a residual dip. I don't
think it uses counterweights.
[OK, somebody blow my argument apart by telling me that they *have* taken
an aircraft compass apart and found counterweights :-)]
This means that the centre of gravity is *not* directly below the pivot
point (there has to be a couple to balance the couple due to dip, so the
weight and the reaction to it must act at different points). Thus if
there's an acceleration in the horizontal plane, again there is a couple
because the inertia force (in the aircraft reference frame) and reaction
to it at the pivot are at different points. So the compass turns...
It's worth bearing in mind that the counterweight argument is very
similar: either way, the horizontal acceleration causes a couple because
the C of G (of the compass) and the pivot are not in the same vertical
line. Whether you achieve that by suspending the compass below the pivot
or hanging a counterweight on one end doesn't matter -- the effect is in
the same direction.
Joe Frisch's argument, which basically seems to be that the apparent
direction of magnetic north is changed if have dip present *and* you bank
the aircraft is, as far as I can see, a completely distinct (but perfectly
valid) effect.
We had this argument before (was it you that set me straight on that one,
Joe?), and decided that although a sensor system with no moving parts
could remove the first effect I describe above (the Strickland Effect),
you can't avoid the Frisch Effect if you bank the aircraft (unless you
give the sensor's computer a table of dip vs position).
Julian Scarfe
ja...@cus.cam.ac.uk
Josef C. Frisch
|> In article <D3G20...@unixhub.slac.stanford.edu>,
|> Josef C. Frisch <fri...@hebe.SLAC.Stanford.EDU> wrote:
|> |> There is an explanation which does not require the existance of a weight. (Note,
|> |> I don't know if there is or is not such a weight).
|> |>
|> |> Imagine you are flying north, compass says north.
|> |> Bank right (to begin a right turn).
|> |> The north end of the compas is attracted somewhat down (by the magnetic dip).
|> |> Down in this case is clockwise. (hold your right hand flat, fingers pointing away
|> |> - north. Now rotate you hand so that the thumb is is higher than the little
|> |> finger. Now the tip of your middle finger - north compas end - tries to go down -
|> |> this is to the right. ) The compas turning right reads as the aircraft turning
|> |> left. The result is that when a right turn is initiated, the compas first shows a
|> |> left turn - this is compas lag. Apply the above rules and you can get all of the
|> |> other effects.
|> |>
|> |> --- Joe Frisch ---
|> |>
|> |> Not a CFI, but I am a physicist.
|>
|> Your explanation does fit for turning error, but it doesn't
|> address the acceleration error that you see when flying on an
|> easterly or westerly heading. The counterweight theory does.
|>
accelerate. The compas will tilt forward (it is weighted so that it is level with
respect to local gravity/acceleration - like a turn coordinater). Now the north
side of the compas was to the left (north). It now wants to point down (to follow
the magnetic dip). This causes your apparent course to move towards north. This
gives the standard Accelerate North, Decelerate South.
I still don't know, however if there is a weight.
BTW, this explanation is (in very abreviated format) in Jeppson - I didn't make
it up (and in fact it took me quite a while to understand it).
--- Joe Frisch ---
Josef C. Frisch
|>
|> We had this argument before (was it you that set me straight on that one,
|> Joe?), and decided that although a sensor system with no moving parts
|> could remove the first effect I describe above (the Strickland Effect),
|> you can't avoid the Frisch Effect if you bank the aircraft (unless you
|> give the sensor's computer a table of dip vs position).
|>
|> Julian Scarfe
|> ja...@cus.cam.ac.uk
If the magnetic sensor were mounted to the AI, it would remain level during
turns, and would work. I don't know if anyone does this.
--- Joe Frisch ---
Les Niles
Here's what I earlier wrote in reply to this question:
In article <3eib9s$7...@cmcl2.NYU.EDU> r...@mchip00.med.nyu.edu (Roy Smith) writes:
> Can somebody explain to me about compass turning error?
