[turmite]

I didn't know where to put this so I though here would fit the best.

does anyone here have any experience with 5 axis. I am primarily interested in the 4th and 5th axis at the moment and in particular which design is best for stiffness. I am not even sure of the proper names but I think it is trunnion (sp) and yoke. What type of system is used to give them their movements. I have seen pics of them that appears to not have a motor on them at all. If someone here has design their own 5 axis from scratch without using the sat units I would appreciate any input.

Mike

[Dan B]

Hi Mike,

We have extensive experience with both trunion style and rotating head type 5-axis machines.

The trunion style is by far the most rigid and accurate of the various designs out there. If the trunion position is fed back to the machine by use of glass scales, then the machine will be extraordinarily accurate. Our Hermle uses this system and we can put holes in all day long within .001". However, we are talking about a $500,000 machine here. The Haas trunion style machines don't use the glass scales, so the accuracy is not as good. And from what I hear, it degrades as time goes on. But to be fair, it is half the price too.

The rotating head designs sacrifice rigidity and accuracy. While some are better than others, I would be surprised to hear of any that can match the rigidity and accuracy of a trunion machine. Our 5-axis head is capable of around .004" accuracy.

Does this help?

Dan

[Evodyne]

Dan,

I'm sorry, but "trunnion style" to me doesn't mean a thing to me. For the newbs could your make reference to some pics of the different 5-axis configs, or perhaps attach a quick drawing? Like Turmite, I'm looking into what it would take to add this to my design. I was wondering how many different approaches there are to solve the same problem-and their merits/faults.

Lance

[Dan B]

Take a look here:

http://www.hermle.de/fs_hermle.php?sprache=en&page=168

navigate around a bit and you will see some drawings and photos of trunion style machines.

Dan

[turmite]

Dan thanks for the info and the link. Keep watching cause I'll be back!

Mike

[Evodyne]

Ditto on the thanks!

Lance

[Dan B]

No problem. Just ask away, and I'll do my best to answer.

Dan

[jcolley]

Mike,

I am in the middle of a design of exactly what you are talking about. My machine will be used heavily for aluminum and steel, and will experiment with "superalloys" if the rigidity proves to be sufficient.

My design is a modified gantry style using four upright columns to support the X and Z axis spindle assembly. The cradle assembly (I think is what you referred to as trunnion) moves in the Y direction on THK linear ways. The cradle has a rotary stage which positions the A axis +/-120 degrees. On the cradle, a servo driven ratory table handles the C axis.

I debated how to drive the A aixs for a while and decided to custom build the rotary drive stage using a standard fine pitch worm gear and dual worm pinions in an anti-backlash configuration. I plan to use dual loop feedback with rotary encoders both on th servo and on a rotary encoder directly connected to the A axis. The software end is posing the greatest challenge atthis point though, as dual loop is not readily available on an affordable hobbyist scale.

As I have said many times before, I base all my work on absolutely nothing and have zero building experience to this point, just some hair-brained ideas and Pro/Engineer Wildfire to hash it all out for me.

Jim

[turmite]

Dan and jcolley after re-reading your posts and thinking about what I ask...........I think I ask the wrong question. Trunion is not what I am wanting. I think Dan called it a rotating head. What I am looking for is ideas as to how to rotate the two axis and where to mount the motors? I think I can get the torque great enough by using harmonic drives but where do you put the motors?

Mike

[Evodyne]

Mike,

As I have said many times before, I base all my work on absolutely nothing and have zero building experience to this point, just some hair-brained ideas and Pro/Engineer Wildfire to hash it all out for me.

Jim

Well, you've got a lot of company! It's amazing how far a little common sense can take you though (as if I'd know). You'll have to keep us up-to-date: it sounds like an interesting project.

Lance

[jcolley]

If it is a dual rotating head you are looking for, check out this conversion ( 5 Axis ) for a Bridgeport style machine. The best I can tell (guess) is that they use cross roller bearings ( THK ) on the head assembly and using bevel gears on each side about the same size as the outer diameter of the bearings. These two bevel gears would be driven by a bevel gear up inside the stationary part of the head. They would then drive a fourth and final bevel gear on the spindle.

Now the actual axis motion would be controlled by a second set of bevel gears on the rotary part of the head, one for the rotating axis and one for the angular axis. I'd have to think about it for a bit, but it doesn't seem that hard to figure out.

I actually have no idea how the one in that link works, but that makes sense to me and if I designed it, that's how it would work.

Again, reference my first post for my actual qualifications to speak here...there are none...

Jim

[Evodyne]

Dan and jcolley after re-reading your posts and thinking about what I ask...........I think I ask the wrong question. Trunion is not what I am wanting. I think Dan called it a rotating head. What I am looking for is ideas as to how to rotate the two axis and where to mount the motors? I think I can get the torque great enough by using harmonic drives but where do you put the motors?

Mike

Mike,

Despite exposing my complete lack of artistic ability, I thought a scribble might stimulate a little conversation. No laughing! Even though this might well be a new low for attached "artwork", it's the thought that counts, right? BTW, it's still a little early to tell, but I think I scored a couple of harmonic drives! Boy will I be honked if I DON'T get them. Any how, take a peek and let's do some brainstorming...

Lance :idea:

[jcolley]

Man, a pictures worth at least how ever many words I just wrote...

[Evodyne]

Man, a pictures worth at least how ever many words I just wrote...

