Foundations of Mechanical Accuracy by Wayne R. Moore

[Tyrone Shoelaces]

My milling experience has been quite different. Of course, your and my technique and tooling may be also be very different. But over the past few years I have milled probably on the order of a hundred straight edges of my own design and manufacture. The majority have been of a prism design and that type of straight edge would immediately reveal any mill-induced warping that might occur as there are three planes intersecting that are milled in succession. My routine is, using carbide-inserted face mills, to first mill the sole, then I mill the 45 degree face and finally the top rail. Of course, these casting are properly thermally stress-relieved prior to milling (or sale) in my oven by me. So, their condition is known.

After milling one face, I often check it for flatness and then go on to the next and then check all three faces when milling is completed. In almost all cases each face checks within 1 thou of flatness after machining on the prism is complete. If the casting were warping them due to either relieved internal stress or induction of new stress through milling, one would expect that perhaps the last surface milled might be straight, but the second face would be warped and the first warped even more. But that is not the case.

Is it possible to induce stress and warping of a casting by milling it? I suspect it is. For instance, I have also noted that milling with high RPM or dull cutters can cause surface glazing and smoking hot blue chips as well as a very very hot part. So, I am careful to avoid those traps. With good inserts and moderately slow SFPM I get a dull grey cross-hatched pattern and I make a light (.005") finish pass after allowing the casting to cool uniformly and to near room temperature. (I have previously written up the methods I use to support the casting to minimize fixturing strain.)

And another but anecdotal report really struck me. I had a guy pretty new to scraping buy an 18 from me a while back. He gave me some updates on his progress (I really do like that) and reported he first mille and then scraped the sole of his 18. THEN he milled and scraped the 45 degree face. I was stunned to hear that his first face blued up essentially the same after milling the second face as it did prior.

For those unfamiliar with my 18" prism design here is a photo.
View attachment 384236
And here is a 26" prism of similar design that I also have machined numerous times (not this one obviously) with similar results.
View attachment 384238

Denis

The finish on the one above is as shiny as a dogs nose. If that would have been planed and then flat tooled it would be a dull grey with a nice pinstripe. Just saying.

Regards Tyrone.

 

[dgfoster]

The finish on the one above is as shiny as a dogs nose. If that would have been planed and then flat tooled it would be a dull grey with a nice pinstripe. Just saying.

Regards Tyrone.

Well, kinda shiny. Photos of machined surfaces are notoriously difficult as above each micro-groove reflects light where it is at the correct angle from light source to lens. You can see the same finish on the inclined face is not reflective. A glazed surface is far different. Wish I had a pic, but I avoid them.

The most important point to this morning’s post was not the photo appearance of the surface but rather whether milling, done correctly, induces significant stress and strain. I do not doubt that a planed surface may be “better.” But we are back to the question of how much better—-in terms of distortion and friendliness to the scraper blade. I do not think distortion is a major player. But I am willing to listen to other opinions and experience.

Denis

 

[marka12161]

I would think any quantitative discussion would need to be put into the context of the magnitude and direction of the residual stress in the material, not to mention the impact of differences in hardness in the material. While there may be differences in the stresses induced in milled vs planed surfaces, are those differences significant relative to these other factors or are they just of academic interest?

 

[Milland]

If you want an advanced discussion on machine tools construction. Look up and read DR. Alex Slocum of MIT. Midland on here is a friend of his and an MIT grad.

Hi Rich,
Thanks for the "promotion", but I worked at MIT, didn't graduate from there. Alex is someone I know, I can't quite claim him as a friend, but an acquaintance. He is a friend of Josh's, the guy who owns the company you did the Somerville, MA training in some years ago.

 

[John Garner]

"The theory that a cast iron part must be exposed to the weather to rest and 'age' in order to stabilize it is a carry-over from the past."

Interesting that people including many on the forum will say that storing castings outside for years stabilizes them. The line above is gospel from the man himself, there have also been numerous papers posted agreeing with Wayne Moore. The line about this misnomer being a carryover from the past is spot on.

