Scraping and Honing

[ConductorX]

I have been reading some interesting books by a guy that built his own lathe and other machine tools. David Gingery is his name.

He says in the book that you can achieve a flat and true surface by scraping the metal. His primary material is aluminum, but he claims it works on steel as well.

Has any one heard of this? Have you done it? Does this method work?

My other question is related to measuring and checking flatness of a piece of metal. How do you determine that a piece of metal is true and flat? I have a lathe bed for my lathe that was in or very near a fire. How do I know if it is true and flat? If it is off a bit how do I fix it?

My last set of questions, I worked many years ago on Ingersoll Rand HHE compressors. We had two 5000 HP and two 2500 HP compressors. The valves in these compressors were reed type valve with spring loaded plates that would open to let the gas in then close with the opposite set opening to let the gas out. We would hone the wear surface of these valves on a huge steel plate using grinding compound. This would insure the valve plates had a true and flat surface to seal against. What is this huge 2' x 2' plate called? It was about 4 inches thick and the top was laid out in squares with a groove between each square where the grinding compound collected.

Is there something like this honing plate that would help me check my lathe bed or help correct small defects?

Thanks for your time as always.
"G"

 

[Mechanicboy]

Scraping the surface will give true geometry and flatness and a lot of pockets of oil.

[ame="http://youtu.be/D1eOQa1gYiU"]http://youtu.be/D1eOQa1gYiU[/ame]

[ame="http://youtu.be/LkdkouWiDFs"]http://youtu.be/LkdkouWiDFs[/ame]

 

[purpleknif]

Machine tools used to be "scraped in" to achieve alignment of the components after being machined. I used to work for a rebuilding service where we would regularly rebuild large machine tools. Nowadays most new machines have ground ways due to the cost of hand scraping. Its interesting to note that one major manufacturer (I can't remember which one) makes a point of Hand scraping as a quality feature.
Hand scraping probably is better but with the dropping cost of new machine tools most of them are considered "disposable" now. You don't see too many $50,000 Haas VMC's being rebuilt.

 

[Terry_S]

There are some fairly quick and easy ways to determine if the lathe is still "in alignment."

First, confirm that the tailstock is in alignment with the head. There are several methods to do this - I will not go into them here.

Next, set up a piece of round stock between centers using freshly dressed centers. Select the round stock to be the longest that the lathe can reasonably handle - and the largest diameter for which you have a micrometer. Take a light cut over the full length of the stock without touching any of the tool settings. Insure that the entire surface of the test piece have been machined. If not, then take another light cut. Light cuts are used to reduce the deflection of the work and machine which will affect the final result. Then carefully check the diameters over the full length of the cut. If diameter remains constant over the full length (or close enough for your purposes), then the lathe is good to go.

If you don't get consistent diameters, you can check the bed for twist using a master machinist level. These levels are able to discern angle changes of 10 seconds of arc, more commonly expressed as 1/2 thousandth per foot of run. From an absolute level surface, a sheet of paper under one end of the level will move and perhaps "bury" the bubble. You would have to find someone who owns such a level. Or, you can buy a Chinese knock off of the US version for about $100. I have found the Chinese version to actually work quite well. Still, the big "S" is better if you have access to one.

You place the level on cross-feed with the bubble running in the cross direction (Y direction to some). Use shims or thin sheets of paper (credit card receipts work well) to get the bubble centered. Then slowly move the carriage up and down the length of the ways - stopping periodically to read the bubble. (Note: you have to stop the carriage and allow the bubble to stabilize for 15 seconds or so before you read it.) Variation of the bubble will indicate "twist" of the ways. If the twist varies both + and - over the length of the ways, you will need some scraping or grinding to straighten out the machine. If the twist is all in one direction, you can often remove the twist by adjusting the loads on the feet of the lathe - either by jack screws (if provided) or be shimming.

Next, use the level to check the ways for sag or bow. The level remains on the cross feed but is now positioned to read in the longitudinal (or X) direction. Start at the center and shim the level to center the bubble. Then, slowly move the carriage towards the head and then towards the tailstock while watching the bubble (stopping to read the bubble). Changes in bubble will indicate a sag or bow. From the bubble deflection, you can estimate the total sag or bow.

Terry S.

