A flat surface

astjp2

TJ Guru
Supporting Member
Joined
Aug 22, 2018
Messages
6,190
Location
Utah and Alaska
A Full Plate

Measurement starts from a reference point or plane. Linear measurement begins at a reference point and ends at the measured point. For angular measurements the reference point must be a plane.

Since a plane is defined by three points, it can be said that a plane has three reference points that define its flatness. In reality these points have only length and width. A three legged stool defines a true plane by the points of contact. A four legged stool with four equal length legs will reveal the flatness of that plane. To make a plane usable, it must have thickness to support both the measuring tool and the part being measured. Planes that provide for useful work are called surface plates.

This reference surface must have known accuracy before it can become the common plane or starting point for angular measurements, height gages and gaging accessories. The surface plate is as essential for positional accuracy as the length standard is for linear accuracy.

History
Ancient Egyptian pyramids and buildings bear testimony to man’s early understanding of flatness as the basic reference plane from which dimensionally accurate heights, angles and squares could be obtained.

The length of the sides of the great pyramid varies no more than 1/20 of one percent from the mean length of 9,069.45 in. This is remarkable accuracy, considering overall size and the number of blocks of stone involved.

It is believed that the Egyptians, in leveling the foundations for their pyramids and buildings, flooded the area with water using the surface of the water as a reference from which to determine overburden removal or fill-in to make the land flat and parallel to the surface of the water on top of it. Nature does not provide a flat surface in the raw. Even the surface of a body of water will tend to be curved by the forces of gravity and atmosphere as the length and width increases. Flatness can not be taken for granted. Remember that in ancient Egypt, as opposed to today, the earth was flat!

In an effort to produce flatter surface plates, we are confronted with devising the best method to use on a material that has properties that will preserve the flatness produced.

In the evolution of methods, gravity has been used to fix a flat surface on material that solidifies after being in a fluid state. Among these are cement, glass, ceramics, cast iron, and steel. Each has properties that make them unreliable for extended use. They lack stability, resistance to wear and corrosion or defy ultra fine surface finishing obtaining greater flatness.

The first machining method that could produce duplicated parallel surfaces on metal was Richard Robert’s invention of the planner in 1817. Today precision surface grinders and lapping machines do the production work in flattening a surface to known accuracy.

Henry Maudslay, who invented the first screw-cutting lathe in 1797, is credited for being the first to produce flatter surfaces by hand filing and lapping with abrasive particles. He produced master-reference surface plates by working three plates against each other.

By 1874 Sir Joseph Whitworth had introduced hand scraping instead of abrading to improve the three plate method of producing a flat surface.

Hand scraping permits greater control of material removal from high spots on the plate’s surface. In the Whitworth method for achieving flat surfaces, three plates were spotted against each other in alternating pairs. The plates were made of cast iron having a ribbed construction to provide rigidity without excessive weight. After normalizing in an outside atmosphere for a year to relieve stresses, the plates were machine finished to the accuracy of the machine tool used.

The following is a description of the three plate method for producing flat surfaces.

The three plates are marked 1, 2, 3, for tracking in the process and scraping operation commences. High spots are seen when bluing from one plate is transferred to the surface of the other plate. Six steps are involved in scraping these three plates.

Step 1 – Plates #1 and #2 are scraped alternately one to the other until they conform to each other.

Step 2 – Plate #1 is now the control plate to which plate #3 is scraped.
Step 3 – Plate #2 and #3 are alternately scraped one to the other until they conform to each other.
Step 4 – Plate #2 is the control plate to which plate #1 is scraped.
Step 5 – Plates #1 and #3 are alternately scraped one to the other until they conform to each other.
Step 6 – Plate #3 is the control plate to which plate #2 is scraped.

Now each of the three plates conforms to each other in relative flatness. To increase the degree of flatness, the above steps must be repeated until the desired flatness is attained.




Demands for greater accuracy during World War II caused extensive research for better surface plate material, more accurate machine tools and instruments to measure flatness.



To solve the material deficiencies of cast iron and steel plates, manufacturers rediscovered what the Egyptians knew 5,000 years ago that black granite was ideal material. Black granite is a form of original rock produced by nature. Composed of gabbros and basalt (or diorites) it is nature’s most enduring material. It is a material that combines wear, shock and chemical resistance; plus hardness and stability, together with machinability, to permit low-cost manufacturing into surface plates.



