[devers6]

Hi all,

I know very little about pneumatics; just enough to know that you need to be carefull what you feed high pressures into. I thought I would pose this question here, because someone out there surely knows more than I do and can save me from making a mistake.

I currently have a pneumatic drawbar that uses a Bimba pancake cylinder to press down on a drawbar with a stack of Belleville washers to release. My current cylinder has a 3 inch bore and gives me about 700 lbs with 100 PSI. I am looking to increase the force by moving to a 4" bore, and also by using tandem/stacked pistons (forgive me if this is not the accepted terminology). I want to increase the holding force of the Tormach toolholder system to something approaching that of the standard R8 drawbar torqued to about 20 ft-lbs. So I have a number of questions:

1) Exactly what is the force equivalent at the collet of 20 ft-lbs on the drawbar? I suspect the formula includes Pi, and the TPI of the drawbar (20); but I don't exactly know where to look. Can someone point me to the formula, or provide the answer of force in lbs given 20 ft-lbs of torque?

2) Once I have the answer to #1 above, I need then to get the Belleville washers that provide this force (I know that they can be combined in series and parallel to achieve desired travel and force). I also need a cylinder with greater force to open the collet. I would like to achieve that through an increased bore (4"), which would get me to about 1100# with 100 PSI. But I have also recently read about using multiple-stack cylinders to also increase the force. I haven't been able to find much about this (which leads me to think I am not using the right terminology in my searches). Exactly how do these work? Does the addition of a piston result in a linear increase in force (2 pistons = 2x, 3 pistons = 3x)? If not, what is the formula?

3) Looking at the construction of the Bimba cylinder, it seems like it could be something I could reproduce. I am thinking of making my own using 4", .065 wall 6061-t6 aluminum drawn tubing (Order Aluminum 6061 Tube in Small Quantities at OnlineMetals.com). Would this tubing be safe at 100PSI? What is the spec to be concerned with in determining a proper tubing size for a given pressure?

4) What is the best method for measuring forces in the greater than 2000lb range? I have bathroom scales that go up to about 350 lbs, and a hanging scale that goes up to 110 lbs. Is a lever arrangement the best way to measure something in this range?

Thanks in advance for any assistance in determining how best to proceed.

Dave

[Enraged]

just some quick answers, but:

1.) http://www.engineersedge.com/calcula...orque_calc.htm

keep in mind that the force exerted by cylinder will be pressure x area, so a 4" bore has about 12.6 square inches area, so at 100 psi you would get around 1260 lb-f, so likely more than enough without going to a more complicated/bigger cylinder.

for cylinder design, this would give you a start to see if that material would work: http://www.engineersedge.com/calcula...l_pres_pop.htm

[SCzEngrgGroup]

1) A fully-torqued R8 drawbar (20-25 ft-lb) will have a tension of 2000-2500 lb. Your air cylinder will need to generate considerably more than this, as you have to compress the Bellevilles further, to overcome friction and get the collet to release. Doing this with air alone is really not very practical. It's typically done with a "hydraulic multiplier", which is a larger diameter air cylinder driving a smaller diameter, long stroke hydraulic "master" cylinder driving a large diameter, short stroke hydraulic slave cylinder, which allows a 100PSI air supply to generate up to several thousand PSI in the hydraulic system. Managing that much force with Belleville washers becomes quite problematic, due to space, and friction, concerns.

2) "Stacked" cylinders give you multiples of the single cylinder force, neglecting friction losses, which are small. If a single cylinder gives you 500#a dual cylinder will give you 1000#.

Regards,
Ray L.

[mattbesquare]

3) Looking at the construction of the Bimba cylinder, it seems like it could be something I could reproduce. I am thinking of making my own using 4", .065 wall 6061-t6 aluminum drawn tubing (Order Aluminum 6061 Tube in Small Quantities at OnlineMetals.com). Would this tubing be safe at 100PSI? What is the spec to be concerned with in determining a proper tubing size for a given pressure?

Have you looked through ebay? surplus center? ? they are almost giving them away when they come up.

I have also been looking at designing a scissor type leverage system. You only need what, 150 thous of movement?

[rwskinner]

There are several factors that people over look when they state 20 ft.lbs of torque on a 7/16" draw bar will cause XXX clamping force.

One, almost all the folks use the dry thread clamping force calculations, and rarely are drawbars assembled dry. Antiseize, grease, oil on the threads will increase clamping force much higher than when dry while using the same torque.

The next thing that really affects the clamping force at a given torque is Drawbar material, and drawbar length. Unless these are taken into account you are pissing in the wind. You have to take into account how much the draw bar will stretch at a given torque.

