Thanks: 0
Needs Pictures: 0
Picture(s) thanks: 0
Results 16 to 30 of 53
Thread: Calibrating a Surface Plate
-
9th Apr 2013, 09:14 PM #16
Hi Michael,
I'll send you a PM about the RepeatOMeter.
The laser calibration is something I've been wondering about, I'm not sure what part of the instrument you would calibrate? I think it's all down to to how accurate the laser frequency is, the wavelength varies a bit in air as temperature, humidity, pressure changes, hence the automatic compensator, but these corrections are second order.. There is an adjustment that you have to do when aligning the instrument that is called laser tuning, but I'm not sure exactly what that is changing ( or by how much )..
So ultimate accuracy comes down to the laser frequency, which derives from the atomic structure of the Helium/Neon atoms.. does that make it a primary standard in itself? I don't know.
As to whether Josh wants to open up "Josh's bargain basement slightly used surface plates" you'd have to ask.... for myself , if anyone wants a surface plate calibration I'd be happy to help out if possible.
Regards
Ray
-
9th Apr 2013, 10:00 PM #17Most Valued Member
- Join Date
- Jul 2010
- Location
- Melbourne
- Posts
- 9,088
Hi Ray,
Wouldnt the base of the mirrors/reflectors etc need to be checked for flatness?
Shouldnt there be four measurements in the middle of your map?
Stuart
-
9th Apr 2013, 11:04 PM #18
Hi Stuart,
Correct on all counts. There is a flatness spec for the mirrors, I remember seeing it somewhere in the HP literature. But I don't think there is anything that can be calibrated. And yes there are 4 points in the center, but since 2 of them were the same
I found a bit of a write up on laser interferometer calibration, http://www.mikes.fi/documents/upload...rferometer.pdf but from what I can understand the calibration involves calibrating the temperature, pressure and humidity sensors, and the laser wavelength checked against an iodine stabilized helium neon standard.. the table shown in the pdf of an uncalibrated system versus a calibrated system shows the major variance coming from material temperature sensing. For surface plate calibration that doesn't enter the equation. and the contribution to error of the uncalibrated laser frequency/wavelength seems minimal.
On thinking about whether or not to lap the plate..
The flatness is defined as the distance between two parallel planes that enclose the surface, and for B grade it's 1400 uinches, for A grade it's 700 uinches, and AA it's 350 uinches, we measured 976 uinches, so it's closer to A than B, whatever that's worth.. so to get it to A grade flatness spec, we would have to knock down the purple section by 300 uinches or better, that means lapping off 7 or 8 microns, ( or 3 tenths if you prefer )..
On reflection, I'm amazed that the laser is good enough to measure the wear pattern and determine how the plate was used in it's past history.. maybe I'm easily amazed..
Regards
Ray
-
9th Apr 2013, 11:15 PM #19Most Valued Member
- Join Date
- Oct 2007
- Location
- Sydney
- Posts
- 3,112
Do you have MH29? There is a section on surface plates in there that was absent in the previous version I had. It deals with using an interferometer, autocollimator, and Repeat-o-Meter to calibrate the plate. I can send you the relevant pages if you don't have it as I have both the hardcopy and dvd version. You probably won't like what it has to say about re-lapping the plate however. The bottom line is they say, "don't try this at home", you'll probably make it worse.
Pete
-
9th Apr 2013, 11:35 PM #20
-
10th Apr 2013, 12:13 AM #21Dave J Guest
With all that equipment I think your a long way from the average home shop Ray.
Dave
-
10th Apr 2013, 07:48 AM #22Philomath in training
- Join Date
- Oct 2011
- Location
- Norwood-ish, Adelaide
- Age
- 59
- Posts
- 6,540
Regardless whether calibration is strictly needed for determining flatness, if you want to be taken seriously by prospective clients calibration will be needed. Defence companies for example won't touch you without it (and they seem to be the ones with the money at the moment)
I think Phil (Machtool) was saying that he had a job where he was not allowed to use uncalibrated dial indicators.
