Jump to content

Talk:G-force/Archive 2

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1Archive 2Archive 3Archive 4Archive 5

Big-G dab?

Should we have a "Not to be confused with Big G" or something early on? Or should that be a matter for G-force (disambiguation)?

Disputed tag

This was added here on January 14 and I think the article has moved on. Any objections if we take the tag down? --John (talk) 05:12, 19 January 2009 (UTC)

  • No objection here. The editor who placed it there has not responded to your request that he state concisely what the exact nature of the disagreement is. He seems to have lost interest and moved on. More importantly, in my view, is the world-wide practices observed by most–WP:Reliable sources are clear as regards terminology and unit symbols. Greg L (talk) 05:29, 19 January 2009 (UTC)

I must admit the article is now rather pretty but just not correct.

I think to start with that we all agree that accelerometers measure g-force.

Where we seem to differ is that the article claims that an accelerometer/g-force measures acceleration due to gravity. But you can't actually measure an absolute acceleration due to gravity. It's physically impossible.

If they did measure that, then I would welcome an attempted explanation, which would explain why an accelerometer resting on the ground gives a positive-g reading away from the center of the Earth, and why an accelerometer that is falling downwards with gravity gives a zero reading. It's almost as if an accelerometer can't measure gravitational acceleration... "An accelerometer never senses gravitational acceleration." isn't it?

I mean if I have something resting on the surface of the Earth, it's not accelerating. No way, no how. There's an acceleration due to gravity pulling it down, and a force due to the surface of the Earth that is pushing it up, that would give an acceleration (from f=ma) that would cancel the gravitational one, causing it not to move/accelerate.

However, if the accelerometer can't read the gravitational acceleration (because all parts of it are subject to precisely the same acceleration due to gravity, giving no strains on the spring), then perhaps you might consider that that might explain things, since you would just be measuring the upward acceleration due to the upward force of the Earth, the downward one being invisible to the accelerometer.

You just can't get an absolute reading of gravity from an accelerometer. You can get a differential reading though, and that's done routinely in spacecraft; but the most correct and most general explanation is that accelerometers measure, not an absolute acceleration (which is impossible), but proper acceleration.- (User) Wolfkeeper (Talk) 12:08, 19 January 2009 (UTC)

Why is this such an important distinction to make? Do we have reliable sources for this? --John (talk) 18:40, 19 January 2009 (UTC)
Because I don't like the idea that we write something that's wrong. I'm sure we could find a really good reference that clears this up (maybe Wheeler's Gravitation covers it). I added a ref to the proper acceleration article that talked about rockets specifically that I found in google books.- (User) Wolfkeeper (Talk) 20:42, 19 January 2009 (UTC)
  1. Where we seem to differ is that the article claims that an accelerometer/g-force measures acceleration due to gravity. But you can't actually measure an absolute acceleration due to gravity. It's physically impossible.
  2. I mean if I have something resting on the surface of the Earth, it's not accelerating. No way, no how.
  3. However, if the accelerometer can't read the gravitational acceleration (because all parts of it are subject to precisely the same acceleration due to gravity, giving no strains on the spring)

You cited a post by a member of a model rocketry club, Dave Redell. In attempting to explain how accelerometers work, Dave used some unfortunate wording in describing how one would look at gravity, but he has correct observations of how accelerometers respond to being handled. Further, a fuller quotation than what you provided above is more illuminating. Dave wrote as follows:

        An accelerometer never senses gravitational acceleration.

or more specifically:

        An accelerometer is a device that senses deviation from freefall.

The last sentence of his explanation paints the complete picture of his message point in that model rocketry article. What Dave (“Lunar” member #332) went on to explain is the same thing this article now explains: when an accelerometer is resting on a desk, it is responding to the earth accelerating the device upwards through spacetime and will read 0 g only when dropped and is in freefall. That is why a “stationary”, single-axis accelerometer reads 1 g on earth and the signal goes to 0 g when you rotate it 90°.
If you don’t believe me, John, Rlsheehan, and Army1987, then please go get yourself a 3-axis accelerometer—even a single-axis accelerometer will do. For that matter, a digital kitchen (diet) scale with a stick of butter taped to its platen will do just fine. You will see for yourself that they behave precisely as described here in the first pargraph of Gravitational and inertial acceleration (as well as with the article you cited). You will be able to see for yourself that the only time there are no strains on the spring of a 3-axis accelerometer is when it is in freefall.
When you see this for yourself, then you will abandon your view of I mean if I have something resting on the surface of the Earth, it's not accelerating. No way, no how. You will see that an accelerometer resting on the surface of the earth is being accelerated—by earth, through spacetime—and the strain gauge within it is “pushing down its its chair” at 1 g. You will see that Einstein has been right all this time.
Alternatively, from the point of view of Newtonian physics, the strain gauge inside an accelerometer resting on earth is being bent down (strained) by gravity—just like the snow-laden branches of a pine tree. We can look at this tree just as Newton did: where gravity is a downward force that pushes down on the branches of a stationary tree. Or we look at the tree just as Einstein did: where earth’s surface and the tree are accelerating upwards through spacetime, causing the tree branches to strain towards the center of the earth. From either point of view, the effect is the same: accelerometers respond equally to gravity and inertial acceleration. That’s why they really and truly output a +1 g signal when stationary.
This is the best explanation I can give. If you, Wolfkeeper, still disagree, then someone else will have to give it a try. Greg L (talk)
No, your interpretation of what is happening is incorrect. If you mount the accelerometer in the nose of a rocket that is pulling 1.01g, and it's 1 foot above the surface of the Earth, then it reads 1.01g. But it will gradually gain speed, and leave the Earth. After it has left the Earth, if the rocket is providing the same acceleration throughout (in other words the thrust is kept a constant multiple of the mass of the rocket), the accelerometer will still read exactly 1.01g, but the gravity has varied. This shows again that accelerometers are completely insensitive to gravity. This is precisely the same argument that Einstein advanced when developing the Equivalence Principle, he specifically talked about rockets, and it is related to why the rocket guys get this right, but most others do not. Rocket guys have to understand this stuff precisely, because if they don't their rockets crash; see Pendulum rocket fallacy for example- it's highly related.
Your argument that it is gravity that causes the accelerometer's mass to strain downwards is indirectly true in a sense, in that particular case, because gravity causes the acceleration that is reacted against, but in general it is not.
Even in Newtonian mechanics it's wrong. Newton's gravity pulls everything with the same acceleration, there's no distortion of the spring due to gravity, it's all due to external forces acting on the accelerometer; gravity can't do that.
These guys also get it pretty right (I know it's not a reliable source, but I'm just trying to show I'm not out to lunch here).[1]- (User) Wolfkeeper (Talk) 20:42, 19 January 2009 (UTC)
Fascinating, and I think I see where you are coming from here. This reminds me of past debates at lift (force). I think we need to keep things (relatively) simple on this article, and we need to stay close to the sources. If there is a decent source which uses this explanation then it could go in the article. But let's not go into arcane levels of complexity here. It seems to me (my opinion, and I did university Physics, albeit quite a while ago), is that the acceleration experienced is always a vector sum of gravitational and acceleration forces, and that GR states that the two are indistinguishable in their effects. If we can agree on these principles (and I hope we can), and they can be referenced, then that is what the article should say. Make sense? --John (talk) 20:55, 19 January 2009 (UTC)
The acceleration experienced is different from the overall acceleration. That's because you can't feel gravity any more than accelerometers can. Einstein commented on that specifically, he told a story of a painter that fell off a roof- he felt weightless as he fell: Zero-g.- (User) Wolfkeeper (Talk) 22:11, 19 January 2009 (UTC)
  • Huhm. Ok, Wolfkeeper, I had to think for a moment there. If you de-tune the rocket’s engine 1%, then the accelerometer will read 1.00 g and the rocket goes nowhere. You will have to do a thought experiment now, and imagine moving the rocket by hand into space (or imagine the earth disappearing). In either case, the accelerometer continues to read 1 g. It’s just that gravitational acceleration (and no inertial acceleration), has been replaced by 1 g of inertial acceleration and zero g of gravitational acceleration. If you do the same for 1.01 g, as in your above post, then you will see the same effect: as the earth slowly recedes away and its gravity degreases, the rate of change in the velocity of the spaceship with respect to the earth (inertial acceleration) increases. This is certainly the behavior one would hope of a rocket, isn’t it(?): if you could instantly put a massive object like earth behind the rocket, then its rate of change in velocity with respect to the earth will instantly and substantially decrease due to all that gravity. Yet, an on-board accelerometer will remain absolutely unchanged. Keen, huh?

    As the article now says, “The connection between inertial and gravitational acceleration is profound. Albert Einstein showed in his 1916 paper on general theory of relativity that gravitational and inertial accelerations are identical and indistinguishable” and “Accelerometers respond equally to gravity and inertial acceleration.” Greg L (talk) 21:10, 19 January 2009 (UTC)

But if you have the vehicle hovering and you turn off the engine, the accelerometer immediately zeros, it doesn't remain at 1g. Again, the accelerometer isn't reading gravity, it's reading the acceleration due to the engine. And I completely agree that the gravity accelerates the vehicle downwards, but it does it in a way that doesn't affect g-force/accelerometers.- (User) Wolfkeeper (Talk) 21:31, 19 January 2009 (UTC)
  • But if you have the vehicle hovering and you turn off the engine, the accelerometer immediately zeros, it doesn't remain at 1g. That is a correct observation. Then you write Again, the accelerometer isn't reading gravity… But wrong conclusion. All I can suggest is that you carefully read the article and what John and I are writing here. Try also getting your hands on an iPhone; you can get g-reading software for it. Gotta go. Greg L (talk) 21:37, 19 January 2009 (UTC)
Your argument continues only to be nonsensical. Additionally, the article where you even mention this doesn't attempt to explain this, nor is it referenced.- (User) Wolfkeeper (Talk) 22:17, 19 January 2009 (UTC)
According to the equivalence principle article what Einstein actually said was:

"we [...] assume the complete physical equivalence of a gravitational field and a corresponding acceleration of the reference system." (Einstein 1907)

The 'reference system' in this context is a frame marking the origin and x,y,z axes and a clock attached to it. He's saying that gravitational fields are equivalent to that frame accelerating away from you- not you accelerating. That's also why accelerometers can't measure gravity- they're not moving according to Einstein- the frame of reference is.- (User) Wolfkeeper (Talk) 22:22, 19 January 2009 (UTC)
Not nonsensical, but perhaps over-complicated. Einstein is good enough for me on this article. Detailed discussion like this may belong more properly on acceleration or General Relativity. Again, we should stay close to what the sources say on this article. --John (talk) 22:59, 19 January 2009 (UTC)
Not if they're actually wrong, we need reliable sources. I'm not going to remove that flag if you insist on 'simplifying' by allowing incorrect facts into the article. That's not simplifying. Einstein also said that 'everything should be as simple as possible, and no simpler' ;-) I'm probably going to have to cruise over to the physics groups in the wiki and get them to look at this, they may have access to copies of Taylor's Gravitation.- (User) Wolfkeeper (Talk) 23:15, 19 January 2009 (UTC)

(outdent) Once again, Wolfkeeper, you accurately quote facts, and draw incorrect inferences. Exactly; when spacetime is accelerating towards the center of the earth, that utterly annoying propensity for humans to stubbornly stand on the earth—rather than fall down a deep hole in the ground so we can stay with spacetime—causes us to accelerate upwards relative to spacetime. The rush of spacetime at us acts like a downwards force of gravity and affects everything that resists its motion. Even massless light is affected (which shows just how profound Einstein’s thinking was).