Most people use a calculator design for a short cap screw with dry threads, and not a stud that is 18 or 24" long. Big mistake.

Now with that all out of the way, seems most folks with 700 to 800 lb springs do fine while others claim to need 2000-3000# spring pressure. This has a lot to do with tool length, feed rates, DOC, material and so on. Lots of variables. Much to do with drawbar, and collet prep.

1000# springs with a 1200# pressure from a 4" cylinder is generally ample for small machines like bench tops. If you want more, you can increase the springs then change the cylinder arrangement such as multi-stage cylinders, or use lever ratios to increase the force.

For Example: I use a 1.500 cylinder @ 100 psi that puts out right around 175#. I push a Class 2 type lever that increases the force by 6:1. This yeilds 1060# on the drawbar which over comes approx 650# of springs.
The above was the math for Hoss's drawbar, I just did all the math.

Angle of the lever can greatly reduce the amount of force applied to the pressure point or draw bar so be sure to do the math on that as well.

On mine, on a SX3, using Tormach Tooling, I have not had a cutter pull out yet, however rarely do I do DOC more than .100 per pass with coolant.

[eartaker]

The cylinder you have now would work for what you want if you add a class 2 lever system. I also am running a 1.5" bore and it works great.

Hoss' lever design
Projects 7

If you are set in making your own cylinder and would like to try it then go for it. I made my own 2" bore double acting before I decided to go with a single acting commercially made. I would go with some tubing with at-least a 1/8" wall for 100psi. You may be able to get away with less but better safe than sorry.

This would give you a bore of 4.75" Order Aluminum 6061 Tube in Small Quantities at OnlineMetals.com

[SCzEngrgGroup]

There are several factors that people over look when they state 20 ft.lbs of torque on a 7/16" draw bar will cause XXX clamping force.

One, almost all the folks use the dry thread clamping force calculations, and rarely are drawbars assembled dry. Antiseize, grease, oil on the threads will increase clamping force much higher than when dry while using the same torque.

The next thing that really affects the clamping force at a given torque is Drawbar material, and drawbar length. Unless these are taken into account you are pissing in the wind. You have to take into account how much the draw bar will stretch at a given torque.

Most people use a calculator design for a short cap screw with dry threads, and not a stud that is 18 or 24" long. Big mistake.

Now with that all out of the way, seems most folks with 700 to 800 lb springs do fine while others claim to need 2000-3000# spring pressure. This has a lot to do with tool length, feed rates, DOC, material and so on. Lots of variables. Much to do with drawbar, and collet prep.

1000# springs with a 1200# pressure from a 4" cylinder is generally ample for small machines like bench tops. If you want more, you can increase the springs then change the cylinder arrangement such as multi-stage cylinders, or use lever ratios to increase the force.

For Example: I use a 1.500 cylinder @ 100 psi that puts out right around 175#. I push a Class 2 type lever that increases the force by 6:1. This yeilds 1060# on the drawbar which over comes approx 650# of springs.
The above was the math for Hoss's drawbar, I just did all the math.

Angle of the lever can greatly reduce the amount of force applied to the pressure point or draw bar so be sure to do the math on that as well.

On mine, on a SX3, using Tormach Tooling, I have not had a cutter pull out yet, however rarely do I do DOC more than .100 per pass with coolant.

The numbers I gave are based on actual measurement when I built a power drawbar for my knee mill a few years back. In fact, when I told Tormach what I was doing, and asked them if they had any good drawbar tension numbers, they said no, and asked me to provide them with my test results! My numbers were confirmed when when they came out with their power drawbar last year, which also targets a 2500# drawbar tension. A smaller drawbar tension (800#) works fine for a small mill, but is NOT adequate for heavy cuts on a larger machine, and WILL allow the Tormach holders to pull out. At a 0.100" DOC, you'll have no trouble, but fast roughing at 0.500" DOC/50IPM with a carbide 3-flute is a whole different ballgame.

Regards,
Ray L.

[eartaker]

.5" DOC @ 50 IPM O_o good lord

[rwskinner]

Ray, in no way was I trying to indicate your figures were not accurate. I was just stating that many people will throw around numbers and many will use online calculators and come up with bogus numbers because they did not use the proper information for the calculations. Sorry if you felt like I singled you out.

I would be very happy with 2500# however I would never need it. If I planned on doing some serious hogging then 2 screws comes loose and the power draw bar folds back and I can then tighten it manually.

[SCzEngrgGroup]

Ray, in no way was I trying to indicate your figures were not accurate. I was just stating that many people will throw around numbers and many will use online calculators and come up with bogus numbers because they did not use the proper information for the calculations. Sorry if you felt like I singled you out.

