What's new
What's new

Lathe - 304 Stainless Surface Finish

droshi

Aluminum
Joined
Dec 20, 2017
Location
NA
Looking for suggestions on what I should do with this operation on the lathe. I'm trying to do a finishing pass using a DCMT Sandvik (2015 grade) insert. Running 960 sfm (880-960 recommended) but I only have 3k rpm on the lathe so it hits that max somewhere on the face of the part. 0.004"/rev and leaving 0.004" stock to take off afterwards.

The speed is on the high end of recommendation, and feed is on the lowest end. Depth is suggested a lot more, but I've found this operation worse at their recommended 0.02" minimum to leave.

My parts as you can see are a pretty tapered surface and I'm just running this on the large taper and face of the part (top). If I finish the OD it comes out beautiful with this DCMT or even a CNMG insert. So it's the taper geometry that seems to be killing me. OD of the part is just under 3.25"

In the past I have used a Kennametal VBMT, but insert grade wasn't ideal for stainless...repeating the finishing pass worked a bit. I also have tried a Sandvik VNMG cermet insert which I had read was supposed to be good...finish only average. The DCMT has seemed about the best, just consistency and edge life is what I'm hoping to improve.

Any suggestions on a specific insert, or what nose radius I want, or feed/speed...I'm open to it all!

Pics go from worst to best finish. The last one still has a bit of waviness in person, but not easy to pick up on cam. The insert wears just a bit and can lose that finish after just 2-3 parts.
 
I'd try 500sfm just to see, but looks to me like something is loose somewhere and causing that vibration.
What size machine and how are the spindle bearings doing in that thing?
 
Other operations do pretty well, including the underside of that where I'm doing a more challenging face grooving.

But you may be right, I do use a threaded fixture that is 1.5" hex bar 1018 steel to hold the part on the ID. I've gotten it pretty well tuned in, and again on the OD it would be a perfect finish so I was thinking it was more geometry or insert choice that was my problem. But it could be that fixture. It threads onto the threads you can see in the pic attached and up against that flat face.

Machine is a small 5hp Chinese CNC lathe, it's nearly brand new and again other operations in stainless come out very smooth.

Attached is a pic of the underside, looks even better in person, there is only a minor seam where the toolpaths meet that is more of a CAM issue.
 
Another issue with setup is that due to space and clearance, I do have a decent amount of stickout on 5/8" square tools, probably 2" or so. I have tried speed slower and feed slower, but maybe I should 1/3rd or 1/2 both at the same time and see.

Almost every other operation and part I just work within the Sandvik recommended parameters and things come out perfect.
 
Yeah this seems strange like it may be a problem in the machine. Of course I say that yet you can try variations on your feeds and speeds. If the machine is in need of repair then this will show up in the future jobs.:)
 
Yeah this seems strange like it may be a problem in the machine. Of course I say that yet you can try variations on your feeds and speeds. If the machine is in need of repair then this will show up in the future jobs.:)

For the most part, I run all the same parts for a product, so it's something I've been dialing in over the course of a while and this is the one operation I just can't seem to figure out 100%. With a new insert it can run fine for a few parts like the one on the right, then can go downhill.

It probably does point back to the work holding, since this is the only operation where I'm not chucked up in the 3-jaw directly. Though at the moment I don't know what to do about that.
 
Another issue with setup is that due to space and clearance, I do have a decent amount of stickout on 5/8" square tools, probably 2" or so. I have tried speed slower and feed slower, but maybe I should 1/3rd or 1/2 both at the same time and see.

Almost every other operation and part I just work within the Sandvik recommended parameters and things come out perfect.

Your first pic looks like chatter marks also to my eye.
5/8" tool with 2" stickout is pushing your luck. Probably works most of the time but it's definitely more prone to chatter.
Sometimes slightest wear on the insert can change the behaviour when you are at the limits of work&toolholding.
(or almost any change for that matter, not just insert wear)
 
It probably does point back to the work holding, since this is the only operation where I'm not chucked up in the 3-jaw directly. Though at the moment I don't know what to do about that.

The first thing I thought was.. "How are you holding it"?

Threaded arbor.. I'm assuming the back of the part is against something...
I'd vote that is your problem right there.

Maybe try drilling and tapping the threaded arbor, and then slot it.. Expanding
arbor style. Looks like you need just a bit more security in the holding, not
a lot, doesn't look like you are far off from a really good finish, just need
to stabilize it a bit more.
 
The first thing I thought was.. "How are you holding it"?

Threaded arbor.. I'm assuming the back of the part is against something...
I'd vote that is your problem right there.

Maybe try drilling and tapping the threaded arbor, and then slot it.. Expanding
arbor style. Looks like you need just a bit more security in the holding, not
a lot, doesn't look like you are far off from a really good finish, just need
to stabilize it a bit more.

Yeah this was a plan at some point, the Expanding 5C seemed like the ideal solution (once I can get a 5C pneumatic closer).

