[mxtras]

WOW! I missed a lot while I was out! I feel left out.... :violin:

I guess the only thing that has not been brought up that really should be mentioned is that 'lapping' as described in this thread will not create a flat surface.

Lapping will match surfaces by removing high spots, but it can not magically make a surface flat. It can, however, magically destroy an accurate surface. Of course, so can scraping!

Lapping is a blending and/or matching process which indicates that the parts' irregularities will be generalized into each other - assuming both are made from similar materials at similar harnesses (again – as it applies to this thread). If one surface is flat and the other has a slight cup for example, when you are done match lapping you will have lessened the cup and ruined the flat surface. You will also have likely changed the running alignment of the component but if one component was cupped to begin with, you likely gained something – they will likely run smoother with slightly less of the greatest prior error and slightly more of the least prior error…does that make sense?

As mentioned before - scraping forces you to compare to a known surface to gauge and guide the result. So if what you are looking for is flat, then you should precision grind then scrape (if it’s even necessary to scrape). If you are looking to improve the running fit and nothing else, then lap. Lapping will offer a surface which looks more uniform and smooth when compared to hand scraping. No argument there.

The end to this mystery would be to blue out a set of components against a known good master and document the results with photos and measurements. Then lap them together as discussed and re-blue and re-measure the same points and compare the results. Has anybody done this?

Scott

[Mcgyver]

Scott, the flatness was brought up, but you had to see past my verbosity to pick up on it. I agree with most of your post, except the grinding part. scraping is superior because it holds the oil better, its my understanding that the high class machines that have one surface ground (hardened) will have the other scraped. Also, a scraped surface (after you knock off the high spots with the ground down piece of file trick) will be very smooth - the rough, frosted look is optical, although as mention you don't want a way perfectly smooth.

This is an interesting thread. I'm no way bothered or testy about the subject, heck fix your machine with a great big sledge hammer if it makes you happy . What surprised me was that scraping vs lapping in this context was considered a debate, like what colour to paint the living room. My background (in the hobby, i'm a finance guy by trade) is traditional machining, and if you posted something about lapping ways in one of the traditional machining forums, I don't think you'd see much diversion of opinion. I may try it for fun, get the asbestos suit on!

Anyway, its all good, just trying to help those that are interested

[wizard]

I concede that scraping is not a difficult skill to learn, how to use scraping to resolve an issue of smoothing some raw machined ways is another matter entirely! As a hobbyist, enthusiast, chip sweeper or any other description you like (don't know much about machine repair), the odds of me correcting my machine's ways by scraping are not good without considerable effort on my part to learn the methodology. Learning how to do this on my machine is not an acceptable option for me.

I think the argument that considerable effort is required to scrap a surface is the one thing I object to in this thread. There are certainly things to learn but it isn't that bad.

I'm sure if you have seen this done in person the light bulbs would shine brightly in your head. Some things are just difficult to describe in text.

You could compare this dilemma to the repair of your auto. If you had some problems with your motor that required repairs there could be a low cost, quick fix we can do it today option and the expensive (skilled labor) best way, 2 week repair option. Not everyone will want the expensive option (most of the pros probably would). If the low cost option was done adequately, the repair may serve the individual well, or the individual may need extensive repairs later. If the low cost or expensive repairs were done poorly, the extensive repair later scenario would prevail. If a DIY approach is taken, the results will be all over the place.

Well if we are to talk about autos, here my perspective is that the lapping of ways is a lot like throwing lapping compound into an engines oil supply and expecting (hopeing) that it will fix a main bearing problem. Not many people would do that but yet we are doing the same thing with our machines?

For many, smoothing or polishing the ways with abrasive compounds is the low cost, quick fix we can do today option to resolve the problem. If done adequately to rough surfaces that were otherwise machined acurately, it may serve the individual well.

No the whole reason for my comments here is that it won't serve the individual well or atleast his machine. There are ways to improve those surfaces but running them together with lapping compound is not one of them.

For some of us, the reliability of our engine is our livelihood and a quick fix is inappropriate.
For some of us it’s just a hobby, and if it blows up, we’ll buy another.
Some of us still don’t know or want to know the difference.

