PDA

View Full Version : Lathe problem - seems to cut too much?



loply
10-02-2011, 11:14 AM
Hi folks,

I've been having a problem with my little lathe - it always seems to remove too much material.

A few days ago I was turning a rod of stainless down to 10mm and as I approached the size I mic'ed the rod and found I had 0.32mm to take off, so I popped a dial indicator on the toolpost and wound it in 0.15mm and took a pass. Prior to this I had taken a few spring passes and the finish was even.

I mic'ed it afterwards and it was 0.03mm undersized :confused:

Just been out experimenting and I played around with some 8mm stainless bar and some 32mm mild steel and sure enough this happens every time.

In one instance on the 32mm bar I had 0.20mm to remove to reach my target, I dialed in 0.08mm and ended up 0.02mm under.

I'm using short lengths of bar, finishing with a high angle CCGT carbide insert, the bearings on the lathe have been warmed up fully, and in all cases I used a dial indicator to measure toolpost movement. The lathe is a heavily modified 7x14 Asian thing. I checked all the obvious things like saddle or cross slide looseness, but it's all snug.

Any ideas what might be causing this? Could this be something to do with head alignment?

Cheers,
Rich

philbur
10-02-2011, 11:53 AM
Possibly your tool post tends to nod forward proportional to the cutting load.

Try to do the last three passes at the same or at least similar DOC.

Phil:)



I popped a dial indicator on the toolpost and wound it in 0.15mm and took a pass. Prior to this I had taken a few spring passes and the finish was even.

chipmaker4130
10-02-2011, 12:00 PM
Is the cutting edge exactly on center height? Is the 'after' measurement the same on both ends of the bar? (taper?)

Black_Moons
10-02-2011, 12:15 PM
hmm, Consider maybe putting a TDI onto your lathe bed, and checking the cross slide movement verus what the dial says.

Also, Most people usally 'sneak up' on the final dimentions.
ie if you had 30mils to go, your next cut would be 25 mils deep, then say theres 3 mils left after measurement, well your next cut would be 1mil (or maybe just a spring pass!), And see where that gets you.

Yea, its awsome just to see 'oh it needs 30 mils, i'll dial in 15 and cut the next 30mils in one pass and get a good finish and be done in 30 seconds' but it never seems to work that way for me!

Scottike
10-02-2011, 12:19 PM
Before I'd be thinking something serious like head alignment, I'd look at some simpler, more likely solutions.
Have you tried using a fresh insert to make your finish passes with?
Are you using the correct speed, feed & doc for that insert?
Most inserts have a recommended feed, speed and doc range they work best at. Check the insert mfr specs for that insert.
Is your toolbit height centered properly? A toolbit that's too low can cause the work to try and "climb" the toolbit causing a deeper cut.
Some materials respond to a given doc & feed better at certain speeds.
And I,ve seen some materials (stainless and some steels) that have actually "work harden" while being roughed, then when you try to take a lighter cut the tool bit will actually grab the "hardened surface and end up "peeling" it off the softer material underneath.
You also might try using a HSS tool also, rather than an insert.

edit: As Black Moons said - Just turning the dial to a doc doesn't mean you'll get it, usually it ends up being more or less than you antipicated, for a variety of reasons. I usually try a few practice finish passes when I'm roughing, just so I know how things are responding, then I can allow for it when I make my final pass(es).

gwilson
10-02-2011, 12:20 PM
Lathe is probably not rigid enough. Tool may be "sucking" into cut some. Try a less aggressively ground tool that won't tend to pull into the work.

vpt
10-02-2011, 01:18 PM
hmm, Consider maybe putting a TDI onto your lathe bed, and checking the cross slide movement verus what the dial says.

Also, Most people usally 'sneak up' on the final dimentions.
ie if you had 30mils to go, your next cut would be 25 mils deep, then say theres 3 mils left after measurement, well your next cut would be 1mil (or maybe just a spring pass!), And see where that gets you.

Yea, its awsome just to see 'oh it needs 30 mils, i'll dial in 15 and cut the next 30mils in one pass and get a good finish and be done in 30 seconds' but it never seems to work that way for me!



^ that stuff. ;)

firbikrhd1
10-02-2011, 01:42 PM
As an experiment, why not try some HSS cutting tools. They tend to be keener and require less power to cut the same material depth and lower forces on the machine. If rigidity is the issue perhaps a keener edge will be the resolution.

Some here have said that Carbide today can have very keen edges and that may be so. I'm a HSS user exclusively so I haven't any experience with carbides, only can pass along what I've read, i.e. anecdotal evidence of what I said above.

Forrest Addy
10-02-2011, 02:08 PM
I suspect part or tool deflection on the previous cut. You took a semi finish cut that resulted n X deflection, dialed in, then a finish cut resilting in a similar deflection.