I tried a smaller expanding mandrel in the past, it didn't work at all for work holding (once spun in 3-jaw, another time part just tilted), but I was only expanding on the smooth bit, not the threads. I think if I get a larger one and thread up to a shoulder same as I'm doing now, just with an expanding bit it would be much better.

My turret is an 8 position, but it's basically a scaled down version of the standard ones on larger lathes. So while it works and all, clearance is really tight due to the diameter and this is what leads me to need a decent amount of stick out.

Pics attached, it's in the mill right now just for engraving, but normally I chuck that up in the 3-jaw.

Thanks for all the help! At least now I know what to try to fix to solve this issue.
 

Attachments

  • IMG_20181102_205727052.jpg
    IMG_20181102_205727052.jpg
    83.6 KB · Views: 153
  • IMG_20181102_205746625.jpg
    IMG_20181102_205746625.jpg
    86.6 KB · Views: 143
Too many variables. Cheap Chinese Machinery is what it is, right? It's like doing surgery with a spoon, it's just not the right tool.

R
 
You state you are using a 5/8 tool. Is that the max size for your turret or is it what fits to your workpiece?

I often turn 316 SS to a finish that is acceptable for a seal. Same inserts you are using but much slower speeds.

Ed.
 
1) New, but cheapest machine known to man.

2) 960 SFM in 304.

3) 2" stick-out on a 5/8 tool.


Man .... You must be a pro to git the results that you have right there! :bowdown:


The threaded arbor may play a roll, but with that small of finish pass, I doubt that it is in the top 3 concerns...

I'd say that your tool stick-out is by far the top place to look for chatter.
Less SFM for tool life.


Generally speaking, I find that manufacturer specs are highly embellished. But then - they have to.
You are normally looking at the rec S and F when comparing tooling. How often doo they publish a tool life to go along with it?
Are you going to buy from the company that says to run slower and back off a little?
They'll werk that way for a while, but will last 5x as long if you scale it back a little.

On the other side, I have found Titex tools to run almost exactly as long as they said at their rec S & F's. They doo offer a tool life chart!


--------------------

Think Snow Eh!
Ox
 
Actually after reading Ox' post I agree, there is some stuff that is just hard to imagine. Then I looked at the Sandvik calculator for grade 2015 in 304L :eek: not exactly in the 900 range.

ToolGuide
 
You state you are using a 5/8 tool. Is that the max size for your turret or is it what fits to your workpiece?

I often turn 316 SS to a finish that is acceptable for a seal. Same inserts you are using but much slower speeds.

Ed.

Max size for the turret. I know it's not ideal, but for now on the lathe it's what I have.


1) New, but cheapest machine known to man.

2) 960 SFM in 304.

3) 2" stick-out on a 5/8 tool.


Man .... You must be a pro to git the results that you have right there! :bowdown:


The threaded arbor may play a roll, but with that small of finish pass, I doubt that it is in the top 3 concerns...

I'd say that your tool stick-out is by far the top place to look for chatter.
Less SFM for tool life.


Generally speaking, I find that manufacturer specs are highly embellished. But then - they have to.
You are normally looking at the rec S and F when comparing tooling. How often doo they publish a tool life to go along with it?
Are you going to buy from the company that says to run slower and back off a little?
They'll werk that way for a while, but will last 5x as long if you scale it back a little.

On the other side, I have found Titex tools to run almost exactly as long as they said at their rec S & F's. They doo offer a tool life chart!


--------------------

Think Snow Eh!
Ox

It's not the cheapest known to man. It's small and sometimes frustrating for that reason, but box ways and can take a pretty decent DOC for it's size in stainless. Glad to hear you are so impressed, others seem less so and automatically blame the tool. I know it's likely my choice in setup that is a problem, so I'm happy that everyone pointed me to the real problem.

Speeds and feeds on most other operations (roughing, grooving) are about 1/2 to 1/3rd what this insert recommends, and I've been going with Sandvik recommendation because it does work so well on pretty much everything else. I thought this was an insert designed for finishing so needed to be run that fast.

Insert is here:
DCMT 3(2.5)2-MM 215

And a pic attached. Why does it seem like people automatically assume I'm lying? You can see there's a bit of variance between what they recommend on the web vs the package, even though package of inserts was bought brand new off Amazon.

View attachment 241466
 
Last edited:
I don't think you are "lying" take it easy. I doubt Ox even cares that much to think something about you or not. I followed Sandvik's recommendation with your Tool and cut parameters and yeah, what you posted, is what they recommend. It is way too fast!! Sandvik can suck on that for a minute.

Because a Tool is designed for Finishing doesn't automatically mean run higher Spindle speeds. You have to know some of the dynamics of Finishing, not just blanket it with go fast. If oyu follow the calculator notice how many Newtons it's taking to do what you are asking. I doubt your machine can push 26 Newtons at 3.25" Diameter, that's where the Chatter is coming from.

