[SCzEngrgGroup]

For years, I've been reading posts where many people here and on several other forums state categorically that TTS is only suitable for very light-duty milling, due to "the pull-out problem". I've always disagreed with that position as I've been using TTS for a number of years now, on several machines having 1-3HP spindles, and I have never once experienced pull-out. I have always argued that TTS works very well on machines at least up to 2-3 HP, provided sufficient drawbar tension is provided. Here is some visual back-up for my position:



There are three operations there, all being performed at 6000RPM, using a cheap, low-quality (Interstate brand) 2-flute HSS endmill:

1) Roughing a 1.006" diameter by 1" deep circular pocket - Cut in two steps, each of 1/2" depth. Helical entry at 77IPM, with w/10 degree ramp angle, followed by 110 IPM constant-engagement cutting at 0.050" stepover.

2) Finishing the same pocket - Cut at full 1" depth, at 65 IPM, in two passes of 0.010" each.

3) Roughing a large, odd-shaped pocket - cutting parameters are the same as operation #1, except depth is only 0.325", though I can just as easily cut the same pocket at 0.5" depth - that's just not what this part called for.

Unfortunately, on the fourth operation a programming error caused the spindle to stall, so I had to hit E-Stop. But, despite the fact that much of this was near the spindle power limit of the machine, and the spindle actually stalled at the end due to tool load, there was absolutely NO pull-out of the TTS holder.

Again, the "secret" is simply providing adequate drawbar tension. My power drawbars actually torque the drawbar to 30 ft-lbs, providing a drawbar tension in excess of 3500# - more than can be reasonably achieved using Belleville springs - which is enough to keep the TTS holder in-place under any load a 2-3 HP machine can muster. Many/most TTS users who have experienced pull-out seem to have power drawbars that provide no more than 2000-2500 pounds drawbar tension (many have much less), which is simply not enough to properly retain the toolholder with much more than a 1/4" tool.

Regards,
Ray L.

[waltpermenter]

how are you physically measuring the torque on the drawbar?

[SCzEngrgGroup]

how are you physically measuring the torque on the drawbar?

With an amazing device I found on the Internet: Let me google that for you

[waltpermenter]

how are you checking this with a torque wrench if the power drawbar is doing the tightening? how do you measure the over 3500# tension?

[Hirudin]

I don't want to contribute too much to the pissing contest that is probably brewing, but I would like to say that I'm pretty sure tension is tension. If your power draw bar, which I think looks very good, can provide 3500+ pounds of tension that is fantastic. That 3500# is superior to the 2500# that the Tormach setup provides. That being said: a belleville setup that provides, say, 4000 pounds of tension would be even better - and entirely possible. If someone wanted to they could make a belleville stack that would rip the draw bar in half. Getting it compressed in the first place, particularly while sitting on to of a mill, would be something else entirely.

I guess I think your power draw bar design has real advantages without continuing bellevilles vs. torque debate.

[Hirudin]

To the question above: it seems conceivable that you could attach the torque wrench to the spindle* and watch the needle. Although I'm not sure this exact method would work, I do know that measuring torque is not rocket science.

*Oops, I mean the PDB

[Jabs1542]

Aren't we really concerned with tensile pulling strength? How would one NDT measure that?

[SCzEngrgGroup]

I don't want to contribute too much to the pissing contest that is probably brewing, but I would like to say that I'm pretty sure tension is tension. If your power draw bar, which I think looks very good, can provide 3500+ pounds of tension that is fantastic. That 3500# is superior to the 2500# that the Tormach setup provides. That being said: a belleville setup that provides, say, 4000 pounds of tension would be even better - and entirely possible. If someone wanted to they could make a belleville stack that would rip the draw bar in half. Getting it compressed in the first place, particularly while sitting on to of a mill, would be something else entirely.

I guess I think your power draw bar design has real advantages without continuing bellevilles vs. torque debate.