Then you took a couple "spring passes" (Rant:: spring passes - I hate 'em. If you have the variables under conscious control you very seldom need them. If you don't, you use them like magic charms, hoping something works) machining away the deflecton and - surprise - the part is U/S by the amount of the deflection.

Using HSS is a good tip. Carbide takes 2 -3 times the force to penetrate the work compared to HSS. Dead keen HSS in mild steel takes very little force. Stainless is harder to penetrate and since it forms a resistant work hardened film from the passage of the tool, light cut finishes can be inconsistant. As a general rule 0.010" is the least practical finish cut in austinetic stainless. As ever, YMMV.

saltmine
10-02-2011, 02:42 PM
It may just be the metal cooling off that accounts for the dimensional change. Aggressive cutting, especially with carbide cutters, can cause this.

The other thing is possible taper cutting. The spot where the measurement takes place might be closer to the axis of the taper, and may actually be the dimension you set it to.

I discovered a long time ago that I was always cutting things undersized, until I started "sneaking up on the final number, and then working the final dimension down with a file and emery cloth. Of course, if you're one of those CNC whiz-bang guys, hand filing and polishing with emery is out of the question (I have yet to see a G-code for hand filing or polishing)

Being an Asian clone, I would suspect either a mis-alignment or a lack of rigidity. Even though tool bit placement can play heavily into a tool "digging in"

GadgetBuilder
10-02-2011, 02:47 PM
I use a 7x12 for much of my work and mostly use HSS because it takes less power, something the 7x12 lacks.

When size needs to be exact I get close to the target diameter, within 0.3mm say, and then switch to the vertical shear bit:
http://www.gadgetbuilder.com/VerticalShearBit.html

The vertical shear bit takes predictable small cuts (while leaving an excellent finish) so it allows sneaking up on the finished size in very small increments. I use the dial for this rather than an indicator because tiny infeeds are easier for me to judge this way (infeed varies slightly vs material), YMMV.

Edit: Fix link

J. R. Williams
10-02-2011, 02:54 PM
Go for a HS tool and have it sharp. Carbide tools have a feed/speed cutting window for each tool. Work outside the window and you will have problems. Your problem may be the tool and not the machine.
JRW

90LX_Notch
10-02-2011, 03:27 PM
Check the alignment of the indicator. It needs to be perpindicular in both planes. When you went to take .30, you were .03 under. When you went to take .20, you were .02 under.

Alistair Hosie
10-02-2011, 04:00 PM
Please also remember the advance rule when requiring to remove 2mm you advance the tool 1 mm as it removes from back and front at the same time .Most metal lathes are set to take this into account .Ok I just wondered ? :confused: Alistair

loply
10-02-2011, 04:30 PM
Hi folks,

Thanks for the very useful contributions.

Just to answer some of the queries-

I am using a carbide insert which is meant for aluminium, it has a very high rake angle and seems to require less cutting force than my HSS tools, I say that because it cuts more silently and leaves a better finish, I do have and use some HSS tools but the finish from these high rake carbides is a lot better, I will repeat my experiment with HSS.

Tool is definitely on centre.

Re the tool nodding forwards or the work climbing, that's what I thought and is why I tried on a 32mm bar, I figured that would be too stiff to climb the tool.

No taper is present and I did try to ensure the part wasn't hot enough to shrink it's way undersize.

I will take some of this advice onboard and have another play with it tomorrow.

Cheers,
Rich

Gravy
10-02-2011, 06:38 PM
In addition to the other ideas, measure the actual travel of the cross slide in comparison to the numbers on the handwheel scale. Some people have found significant mis-calibration that way. You may be getting .032" movement when the scale says .030"

jugs
10-02-2011, 06:51 PM
In addition to the other ideas, measure the actual travel of the cross slide in comparison to the numbers on the handwheel scale. Some people have found significant mis-calibration that way. You may be getting .032" movement when the scale says .030"

Thats why I use DRO, measure the tool movement, not the dial movement.

Your cross slide screw maybe worn/badly made

philbur
10-02-2011, 06:55 PM
That doesn't account for the tool nodding forward.

Phil:)



Re the tool nodding forwards or the work climbing, that's what I thought and is why I tried on a 32mm bar, I figured that would be too stiff to climb the tool.

Hopefuldave
10-02-2011, 07:42 PM
It may just be the metal cooling off that accounts for the dimensional change. Aggressive cutting, especially with carbide cutters, can cause this.



I've found this when cutting lots of 8" cast-iron backplates down to 6" (don't ask, wasn't me who ordered the wrong size!) - positive-rake carbide tool at high speed (300SFM!), repeated 1.5mm DOC, fine feed (3 thou"/rev)... the workpiece gets too hot to handle, and I've found leaving them about 8-9 thou" oversize puts 'em right on 6.000" once they're down to room temperature...

