Issue with part pulling off sub jaws at transfer

[Ju1ien]

Hi guys,

I'm currently using an Okuma LB3000-EX2 lathe with an S30 TRUE LENGTH chuck on the main (ATS 782AS1-C321) and 3 jaw on the sub (ID jaws). My issue is during the transfer, when the sub (W axis) comes into the part and seats against the wall and clamp on the part, when the MAIN jaws OPEN the part will pull OFF of the sub jaws by anywhere between .001 and .006. This seems to be an intermittent issue. I've seen it get slightly better over time running the machine, and it seems worse in the morning. Adjusting the S30 collet chuck sleeve location on the thread seems to have some effect, but nothing dependable or repeatable.

I'm trying to understand what's going on mechanically here and wondering if anyone has had this issue. It seems especially bad when trying to clamp on the sub jaws with minimal pressure (~100lbs).

Any insight very appreciated!

 

[wmpy]

The part is pulled back towards the main spindle when the main spindle collet opens? That's weird. You're transferring a part or pulling a bar? What condition is your collet closer in?

 

[eaglemike]

Is the collet fully open before the sub/part retracts? If the collet is sticking closed a little, can you program a short dwell, and then a tiny movement toward the main to be sure it's released? Sorry if this isn't appropriate or a good idea for you. Just an idea. I have my first dual spindle on the way, will be here Tuesday.

 

[XanderB]

In my experience:

Assuming it is not an issue of debris interfering with the part seating or clamping...longer dwells are sometimes required at lower clamping pressure. Possibly increase both the main dwell (during unclamp) & especially the sub while clamping.

Intermittent sounds like debris to me.

 

[Ju1ien]

The part is pulled back towards the main spindle when the main spindle collet opens? That's weird. You're transferring a part or pulling a bar? What condition is your collet closer in?

Pulling a bar. The sub comes in, syncs with the main, stops, then the sub jaws move onto the part and clamp. After the clamp, the main jaws OPEN, and the moment they open the part pulls back .001-.006 away from the sub jaws. The bar pull and transfer is otherwise perfectly normal, and no dwell anywhere would make a difference. It happens when single blocking.

I have removed the S30TL head and taken it apart, cleaned it and re-greased it. The spindle liner is clean and free of chips or obstruction. The bar itself is +/-.005 of the diameter and moves freely through the open collet.

 

[Vancbiker]

Is the bore that the sub jaws are holding a through hole? If shouldered or blind, are the jaws contacting a radius at the bottom?

 

[Ju1ien]

Is the bore that the sub jaws are holding a through hole? If shouldered or blind, are the jaws contacting a radius at the bottom?

It is a through hole, and there is clearance. The part will seat perfectly against the sub jaws before the main jaws open. The moment the main jaws open, it pulls away very slightly.

 

[Mtndew]

It is a through hole, and there is clearance. The part will seat perfectly against the sub jaws before the main jaws open. The moment the main jaws open, it pulls away very slightly.

Same machine here, same issue.
Are you using the default transfer settings that are in IGF? We haven't changed them.

 

[bigjon61]

Put an indicator on the master collet on the main spindle with the jaws clamped. Unclamp the chuck and note any movement. I bet your collet is moving forward when the chuck is closed and the slight bit of movement drags your bar back before the collet lets go.

 

[pcasanova]

Pulling a bar. The sub comes in, syncs with the main, stops, then the sub jaws move onto the part and clamp. After the clamp, the main jaws OPEN, and the moment they open the part pulls back .001-.006 away from the sub jaws. The bar pull and transfer is otherwise perfectly normal, and no dwell anywhere would make a difference. It happens when single blocking.

I have removed the S30TL head and taken it apart, cleaned it and re-greased it. The spindle liner is clean and free of chips or obstruction. The bar itself is +/-.005 of the diameter and moves freely through the open collet.

dwell will make a difference. if it runs better after hydraulics have warmed tells me the chucks are sluggish with cold (thicker fluid) and not getting enough time to get the movement complete.
put a 1 maybe 2 sec after the sub clamps
open main, another dwell make sure its open before you pull the sub
clamp another dwell for a sec then part off

 

[garyhlucas]

Since you say it is happening in Single Block then that suggests that observation of what is moving is incorrect. Mount an indicator to the machine frame and measure movement of the main collet, sub chuck and part, and even the tailstock, and the one that actually moves is the culprit.

