PDA

View Full Version : Lapping A Granite Surface Plate



Neil Jones
05-25-2009, 06:06 AM
Do you need multiple grits of lapping compound? Is it diamond lapping compound that is used? How many lapping plates are used and how big do the lapping plates need to be?

SteveF
05-25-2009, 06:27 AM
A few years ago I took my surface plate to Starrett to have it certified. When I dropped it off the guy asked me "Do you want a quick tour"? Hell, yes! The plates getting recerted were measured with an autocollimator and then turned over to a guy who lapped it according to the map of measurements. The lap he used was a block of granite maybe 24" x 6" x 4" with a handle closer to one end. He moved it around in a circular pattern and checked his work with selection of Aluminum I-Beams with a mounted dial indicator. Some had pads on the end with a dial indicator in the middle sticking down, some had the indicator on the end. Once he felt it was flat again it went back to the guy with the autocollimator for another mapping. Don't know what type of compound was used.

Steve.

tdmidget
05-25-2009, 07:59 AM
Yes it's diamond paste. No do not even think about trying this on your own. Our calibration service does this to way less than .00005, more like .00001 and you need the proper tools which cost a fortune. Remember it is done in place for a reason. If you move or change anything on a large plate you will likely affect it's flatness.

Forrest Addy
05-25-2009, 10:22 AM
Yeah what td midget said. Wllingness to lap a surface plate does not equate to success after the plate has been lapped.

There's some fine technique and elaborate instrumentation required to ensure success and acheive cerfifiable accur4acy. At the minimum you would need the basic lapping materials plus a Rhom Planikator (?) system to quantify progress as you procede. You will also need the Federal Specification GGG-P-463c because it has in it the info you will need to properly support the plate for inspections and the acceptance criteria for the various grades and sizes. Unless there is calibration facility near you to document your finisheded surface plate you're better off in these days of cheap immports to demote the suspect surface plate to a paker;'s kitchen and purchase a replaement.

I'm not saying a rank amateur couldn't do a servicable job of lapping and calibrating a granite surface plate. I mere merely suggesting that it's not too practical given the low cost of new replace ment import plate Vs the time to acquitre and the cost of the calibraion equipment and devevelopment of the skills. For exam0ple Iit took me over three years to dinf and urchase a Federal electronic differential leveling system resolving 1/2 arc second for l3ss than $1500.

lazlo
05-25-2009, 10:35 AM
Out of curiosity Neil, how were you proposing to measure the flatness of the surface plate? :)

Neil Jones
05-25-2009, 11:07 AM
Yes it's diamond paste. No do not even think about trying this on your own. Our calibration service does this to way less than .00005, more like .00001 and you need the proper tools which cost a fortune. Remember it is done in place for a reason. If you move or change anything on a large plate you will likely affect it's flatness.

For sure I'm going to try it "on my own". I don't consider less than $1,000 expensive to buy a used Rahn Repeat-O-Meter and a used Hilger and Watts Autocolliminator to be a "fortune".

Neil Jones
05-25-2009, 11:11 AM
You will also need the Federal Specification GGG-P-463c because it has in it the info you will need to properly support the plate for inspections and the acceptance criteria for the various grades and sizes. Unless there is calibration facility near you to document your finisheded surface plate you're better off in these days of cheap immports to demote the suspect surface plate to a paker;'s kitchen and purchase a replaement.

Anyone can download the Federal spec you mention. I have the link at work if anyone needs it.

When is your next scraping class going to be held?

lazlo
05-25-2009, 11:12 AM
A Repeat-O-Meter just tells you localized flatness, and an autocollimator will tell you global flatness. It's going to be a bitch to extrapolate a flatness map to hand lap it...
In Moore's book, they show the giant surface plates scraped together with the three-plate method, and the plates are checked for flatness by suspending a straight edge above the surface place, and exhaustively measuring each point with a gauge head.

I've read that granite surface plates are lapped on a large rotary lap.