Compass turning error arises because the magnetic lines of force that
the compass magnet aligns with are not horizontal, but "direction" is
really defined as an angle from magnetic north *in a horizontal plane.*
(Nota bene: "Plane" refers to the two-dimensional geometric object. A
flying machine will be referred to as an "airplane.") In unaccelerated
flight, the compass is constrained to rotate in a horizontal plane and
the magnet aligns as closely as possible with the magnetic lines of
force, which point toward the magnetic pole but dip downward instead of
being horizontal. The compass will then accurately point toward
magnetic north.
The error arises in accelerated flight, either when turning or changing
speed. Then the local apparent "horizontal" plane that the compass is
constrained by is no longer truly horizontal, but is tilted. (The
airplane does *not* need to be banked for this to happen, but in a
coordinated turn the tilt of the apparent horizontal will be exactly the
bank angle.) The compass magnet will now align as closely as possible
in this tilted plane with the dipping magnetic lines of force. Because
of the dip, this alignment will generally be in a different direction
from what it would be if the compass plane were truly horizontal -- this
difference is exactly the compass turning (or acceleration) error.
To see how the directions are different, consider an extreme case, where
the acceleration is so strong that the local apparent "horizontal" plane
is tipped over nearly vertical -- perhaps an 80 degree bank aerobatic
turn. Suppose one starts out unaccelerated on a northerly heading. The
compass indicates north and all is well. Then quickly roll into the
sharp turn. Now the compass will point not straight ahead, but downward
at roughly the dip angle of the magnetic field lines, something like 30
degrees. If the turn is to the right, from the cockpit the compass
magnet points to the right, which is really downward, by about 30
degrees. This would be interpreted as a heading of 30 degrees west of
north (330), even though the heading is actually north and the airplane
is turning east. Voila, compass lag on northerly headings. For
southerly headings, the directions are reversed and the compass leads.
It's only the tilt of the apparent horizontal plane along a north-south
axis that causes these sorts of errors. Turning through an east or west
heading, there will be no such tilt and thus no turning error. But a
change of velocity that has an east-west component will tilt the
apparent horizontal along a north-south axis (the way water sloshes fore
and aft when you speed up or slow down); this results in the compass
acceleration errors.
The reason you don't see the compass turning error in a boat or when
swinging a compass around by hand is that the speed is too low. The
error is proportional to the tilt of the apparent horizontal plane,
which is proportional to the acceleration. To get significant error, a
centripetal (rotational) acceleration that is a significant fraction of
one g is needed. To get high acceleration at low velocity requires a
high rate of turn. (Just like in an airplane, a given bank angle
results in a higher rate of turn at low speeds than at high speeds.)
When you're already turning rapidly, a little bit of compass turning
error is hard to notice. But you should be able to see the acceleration
error, if you put the compass in a car that's got decent acceleration
and punch it on a road that runs roughly east-west. (ObLegalDisclaimer:
This demonstration is conducted by professional drivers on a closed
course. Do not try this at home. The Secretary will disavow any
knowledge of your actions.)
-les ni...@parc.xerox.com
Mooney N9752M
Julian Scarfe
> |>
> |> We had this argument before (was it you that set me straight on that one,
> |> Joe?), and decided that although a sensor system with no moving parts
> |> could remove the first effect I describe above (the Strickland Effect),
> |> you can't avoid the Frisch Effect if you bank the aircraft (unless you
> |> give the sensor's computer a table of dip vs position).
In article <D3n6r...@unixhub.SLAC.Stanford.EDU>,
fri...@hebe.SLAC.Stanford.EDU (Josef C. Frisch) wrote:
> If the magnetic sensor were mounted to the AI, it would remain level during
> turns, and would work. I don't know if anyone does this.
Then I misunderstood your argument. Sorry, I'll think about it some more.
--
Julian Scarfe
ja...@cus.cam.ac.uk