I truly think you’ve got an image that is locked in your mind as tenaciously as Greenpeace protesters when chained to the fences of a furrier. You really should see for yourself how accelerometers work. I strongly suggest you round up anyone’s iPhone and install the free (iTunes link) G‑meter application. You might also install the $8.99 (iTunes link) gMeter. The latter is nice. I own it. It allows you to calculate 0–60 times and quarter-mile times, horsepower, etc. But the former one, G‑meter, simultaneously shows you three g read-out dial faces—one for each of the iPhone’s three axis. Please do as I suggest; you will see for yourself that accelerometers respond to gravity and inertial acceleration alike. This will save us all here a metric butt-load of time.

Please also directly respond to John’s 18:40, 19 January 2009 and 20:55, 19 January 2009 posts. He’s not asking too much. Greg L (talk) 23:17, 19 January 2009 (UTC)

Greg, I am sorry if you feel I am asking too much; or was there a missing "not" in there? Wolfkeeper, I like the idea of bringing in others' views. Per Einstein as you quote him, I agree we need to write this in the simplest possible way without being misleading. However, I'd point out that if it is just one editor dissenting, it isn't fair to hold the article hostage with a disputed tag. Some reliable sources for your interpretation of Einstein and acceleration would be good because it differs from mine (and Greg's). I'm sure that if we stay with the sources (and everything I've ever read supports the view that gravitation and acceleration are indistinguishable), we can come up with a compromise version that will satisfy everyone. --John (talk) 23:26, 19 January 2009 (UTC)
  • I note that our Accelerometer article begins with this line: An accelerometer is a device for measuring acceleration and gravity induced reaction forces. I checked Honeywell’s FAQ page but this issue is so profoundly basic that they don’t address it. Why are we discussing this any further? This is basic logic and physics. Further, anyone who has spent five minutes with an accelerometer can see that it responds to both inertial and gravitational acceleration and hasn’t a clue how to distinguish between the two. Wikipedia’s processes don’t effectively deal with circumstances like this. Allowing a single editor to junk up an article with {disputed} tags after a good handful of editors with strong physics and science backgrounds believe it is correct frankly seems absurd to me. Greg L (talk) 00:57, 20 January 2009 (UTC)
Well, I've asked for a couple of the relativity task force people that have edited recently to join us. I'm very sure they'll back me up, although I have no personal connection to them, and I'm hoping they can find a good reference as well.- (User) Wolfkeeper (Talk) 02:49, 20 January 2009 (UTC)
Ah, I've just found a pretty good reference "Introduction to Space Dynamics by William Tyrell Thomson (1986)". It's a fairly widely used textbook.
In the context of section 6.10's an inertial navigation system which has 3 gyros employed to produce a angularly stable platform containing 3 axis accelerometers, Section 6.11 begins:

"The accelerometers mounted on the vehicle measure only the non Gravitational force F_ng acting on the vehicle, and therefore one must add to it the gravitational force F_g in order to determine the total force which determines the acceleration of the vehicle, F_ng + F_g = m a_y... The gravitational acceleration a_g, which depends only on the position is computed and added to the output of the accelerometers to give the vehicle acceleration a_v"- (User) Wolfkeeper (Talk) 02:49, 20 January 2009 (UTC)

I think that's clear, as I stated, in general, accelerometers don't measure gravitational acceleration/force otherwise they wouldn't need to add on an estimate for it to the output from the accelerometers for this system.- (User) Wolfkeeper (Talk) 02:49, 20 January 2009 (UTC)
Also, the accelerometer article says it measures gravitational reaction forces (sic). It doesn't say it measures gravitational accelerations, only the reaction to them; but that doesn't always happen.- (User) Wolfkeeper (Talk) 03:08, 20 January 2009 (UTC)
Oh, wait. I see what's wrong. Has your 3-axis accelerometer got 3 axis gyros in it or something as well? They're probably using a computer to deal with this issue, essentially making an inertial navigation system. You can't make assumptions based on how such a system behaves and then try to extend it to accelerometers and g-force. And that's a problem as well, you're using OR to write the article.- (User) Wolfkeeper (Talk) 03:08, 20 January 2009 (UTC)
I agree with Wolfkeeper and think that the article should state that accelerometers do not measure gravity's acceleration. It's true that within General Relativity's framework gravity is an inertial force (like centrifugal force and coriolis force), but that seems to be an unnecessary level of complication for that article. I'm in favor of keeping it simple enough that the average reader will be able to understand. Dauto (talk) 03:17, 20 January 2009 (UTC)
So how does the accelerometer "know" not to measure gravity, if the two forces are indistinguishable? I agree we should try and avoid over-complicating this. I came into this understanding the area perfectly and am now rather confused. Have I been fundamentally misunderstanding one of the great works of the 20th century all this time? --John (talk) 05:28, 20 January 2009 (UTC)
It literally can be considered to be as simple as just that all parts of the accelerometer accelerate at the same rate under gravity at all times, giving zero reading, and then you add on the effect of other accelerations linearly. Einstein has a cute explanation for why the whole accelerometer go at the same acceleration due to gravity, but it comes to the same thing really as Newton for our purposes. The only time you get a reading from an accelerometer is if there's an external force acting on the outside casing of the accelerometer and then the test mass moves on its spring. That causes a relative displacement between the casing and the test mass and causes a readout.
When it's resting on the ground, the weight of the accelerometer itself pushes on the ground and the ground pushes back on it (equal and opposite force) and that external force from the ground causes an acceleration on the accelerometer that makes a strain on the spring and that triggers a readout. That's also why the acceleration is registered as upwards. If you know it's just resting on the ground and not moving you can use it to infer what the acceleration due to gravity because it's equal and opposite.- (User) Wolfkeeper (Talk) 06:08, 20 January 2009 (UTC)
  • Why are editors coming here to say that “accelerometers don’t measure gravity” when that’s the first thing they do out of the box? The only accelerometer that doesn’t respond to gravity is a broken one that can’t respond to anything. Why did I have to write that? This is absurd. Greg L (talk) 05:45, 20 January 2009 (UTC)
If you open it on a plane, it doesn't measure gravity, it measures the g-force of the plane. You actually can't measure gravity while you're on the plane, all you're measuring is the instantaneous g-force due to the aircraft's lift, but the accelerometer isn't broken.- (User) Wolfkeeper (Talk) 06:08, 20 January 2009 (UTC)
Thanks for explaining what you meant, Wolfkeeper. I think what you are saying could maybe be used to explain why an accelerometer at rest (or in constant motion) registers 1 g upwards, the opposite of the direction of gravitational force. I don't think there is as much difference as you think between your explanation and Greg's. I really don't think this article is the place to get into these complex explanations, because as we see here in talk, there are many different conceptual models and we would then have to have several different models in the article. It's easier to keep this simple I think. --John (talk) 06:19, 20 January 2009 (UTC)
  • Yes, Wolfkeeper. I completely agree where you wrote When it's resting on the ground, the weight of the accelerometer itself pushes on the ground and the ground pushes back on it (equal and opposite force) and that external force from the ground causes an acceleration on the accelerometer that makes a strain on the spring and that triggers a readout. That's also why the acceleration is registered as upwards. Greg L (talk) 06:26, 20 January 2009 (UTC)
Yes, but you can't go further than that because it's deceptive to say an accelerometer measures the gravitational acceleration, only that it can be used in some (albeit common) circumstances to determine that. I mean, you can use a barometer to measure the height of a tower by measuring the difference in air pressure, but that doesn't mean that it's a ruler, and there wouldn't be an article that read 'a barometer is a way of measuring air pressures and the heights of buildings' to have an only slightly forced example.- (User) Wolfkeeper (Talk) 07:00, 20 January 2009 (UTC)
Again though, we are going too far into the philosophy of science here. An accelerometer measures (can measure) gravitational force, which is indistinguishable from acceleration g. Your barometer example is a nice one; an aircraft's or a mountaineer's altimeter is exactly that, a specialized barometer calibrated for height, though not of buildings. I would agree that it isn't common for an accelerometer to be purposely used to measure gravitation as it is pretty invariant. But an accelerometer is surely always measuring the vector sum of gravity and acceleration, isn't it? Anyway, I feel we are in danger of arguing about angels dancing on pins here. Wolfkeeper, can you suggest what wording you would like to see? --John (talk) 07:28, 20 January 2009 (UTC)
No, it doesn't add the vector sum of gravity and anything, since gravity always reads zero. And yes, a barometer can be made to make, or operated as, one kind of altimeter, but they're not logically or necessarily precisely the same thing. Similarly an accelerometer is not necessarily a gravimeter but they can often be used to make one, not necessarily a good one. An accelerometer always measures g-force. G-force is always and only a property of the external forces, not the 'internal', inertial forces like gravity. Accelerometers read zero, for all inertial forces, every time.- (User) Wolfkeeper (Talk) 12:50, 20 January 2009 (UTC)
1) What is your source for this assertion? ("Accelerometers read zero, for all inertial forces, every time") It seems to contradict what you said above when you said "it's deceptive to say an accelerometer measures the gravitational acceleration, only that it can be used in some (albeit common) circumstances to determine that"
2) Why is it so important to put this in the article, as opposed to say in proper acceleration?
3) An aneroid barometer is logically and philosophically precisely the same thing as an altimeter. It is just calibrated differently.
4) I ask again, what wording would you like to see in the article, and with what sources? --John (talk) 14:25, 20 January 2009 (UTC)