The other one is checking the dimensional relationship of large machine tools. It is assumed to be perfect and proven by acceptable parts being produced but again there is uncertainty where any variations are coming from.
Michael
-
10th Apr 2013, 09:20 AM #23Golden Member
- Join Date
- Feb 2010
- Location
- Ballina, NSW
- Posts
- 900
Hey Ray, if you post the excel file and dimensions, I'll do you up a surface gradient and contour map if you want.
Cheers
- Mick
-
10th Apr 2013, 09:36 AM #24Most Valued Member
- Join Date
- Jun 2011
- Location
- Australia east coast
- Age
- 71
- Posts
- 2,713
That's my understanding as well. One of the clients I've kept on (he's one of my ex-staff) does Defence Dept work. I wrote him a database application to keep track of everything including when his equipment was last calibrated and how long the calibration is good for. Easy job as we used to take all the sensors off of our ships every year and send them off to the manufacturers for testing & recalibration so I'd done it all before....
PDW
-
10th Apr 2013, 10:07 AM #25Most Valued Member
- Join Date
- Oct 2007
- Location
- Sydney
- Posts
- 3,112
As an aside Ray I think this edition is a big improvement on previous editions, it seems to cover metric much better than it used to (I always felt it considered metric somewhat a novelty) and seems to have additional sections my previous one doesn't. I think you'll find it a very worthwhile investment over what you have now. I'd been considered it for a while, but at over 200 bucks for the large print plus CD edition it's not like going down to the local store and buying a paper! I'm old school, and still prefer to read from books, and for work I study from books, but the CD has additional information on it, including material removed from later editions. Plus, I can refer to it while I'm away, as that's when I plan out my work for when I'm home. I don't recall anything on surface plate calibration in my previous edition, and my decision to buy it was made when I saw the section on using an autocollimator to calibrate plates, as I am currently buying one. But the Black Book is the one that gets used daily within the workshop.
Machinery's Handbook isn't something pitched to the home workshop, the paragraph referring to lapping plates followed the section on interferometer calibration! They are very economical with space in the book, and I figured if they went to all the trouble to dedicate a whole paragraph to saying "don't try to lap a plate yourself", there must be a good reason for saying it. I have no idea what that reason is, and why they would specifically make that warning. But I think it worth pointing that out and finding out why they would say that, then making a decision as to what would be involved in overcoming it. As mentioned, I can't see why it should be such a big deal, at least in theory, but clearly the MH editors do, as I have not seen any similar cautions anywhere else within it.
For what it's worth, I think if the plate's errors are known, you're half way there to being able to use it accurately anyway, and it's probably not worth losing too much sleep if it's decided that it's not worth attempting to lap it to tighter tolerance.
Pete
-
10th Apr 2013, 11:54 AM #26Distracted Member
- Join Date
- May 2010
- Location
- Lower Lakes SA
- Age
- 58
- Posts
- 2,607
This may be of interest. It briefly shows several lapping processes.
-
10th Apr 2013, 12:05 PM #27
Not sure I'm going to be doing much defence dept contracting, I currently do software and hardware for state govt that satisfies my need for mindless bureacracy...
On the CNC front, I have a Renishaw QC10 ball bar calibrator, that is a pretty slick bit of gear for CNC machine calibration, it uses a calibration bar made from zero temp co-oefficient ceramic "Zero Cera"
Hi Mick,
Thanks for the kind offer, I've attached the excel spreadsheet, I exported it from Openoffice as Excel-97 so I hope it retains the formatting. go to the Summary tab and you'll find the data you need for plotting, the grid is a 4"x4" grid, which is determined by the footprint of the optical fixture. So the spacing along the diagonal lines is in 4" intervals, you'd have to convert those diagonal co-ordinates to x,y,z format.
The other tabs show the raw angular data, and the calculations to convert the angles into corrected elevation data.
Hi Pete,
I use the "black book" series mostly, and Machinery's Handbook only for the more unusual stuff or greater detail, but I hadn't realized I was so far out of date...