Neil Jones
05-25-2009, 11:17 AM
A Repeat-O-Meter just tells you localized flatness, and an autocollimator will tell you global flatness. It's going to be a bitch to extrapolate a flatness map to hand lap it...
In Moore's book, they show the giant surface plates scraped together with the three-plate method, and the plates are checked for flatness by suspending a straight edge above the surface place, and exhaustively measuring each point with a gauge head.

I've read that granite surface plates are lapped on a large rotary lap.

The way I seen it done is to use a straight edge and a mirror mounted on carbide points with the autocolliminator to map out the plate and you follow the Federal spec

lazlo
05-25-2009, 11:47 AM
The way I seen it done is to use a straight edge and a mirror mounted on carbide points with the autocolliminator to map out the plate and you follow the Federal spec

An autocollimator is used to detect global flatness (for example, twist): it tells you how planar the two points are. The problem is that you need the flatness for all the points in between. You could conceivably move the autocollimator around thousands of times, and map out the entire surface, but it takes awhile to setup and register each time

I'm guessing you bought a big granite surface plate and it's not flat? You might be better off with a Talyvel -- an exquisitely sensitive electronic level (really a gauge head with a pendulum sensor). With a Talyvel and a lot of patience, you could record a 2D map of the flatness of the surface plate.

http://reliabletools.cachefly.net/itemimages/JanuaryV09/v21d.jpg

Mine has an analog display head, but the new ones have a digital interface with software that maps the 2D flatness of the surface (you can do this by hand with the analog head, of course):

http://www.spectrum-metrology.co.uk/images/grid.gif

sidneyt
05-25-2009, 12:01 PM
What size and grade of surface plate are we talking about? I purchased a Grade B 12 x 18 surface plate from Enco for $25 with free shipping and you can buy a grade A plate (.0001") with two ledges for $36.95. Of course the larger size surface plates are more expensive (to buy and ship), but even so it makes it difficult to imagine how you could save much money especially if you are paying the shipping cost plus the redo of the plate. If you do it yourself wouldn't you need another surface plate to use as the lap?

Neil Jones
05-25-2009, 12:05 PM
An autocollimator is used to detect global flatness (for example, twist): it tells you how planar the two points are. The problem is that you need the flatness for all the points in between. You could conceivably move the autocollimator around thousands of times, and map out the entire surface, but it takes awhile to setup and register each time

I'm guessing you bought a big granite surface plate and it's not flat? You might be better off with a Talyvel -- an exquisitely sensitive electronic level (really a gauge head with a pendulum sensor). With a Talyvel and a lot of patience, you could record a 2D map of the flatness of the surface plate.

I have not purchased a big granite surface plate that is not flat but I'd like to. :D

Lapping a surface plate in is a skill I'd like to acquire. Same with scraping.

When I've seen surface plates inspected and lapped the only tools that were used is what I've already mentioned. I've never seen anyone use a Talyvel. Seems like it might make the job a lot easier.

Speaking of scraping are you familiar with this book and video and if so do you think they represent good value for someone who would like to learn how to scrape?

http://www.machinerepair.com/orders.html

Neil Jones
05-25-2009, 12:10 PM
What size and grade of surface plate are we talking about? I purchased a Grade B 12 x 18 surface plate from Enco for $25 with free shipping and you can buy a grade A plate (.0001") with two ledges for $36.95. Of course the larger size surface plates are more expensive (to buy and ship), but even so it makes it difficult to imagine how you could save much money especially if you are paying the shipping cost plus the redo of the plate. If you do it yourself wouldn't you need another surface plate to use as the lap?

I've always liked the idea of spending thousands of dollars to try and save a few bucks. You end up with a skill that can reward you (mentally) many years later when you forget about how much it really cost you to acquire that skill. :D

lazlo
05-25-2009, 12:19 PM
Speaking of scraping are you familiar with this book and video and if so do you think they represent good value for someone who would like to learn how to scrape?

http://www.machinerepair.com/orders.html

I have the book and video, and they're both excellent. The video covers predominantly hand-scraping, and the book -cough- dovetails nicely with technical background, instructions on sharpening the blade, to to measure for aligment, etc. Michael was a little disorganized about prices and shipping costs when I called, so I bought both from a Dapra distributor.