Those that don't know or don't want to know are not likely to even be concerned. Those that want to improve their machine for any reason are likely to want an approach that does not have a negative on the machines performance.

Let’s call the quick fix “smoothing” or “polishing” or something else so everyone can get back to work.

Not so fast i have the whole week off :banana:

I still have a huge problem describing this activity as even a quick fix. Even when a quick fix is called for this isn't how it is done.

Twenty five some odd years ago I had the good fortune to work with a machine tool builder that frankly knew more than I'm ever likely to know about machine tool maintenance. He could take a machine apart and tell you what the last guy that worked on the machine did with it and where all the short cuts where taken. It was a very enlightening two years for me. Still with everything pointed out to me, the idea of running lapping compound between bearing surfaces was to abhorrent for even production repairs. There are things done to keep a machine running and things done to rebuild it properly, sometimes the two are at odds with each other.

Thanks

Dave

[mxtras]

Oh, no doubt - I agree 110%, Mcgyver. I should have elaborated. I also should have read every thread carefully, but I had been away from The Zone for 5 days - I was out of practice....sorry.

During motion the ways should be running on a film of lube - not metal to metal.

Scott

[MikeAber]

It's been interesting.

[MikeAber]

Wizzard, ?
Maxtras, thanks for your help, your perspective is refreshing.
Mcgyver, thanks for the great photo essay and the book reference.

MACHINE TOOL RECONDITIONING by Edward F. Connelly Copyright 1954
This book is much more than I expected. - Priceless! -

Once you've read the first two chapters of this book (7 pages) you'll understand why all other methods are inferior to scraping and that everyone is not up to the task.

Read the book!

[ChrisJ]

Hi Mike,

I did a quick search on Amazon and it did not come up with anything. Do you know if this book is still in print?

Chris

[nikolatesla20]

Geez, copyright 1954? By now someone should be able to just scan and upload that book I would think!

-niko

[nikolatesla20]

I do have one question though about making parts straight

I've read the notes on how they originally made straightedges and flat surfaces by using three surfaces all mated together in round-robin style. It makes sense of course.

But my belief has always been that yes, if you took two edges of something and worked them together, the friction would over time cause those two edges to completely match up. I'm not talking about machine accuracy here; I'm talking about just trying to make one edge straight along with another. For example, say you just wanted to make a straight edge on something. Say that one edge was the only edge you cared about (you aren't trying to align it with anything else on the part, maybe say you just want a straight edge to use for marking or something). Why couldn't you simply take two pieces of that material, and basically mechanically fasten them so they rub together using a reciporicating mechanism, and then put in some lapping compound? Over time any high spots between the two surfaces would be "sanded" down until both surfaces almost had full contact. Hence, making the edges perfectly straight and flat. Of course they would not be flat to some reference point, but only flat in their own right (And they would be flat to each other).

I keep thinking this should work fine, but obviously in the past they chose to use three surfaces.

What am I missing?

I know that ball bearing making machines use a technique similar to this; they have two rotating grinder discs, and the balls drop down in between them, and then rotate and get ground at the same time. Then they move around because a disc inside forces them side to side (or something like that) so eventually the whole outside surface it round and accurate to the width of the grinding discs.

-niko

[MikeAber]

I purchased a used book (9th edition 1974) in very good condition on Amazon.com.

No fancy color pictures, excellent illustrations and descriptions of everything in terms I can understand. This is a repair manual that assumes you are an apprentice. Very technical, yet not intimidating - does that make sense?
Don't bother buying any other book on scraping or machine tool repair, as Mcgyver said this is the tome.
Powell’s has new reprints of this book.
http://www.powells.com/cgi-bin/bibli...4-1114266612-0

[Mcgyver]

niko, I think what you are missing that in generating a flat or square, A, B & C are compared to each other, not worn against each other. Guy Lautard in one of the bedside readers describes the process. Lets say you could wear two pieces away until there was complete contact, there would be no reason for it to be flat, ie. convex/concave.

the idea of wearing two pieces until they are a perfect mate I assume requires them to be the same length. In this case lapping would tend to remove more material from the middle as it tough to avoid there being more contact there per stroke.

for the original generation of flats or squares, the A, B & C thingy wasn't a choice that was made, its the only way to do it. Mike's probably finished that book by now and can pipe in and give the master mechanic explanation

[MikeAber]

Mike's probably finished that book by now and can pipe in and give the master mechanic explanation

OUCH!

