What's new
What's new

A way to force a 1 degree B axis HMC into fractional increments?

NTM

Cast Iron
Joined
Sep 4, 2009
Location
Mooresville, NC
I have an early 2000s Mori Seiki NH5000 with a Fanuc 18i control on it. This machine came with a B axis that will index in 1 degree increments. Sometimes I need a fraction of a degree and I have to do this with fixturing which is a pain. Especially when I consider the machine should be capable of doing this?

Here's my reasoning. I can program a B axis move and pause the move at a fraction of a degree. Lets say I stop it at 1.567 degrees. I can then go into G91 (relative mode)and make moves in one degree increments relative to that 1.567 degree position that will continue to remain in that fractional increment, so I can get 2.567 and 3.567 and so on. So clearly the machine is capable of it. So what am I missing here? Is there a way to get a 1 degree indexing machine to go directly to a specific fraction of an angle? It would save me a lot of pain and suffering.
 
do you have drawings for the mechanicals? parts lists perhaps? to get an idea what sort of locking/holding is used for the axis, because there are some that won't lock at fractions, measuring and getting to stop it may be one thing, holding under load may be another...
 
On some Mori Seikis of that vintage you can split degrees. There were a few configurations. A 5 degree index, a 1 degree index, a 1 degree index and split degrees, and a full 4th axis. If you have the 1 degree and split degree configuration, at the 1 degree indexes the curvic coupling engages. At any position other than the one degree indexes you only have a brake so some rigidity is lost.
 
On some Mori Seikis of that vintage you can split degrees. There were a few configurations. A 5 degree index, a 1 degree index, a 1 degree index and split degrees, and a full 4th axis. If you have the 1 degree and split degree configuration, at the 1 degree indexes the curvic coupling engages. At any position other than the one degree indexes you only have a brake so some rigidity is lost.

Yeah, this one is set up for 1 degree, but it has demonstrated it will do fractional increments. I'm wondering if anybody has ever heard of tweaking a parameter in an 18 to achieve this?
 
Yeah, this one is set up for 1 degree, but it has demonstrated it will do fractional increments. I'm wondering if anybody has ever heard of tweaking a parameter in an 18 to achieve this?

I would look through the paper ladder, in the keep relay section. I have looked through mine (nh5000) and dont recall seeing anything about 1/2 degree.

But i was not looking for that, so.
 
So what happens when you try to run a program with a split degree command?

IIRC, there was an M code that had to be commanded to engage the brake after commanding a split degree movement. At the 1 degree indexes, the curvic coupling automatically locks at the index completion, but that is a different mechanism than the brake. Shop I retired from used to have a Mori MH630 with the split degree function, but it's been a long time ago so don't remember all the details.
 
So what happens when you try to run a program with a split degree command?

Normally I hear the hydraulic system activate a solenoid valve and disengage the brake and the sound of the servo index, then re-apply the brake, but a split degree does nothing, it just gives an error code.
 
If it alarms when you command a split degree, that means the option was not spe'd at the machine purchase. The solenoid you are hearing during a normal 1 degree index is not a brake, it is the curvic coupling unlocking and locking.

Is the error code an EX followed by a number?
 
My Miyano (old one) has 5º indexing. I can get it in 1º increments no problem haven't tried 1/2º.
so one day I was looking at it cause my angles were off a tad Turns out the solenoid assembly was loose, there's a sprocket on the back of the spindle has a v shaped solenoid it will index then shoot the v shaped piston into the sprocket to lock it. I was going to make a sprocket with 1º increments one of these days and try it out.
 
I don't believe you can with a 1 degree machine. The solenoid engages the curvic is the way I understand it. I don't believe there's a brake on the b, I think its just the drive holding position till the curvic engages. The company I used to work for had about 20 NH8000's and only a few were true B axis ones that I heard of. Had to be ordered that way. Also had about 25 633's and they were all 1 degree machines. When you pause the machine at 1.5 degrees will it let you manually index the b and restart the program?
 
If it alarms when you command a split degree, that means the option was not spe'd at the machine purchase. The solenoid you are hearing during a normal 1 degree index is not a brake, it is the curvic coupling unlocking and locking.

Is the error code an EX followed by a number?

I don't see a Fanuc side error message, but the front end says 3001 B command Error when I try to program a fractional move directly. Although if I simply interrupt a move at the desired angle and make future moves in incremental it will remain at that fractional angle and run.
 
When you pause the machine at 1.5 degrees will it let you manually index the b and restart the program?

Yes it will. The machine appears to be completely capable of doing what I need. It actual will do what I need if I "cheat" it. I'm wondering what do I need to do for it to simply go directly to a fractional angle without fussing about pausing it at the angle I need (which takes 100 attempts to time correctly).

The mill can do it, it just won't.

I'm beginning to suspect it's like a Fanuc machine I bought in the late 90s that could do helictical interpolation, but it just wouldn't (until a parameter was changed).
 
The word that scares me is "force." As in you are trying to make the machine do something it isn't designed to. If it were me, I'd get a little more creative in fixture design, or start scanning the ads for a machine with a full B axis. Anything else, is just a tad too scary.
 
The word that scares me is "force." As in you are trying to make the machine do something it isn't designed to. If it were me, I'd get a little more creative in fixture design, or start scanning the ads for a machine with a full B axis. Anything else, is just a tad too scary.

Yeah I hear you. And that's what I do now, creative fixturing. But a part with multiple moves would require multiple fixtures which is a lot of unnecessary handling.

