C axis oriented transfer with cutoff
Hi guys
I will be trying in a few days to do something I haven't done before and would like to check the solution with ones that have experienced it already.
The machine is LB3000MYW and it has a specific profile loaded in the barfeeder which also means that both spindles have collet chucks with matching profile. My question is how would you program the transfer as both spindles must have matching angular orientation prior to the cutoff tool to cut the workpiece?
Re: C axis oriented transfer with cutoff
M151 will match the spindles speeds and angles. There is a second spindle zero set parameter that can be used to adjust the angle of the relationship. Adjusting of the parameter can be done “live” with the M151 engaged.
Re: C axis oriented transfer with cutoff
Clear enough, thanks for the quick reply!
1 Attachment(s)
Re: C axis oriented transfer with cutoff
hi gunda :) M151 targets S axis, not C axis
in C mode you have control of 0.001 rotation angle
in S mode, you can also trigger a comand M19 C19 or 19.1 or 19.01 or 19.001
if you go like this M19 C19, put the dial on the jaw, then M19 C19.001, the dial may not move
if you go like this M19 C19, put the dial on the jaw, then M19 C19.010, the dial may move
if you go like this M19 C19, put the dial on the jaw, then M19 C19.005, the dial may move
* thus you can program a 3rd digit, but M19 is not so sensitive as the C axis; however, i guess it is sensitive enough for your setup :)
ok, so :
... C axis can sync the phases with high accuracy, but it won't reach high rpms
... S axis + M151 syncs with lower accuracy, but it allows higher rpms
if you are looking for high rpm & high sync accuracy, then replace M151/M150 with M697/M696
whatever you use, M151 or M697, try lower rpms + low cutting specs ( low cutting force ), because the differences are increased by high rpms & high cutting forces
to adjust the spindle phases, look for these parameters ( try para \ other function ) :
... spindle orientation zero offset
... 2nd spindle orientation zero offset
pff, i just crashed my thermos :)
you can find attached pdf, about workpiece transfer program
check also this tip from mr Wizard / kindly :)
Quote:
Originally Posted by
OkumaWiz
Your chuck has clearance in it in order to operate. As it clamps, the tolerances are taken up and the part moves slightly in Z. This shows up as motor load as the axis sees the motion. This occurs even on a dead length collet chuck. That is why after clamping, you may want to back off slightly with a W axis to avoid pinching the cuttoff tool when the load is released. Use the load meter to determine how far to back off. Move in W+ until the load decreases to zero and then starts to climb again to about 20%. Then record the amount W moved. Put a G91G94 W.00xx F50 in your program to accomplish this after clamping. It will make your groove tool last MUCH longer
Re: C axis oriented transfer with cutoff
Good remarks deadlykitten! OkumaWiz also has the right
The company that owns the machine produces door locks and they have brass bar with the geometry of a locking mechanism so I doubt I can be as precise as microns when I try to match the sub C axis angular position. Just another quckie - if I try like in the link below I believe I only need to use M19 prior to the chucking and then program M151 and M3/M4, right?
https://www.youtube.com/watch?v=JY4Uk0uTAR8
Re: C axis oriented transfer with cutoff
Quote:
OkumaWiz also has the right
mr wizard is always right :)
Quote:
brass bar with the geometry of a locking mechanism
hy / if the jaws were milled, then is good that the finish cut to be done with a tool that had been alligned before, thus to be sure that it is coaxial with the spindle
obviously, this is ok to be done for both spindles
otherwise, you may end up with almost identical jaws geometry ( on both spindles ), but excentric; you can still clamp the part ok, if the excentricity is the same, thus if it has same value and same spindle phase, but this is pure luck :)
Quote:
so I doubt I can be as precise as microns when I try to match the sub C axis angular position
it is possible, maybe not when you try to reach 0.001, but somewhere arround 0.025-0.030 should be easy / just saying :) is easier to align the machine after crafting/finishing the jaws, and a bit harder if the machine has worked for a while, because there is wear on the jaws, but ... whatever :)
Quote:
I only need to use M19 prior to the chucking and then program M151 and M3/M4, right?