 

[dman]

yeah scraping works. you need a master flat though... if you have a good sized surface plate you can scrape all the flats you need to then go scrape the ways. aluminum will work but cast iron is mostly what the industry uses when they scrape. you can scrape some steels but steel can be very hard depending on alloy and temper conditions.

there is a forum member named nick muller i think in germany who has excelent scraping videos on youtube. i'd recommond checking them out.

 

[bb218]

When I set up my LeBlond lathe I first leveled the machine front to back and side to side, trying to keep the leveling screws supporting close to the same amount of weight.
Chuck a piece of free machining brass or steel in the chuck, use as large a diameter and reasonable length as you can, take a light cut for the entire length, measure the diameter on each end. If the diameter is small at the tailstock end adjust the right rear corner some amount maybe 1/2 turn on the leveling screw and take another cut. You should be able to get the lathe cutting straight. Once you get the machine cutting straight you can adjust the tail stock the same way by cutting stock between centers an adjusting as needed.

Some lathes are to short and ridge to adjust with the above method.

Mike

 

[portlandron]

The flat plate you were asking about is called a "Lapping Plate".

 

[ConductorX]

The flat plate you were asking about is called a "Lapping Plate".

Lapping plate... DOH!! and giving myself a dope slap.

Thanks much!!
"G"

 

[Wizard69]

I have been reading some interesting books by a guy that built his own lathe and other machine tools. David Gingery is his name.

Dave has a whole series of books, very well done considering his theory.

He says in the book that you can achieve a flat and true surface by scraping the metal. His primary material is aluminum, but he claims it works on steel as well.

This is the only way to get a truly flat plain bearing surface.

Has any one heard of this? Have you done it? Does this method work?

It is the way high quality machine tools have been made for centuries now. In my youth I spent about a year working with a machine tool rebuilder doing just that. Done right it is the way to a high quality plain bearing surface.

My other question is related to measuring and checking flatness of a piece of metal. How do you determine that a piece of metal is true and flat? I have a lathe bed for my lathe that was in or very near a fire. How do I know if it is true and flat? If it is off a bit how do I fix it?

I'm not sure which of Dave's books you have but he does explain that in one of the books. Maybe not in the depth that maybe he should but the reality is you can find other books focused on the art of scrapping and only that art.

My last set of questions, I worked many years ago on Ingersoll Rand HHE compressors. We had two 5000 HP and two 2500 HP compressors. The valves in these compressors were reed type valve with spring loaded plates that would open to let the gas in then close with the opposite set opening to let the gas out. We would hone the wear surface of these valves on a huge steel plate using grinding compound. This would insure the valve plates had a true and flat surface to seal against. What is this huge 2' x 2' plate called? It was about 4 inches thick and the top was laid out in squares with a groove between each square where the grinding compound collected.

Sounds like a lapping plate to me. The squares are very important to getting a flat surface. Without those squares your lapping efforts would have produced a fine curved surface.

Notice that there is a difference between lapping and scrapping. When scrapping your intention is to get a fine flat surface however that is over the distance. Scrapping actually introduces very fine pockets in to the surface to aid in lubrication and reduce stiction.

Some times grinding is used in place of scrapping these days. To get around the stiction problem the ground ways are often flaked. Flaking adds those tiny pockets to the ground surface to prevent stiction.

Is there something like this honing plate that would help me check my lathe bed or help correct small defects?



Thanks for your time as always.

"G"

You would want surface plates and straight edges. These straightedge so though are not what you may be thinking. Look up humpback straight edge for one example.

Time to go back to work!



Sent from my iPad using Model Engines

 

[Dunc1]

Home Shop Machinist has a series on scraping by Michael Ward. Started late 2011 or early 2012. I expect that back issues/reprints would be available. Usual disclaimers

 

[barnesrickw]

I have a stone lapping plate I use for sharpening my woodworking tools. Lay sandpaper on it, sharpen. I recently lapped some of my Stanley combination squares. It works pretty well now at showing if a surface is true now.


Sent from my iPad using Model Engines

 

[purpleknif]

Well... Flat is easy. The trick is to get everything aligned when you put it back together. Not too hard on a lathe but try it on an old 6" G&L boring mill with 12 feet of column and 15 feet of saddle. Took about 7 months after prelim machining. That's why I don't do that anymore. Oh yeah and 16 feet of table. Finally did get it within .0004"
everywhere.