Even THAT one? Ugh! The reason I was thinking of belts vs. gears is that you might be introducing the very lash that the harmonic drives can eliminate. O.K, I lied: I saw it on the vendor's web site. But it sounded good! He he he. What I didn't draw was that, in my mind, the whole Y would raise and lower-probably via twin screws in the end uprights. That way you could raise up very high or dive down "to the deck" without torquing the Z any differently. Hmmm. Did you ever look at a word and think "That doesn't look right", but your not really sure? Need a spell checker CNCadmin! "Torquing" just looks odd. Oh well. Crap! Now "Oh" looks odd. Guess I'd better get some extra sleep tonight, huh? Bie!

Lance -

[jcolley]

The only problem I can see using belt drive is if the machine is designed to cut harder materials, then you need to be able to put more force on the axes to keep the cutter doing its job. I thought a bout using a belt drive on my 'A' axis, but couldn't see how I would be able to use it for much more than cutting foam if the cradle it was trying to move weighed 100lbs.

So, for me, a harmonic drive or a worm drive will provide a torque multiplication as well as the speed reduction I need. The only way to overcome the backlash problem I have figured out yet is anti-backlash gears which involve some sort of dual gear with a preload. At least on an affordable scale...

Jim

[miljnor]

belts might or might not work in this application but if your just worried about harder materials most of the hobby machines I have seen using servos use them and there are some out there cutting hard materials with very little problems(as far as hobby machines go) So I dont see why belts would be that bad on a rotrary axis.

Personaly I think the biggest challange will be the wiring and keeping it from tangling. The post earlier on the add on head for the fedal (*****en link by the way! thanks!) looked like the head was spining around more than once insted of unwinding in the reverse direction. This could pose some unique problems with the wiring runs.

edit: as far as horse power goes I have seen some 200hp hardley davidson customs with timing style belts so they can be built for the HP of your app. Although I must say I am just playing devils advocate as I personaly would go with gears for any machine I made in with this style head.

[turmite]

If it is a dual rotating head you are looking for, check out this conversion ( 5 Axis ) for a Bridgeport style machine.................but it doesn't seem that hard to figure out. Jim

Well it sure is hard for Evodyne an me to figure out! (nuts) He can't draw and I can't figure! Just kidding EVO.

Jim the 5 axis for the bridgeport is high $$$ if I remember correctly. That is why I have not even considered it. Second you would have to have a z axis built much like a bridgeport or it wouldn't work.

Evodyne could you email me directly the print you drew? When I clicked on it my eyes didn't seem to pick up some of your writting. Getting older by the minute!

I am actually wanting to build two machines. One a very small one that can be taken to classrooms and demonstrated and the other a very large one but light weight and relative fast. Neither will carve anything harder than mdf but most of the time it will be foam or wax. The small machine I can drive directly with 11, 14 or 17 size high torque steppers. The large machine I need a spindle that can handle at leas a 1/5" shank but I would prefer a 3/4" and that requires me to move a 80 lb. spindle in addition to the head, the z axis and whaterver I put on it at speeds approaching the speed of light if I can get it!

The larger machine I want for carving full size car or boat forms for making bucks from. There would be many other things it would be able to do if it was that size. Now one of the designs I shared with Skippy from here at the zone was a 5 axis but only 3 were programmable. The other two you manually turned and indexed with pins and locaters so it would be repeatable. This would require much more cam work but it could be done and if I can't come up with a satisfactory system to make it all programmable this is how I will build it.

Mike

[Dan B]

Hi Guys,

We have a Tri-tech head attached to a Fadal. It would not fit on a Bridgeport. The cost of the head is around $50,000 US, and the controller needs some major updating too, at big $'s. When all is said and done, you will have close to $90K invested into it.

The principle of operation is quite a bit simpler than discussed here. There are less gears in it than you would think. In fact when you rotate the head (A), you need to back-feed the spindle (C) just to maintain position. The "magic" of this head lies in the post-processor.

The downside of this type of design (head vs trunion) is it's lack of rigidity. We have busted this thing at least a dozen times (luckily we have a local site that repairs it). We use it primarily for drilling compound holes, but we also use it for simultaneous cutting as well (light duty).

I have to give Tri-tech credit though. This is a relatively cheap way to get into 5-axis on an industrial level. It opened our eyes as to the advantages of 5-axis machining and made the decision to buy a "real" 5-axis machine a no-brainer even at $300K.

Dan

[Evodyne]

The only problem I can see using belt drive is if the machine is designed to cut harder materials, then you need to be able to put more force on the axes to keep the cutter doing its job. I thought a bout using a belt drive on my 'A' axis, but couldn't see how I would be able to use it for much more than cutting foam if the cradle it was trying to move weighed 100lbs.

So, for me, a harmonic drive or a worm drive will provide a torque multiplication as well as the speed reduction I need. The only way to overcome the backlash problem I have figured out yet is anti-backlash gears which involve some sort of dual gear with a preload. At least on an affordable scale...

Jim

Well my cruddy picture might not have made it clear: the belt is turning the input side of a harmonic reducer. Given the incredible reduction, the feedback to the drive side will be minimal-I doubt the motor would even feel the extra load as you ploughed into some material. The belt between the motor and the harmonic drive would, of course, have zero lash if tensioned properly. And given the cost vs. gearing, I can't see that I'd go any other way.

Lance

[jcolley]

you know, I have been thinking about this for a couple of days now. I never even thought about the torque that a belt drive would have to be able to handle until you mentioned the Harley.

It seems to me that maybe I should research timing belts more. I have looked at the SDI site a little bit, but never really found anything about torque. Now, I know that any time you have a reduction ratio for speed, you will have a multiplication for torque. But, if you go in the other direction, with say a 50:1 ratio, then the countertorque developed by the cutter would be reduced by a factor of 50 to the servo.

Has anyone actually tried using a belt drive for a rotary axis? It seems that if it was a viable solution that it would be used more by manufacturers...

Jim