Nut is not self locking on the threads 1

[aaansari97]

Hello,

I recently designed a mechanism with a 4-start left-handed trapezoidal thread where the nut translates linearly.

The nuts is brass H62 and the leadscrew is SS 304.
Pitch is 1.5mm; lead is 6mm
Major diameter is 8mm.
Coefficient of friction (brass on steel, dry): 0.51

According to my calculations as per the equation f>l/(pi*d_m) where d_m is mean diameter, the nut should be self-locking. However, in actual it is not. :/

Any ideas why? Is the friction coefficient inaccurate, or is the equation generally unreliable?

Any help will be appreciated because we're considering reverting to a single start thread if we're unable to work out why this design is failing.

Thanks!

 

[3DDave]

More starts - less able to self-lock. Basically look at the slope - 6mm rise over pi*8mm run = .238 coefficient of friction (roughly, a number you could have supplied) to stop it sliding. You think it is .51 brass on steel, if it is completely clean, but I expect that there is oil involved. It will take nearly no oil to reduce this to some far lower number.

Try cleaning the parts with brake or carburetor cleaner first and try again. Also - you may have an assembler helping you out. I would expect that, since it's a screw, this is more than a one-time use and will wear out the brass nut rapidly without lubrication. One drop of oil would be plenty to reduce it to <.2.

 

[aaansari97]

3DDave, thanks. I was aware of the reduced ability to self-lock because of the greater lead angle during the design phase but I thought I was playing safe given the calculation.

I also thought that the oil was the reason. But I confirmed with our assembly vendor that they put no oil there.

Maybe, I used an incorrect coefficient of friction for this particular brass?

Is there any reliable source for the coefficients of friction?

Any solution for making it self-locking and still have a good life? I was thinking of having a snap lock mechanism where it's supposed to lock.

 

[3DDave]

Don't confirm you assembly vendor did or didn't do anything. Check it yourself, with a can of degreaser.

Put up a picture of what it is and why it is self-locking.

This guy did some testing - polishing lowered the friction considerably and he felt oil made no difference.

https://www.varmintal.com/afric.htm

 

[Windback]

If there is resource, preparing 30 specimens for testing against the equation and refining the assumed parameters based on the normalized test results might work. Tackling with friction coefficients might be frustrating.

 

[aaansari97]

So I think it's best to go with a trial and error approach.

Thanks, everyone!

 

[mfgenggear]

windback
multi lead or multi start lead screws are designed to slip and not lock. why did you use a multi start and what are the design parameters , angle of the thread.

 

[desertfox]

Hi aaansari97

Can you elaborate more on your mechanism and how the screw isn’t self locking?
If the mechanism see’s a vibration or vibration’s during operation then that can defeat the self locking of the thread as friction can be reduced to zero, self locking threads can only lock and prevent rotation against axial forces.

Can you clarify whether the 8mm diameter in your post is actually the major diameter of a 8mm metric thread or did you mean the pitch diameter of the thread in order that the lead angle can be calculated. According to my calculation the friction coefficient angle is less than your lead angle and therefore is not self locking, unless I have made a mistake

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[hydtools]

Slide a block of the brass down an inclined plane of the stainless steel to get a idea of the coefficient of friction. The arctan of the angle of inclination when the brass just slides will be the coefficient of friction. See any machine design textbook for the example of a body sliding on an inclined plane.

Ted

 

[Scuka]

What's your factor of safety in the equation? If your thread just barely enters the non self-locking zone, that's where your problem might be. Coefficient of friction is not exact science. The number could be off, and it could easily be altered by a variety of factors.

 

[desertfox]

Hi aaansari97

I think I made a mistake in my calculation and in theory at least your thread should be self locking. It really now depends on how the nut and screw are loaded ie back to my vibration comment earlier. Could also be reliability on the coefficient of friction as others have stated

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[btrueblood]

What brass alloy did you use, and does it contain a small amount of lead?

from this -

https://www.engineeringtoolbox.com/friction-coefficients-d_778.html

look down below brass to the listing for "copper-lead alloy", which is more likely what you have in a machinable brass alloy...I see 0.22, and from experience, that's more in line with a typical C36000 free-machining brass on steel couple.

 

[aaansari97]

Thanks, everyone!

I rechecked the calculations, and most likely the issue is with the coefficient of friction that I used.

The brass we are using is H62.

I'll most probably proceed with this arrangement and introduce a locking system (some sort of snap mechanism, maybe).