By the way, you asked on the other thread about the "Scraping for Alignment" DVD -- it was also excellent. Here's my review:

Bridgeport Scraping/Rebuilding DVD Review (http://bbs.homeshopmachinist.net/showthread.php?t=29819)

dp
05-25-2009, 12:47 PM
Short of attending a class the video looks like it will provide a decent framework for learning. As I recall the class I took from Forrest the new material I picked up included work holding, tool usage, marking and analysis, and the need for discipline in tool usage including stroke length, stroke angle and pressure, posture, preparing the work for marking (India stone, remove all debris!).

In addition I learned about the pace of progress which is surprisingly fast for the small blocks we were working on. Finally I became conscious of simultaneity. The object we were working on was intended to come to size, perpendicularity, and surface quality all at the same time, ending with a final stroke. Knowing when that final stroke has happened is elusive to say the least, but it is the most important thing to know - when to quit.

Along with that is recognizing you are on or not on a path to successful completion. If you are not on a path to completion you are on a path to failure, so you need a mental map of where you are and where you are going, and what the lay of the land is, and that your progress is taking you toward success. The understanding of simultaneity is useful in this analysis.

Not included was actual scraping processes such as frosting and working with dovetails and round work such as bearing surfaces.

Jpfalt
05-25-2009, 02:05 PM
I agree that the Chinese surface plates are really cheap. After checking a few, they are also fairly accurate.

If so, the why are granite straight edges and squares so darn expensive?

I have used fine pink granite material to make hand flats up to about 6 x 6, straight edges, angled straight edges for dovetails and have worked some pieces for square edges. The measurement methods are the same as for hand scraping cast iron and the diamond abrasives are fairly effective in cutting.

I use grinding wheels and pads intended for working granite countertops. (I paid for the tools by doing counter work for our kitchen and makig granite top end tables from the remnants.) and using the diamond plated files and hones that are now available for very little money. HF has a 4 sided hone with 2 x 6 surfaces in 200, 300, 400 and 800 grit that can be removed and glued to a more rigid backing. The final lapping for accuracy can be done with a single grit of 10 micron diamond lapping compound that I originally bought for lapidary work used on a cast iron or copper lap.

I limit myself to 100 millionths and use a .00005 dial indicator on a base and measure per the federal standard. It's more than good enough for my work. For spotting, I use cadmium yellow pigment in light oil and spot just like with cast iron scraping.

SteveF
05-25-2009, 03:02 PM
You could conceivably move the autocollimator around thousands of times, and map out the entire surface, but it takes awhile to setup and register each time

And that's the general concept of what they do in the Starrett plant. The autocollimator is setup just off the corner of the plate parallel to the short axis, and a 45 degree mirror to reflect the view down the long axis. Then they map multiple points down the long axis, move the mirror, map another line, etc.

Steve.

Neil Jones
05-25-2009, 04:55 PM
And that's the general concept of what they do in the Starrett plant. The autocollimator is setup just off the corner of the plate parallel to the short axis, and a 45 degree mirror to reflect the view down the long axis. Then they map multiple points down the long axis, move the mirror, map another line, etc.

Steve.

I believe the mirror rides up against a straight edge. I think the autocollimator stays on the table. I'll check this as soon as I get a chance.

Neil Jones
05-25-2009, 05:13 PM
Short of attending a class the video looks like it will provide a decent framework for learning. As I recall the class I took from Forrest the new material I picked up included work holding, tool usage, marking and analysis, and the need for discipline in tool usage including stroke length, stroke angle and pressure, posture, preparing the work for marking (India stone, remove all debris!).

In addition I learned about the pace of progress which is surprisingly fast for the small blocks we were working on. Finally I became conscious of simultaneity. The object we were working on was intended to come to size, perpendicularity, and surface quality all at the same time, ending with a final stroke. Knowing when that final stroke has happened is elusive to say the least, but it is the most important thing to know - when to quit.

Along with that is recognizing you are on or not on a path to successful completion. If you are not on a path to completion you are on a path to failure, so you need a mental map of where you are and where you are going, and what the lay of the land is, and that your progress is taking you toward success. The understanding of simultaneity is useful in this analysis.