Were it me, I would slow it down to 400 SFM, leave .004" for the Finish pass (on the Face and the Taper) and I would start the Tool at Center of the Part and drag it out to the OD. I'm assuming your China machine has Coolant and the mixture is above 5%.

R
 
Maybe he took my "pro" comment as a dig, but actually the opposite was the proposed intent.

Meaning that to git those good of results under the parameters presented, you are VERY lucky to get anything as good as you have there. I would have expected worse results.

We are all saying that - regardless of what the MFG says to doo with those - we feel that you need to cut your R's in half and try aggin.

If you want to run the fast speeds, spend the next week conversing with a tech from Sandvik and not us.


------------------

Think Snow Eh!
Ox
 
Yeah this was a plan at some point, the Expanding 5C seemed like the ideal solution (once I can get a 5C pneumatic closer).

I tried a smaller expanding mandrel in the past, it didn't work at all for work holding (once spun in 3-jaw, another time part just tilted), but I was only expanding on the smooth bit, not the threads. I think if I get a larger one and thread up to a shoulder same as I'm doing now, just with an expanding bit it would be much better.

My turret is an 8 position, but it's basically a scaled down version of the standard ones on larger lathes. So while it works and all, clearance is really tight due to the diameter and this is what leads me to need a decent amount of stick out.

Pics attached, it's in the mill right now just for engraving, but normally I chuck that up in the 3-jaw.

Thanks for all the help! At least now I know what to try to fix to solve this issue.


Somehow I missed this post. :confused:
(and 90% - the pics!)

Are you chucking on the turned portion of that shank?
If so (as it's the only reason that I can think of it being that way) - why?
Is that max D of your chucking ability?
Not able to git a hex collet that size? (or just trying to git by without?)

Even if you are, and need to chuck on the turned portion, why is it not turned all the way?
So it would seem that your arbor is sticking 1.5" (?) out from the chuck or collet face?

Now that I see the back side of the part - you are going to have bad harmonics in that part it'self, let alone the other param's.

What would be best is if you could set the part back tight up agginst the chuck face or ??? for support from that direction. Even better if you have a pull-back collet that it actually pulls the part up agginst the closer face!

If not possible, a more better, but slightly more involved option is to make a more better arbor for it.

Your arbor should come out to the biggest diameter on the back side, and preferably stick out .005 (?) further than the surface in the center of the part so that when you tighten it up, that outside surface is seating first and put in slight tension agginst the arbor.

This aint gunna help that 5/8 x 2" stick-out tool, but at least you can git the harmonics out of the part it'self!

Now I am REALLY amazed at the results that you have gotten thus far! :eek:


---------------------

Think Snow Eh!
Ox
 
Thanks guys, I'll try slower back closer to my roughing SFM, which is around about 450.

The turned portion of the shank was just to fit further into the chuck body...I would like to shorten up my fixture even more. I don't yet have a collet chuck, so that hex bar is just being held in the 3-jaw. Clearance will be the issue for my tool stick out, I'll try to see the minimum I can go and still be able to face the part. I have the 3/4" through hole, so it doesn't need to face all the way down to X0.

Had I known what I know now (at the time I believed this machine had 3/4" tooling) I would have stepped up a little larger on model. For me, I would have loved a better brand lathe, but since I had never machined any metal prior to my purchase, it was the risk involved that kept me from getting a Doosan or Takisawa (as well as cost). $20k plus freight/rigging was a lot easier to risk knowing I may have to scrap the whole project.

Since getting some confidence on the lathe, I have gotten a Brother Speedio and certainly the quality and support difference can't be compared. Even still I continue to be impressed by my "spoon" of a lathe.
 
Man you guys run fast, roughing 304ss at 450sfm and finishing at 900SFM. I know I could never making my stuff last running that fast. I don't know about anyone else but I've never found the insert manufacturers recommendations remotely accurate. Maybe to make one part but not in a production environment.
 
There is a whole lot less "risk" in buying used.
Here is what $20K buys:

eBay - Page Not Found

There is a whole lot LESS risk in buying most any of these.

Any of these (that run) will produce your part a whole lot better, and if'n it don't werk out, you can expect to re-sell it without too much loss.

What's the resale value of that thing that you got there?

The only thing that I know of that uses 5/8 stick tools is a swiss with maybe 3/4" capacity at the most.
(read - no stick-out)

I just will never understand the thought process that ends with buying new junk over used real stuff for the same $.

BUT - I still think that if you make an arbor that will pre-tension your part, you will have better luck.


Well, IDK why the link doesn't werk, but here is the cut/paste option:
just take out the spaces...

e bay . com /sch/i.html?_from=R40&_nkw=CNC+Lathe&_sacat=0&_udlo=9%2C450&_udhi=19%2C700&_sop=16


---------------

Think Snow Eh!
Ox
 








 
Back
Top