allign you spindles using the parameters from "other function", and after that run a trial program at slow rpm, like 40-50 ( there is a minimal value, behind which you can not go )
don't worry, you may use the torque functions to protect your machine / kindly :)
Re: C axis oriented transfer with cutoff
one more thing : M151, M697, or whatever code for syncinc oposing spindles, has a huge chance to be there only for the constant rpm mode, thus you accelerate both spindles at same rpm, and after that you issue the M code, then clamp the part
it may not be ok to change the rpm while an M sync code is active, or at least trigger small changes in rpm
i don't know how okuma behaves, but on some machines, with dual collet chucks, the collet may rotate inside the chuck, because of a sync lost
some machine tools providers do not recomand syncing above a specific rpm ( lower than normal spindle rpm limit ), so to still be sure that the sync is in 'full' effect
generally, when an error of out-of-sync appears, it may be too late
also, the program shared in a previous post do not uses rpm changes, but only chuck intelocks ( M185 , M247 ), and this means that that code requires that both spindles are allready spinning at same rpm, before M151 is executed
there are persons trying to gain "precious" seconds, by using high rpms when a lathe is in a dual-state mode, and sometimes they hit the rpm limit of the sync mode, and some controllers (<> osp) allow using an interlock to ignore this warning, so to achieve higher rpms, but with less sync accuracy
kindly :)
Re: C axis oriented transfer with cutoff
The Okuma is the best that I've seen in the industry for synced rotation. It can sync when stopped, while rotating, can start and stop while synchronized without losing sync.It will work in either G97or G96. It can go up to the max rpm of it's slowest spindle. It can have the phase angle adjusted by parameter or even in the program, and can even optionally sync both C's when they are active. They can also sync without phase angle to reduce cycle time. They definitely did their homework on this one. I've personally seen them sync pie jaws and bring them together like a lovejoy connection in any of the above situations. It also checks to make sure that at least one chuck is open when syncing to avoid "twisting" the part during sync.
Re: C axis oriented transfer with cutoff
Quote:
Originally Posted by OkumaWiz
The Okuma is the best that I've seen in the industry for synced rotation ... They definitely did their homework on this one
hello mr Wizard :) i am glad to hear such news; a normal spindle do not react at rpms <40-50, thus a minimal rpm is required. Also there is not so much torque until circa500rpm. If both spindles can accelerate in sync, then there should be an improved motion controller for small rpms. Maybe something similar to the special function for " spindle extreme low speed cutting " ?
Quote:
Originally Posted by OkumaWiz
It also checks to make sure that at least one chuck is open when syncing to avoid "twisting" the part during sync
well, if it syncs so good, then it should definetly do it with both chucks closed. A spindle accelerates at circa175% load, at least for some 3000 models, and maybe, when both chucks are clamped, the acceleration time is greater, so to reduce required acceleration load ... so to avoid breaking the neck :)
Quote:
Originally Posted by OkumaWiz
Your chuck has clearance in it in order to operate. As it clamps, the tolerances are taken up and the part moves slightly in Z. This shows up as motor load as the axis sees the motion. This occurs even on a dead length collet chuck. That is why after clamping, you may want to back off slightly with a W axis to avoid pinching the cuttoff tool when the load is released. Use the load meter to determine how far to back off. Move in W+ until the load decreases to zero and then starts to climb again to about 20%. Then record the amount W moved. Put a G91G94 W.00xx F50 in your program to accomplish this after clamping. It will make your groove tool last MUCH longer
it is possible to code entirely that behaviour, so to avoid the manual setting time :
Code:
W @ approach position
torque skip until contact
record W position and check if position is good (*3)
clamp 2nd chuck
G29 PZ=15 (*1)
G04 F0.3
G29 PZ=10
G04 F0.3
G29 PZ=5
G04 F0.3
G29 PZ=0
G04 F0.3
G22 PZ=20 Z=small_positive_value F+15*5 G94 D G91 (*2)
record W position and check if position is good (*3)
NOEX G90
cut
(
*1 starting to reduce W load
sequential reduction should deliver smoother behaviour
maybe is not needed to reduce until 0 load, but it is needed
to reduce under the reversal-spike load, which also depends
on feed
*2 that code will output 15mm/min; such a smaller feed is needed, because
also the travel is short, and it should be executed slow, smoooth
*3 stats for nerds :) if implemented right, it will save the day
)
Quote:
Originally Posted by
deadlykitten
don't worry, you may use the torque functions to protect your machine
someone crashed his machine during a transfer; his code was ok, only that it hit into a case that does not frequently occur :)
part was still into the 2nd spindle when the transfer started, and the crash required service intervention, bla-bla-bla
so, to avoid this :
... use the torque skip function not only from a small clearance ( c ) in front of the 1st chuck, but from a greater clearance ( c+part_length )
... use load monitor, at least during W approach ( so to monitor entire travel )
... check if 2nd spindle is empty ( torque limit, load monitor ) before begining the transfer sequence / kindly :)
Re: C axis oriented transfer with cutoff
Hello deadlykitten
I'll be trying to make the code work tomorrow which I strongly believe I'll succeed in. However I am trying to make myself a note of those tricky moments you mention about. What do you actually try to show in the code example in your last comment? Do you think backing off slightly as OkumaWiz advises is beneficial?