I would be very happy with 2500# however I would never need it. If I planned on doing some serious hogging then 2 screws comes loose and the power draw bar folds back and I can then tighten it manually.

No problem. I wasn't offended, just pointing out that what I gave was as close to a "real" number as you'll find on this topic. The original question was how much force is required to duplicate the clamping force of a drawbard torqued to 20 ft-lb, and that's basically what I measured. IIRC, the normal calculation yields a much higher number.

Regards,
Ray L.

[pete from TN]

This is something I have yet to tackle, I did manage to buy some bellvilles for the drawbar and do some figuring but so far I have been more using the machine than actually working on it. It sure would be nice to have a power drawbar with the TTS system and I am sure I will get to it soon. I have now been using the TTS system on my RF45 for some time now. Hell I ran the machine for basically the whole day today on a small production run. I am currently using a pneumatic impact wrench to draw it down but that would be nice to not have to do. I still have aspirations of a toolchanger of some sort at some time in the future but really just a quicker toolchange would go a long way.

I had actually talked with Art quite a bit about a homemade cylinder setup and I think it is definitely doable altho I still have this electric actuator I bought for the project. Anxious to see others designs on an RF45.... PICS PLEASE!!! I am looking at making it a build onto the belt drive assembly I made for the machine. This machine does the kind of work that makes this a valuable accessory and would speed things up a bit for sure. Trying to get my finances back in order now that I am starting to feel better after my surgery but it is slow process. Good luck with this and let us see your progress.... peace

Pete

[devers6]

Thanks all for your replies.

As I mentioned, I currently have a pneumatic drawbar in use with 500# bellevilles and a 3 inch cylinder. It works fine, but I have been experiencing some instances of tool slippage. I have read the Tormach white paper, and there are probably things I could do, such as make sure the bore of the Tormach collet is clean and grease/oil-free, that would minimize this; but I came away thinking that my 500 # holding force was likely a little low.

I don't usually do heavy cuts (typically 1/2" endmill at .030 DOC), but I noticed anytime there was any vibration I could usually find the endmill slipping somewhat. So the natural assumption (especially reading Ray's many posts on the subject - thanks for those, Ray) was that I needed to up the spring force and upgrade my cylinder. I have been attempting to stay away from mechanical leverage just to simplify things; so the question was if 500# was too low, what was likely high enough?

I think I might opt for an interim approach and try upping the bellevilles to 1000# by doubling the number and stacking them together like <<>><<>>.
Then I could use a standard 4" bore at 100 PSI. There is bound to be an improvement; and maybe for my mill that would be enough.

As far as constructing my own cylinder; I'm not dead set on doing it; but I already have all the materials and the construction seems easy enough. It would be single-acting with a light spring return; and I could get exactly the travel I need (no more than .25") and have the lowest possible profile. I am also trying to remedy the problem of my current system, which simply pressed down on the spindle and therefore subjects the bearing to the force. I have a design that solves that problem; so I am ready to start. I just don't want to do anything unsafe.

I went back and looked at the specs for drawn 6061-T6 tubing in .065 wall thickness, and saw specs for Ultimate KSI = 42, Ultimate Shear strength KSI = 27, and Yield KSI = 37. If I am correct that KSI equates to 1000 PSI, that seems to be a safe margin vs 100 PSI even though I'm not sure which spec is most applicable. And my Bimba cylinder wall thickness seems quite thin (I didn't measure it when I had it apart) but I have no idea what material it is made from.

I guess I'll go check ebay for any good deals on a 4" bore cylinder, and if I find any maybe that's the safest approach. But I thought it was worth posing the question here before I went off and made myself a bomb.

Thanks again, everyone, for the helpful responses.

Dave

[ledbed6b]

Not sure if this was mentioned, some posts were a little long to follow, but what about bumping up your psi? 200psi compressors are not uncommon and most of the pneumatic solenoids and tubes I have are rated to 200psi or so.

Just a thought

[ninefinger]

I just don't want to do anything unsafe.

I went back and looked at the specs for drawn 6061-T6 tubing in .065 wall thickness, and saw specs for Ultimate KSI = 42, Ultimate Shear strength KSI = 27, and Yield KSI = 37. If I am correct that KSI equates to 1000 PSI, that seems to be a safe margin vs 100 PSI even though I'm not sure which spec is most applicable. And my Bimba cylinder wall thickness seems quite thin (I didn't measure it when I had it apart) but I have no idea what material it is made from.

