Vibration due to orientation of tooling

[RaisedByWolves]

You're right, some pictures would help.

No chatter
View attachment 496738

Chatter
View attachment 496739


Measuring with the carriage locked, I'm getting +/- 0.003 in the direction of the cross slide. Same results if I push on the tool post or on the tailstock.
View attachment 496740

Maybe it's torsion in the bed?

I had half anticipated this problem and made a weldment to mount the lathe on. The idea is to "bed" the lathe to the weldment using Moglice. The weldment was specifically designed to resist torsional forces and is constructed from 1/4" sheet. Very heavy. Probably overkill.However, the weldment is not quite not finished and I need to finish this job before tearing the shop apart.

So much info in those pics.

First, your approaching or exceeding the capability of this machine. Your part is larger than your machines spindle and therefore its bearings. Any play in the bearings will be amplified.....how far away from the headstock, 4-5" therefore even further away from the closest bearing?

Second, you should be running more like 250-300rpm with that part in that machine.

3rd, carbide is not your friend here. If you're running 1000rpm I can see why you're using carbide, but either way, slow it down.

Is your insert on center?

The basic problem is the force of taking the cut is overwhelming the ability of the machine to keep everything in line. The work is trying to climb over the insert hence the chatter and this is why it cute better with the insert face on an angle.

Its literally pushing the bearing races and balls closer together via the axial load, hence less chatter.

ETA: I just saw the chips in your tray....Holy cow, on that machine?

That's aggressive.

 

[C-Bag]

Chatter/no chatter is a very deep and involved subject on a 9x20.

I don’t use carbide unless dealing with something really hard. I use HSS and sharpen often when chatter rears its ugly head. When I couldn’t deal with chatter no matter what I did I happened upon Savarin’s post about the fact that it’s impossible to adjust the headstock bearings as it came from China. Even though it’s a knockoff of an Emco it lacks the attention to detail. One major one was that the back tapered headstock bearing is a press fit just like the front bearing. That’s not how the original Emco was. The front bearing(closest to the chuck) is a press fit but the back bearing needs to be a SLIP FIT, not press fit. This is why it’s so hard to set the pre load on the stock head. As it gets hot it will change. Once I pulled it and ground it to a press fit(no radial play) it made all the difference. Now I can set the pre load correctly. Beside getting my grinds right on the HSS bits and making sure there was little play in the cross slide and a number of other details I can now work chatter free. YMMV because the list of possible problems with a Chinese lathe are varied and endless.

 

[WobblyHand]

I have both a 7x16 and a 10x22, which bracket a 9x20. In my humble opinion, it's possible to do decent work on both of them. But you do have to live within the limits of the platforms.

This is how I recommend the tool should be placed. In your case, I'd reduce the stick out to the minimum. I use TCGT inserts most of the time, except when doing rougher stuff, then I use TCMT. The TCGT inserts give a nice finish. They also reduce cutting forces, which can reduce chatter. I just went into the shop to stage these two photos. This tool angle faces well.

Here is the big view of my 10x22. FWIW, I made the ER40 set true back plate (that's in the photo) from a raw chunk of cast iron on this lathe. Used TCMT inserts to peel away the rough outer layer, then switched to TCGT for all the rest. Single pointed the back plate's 8 TPI thread on this lathe as well. I'm using 3/8" tooling so I can swap it between my mini-lathe and the 10x22.

Try this tool orientation in the first photo for facing before going down the rabbit hole of other things. It might minimize the amount of chatter you get. If you can find some GT inserts check them out. Check your gibs as well - sometimes they loosen.

 

[682bear]

You are using a 321 insert... which has a radius of .016" and taking a .020 depth of cut.

The pic with no chatter, the insert is oriented where it is cutting on a 'positive' angle... or 'pushing' the chip toward the workpiece... if that makes sense.

The pic with chatter, the insert is cutting a larger DOC than the radius... the chip wants to ride along the backside of the angle... rather than be cut off cleanly. In effect, it is tending to sort of 'pinch' or 'squeeze' the tip of the insert, causing more drag... and more chatter.

Reduce the DOC to .012 or try a 322 insert, which has a .032 radius. The 322 should take a .020 DOC if the lathe is rigid enough to use it.

-Bear

 

[durableoreo]

First, your approaching or exceeding the capability of this machine. Your part is larger than your machines spindle and therefore its bearings. Any play in the bearings will be amplified.....how far away from the headstock, 4-5" therefore even further away from the closest bearing?

Second, you should be running more like 250-300rpm with that part in that machine.

3rd, carbide is not your friend here. If you're running 1000rpm I can see why you're using carbide, but either way, slow it down.

I wish I had space for a larger lathe. Yes, with the chuck, the cut is 7" in front of the bearing. It's a significant moment arm.

I'll try a HHS tool tomorrow morning, with the correct surface speed.

Is your insert on center?

Yes, I got one of those center finders from Edge Tech. www.amazon.com/dp/B01MQ4FHML Also, because it's a facing operation, I can see that there's no tit left in the center.

The basic problem is the force of taking the cut is overwhelming the ability of the machine to keep everything in line. The work is trying to climb over the insert hence the chatter and this is why it cute better with the insert face on an angle.

Its literally pushing the bearing races and balls closer together via the axial load, hence less chatter.

Good point. I'll try some of these other suggestions first, but what do you think more pre-load on the spindle bearings? Even if I'm sacrificing higher speed operation, it might be worth it.

ETA: I just saw the chips in your tray....Holy cow, on that machine?
That's aggressive.

This is from drilling, so the chips look more impressive than any turning or facing operations. The chips from drilling are shiny on one side, dull on the other. I filled 2 trash cans with chips today!

 

[RaisedByWolves]

I wish I had space for a larger lathe. Yes, with the chuck, the cut is 7" in front of the bearing. It's a significant moment arm.


7”, wholly crap!

I'll try a HHS tool tomorrow morning, with the correct surface speed.



Yes, I got one of those center finders from Edge Tech. www.amazon.com/dp/B01MQ4FHML Also, because it's a facing operation, I can see that there's no tit left in the center.

Nice! I usually go with a setup stick, but I’m a tightwad.

Good point. I'll try some of these other suggestions first, but what do you think more pre-load on the spindle bearings? Even if I'm sacrificing higher speed operation, it might be worth it.

Judging by what C-bag said I don’t think that is advisable without following his modification. Even at that you’re asking a lot of that design.

His findings do back up my theory of why it cuts better in one orientation than the other, ie it’s forcing the bearing races and ball into forced concentricity vs them being set perfectly.

This is from drilling, so the chips look more impressive than any turning or facing operations. The chips from drilling are shiny on one side, dull on the other. I filled 2 trash cans with chips today!

I was hoping those were drill chips.

Being 7” from the headstock the only suggestion I can come up with that hasn’t been mentioned is using a live center if at all possible.