1) it's widely known that gravitational accelerational in GR is an inertial acceleration. Inertial accelerations generate no g-force. Accelerometers measure g-force. 2) so: "don't bother me with facts, this is only the wikipedia!" is your attitude? 3) No, one way of constructing one type of altimeter is to calibrate a barometer for that purpose. It's one application of a barometer. If you don't believe me, by all means call for merge of the articles in the wikipedia. One application of a accelerometer is as a gravimeter. 4) just state that they measure non inertial accelerations, and point out that gravity is an inertial acceleration. - (User) Wolfkeeper (Talk) 14:54, 20 January 2009 (UTC)

  • I don’t understand why you keep on writing stuff like since gravity always reads zero. The only time an accelerometer reads zero in a gravitational field is when it is falling. If an accelerometer isn’t accelerating upwards or downwards with respect to earth’s surface, it reads 1 g upwards due to the force of gravity. You obviously don’t believe that what the article is now saying is correct. All I can say is that it is clear enough for the rest of us here. Greg L (talk) 16:53, 20 January 2009 (UTC)

Why is there a dispute at all?

Accelerometers measure proper acceleration. The proper acceleration of a body is its acceleration minus the gravitational field at the place it is. (For example, my proper acceleration right now is approximately 9.8 m/s upwards.) According to general relativity, both the acceleration of a body and the gravitational field on it depend on the frame of reference used, but their difference, the proper acceleration, doesn't. (For example, in the geocentric reference frame, I am stationary and subject to a downwards gravitational field of approximately 9.8 m/s; whereas in the free-falling frame, there is no gravitational field (by definition), but I'm accelerating upwards at approximately 9.8 m/s.)

Hence, the question about whether the "1 g" an accelerometer placed on a table is an acceleration or a gravitational field is meaningless unless we specify a frame of reference. As for me, I don't like using a frame of reference in which Italy and New Zealand are accelerating away from each other but their distance isn't changing, but YMMV, and according to general relativity there is no rationale for my dislike.

What part of this doesn't everybody agree with? -- Army1987 – Deeds, not words. 17:24, 20 January 2009 (UTC)

Sounds good to me, Army 1987. For the eleventy-seventh time though, this article should not be the place for a detailed discussion of general relativity or proper acceleration. They have their own articles which we can link to and summarize briefly. Neither should it deal with competing conceptual models of "why" an accelerometer reads a particular value. My physics is a little rusty but I do recall that according to GR there is no way to distinguish gravitational force from acceleration force, which makes it improbable that an accelerometer can accomplish this. I like the proposal you are making. --John (talk) 20:31, 20 January 2009 (UTC)
  • No Army. Accelerometers do not measure only “proper acceleration”. Where in the world did you get that idea? If they did, they wouldn’t read 1 g when stationary, would they?

    As for Hence, the question about whether the "1 g" an accelerometer placed on a table is an acceleration or a gravitational field is meaningless unless we specify a frame of reference. I am baffled why you would write such a thing. The answer is simple: relative to where the accelerometer is sitting. Since that is the common-sense way of looking at it, it need not be specified that we don’t mean relative to someone on the other side of the world.

    Single-axis accelerometers can’t distinguish between off-angle orientations, gravity, inertial accelerations, or any combination of the three. They know only acceleration along a vector and they respond equally to gravity and proper acceleration.

    There is simply no way of properly denying this unless you deny that a single-axis accelerometer reads 1 g when oriented vertically and 0 g oriented horizontally. And off course I am talking about when it is sitting in your frame of reference—not Mars or something weird. If you agree that this observation is true (1 g when vertical, 0 g when horizontal), then you must agree that accelerometers respond to gravity. This is just so basic. Greg L (talk) 06:29, 21 January 2009 (UTC)

Unfortunately, you are demonstrating only too well that it is not so basic. A single axis accelerometer does not read 1g downwards in that situation in a gravitational field. It is therefore not reading gravity which is a downward acceleration.- (User) Wolfkeeper (Talk) 06:38, 21 January 2009 (UTC)
So it's reading the reaction to gravity then? The upward acceleration? --John (talk) 08:10, 21 January 2009 (UTC)
  • Yes. It is reacting to the force of gravity. In the plain ol’ common-sense way where you open up the box and turn it on and it says “1 g” and you can’t make the signal go away unless you jump off a roof. The issue is about writing crap like “an accelerometer responds to non-gravity accelerations.” This is what Wolfkeeper keeps on trying to do. Greg L (talk) 08:34, 21 January 2009 (UTC)
You're precisely correct John, it's reading the reaction to gravity- reaction in the Newtonian action and reaction sense. It can indirectly read gravity, but never, ever directly. Note that Greg_L when apparently he agrees above, is actually not saying the same as you. He doesn't understand it.- (User) Wolfkeeper (Talk) 16:19, 21 January 2009 (UTC)
Pedant-point: this isn't an action-reaction pair, which is always a mutual interaction between two objects. (So, the reaction to the gravitational force on the device is the reactive gravitational force on the earth.) --Starwed (talk) 05:31, 22 January 2009 (UTC)
Proper acceleration is essentially (if you're not going near the speed of light) acceleration minus the gravitational field. So an accelerometer on a table, in the frame of reference of the table, is subject to zero acceleration and to a gravitational field of g downwards, so the proper acceleration is g upwards. BTW, acceleration relative to where the accelerometer is sitting is zero by definition (or did you mean something else by sitting?), so in that particular frame what the accelerometer measures is just the negative of the gravitational field (or the reaction to the gravity, as you prefer to call it). We're not disagreeing about the physical phenomena, just about how to call things. -- Army1987 – Deeds, not words. 14:51, 21 January 2009 (UTC)
The thing is, if we use the wrong words in the article it can actually give people the impression they understand it when they don't. Of course we can confuse people by being too technical also, and I don't want to do that either. Greg_L is probably a really good example of what happens when people just don't quite get it.- (User) Wolfkeeper (Talk) 16:19, 21 January 2009 (UTC)

I hate these mile-long threads. Surely this dispute is merely about terminology. I think everybody here agrees on how an accelerometer will behave in every actual physical situation. An accelerometer in freefall will read zero (or close to zero—it might notice tidal acceleration). An accelerometer sitting on a table will read 9.8 m/s² vertically. An accelerometer measures (approximately, in ordinary circumstances) the magnitude of its own deviation from gravitational freefall. So, does an accelerometer measure gravitational acceleration? On the one hand the readout uses gravitational freefall as a baseline, so no. On the other hand an accelerometer can be used to measure the so called local acceleration of gravity at any point on the Earth by setting it on a table and looking at the readout, so yes. These are not contradictory answers, they're answers to two different questions that happen to be expressible with the same sequence of English words. I think the solution is simple: don't use that sequence of English words, or any other sequence that can be interpreted both ways.

I think it's also worth pointing out that accelerometers don't really measure proper acceleration, or any externally meaningful quantity, as such. A froglevved mechanical accelerometer will read zero, even though an identical accelerometer sitting right next to it on a table, and clearly at relative rest, will read 9.8 m/s². That said, you could design a sophisticated accelerometer that could compensate for all other influences by using a variety of differently-responding materials (like those amazing mechanical clocks of John Harrison's) but you can never avoid the freefall reference (unless you're allowed to use an external reference body like GPS), so in that sense there is something deep going on here. -- BenRG (talk) 15:41, 21 January 2009 (UTC)

I really agree with you, I've hated every minute of this talk page. But it's not really terminology, we're not arguing colour or color, it's more backup (reversing) and backup (saving stuff). The frog levitation only applies to water, and a few other materials, if you made an accelerometer out of non diamagnetic material it would actually work perfectly, and wouldn't levitate. Hmm, come to think of it, if you were to accelerate the huge magnet, the frog would make a perfectly fine accelerometer already ;-)- (User) Wolfkeeper (Talk) 16:19, 21 January 2009 (UTC)
I believe that proper acceleration is completely the correct term though, even if you're going near the speed of light.- (User) Wolfkeeper (Talk) 16:19, 21 January 2009 (UTC)

Capitalizing “earth” (or not)

The Associated Press observes this practice world-wide regarding when to capitalize “earth”. This is from Grammar.ccc.comment.edu:

• [Do capitalize] Names of celestial bodies: Mars, Saturn, the Milky Way. Do not, howver, capitalize earth, moon, sun, except when those names appear in a context in which other (capitalized) celestial bodies are mentioned. "I like it here on earth," but "It is further from Earth to Mars than it is from Mercury to the Sun.

I always wrestled with this issue. I see now that the practice is nuanced. Greg L (talk) 19:31, 19 January 2009 (UTC)

Hmm. But our article on Earth consistently uses capital E, and it is featured. --John (talk) 20:13, 19 January 2009 (UTC)
  • <tone of being facetious>You mean, a Wikipedia article isn’t correct?!?</tone of being facetious>. Our own WP:MOS has it correct at #Celestial_bodies. It’s a grammar & punctuation thing; not everyone gets it right, but professional writers (who read their manuals of style) apparently do. Like I said, I just learned this myself. Greg L (talk) 20:50, 19 January 2009 (UTC)
  • The MOS has "capitalized ... in an astronomical context when referring to specific celestial bodies (our Solar System, Sun, Earth, and Moon): The Moon orbits the Earth, but Io is a moon of Jupiter." It therefore seems to me that Earth should be capitalized on this article. --John (talk) 20:59, 19 January 2009 (UTC)
  • The MOS gives these examples: The sun was peeking over the mountain top. And The Moon orbits the Earth, but Io is a moon of Jupiter. This is consistent with what the above referenced grammar site says: One writes: I like it here on earth and also writes "It is further from Earth to Mars than it is from Mercury to the Sun. Both sources are consistent with each other as they both conform to this rule of punctuation: Do not, howver, capitalize earth, moon, sun, except when those names appear in a context in which other (capitalized) celestial bodies are mentioned. The MOS version of this principle is as follows:

Sun, earth, and moon are not capitalized when used generally: The sun was peeking over the mountain top. They are proper nouns and capitalized when personified: Sol Invictus ("Unconquered Sun") was the Roman sun god. and in an astronomical context when referring to specific celestial bodies (our Solar System, Sun, Earth, and Moon): The Moon orbits the Earth, but Io is a moon of Jupiter.