Not included was actual scraping processes such as frosting and working with dovetails and round work such as bearing surfaces.

I'd like to be as prepared as possible before taking the class and I've been working up to this for some time now. You refer to this as a decent framework and I agree. I see this video as yet another step toward acquiring a decent framework.

I just finished reading Robert Wade's The Art Of Hand Scraping .pdf. The pictures in the .pdf I was sent are horrible and I can't really make out what he's doing in many cases. One thing I wanted to ask was that Robert Wade seems to imply that you need to sharpen the carbide blades you use for scraping with coolant flowing. I don't see this happening in the pictures I've seen posted of homemade carbide grinders.

I can't see really starting to scrape without taking a class. In my view you can read and watch videos all you want but without the actual hands on practice you can't really develop the skills of sharpening and honing a carbide blade, "the rock test", step cutting, properly holding a straight edge, etc. without apprenticing under as skilled mentor so as not to pick up lots of bad habits.

dp
05-25-2009, 06:15 PM
One thing you can do is practice on a piece of sacrificial iron. If you have an old rusty faceplate around, give it a try. You can't hurt it but you can surely learn from it. Or get a starter piece of iron and make a 1-2-3 block. That is what Addy's class does, in fact. As you go you can ask questions and take notes of your scraping experience. Had I known then what I know now I'd not have hesitated to try a hand at scraping. Every stroke is a teacher.

In our class we used diamond paste to sharpen the tools against a spinning flat iron faceplate on a motor shaft. There is clearly a better way to put an edge on a tool but that does a good enough job for the classroom.

In the case of my own scraper it was sharp enough, but at a very magnified view it was ragged and that showed up in the scraping. Visible to me only with high resolution photographs, I should add. I doubt there's anything about scraping that requires unusual methods of tool sharpening, but in the case of any finish work you want the best edge possible. Getting that is already widely understood by anyone who grinds their own tools.

lazlo
05-25-2009, 07:55 PM
I can't see really starting to scrape without taking a class. In my view you can read and watch videos all you want but without the actual hands on practice you can't really develop the skills of sharpening and honing a carbide blade, "the rock test", step cutting, properly holding a straight edge, etc. without apprenticing under as skilled mentor so as not to pick up lots of bad habits.

Stephen Thomas, the co-instructor for the first scraping class, is a self-taught HSM'er.

This is the table of a mortiser he recently posted on PM. Amazing!

http://www.metalworking.com/dropbox/smt_mortisescrape7.jpg

Just get a piece of cast iron, make a scraper, and start scraping -- the process is incredibly simple. It just takes lots and lots of practice :)

Neil Jones
05-25-2009, 08:09 PM
In our class we used diamond paste to sharpen the tools against a spinning flat iron faceplate on a motor shaft. There is clearly a better way to put an edge on a tool but that does a good enough job for the classroom.

In the case of my own scraper it was sharp enough, but at a very magnified view it was ragged and that showed up in the scraping. Visible to me only with high resolution photographs, I should add. I doubt there's anything about scraping that requires unusual methods of tool sharpening, but in the case of any finish work you want the best edge possible. Getting that is already widely understood by anyone who grinds their own tools.

Is there a reason that the Baldor 6" carbide tool grinder with two diamond wheels, 280 grit for roughing and shaping a new blade and 400 grit for finishing and sharpening with it's gravity feed wet system wouldn't be ideal? Wade suggests 80% kerosene and 20% light machine oil. The Baldor 6" carbide tool grinder spins at 3600 RPM.

lazlo
05-25-2009, 08:16 PM
Is there a reason that the Baldor 6" carbide tool grinder with two diamond wheels, 280 grit for roughing and shaping a new blade and 400 grit for finishing and sharpening

For initial shaping that's great, but for sharpening a carbide scraping blade, you hone it with 9 to 12 micron diamond lapping compound.

dp
05-25-2009, 08:48 PM
Lazlo is correct - the grinder gets you to the shape, the diamond lap puts a nice finish on it, and I think a diamond sharpening steel can be used to refresh or strop the edge to remove any difficult to see irregularities thrown up by the grinder.