Plate Accuracy

Cast iron and steel, as a material for surface plates, has been outmoded by granite. Granite surface plates having flatness within millionths of an inch can be produced at far less cost than metal plates. In preserving this flatness, granite plates do not corrode or rust. They are not subject to contact interference because they do not burr or gall or crater. To assure greater measurement accuracy, granite surface plates are nonmagnetic and have exceptional thermal stability. To resist wear, granite plates are harder than metal plates. They are easier to clean and the non-reflective surface is easy on the eyes.



Since perfect flatness is unobtainable, the problem is solved by knowing how flat a surface plate is. For this purpose, manufactures calibrate the flatness of their surface plates by using electronic indicators capable of measuring to .000010 in. or with autocollimator. These measuring tools permit precise reading of critical areas at critical points on the surface of a plate. By plotting these readings as a graph, the user knows where and how much the surface plate is out of flatness.



To determine overall area flatness even more precisely, an absolute standard, the wave length of light is used in the form of a laser interferometric surface contour projector. This measuring tool permits viewing and photographing areas up to five inches wide by forty-eight inches long at one time. The photograph pictures the surface as a series of interference bands which can be read to accuracy within a few millionths. Deviations from straight, horizontal or vertical bands denote high or low points on the surface plate.



Surface plates are made to three grades of accuracy. Grade AA is used for laboratory work where greatest accuracy is required. These plates are made to a flatness tolerance of ±25 millionths of an inch per two foot square area. Grade A is an inspection quality grade surface plate having a flatness tolerance of ±50 millionths per two foot square area. Grade B is a surface plate having a flatness tolerance of ±100 millionths per two square foot area. It is used in the shop for tool room work.



Accuracy of surface plates is given as a bilateral tolerance. Bilaterally is the tolerance above and below a perfectly flat plane. This accuracy is expressed at “No point on the work surface shall vary from a mean plane thereof by more than the amount specified. Measurements of accuracy should be not be made nearer the short and long edges than 3% of the width and length or in no case closer than ½ in. from edge.”







TABLE: SIZE AND ACCURACY



Size​
Work Surface Accuracy​
Work Surface


Width Length
In. In.
Grade A
Thickness
(Minimum)
In.​
Grade B
Thickness
(Minimum)
In.​
Grade A
Tolerance
Plus or Minus
In.​
Grade B
Tolerance
Plus or Minus
In.​
3 ½ 4
1​
1​
.000025.0001
8 12
3​
3​
.000025.0001
12 12
3​
3​
.000025.0001
12 18
4​
4​
.000025.0001
18 18
4​
4​
.000025.0001
24 24
4​
4​
.000025.0001
24 36
6​
5​
.000025.0001
24 48
8​
6​
.000050.00015
36 48
8​
6​
.000075.00015
36 72
12​
10​
.00015.0003
48 96
14​
12​
.0002.0004
48 144
24​
20​
.0004.001


Surface Plate Design

Granite surface plate design has followed that of metal surface plates in having ledges for clamps and surfaces that are square or rectangular in shape. Today granite surface plates are made in any shape desired. Standard sizes range from the small 8 x 12 in. to those 72 x 144 in. To obtain a surface area of much larger size several plates are linked together by proper alignment to each other.



The geometry of many parts, fixtures and measuring instruments require their being clamped to the surface plate during measuring operation. To provide this facility, surface plates were originally furnished with two or four ledges. These ledges were formed by undercutting the sides with diamond saws. Ledge thickness was one-half the thickness of the plate itself so that the ledge would support whatever was clamped to the edge of the plate by means of a C-clamp. Unfortunately, when manufactures of granite plates originally began making them, they related plate accuracies to the number of ledges. For example, 4-ledge plates were made with laboratory grade, 2-ledge plates with inspection grade and plates without ledges to shop-grade accuracy. Users for years purchased the more expensive 2- and 4- ledged plates without need for the ledges because they required a higher degree of accuracy than shop grade.



Ledges are not only expensive, but a great cause for inaccuracy. Experimentation and research reveal that no-ledge plates retain their accuracy better than ledged plates. Deflection caused by thermal expansion of unequal lengths of top and bottom portion of ledged surface plates can result in inaccuracies from deflection. Even a bulky and thick ledge will deflect under heavy load or clamping pressures and cause inaccuracies. Clamping on ledges restricts the gaging area to edges of the plate which causes excessive, concentrated wear. Based on these findings, no-ledge plates are now being manufactured to the highest degree of accuracy attainable. Users are now benefiting from no-ledge plates in several ways. First, they can purchase the plates at a lower price because of the ledge elimination. Second, the plate can be made more accurately by the manufacture and the accuracy is retained longer. Third, the use of inserts for clamping distributes the wear more uniformly over the surface of the plate and provides greater versatility and usefulness.



