View Full Version : Dumb question I have.

03-09-2012, 12:28 AM
So let's say I have a chunk of whatever in my 3 jaw lathe chuck. Let's says its steel. I advance the cross feed 10 thou. If I make a cut one way, by cranking the saddle along the bed. Then don't touch the compound or crossffed, I move the saddle back along the piece the opposite direction, it cuts again. Why does it cut in both directions? The tool is no closer to the work.

I don't get that.

J Tiers
03-09-2012, 12:34 AM
1) it cuts whatever it didn't the first pass, which can include the little ridges between the cut areas.

2) It IS closer, the "spring" in the toolholder caused the cutter to be slightly farther away as it made the first cut and was forced away by the cutting force. That's why the repeated cuts are often called "spring passes".

3) the saddle may have a slight "slop" worn in it, and be twisting a hair as the direction is reversed.

03-09-2012, 12:39 AM
Metal is as rigid as a wet noodle. The more you repeat that, the better you will understand what goes on when 5 passes later with a boring bar and shavings are still being produced dispite the fact you have not yet advanced the cross slide.

03-09-2012, 12:41 AM
The reason the cutter cuts when backing up is that there is flex in the tool holder as well as the piece of steel you are machining. When you make the cut in one direction if the setup is not rigid enough the tool will push away from the part as it tries to cut, when you back the cutter up the pressure is less then on the forward pass and the tool and the part will spring back causing the cutter to cut even though the cross slide was not advanced.

Hope this helps.


I was slow typing and see I have repeated what is above

03-09-2012, 12:41 AM
I'll take a stab at this, although others may be able to explain it better. You have two possible factors causing this. Flexibility and heat. When you take a cut on a lathe, you are generating quite a considerable force between the cutting tool and the workpiece. Depending on the size of the workpiece and how much of an unsupported length you have out of the chuck, the workpiece will flex away from the tool just a little bit. This is why finish passes are quite small, and you will often hear people speak about "spring cuts", which is taking another cut without advancing the feed any, to clean up that amount of material which would have been cut had the workpiece not "sprung" away from the cutting tool on the previous pass. Also as the tool cuts material from the workpiece, heat is generated, sometimes quite a considerable amount. The workpiece will actually expand to some degree, depending on the type of material, and contribute to what you are observing. Generally, you don't worry about it until you get close to your finished dimension, then you take a very light pass or two, sometimes without even advancing the cross slide, to compensate for these effects.

03-09-2012, 12:48 AM
It's not a dumb question. It's something you learn about and compensate for during the course of becoming a better machinst.

There are a few things that happen. One is that when you make a cut, often the tool is flexed away from the workpiece by the forces. When you stop the feed, the spring in the setup wants to relax, and pushes the tip of the tool back into the work. If you take another pass at the same settings, or reverse the feed, it cuts again. When the flex has relaxed, the tip of the cutting tool no longer pushes against the work.

Another factor that comes into play is that the leading edge of the tool normally does the cutting, and it will lose sharpness over time. When you reverse the feed, a fresh part of the tip comes into play and it will shave more material off.

The geometry of the cutting edge is interesting. With a given material, there will be certain angles that you could grind on the cutting edge that would result in the tool neither deflecting away from the work or pulling into it. Either thing can happen, depending. Ideally, you might want the forces to balance so that the tool doesn't deflect one way or the other. That can happen, but mostly I think the tooling is pushed away from the rotating work piece to some extent, with normal feed, speeds, cutting angles, and materials. You definitely don't want the tool to be pulled into the work.

03-09-2012, 01:13 AM
Metal is as rigid as a wet noodle.

I like that explanation, work piece, tool holder, tool post, the whole damn fricken lathe, its just metal, and it moves, just like play-doh.

Think about your everyday life. What really is 10 thou? Twice the thickness of your hair? about that.

Think about the solids you encounter in your daily life, your desk, your chair, a wall, your steering wheel, your car door. They are all "solid", you can move those .010", easily, most likely a large multiple of .010".

Now think about your lathe, you moved 2 hair widths and it pushed off a bit due to everything being like play-doh, and when the carriage came back it took off a small tiny fraction of the width of a hair.

Pretty amazing stuff, where else in life can you actually witness a measurement the fraction of a hair width?

It happens on the biggest million dollar machines as well as the cheapest harbor freight machines. Metal does move. Crazy amazing stuff, why I like metal.

Wood on the other hand, if you let out a juicy fart 3 rooms over, it warps, grows, shrinks, and then cracks and all your fancy measurements go out the window.

03-09-2012, 07:50 PM
Relatively speaking when we make a cut we are in essence cutting a very fine thread. Part of what we do to fix this us use a slight round on the nose of the tool so that in effect we get a little overlap of the cuts and as a result a smother finish more like a series of scallops rather than peaks like a thread.

It is also my understanding that for a cut to be effective you need to have a minimal cut for the tool to prevent it simply rubbing. Hence my question: is it better to take the same cut several times trying to "lop off" the heads of the "threads" or to take a cut that is just close to the diameter you want and finish with a very fine cut that will reduce the flexure in the tool and have enough overlay of the cut (by using a slow speed) to get a more true cut.

I realize that at some point you just change processes and move on to either honing or lapping.

03-09-2012, 10:24 PM
So let's say I have a chunk of whatever in my 3 jaw lathe chuck. Let's says its steel. I advance the cross feed 10 thou. If I make a cut one way, by cranking the saddle along the bed. Then don't touch the compound or crossffed, I move the saddle back along the piece the opposite direction, it cuts again. Why does it cut in both directions? The tool is no closer to the work.

as others have said, work flexes, tool flexes...but even with a stout piece close to the chuck and solid tool post you still notice it.

think of all the cross slide and compound parts as a stack. If each is very flat and fits perfectly to its mate the stack is solid and supports itself very well. However, wear or a poorly made lathe to begin with, does not have perfectly flat bearing surfaces and their mateing surfaces do not fit well...think a stack of belleville washers. The load from cutting compresses the belleville washer lathe and it takes another cut when you go back over it.

Getting these bearing surfaces right is not trivial; its why high end machines cost so much and why so much effort is put into scraping machine tools.

If you are curious the state of yours, take it apart and check the flats of the dovetail using blue on the surface plate. Then blue one part and check to the other. As I've never bought a new HLV or 10ee or other high end lathe you'd expect to be dead on, I've never bought a lathe that passed this test well when it first arrived in my shop :). But it can be fixed