So was Wayne Moore contending that "weathering" to relieve stresses in cast iron doesn't actually relieve those stresses, or that weathering is not the ONLY way to relieve those stresses?

 

[Tyrone Shoelaces]

So was Wayne Moore contending that "weathering" to relieve stresses in cast iron doesn't actually relieve those stresses, or that weathering is not the ONLY way to relieve those stresses?

We used to “ weather “ castings by putting them into a stress relieving oven.

Regards Tyrone.

 

[MCritchley]

So was Wayne Moore contending that "weathering" to relieve stresses in cast iron doesn't actually relieve those stresses, or that weathering is not the ONLY way to relieve those stresses?

Wayne was a smart guy so he and every smart metallurgist understands weathering iron castings does nothing. Heat does everything!

One of the papers that’s been posted here talks about the effects of weathering and dismisses it entirely.
The paper did mention that castings do rust outside and were typically shot peened to get the rust off. Shot peening does tend even out the surface stress of the material. So the equivalent of outdoor weathering is just a trip to a shot peening booth.

One might ask why did machine tool vendors stack castings outside for years? They simply bought a bunch of castings at once and outside storage is cheap.

 

[Richard King]

Storing parts properly is the key. When I toured machine builders plants in Taiwan and Turkey was asked to figure out why the castings moved so much after their initial machining. I asked to see where the castings were stored when they came from the foundries and heat treaters. In Turkey at Spinner a German company that has a plant in a suburb of Istanbul they stored the castings outside in the open weather. They either set them on the ground or on wood boards with no specific place. Then the piled more castings or layered on top of each other 3 or 4 castings high. In Taiwan most of the factories had warehouses where they stored the castings as it rained a lot. Also they set them on the concrete floors in no rime or reason, some with wood blocks and some without.

At YCM (old Supermax) they made machines that looked like Giddings & Lewis Horizontal Boring machines. The columns were approx. 12' long x 36" x 36" at the base and tapering smaller the higher it went to 30". This had a real issue moving after they machined then grinding and then assembling the headstock on them and scraping. YCM added 3 points in their patterns so the casting had built in 3 points.

I showed them - in Turkey and Taiwan something I learned when I was a kid - taught by my Dad when I was his Apprentice. Simple and something many of the old timers did all the time. We stored casting, machines and parts as we dismantled machines, when we scraped machines and never had issues. We put them on 3 points or on 30% on long castings like Mill tables. I would show them a board like a 2 x 4 and how it bent laying flat, but how much stronger it was sitting on the side.

These people had degreed Engineers, but they assumed cast iron was not going to move. In Turkey the built steel skids racks so they could set mill table on their sides at 30% and it would hold 6 per skid. Those skids set on 3 points on the ground. I YCM in Taiwan They cast into the columns 3 legs on the back side so it set on the same 3 points when it was stored, machined and ground. Giddings and Lewis did that. Next time you see a G&L look at the column and remember who told you...lol

The issues they had seen was caused by not storing the parts properly. Most of the machinists understood 3 points and I would see the parts on big Planners or grinders on 3 points, but after they sat twisted in the storage areas for months or weeks they developed a twist. Even though the machinists set them on 3 points the castings had set in a twist and as time after machining they castings were set back on the floor any old way and would twist again.

In a recent thread on one of the forums I write on, someone showed a cylindrical grinder on 2 wood 4 x 4's and asked why it was grinding a taper. It was obvious that the machine needed to be set on the floor and on it's leveling screws. The owner argued that the machine would not twist. I said "why did the factory put in the leveling screws. He wrote how he would set it on the leveling screws and he argued it would not help. That was 3 weeks ago and now he is silent. I think he is to embarrassed to write again.

Van Norman Crankshaft Grinders were made to set on 3 points as a Blanchard Grinder, Heald Grinders, Moore Jib bores, Pratt & Wittney and Sip Jig bores all set on 3 points from the get go. Many others too. Use 3 points and see the difference.