As far as the surface plate goes, it was a gamble to buy it considering the poor general condition it was in, and I wouldn't have considered buying it without having the gear to calibrate it. Likewise I'm more confidant of taking on the lapping of a plate when I know I can measure it and see what's happening. One of those rare occasions when you can be lucky enough to have the right gear, I guess, anyway within 1 thou over a 6'x4' table isn't too bad. Good stuff granite.
Regards
Ray
-
10th Apr 2013, 01:20 PM #28Golden Member
- Join Date
- Oct 2012
- Location
- Australia
- Posts
- 621
Creating your own primary standard such as a surface plate is an interesting problem.
Scaping is the only technique I know of that will let you auto generate a flat plane with three identical plates, any other technique requires a "measured map" of the surface or comparision existing master of a known standard so that you know where to take the material off.
So as I see it the main problem of lapping at home is that home workshops normally do not have the means to map the geometry to a sufficient accuracy. The cost of autocollimators, laser interferometers, differential levels and planekators put them almost purely in the domain of industry, physics, metrology, calibration labs.
The other problem for the home workshop is control over the environment, it would be quite an impressive home workshop that had a clean room with climate control. That being said it should be possible to control the conditions sufficiently to achieve an A/0 grade plate, AA or 00 or better would be much harder(or impossible) to reach without a finer control over the environment.
The third and fourth problem I can see is that the final lap itself needs to be of a good quality and flatness and the technique right if the plate is to pass the repeatability calibration.
The fifth problem is the time required to calibrate, lap, calibrate, lap, calibrate ad infinitum would be considerable for a plate this size. After some practice Ray and I got the calibration time down to about an hour and a half for a 8 line union jack calibration with some 375 measurements for a 125 data points and 9 grid points and two degrees of freedom to solve (easy enough to do with a spread sheet).
The 8 line is the minimum for the calibration according to the specification, This is maybe an insufficient a map to lap by?
A full 12x17 grid (4"x4") by my calculations would have 234 grid points, 924 data points, 2772 measurements and 527 degrees of freedom to solve (a much more involved problem, that I'm not sure where to start).
Incidentally the finest resolution map possible with the optics would be a 2"x2" grid, which would have 875 grid points, 3488 data points, 10464 measurements and 2276 degrees of freedom and would take so long as to be impractical.
-Josh
-
10th Apr 2013, 04:33 PM #29Most Valued Member
- Join Date
- Oct 2007
- Location
- Sydney
- Posts
- 3,112
The thing that strikes me about granite plates is that some that appear to be in "poor" condition may actually be quite accurate, conversely some that may appear pristine, could be totally useless as far as measurement. Chips and so on shouldn't really affect accuracy unless they're tragic, but a worn section clearly will. I thought the data from your plate looked to be pretty good Ray. I think I'd turn it around and call the job done personally
Josh, it could be as you say, I have no idea as to why they specifically provide that warning. They aren't assuming any auto-generation or lack of measurement ability; it follows on from the discussion on the interferometer software, and says that once the software has provided the map, a determination on whether lapping if required can be made. To save any further confusion, this is specifically what they write:
Lapping a surface plate is a very specific skill and should not be attempted by an amateur or enthusiastic technician however good the intentions may be. Lapping requires specialist tools and compounds as well as skill. Attempting to lap a surface plate by an inexperienced technician may well cause more damage than is already present. Always consult an experienced and proven technician in this field.
Pete
-
11th Apr 2013, 03:02 AM #30
The things you do when you're easily amused..
Surface Plate Effects - YouTube
Regards
Ray
Similar Threads
-
Surface Plate Substitutes
By Oldneweng in forum METALWORK GENERALReplies: 41Last Post: 28th May 2012, 08:41 PM -
what do you think of my surface plate
By welder in forum METALWORK GENERALReplies: 11Last Post: 21st Nov 2011, 05:33 PM -
surface plate Brisbane
By mic-d in forum METALWORK GENERALReplies: 25Last Post: 6th May 2010, 10:01 PM