Surface Plate Maintenance

Good housekeeping is a basic practice essential to taking precision measurements. The accuracy of the measurement being made is in direct proportion to the degree of cleanliness in the shop or metrology lab.



Temperature and dirt will displace measurement accuracy. The surface plate presents the biggest target for contaminants. Its broad surface is a natural for collecting dust, dirt and oil. For this reason it must be cleaned before taking any measurements. Use a non-water base cleaner to spray or cover the plate surface. Rub small areas at a time with a damp cloth or disposable wiper. As they say turn the cloth frequently as the plate is wiped clean. Using a non-water base cleaner will help seal the fine silken texture of a granite plate. The sealing process prevents moisture absorption which can catalyze corrosion of parts or gages left on a plate overnight or longer.



Utilize the full surface of a plate so that wear is distributed and not concentrated in one area. One of the most common abuses of surface plates is a stool or chair. Placement of sitting appendages will concentrate wear in one area of a plate, human nature being what it is.



Surface plate covers should be used religiously for protection. Covers should hold their shape and not absorb moisture like wooden covers. Soft covers can cushion shock and prevent damage by an accidental dropping.



The condition of the accurate plane is an integral factor in the measurement being made.
 
Very interesting article, thanks for posting. I have 2 small granite surface plates, one made in China and the other a Starrett. The one made in China is not flat on the bottom side!!!! When placed on any reasonably flat surface (mill table, for instance) it will tip back and forth slightly on the diagonal. The Chinese plate came with a small round felt "foot" attached to the bottom on each corner I guess to hide the tipping. You get what you pay for!!
 
my 6'x4'x10" surface plate sits on 3 rubber pads, the nist/astm/anis spec states the position of the pads based on a percentage from left to right, what side faces north, etc. The chinese ones can be good, just need a basic inspection to determine the flatness/grade. Tim
 
my 6'x4'x10" surface plate sits on 3 rubber pads, the nist/astm/anis spec states the position of the pads based on a percentage from left to right, what side faces north, etc. The chinese ones can be good, just need a basic inspection to determine the flatness/grade. Tim
If you like this type stuff, the book "The History of the Screw" is a must read . It's the story of how man achieved precision machine work basically.
 
I have machine tool reconditioning, CNC machine tool theory and design, machinist handbook, metal spinning, know your lathe, 2 different testing of machine tool manuals, a mathematics for mechanical accuracy, foundations of mechanical accuracy if you need them, all on PDF if you want them. tim
 
Wow! Nice collection of reference materials! I'm currently restoring my dads 1948 Atlas 10F lathe. It sat for around 40 years without being used (he had purchased a larger, better machine) and was absolutely black with filth and crud. No rust though. It's the lathe that I first learned to work on so it's got quite a bit of sentimental value. I'm in the assembly stage now, hope to have it up and running in a week or so. I have the original owners manual for it. It has a surprising amount of operating info in it. Pretty clearly written for a beginner to use.
 
  • Like
Reactions: jgaz
Did you scrape anything in or just clean and paint?? Join the hobbymachinist.com, there are a butt load of references there that you will find helpful. Tim
 
I think that the bed ways are OK but sadly, none of the ways are hardened. I can't feel anything with my fingernail but I won't know for sure until I put an indicator on it. The crossfeed and compound ways look new. I've never scraped in ways before but I think that I know someone that can supervise me if it needs to be done. I took the lathe apart down to the last screw and cleaned and reconditioned everything. Mine is a '48 so the spindle has tapered Timken roller bearings in it instead of the earlier bushings which is nice. Thanks for the tip about hobbymachinist, I'll check them out.
 
I think that the bed ways are OK but sadly, none of the ways are hardened. I can't feel anything with my fingernail but I won't know for sure until I put an indicator on it. The crossfeed and compound ways look new. I've never scraped in ways before but I think that I know someone that can supervise me if it needs to be done. I took the lathe apart down to the last screw and cleaned and reconditioned everything. Mine is a '48 so the spindle has tapered Timken roller bearings in it instead of the earlier bushings which is nice. Thanks for the tip about hobbymachinist, I'll check them out.
Be sure to post some pictures after you get your Atlas resto done.

Bringing an old tool back from the dead is an interesting diversion/hobby.

This is a picture taken during my restoration of a 1943 Power Craft lathe sold thru Montgomery Wards. It’s a rebadged Logan lathe. This was a common practice for years. Sears rebadged some Atlas lathes as Craftsman brand.
76746