[Geof]

An example of generating conves/concave surfaces by lapping two parts together is the preparation of mirror blanks for small astronomical telescopes When two glass discs are lapped together you generate perfectly matching convex/concave surfaces which is why astronomical telescopes could be made long before sophisticated machinery.

I think the one time you do actually lap flat things against each other in threes and generate a flat surface is the preparation of optical flats in glass. Also ball bearing manufacture where the balls are lapped between two large discs is more or less three surfaces lapping against each other. In this case the surface of the discs become flat and the balls all the same size.

[nikolatesla20]

niko, I think what you are missing that in generating a flat or square, A, B & C are compared to each other, not worn against each other. Guy Lautard in one of the bedside readers describes the process. Lets say you could wear two pieces away until there was complete contact, there would be no reason for it to be flat, ie. convex/concave.

the idea of wearing two pieces until they are a perfect mate I assume requires them to be the same length. In this case lapping would tend to remove more material from the middle as it tough to avoid there being more contact there per stroke.

for the original generation of flats or squares, the A, B & C thingy wasn't a choice that was made, its the only way to do it. Mike's probably finished that book by now and can pipe in and give the master mechanic explanation

Thanks, 'tis what I wanted to know! Yep I knew that they didn't rub the things together for A B and C I was just wondering why rubbing two parts together wouldnt make them flat.

-niko

[sblack]

I am looking seriously into the IH mill and have discovered this discussion in my research. I have read through enough of the Connelly book to understand the principles behind scraping and I have had some precision grinding done on an old lathe so I think I have some idea about what is (was) being discussed but I have a question: What is the issue with the ways on this mill? Are the ways simply milled or are they ground? If they are ground is it simply a question of knocking off some burrs? Or are there fundamental alignment problems?

While scraping in theory does not seem that hard, there is an initial investment required that could be substantial i.e. surface plate, indicators etc. If there were ways that were not parallel or planar I think it would be an intimidating thing for a guy to do on a new machine. But if the overall alignment is good and it is just a question of knocking off some high spots to get good contact I would not be too apprehensive about it.

So if anyone can accurately describe to me the real problem with the ways as they come out of the box it would be very helpful. Thanks

[wizard]

I am looking seriously into the IH mill and have discovered this discussion in my research.

The IH mill is a nice mill there is no doubt about it. I'd have to say it is the ideal machine to CNC at the low end. One shouldn't confuse it with an industrial machining center but I don't think that is an issue on these boards.

I have read through enough of the Connelly book to understand the principles behind scraping and I have had some precision grinding done on an old lathe so I think I have some idea about what is (was) being discussed but I have a question: What is the issue with the ways on this mill?

I'm not sure the ways are as much of an issue as the suggestion to lap them to increase performance or ease of axis movement. Lets face it, the mill is Chinese and built at a price point to support a cdertain market segment. It is a good machine for the money but it is not something to get excessively excited about either.

Are the ways simply milled or are they ground? If they are ground is it simply a question of knocking off some burrs? Or are there fundamental alignment problems?

Having no familiarity with the manufacture and only having a brief glipse of the ways at one of the shows I suspect that they are milled or roughly ground.

While scraping in theory does not seem that hard, there is an initial investment required that could be substantial i.e. surface plate, indicators etc. If there were ways that were not parallel or planar I think it would be an intimidating thing for a guy to do on a new machine.

Everybody has their first time. The thing with scraping is that you get better at it as you age. Well atleast until your eyesight and back give out.

But if the overall alignment is good and it is just a question of knocking off some high spots to get good contact I would not be too apprehensive about it.

I can only give an impression here and that is that overall the machine is not that bad.