I don't manually home this machine when it's turned on because it has absolute encoders. The machine will usually have a pallet change before running a program and this pallet change cycle will home the B. And, most programs have a G90 B0. etc. But, in situations where the machine is turned on and the tombstone I want to work on is loaded and facing the right direction I will need to home B because it can be off a few thousandths of a degree when it's turned on before its homed. This tells me that angle isn't being tightly controlled by a coupling. And the fact the machine is perfectly happy to run (and repeatedly index) with that whatever angle makes me think it isn't doing anything too particularly weird. So I've started thinking this might just be a parameter issue and nothing more?
 
I can give you some theories on why they dont want you to do half indexes on it without more hardware.

The indexer machines rely on the curvic to provide accurate positioning. The servo motor is coupled to the rotary via a pinion gear, not a worm.
There is not a scale on the B. There is only the encoder on the motor.

I know from firsthand experience that if you get the curvic to clamp without being in position, the clamping member will get out of time with the other two pieces of the curvic. The reason for this is that the clamping member is not restrained from rotating. So clamping while out of position makes the clamping element sit on top of the teeth of the moving member. On a properly adjusted switch this will give an alarm as the clamp element is too high when it is supposed to be clamped.

I have an SH403 here that the previous owner adjusted the switch so it would think it was clamped when it was out of time.
 
I can give you some theories on why they dont want you to do half indexes on it without more hardware.

The indexer machines rely on the curvic to provide accurate positioning. The servo motor is coupled to the rotary via a pinion gear, not a worm.
There is not a scale on the B. There is only the encoder on the motor.

I know from firsthand experience that if you get the curvic to clamp without being in position, the clamping member will get out of time with the other two pieces of the curvic. The reason for this is that the clamping member is not restrained from rotating. So clamping while out of position makes the clamping element sit on top of the teeth of the moving member. On a properly adjusted switch this will give an alarm as the clamp element is too high when it is supposed to be clamped.

I have an SH403 here that the previous owner adjusted the switch so it would think it was clamped when it was out of time.


Gotcha. I guess that settles it. I appreciate your input and everyone's time on this.
 
I can give you some theories on why they dont want you to do half indexes on it without more hardware.

The indexer machines rely on the curvic to provide accurate positioning. The servo motor is coupled to the rotary via a pinion gear, not a worm.
There is not a scale on the B. There is only the encoder on the motor.

I know from firsthand experience that if you get the curvic to clamp without being in position, the clamping member will get out of time with the other two pieces of the curvic. The reason for this is that the clamping member is not restrained from rotating. So clamping while out of position makes the clamping element sit on top of the teeth of the moving member. On a properly adjusted switch this will give an alarm as the clamp element is too high when it is supposed to be clamped.

I have an SH403 here that the previous owner adjusted the switch so it would think it was clamped when it was out of time.


Nicolas Correa design of a double hirth coupling that can get 18,000 positions in a circle or 0.02 degree positional increments,

but seems to swing a head one way and then clamp , then swing a different ratio plate (sort of a indexing Vernier effect) in the opposite direction for final "fractional" positioning. [I think ???]. International-ish patent [EP1 386693 B1].

Kinda off topic with OP's conundrum but also worth seeing what it can take to solve the problem in another way.

It is a little frustrating on machine like an integrex J series that indexing (although very precise and accurate curvic coupling) is in 5 degree increments when a lot of stuff / engineered parts require fractional degrees + a very rigid and repeatable machining process.
 
A couple questions.

Do you often need an odd angle, like 1.34? Or you just want 1/2deg increments?

Can you make a subplate of sorts milled at 1/2deg angle and bolt that onto the tombstone face?

Is 3d milling an option to get your angles, or are we talking holes and such? 3d is typically slower, but if you can eliminate a fixture and/or part load it might be worth it.
 
A couple questions.

Do you often need an odd angle, like 1.34? Or you just want 1/2deg increments?

Can you make a subplate of sorts milled at 1/2deg angle and bolt that onto the tombstone face?

Is 3d milling an option to get your angles, or are we talking holes and such? 3d is typically slower, but if you can eliminate a fixture and/or part load it might be worth it.


I didn't originally use this horizontal for multiple angle cuts, but when I figured out how to do it I enjoyed reduced setups and better accuracy. But only on 1 degree increments.

Take this chamber for example.

PuGWuWp.jpg



The end seals are at 60 degrees. <--- Yay!

But the doctor blades are inclined at 7.833 degrees. That's a two additional setups, and those inclined surfaces can not be at all skewed or the intersection of the cylinder becomes a conic which can not be sealed with a straight blade.

This part is six setups on a 3 axis vertical. It might be only two on a 5 axis. It's 4 on my horizontal but it could be only 3 if I could do sub fractional. Which apparently (despite doing it if I cheat it) isn't a good idea. It would save fixturing, handling and it would be more accurate. I guess I need to start saving up...
 
I didn't originally use this horizontal for multiple angle cuts, but when I figured out how to do it I enjoyed reduced setups and better accuracy. But only on 1 degree increments.

Take this chamber for example.

PuGWuWp.jpg



The end seals are at 60 degrees. <--- Yay!

But the doctor blades are inclined at 7.833 degrees. That's a two additional setups, and those inclined surfaces can not be at all skewed or the intersection of the cylinder becomes a conic which can not be sealed with a straight blade.

This part is six setups on a 3 axis vertical. It might be only two on a 5 axis. It's 4 on my horizontal but it could be only 3 if I could do sub fractional. Which apparently (despite doing it if I cheat it) isn't a good idea. It would save fixturing, handling and it would be more accurate. I guess I need to start saving up...

This all sounds like justification to use with the brass to trade up. Reduced setup time = lower cost. increased part quality.
 








 
Back
Top