Cutting undersize, what to check on Hercus 9” Lathe

[neevo]

I’ve just finished an install of a DRO on my Hercus 9” C lathe and I’m noticing some weird behaviour that I need some help to further test. I suspect having a DRO has highlighted a problem that has always been there but now I have the ability to easily measure it, it’s become apparent.

Issue: I’m finding my lathe is cutting smaller than I’m reading on the DRO. To set my diameter I take a cut, leave the crosslide where it is and enter that measurement into the DRO. Then as I reduce the diameter, the actual diameter drifts away from the DRO and I end up overshooting the dimension by up to 0.2mm on a 10mm diameter.

What have I done.

1. I’ve tightened up all the gibs as one of them on the crosslide was loose after the install.
2. I put an indicator on the tool post to measure the actual change vs the DRO. This was perfectly in sync over 6mm but I did not check more than that as I assumed it was good.
3. I have checked my spindle bearing clearance and found it’s a bit loose. 0.04mm when I pull up (my lathe has plain bearings). Whilst I can imagine this doesn’t help, it seems too small to be the primary cause of my issues

What could be going on? What tests can I run to try and understand the cause of my issues. Am I using the DRO wrong? Not being able to rely on hitting a dimension is really frustrating, especially when I’m going smaller.

I thought the bed could be twisted but that seems like a huge twist to be missing dimensions by that amount. I’ll likely make a test bar tomorrow though to triple check.

 

[technocrat]

how much slop is in your cross slide leadscrew nut? Put an indicator from the bed to the cross slide and push the cross slide in and out.

 

[EmGo]

Tool is not on center or the cross slide isn't at right angles to the bed or the scale is mounted wrong - something is geometrically kerflooey because if everything is on center and square, with a dro this is not possible. DRO measures actual position, if it's mounted right.

 

[Bill in PA]

Another vote for tool not on center.

 

[neevo]

how much slop is in your leadscrew nut? Put an indicator from the bed to the cross slide and push the cross slide in and out.

0.24mm

 

[neevo]

Tool is not on center or the cross slide isn't at right angles to the bed or the scale is mounted wrong - something is geometrically kerflooey because if everything is on center and square, with a dro this is not possible. DRO measures actual position, if it's mounted right.

Yeah that’s why it’s bending my brain. Particularly when I have confirmed it’s moving perfectly and the correct distance when I’m measuring the tool post travel with a DTI.

Pretty sure I have the tool on centre height too. It’s removing all the centre nubbin when facing too.

 

[neevo]

If I have a bit of stick out, could part deflection cause the part to turn smaller than expected? I would have thought it would turn bigger.

10mm diameter with 40mm stick out.

 

[Limy Sami]

If I have a bit of stick out, could part deflection cause the part to turn smaller than expected? I would have thought it would turn bigger.

10mm diameter with 40mm stick out.

4 x D is about the limit - especially on small dia's ** What dia are you turning down to? What is the top take on the tool?

**Yes I know dia shouldn't have any affect but IME in real life it does.

 

[neevo]

12mm rod, turned down to 10mm with 40mm of stick out. Using a CCMT insert.

 

[EmGo]

Easy enough to tell - put a enter in it.

 

[eKretz]

Yes, put a center in and cut something at a larger diameter so it's stiffer. Your tests so far are inconclusive. Tool pressure can easily move a size that much, especially with flimsy material and if your tool is above or below center. Being high or low is effectively decreasing or increasing the tool rake, which can absolutely cause the tool to push away or pull in, especially so with heavier cuts and/or flimsy material.

 

[farmersamm]

Sometimes, on a less than rigid lathe, CCMT inserts will rub to some extent instead of cutting.

If you're getting a healthy spiral gouge when moving the carriage back to the right after taking a cut, it means that you're seeing tool push off. It didn't cut to the desired depth on the first pass. Try a spring cut to see just how much you missed on the first pass.

Because the insert isn't taking the correct amount, you will set the cross slide for the next cut, not knowing that you didn't get the full depth on the previous cut. At this point sometimes the insert will dig in, and actually take more than you wanted to take. CCMT's are sometimes unpredictable on small machines, or light cuts. You can get CCMT finishing inserts that perform better than medium, or rough cut, inserts.

If you have them in the shop, try some TCMT inserts, and see how they perform on light cuts. They're generally better at taking small cuts. 1/64 radius being the best. TT inserts also excel at light cuts. TT's are the bomb for small machines.

Also try some larger stock to see how the lathe performs. The small stuff you're cutting is sometimes hard to deal with, in terms of springing.

 

[farmersamm]

I gotta lathe that's akin to a rubber band. I've found that insert geometry is the key to success.

A well ground HSS tool is best, but it doesn't mean that you can't use carbide inserts.

Your CCMT's are good positive rake inserts.

Even better are the Txxx inserts.


TPGH inserts on a Dorian bar.


TT insert on a Borite turning tool.

You can get a lot done with small machines if you figure out how to smoogie them.

 

[hvnlymachining]

Along the same line as the previous comments, deflection or varying cutting tool back pressures can cause a lot of trouble.

Try a sharp cutting tool, make the same relatively shallow depth of cut for three passes measuring carefully after each pass, are those passes consistent?

 

[technocrat]

0.24mm

So as you rotate the handle to move the cross slide, the thread is pushing the nut. As the tool touches the work, it digs in and pulls the cross slide to the front of the thread. The dial still shows the same place but the cross slide has moved closer to the center. You might see this on the DRO if you watch when you touch.

 

[AA+]

Lathe tool cuts diameter. DRO reads radius. 2:1
Best wishes --- Allen Anway

 

[hvnlymachining]

Lathe tool cuts diameter. DRO reads radius. 2:1
Best wishes --- Allen Anway

Most DROs can be set to read on diameter or direct read mode.

 

[neevo]

A few things based on the comments.

1. DRO is definitely set to diameter mode. I’ll try it on radius to see if it’s a read out issue.
2. I’ll try putting a center in and taking a cut and see if that makes any difference.
3. If the tool is digging in and or using up the backlash in the crosslide nut, wouldn’t I see that on the DRO as the read head should register that movement? I’m not seeing that at all, the DRO holds the same value. Unless it’s pulling the compound. I’ll try locking that down further today.

The odd thing is that the part is cutting undersize. I would have thought that part or tool deflection would cause it to cut oversized.

Today I’m going to try:

1. Locking down the compound gibs as much as I can
2. Testing the tool post movement relative to the DRO over 25mm and make sure it’s working properly over a much bigger range
3. Try tail support and see if that does a better job. Plus try some shallower cuts
4. Adjust my spindle to headstock bearing clearance. I think my clearance is way too big

 

[hvnlymachining]

neevo said;

"If the tool is digging in and or using up the backlash in the crosslide nut, wouldn’t I see that on the DRO as the read head should register that movement?"


Yes, you definitely should. That's what makes a DRO so useful. ( Among other things)

 

[neevo]

neevo said;

"If the tool is digging in and or using up the backlash in the crosslide nut, wouldn’t I see that on the DRO as the read head should register that movement?"


Yes, you definitely should. That's what makes a DRO so useful. ( Among other things)

That’s the weird thing. I set my DRO to the diameter I want to cut. It doesn’t move the whole cut, but when I measure the new diameter it’s smaller than the DRO.