The g-force article doesn’t speak of any capitalized references to other celestial bodies (the Earth, Moon, and Mars); it just talks about issues like “earth’s surface”. That’s why it’s properly done lowercase. Greg L (talk) 21:18, 19 January 2009 (UTC)
I don't have a problem with it being lower case on this article. Maybe a post at Talk:Earth is in order though? --John (talk) 23:00, 19 January 2009 (UTC)
Done. --John (talk) 23:36, 19 January 2009 (UTC)
  • Articles like Earth are frequented by a diverse lot. When my middle daughter was two years old, I was sawing some wood out back in the scrub brush portion of my property just off the lawn. I heard a noise of discontent from her and looked up. She had literally stirred up a hornets’ nest and they were swarming around her. Without thinking, I bolted towards her (wondering what the hell I was going to do when I got there). I had about 1.5 seconds to figure it out. Simple solution: I snared her arm like a steaming train catching a mail bag with its hook. I slowed down a bit when I reached the lawn, lowered her down and dragged her along on the grass as I slowly rolled her to scrape off any hornets. I stood her up and checked her out. She wasn’t crying. She didn’t even seem bothered. She didn’t have a sting on her.

    Odd; your suggestion that someone go to Talk:Earth with this observation about capitalization just now reminded me of that incident. Busy chopping wood here. I’m glad you did. Greg L (talk) 23:43, 19 January 2009 (UTC)

Duration

We need to clarify that g-force usually refers to a sustained acceleration. Shock and Impact are used to describe short term transient accelerations. All can be reported in multiples of g. This should go in the introductory section. Rlsheehan (talk) 02:47, 20 January 2009 (UTC)

Yes. I can bring sourcing to this. --John (talk) 03:14, 20 January 2009 (UTC)
  • Woa, woa. How’s that?? G-force is used primarily for sustained accelerations only if you are in the business of sustained accelerations. G-force is used all the time in vibration testing and other transient phenomenon—including shock. I designed industrial instrumentation and worked with UL and CSA. Shaker tables can be found big & small and operate a very wide range of frequencies depending upon the need. Nothing needs to be clarified regarding “sustained acceleration”; that notion is completely incorrect. There is no magic time period in industry over which g-force becomes less common. It might only seem that way to the general public that is primarily exposed to g-forces when watching Red Bull air races. If anything, I need to add a few line items to the table to give some lip service to vibrational g-forces. Greg L (talk) 05:35, 20 January 2009 (UTC)
  • Nearly all of the current article deals with sustained g-forces though. We would need (I think) to have a short section with a sourced discussion on transient versus sustained, with links out to shock (mechanics) and impact force where most of the material relating to short-acting g would best be placed. There's a bunch of stuff in that air crash book I could bring to this. --John (talk) 05:46, 20 January 2009 (UTC)
  • Agreed. Something about vibration testing would be good. Vibration testing is really common and needs fair play here. Underwriters Laboratories and IEC (I’ve redesigned American equipment and certified it to IEC for sales overseas) both embody standard industrial practices in their standards. So both would be good sources for gathering up background upon which to base that section. Do you want me to do this or do you want to, John? I wouldn’t mind. The only reason I write stuff is because I want to brush up on it and understand in serious depth.

    And, yes, something about jerk. If one didn’t know the term, you’d never find it. This is a great place to mention it. Greg L (talk) 05:54, 20 January 2009 (UTC)

  • (outdent). Do you want to tackle jerk, John? I can keep myself real busy giving a proper treatment to transient. It will also take several days for me to formulate what I want to write in my mind. Greg L (talk) 05:58, 20 January 2009 (UTC)

What “air crash book” are you referring to, John?

    • Beyond the Black Box: the Forensics of Airplane Crashes by George Bibel, John Hopkins University Press, 2008 (ISBN 0-8018-8631-7) A great book with an unusually detailed coverage of scientific principles. Loads of good stuff on metal fatigue and g-loads in crashes. --John (talk) 06:22, 20 January 2009 (UTC)
  • Books published by reputable publishing houses have an editorial process which makes them good reliable sources, at least in theory. I'm off to see what conceptual model Bibel uses to explain g. Maybe he can shed more light on the best way to explain it accessibly on this article. --John (talk) 07:34, 20 January 2009 (UTC)
Vibration, shock, jerk, and impact all have to do with the dynamic response of an item to input. Sustained loads are easier. I suggest that this article NOT address the short term loads and vibrations which involve dynamic response. Rlsheehan (talk) 14:58, 20 January 2009 (UTC)
  • I don’t have a rocket up my butt to address vibration in depth. As you point out, it can be complex. However, the role of any encyclopedia is to educate readers with minimal confusion and ensure they are well prepared to absorb information in their studies elsewhere on the subject. We also properly and honestly use terminology and symbology so readers can be conversant with others experienced in the art. We could probably use something about vibration here that informs readers of sinusoidal g-forces without getting bogged down in the minutia of reactive harmonics and other unnecessary complexities. In the mean time, I’m intent on directly and succinctly addressing the issue of “force” and how it relates to acceleration and gravity and the way Newton understood forces. Greg L (talk) 16:45, 20 January 2009 (UTC)

What, exactly, is this article talking about?

The problem with this article is that it does not define its terms, and seems to be talking about a blend of different topics under the same heading. In what we think of as the real world, there is less-than-complete consensus as to what “gravity” is, and what the symbol “g” means. This article does little to clarify it. It seems to be discussing a mixture of:

The last, Gf, seems to be closest to the main topic, "g-force". If you substitute m (for mass) for mg (for weight), the units work out to be those of acceleration, which seems meaningful.RockyMtnGuy (talk) 01:23, 21 January 2009 (UTC)


Editing against common sense

Wolfkeeper. With these edits, you keep on promoting the notion that accelerometers respond only to proper acceleration. This is absurd. To claim such a thing, you must provide an explanation for why an accelerometer will read 1 g when oriented upwards, and 0 g when oriented sideways. If you deny that this is true, then you deny an indisputable fact of the way accelerometers work. If you agree that this is the true behavior of accelerometers, you therefore are agreeing that they clearly respond to the acceleration of gravity.

It would also be nice if you didn’t ruin articles with links that look like this. Greg L (talk) 06:10, 21 January 2009 (UTC)

Sorry, but I removed uncited material, and inserted cited material that is referenceable to a WP:RELIABLE source. You can read the text at: [2]. Your claims of 'common sense' are OR.- (User) Wolfkeeper (Talk) 06:20, 21 January 2009 (UTC)
  • What is wrong with you? You just cited a reference that proves my point and disproves yours. Didn’t you read it? It is talking about a nuance of its output signal. It reads as follows:

[An accelerometer] cannot distinguish between inertial acceleration [a-bar] and gravitational acceleration

Stop pushing this notion of yours. It is absurd and you just found a reference that speaks of accelerometers having no ability whatsoever to distinguish between proper (inertial) accleration and gravitational acceleration, which is what I’ve been saying all along (and what the article says). Greg L (talk) 06:37, 21 January 2009 (UTC)
You've taken it out of context and misread it as well.- (User) Wolfkeeper (Talk) 06:45, 21 January 2009 (UTC)
Look at the equation.- (User) Wolfkeeper (Talk) 06:45, 21 January 2009 (UTC)
  • No. The statement:

[An accelerometer] cannot distinguish between inertial acceleration [a-bar] and gravitational acceleration

It goes on to talk about how to separate the inertial acceleration from the mix of inertial and gravitational signals by subtracting out the gravitational signal from known navigator position and attitude. It reads as follows

Hence, the accelerometer cannot by itself provide the determination of inertial acceleration required for navigational purposes. Therefore, to get the total acceleration vector, accelerometer triad outputs must be added to the gravitational acceleration vector, which is calculated from the known navigator position and attitude;

This is beyond absurd. You’ve never seen an accelerometer before in your life, have you? Greg L (talk) 06:51, 21 January 2009 (UTC)
Yeah, I have actually, thanks. Oh and I checked with somebody who is consulting for an aircraft manufacturer to build a gradient gravimeter, and he absolutely and completely confirmed it.- (User) Wolfkeeper (Talk) 06:54, 21 January 2009 (UTC)
The equation given in the book is f = a - g
where f,a,g are vectors. g is the acceleration due to gravity and points downwards (0,0,-9.8). a is the acceleration in the inertial frame, say (0,0,0), and f is the acceleration that the accelerometer gives f= (0,0,9.8)- (User) Wolfkeeper (Talk) 06:54, 21 January 2009 (UTC)
As it states, f does not read the acceleration due to gravity and you have to add it on to get the right answer. In the case of your accelerometer just sitting there on the table, you have to calculate and add on the downwards g to get the correct motion (in that case it's not moving.)- (User) Wolfkeeper (Talk) 06:59, 21 January 2009 (UTC)
  • Apparently, every single bit of “[An accelerometer] cannot distinguish between inertial acceleration [a-bar] and gravitational acceleration” confuses you. Greg L (talk) 07:15, 21 January 2009 (UTC)
The only thing I don't understand is why you think it supports your position, when it clearly doesn't. I also don't understand why you though you could in any conscience cut it out of the full sentence which reads:

