View Full Version : Surface plate mounted on five points?

11-11-2016, 10:24 AM
Has anyone seen this before? I thought that maybe it's resting on three points, and the other two are just shy of the surface to stop it from tipping over if someone leans on it. It's 60 X 60 X 10 cm, about 2' X 2' X 4".


11-11-2016, 10:51 AM
Probably backed off on 2 as you said, guys used to map the plates in work, autocolimeter job, even actually especially brand new ones, they used to do an optical bench too, that thing was full of holes.
Mine is 3/4 plywood so it's guaranteed to be as flat as it is!
The supports look more like machists jacks I used to own, same thing I suppose

11-11-2016, 11:10 AM
Yes, I've seen support installations like that. It's essentially a three point support at the Airy point with two outboard of the single support end to prevent tipping if a heavy workpiece were placed on the corner. They're usually just snugged up or left slightly slack to catch if there were any tendency to tilt.

old mart
11-11-2016, 01:16 PM
I'd not be happy with that setup, one hard knock and you could have half a mountain falling on you.

11-11-2016, 02:15 PM
I'm not sure what you mean. What's wrong with the setup?

11-11-2016, 02:32 PM
I'd not be happy with that setup, one hard knock and you could have half a mountain falling on you.

I don't think someone bumping into it with their hip would move it. But something with more momentum could result in it moving off the intended support points. Not to mention the edges are exposed to being chipped by mishandled heavy items being moved around or onto the plate. A somewhat protective frame around it attached down to the stand would not be a bad idea at all.

I also assume that the jacks are secured to that frame. If not a hard enough nudge would indeed see the plate go tumbling if the jacks were bumped off the frame.

Paul Alciatore
11-11-2016, 02:39 PM
I have to agree with BC. Some kind of protective frame (angle iron) around the plate is definitely an advantage for both the plate and the people working around it. And it would probably eliminate the need for the extra two "supports".

11-11-2016, 03:16 PM
Ok, got it. Probably not going to be an issue for me. It will be in a corner with a cabinet beside it, so it will be boxed in from three sides. The only way you could make it fall off would be to pull it towards you.

This is a MeasureMax surface plate, probably not the best quality. Couldn't find much info about the company, but I think it comes from either China or India. Would prefer something better, but hard to find around here.

Thanks for all your replies.

11-11-2016, 03:39 PM
I wasn't thinking of an angle iron frame as Paul suggests but such a frame would kill two birds with one stone. The lower flats of the "L would perform the safety function given by the added jack screws. All that is needed is for the three jacks to lift the plate slightly clear of the lower inside of the frame.

Pinstripe, you don't even really need jack screws for the plate. Three old hockey pucks or three small pads of plywood at the right spots works just fine. Seems to me the only upside of using jacks like this is that you can level the plate at the same time you support it on the three points. If you have a good machinist's level then it's something that would be nice because it gives you the option of using the level during any work. But otherwise any old three pads to lift the plate a touch off the bench top is all you really need.

11-11-2016, 03:51 PM
A brainwave occurred to me just after hitting "post".

If I had a suitable machinist's level I think the way I'd go is put the single point at the rear where it's hard to reach. And for giggles stick the slightly shorter "safety posts" out to the sides as well. The single rear support does not need to be adjustable. The other two would be jack screws that are able to adjust to a little shorter to a little longer than the fixed rear support. With the two on the front able to be reached you can adjust for level in both axes easily.

They COULD be just two independent jacks with three pads of something like 1.5" round bar for the rear pad and safety pillars. But in the end I think I'd use a sheet of plywood and fit it with some 1/2" UNF inserts that stick out the bottom slightly to rest on washers as pads on the benchtop. Then I'd use short 1/2" UNF screws with the top of the heads skimmed in the lathe to remove the markings to support the plate. I'd likely use a pad of something like 1/8 masonite so the hard contact points of the bolts are padded slighty against the stone. A mounting and leveling plate of this style would aid in preventing the plate slipping around or tipping over the jacks.

11-11-2016, 04:24 PM
Thanks BCRider. I'm having a bit of trouble understanding the benefit of your setup though. The Masonite being more grippy than steel on stone I get, but not the rest of it. This won't be sitting on a bench, it comes with a stand and I think the jacks are attached to the stand. There is also an adjustable levelling foot on one leg of the stand. It looks like this.


Mark Rand
11-11-2016, 04:34 PM
This is a completely normal way of mounting a surface table. Both my 36"x48" and my 24"x36" Crown Windley tables are mounted this way... With one minor difference, the metal support on the 36"x48" has five legs as well to avoid sagging with the centre support of the granite. All of the ones at work are the same. The 'jacks' are bolted to the metal frame. The feet should be adjustable as well on a good design for leveling equalization of load.