 

[Richard King]

When I store or scrape a grinder table or saddle I don't set them on horses or a bench and shim up around the edges so they don't rock. I look at the weight or design of the casting or part and put it on 3 points to eliminate the twist when storing that part or scraping it. I may block around it, so it doesn't move if I have to push hard when I'm scraping. On a surface plate I put it on 3 points and I also put small jack screws around the perimeter to help support the edges if I set something heavy on the plate. Finger tight pressure on the surrounding screws. One does not need a Engineering Degree to do it the right way. You need to have common sense and have a good teacher. :-)

 

[Richard King]

When I was an Apprentice, my Dad would do the service calls. and would be out of town for weeks. I now am gone for weeks. When I taught in Taiwan I would work there 90 days at a time. It was hard on the family back in MN but that's what I had to do.

When I was a kid, My dad had to go out for 2 weeks to Quincy IL and left me at the shop to dismantle a Do-All surface grinder. I took the table off and stood it up lengthways up in a corner behind the base. Then I took apart the rest of the machine setting it on boards on the floor. When my Dad returned he blew a gasket seeing the table in the corner. That was 50+ years ago and I remember it as if it was yesterday. He and I took the table out and put it on metal saw horses and he showed me in an angry tone how to put the table on 3-points. Then we checked it with out King-Way gage which is a sled with a level vial that checks for twist. The table was twisted out of the bubble read. He was furious. He was a red headed Irishman. I apologized and said I could scrape it and he said "NO, we will let it sit there for a dew days and it will twist back to good".

That's what we did and in a few days the twist was gone. A lesson I will never forget. I write in my class manual that they say scraping is a lost art, so is know how and using 3-points. :-)

 

[dgfoster]

When I was an Apprentice, my Dad would do the service calls. and would be out of town for weeks. I now am gone for weeks. When I taught in Taiwan I would work there 90 days at a time. It was hard on the family back in MN but that's what I had to do.

When I was a kid, My dad had to go out for 2 weeks to Quincy IL and left me at the shop to dismantle a Do-All surface grinder. I took the table off and stood it up lengthways up in a corner behind the base. Then I took apart the rest of the machine setting it on boards on the floor. When my Dad returned he blew a gasket seeing the table in the corner. That was 50+ years ago and I remember it as if it was yesterday. He and I took the table out and put it on metal saw horses and he showed me in an angry tone how to put the table on 3-points. Then we checked it with out King-Way gage which is a sled with a level vial that checks for twist. The table was twisted out of the bubble read. He was furious. He was a red headed Irishman. I apologized and said I could scrape it and he said "NO, we will let it sit there for a dew days and it will twist back to good".

That's what we did and in a few days the twist was gone. A lesson I will never forget. I write in my class manual that they say scraping is a lost art, so is know how and using 3-points. :-)

Very interesting, Rich. And worth replicating and measuring the results to actually see what is going on.

So, based on your 50 years of experience, tutoring by the Red-Headed Irishman, and common sense, how long would it take to establish a “set” in a cast iron casting and after the set was established, how long would it take to resolve?

I would propose an experiment along the lines of my testing thermal expansion using the mu checker loaned (and now given) to me by Carbidebob. I could set up a thermally stress-relieved grey iron 36” camelback casting on 2-4-6 blocks with careful attention to 3-pointing it. Then that setup could be established as “zero” deflection. Then I could load the casting with something like it’s own weight and observe deflection. The setup could be allowed to sit for hours or even a week and the the load removed to see if it springs back and by how much. That would show how much set the casting had taken and then observations made over time to see it relax to zero.

I am thinking loading and unloading with water poured into or drained out of a suitable container could ensure no disturbance of the setup due to load manipulation.

Anyone interested? Actually observing the behavior of cast iron taking a set might be really helpful to our cumulative understanding.

The checker is quite stable with typically only 1-2 millionths drift (likely temp related due to elongation of the probe) over days’ time.