Unfortunately, an accelerometer measures the nongravitational specific force, output = f = a - g, and cannot distinguish between the inertial acceleration, a and the gravitational acceleration, g. Hence the accelerometer cannot by itself provide the determination of inertial acceleration required for navigational purposes. Therefore to get the total acceleration vector, accelerometer triad outputs must be added to the gravitational acceleration vector, which is calculated from the known navigator position and altitude a = output + g(r). Specific force is simply another word for acceleration, (specific means per unit mass, specific force is force per unit mass; f/m = a; and it actually says that specific force is acceleration a little further down). a is the acceleration relative to the rest mass of the Earth, and is what they're trying to calculate (it's an inertial navigator). It's very, very clear. And you're removing it purely because of your own OR. I mean does it ever occur to you that you're actually wrong. There was also the other two references I found: [3]"From a classical physics perspective, an accelerometer is a device that measures the acceleration due to all forces acting on the accelerometer case except gravity." and there was the astrodynamics text book I mentioned above. - (User) Wolfkeeper (Talk) 07:38, 21 January 2009 (UTC) I also today got an email from Henry Spencer who is a professional aerospace engineer, noted for being absolutely precise about every goddamn thing. There was also half a dozen others that said exactly the same thing in various phraseologies. All you've got is a bunch of OR, a rather severe misunderstanding of the physics involved and a revert button, and claims that playing with an accelerometer for a bit makes you an expert. Um. No.- (User) Wolfkeeper (Talk) 07:38, 21 January 2009 (UTC)

  • Look up and what the letter writer wrote. How many times does it say it reads 1G when sitting on the ground? Lots, right? You fail to see what the above letter is saying. The same to for what the article says. The letter writer you quoted above isn’t cluing in on the fact that you question whether accelerometers respond to gravity. They are assuming you know this and are asking about sensing the *gravity itself*. They’re trying to answer an abstract issue of detecting the the gravitational field. And the letter writer is telling you that an accelerometer reads 1 g when it is stationary on the ground because the earth holds it up in gravity. The article makes all this clear to: it is the earth pushing the accelerometer upwards through spacetime that makes it read 1 g on the ground. The whole point of the book you are citing is talking about how difficult it is to tease out small inertial signals when there is the big gravitational signal. The key point you should have picked up on here is where it says “An accelerometer] cannot distinguish between inertial acceleration [a-bar] and gravitational acceleration”. This is almost verbatim to what the our article says. Accelerometers respond to both gravity and inertial accelerations and can’t tell the difference between the two. This is precisely as the book says. It is simply absurd to claim that accelerometers respond only to inertial acceleration and not to gravity. The entire above letter makes it clear that they respond very much to gravity when they are stationary. Greg L (talk) 08:11, 21 January 2009 (UTC)
    • Nice post. I wish you two could depersonalize this a bit; you are not really that far apart from each other but you have become entrenched. I know you've been in conflict for a while now, but if you could step away from the personalities for a moment. Language like "All you've got is a bunch of OR, a rather severe misunderstanding of the physics involved and a revert button" and "You’ve never seen an accelerometer before in your life, have you?" isn't helping us move forwards here. Per WP:EXPERT, our own personal qualifications and experience don't count for diddly here, and we must instead go to the sources. Blog postings aren't the best sources here, and we should stick to a brief description per the sources, and try to avoid getting too personal here. With all respect, --John (talk) 08:24, 21 January 2009 (UTC)
      • Look, basically, it won't take many seconds of anyone fairly educated in physics looking at the source I added and the article is going right back to my version. What I wrote isn't in any way debateable. I've asked User:SBHarris to look at it, he's a doctor but he knows a lot about physics and he's bloody smart. If he says I'm wrong, I'm wrong. I'm going to ask a few others as well.- (User) Wolfkeeper (Talk) 08:40, 21 January 2009 (UTC)
      • I've also called in User:Georgewilliamherbert he's an admin as well as an aerospace engineer, but I've asked him to comment in a non admin capacity.- (User) Wolfkeeper (Talk) 08:52, 21 January 2009 (UTC)
  • Fine John, let’s go with the sources. He found one that says as follows:

[An accelerometer] cannot distinguish between inertial acceleration [a-bar] and gravitational acceleration.

From this he writes “an accelerometer which gives a measurement of an object's non-gravitational acceleration.” This is false. Further, even his above-quoted letter is chock full of “1G” due to the influence of gravity. The article properly says that “Accelerometers respond equally to gravity and inertial acceleration.” It is time to stop letting Wolfkeeper hold this article hostage with a {disputed} tag. Greg L (talk) 08:31, 21 January 2009 (UTC)

Let's clarify what exactly the dispute is about

Simple issue: is the following paragraph correct?

An accelerometer measures acceleration in one or more axis. It responds to both gravitational and inertial acceleration. If you orient a stationary, single-axis accelerometer so its measuring axis is horizontal, its output will show zero gee. Yet, if you rotate the accelerometer 90° so its axis points upwards, it will read +1 g upwards even though still stationary. If you mount the accelerometer in an automobile with its axis aligned forward with the vehicle’s direction of travel, and drive down the road at a constant speed, it will read 0 g. Yet, if you hit the brakes, it will read about −0.9 g. Accelerometers respond equally to gravity and inertial acceleration.

You, Wolfkeeper, found a book that says as follows:

[An accelerometer] cannot distinguish between inertial acceleration [a-bar] and gravitational acceleration.

Let’s have you, Wolfkeeper, start off with real clear explanation of exactly what you think is wrong with the above paragraph. Greg L (talk) 08:51, 21 January 2009 (UTC)

I already have repeatedly, I'm not commenting further, it's for others to do so.- (User) Wolfkeeper (Talk) 08:54, 21 January 2009 (UTC)
  • That’s fine. Here is the edit you’ve been repeatedly making. This is the crux of the issue: You’ve been saying here that accelerometers measure only inertial acceleration, not gravity-based acceleration. The question is: are you right about that? Greg L (talk) 09:20, 21 January 2009 (UTC)
  • For the record, he is. Whenever you THINK you're measuring gravitational acceleration (strength of g-field) with an accelerometer, all you're really doing is measuring the force (ordinary mechanical force) needed to keep the accelerometer motionless in that field. But you can turn THAT mechanical force off and on, and the accelerometer will just measure it (or not) and though it all, it still won't measure the g-field, which could be anything (any value) though any of it. Is that simple enough for everybody? SBHarris 13:39, 21 January 2009 (UTC)
  • Hmm: Check this. This is from MEMSIC.com. They make sensors for “consumer, automotive, medical or industrial product applications”. It is titled ACCELEROMETER PRIMER. And it begins with this:

Accelerometers are used to convert an acceleration from gravity or motion into an electrical signal.

(My emphasis). I really do hope this is clear enough. Greg L (talk) 09:08, 21 January 2009 (UTC)


Ahhhhhh (thud) - the sound of George's head hitting the keyboard again and again
Look... This was slightly lame when the discussion started a couple of days ago on the arocket mailing list, and isn't helping here. The terminology is not sufficiently precisely consistent between different branches of physics, and between different branches of engineering, and between physics and engineering. It's possible to use reliable sources to pedantically prove that 1 = -1 quite easily on this topic, using sources that disagree on precise definitions or are internally not rigorously consistent. That exercise is trivial and not helpful.
It is not proper to take articles off into long pedantic fights. Wikipedia articles are part of a general encyclopedia. They have to explain things so that normal people have some chance of understanding. They most particularly must not cause people to come away from the article confused or with an incorrect understanding.
It is particularly important for editors to be aware of this on topics which are, even for experts, badly or imprecisely defined, and particularly again for topics popular or widespread enough that a lot of people are likely to come read it. We have in our hands the power to go off in any arbitrary direction on any set of reliable source information and in doing so commit evil upon the intellectual development of the world writ large.
In clear terms: There is no difference between a gravitational and an inertial acceleration, per Einstein. We're sitting in a 1-G static gravitational field. There's a force exerted by the ground, floor, and chair which counter the gravitational acceleration force downwards.
If we try and get too pedantic on this it will nuke the conversation right out of people's ability to follow, which is a sufficiently fundamental error that I call foul and ask for time out.
[ I particularly object that this came up on a topic where I persistently for no clear reason want to misspell accelleration with an extra l, which my brain keeps insisting should be there despite clear evidence to the contrarry... ]
Georgewilliamherbert (talk) 09:18, 21 January 2009 (UTC)
  • Thank you, Georgewilliamherbert. Where you write, In clear terms: There is no difference between a gravitational and an inertial acceleration, per Einstein. We're sitting in a 1-G static gravitational field, that’s what I’ve said over and over and over here on this talk page. That’s what the article says. But as the principle applies to accelerometers, Wolfkeeper has been saying it is WP:OR to say that “Accelerometers respond equally to gravity and inertial acceleration.” Quite frustrating. I’m sorry you got dragged into this. Greg L (talk) 09:24, 21 January 2009 (UTC)
I'm really sorry to drag you in George, and I agree with not making it too difficult and everything, but I don't think I was, and we're talking about g-force, which is always the non gravity bit. You can't in any case ever feel a gravitational acceleration, and the reverted article starts with stating you can. I have no problem at all with writing this as clearly as humanly possible, but I don't see that we can't write things that are simply wrong.- (User) Wolfkeeper (Talk) 09:31, 21 January 2009 (UTC)
I hope you don't think I picked and chose my sources either George, I just looked for the best quality source I could find, and even this one is hard to read; if I genuinely had found any real discrepancies I would never have gone with this source, I would have added multiple references to different POVs in some way.- (User) Wolfkeeper (Talk) 09:36, 21 January 2009 (UTC)
  • What is your mental block here? George was clear. And yet, here you are writing You can't in any case ever feel a gravitational acceleration. It doesn’t matter whether or not you can feel it; I can. It makes my apple fall to the ground and the food stay on the bottom of my stomach. It makes snow-laden branches on trees sag down. When accelerometers are sitting around on earth, like pretty much anything else you see, including yourself, they do feel gravity—exactly like tree branches. And accelerometers send out a 1 g signal testifying to the fact that they’re being exposed to the force of gravity; they all do. Get that into your head please. Accelerometers detect gravity. They send a signal saying they do. When engineers (which I happen to be) do important work with accelerometers in inertial navigation units, they have to figure out clever ways to get that damned gravity signal out of the way. If you turn a single-axis accelerometer sideways, the gravity signal disappears. Moreover, the simple fact is that the above paragraph is true. Accelerometers respond to both gravitational and inertial acceleration and cannot differentiate between the two, just as George said. NOW STOP EDITWARRING please. Greg L (talk) 09:46, 21 January 2009 (UTC)
  • You're not feeling gravity, either. All you feel is mechanical pressure from an electromagnetic force (there are four basic forces in nature: this is not gravitational, strong nuclear, or weak nuclear. Thus, it's electromagnetic). Food stays on the botton of your stomach because your stomach pushes up on it. It would do the same if you were in a rocket, with no gravitational field whatever. And your accelerometer would not know the difference, either. SBHarris 13:42, 21 January 2009 (UTC)
  • No, SBHarris, gravity is not an “electromagnetic force” by any stretch of the imagination. Please cite where you got that bit of info. Nor is the effect of gravity a “pressure” in any fashion. Sorry. Greg L (talk) 18:43, 21 January 2009 (UTC)
  • Gravity is not electromagnetic, true, but that is okay, because you're not feeling gravity. You never feel gravity. You can only feel electromagnetic stress (mechanical pushes). The same with an accelerometer. When nothing but gravity acts on you, you feel nothing. You're floating at what feels exactly like being under no-force. Accelerometer reads "zero."

