Grinding vs. Scraping

[Paul Alciatore]

Some years ago I made a small telescope mirror. GROUND it from scratch. Now that I am more active in machinning metal, I have read over and over that precision flats and machine ways are made by SCRAPING. I am curious about the relative merits and problems of each of these processes.

Now I am not talking about grinding with a mega bucks grinding machine. I am talking about grinding with a powdered abrasive, perhaps in an oil or water mix, applied between two of the three surfaces and then they are ground together - by hand. Successively finer grades of abrasive would produce a finer and finer surface.

Obviously, there must be a reason why machinists scrape instead of grinding. But in theory, three flats could be made by alternatively grinding them together with progressively finer grades of abrasive just as well as they can be scraped to flat.

In making my telescope mirror, I started with a coarse abrasive, perhaps 80 grit and worked my way down. Since I was making a spherical surface, only two pieces of glass were needed, instead of three. And I guess that scraping would not work with glass or pyrex because they are too hard. And yes, I know glass can be cut with diamond tools.

But why is scraping prefered with metal? Would grinding with a loose abrasive between the surfaces not work? Perhaps mixed with some oil. Has anyone ever done it this way? I would love to hear some discussion on this.

Paul A.

[brunneng]

Basically, scraping is used because it produces bearing points which oil can move around. This aids lubrication. You scrape to a minimum number of high bearing points per sq inch based upon the use it will be put to.

Grinding produces a polished surface that squeezes the oil out from between the surfaces. This impedes lubrication. Think of lapping.

That's what I've read / experienced.

I've been doing telescope grinding too. All the grits above the cerium oxide pitch lap actually grind by pitting the glass while rolling around. They just produce finer and finer pits. The oxide lap actually scrapes.

Each process could and is used on metal - it just depends on the desired characteristics of the surface left.

there's my two cents (worth 1-1/2)
kevin

[mochinist]

While I have not personally done any scraping, I have watched it many times at work. I think scraping is more effective for machine tools because of the fact that you are scraping the machine surfaces not only to be flat but also to be square and parralel to the other surfaces. I think this would be alot harder to do with the method you described. I also think it would be killer on your back, you would pretty much need a piece of metal the same size as the surface you were trying to get flat. Another thing that brunneng touched on was how the scraped surface produces bearing points so that the oil can flow. If two surfaces are highly flat and polished they will wring together (think gage blocks), this would not be very good for your arms when you are trying to turn the handles on your mill. Hope I didnt say anything dumb and I am sure others will have more to add. My one cent.

[J Tiers]

<font face="Verdana, Arial" size="2">Originally posted by Paul Alciatore:


But in theory, three flats could be made by alternatively grinding them together with progressively finer grades of abrasive just as well as they can be scraped to flat.


Paul A.</font>

Which would be one way granite flats are generated.
Probably the best way, and the only way which MUST converge on flat.

The thought of doing that on every surface, and keeping to alinement, is daunting. I'd rather scrape, and that is hard enough.

[Paul Alciatore]

Sounds scraping is prefered for machine surfaces for a combination of reasons including getting a rougher surface finish, faster, smaller "tools", and better control of the geometry.

How about the abrasive itself. Would metal tend to retain some grains of abrasive more so than harder glass or pyrex? Therefore causing future wear?

Paul A.

[Jpfalt]

The free abrasive lapping process you are talking about for working granite or glass has one important feature. The harder material is what gets cut during lapping. The abrasive embeds in the softer material and cuts the harder material. So to make a cast iron flat using lapping compound you would need to use a lap that is significantly softer than cast iron but still capable of holding it's dimensional stability. Something like asphalt or pitch on a backing plate could work, but would deform and flow over time. Grinding telescope lenses and optical flats depends on the pitch flow to conform to the shape and requires specific movement patterns to affect the overall geometry of the optical surface. It's somewhat difficult to target material removal at specific high spots using lapping.

Scraping is very specific in the location of material removal and by the nature of a surface plate you are talking tenths of thousandths for final geometry rather than the angstroms and wavelengths used to measure optical surfaces.