Yes, you could build a drawbar to provide virtually any level of tension, but generating the force to RELEASE that drawbar becomes impractical after a certain point, as you'll have to sink an awful lot of money into a multi-stack air cylinder, or air over hydraulic system.

In any case, that is NOT the point of this thread, and I have no intention of engaging trolls. The point is, despite what many seem to believe, toolholder pull-out does NOT have to be an inherent part of life with TTS. TTS is perfectly capable of far greater performance than it is generally given credit for, and certainly capable of handling any load these machines can throw at it. I am trying to de-bunk the persistent myth that it is not.

It's just like the other myth that won't die, that stepper motors ALWAYS lose steps. That one just won't go away, no matter how much evidence there is that it is simply not true. I guess its easier for some people to blame the components than to admit that they did a poor job designing their system. TTS, like stepper motors, works great when used properly.

Regards,
Ray L.

[Hirudin]

Letvs all build these!
http://www.cnczone.com/forums/genera...rce_gauge.html

[Jabs1542]

Ok, I buy in on the TTS strength against pullout - I never had a problem when I was manually tightening the drawbar. I saw a screw based power drawbar video (it might have been yours) but unfortunately I have the Tormach belleville based PDB. Now I have tool pullout issues, usually aluminum, usually larger cuts (like the video you posted in this thread). Other than maintaining a well lubed collet shoulder is there anything that can be done to increase the "pull" of the drawbar?

[waltpermenter]

i would still like to know how the 3500# pull that is claimed is measured, theory is just theory until it's proven by physical testing. it would be nice if the procedure for it's testing is described or shown by the claimant so that the same test can be performed on the tormach power drawbar for comparison. i think a more prudent test would be the grip strength on the toolholder as that is what ultimately retains the toolholder.

[johnedward]

i would still like to know how the 3500# pull that is claimed is measured, theory is just theory until it's proven by physical testing. it would be nice if the procedure for it's testing is described or shown by the claimant so that the same test can be performed on the tormach power drawbar for comparison. i think a more prudent test would be the grip strength on the toolholder as that is what ultimately retains the toolholder.

I agree that it would be nice to hear or see the methodology at arriving at 3500 lbs but it's my opinion that it was guesswork not a practical procedure that could be repeated. I would also call the claim about reports "that TTS is only suitable for very light-duty milling" a bit spurious, have never seen that myself here and I've read a lot.
john

[vertcnc]

Ray,

On the stepper subject, are you still using the original steppers or have you changed to servos? I remember you saying you were going to upgrade your Torus Pro. What brand steppers and drives do they use.
Thanks

[SCzEngrgGroup]

Ray,

On the stepper subject, are you still using the original steppers or have you changed to servos? I remember you saying you were going to upgrade your Torus Pro. What brand steppers and drives do they use.
Thanks

I switched over to AC servos a few months ago. The only functional advantage is they are quieter. I still run the machine at the same 350 IPM I ran with the steppers, though the servos are capable of 500 IPM, but I'm just not comfortable with that speed. The stepper drives were, I think, Leadshines, and worked really well. No idea what brand the motors were, but they were NEMA43s - BIG!

Regards,
Ray L.

[FuriousGeorge]

With an amazing device I found on the Internet: Let me google that for you

It would be nice to see an appropriate answer to a valid question rather than sarcasm, so I'll ask it again.

Could you tell me in a detailed and meaningful way how you measured the torque on your PDB so that I could follow the same procedure to measure mine?

[SCzEngrgGroup]

It would be nice to see an appropriate answer to a valid question rather than sarcasm, so I'll ask it again.

Could you tell me in a detailed and meaningful way how you measured the torque on your PDB so that I could follow the same procedure to measure mine?

I thought use of a torque wrench to do that would be pretty obvious. There is a socket on the PDB, there is a socket on the torque wrench. You simply use a suitable piece of hex stock to connect the two. Could not be simpler. You can also do it even more simply, by using the PDB to tighten the drawbar, then put the torque wrench on it, and see how much torque it takes to get the drawbar to being to tighten further. Either method will get you the same answer, within a couple foot-pounds.