    Stand on a table out in space, being boosted underneath by a rocket at 1 gee. You're not feeling gravity (there isn't any), you're feeling mechanical stress on you body transmitted though your feet. That's all your body's resistance to changing v; or inertial force. It's electromagnetic. Ramp it up and it will squash you. Hover that platform over the Earth, and you still do not feel gravity. All you feel is 1-gee inertial force. Turn the rocket off and land and you still do not feel gravity; again just the inertial force of the terrain pushing up on your shoes, keeping your body from following its natural geodesic path though spacetime, by applying an EM force to your feet (which transmits up to the food in your stomach). In all cases, the accelerometer on your wrist feels the electromagnetic inertial force too, and always misses the gravitational component completely. SBHarris 19:47, 21 January 2009 (UTC)

Your behaviour in the wikipedia is inexcusable, in every case, in every way.- (User) Wolfkeeper (Talk) 10:38, 21 January 2009 (UTC)
It's got the italic 'g' everywhere. You really are a piece of work.- (User) Wolfkeeper (Talk) 10:38, 21 January 2009 (UTC)
  • On the (perhaps shaky) assumption that you aren’t going to claim that MEMSIC doesn’t know what they are doing, I’ve removed the {disputed} tag. If you still have stuff you want to dispute, discuss it here please. I’m not saying I’ll always be right here. But the simple stuff, is, well… simple stuff. I’ve got 15 patents on some of it has been on technical issues that were a hundred-thousand times harder to figure out that this. Discuss please. Keep an open mind. Get rid of old notions. And if I mess up, I guarantee you that just one—at most two—logical posts is all it will take to get me to admit I was wrong and turn on a damn dime. Don’t mess up articles with tags; there is no need for it. Greg L (talk) 10:09, 21 January 2009 (UTC)
Well JRSpriggs says that I am [4] correct, twice: [5].- (User) Wolfkeeper (Talk) 11:03, 21 January 2009 (UTC)

This argument is well and good. Why does it matter, at all, to this article, exactly, except that you all want so desperately to adress this argument in the article? Just excise all mention of gravity. Hipocrite (talk) 19:50, 21 January 2009 (UTC)

New Section on what accelerometers measure

I hope you all don't mind me starting a new section. I've read the old ones, plus Wolfkeeper's cited book. He asked for my opinion, and here it is.

The text in question merely looks at the inertial acceleration "a" of an object in a g field. This "a" is what you're interested in, in inertial navigation, because the double time integral of "a" gives you the distance you've moved relative to the surface of the Earth, ie, the distance traveled in your 1-gee accelerated frame. This "a" is the sum of two things: one is g, the acceleration of gravity, if there is any. The other is what the book calls "f", which is the total force on the object, as you ride along with the object in it's frame (which may be accelerated, or not). This total acceleration of the object, caused by unblanced forces in the object's accelerated frame, is what gives you the total acceleration you feel in your stomach as you ride with the object, and it is what the accelerometer (mounted in the dashboard of the object) measures. It's quite correct that accelerometers cannot measure the acceleration of gravity, and neither can you feel it. All you feel is this residual total of the forces on the object in the freebody diagram, which if they are unbalanced, lead to an acceleration. None of them are gravitational: you can't feel gravitational force. What you feel is the residual of OTHER forces, if they aren't balanced. If they ARE balanced, you feel nothing--you're in an inertial frame, zero-gee, and everything floats. If they are NOT balanced, you feel that because the wall or the floor or something hits you in the feet or the rear.

Now, the text makes the matter difficult, because it simply calls all these non-gravitational forces/accelerations, "f", non-gravitational, (indeed they ARE that) and it notes that this is what accelerometers measure. Many people "know" this isn't true. If you're just sitting on the launch pad, doesn't your accelerometer register 1-gee, and you feel that? Yes, it measures one-gee, but it's NOT from the Earth's g field. What it measures is the 1-g electromagnetic acceleration of the Earth hitting against the base of the rocket, transmitted up to the floor, your shoes, and into the mechanism of the accelerometer itself. Not gravity. It's measuring an electromagnetic acceleration and EM push.

You don't believe me? Okay, if you want that same 1-gee feeling out in flat space, your rocket has to be blasting at 1-gee, and then it's clear that what your accelerometer is feeling is an electromagetic push from the nozzle on up through the rocket, and it all comes from reaction from the push off of rocket gasses. Same thing in both cases. So accelerometers on rockets blasting or hovering are not feeling gravity, and neither are they when the rocket is turned off.

Now, back to the rocket on the ground. Dig a hole under it, fire the rocket, and let it hover at 1-gee. The accelerometer still feels 1-gee and so does your gut, but now you may attribute it to gravity, same as if the rocket was sitting on the ground. But as we see, it's not. You still can't feel the gravity field. You're still feeling the EM kickoff.

Now, turn the rocket jet off, and fall into the hole which has been dug under the gantry. As you fall, you're in Einstein's elevator with the snapped cable. You don't feel the g-field. As far as you're concerned, you could be out in flat space.

Now, where does the inertial gravitational correction come in? Since accelerometers can never feel the g field, you have to know what the value of it is, and "add it" by hand to the acceleration/force the accelerometer DOES feel (which is zero), so you can tell if you're moving. Out in space, if you're blasting along at one-gee, you note that there's no g-field to subtract out, so your 1-gee is entirely "a", which means it's all inertial acceleration and you're going somewhere fast, at 1-gee inertial acceleration. But your acceleromater by itself can't tell you that, without the "hand" correction. All the accelerometer tells you is what your stomach does, and without looking through the window you can't tell if you're blasting through space at 1-gee, or hovering at one-gee on the launch pad. Do the calculation there and you have to subtract a g field of 1-gee from the 1-gee your accelerometer measures, and when you do, you find out that your "a" is zero: you're going nowhere. That's true whether your rocket is blasting to hover you over a hole, or your rocket is turned off and you're just sitting on the gantry. The accelerometer reads the same in both cases (we ignore the vibration-- maybe the simulators are shaking the thing-- how would you know?).

Finally, we look at the case where you're falling down the hole. Your accelerometer reads zero and so does your gut, but adding in the 1-gee grav field by hand, you find that your inertial "a" is 1-gee, and you're going down a hole, fast. Out in space, you don't have to add the g-field, so your accelerometer reading of zero means an inertial acceleration of zero, and you're again going nowhere.

Okay, now again note that the book says flatly that accelerometers measure a net "non gravitational" set of forces: lift, thrust, and drag. Of course the balance of thust and drag give you your horizontal acceleration, but note that the lift-force has to be large enough to balance out the gravitational force, or (if you like) the weight. The accelerometer measures the sum of them all. Now, again you can confuse yourself and consider this to be the horizontal accelerations ADDED to a 1-g downward gravitational acceleration, but that's again wrongheaded. The accelerometer again cannot see the downward gravitational component. All it measures is the airplane's resistance to it, which is electromagnetic. You have to take the downward acceleration and subtract out the gravitational field which you know to be out there, but can't measure. If you're a little light, that may mean that you're falling, or perhaps that you're flying really high and the gravitation outside is a little bit smaller than 1-gee. In either case, the inertial result (you're falling or you're in level flight) cannot be told by the accelerometer. All it measures is a total EM force, and you must know the gravitational acceleration to correct that, in order to get the inertial up/down acceleration.

Okay, I'm quitting for now. Is that clear? You should know enough to be able to change the article now so that everybody agrees. SBHarris 11:38, 21 January 2009 (UTC)

