PDA

View Full Version : feeding in or feeding out when facing



rmatel
09-05-2004, 10:52 AM
Sounds like a song lyric doesn't it?

Anyway, the "auto cross feed" thread reminded me of a question I've wanted to ask. What are the pros and cons of feeding in vs. feeding out while facing? Does it depend on the size of the machine or the type of material, or maybe the size of the material?

Bob

[This message has been edited by rmatel (edited 09-05-2004).]

BillH
09-05-2004, 11:07 AM
My southbend book shows you to face feeding out. However I have always did it going in.
I would think it would depend on your tools geometry more than anything else. I guess feeding in with a sharp angle would be good for roughing, and feeding out with a very small angle would be good for finishing.
This is my educated guess.

[This message has been edited by BillH (edited 09-05-2004).]

Forrest Addy
09-05-2004, 11:50 AM
The traditional fitting technique for engine lathes is to face flat to slightly convex. The theory is that slightly convex faces joint up and seal better than concave.

If I want flat and the lathe faces convex I feed from the inside out counting on tool wear to flatten the convexity.

Ryobiguy
09-05-2004, 12:06 PM
I think this came up some time back, and the thing that totally sold me on feeding outwards is that if you goof up (and engage the cross feed when you shouldn't) you're not going to screw up the part. But, I guess it's kind of like those chuck keys with the springs on 'em - if you need the extra safety, you're probably an idiot anyways. http://bbs.homeshopmachinist.net//smile.gif

Paul Alciatore
09-05-2004, 01:03 PM
If it's just a facing, I feed in for the rough cuts and then out on the last, finishing cut.

Many times a facing cut is to an inside corner and meets a turned OD. I will turn the OD with a series of rough cuts that end about 0.020" from the desired face. Then when that section is within 0.005" or so, I will make a in cut on the face to bring it within about 0.005". Finally, I will make one more pass reducing the OD to the final size (perhaps a bit over for abrasive finish) and then a outward facing cut to do the same for the face. It seems to work well and produces a good finish on the face.

Paul A.

JRouche
09-05-2004, 01:22 PM
If a tube I'll face in to out. If solid I'll face out to in. But as always, nothin is set in concrete (cept Jimmy Hoffa). JR

wierdscience
09-05-2004, 05:41 PM
With me it all depends on where I'm at on a tool change,if I can turn the od then face to the center I go for it.If the part is a tube or has a bored hole I face out,simply because I don't want to roll a burr on the id.

rmatel
09-06-2004, 08:56 AM
So again, it depends on the cat skinner! Thanx 4 ur responses.
Bob

darryl
09-06-2004, 08:00 PM
Feeding in for the roughing, and out for the finishing- essentially you're scraping to a finish with a sharper edge feeding out. More of the edge is in contact with the work, evening out the irregulatities to a greater extent. There's a fine line though, if the edge starts to grab, the finish will be worse. While feeding in, you're digging rather than scraping, which will not leave as good a finish.
A good analogy is using a pencil. Dragging a pencil away from it's point will leave a better line than pushing it towards the point.

kneislyd
09-07-2004, 08:44 PM
At my real job we make tire molds. When facing large heavy parts we feed out. With constant SFM enabled it takes less power since the table need to slows down as the tool moves out. These parts have a lot of inertia. We've done test and find we can take a heavier cut feeding out.

entoffice
09-07-2004, 10:52 PM
you also have to consider the tool geometry-feedout with a right hand tool-feed in with a left hand tool

entoffice

Evan
09-07-2004, 11:55 PM
Forrest, I will have to call you on this. The traditional lathe faces from flat to slightly concave, not convex. Convex would produce a poor seat/fit/seal.

From the UK lathes website re: the Wade precision toolmakers lathe:

"Wade claimed very high standards of accuracy for the lathe: the taper hole in the nose of the spindle, and the chuck seat on the outside, concentric and true to within 0.00001". The spindle alignment with the guideways on the bed, carriage and tailstock to within a tolerance of 0.003" over 12" and a facing cut made across 8" in diameter would produce a surface within 0.0002" concave and 0.0000" convex."

dsergison
09-08-2004, 10:35 AM
yeah. I don't know jack but I have read on many occasion that lathes were supposed to be slightly concave. did you perhaps just transpose your terms Forrest?

Dan

metal mite
09-08-2004, 02:05 PM
Maybe Forrest is having a senior moment!
Or, maybe it's me.

Or maybe, I'm out of the loop again.

mite

[This message has been edited by metal mite (edited 09-08-2004).]