Regards,
Ray L.

[johnedward]

Yes but checking drawbar torque is pointless for the tormach drawbar, how did you arrive at the 3500 lb pull force figure? I know how we check ours but would like to hear how you did.
john

[SCzEngrgGroup]

Yes but checking drawbar torque is pointless for the tormach drawbar, how did you arrive at the 3500 lb pull force figure? I know how we check ours but would like to hear how you did.
john

I spent considerable time about 4-5 years ago testing, to see how much torque was required to positively retain a TTS tool under worst-case conditions, by applying different drawbar torques, then testing how much torque was required to get the tool holder to move. All done using simple torque wrenches and muscle power. I contacted Tormach, to see if they had any data, and even talked to one of their engineers who say they did not, though he was very interested in hearing what numbers I came up with. This was about a year or so before they came out with their PDB. I came up with about 20 ft-lbs drawbar torque/2500 pounds drawbar tension as the absolute mininum required, while they apparently came up with 2500 pounds as their target alue. To me, any real-world solution, to be truly reliable, would require considerably more to account for varying friction and other factors. I considered 30% to be a reasonable "safety factor", which meant setting my target at 3500-4000 pounds, which corresponds to roughly 30 foot-pounds drawbar torque.

It was at that point that I concluded a Belleville solution was really not viable, as the size, and cost, of a suitable pneumatic cylinder was unreasonable, and hydraulic intensifiers and other solutions (which is what is used on larger machines, like VMCs, which almost always have Belleville drawbars consisting of, typically, well over a hundred Belleville springs to achieve a drawbar tension of 3-4,000 pounds) are similarly too complex and expensive to be practical on a small machine. Hence, the motor-driven approach. I built a PDB using a NEMA34 stepper motor and off-the-shelf NEMA planetrary gearbox, which has been in use on my knee mill for years now, without a single instance of pull-out, except one time when I failed to properly lubricate the drawbar thrust washer. That PDB was capable of putting almost 75 ft-lbs on the drawbar - far more than required. I took what I learned from the stepper design to create my current design, which uses a very small DC motor, and much more sophisticated control, to create an extremely compact, reliable, highly efficient PDB with a target torque of 30 ft-lbs, to provide the 3500-4000 pounds drawbar tension needed for reliable operation at all times. Daily use over the last year, and by customers over the last several months, has proven this to be a very consistent, reliable system.

Checking the actual drawbar tension with my system is completely unnecessary, as tension can be directly related to drawbar torque. I frankly no longer care what the actual drawbar tension is, as long as I can demonstrate that it is more than adequate for even worst-case conditions. I am quite content to know what the approximate tension is, derived from the simple relationship between torque and tension in a threaded fastener, and to know through years of actual use that the tension I provide is more than adequate, even under worst-case conditions. Drawbar torque is the parameter that I directly measured in all of my testing, and it is the parameter that I directly, and actively, control with my PDB controller. Besides, torque is quite easy to measure while drawbar tension is not. Several years of daily use on several machines, without a single instance of pull-out, has proven the effectiveness and reliability of this method, as does the video I've posted here. If anyone wants to try that program with their own system, I'd be happy to post the G-code.

As far as how to measure the tension in a Belleville system, my best recommendation would be to either build a hydraulic system similar to what's shown in the thread someone linked to earlier in this thread, or buy a commercial load cell, and test a stack of springs on a suitable test fixture, like a hydraulic press. Testing on the actual machine will be difficult, and really offers nothing in the way of improved accuracy. You need to map out the force/displacement curve for the specific stack of springs actually used - testing a single spring would be close to meaningless. Once you have that, you simply measure the actual displacement when you install the stack on the machine, and you'll know the force/tension, provided the conditions are the same. One of the problems with Belleville springs is that there is a lot of frictional loss between the springs, due to the sliding contact as the stack is compressed. This is especially true when two or more are stacked facing the same direction. That friction will significantly alter the force/displacement curve as any lubrication is squeezed out or otherwise lost over time. The result will be that while a well-lubricated stack may work as you want, over time the the force the stack applies to the drawbar at any given displacement will decrease as more and more is lost to friction internal to the spring stack itself. This means the required release force will also increase over time, making it harder and harder to compress the stack enough to release the tool properly. Proper pre-load, and frequent lubrication is essential to maintaining proper operation of a Belleville system.