Thanks Steven, we won't be able to refer to it in terms of EM forces in the article, but what you wrote definitely is right. Everything in the world is held together with EM forces, and the paths of the forces and accelerations you've described is precisely right.- (User) Wolfkeeper (Talk) 14:07, 21 January 2009 (UTC)
There is an existing article on accelerometers. We do not have to have a redundant discussion here. Rlsheehan (talk) 14:49, 21 January 2009 (UTC)
Except that the same argument is now going on, there. Somebody has added the patently untrue fact that accelerometers measure all accelerations on them. Wrong. A falling accelerometer is OBVIOUSLY being accelerated, but it will read ZERO. Get your heads around this, and you'll be fine. SBHarris 20:10, 21 January 2009 (UTC)
  • I agree with the notion that the article could likely benefit from the inclusion of a section that connects the dots between acceleration and the notion of “force.” I’ve got a draft version of something that explains the Newtonian way of looking at g-forces (which is the intuitive, natural way humans think of gravity the forces we feel as a consequence). Right now, it mentions all three of Newton’s three laws of motion. Though quite complete, it’s simply too long right now and needs a massive trim so John won’t have so much work to do. ;·) This morning, I had an idea on how to streamline it down to a tight nugget; in fact, there might not be enough room for the picture I had in mind for it. I think you’ll all be happy with it. Greg L (talk) 18:40, 21 January 2009 (UTC)
  • We've been sidetracked by the idea that gravity is a "force" when in some ways it does not act like a force. A force by experience is usually something you feel, and you can't feel gravity (put you in a spaceship and you can't tell if there's a gravitational body near or not). Forces cause mechanical stresses, and gradient-free gravity (no tides) doesn't. All this is part of what led Einstein to the conclusion that gravity isn't a force at all. In our case, we merely need an intuitive reason why accelerometers don't "see" gravity. Proof: falling accelerometers, which are being accelerated, don't "know" it, and read zero. Accelerometers only read/feel "inertial forces", which are those that cause mechanical stress. That's it. But we need to lose the BAD idea that accelerometers measure all sorts and all kinds of accelerations. They don't. They only measure the one kind-- the nongravitational kind that produces mechanical stress. Example: An accelerometer sitting on the ground is not being acclerated (really, it isn't-- do you see it move? No, you don't), but it reads "1-gee, upward." An acceleration of 1 gee upward should cause it to move upward, but it's not moving upward. Why? It's reading wrong, and missing gravity. All it's reading is the force from the ground which causes the acceleration which counteracts gravity. But the gravity is not seen. SBHarris 20:28, 21 January 2009 (UTC)
  • Please stop continuing your debate here. This is a page for improving an encyclopedia article on g-force in a General Purpose encyclopedia, not for debating and discussing what accelerometers do and do not measure. Hipocrite (talk) 20:31, 21 January 2009 (UTC)
  • Right. And you made the same argument on the accelerometer page, when you said we should excise all mention of g-forces THERE. You can't have it both ways. Someplace, somebody needs to say what accelerometers do, and what they measure, and whether g-force is included. Where would you like? SBHarris 20:37, 21 January 2009 (UTC)
  • That is a page for discussing the accelerometer article. Feel free to discuss the accelerometer article there. Do not engage in debates unrelated to the article. Hipocrite (talk) 20:44, 21 January 2009 (UTC)
  • Well, you should have used that page before making changes to the accelerometer article which caused it to be in error, then, shouldn't you? I agree that THIS article is not about accelerometers, but when it mentions them and their function (*I* didn't put that in!), it should do so correctly. I've fixed it (it was almost correct, as it was). If you don't want to accurately portray accelerometers in this article at all, feel free to delete mention of them completely! But if you must have them, please discuss what they measure correctly. Thanks. SBHarris 20:55, 21 January 2009 (UTC)

Noteable car crash

The "noteable" section needs care. The example of the race car driver surviving 180 gs is amazing but questionable. The citation is not from an engineering perspective. Was the car instrumented (how?) or was this just a calculated value? Was this a "spike" in the shock pulse or was this peak observed for a while (how long)? What was the actual g-level on the driver - - not the car? This is an example of a short term shock which was not properly described and, in my opinion, not very useful. Rlsheehan (talk) 17:07, 21 January 2009 (UTC)

  • You raise a very valid point, Rlsheehan. They appear to have calculated it from 108mph to zero in just over half a meter. That lessens the accuracy some and the error is frequently on the plus-side because elastic deflections in the race car’s body spring back—all the race team took into account was the permanent crunching remaining.

    I suggest we all just keep our eyes open for a better example. About a year ago, I saw a lengthy educational-channel show (I think it was the Discovery Channel ) about crash physics. They showed an Indy driver who had gone backwards into a wall and he had an accelerometer under his seat. It measured 100 g. Apparently now, many Indy car and Formula 1 cars now have accelerometers so engineers can improve crash safety.

    I don’ think I’ll be able to cite that Indy-car crash since it’s all so vague—I just clearly remember that g-force value. I suggest we all just keep our eyes peeled for an accelerometer-based crash citation in the Indy and Formula 1-car world (where all the good stuff is coming out of on this issue). And in the mean time, I see no harm in leaving the existing citation up, “108mph to zero in just over half a meter” isn’t exactly non-science; it’s a very accurate measure of the acceleration the parts of the race car nearer to the crunch zone. However, it is way, way too high a precision for stating what the driver experienced. That’s one reason I rounded it off in the chart. Greg L (talk) 18:32, 21 January 2009 (UTC)


  • Here: NHTSA: [6], is a report by…

    Augustus "Chip" Chidester, National Highway Traffic Safety Administration
    John Hinch, National Highway Traffic Safety Administration
    Thomas C. Mercer, General Motors Corporation
    Keith S. Schultz, General Motors Corporation

    It says as follows: In 1992, GM installed sophisticated crash-data recorders on 70 Indy race cars. While impractical for high volume production, these recorders provided new information on human body tolerance to impact that can help improve both passenger vehicle occupant and race car driver safety. As an example, the data demonstrated that well restrained healthy, male race car drivers survive impacts involving a velocity change of more than 60 mph and producing more than 100 g's of vehicle deceleration. Such information will be helpful to biomechanics experts refining their understanding of human injury potential.

    I’ll keep on looking, but I suspect actual measured values are going to be closer to 100 g. And, like the report’s authors did (although those would be just more examples of *stupid ignorant American* practices), but also as NASA, ESA, JPL, Honeywell, Sensr, and others), it’s lowercase, roman g, (although roman uppercase G is quite common too). Greg L (talk) 19:01, 21 January 2009 (UTC)


  • This report by the Vehicle Safety Research Centre Loughborough University speaks clearly to the subject of brief (millisecond) accelerations in crashes. I can tell you from personal experience, that millisecond accelerations are extraordinarily common in industry. As we had discussed before, this article needs to touch briefly on that issue to ensure our readers understand that accelerations are not just a measure of persistent gees. Oh, BTW, this study by a British University also properly uses (not surprisingly), lowercase roman g. Did I mention that in an earlier engineering job, I designed electronic equipment and had to design rubber bumpers to mitigate g-forces on the equipment in drop tests? Greg L (talk) 19:10, 21 January 2009 (UTC)


  • Getting closer. Probably the History Channel. Here is the Delphi Accident Data Recorder 3 (ADR3) MS0148. It is the accelerometer mounted in Indy cars. BTW, lowercase roman g.


  • Done. Revised to “>100 g” and properly cited. The old B.S. about 179.8 g has been removed from the body text. Bad science. Thank you very much, Rlsheehan, for pointing this out. You have a keen eye for stuff that doesn’t pass a proper, scientific *grin test.* Greg L (talk) 19:43, 21 January 2009 (UTC)

Don't screw up article

Whoever is screwing this article up, needs to stop. It was completely correct before. I use accelerometers in daily life. This article is completely correct--at least now that I've restored the text someone removed. Deegee375 (talk) 22:08, 21 January 2009 (UTC)

As the screwer uper, I don't dispute the information is correct. Some other correct information is that Barak Obama is the 44th president of the US. Neither the facts about accelerometers or Barak Obama belong in this article. Consider going back to my version, please. Hipocrite (talk) 22:15, 21 January 2009 (UTC)
  • It’s oh-so easy to delete stuff. John and I spent hour and hours and hours collaborating on getting that text succinct, encyclopedic in tone, and correct. It speaks straight to the issue of explaining how one is at 1 g sitting in a plane and how if you pull back on the stick, the g reading goes up from there. It is an important concept in understanding the nature of accelerations. Thanks Deegee. Greg L (talk) 23:51, 21 January 2009 (UTC)
But we have textbook references that say it's not correct; and you arbitrarily deleted them, in violation of WP:NPOV and giving undue weight. You can't do that under the core values and I don't care how nice the 'encyclopedic tone' is afterwards.- (User) Wolfkeeper (Talk) 00:02, 22 January 2009 (UTC)
  • No, you are simply refusing to get the point and think you WP:OWN this article. Nor can you be selective with which parts of citations you want to use. You want to use MEMSIC, which is a world-wide manufacturer of accelerometers to buttress your argument that the unit symbol is g (which is fine, because there is now some citable use), but you then can’t conveniently chose to ignore what they say in their ACCELEROMETER PRIMER. What they have is so clear, even a five-year-old can figure it out: Accelerometers are used to convert an acceleration from gravity or motion into an electrical signal. Please stop with your WP:Tendentious editing, it is simply wrong to cherry-pick the parts of citations you think serve your needs. The primer is clear as glass. Greg L (talk) 00:12, 22 January 2009 (UTC)
    The problem is that truth and clarity are conjugate variables in writing. The primer may be clear as glass, but on this point, it is wrong. An accelerometer does NOT convert acceleration from gravity into an electrical signal. An accelerometer in free fall, or in orbit, where the only acceleration on it is that produced by gravity, will produce exactly the same electrical signal as an accelerometer at rest in deepest interstellar space, where there is no acceleration on it at all, from any source. Which is to say, zero signal. Zip. Nada. It registers zero acceleration. Do you deny this? If not, then something has to give. SBHarris 01:19, 22 January 2009 (UTC)
You do have to follow NPOV, and you don't get to successfully claim that anyone that insists you follow NPOV is being disruptive. Well, you can claim it all you like on ANI; but the admins will continue to ignore you, because it's not true. And an accelerometer's job is not to convert an acceleration due to gravity to an electrical signal. That's a gravimeter; and yes you can build a gravimeter with an accelerometer, but there's other system components you need to make it read correctly. Accelerometers in and of themselves are completely insensitive to gravity, and their operation and behaviour cannot be correctly understood in ignorance of this fact. I'm therefore finding your removal of valid references that state this to be extremely disruptive and harmful to the accuracy and values of the wikipedia.- (User) Wolfkeeper (Talk) 00:34, 22 January 2009 (UTC)
  • No, you are simply refusing to get the point and think you WP:OWN this article. Georgewilliamherbert was quite clear when he wrote In clear terms: There is no difference between a gravitational and an inertial acceleration, per Einstein. We're sitting in a 1-G static gravitational field. Further, a world-wide manufacturer of accelerometers says Accelerometers are used to convert an acceleration from gravity or motion into an electrical signal. You just don’t understand this technology at all well. A gravimeter is simply a high-sensitivity, high-precision accelerometer. Then there is DeeGee375 above, who uses accelerometers in daily life and he says the article is now correct. Greg L (talk) 00:47, 22 January 2009 (UTC)
  • A gravimeter is a high precission accelerometer used in a certain way (at rest, and with the understanding that the results are due to the fact that in this special circumstance, force measured is the acceleration of gravity). But as soon as you do anything else with it, it is no longer a gravimeter, because it doesn't measure gravity and never did.

