[scpost]

Hello,

we have PUMA TT1800SY mill turn machine with two spindles. I have one question about part transfer from main spindle to subspindle.
When we need to pull part with subspindle out of main spindle before cut-off it is better to stop spindles or both spindles can turn
during pulling the part from main spindle.

Can second approach cause some damage to part or machine, or bar feeder components?

I am in doubt because all videos i have found on internet shows bar pull process when both spindles are stopped.

[deadlykitten]

hy i can't provide codes for your machine, but pls take a look at this scenarions :

[ round stock pull of with 2nd spindle at 0 rpm, then continue machining with both spindles ]
... stop s1
... stop s2
... unclamp s2
... move s2
... wait, so s2 to be in position
... clamp s2
... wait, so s2 to be clamped
... unclamp s1
... wait, so s1 to be unclamped
... move s2
... wait, so s2 to be in position
... clamp s1
... wait, so s1 to be clamped
... activate syncronize mode ( s2 will be slave to s1 )
... rpm on s1 ( s2 will copy the rpm in syn )
... cut
... turret away
... deactivate syncro

the reason for the wait codes / delays, etc, is that they act as confirmation, for example to be sure that pull-off does not begin while s1 is still clamping

it is critical to understand the folowings :
... if machines confirms that rpm is 0, this does not mean that real rpm is 0, but close to 0
... if machine confirms that movement is completed, this does not mean that real movement is done, but almost finished
wait codes are used to avoid conflicts, increase safety

[ round stock pull of with 2nd spindle at rpm<>0, then continue machining with both spindles ]
... activate syncronize mode ( s2 will be slave to s1 )
... unclamp s2
... move s2
... wait, so s2 to be in position
... clamp s2
... wait, so s2 to be clamped
... unclamp s1
... wait, so s1 to be unclamped
... move s2
... wait, so s2 to be in position
... clamp s1
... wait, so s1 to be clamped
... cut
... turret away
... deactivate syncro

this one is faster, because it does no longer loose time with s1 stopping

finding the g-code for each action above is up to you

---

most nasty lathe crashes happen during transfer; some scenarions can be prevented if machine has monitoring functions, but some cases just happen

is good when 2nd spindle aproaches, to move in rapid to a safe clearance, then switch to feed in mm/min and monitor that travel, so to be sure that no extra load occurs :
... sometimes there is still a part in 2nd spindle ( it can simply stay in there, even if the jaws are open ), so adjust your travel to cover this case
... there may be hard chips, that will be squishedd between your part and chuck, so blow air/coolant, and if possible, check for overload during last few mm

sometimes a crash occurs, and only turret is checked, while also the 2nd spindle axis origin may have shifted; this happens when the service team does the mechanical alignment, and after that they rely for parameters declaration to the client, or to somebody else; an unskilled, or unaware operator may start that machine with wrong origins

sometimes, the cut off operation does not complete succesful, and here are these scenarions, when the 2nd spindle leaves :
... the part may remain still in s1, and the 2nd spindle won't trigger an error, because of whatever reason, and the turret may come and hit the material, especcially when the lathe is short; such a thing can be tested with monitor functions, and/or checked confirmation sensors for 2nd spindle
... the part remanins in s1 for a while when s2 is moving, then it snaps and remains in s2, but at wrong position, and when turret goes to 2nd chuck, it may hit it

is good to know that cutting between spindles puts a lot of pressure on the insert, because both chuck tend to push the material towards each other, so the insert is pressed in between; to prolong insert life, before cutting, while s2 is clamped, simply move s2 a few microns to a few 0.01s ... travel depends on real case, and it's purpose is to reverse the tension inside the material, from compression, to extension : thus there will still be tension, but the forces will tend to pull the mateiral away from the insert, and not towards the insert

in syncronize mode, spindles may not go above a limit rpm, while in free mode, they may spin faster; this is because in sync mode, the controller needs to be sure that spindles phases are matched, so to avoid twisting the bar; as a consequence, any rpm change in syncronize mode will take longer, so the insert will be in useless contact with the material for too long, and this leads to wear; a quick solution is to cut in constant rpm, not constant speed

when there is a barfeeder involved, but pulling is done with 2nd spindle, it means that the feeder will only advance one time / bar, then s2 will keep pulling, until a small remnant remains, that is ejected, and cycle goes again if coding is not ok, or some confirmation fails for whatever reason, you may end up with a very long remannat, or with bars pushed through s1 directly deep into s2 drawbar, etc, and the turret will think that it has to cut a face, and in reality will smash hard ... such scenarions are harder to debug, since they involve accest to barfeeder logic, and some barfeederes simply loose they origins

all above cases can happen, so is good to know what to expect, and be prepared; some things happen gradually, so is possible to see them as they go wrong, while others simply happen in an instant most nasty cases may happen with a low chance, like, for example, during mass production, may happen after several months, or while lights out machining

a lot of safety can be implemented if machine has monitoring functions

When we need to pull part with subspindle out of main spindle before cut-off it is better to stop spindles or both spindles can turn
during pulling the part from main spindle.

Can second approach cause some damage to part or machine, or bar feeder components?

so, if damage occurs, is not only caused by the rpm status there are many other things to consider

i recomand transfering for a while at 0 rpm, see how things go, adjust the code, then go for transfer at rpm <>0

be sure to know the upper limit of rpm during syncronize mode, and try to transfer at a smaller value; for example, for a random okuma lathe, for main spindle :
... max rpm is 4000 in free mode
... max rpm is 2000 in sync mode
... i transfer at 1500

if you wish to transfer at rpm <> 0, but you have doubts, begin with tranfer at slow rpm, like 200 for example, and try to estimate if it is worthy to tranfer at higher rpms

advantages are less power consumption ( no more spindle decc acc, etc ) and shorter cycle time, but in the bigger picture, such advantages are not always having a meaningful impact they mostly have for small parts, like gang lathes, etc that operate at really high rpms, while a big chuck 200-300mm, may have a less chance to be profitable / kindly

[XanderB]

I have done it both ways. If your going to part off mid transfer, then sync up and keep the spindles turning. If you running slugs and don't need to maintain part orientation... The it hardly matters. I do find that keeping the spindles turning during transfer helps to reduce debris getting in the wrong places.