[This message has been edited by metal mite (edited 09-08-2004).]

tattoomike68
09-09-2004, 12:55 AM
darryl has it right, rough in finnish out.

the trick to high quality finish is always chip control, if chips pack into the workpeice & tool it will look bad and you will not get a flat face.

this pic will show you, you may have to mess with yahoo.

http://photos.groups.yahoo.com/group/machinistnmanufacturershdqtrs/vwp?.dir=/shop+photos&.dnm=clipsfly.jpg&.src=gr&.view=t&.hires=t

kap pullen
09-09-2004, 07:22 AM
That's something out of those obselete South Bend, and Atlas lathe operators manuals from the thirties.

Don't see much use to feeding out.

Feeding out will, depending on the tool, give you a much wider cutting area causing chatter on those light lathes many of you use.

The only reason I can see for feeding out is cutting iron castings where you don't want to ruin the tool on scale before starting the cut.

Machining inside out will break the scale away instead of grinding your tool down.

You need a tool to curl the chip away from the face to avoid the problem tattoomike talks about.

I got fifty lathe dogs and I haven't used one of them for years either.
Obselete junk!

kap

John Stevenson
09-09-2004, 08:13 AM
<font face="Verdana, Arial" size="2">Originally posted by kap pullen:
That's something out of those obsolete South Bend, and Atlas lathe operators manuals from the thirties.

Don't see much use to feeding out.

kap</font>

Problem is Kap those obsolete lathes still haven't changed much over the years.
I find it's horses for courses, depends on a lot of factors, tool and material mainly.

Noting written in stone, did two jobs yesterday where they were opposite, one got best finish going in, t'other best finish going out.

I never say that there is a right way and a wrong way to do a job, just many DIFFERENT ways.

John S.

kap pullen
09-09-2004, 03:43 PM
Quote:

John S.

"I never say that there is a right way and a wrong way to do a job, just many DIFFERENT ways."

Don't think I said there is a right or wrong way.

There are ways bound to get you in a heap of trouble though.

It depends what you have to work with and your experience.

How you face dosen't make a lot of differance. That's not the hill I want to die for.

Some of the "self proclaimed experts" on these forums, expect us to believe what they say, like gospel.

Like one "Expert" claiming lathes are made to face concave on another post.

Or the guy told me throw away my indicator and use a level to line jobs up.

You guys do it like you want, but I'm gonna face it outside-in, like I have for the last thirty five years (usually).

O/T

I think Tony Blair is great.
You guys make great allies.
Thank you.

Kap

Still got a box of obselete dogs (1/4' to 3" by eighths), never been used, going on EBAY

nheng
09-09-2004, 04:19 PM
Kap: You're absolutely right in that they are not MADE to cut concave but the tolerance limits on inspection
sheets show "concave only". I think this is where the myth came from.

So ... you more than likely will cut a few
tenths concave by being within the center of the tolerance limits for most machines (Mori-Seki, Harrison, Emco).


If your machine cuts a perfect 90 degrees, it's on the way out http://bbs.homeshopmachinist.net//wink.gif

[This message has been edited by nheng (edited 09-09-2004).]

Evan
09-09-2004, 05:32 PM
Lathes are made to cut slightly concave since only a fool would design something with no tolerance allowance. If the stated tolerance of something is +0.0 and -0.001 then the manufacturing design target will not be 0.0, it will be -.0005.

This doesn't just apply to lathes either. The specs for a Yuasa rotary table calls out a surface flatness of 0.0 convex (flat) to -0.0006 concave. I'm sure it would be nice if they could make them all to 0.0 but they can't so they will aim for slightly concave. You can't build to the edge of the allowable error band intentionally and there is always an error band.

If you did try to shoot exactly for the edge of the tolerance band then statistically half of your product would be out of spec.

nheng
09-09-2004, 06:58 PM
For normal manufacturing I would totally agree but does it apply here ??

Lathe makers were scraping in the cross slide, carriage, etc. for near perfection.

I wonder if you wouldn't see a distribution that was heavily weighted toward the "nominal".

Just thinking out loud here and somewhat embarrassed that I didn't consider the targeting of -0.00025 as being "designed in" concave http://bbs.homeshopmachinist.net//frown.gif

Of course, a "designed in" concave today would allow the machine to wear closer to spec, rather than worse (this was just discussed in another thread, possibly on the PM forums).

spope14
09-09-2004, 08:44 PM
Feeding out was mainly for the Armstrong tool holders where the "flex" of the holder would be less starting at center where surface footage would be less overall during the start of the cut. This was in a few books I have read over the years. Tool geometery was adjusted for this as well.

I used to feed out on finish cuts as well by teaching and habit. I still do, as mentioned, on tubing and drilled pieces due to the burr on finish. However, new inserts and tool geometery and more solid machines and posts do not give as big of a difference on this. Some materials and situations will still do this, but for the most part, this is not the issue it was.