Another factor, which seems significant in many of the cases of pull-out noted here, is that a Belleville stack remains somewhat compliant - it is, after all, a spring. A fairiy common problem appears to be chatter-related pull-out. It's not hard to imagine chatter-induced vibration causing the tension exerted by the toolholder to exceed the tension of the Belleville stack if the machine is already running near the tension limit of the Belleville stack. When this happens, the Bellevilles will compress further with a more-or-less linear force-displacement curve, while the non-compliant drawbar in my design will provide a far stiffer response, and will yield far less under any given load. I've run some very aggressive cuts where the tool was chattering very badly (badly enough, in fact, it was painful to listen to), for a considerable period of time, and still never saw a hint of pull-out.

The beauty of a Belleville system is its elegant simplicity. But consistent, reliable, ultimate holding power is very difficult to achieve beyond a certain point, without resorting to much more complex, expensive components. If the application requires 1000 pounds tension, Bellevilles are the only logical choice. Beyond about 2000 pounds, it starts getting difficult, and dicey. Above 3000 pounds, it becomes really, really difficult, without resorting to very long Belleville stacks, and expensive air-over-hydraulic release mechanisms.

All of which, still, has nothing whatsoever to do with the main point of this thread....

Regards,
Ray L.

[johnedward]

Very wordy response but you still haven't answered how you tested for pull strength which tells me that it never was so your figures then and now are only guesstimates. Yes we use a system by Clamprite for ours which tests what the real physical draw is, had hoped you had done something similar, pity. I get what the point of this thread is and it's not to promote tts.
John

[SCzEngrgGroup]

Very wordy response but you still haven't answered how you tested for pull strength which tells me that it never was so your figures then and now are only guesstimates. Yes we use a system by Clamprite for ours which tests what the real physical draw is, had hoped you had done something similar, pity. I get what the point of this thread is and it's not to promote tts.
John

If you'd actually read my response, you'd see I DID answer your question - I did not measure the "pull strength" because I simply don't care what it actually is. All my testing and design was based on measuring and controlling torque, not tension, since that is the controlling variable in my design, and it is much easier to measure accurately. I have proven through both extensive testing and years of daily experience that adequate torque results in reliable drawbar tension, so why should I care whether the drawbar tension is 10 pounds or 10000 pounds? What matters is the end result, and the end result is I NEVER have pull-out, even on cuts that are severe enough that they are at the very limit of the spindle power on machines up to 2-3HP. I think I made it perfectly clear the tension is an estimate, calculated using basic, first-year engineering relationships. If you have a problem with the numbers I gave, then how about presenting your own, with the engineering basis for them? I can cite dozens of places where you can easily validate my numbers, without even having any engineering background. If that does not satisfy you, well, that's really not my problem, now, is it? I don't owe you, or anyone else, any more explanation than that. If you disagree with my methods, so be it. I'm sure I might well disagree with your methods, but I've never seen you post anything comparable that you've done yourself. I prefer to stand on my accomplishments, which I think are speak for themselves. I have spent considerable time doing testing, and taking measurements, over a period of years, and I have produced a successful product (several, actually), which has very happy customers. What have you done that qualifies you to suggest that my methods are not just as valid as whatever you might have done? How many power drawbars have you designed? How much testing have you done on TTS tool retention? I'm guessing the answer is "none". But I guess it's much easier to just lob hand-grenades from the sidelines., isn't it?

Regards,
Ray L.