 

[XanderB]

It is a through hole, and there is clearance. The part will seat perfectly against the sub jaws before the main jaws open. The moment the main jaws open, it pulls away very slightly.

Maybe the sub jaws have a taper? Allowing the part to slip off once the main releases? And/or the bore has taper as well. Does your transfer process use any dwells? Surely... I only ask because you don't feel it would help.

 

[XanderB]

It is a through hole, and there is clearance. The part will seat perfectly against the sub jaws before the main jaws open. The moment the main jaws open, it pulls away very slightly.

Is there an undercut at the shoulder of the sub jaws? Could make a difference.

 

[Ju1ien]

Put an indicator on the master collet on the main spindle with the jaws clamped. Unclamp the chuck and note any movement. I bet your collet is moving forward when the chuck is closed and the slight bit of movement drags your bar back before the collet lets go.

This seems to be exactly what's going on. I indicated the front of the collet during open/close with the bar in the spindle.
If I screw the collet completely to the top of the stroke where the bar is tight going through even when open, then the collet actually moves forward approx .004 during open/close. When I went to the other end of the stroke the collet pulled back about .006. There's a sweet spot about a turn and a quarter away from the top of the stroke where it seems to push the collet forward about .002. Will run parts and see if that fixed it!

 

[Ox]

Well, I think that most folks here missed the part of the "True Length" collet chuck on the main.
I am not familiar with this chuck, but it sounds like this is a "Dead Length" chuck eh?

Typically the collet body on these is floating, with a spring keeping it FWD, but even with it perfectly clean and lubed, this set-up is likely to have a wee bit of relaxation of the collet when the chucking sleeve is pulled back, and the bar will likely want to follow that for a bit until the collet loosens enough to let the bar slide.

As long as your main collet has a better grip on the bar than your sub has on the blanked part, this could always be an issue. Flex in the machine would actually be a good thing at this juncture.


If you were to go a Pull-Back collet system, that would not pull the part out of your chuck, BUT - if you were not completely seated to the stop on the sub, it would PUSH the part INTO the sub as much or more than you are seeing it pull out now.

Is there any reason that you can'rt put a collet chuck on the sub?
You say that you have low pressure doo to thin wall, well a collet would Shirley help this situation for sure!
This would be your most likely "best fix" for this application.


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

Think Snow Eh!
Ox

 

[Mtndew]

This seems to be exactly what's going on. I indicated the front of the collet during open/close with the bar in the spindle.
If I screw the collet completely to the top of the stroke where the bar is tight going through even when open, then the collet actually moves forward approx .004 during open/close. When I went to the other end of the stroke the collet pulled back about .006. There's a sweet spot about a turn and a quarter away from the top of the stroke where it seems to push the collet forward about .002. Will run parts and see if that fixed it!

We spoke with Gosiger last night about this issue, they said to change Parameter (Long Word,Word,Bit) Long Word #26 to 70. They don't recommend going higher than 70.
So, in the IGF pickoff cycle, you have 3 default values (30, 25 and 5). We set those all to 70 and it did help. It's now only .0005" away from the face.

 

[Ju1ien]

We spoke with Gosiger last night about this issue, they said to change Parameter (Long Word,Word,Bit) Long Word #26 to 70. They don't recommend going higher than 70.
So, in the IGF pickoff cycle, you have 3 default values (30, 25 and 5). We set those all to 70 and it did help. It's now only .0005" away from the face.

This is for the torque sensor?

Anyway, thanks for the replies guys. Eventually what did it was brass tipped grub screws screwed both into the channels of the sleeve as well as on top of the thread. Really torqued them down in order to get the collet to not move at all.
I always called it "dead length" but they are referenced as "true length" by suppliers.