Maximum unloaded speed of a stationary leadscrew with rotating nut

[Strostkovy]

I am building a press where I intend to use a 2"-4 acme leadscrew with a large bronze nut spun by a motor for rapid travel. It will have 1000 pounds on it during rapid travel, and that will be passed through a thrust bearing to the nut. A hydraulic cylinder will push on the end of the leadscrew for the actual pressing action.

I don't know what rapid travel to expect. I can obviously calculate motor size based on the torque required to raise the load and the desired speed, but I don't know where the speed limit is. The nut is spinning, so there is no critical shaft speed. Pressure-velocity product goes out the window when such low loads are applied. The nut will be well lubricated with a zerk directly to the middle of it.

Ideal speeds are 1000 IPM, but 500 IPM is acceptable. This spins the nut at 2000-4000 RPM, which feels high, which is why I ask. Inertia of the motor and nut aren't a huge deal. I may be able to jump to a larger thread pitch, but I need to make sure the nut is always self locking.

 

[DDoug]

Maybe I missed it, but how long is the threaded rod, and is it vertical or horizontal ?
Thrust is compression or tension loading ?

 

[Conrad Hoffman]

Zerk fitting suggests grease and I'd think max speed would be lubrication dependent. Flooded with oil probably does better at high speeds than grease.

 

[Garwood]

Yep. Depends on lube. I have a press that was converted from oil to grease without running the calcs and the higher speed (400 rpm @ 4.5") shafts destroyed the bearings immediately.

I don't understand using a nut for rapids like that. Why not do it hydraulically? I have a 5hp 40 ton hydraulic trim press that's "holy shit fast". It runs the discharged oil into the power side of the ram until it hits a few hundred psi.

 

[ViktorS]

Yeah I'd say thats way too fast for using grease, oil will be needed for heat dissipation at least. Probably air-oil mist.

I converted to metric in my mind so might be off with my calculations but that's about a 10 kW motor, and an acme leadscrew is somewhere around 20-30% effective. That's a lot of heat!

Look at hydraulics to get self locking, otherwise maby a roller screw. I know they are often used in fast moving presses but have no experience of them.

 

[DanielG]

You'll need to check P and PV for whatever your material is. That sounds too fast to me. I would probably just call Nook and run it by their engineers.

 

[Strostkovy]

The original thinking was I could do rapid, high accuracy travel with the screw, and then use a fixed stroke hydraulic cylinder to push to a known depth. This would be cheaper than a ballscrew or hydraulic servo, but my speeds and ram mass are too high to make this work. I think I may work to reduce the load, and possibly increase screw lead but we'll see.

Heat dissipation is the think I had not considered. That's pretty critical with the 25% or so efficiency of lead screws. I was planning on using a 1kW servo on the screw but I'm off by an order of magnitude.

 

[ViktorS]

What kind of accuracy are you looking for?

Depending on this, hydraulics might not be so far off. I have designed several systems with precision in the +/- 0.2 mm region with nothing more than a hydraulic cylinder, direction valve and a regular gear pump. Gear pumps actually work quite well in slower speeds too regardless of what most say, and when driven by a VFD controlled motor can be run from standstill to max speed with linear output.

 

[Strostkovy]

What kind of accuracy are you looking for?

Depending on this, hydraulics might not be so far off. I have designed several systems with precision in the +/- 0.2 mm region with nothing more than a hydraulic cylinder, direction valve and a regular gear pump. Gear pumps actually work quite well in slower speeds too regardless of what most say, and when driven by a VFD controlled motor can be run from standstill to max speed with linear output.

I'd like +-0.001". The flex of the press frame can be calculated and compensated for. This is basically a press brake.

I did find that same place I was going to buy the screws from (Roton) also sells 1-1/2" ball screws with low dynamic load but very high static load, so I can do the same thing. I think a servo with a brake will be able to handle the back driving force, but that and reducing the ram weight or counterbalancing it *should* bring this into the 1kw servo per side.

I'm also getting cheeky with synchronizing cylinders by using identical gear pumps powering each cylinder, run off the same motor shaft. Synchronization is actually less of a concern because the cylinders bottom to a known position at end of stroke. I just need the return stroke of the cylinders to be accurate to 0.05" or so.

 

[Garwood]

This all sounds super hokey.

If you want repeatability, speed, tons and safety you need a real press.

What are you trying to do? How much stroke, tons and platen size do you need?

 

[CarbideBob]

The nut is spinning, so there is no critical shaft speed.