So if anyone can accurately describe to me the real problem with the ways as they come out of the box it would be very helpful. Thanks

Rough - I think that about wraps it up in a nut shell. The reality is this thread got away from discussion of the ways and drifted into what was the right way to work on the ways. The reality is one needs to look at what you are buying for the money. Even with Chinese rates one can not expect to get $10000 dollars worth of machine tools new for one tenth the price,

Thanks
Dave

[MikeAber]

I owned two of the gear head mills, a Harbor Freight round column and the Industrial Hobbies Mill. Both of them had MILLED ways and accuracy of the ways was not an issue for me. The HF mill's ways were OK; the IH's ways were very rough. Both of them also have issues with flatness of the tables and many other issues compared to an industrial quality machine. I consider these machines stepping stones and for some these will be all they need. It is possible to do nice work with this level of equipment but not easily. The biggest problem is the rigidity between the column and the base. I still have the IH mill and am using it for manual operations. If I could do this all over again, I would purchase a turret head Bridgeport style mill for manual operations.

MikeAber

[jderou]

Been reading intently, good reading! I still don't understand a couple of things. Number one, how do you scrape the saddle way such that it is perfectly perpendicular to the knee/column way? I haven't heard anyone mention scraping a right angle reference tool yet. If a right angle tool does need to be scraped, how the heck do you reference it? and if I had bought a right angle reference, why wouldn't I use that for a way scraping reference instead of scraping a reference tool. And the biggest question in my mind, is why you cant grind a flat reference tool on a surface grinder? I would think this would be plenty accurate place to start.
I am preparing to scrape my BP ways, and I am trying to get as much info as possible before I start. I have wear in the center of my y-axis. I know I can get it flat, but getting it horizontal is the hard part.
Joe

[ZipSnipe]

No Doubt this is absolutely the best post on this forum.... BRavo gentlemen, Bravo

[Mcgyver]

Zip, its a good subject that doesn't seem to get much coverage on the net, maybe it's getting bumped will bring in some more perseptive from some of the pro's here. Joe those questions so staccato, not sure where one starts and one stops. here's the verbose response….

Getting things square. Lautard in one of the bedside readers gives a good description of generating a square - interesting but not something I'd considered doing so until your question of getting the knee square to the column. Highly accurate (to say a tenth) 6” squares are reasonably priced, but how do you come up with a 18 or 24 inch square - commercially it might cost as much as the restoration project (like the big starrett granite square). I’d probably give generation a shot, will it looks daunting, you are making the square out of thin stuff so it shouldn’t be that bad. Restoring that mill is both a major undertaking and the quality of the work you do on it is the foundation to accracy thereafter- so it may be worth going the distance.

I’ve not done it, but if I were facing the task I’d scrape the column with a reference and indicator, then the knee to the column. Then using a combo of reference, indicator and big 18” home made square get one of the saddles dovetails perpendicular to the column. It doesn’t sound like an easy job because the parts are heavy, but in approach it’s the same as most scraping – pick a reference, bring things into relation to the reference. Better yet, buy Connelly’s book and learn it from a pro, although as Robert Bastow said, Connelly's probably the only guy that was every able to stay awake through the entire book.

As to why you scrape a reference vs grinding, lots of reasons. Most grinders don’t have the capacity and the magnetic chuck itself distorts the work are two obvious ones. You are holding a tenth or so over maybe 24” or longer, even a big cross section of cast iron gets pulled out of shape on the chuck, and that’s if you have a big enough grinder. For example, put a tenths indicator on your lathe and lean heavily on the end of the bed and watch it move – and the reference is a lot more “flimsy” than the lathe bed. I would guess commercial ones are done on a grinder then finished by scraping to surface plate, but hey if you got a 24” grinder, go to it and check it with blue and the surface plate.

A lot of talk so far on reference has been dovetail ways which have an angle less than 60, not square – so don’t’ really understand the “why wouldn’t I use the 90 reference I bought” question. You can buy a way reference, just not sure where and they’re expensive. When I say reference, I’m talking about something like the pair in the grey box on page 2 or the camelback, not common items at the local tool store.

Also just for clarity, when scraping, the part being scraped is compared to a reference – that reference may be a surface plate, precision square, precision flat, camelback or the matching machine part that you just scraped flat. Point is, if you bought a right angle reference, and you’re using it for scraping, it’s a scraping reference in this context. It’s just given very high precision is very expensive to buy, and the reference needed in are large (even more expensive) but they can be very accurately home made.

take lots of photos of the bport job