Denis

 

[IceCzar]

"NO, we will let it sit there for a dew days and it will twist back to good".

valuable advice, my 10ee shipped without the three point levelers and I've no idea how long they have been off.
what procedure would you recommend in a case like that?

 

[Richard King]

I believe the feet are cast into the base. So just sitting on the floor it's on 3 points. If your floor isn't level you can shim it. Level isn't that important as long as your close and the coolant runs to the TS end.

 

[dgfoster]

valuable advice, my 10ee shipped without the three point levelers and I've no idea how long they have been off.
what procedure would you recommend in a case like that?

Your EE is like my EE and every other one. It has only three feet and they are integral to the lower casting. So, it is always 3-pointed. The lathe base and upper are of very robust design so that even if trying to twist one it would be hard to deflect it significantly. They were often part of the on-board ship tool shop as they are compact and robust. Leveling was not a part of the program on a battleship or carrier and is not an issue with the EE like it would be for lathes that are set up on 4 or more pads.

Denis

 

[dgfoster]

I see Dr. Foster is writing. I don't know what he is talking about as I have him on ignore as he has very little experience in the machine rebuilding business. He is probably very talented as a Physician and examining some who has issues. I have 50+ years in my trade and have been successful. I have taught at new machine BUILDERS in Germany, The USA, In Taiwan and Turkey.

In 2 weeks I am teaching Scraping and Machine rebuilding at Defense Logistics East - a USA Navy facility in Mechanicsburg PA that rebuild machine tool from all the USA bases, Navy ships, etc. This will be my 4th time in 25+ years, teaching their Apprentice Classes. In March I am teaching at All World Machinery in IL, All World imports, sells and rebuilds Okamoto Surface Grinders. I teach on here. Several of the members on Practical Machinist have taken my classes and have been successful. Our mutual friend Lucky 7 has taken 2 of my class. I can tell people "HOW to REBUILD Machines" look at the stickies above. If you want references sign in to You Tube and search "Richard King Scraping Biax" also look at other You Tube shows my famous students have made several shows: Tom Lipton, Adam Booth, Keith Rucker, John Saunders, Jon Brooks, Jan Sveere, etc, etc.

Dennis makes cast iron straight edges and I doubt he has ever rebuilt a complete machine tool. If he has, can he tell us how many and does he have any photo's. Can he even scrape? If so lets see some photo's The readers can Google my Name and on the same line and type "practical machinist" and see the hundreds of "how to" over several years on here helping.

So you -THE READERS - decide who your going to believe and trust when it comes to Machine Reconditioning., Machine Alignment, 3 points, Scraping MOORE Jig Bores, etc.

Thank You DR. Denis Foster. letting me Tell it like it is! Oh how many awards have you received from Machine Tool Builders? Oh and I have been working with DAPRA..com for 30+ years. They advertise my classes on their web-site. They sell my DVD on their website on "How to Scrape" "Who you going to trust"?

My interest was piqued when an experienced member of this forum started reporting that he had seen cast iron "take a set" from improper storage. That is contrary to my understanding of the behavior of cast iron. So, not considering myself an expert in the field of metallurgy or mechanical engineering, I contacted a degreed metallurgist who is employed at a foundry that has been kind to me over the years and provided a fair bit of mentoring. And I consulted a mechnical engineer who is a prominent member of the engineering department of a firm whose name everyone here would recognize. Neither had heard of this phenomenon occuring.

So, it seems that some investigation of the facts is in order. I am planning to do some testing of cast iron in the same vein as the testing I have been doing on thermal expansion of iron.

I will carefully design the test to the best of my ability and describe my methods and results clearly (and as was done in the thermal experiments, even provide video of the tests) so there will be no mystery and so that anyone wishing to do so can replicate the testing and either confirm or call into question the results. I really hope someone else will do similar or better testing and report it. That way there will be no need to "believe" results. The results will be clear and verifiable. Anyone who wishes to take exception to my methods is free to do so and I invite legitimate criticism or suggestions. I do not recommend ad-hominem attacks on me or anyone else. That is a cheap substitute for legitimate discussion and reprehensible.