The hot metal hailstorm's worth standing back from, too!

Dave H. (t'other one)

darryl
10-02-2011, 09:09 PM
It's definitely worth setting up an indicator to compare with the readings on the dial. Once you know that your dial does advance the slide by the amount indicated, you can discount that as a source of the problem. If you find that it doesn't, then IMHO it would be best to do something about this before you get into too much making of parts.

I've checked machines on the showroom floor and found some dials that were way off. On one machine, the crosslide leadscrew had a periodic error of 12 thou- how could you possibly be happy using such a machine?

The leadscrew error can be from the threads themselves being drunken, or from the collar that positions the leadscrew in the crosslide. If the collar is off, the leadscrew will move back and forth as its rotated. This isn't too hard to fix, and the machine would be way more friendly to use. If it's the threads that are bad, get a new leadscrew under warranty if possible.

You might want to do this same test on the compound. Align the compound parallel to the spindle axis, set up an indicator to bear against the chuck, then wind the leadscrew (and hope the dial markings coincide with the indicator readings).

My advice is to get this out of the way first, then try to find where the problem lies. Another test that's worthwhile to do is to turn a test piece, using a sharp tool, then see how many spring passes it takes before the tool stops removing material. This might show that there's a source of play still there somewhere.

On my lathe, I found that the rear of the carriage easily raised from the guideway. The effect of this is to back the tool away from the workpiece. If the carriage was touched down on the rear way, the tool would cut more. There is a tab at the back that prevents the carriage from lifting off too far, but unless the play there is zero there is going to be an effect on tool position. If the tab was shimmed to eliminate play, then the carriage would be hard to move in some spots. My solution was to make up a spring-loaded wheel that would run on the underside of the rear way and keep the carriage touching the rear way at all times.

Every time I eliminate a slop somewhere, the lathe gets nicer to use-

franco
10-02-2011, 10:10 PM
It might also be worth checking whether you actually have a metric pitch screw and not an imperial one which has been fitted with a metric dial. Stranger things have happened. Some common metric pitches are very close to imperial pitches, e.g. 2.5 mm pitch is very close to 10 TPI, but if a metric dial has been fitted to a 10 TPI screw, there will always be a small but consistent error in the indicated cross slide travel.

franco

saltmine
10-02-2011, 11:51 PM
When I first got my lathe, I assumed(you know what happens when you assume....you make an Ass out of You and Me)

It wasn't long before I ran into the same problem, and sometimes a severe chatter when cutting 4140 steel. I switched over to carbide with no luck.
Then, as luck would have it, I was re-adjusting the compound, one day and I noticed the carriage move. A closer look revealed that the carriage was lifting up about .020" off of the ways, and I could easily rock it back and forth. So much for factory set-ups. I cleaned everything off, and took it apart. The gibs were sloppy as all get-out, there was swarf and dirt everywhere, even the cross-feed nut was loose, and the apron allowed the drive gear .025" of slack (instead of .010") I spent all day, and part of the next adjusting everything. This included removing the bottom of the apron, clamping it in my mill and machining off .015". I also fabricated an aluminum cover for the back side of the apron to keep swarf out of the gears (yes, there must have been a whole bunch of it in there, the gears and handwheel would hardly move.) After lightly oiling and adjusting everything, the machine has been cutting quite well for over three years.
I use both HSS and carbides depending on what I'm cutting. I've even machined stainless steel and hard chromed parts.
Were I to do it all over again, I'd take the thing apart and clean and adjust everything before I even tried to use it.

Jaakko Fagerlund
10-03-2011, 03:56 AM
A few days ago I was turning a rod of stainless down to 10mm and as I approached the size I mic'ed the rod and found I had 0.32mm to take off, so I popped a dial indicator on the toolpost and wound it in 0.15mm and took a pass. Prior to this I had taken a few spring passes and the finish was even.

I mic'ed it afterwards and it was 0.03mm undersized :confused:
There is your major problem - taking spring passes and/or not taking proper depth of cuts to get to the diameter needed.

Generally you turn to about 1 mm oversize, mic it, take half, mic it, take the rest and it comes spot-on. In different materials and with different tools you may be required to take smaller finishing chips than this example.

beanbag
10-03-2011, 05:23 AM
Maybe the rake on your insert is TOO high, and pulls the tool and the work towards each other. If you want to use an indicator to check deflection, you need the indicator mounted directly on the tool holder, and the other end touching the actual spinning stock. Note the measurements both with the lathe on and not doing anything, and while cutting metal. Once you have this figured out, you can mount the indicator in other places and measure this deflection with respect to other deflections.

It is also possible that you are getting "built up edge" on the tip of your carbide tool, which causes it to cut deeper. Typically, a coated, slightly positive rake insert is used for steels.