I did a search for any discussions related to shaft whip and came up with little.

What max RPM can one reasonably spin a 5’ 1/2" ACME threaded rod without developing any shaft whip or resonance?

Does anyone out there use screw drive on a larger (4’x8’) table? Shaft whip problems? What RPM?

Tim T

For half inch it's around 500 rpm at 30" long (from experience). I am guessing a bunch less for 5".

I don't know the answer to your question, but as an alternative(if there is one), set up the drive in the middle of the screw, turning the nut/nuts instead of the rod.

I don't know about the 1/2" but Mariss from Gecko did a calc for me on my 1.125" X 48" at the fastest rapid I wanted, 300ipm, and he told me I would need 1500 rpm and the screw would whip so bad it would be useless. Maybe between the two sizes given here at by balsaman you can get close.

turmite

This will give you the information you need.

http://www.nookind.com/pdf/powerac1.pdf

Read it all, there is a lot of information in there. The speed chart is page 11.

Gerry

I currently run my 3/4" acme at 720 rpm for 120" per minute, down from 150" per minute because my motors can handle the speed but the rods whip to much for me. It's smoother at 720 than at 900 rpm. My longest screw is 48" long.

One way to get around this problem is to get 1 pitch screws and gear it way down. Then you only need 100 rpm to go 100" per minute. Rack and pinion or belt drives are also an option, but it's hard to get the backlash out of the gearbox you need for those systems.

Eric

Thanks for the info everyone. I was afraid you were going to say exactly what you have.

I imagine the efficiency of 1 pitch screws is a lot lower than a higher pitch. Maybe I'll investigate ball screws a bit more.

I have been designing for belt drive, but don't want to go down a path without considering all possible routes.

I have seen the photos of the Axxus belt 4’x8’ table. Does anyone know of any other examples?

Tim T

I was going to tension my 36" threaded rod shaft. Will that not help a bunch?
co

There's another possibility that you're not considering. The lead screw only has a problem with whip, if it's turning!

Keep it still!

I saw mention of a design on CCED, which basically turned the nut, rather than the screw. Of course, keeping backlash down was a concern, since there's a rotating mechanical linkage between the nut and the stepper, but the general tone on the list was that it was a workable solution.

You should be able to get away with a MUCH smaller screw, using a system like this.

-- Chuck Knight

I would think that a lot of the whip would depend on how well the shaft is suppored..or am I mitaken?

Originally posted by yukonho
I was going to tension my 36" threaded rod shaft. Will that not help a bunch?
co

I would think that yes, tensioning the screw would help a lot. This will require a sturdy frame and real thrust bearings, though, because the preload will be quite high to actually create significant tension.

Originally posted by avsfan733
I would think that a lot of the whip would depend on how well the shaft is suppored..or am I mitaken?

You can only support it at the ends. Therin lies the problem.

Eric

For maximum shaft speed, support the shaft in two places on each end rather than one. The two bearings don't need to be far apart, but the critical speed increases dramatically. This is referred to as "fixed" support, one bearing is referred to as "simple" support. Check out www.pic-design.com , they have a technical information section with calculators for this and many others.

Steve

Just installed my X axis screw. I am using 1/4-20 rod for now. The axis is 36" long. I put a bearing at each end and put it under a bit of tension. It is significantly stiffer than when not under tension, and still turns very freely. I will have to see what happens when Iactually spin it up.
Hu. You are probably right about needing thrust bearings, but from what I observe right now, it doesn't need a lot of tension to make it a lot stiffer. Also, my machine is MDF, but the ends are supported top and bottom, and is very rigid, definately rigid enough to tighten this screw a lot more than I have done. I will have to wait until probably after Christmas for my steppers to arrive so I will give an update when I can.
co

I noticed that shaft whip doesn't occur instantly at high speeds but builds up slowly over time. That means that actually not much energy is fed in a resonant system. The energy just accumulates until the whipping is so bad that the steppers stop. Maybe just a little bit of damping could drain sufficient energy to stop the building up of resonance? I just got my X-axis running and can run my steppers a lot faster than the speed where whipping occurs, so I found a good reason to give this damping theory a try :D.

One way I can think of is to slide some PVC tubes over the leadscrew and support these in the middle in a cork mount. The tube could be a bit loose- fitting and will align itself to the leadscrew. Any offset motion of the screw will feed energy to the cork mount. See drawing.

Will this work? Any other ideas?

John

Limbo, I've seen something like your idea on a Thermwood router. (I think its patented).

I going from memory, but it seems that they would mount a pivoting delrin support near the center of the ball screw. An 8' screw would have the support at the 4' mark.

The support did not move with the gantry, but stayed in place. Where it met the ballscrew it looked like a U. When the gantry went by, it would push on the support which would pivot out of the way. After the gantry went by, the support would simply pop back up in place.

Very simple and greatly reduced the screw whip.

I have also seen Chuck's idea about rotating the nut and keeping the screw stationary.
I think it was on a Routech router.

Last but not least, would be to increase the diameter of the shaft. I saw someone use a large diameter aluminum screw. The large diameter solved the whip. They had the screw hard anodized to create a good wear surface.

I use a nut at the far end of the support and a guide at the other end. Like what you have drawn but the nut on one end, and a guide on the other. Nothing in the middle.

Eric

Limbo that was kind of my idea...except i had the thought of using two nuts...one on each end where you supports are to spread it out a little...also allow me to use less tension and still have less lash

Avsfan,

I didn't have much luck with two nuts in my first machine. When the gantry is far to the right, the thread will sag in the middle. This way it will lean hard on the left nut and push upwards in the right nut. The added friction reduced efficiency and max. speed a lot. I'm not trying to scare you though, of course it depends a lot on how the rest of your machine is set up, so it might work just fine in your machine.

What I'm going to try with these dampers is not support the thread (the dampers will deform and move away under the weight of the thread), but just drain energy out of the resonant system. A rigid support (like a second nut) can absorb no energy. Its a mass-spring system with low losses. I think you need deforming to add these losses.

John

creative, odd, possibly stupid idea...but i'm just throwing things out here. and then supporting the shaft from the bottom in another fashion basically using it like the open bushing are used open supported guide rails...not sure how you would support the bottom but its a thought

The following messages contain the details of the technique Steveald was discussing several messages ago.

Read the sections on "End Fixity" and "Critical Speed". These pages are out of the Nook catalog.

If I read this correctly, a 30" 1/2 x 10 lead screw with Simple Fixity should be capable of slightly less than 100 ipi. The same with Fixed Fixity should be capable of approx 200 ipi, but please review this info and confirm these numbers.

Next Page

Ok, this really is the second page :D

Last Page. If anyone can't read any of these, please PM me with your email address and I'll send the full image. they are approx 800kb.

Better detail of the right hand side of the chart...

It looks like about 30 ipm to me for the simple (A), about 120 ipm for fixed (D).
Gerry

Perhaps I wasn't clear or perhaps the document isn't clear, but I was thinking of “Simple” as representing one bearing at each end, and “fixed” as representing a double bearing at each end. This would then correspond to line “B” and line “C’ on the chart, respectively.

Looks like you're about right, then.

Gerry

[QUOTE=ger21]It looks like about 30 ipm to me for the simple (A), about 120 ipm for fixed (D).
Gerry[/QUOTE


Dear Gerry,

This may be a daft idea, but has anybody thought of spacing the two halves of their anti-backlash nut several inches apart? If you did, the AB assembly itself would become fixed, I think, and would greatly lessen any tendency of the leadscrew to whip because the two halves of the nut would keep the screw in line.

Regards

Martin