The outer two supports on the three-support side are snugged up so that there is no gap at all. That way the plate won't tip when you put a heavy item on one of the corners at that end. The heaviest single item I've worked on with my 36"x48" table was about 400lb. That's about 40% of the weight of the granite. That weight on an unsupported corner might have been marginal.

11-11-2016, 04:43 PM
OH! That's a bit different. I'd assumed you ran across the picture and were asking about it out of basic curiosity.

I see that the stand has one foot of the base that is adjustable. That's excellent since I've yet to run into any truly flat and curve free floors. The adjustable foot will allow you to more or less equalize the load of all four legs against the floor for stability. Turn the foot down so it's in contact and give it your best feel/guess at dialing in a pressure that you estimate is equal to the other three legs. Checking by lifting hard but not hard enough to lift clear on the one with the foot and the leg next to it will pretty quickly show you if the pressures are near enough to equal by how it seems to want to rock or not rock on you.

If the jacks are indeed secured to the stand like they SHOULD be then there's really not much else to do other than you can use them to set up the plate so it's level with the aid of a machinist's level. Once leveled just bring up the safety jacks until they touch and back off a quarter to eighth turn so they are not a supporting factor that might affect anything.

If you don't have a precision level I'd still level the plate as best you can with a builders level even if it's only the one commonly found in the handle of an adjustable square. It'll aid in slowing down stuff that wants to try to roll off the plate. And if using it for surfacing with wet sanding it'll aid in keeping the pool of lubricating water or oil on the paper.

For whatever reason I don't like the idea of a hard point like a metal screw jack sitting against something as hard as the granite. So I might still be inclined to use that piece of masonite to soften the pressure and spread the contact load out a little. The added friction should also aid a little in keeping the plate from skidding around at all and in keeping the screws from rotating with any vibrations from use of the plate itself or from any vibration up through the floor from other machines. But I may be worrying about nothing and it's my OCD kicking in again.

EDIT- I see from Mark's reply that the instructions for his plates call for the tipping support jacks to be in contact but with, I assume, no real pressure. Sounds like a good time to go with the option that is directly from the manual for a change.... :D Might be worth checking them at some point a week or two afterwards to ensure nothing settled at all in the stand and ends up putting too much load on either of the tip supports.

Very nice score on the plate and stand too. It might be an import but if it's more accurate than you require by a factor of 10 then it's accurate enough for anything you might need it for. Anything more becomes a matter of bragging rights and feeling posh.

11-11-2016, 05:09 PM
Thanks Mark and BCRider. No gap definitely sounds like the way to go. I've got a 6-inch machinist's level, so I will use that to level it.

11-11-2016, 11:02 PM
When you align motors to pumps the very first step is to identify the "soft foot" which is the one making full contact. You then shim that one before proceeding with the alignment.

11-12-2016, 08:39 AM
The plate plus the stand is too high for the wheelbase. The legs of the stand should be outboard of the plate extremities. Think of the center of gravity.


11-19-2016, 10:20 PM
Unless you really do have a "need" (as opposed to just a "want" then as previously discussed as regards "levelling lathe" beds et all, there is no real need for the plate to be "levelled" - unless you really "need" it to be.

All the plate needs to be to perform as the manufacture made it and intended it to be then it only needs to be supported on- and only on - the three "airey" points" (on the under side of the plate).

As I recall, no manufacturer specification or catalogue required the "plate" be "level" - just as said - on the three "airey points".

"5-point support" seems more in keeping with the legs on a good modern office chair (as opposed to the older modern "3" or "4" legged chair).

11-20-2016, 02:31 AM
I wouldn't like to use the setup shown in post #11. The stand would be adequate for a smaller plate. With this large plate it won't be stable enough to my taste. Besides, I suppose that one or two of the Airy points are located under the unsupported hanging portion on the right (unless the plate is hovering above the stand on the picture and doesn't reflect its intended position on the stand).

P.S. I'm sure the picture was created by an artistically inspired "photoshopper" rather than by somebody with a common sense. Putting the plate vertically on its edge would be even more interesting. ;)

P.P.S. I've just visited the site where this photo came from and read on the specs. The plate is 2'x2', and the hanging portion of the plate (if it's centered) will be about 4". This sounds acceptable.

11-20-2016, 03:29 AM
The stand comes with the plate. I checked the manufacturer's site, and it shows the same stand. I'm not going to use the stand because I want to make better use of the space underneath it. How are the Airy points normally marked at the bottom of the plate?

Edit: Even though I'm not planning to use the original stand, I will be using the same support points.