I guess I'll go check ebay for any good deals on a 4" bore cylinder, and if I find any maybe that's the safest approach. But I thought it was worth posing the question here before I went off and made myself a bomb.
Dave

Do not equate the pressure in the cylinder to the stress the cylinder walls see at that pressure! Just as a cylinder can generate 1000lbs force from 100psi so too can that 100 psi generate 1000's of psi strain in the cylinder wall. For a quick example using your 4" cylinder (2" radius O.D.) and wall thickness of 65 thou (ie radius 1.935 ID - the "hoop" stress is 3151 psi in the wall of the cylinder. Stress in Thick-Walled Tubes or Cylinders - so for this particular application you're probably ok but don't take anything for granted - increasing the diameter has an effect on the stress in the cylinder wall for the same psi, a linear relationship (ie 2x the pressure is 2x the strain, 2x the diameter is 2x the strain)

Go safe and buy one if you're not comfortable with the design aspect.

Mike

[SCzEngrgGroup]

For the benefit of any adventuous souls out there, I'll pass along the conclusion I reached after attempting to make an R8 power drawbar that would duplicate the drawbar tension of hand-tightening to 25ft-lbs. While it is possible to do it with air alone, or an air-hydraulic system, I concluded neither solution is really ideal for several reasons. First, the complexity of a mechanism that can deal with that much force. Second, the fact that you WILL sometimes want to put a different R8 tool in, which the drawbar will make more difficult. I came up with another solution that solves both problems: Use a NEMA23 stepper motor driving a high-ratio planetary gearbox to tighten and loosen the stock drawbar. The motors are cheap ($50), the gearboxes reasonable ($150-200 for one that will bolt directly to a NEMA motor), and the right combination is capable of generating more than enough torque - up to about 75 ft-lbs. This would allow you to very easily loosen the drawbar just a turn or so for releasing Tormach tools, or completely unscrew it to release the collet and install an R8 tool, and would require no Bellevilles at all. Mach3 macros could very easily be written to control the stepper to accomplish this. You'd just want to limit the stepper current when tightening to control how tight it gets, but allow full current for loosening. A simple relay connected to the DIR line will do that. That's the approach I would now take, had I not decided to bite the bullet and swap out my spindle for an ISO30.

Regards,
Ray L.

[rwskinner]

After the stepper loosens the drawbar what drives the drawbar down so the collet releases? Really, sounds plausable. The only two items I don't like about it is your using valuable pins for the stepper which I'd rather use for something else, then the other concern is knocking the collet loose.

I guess you could use an air cylinder to keep the stepper and socket pressed down on the drawbar so 1, it engages and 2, the cylinder causes the collet to break loose?


Richard

[SCzEngrgGroup]

After the stepper loosens the drawbar what drives the drawbar down so the collet releases? Really, sounds plausable. The only two items I don't like about it is your using valuable pins for the stepper which I'd rather use for something else, then the other concern is knocking the collet loose.

I guess you could use an air cylinder to keep the stepper and socket pressed down on the drawbar so 1, it engages and 2, the cylinder causes the collet to break loose?


Richard

You have to have something to engage the stepper with the drawbar. Whatever does that, could also provide the down-force to release the taper. Done properly, simply unscrewing the drawbar can force the tool down, due to the drawbar threads - all it requires is preventing the drawbar nut from rising as it's unscrewed.

Regards,
Ray L.

[pete from TN]

I like the idea of the stepper as well but I would have to do a lot of rewiring and setup a fifth axis. I have also looked at making a dedicated air or air over hydraulic cylinder for the machine with a remote power supply hopefully controlled by mach at some point. The pneumatic setup is really not that complicated and Tormach has already proven that it works because there are several people using it. I am wondering if a very large homebuilt cylinder with some delrin guides for the ram could be built on my lathe out of a large chunk of steel and using a stainless ram it would weigh enough that it could just sit atop the flat millhead inside some guides to keep it in place and when actuated would be heavy enough to knock the collet loose. I have mulled it over with art a few times and just have not gotten around to it yet. Sure would be nice to have an electrically actuated pneumatic drawbar on my machine. Then I can make my toolrack like Hoss did on his machine mounted to the table. That would really be cool... May have to revisit the drawbar now that I am feeling better. Peace

Pete

[tooManyHobbies]

the gearboxes reasonable ($150-200 for one that will bolt directly to a NEMA motor)

Ray,

Can you tell us where you found reasonable gearboxes?

tom

[SCzEngrgGroup]

Ray,

Can you tell us where you found reasonable gearboxes?

tom

At the time, I easily found a number of sources, but don't recall who they were. Do some Googling, and you'll find them.

Regards,
Ray L.