SteveF
05-25-2009, 09:22 PM
I believe the mirror rides up against a straight edge. I think the autocollimator stays on the table. I'll check this as soon as I get a chance.

That could be. I was there about 4 years ago and the old memory ain't what it used to be.

Steve.

Neil Jones
05-25-2009, 09:22 PM
So Dennis and Lazlo you are saying you need both a carbide grinder (with rough and finish diamond wheels) and you also need to make a diamond lap tool?

I am struggling to grasp how the "rock test" that Robert Wade refers to works.
How are you suppose slide the part you are scraping and that has the red lead on it if the part you are scraping is convex, concave or flat so you don't get false readings?

Wade says that both red lead and bearing blue need to be prepared and must sit for a week. What a pain if you run out. You have to wait a week to scrape again? What's the shelf life of this stuff once prepared and aged for scraping?

Neil Jones
05-25-2009, 09:24 PM
That could be. I was there about 4 years ago and the old memory ain't what it used to be.

Steve.

Steve I saw it done at our shop 2 weeks ago and I can't remember. Feel better now? :D

SteveF
05-25-2009, 09:25 PM
I don't know, what were we talking about? :D

Steve.

Neil Jones
05-25-2009, 09:30 PM
I don't know, what were we talking about? :D

Steve.

You win, Steve. Too funny. That made my day. :D

Greg Q
05-25-2009, 09:40 PM
Wade says that both red lead and bearing blue need to be prepared and must sit for a week. What a pain if you run out. You have to wait a week to scrape again? What's the shelf life of this stuff once prepared and aged for scraping?

I think Wade is referring to mixed pigment in oil type spotting compound. Canode is available from Dapra (and probably others). It is ready to go and water soluble which makes clean up much easier. Otherwise there is the traditional blue spotting medium available from Dykem and Permatex and I'm sure others. Oil based and a bit messier.

Red lead is I think banned owing to its heavy metal poisoning effects.

One more advantage of Canode brand is that it also comes in Red and Yellow (although perversely its all called blue spotting medium-go figure). Having contrasting colours available is handy sometimes, and I find the red easier to see than the blue in any event.

Greg

Neil Jones
05-25-2009, 10:01 PM
Thank you, Greg! I'll call Dapra in regards to Canode.

dp
05-25-2009, 11:08 PM
I bought Prussian blue at the art store where I bought my brayer. Also bought yellow and red - all oil-based. They work great and clean up easily. Many of the methods, tools, and supplies used by factories can be duplicated more cheaply by the HSM with good results using off the shelf supplies from a good art supply store and fortunately, Seattle has several. Once you start marking your work it will become obvious what works and what doesn't.

Greg Q
05-25-2009, 11:22 PM
Dennis is probably right in that regard, but I am biased by high prices for everything here. At the local art supply a few tubes of paint cost about as much as two bottles of Canode shipped from the US. (I don't have a smiley to convey the frustration of that situation).

Spotting medium is just an evenly milled pigment in a non-drying carrier. I am still working on the zen issues of scraping and didn't want to get into more variables quite yet.

Speaking of variables, I bought a granite surface plate at an auction that included several obviously well cherished machines. I still wanted to get the plate certified rather than assume its flatness. The local calibration shop wants $600 for this service. How much are people paying for those Talyvels?

Greg.

lazlo
05-26-2009, 12:10 AM
Dennis is probably right in that regard, but I am biased by high prices for everything here. At the local art supply a few tubes of paint cost about as much as two bottles of Canode shipped from the US.

The big advantage of Canode is that it's water based, so it washes off. With Prussian blue, you have to wait for your skin to wear off :)

By the way, red lead is widely available at chemical supply houses. I don't use it because I have little ones in the house that occasionally make a break for the shop.

The Artful Bodger
05-26-2009, 02:23 AM
I am wondering about the use of expensive instruments to test these granite 'flats'.

Amateur telescope makers have very precise testing procedure down to optical wavelengths using quite simple equipment. Of course the surface of the granite would have to be reflective, is it?