Is this so?? ... my guitar strings do not spin but they do sing and they sing differently on end tension.
Just a thought.
Spin a long shaft and the whip problem which you find everywhere. Not so much information out there on high speed spinning nuts and their problems.
Yes I am just doing the bad,bad poking thing.

 

[Strostkovy]

Is this so?? ... my guitar strings do not spin but they do sing and they sing differently on end tension.
Just a thought.
Spin a long shaft and the whip problem which you find everywhere. Not so much information out there on high speed spinning nuts and their problems.
Yes I am just doing the bad,bad poking thing.

Yes, there absolutely is resonance. But it doesn't amplify like critical shaft speed does

 

[Strostkovy]

This all sounds super hokey.

If you want repeatability, speed, tons and safety you need a real press.

What are you trying to do? How much stroke, tons and platen size do you need?

40 tons on a 4' wide bed. 500-1000 IPM rapids, 10" stroke for nonpressing and 4" stroke for pressing.

 

[Garwood]

40 tons on a 4' wide bed. 500-1000 IPM rapids, 10" stroke for nonpressing and 4" stroke for pressing.

What do you need 14" of stroke for? Usually if your tool is active for 4" you would shop for an 8 or 9" stroke press.

It sounds like you don't really want to share enough detail to get suggestions, but it really seems like you're going about this all backwards with the screw and servo stuff.

If you really need 10" of stroke just for positioning and need it accurate within a thou what you need to build or buy is a toggle joint. You have one little cylinder that runs the toggle for the 10" motion and then your big boy cylinder does the hard work.

 

[Strostkovy]

What do you need 14" of stroke for? Usually if your tool is active for 4" you would shop for an 8 or 9" stroke press.

It sounds like you don't really want to share enough detail to get suggestions, but it really seems like you're going about this all backwards with the screw and servo stuff.

If you really need 10" of stroke just for positioning and need it accurate within a thou what you need to build or buy is a toggle joint. You have one little cylinder that runs the toggle for the 10" motion and then your big boy cylinder does the hard work.

There isn't that much to it actually. I'm just frustrated with available press brakes and their numerous downfalls. I want more open height for manipulating parts (especially large ones) with fewer handling errors due to poor visibility, and I want no pauses between rapid travel and bending, and a few other things, mainly backgauge and work height and safety system related that just aren't though through well on production machines within a reasonable budget.
The purpose of this particular mechanism was that it is a very economical way of achieving precision and force where needed and speed where that is needed.
I am in the process of manufacturing a pressure controlled rapid cycle time all hydraulic press for inserting hardware as an actual product, and while everything in that is great, trying to adapt that approach to achieve positional accuracy makes it insanely expensive.

 

[Garwood]

OK. Nevermind. You could have saved a lot of wasted thought and typing by just saying "press brake" from the start.

Sounds like a solution in search of a problem.

BTW, the press world speaks in strokes per minute, not inches per minute.

 

[Strostkovy]

OK. Nevermind. You could have saved a lot of wasted thought and typing by just saying "press brake" from the start.

Sounds like a solution in search of a problem.

BTW, the press world speaks in strokes per minute, not inches per minute.

I omitted the press brake information because I try to keep questions general and useful to someone searching the forums. And I think it was useful, as someone could easily find they were not considering the heat generated in the screw being a significant concern.

Strokes per minute is a garbage metric because they just program their open height to nothing and say the press is fast. My presses can do 400 strokes per minute but I would never advertise that because the return height is the minimum setting and the working stroke is zero, meaning no work was actually done.

 

[Garwood]

I omitted the press brake information because I try to keep questions general and useful to someone searching the forums. And I think it was useful, as someone could easily find they were not considering the heat generated in the screw being a significant concern.

Strokes per minute is a garbage metric because they just program their open height to nothing and say the press is fast. My presses can do 400 strokes per minute but I would never advertise that because the return height is the minimum setting and the working stroke is zero, meaning no work was actually done.

I think you need experience with presses outside of press brakes. This is very limiting to your problem solving approach. Press brakes are a tiny percentage of presses out there. They are versatile, but a compromise at everything else.

By the sound of what you're trying to do you're going further out into that versatility direction, but you need to watch what you are compromising.

I suggest a toggle instead of a screw because it is inherently ultra rigid, fast and accurate. The more shit you have moving the less rigidity and more slop you're going to chase.