I have made both cast iron and granite flats and straight edges as well as telescope mirrors and found that plated diamond hones were of use on both the cast iron and granite, but largely worthless for the mirror except for the initial roughing in to get the initial cride form of the mirror. Free abrasive lapping was the only way to go on the mirror and was of some use for area touch ups on the granite, but useless on the cast iron. Scraping was useless on anything but the cast iron. I had given some thought to buying a polycrystaline diamond lathe insert to turn rough mirror blanks, but haven't worked up the financial nerve to give it a shot yet.

[DR]

"Basically, scraping is used because it produces bearing points which oil can move around. This aids lubrication. You scrape to a minimum number of high bearing points per sq inch based upon the use it will be put to."

This is pretty much my understanding too.

Where do the precision ground box ways on newer machines tools fit into this scenario? They're hardened and ground and don't have the characteristic scraping marks to hold oil.

[Forrest Addy]

I gotta make some pompous statements.

We have confusion in terms. The process by which lensmakers usually generate telescope mirrors is "loose abrasive lapping". While lensmakers might call it "grinding" among themselves thansliterating their particular use of "grinding" as a term to another manufacturing arenas can cause confusion.

In metalworking "grinding" is accomplished by means of rigid bonded abrasive wheels rotating at speed acting on the work where the work is abraded by the individual abrasive grains acting as cutting tools.

"Precision grinding" is accomplished by precision shaped (dressed) rotating grinding wheels where the precision is imparted to the work is obtained moving wheel and/or work in precisely guided paths.

"Lapping" is accomplished with loose abrasive and usually a vehicle between a consumable tool and the work. Flatness or shape of the work is determined by the conjugate action between the tool and the work.

Scraping is acccomplished by use of a hand or power driven edged tool that when guided by the operator removes selected material in by reference to inidcations transferred from a reference tool. The indications or marks are transferred directly from the reference tool to high spots of the work by use of a pigmented transfer medium.

The three methods all have their advantages in acheiving high precision if worked by skilled operators. Loose abrasive lapping has some distinct disadvantages for machine tool fitting and entrained abarasive in the metal destined to be bearing surfaces is only one of them.

Precison ground assembly joints are no better or worse than hand scraped.

Precision ground way bearings can be very satsfactory provided the design of the bearings abates the dynamic film forming properties of oil.

Scraped way bearings are generally considered to be superior to precision ground because the interrupted surfaces so generated feature lower "stiction." They have been demonstrated over generations to be cooler running thanks to lower fluid shear and the "point to point" transfer of load as opposed to large areas of intimate contact. There is also less vertical "float" because dynamic film forming is abated by the scraped surface when in motion.

[This message has been edited by Forrest Addy (edited 01-20-2005).]

[Paul Alciatore]

Thanks to all. I think my curiosity has been satisfied, at least for now. And I have learned things to think about.

Forrest, I am one who believes in using the correct, precise terminology. So I do appreciate the lesson. It is interesting how the same term can have really different meanings in different fields. I was aware of the differences between grinding as used in the machinning world and the process used in making optics. That's why I explained what I meant. I now know what the machinning world calls it. Thanks.

It seems each process has it's advantages and uses. I just needed to understand each better. One of these days I will have to try some scraping. Sounds like "fun".

Paul A.

[MJMIKE]

I only know about machine tools - if you were to try to grind a slide flat on a surface grinder you have to hold on to it and despite your best efforts at clamping it down or holding it evenly it will bend, twist, or deform some minute amount that will result in it not being flat when it is released after grinding. Hand scraping is done with the slide laying on a surface in a relaxed condition and when you finally get is scraped flat is stays flat. The little scallops in the surface from a convex scraper tip aid in holding the lubrication. A lot of new machine tools have various forms of plastic sliding surgace in contact with the ground ways. The plastic is sometimes still scraped just as the cast iron was but the finished fit does not have to be quite as exact since the plastic will compress or deflect a little. Plastic does not tend to gall like cast iron. A scraped cast iron slide has to have enough load bearing high spots to lower the pounds per square inch load on the slide below where it will gall. If all of the weight is on too few spots it will gall.

------------------
Mike