However, to each their own, and yes, I do intend to try this in my shop tomorrow, and if I find something different, using many of the great suggestions above, I will report this.

One thing I love about this page, I get great ideas and learn new ideas.

kap pullen
09-10-2004, 08:06 AM
I goofed in my previous post.
The expert said lathes are made to cut convex.

The manuals say, at least for retrofits,
lathes to 12 " should face .000 to .001 low in the center, and larger machines .000 to .0015. (the 12" dia may be wrong)

Check that machine tool reconditioning book you all waited so long for.

When I was a dfa (dumb Fuggin apprentice) procedure was when facing, turn to the carriage stop, leaving a skim cut for finish facing.

In the days before readouts, that put the wear on the pinion and rack, instead of the crosslide screw and nut.

You were more dependant on the screw and nut at that time for your accuracy.
Now many of you have readouts.

To this day that's how I operate.

Those old Monarchs were graduated both up and down on the crosslide dial, so your dial went down as the size of the stock did.

You could preset your dial at say .998 what you measured on the part, and when the dial read .750 that's where you were cut wise.

Maybe Monarch had a patent for that.

To my way of thinking all other dials are graduated backwards.

That's about a nickle's worth.

kap

J Tiers
09-10-2004, 12:39 PM
Kap:
I might just have another back-burner project, now that I read that post of yours....double dial is a good idea.
.
.
As far as the which way to feed issue?

I was reminded by one of the posts above, that if you feed OUT, the force on the tool is away from the work for many tool-holder styles.
If it shifts, it is likely to move away and cut less deep. You can fix that.

If you feed IN, and the tool shifts for any reason, it will dig in deeper and probably ruin the work and tool.

Depends if you are turning and facing, or just facing. If just facing, you can set the tool generally parallel to the ways, so it doesn't matter which way you feed.

If you use a QC toolpost, and you are turning and facing, you could have a second cutter and swap to put that holder in the boring position, of course. Only a few seconds to swap.

For the block or rocker posts, this is less convenient if you are turning and facing. You would have to move the tool and re-set to face with tool parallel to ways. Or just deal with it and face outwards...since you were probably turning right to left

darryl
09-10-2004, 02:46 PM
I found it interesting to read that lathes are made to face slightly concave. I can understand that for the reasons given, but doesn't that suggest that everything you face, you must want concave? What if you actually want it flat?
Anyway, consider another scenario. As suggested just above, you can always orient the tool to be parallel to the ways, and if it's sharpened to the same angle on both sides, it should cut the same in either direction. Well, two things come to mind with that. One, what happens to the actual position of the cutting edge when force is put on it from either side- will it extend towards the chuck slightly from the twisting force on the tool holder, compound, crosslide, etc, or will it retract away from the chuck slightly. It's a bigger issue, involving the whole lathe, including the spindle bearings. Two, when you're feeding in, the tool is touching the outside of the arc formed by it's cutting action, and when feeding out, it's touching the inside of the arc. Feeding out, there's slightly less rake and less clearance, feeding in there's slightly more of both. I can imagine there being an effect on the amount of material removal. And as well, this could change with the distance to the center of the workpiece. This would mean that you could be turning a slight concave if feeding one way, and a slight convex if feeding the other way, and this is only one reason for this possibility to exist. You might in fact, actually be turning a curved face, probably toward the convex nearer the center. Seems to me that if the proper alignment of the lathe is supposed to allow a slight concave, within a fraction of a thou over an inch or more, that it would depend more on operator skill than anything else to achieve the result wanted, whether a concave or convex facing. Bear in mind I'm talking fractions of a thou here, a level which might be small compared to the flex in the tool bit, the holder, etc, down to the carriage itself.

[This message has been edited by darryl (edited 09-10-2004).]

Evan
09-10-2004, 03:34 PM
The variables you mention are all real and will affect the result. Another way to think about the issue is to consider what is trying to be achieved. First, if you consider how many parts fit together or work then it is obvious that given a choice between a concave face and a convex face the concave is prefereable. A washer, the face of a slitting saw or simply an item that sits on a "flat" surface.

Trying for 0.0 dead flat is an impossible target, it is a single point in the range of tolerance and is infinitely small. Nothing is perfectly flat. Given that then concavity is the prefered condition. Keep in mind that for a good quality lathe properly set up the amount of concavity will be unmeasurable by most common shop instruments. It's not going to look like a telescope mirror or something. For our purposes it is flat.

[This message has been edited by Evan (edited 09-10-2004).]

J Tiers
09-10-2004, 05:12 PM
Makin it simple.....

One way (convex) a faced item would rock on a flat surface because it is sitting on a small diameter while the outside does not touch teh surface.

The other way it will rest on the outside edge and won't rock.