    Your bathroom scale is the same, actually. It's an accelerometer, too. And also can be used as a really low-precision gravimeter if you stick it under you on the airplane seat while on the tarmac. But That doesen't mean it's really measuring gravitational fields-- only that it measures mechanical acceleration, since it's really only measuring measuring weight. It will tell you how many g-s you're pulling (g-force!), but not a thing about how much gravitational field is around you. It's insensitive to that. All it feels is pressure. SBHarris 11:34, 23 January 2009 (UTC)

You see, you still don't get it. The wikipedia works on not giving undue weight; you're cherry picking certain references and systematically remove all the other ones you don't like. And the most reliable sources say the polar opposite to your position which is solely based on poor quality sources, like manufacturers brief explanations they use to shift product.- (User) Wolfkeeper (Talk) 01:15, 22 January 2009 (UTC)
Wolfkeeper and Sbharris stop with this personal attack nonsense. Either come up with a better argument to support your point of view or stop editing the article. WorkingBeaver (talk) 11:38, 23 January 2009 (UTC)

Refusal to get the point with {disputed} tags

Ok, Wolfkeeper. Please explain precisely what is wrong with the following paragraph:

An accelerometer measures acceleration in one or more axis. It responds to both gravitational and inertial acceleration[1]. If you orient a stationary, single-axis accelerometer so its measuring axis is horizontal, its output will show zero gee. Yet, if you rotate the accelerometer 90° so its axis points upwards, it will read +1 g upwards even though still stationary. If you mount the accelerometer in an automobile with its axis aligned forward with the vehicle’s direction of travel, and drive down the road at a constant speed, it will read 0 g. Yet, if you hit the brakes, it will read about −0.9 g. Accelerometers respond equally to gravity and inertial acceleration.

1^ MEMSIC: ACCELEROMETER PRIMER

Please explain clearly and exactly what it is you dispute. Then we can go to dispute resolution if necessary. Greg L (talk) 00:55, 22 January 2009 (UTC)

It's pretty clear, is it not? If you mount a one-axis accelerometer sitting on the ground, so that it is looking at the vertical axis, it reads that there is a 1-gee acceleration upwards. That is wrong. There is no acceleration upwards-- the thing is clearly sitting at rest, on the ground, d(height)^2/dt^2 =0. The definition of "acceleration" does not leave any room for compromise. Such an accelerometer as part of an inertial navigation system would tell you that you are accelerating into the sky and thus maing rapid upward distance changes. That is in error.

Now, what's wrong with this picture? I and many others have attempted to explain, and a textbook has been offered which says the same. Accelerometers do not sense gravitational force or gravitational acceleration. All they sense is non-gravitational forces and non-gravitational accelerations. In this case, the device senses the +1 g mechanical force FROM THE GROUND which is pushing it toward the sky, and reports this as +1 g, upward. It does not "see" what in the Newtonian language is -1 g of gravitational force or acceleration, balancing this to keep the device to an acceleration of zero. If you let the device fall, it does report acceleration zero, but of course that's because it's not seeing gravity then, either.

In Einstein's language, gravity isn't a force, but near Earth time and space are bent so that a force mechanical is necessary to keep something "sitting still" in curved space and slowed time, where it follows a path through 4-space which isn't the natural one, which is a geodesic (produced by free fall). But that single force is also an upward force sufficient to produce +1 g, and again the accelerometer doesn't measure anything else, because (surprise) in Einstein's view there isn't anything else. No matter how you look at it, accelerometers don't measure gravity. SBHarris 02:58, 22 January 2009 (UTC)

There's no hope for you. You're deliberately creating a non neutral article, in numerous ways, and have been doing so since you started editing here. You've also been abusive, obnoxious, insulting, you've accused me and others of being meatpuppets, you've edit warred, you've selectively removed citations, you've accused me of ignorance of basic physics, I can go on and on and on about this. I can honestly say I've only seen one other editor behave as badly as you, and he later admitted to being a paid to represent the interests of corporations. It's been real educational. Contrasts strongly with the artificially dumbed down article you're trying to create, which is actually wrong on key points.- (User) Wolfkeeper (Talk) 01:10, 22 January 2009 (UTC)
  • I see. Well, I think there may be hope for you. But I think you’re busy attacking another editor rather than focusing on the single, important issue here: What the article says is the central point of any dispute. Please explain clearly and exactly what it is about the above paragraph that you dispute. You nebulously refer to something which is actually wrong on key points. Nothing can be resolved until it is clear what you think is incorrect. Perhaps then, we can learn from each other and produce wording that seems clear to us both. If not, we can go to dispute resolution if necessary. Greg L (talk) 02:01, 22 January 2009 (UTC)
Just as a naive first reader, I'd say that what is "wrong" with the paragraph is that it fails to distinguish between the facts that an accelerometer responds to gravitational force, but registers changes in inertial forces.. That is the crucial distinction. I believe that Sbharris above noted that an accelerometer in free-fall responds not to gravitational forces (other than that the accelerometer itself is accelerating, but you can't read that on the display screen - correct me if I'm wrong, but don't we enter the mysterious Mach/Einstein realm here?). The accelerometer has to be attached to something else to register a force/acceleration. An accelerometer at rest (in the local inertial frame) is a gravimeter. An accelerometer accelerating freely in a gravitational field is useless. An accelerometer undergoing non-uniform acceleration is a useful instrument. I'm on the bleeding edge of my physics knowledge, but is that helpful at all? Franamax (talk) 01:36, 22 January 2009 (UTC)
No, you've got the idea. You could be falling into a black hole, and your accelerometer would read "no acceleration" all the way down (neglecting tides). That doesn't mean it can't tell gravitational acceleration from other kinds of accelerations. It means it can't read gravitational acceleration at all. SBHarris 03:14, 22 January 2009 (UTC)
  • The point, Franamax, is that an accelerometer can not distinguish between gravitational acceleration and inertial acceleration; they are all acceleration from the point of view of an accelerometer. This is what the article says, which is also what Georgewilliamherbert said above: In clear terms: There is no difference between a gravitational and an inertial acceleration, per Einstein. We're sitting in a 1-G static gravitational field. What SBHarris says about accelerometer is freefall is what the article says: they read zero in freefall. Greg L (talk) 02:01, 22 January 2009 (UTC)
  • No. They are DIFFERENT from the point of view of the accelerometer. The accelerometer reads one kind, not the other. An accelerometer in free fall in a g field reads zero, even though it's clearly accelerating. Produce that same acceleration with a rocket, and the accelerometer would read it fine. So there's a difference. An accelerometer in free fall, far from any mass and in no g field, reads the same as an accelerometer in free fall in a strong g-field. Accelerometers do not see g-fields, or accelerations due to g-fields. They do see accelerations due to forces other than gravity, but not those due to gravity. I think I've said this every way it can be said. SBHarris 02:58, 22 January 2009 (UTC)
Sorry to interject, Sbharris, but you have disproved your own point - as you say, an accelerometer that is not being accelerated and far from any mass will read zero, the same reading that it will have when accelerating at one g in a one g gravity field. Hal peridol (talk) 19:30, 22 January 2009 (UTC)
Greg, sit inside the box with your horizontal accelerometer and watch the dial. If you're free-falling in space near the Earth, it will continue to register 1G or so (give or take a distant star or near planet). But this is the value you calibrated the instrument to when it was sitting on the table, so it is effectively zero, it is the reference value. If the value doesn't change, the instrument measures nothing. In a pure gravitational field, an accelerometer only measures what you set it to in the first place. It and you are just travelling along the shortest geodesic line in spacetime, which near the Earth happens to be pretty much straight downwards. Newton says that's a gravitational force that accelerates a mass, Einstein says that's the way things travel in spacetime when they're not subject to "other" forces. The floor you walk on looks flat, but it's not flat, it goes downward. Now if you get pissed off and throw the accelerometer across the falling box, it has a different inertial frame than you - and/or it is accelerating in your own inertial frame - now we can use the accelerometer and we can compare stopwatches if you threw it fast enough to approach the speed of light. These are difficult concepts... Franamax (talk) 03:45, 22 January 2009 (UTC)
  • Certainly I understand all this, Franamax. The article is clearly not addressing such complexities as zeroing-out the output of a 1 g signal from (a properly calibrated) accelerometer when on earth and then going into space and reading −1 g. Any reasonable interpretation is that we’re talking about the raw, absolute output. To do otherwise is as valid as saying my AC voltage from my 120 V outlet is only 5 V (after I had zero’d it at 115 V).

    Accelerometers measure strain on internal components. Just as the article says, the only way to get a single-axis accelerometer aligned vertically with respect to the barycenter of earth to read zero is to go into a free fall. If you stand on the earth with the accelerometer aligned the same way, it will register the force of gravity just as it will if you then go into a proper-motion inertial acceleration upwards. The force of gravity when stationary on earth affects an accelerometer precisely as does accelerating at 1 g inertially from earth in space. No accelerometer, not even a light beam, can distinguish between the two accelerations; they are one in the same phenomenon. Greg L (talk) 04:00, 22 January 2009 (UTC)

  • No, the force of the table or whatever is holding the acccelerometer up on the Earth, affects it precisely as does a force which accelerates it inertially at 1 g in space. Not surprising, because both are mechanical forces. The instrument is not measuring the g field, it's measuring that mechanical force. That is all these devices CAN measure. Keep the force in place (as when your accelerometer is held up a few feet in the air by a hovering rocket), and the reading doesn't change-- even if you remove the Earth. But remove the rocket and the reading goes to zero whether the Earth is there or not. So the Earth and its field has nothing to do with this reading. All it did was calibrate it, when you started by choosing your force to be large enough to make the instrument hover. SBHarris 06:23, 22 January 2009 (UTC)
Sbh, I'll take the liberty of correcting Greg to say that instead of the +1V signal changing to -1V, I'd prefer +1V changing to 0V; and where Greg compares a stationary accelerometer on Earth above, he should be rephrasing "at 1g ... from earth in space" to "at 1g in empty space" - whereupon we could all agree. We're all very close, it's just plus/minus one of something or other. The gravitational and inertial frames are exactly equivalent but subtly different. I can demonstrate this by twirling my nephew around by his ankles. Franamax (talk) 06:55, 22 January 2009 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Talk:G-force/Archive_2&oldid=1079340772"