Because of travel plans and other obligations, my reporting may be a bit spotty over the next couple weeks. But, I will get it done. Those travel plans actually should provide a good opportunity to leave a casting in a strained condition for a week to see if it can take a "set."

I was hoping that Mr. King may provide some guidance on testing based on his years of experience---see my post above. But, if he chooses to "ignore" ( but not really) the gathering of information and not help us get at the facts, that is his choice.

Tomorrow I will start setting up a thermally stress relieved casting on my surface plate and put together the apparatus that, while somewhat crude, should still allow measured smoothly-applied and persistent strain as well as unloading of the casting without somehow dislodging or disturbing the casting.

Let's see what we can learn together about cast iron taking a set and then relaxing after a period of time.

Denis

 

[IceCzar]

well...err
you both agree I should have spent more time peering under the 10EE when it was in the air

thank you both

 

[dgfoster]

I am getting started on the cast iron elastic deformation testing better known as "does cast iron take a set when subjected to low-level (below elastic limit) strain for moderate periods of time. I'd do the many year's test if time allows. But for now, a few hours to a few days to week or two should do as that is the sort of time interval cast iron was observed to take a set in the posts around #30 above.

So, the first order of business is to get a casting setup solidly on 3 points and then load it while measuring its initial location, load-related deflection, and subsequent return to its initial location after removal of the load. Will the casting return immediately to its initial position or will it retain some deflection which resolves over hours, or days is the question to be answered.

It seems important that the supporting 3 points should be themselves very stable and not subject to deflection due to making poor contact with the surface plate or casting. In my mind, simply setting it up on 3 simple steel collumns or pads might be at risk for being not very solid should some random metal spicule or bit of dust or dirt be between the column and the casting or plate. So, to minimize that chance I turned a few supports consisting of center-drilled 3/4" diameter steel discs relieved centrally on the sides that will face the plate or casting. Those relieved surfaces were quickly lapped on a 600 grit diamond plate to reduce surface irregularity.

To reduce the chance of a random bit of metal sticking up in the center-drilled depressed cone, I put the ball bearing in place in the top and base pieces and then struck them a couple moderate blows with a hammer.

I needed a few of these supports for other purposes in my thermal experiments too. So I do not mind taking the few minutes to make them up.
Denis

 

[John Garner]

dgfoster --

Your "swivel pucks" are materized wisdom. Let me suggest that use a few drops of hot-melt glue to hold the top pads in place on your test object.

Beyond that, I am a huge advocate of using some form of mechanical pivot at attachment points generally, to specifically include "three-point" mountings. Yes, Euclid's "Three [non-colinear] points determine a plane", but I've never seen a mechanical attachment point that is a physical manifestation of a Euclidian point.

John

 

[dgfoster]

dgfoster --

Your "swivel pucks" are materized wisdom. Let me suggest that use a few drops of hot-melt glue to hold the top pads in place on your test object.

Beyond that, I am a huge advocate of using some form of mechanical pivot at attachment points generally, to specifically include "three-point" mountings. Yes, Euclid's "Three [non-colinear] points determine a plane", but I've never seen a mechanical attachment point that is a physical manifestation of a Euclidian point.

John

Thank you for your comments. I intend to use light oil on the balls. I am divided on the hot glue idea. I do not want to have anything "organic" potentially disturbing the contact bearing between the cones and metal surfaces. I considered cyano glue. But, I am not certain of its behavior on a millionths scale with varying humidity, time, and temperature. I do not want to unwittingly foul the results somehow so I am favoring maximally reducing variables.

Along the same lines of thinking, though I will be using a raw casting, I will lightly mill the contact points so there is good bearing there as well.

Denis

 

[Richard King]

To all the readers. If you would care to learn about 3 points please email me as I have had success using it for my career of 50+ years. Plus I have rebuild a handful of Moore Jig bores and have owned a copy of The Foundations of Mechanical Accuracy for 40+ years. [email protected]