Why is one mechanism more prone to jamming? 4

[Jonahtang]

Let's say i'm designing a rectangular button with these two different cross sections. The button is constrained on the left and right by a vertical wall and supported on the bottom by a spring directly below the CG. The only difference between options 1 and 2 is length l.



Three questions:

1. If a force is applied off center, why is option 2 more prone to "jamming"?

2. Why can be done geometrically to reduce likelihood of jamming?

3. What if the button was a cylinder instead of a rectangle, would that help and if so, why?

[URL unfurl="true"]https://res.cloudinary.com/engineering-com/image/upload/v1652147076/tips/IMG_2919_brvepq.heic[/url]

 

[TugboatEng]

The wider button has a greater moment around the spring for a given F. The button is rocking in the bore. To fix this you need think like a bearing. Clearance and supported length are important. Less clearance prevents rocking, more supported length increases the resistance to rocking. See the width/height ratio in your drawings. If you can't make the button itself taller, consider adding a shaft below the bottom guided by a second bearing.

 

[Jonahtang]

Thank you TubboatEng, your comments are aligned with my thinking. Can you clarify how the bearings would be incorporated to help fix this?

And I'm also struggling with the free body diagram that explains why a more supported length increases the resistance to rocking/jamming. Can you help explain it through a free body diagram?

 

[3DDave]

The shorter the support the greater the allowable angular movement so the greater ability to have an off center load produce a moment, which is reacted by a couple in the support, increasing the contact force and the contact friction. You can demonstrate this with almost any drawer that doesn't use a ball bearing. Slip fit wood drawers are really good at jamming.

 

[Jonahtang]

Thanks 3Dave,

So does jamming occur because the contact friction force becomes too high?

If this was a frictionless system, would jamming still occur?

I'm getting confused on the reaction forces - It seems like the reaction forces R1 and R2 tend to push the button back horizontal CW while the external force F tries to push it CCW.

https://preview.redd.it/xh399sghbky...bp&s=f0bf4e0dcd9c6eb75a6e81f67a44f2562f0551eb

 

[3DDave]

Pretty good - but R1 and R2 are perpendicular to the sidewalls. There could be F1 and F2 that represent the friction component and are parallel to the sidewalls. While I guess they could be summed to a vector you have, keeping them separate eases analyzing whether the button is pushed or pulled while the perpendicular load is described next:

What also figures in this is the stiffness of the parts. For an increased amount of turn the corner-to-corner distance increases, but the width of the slot tries not to. So you can look at how tough it is to squeeze the button enough to fit. The more it turns the more resistant it is to turning.

This is the source of the initial sticking - that attempt to squeeze the drawer or the button into a hole that is not wide enough.

I haven't had a need to work this in detail - all my work depended on never getting anywhere close to sticking - so. low friction, close fit guide surfaces, lead-in chamfers to reduce the width at the start when installing.

 

[desertfox]

Hi jonhatang

What you have is the classic desk draw situation, have you ever pulled a draw out on its slides (not rollers) and it sticks? Then you have to pull slightly on the side Of the draw to pull it square before you can open it further.
Basically the draw wedges a because when it tips and contacts the side of the slides you need an increased force to overcome friction.
Have a look at this site it explains friction locking with wedges and basically your piston in option 2 is tipping and acting like a wedge.

https://www.hkdivedi.com/2020/04/wedge-friction-and-self-locking-in.html

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

 

[TheTick]

CORNERS!

Chamfer the corners that are in contact with the side wall.

Make sure the edge corners (when looking down from top of button) have plenty of relief. Rounds on hole corners should have larger radius than button edge corners.

(This in addition to advice about side walls)

 

[zeusfaber]

Rounds on hole corners should have larger radius than button edge corners.

Is that the right way round?

A.

 

[BrianE22]

Sometime ago a poster asked what L/D was required to prevent cocking. I had thought it was L/D > mu (coefficient of friction). It's a bit more than that though. Maybe I can find the thread.

 

[TheTick]

zeusfaber--my way keeps two sides of a corner from contacting at once.

Honesty may be the best policy, but insanity is a better defense.

http://www.EsoxRepublic.com

-SolidWorks API VB programming help

 

[BrianE22]

Take a look at this thread:

thread404-474483

 

[zeusfaber]

@TheTick

Thanks. I had to draw that to see what you were getting at.

A.

 

[Jonathan Tang]

CORNERS!

Chamfer the corners that are in contact with the side wall.

Make sure the edge corners (when looking down from top of button) have plenty of relief. Rounds on hole corners should have larger radius than button edge corners.

(This in addition to advice about side walls)

Can you explain why the chamfers would help. I saw on a following thread you said it "keeps two sides of a corner from contacting at once" - Why does this help.

From a top view of the button, I'm imagining you recommend filleting the edges and chamfering the corners as shown attached?

 

[TheTick]

 

[TheTick]

 

[Jonahtang]

Thank you for the images TheTick. I understand that adding the chamfers would help with positioning and inserting the button into the hole/slot. But once the button is inserted, in terms of jamming - won't the added chamfer effectively just affect the geometry as if the button was shorter (to the top edge of the bottom chamfers) and the side walls shorter (to the bottom edge of the top chamfer) since those will be the new edges that will make contact first?

I don't understand why added chamfers would help with jamming.

What i've learned from everyone (thank you everyone!) is that increasing the side wall length of the button aids in reducing angular movement and increases the lever arm such to reduce the reactive force on the wall (thus reducing friction). Wouldn't adding the chamfer reduce the side wall length thus worsen it?

 

[3DDave]

At the least the chamfer removes a sharp corner that can gouge in. It also helps make the item behave a little bit like a shape of constant width - if there was a radius from the centered at the top edge/outside of the button shell to the opposite corner of the button, then the button would not increase in width and touch both sides at the same time, so it could not jam. The easiest insertion depth to do this is at the start when there isn't much interaction to guide the item straight, so a small chamfer can approximate that.

That little red sliver off the corner allows the button to rotate into position without jamming.

 

[LittleInch]

I think also if F x the distance to the C0g/ spring fulcrum is more than R x the distance then the risk of jamming occurs.

If you look at the two options, the first is long and thin and the distance of R to the Cof G is longer than that of F.

On the thin wide cylinder the opposite applies.

And I think this makes sense in reality and maybe helps explain why pistons in engines tend to be slightly longer than they are wide and/ or almost as long as they are wide.

Most of us can see that a thin wide disc inside a cylinder is going to be more likely to jam than one whose length is more than the width.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[TheTick]

You won't have a sharp 90° corner digging in to the sides with the chamfer. It doesn't take much. Just enough to deliberately blunt the edge. 0.25mm will do.

Also more mold-friendly than a fillet. If your button is molded, the part line will likely be on that edge, which may have flash as a result.