11-20-2016, 06:11 AM
This should clarify the "airy points" situation somewhat:


https://www.google.com.au/search?q=surface+plate+airey+points&rls=com.microsoft:en-AU:IE-SearchBox&rlz=1I7IRFC_enAU360&biw=1280&bih=585&source=lnms&tbm=isch&sa=X&ved=0ahUKEwj8v7KAmLfQAhUDULwKHeryAxoQ_AUIBigB#tbm= isch&q=surface+plate+airy+points

Mark Rand
11-20-2016, 09:30 AM
It should be noted that for a surface or straight edge to have the minimum deviation from flat, it should be supported at its Bessel points, not its Airy points. The Airy point support will ensure that the ends of the piece are horizontal, but the sag in the centre is larger than that caused by supporting at the Bessel points.

Odd that people keep referring to the wrong locations...

11-20-2016, 12:34 PM
How are the Airy points normally marked at the bottom of the plate?
It will either have clear marks underneath or some sort of pre-attached bumps. Those locations were used to support the plate when it was lapped, and they should be used for keeping the plate mounted.

If there are no marks (unlikely even for the Chinese plates), support your plate on three points located about 1/5 (but not more than 1/4) of the total length or width off the sides. For the 60cm x 60cm plate the points should be 12cm (and not more than 15cm) away from the edge. If you have one point on the left side and two points on the right side, the one on the left should be located 12 cm from the left edge and centered between the back and front sides (30 cm away from each). The two points of the right will be 12cm off the right edge and 12cm from the front (for one of them) and back edge (for the second one).

P.S. Mark is right on the proper name if applied to surfaces vs. straight edges. I was going to mention it too, but decided against it, because the term "Airy point" is used much more often (although often incorrectly).

11-20-2016, 12:53 PM
According to this source...


The spacing from the edge of a plate of length "L" should be:

Airy => 0.21132 * L

Bessel => 0.2203 * L

Minimum Deflection => 0.2232 * L

Thus the difference in point spacing between Bessel and Airy for a 600 mm plate would be:

(0.2203 - 0.2113) * 600 = 0.009 * 600 = 5.4 mm = 0.21 inches

I'm having trouble believing that difference matters much in any home shop.

11-20-2016, 01:12 PM
It should be noted that for a surface or straight edge to have the minimum deviation from flat, it should be supported at its Bessel points, not its Airy points.Both Airy and Bessel were only concerned with straight bars, not surfaces.

... because the term "Airy point" is used much more often (although often incorrectly).
The use of the term 'Airy points' applied to surface plate mounts is terminology popularised only in the echo chamber of internet forums. I've not yet heard it used in the wild, nor have ever seen any reputable manufacturer's literature referring to their mounts as Airy points. Has anyone ?

The support points on a surface plate are not calculated, they are not special. They don't need a fancy name. They were simply chosen to be 'about right'.

Historically the mounts may once have been arranged by some manufacturers in 'Airy ratios' (in each axis), but this was ultimately an arbitrary and baseless 'best guess', and doing so doesn't make them 'Airy points' in any functional sense.

The flexural analysis of a two dimensional surface supported on three points is immensly more difficult than a bar on two points, and was far beyond the capabilities of the 19th century. Neither Airy nor Bessel could have possibly had anyhing to contribute on the matter.

A hundred or more years or so later, numerical finite element analysis computations demonstrated that the three point mounting required to achieve the minimum possible distortion of a flat plate does not directly equate to any of the commonly used support positions. Nobody concerned with surface plates became the slightest bit excited however because this was of purely academic interest, and would only be significant in the support of a plate that was perfectly flat in a free state, e.g. in zero gravity.

A real surface plate is not made to be flat in the free state. As has been said many times it should become flat when distorted by gravity while supported on the mounting points used during manufacture and calibration, which are therefore the points which are optimum in use.

That's it, plain and simple. Mr. Airy not involved in any way shape or form. Not at all. Never was, never will be...



11-20-2016, 01:31 PM
I'm having trouble believing that difference matters much in any home shop.But when we rest our Chinese plates on the proper points, we feel much more confident. Like when we figure out how parallel the sides of out Chinese gage pins are.

The best of us buy Starrett plates to compensate for the shortcomings of an unheated garage.

Psychological aspects of the process is what matters the most, after all. ;)

11-20-2016, 03:56 PM
The thickness of a ("Starrett") surface plate is governed by its deflection under load at the plate mid point - see page 2 of the Starrett pdf file:


The applied load is actually 50 pounds per square foot with the plate supported at its Airey points.

I think that some may need to get acquainted with the referenced Starrett *.pdf file - above.

The deflection is governed by the plates mechanical properties and the thickness is governed by the deflection of the plate.

Plates are normally "rated" ("graded" and/or "calibrated") for "flatness" with the plate supported on its airey points with no applied load on the plate at its mid-point or else-where.