I've done part transfers many times on swiss type lathe machines but using just a straightforward approach has worked for me so far - no workarounds with backing off the sub spindle or whatsoever ...
Cheers
2 Attachment(s)
Re: C axis oriented transfer with cutoff
hy gunda :)
i recomand :
... mrWizard's tip
... torque skip when 2nd chuck clamps the part + logging/writing clamp position to file + checking it to be within safe limits
... using load monitor along entire W travel
... clamping the part at low rpm, and after that raising the rpm and begin cutting
depending on setup length, how long will it be there ( hours - days - weeks, etc ), you may decide what to do :
... a longer setup is justified for using more safeties, because stuff may happen during the long run :)
... a shorter setup may only require operator surveillance, because adding extra code requires too much time
* in time, if you build parametric custom safe code, you may configure it fast also for short setups :)
Quote:
Do you think backing off slightly as OkumaWiz advises is beneficial?
of course :) mr Wizard had been in this bussines for long, and he know what he is saiyng
imagine that you clamp inside the 1st chuck a piece of plastic ( something softer ), and you start to cut it : all should be ok, the part will fall, etc :)
now pls imagine that you also push it with the tailstock, with huge pressure : when the cutting tool will be close to center ( but not yet at center ), the part may get moved by the tailstock, and it may end up like in image 01; because is plastic, nothing will happen :)
now replace plastic with metal : as a result, tailstock live center may get damage, tool insert & tool shank may get damage
all these happened because the part was tensioned, it was compressed :)
the 2nd chuck may actually compress the part :
... if compression is big, part will fly off during cutting ( this is rare )
... if compression is low, it may shorten the insert life, because it will be pushed sideways; in this case :
...... if insert life is long enough, it is possible to ignore this phenomen ( this is why few people use this technique )
...... if insert life is short, or if you have been arround cnc machines for long time, then you may notice this phenomen
one more thing : if you keep a record of insert life spam, and parts / edge, you may notice that insert life is longer when cutting without tailstock, or without a 2nd chuck :) this is easy to say, but in reality it really requires time to gather all this data
Quote:
I've done part transfers many times on swiss type lathe machines but using just a straightforward approach has worked for me so far
ookey, i would like to say that, even if things work, this does not mean that there is no problem : there are problems with spontaneous effect, and problems with effect manifested among long time
swiss lathes generally use collets, and the collet has the tendency ( during clamping ) to drag the part inside, into it, while a jaw chuck has the tendency ( during clamping ) to spit the part out; as a result :
... a part clamped between 2 chucks has a big chance to be compressed
... a part clamped between 2 collets has a big chance to be stretched
... when you open a collete, the part may shift towards, or fly out at speed, like a mini-bullet
... when you open a chuck, the part may shift inwards
there are special chucks that try to drag the part towards them when closing, so to achieve higher repetability and minimeze the tendency to spit the part out :)
on a swiss i had experienced an uneven wear on the cutting insert ( image 2 ) : if i inspected the insert with a magnifier, i could see that one corner was dulled very much, while the other corner was ok; the difference was hard to notify ... even so, the insert would last for more then 8 hours, and parts would seem acceptable ... in the end i solve it ( reduced tool hangout, thicker insert, etc )
Quote:
What do you actually try to show in the code example in your last comment?
the method presented by mr Wizard requires a bit of set-up time, that needs to be done in manual mode; that code is doing just that, thus it is adjusting the loads in automaticall mode, so there is no longer need to adjust the setup in manual mode ... it saves a bit of time / kindly :)