Greg Q
05-26-2009, 03:11 AM
No, granite plates are very dull-I guess from the lapping process. Look as flat as they are. I sure would love to discover a way to do this check with the tools that I already own (straightedges, 199 level, tenths indicator, surface gage.)

Greg

SteveF
05-26-2009, 03:31 AM
This method might be worth a try.

http://www.cnccookbook.com/MTLaserMetrology.htm

Steve.

Greg Q
05-26-2009, 04:01 AM
I tried that laser test on a woodworking machine and the results broadly agree with the straightedge/feeler gauge method. I couldn't get better than +/- 0.001 or so resolution. The other problem with the surface plates is that all of these are first order tools and should be better than 0.0001.

Since the accuracy rule of thumb is 10:1 for metrology steps, the result of scraping a spotting tool from the markings of a 0001" flat reference is a tool not quite that good. This will yield a scraping result on the machine of even less accuracy.

50 millionths (0.00005") is the number that you see on the certs for these tools, which is (I guess) the economical limit of accuracy/precision/repeatability for lapping these stones. Moore Special Tool Company (Jig Bores and grinders etc) built their reputation for unreal accuracy on their reference tools. They had hand scraped cast plates flat to better than 20 millionths

I was thinking of the results of scraping in my cross slide and saddle and gib strips. To get the celebrated results that are apparently possible I wanted not to assume that my plate was flat.

(I might however be making the new guy mistake of chasing crazy small numbers for no real world gain.-but I don't think so given what I've read and been shown)

Greg

Paul Alciatore
05-26-2009, 04:20 AM
I am wondering about the use of expensive instruments to test these granite 'flats'.

Amateur telescope makers have very precise testing procedure down to optical wavelengths using quite simple equipment. Of course the surface of the granite would have to be reflective, is it?

Yes, amateur telescope makers use a device called a knife edge tester. Trouble is, it does not work directly on flat surfaces as the light starts at a "point" source and is reflected back to a "point" image which the knife edge intercepts. I have a home made knife edge tester that cost all of about $25 and it can test spherical and parabolic mirrors to 1/4 wave or better. That's microns folks.

However, in order to test a flat with it, you would need a mirror, preferably spherical, that is already as accurate or better than the accuracy you want to test to AND that is as large as the flat itself. That's why the commercial, optical test instruments cost so much. I guess you could test in smaller sections and map them together, but you still need a good mirror. And the setup would be challenging. If the flat is horizontal, then the light beam would be vertical or perhaps at a 45 degree angle between horizontal and vertical. The known good mirror will need to be mounted so that it would not be distorted. Thermal effects are hell and it pays to set things up and then come back in 12 to 24 hours. Of course, you can not touch the mirror or the flat as the heat from your hands will distort things.

But it could be done for under $200, including grinding and aluminizing a 6" or 8" mirror, I would think. A 12" mirror would be better but more expensive. If you buy the mirror ready made, it could cost hundreds or even thousands.

An optical flat and monochromatic light source might be a better way.

DMF_TomB
05-26-2009, 04:20 PM
Most high precision levels at 0.02mm/m can check flatness.

A 100mm long level will show 1 line bubble movement if it dips 0.002mm per 100mm.

http://www.leveldevelopments.com/engineers-block-levels.htm

clutch
05-26-2009, 05:57 PM
I'm guessing you bought a big granite surface plate and it's not flat? You might be better off with a Talyvel -- an exquisitely sensitive electronic level (really a gauge head with a pendulum sensor). With a Talyvel and a lot of patience, you could record a 2D map of the flatness of the surface plate.

http://reliabletools.cachefly.net/itemimages/JanuaryV09/v21d.jpg

Mine has an analog display head, but the new ones have a digital interface with software that maps the 2D flatness of the surface (you can do this by hand with the analog head, of course):

http://www.spectrum-metrology.co.uk/images/grid.gif

We had a gent in recently, he had two Mahr units
http://www.mahr.de/index.php?NodeID=10257

He took readings across and diagonally using two sets of those feeding a computer that mapped the granites and then he executed a lapping schedule followed by cleaning and rechecking.

Clutch