 

[ViktorS]

40 tons? At 1000 IPM?
I wouldn't have my body anywhere near that thing!
There is a reason press machines of all kinds have ridiculous safety, there are quite a few accidents involving them.
But anyhow, you already know that.

We had an older 150 ton press brake some years back. All hydraulics in that one, two rams and servo valves.
It kept alignment and precision between the two rams with glass scales, one on each cylinder. Actually worked pretty good, especially considering it was from the 80's.

But I keep wondering, do you really know what spec you actually want?
14" travel at 1000 IPM, that's full travel end to end in about 0,8 seconds!
With the weight of a structure that can support 40 tons over 4 feet and keep straight.
Well, I can tell you that you will need a lot more than 1 kW servos for this!

But, let's have some fun with some hydraulic calculations, imagining that we go all hydraulic. Whatever you choose the numbers will be roughly the same. I'm using metric since that's my region..
Using the smallest cylinders that can press with more than 40 tons at a not-to-high pressure, that's two 120/70 cylinders at 200 bar (2900 psi). That's giving us about 45 ton combined.
We want to keep the cylinders small so we can move with speed too.
Say that the structure weighs about 1 ton, we will need to lift that in the returning stroke. This limits the lowest pressure we can use.
We want to keep the pressure low for positioning moves to keep the power requirements as low as possible.That is in the region of 8 bar/116 psi for two 120/70 to lift slightly more than a ton.
We're looking for a velocity of 25.400 mm/min or 423 mm/sec for 1000 IPM. That velocity with those cylinders give us a total flow of 575 liters/min @ 8 bar.
Nice! We're also approaching max speed for most standard piston seals, which is around 500 mm/sec.

Now, a pump that gives 575 lit/min at 8 bar will consume about 7,7 kW, at 100% efficiency. A more realistic number is probably about 85-90%, so 9 kW.

As I don't know the velocity for the pressing move I can't calculate that, but I imagine that it is somewhat lower, but let's say 7 seconds for 4"
We'll definitely need to use a slave pump to get higher pressure at lower speed, so that sounds reasonable. As the pump is VFD driven we can over speed the motor during light load for better performance.
Those numbers give us 20 lit/min at 200 bar. That's almost 7 kW at 100% efficiency.


While we're at it, let's calculate some basic screw data too.
I found that Hiwin has a 48x10 screw that actually can handle (theoretically at least) the load. Load is 196.200 N/screw. That screw can handle 242.800 statically, but the next one down was way under the target.
Now, with 90% efficiency in the ball screw we will need about 340 Nm of holding torque. When I'm looking in BEVI's catalog for a 4 kW 3 ph motor with brake, I get the number 10-50 Nm, adjustable. And that's a 4 kW motor, not a 1 kW.
A 22 kW motor has 60-300 Nm, still 40 under what we need. I'm having a hard time believing that a servo motor has more brake torque than any "normal" brake, especially considering that servos usually have quite small diameters.
So there is that..

Let's look at positioning, as that will give us some better numbers, right?
Well, if we go with the one ton load of the frame again, we're looking at 4905 N per screw. That's 8 almost 9 Nm. That with a speed of 423 mm/sec gives 2538 rpm on the screw, or 2,3 kW.
Combined that give us 4,6 kW, so a bit better than the hydraulics, although some of this is due to rounding errors on my side.

One thing to note though, all these calculations are static, IE no consideration for acceleration or deceleration. Add that and power requirements will jump.

 

[Strostkovy]

I think you need experience with presses outside of press brakes. This is very limiting to your problem solving approach. Press brakes are a tiny percentage of presses out there. They are versatile, but a compromise at everything else.

By the sound of what you're trying to do you're going further out into that versatility direction, but you need to watch what you are compromising.

I suggest a toggle instead of a screw because it is inherently ultra rigid, fast and accurate. The more shit you have moving the less rigidity and more slop you're going to chase.

I do agree that press brakes are a small fraction of presses that are out there, but they are my main source of frustration and an area that I feel could use far more thought out development. I intend the machine that this mechanism was proposed for to solely be for folding metal.

Forgot to mention that the press I was talking about in the second half of my reply is the actual small hardware insertion press I am developing for sale. Two totally different applications.

A toggle mechanism is interesting, however has some complexities regarding controlling depth of stroke, though that is not necessarily insurmountable.

I will say that while I do intend to ultimately develop a press brake for sale, I am years away from even designing a production version, and am focusing on much smaller specialized machines to start with. This screw idea was mostly just off the top of my head, with little calculation involved