Rotating, positive lock design

[mbuis]

I am new to working with polymers/plastics. I am currently trying to design some sort of mechanical, positive lock joint. I have two parts, one of which will be able to rotate from 0 to 125 degrees in the other part. I want to design some sort of interaction between the parts such that at the extremes of rotation (0 deg & 125 deg) the rotating part is "locked" into place with a solid snap. However, the part needs to be able to "unsnap" and rotate to the other position. Currently, I have two thoughts/designs for this problem. See attached images.

In the first one there is a boss on the face of the non-rotating part, and multiple grooves on the face of the rotating part. This one would have a "ratcheting" feel. The second one has two grooves around the outer face and around the axis of rotation of the rotating part, with a long boss on the non-rotating part. Hopefully, the pictures will help clarify. My questions are basically contained to how do I know the part will actually rotate and not just become permanently snapped into place once assembled? There are multiple articles about snap-joint designs, but I have really found nothing which pertains to this situation. I can, of course, have some rapid prototypes printed and "try out" different designs; however, I am hoping there is a more logical approach in order to eliminate unnecessary spending.

Thank you for any help with this situation.

 

[MikeHalloran]

You are trying to make a detent, not a snap.

To do so, you will have to analyze the details of the detent throughout its stroke, looking specifically at forces and pressure angles.

The detent's force comes from a leaf spring, which is also part of the necessary analysis. If I am reading the image correctly, the spring is probably too short, and will be overstressed in bending as the detent is forced out of its notch. The amount of force available/necessary to do that is also part of your analysis.

You might consider modeling sections of the problem with paper dolls cut from posterboard. For me, that's sometimes faster than modeling mechanisms with CAD.



Mike Halloran
Pembroke Pines, FL, USA

 

[mbuis]

Thank you Mike for your response. And I agree, a detent is probably a more apt description. However, I shied away from using the word "detent" because a detent is often something you have to release, as in push a button. I do not wish to add any complexity (aka cost) to the design unless necessary. I was using the word "snap" because the user would apply a force to the rotating part and it would "unsnap" and rotate to the next position and snap into place there. A big part of this design is the "feel" of it. Which, admittedly, is an intangible quality which is difficult to define. But, again, I agree the use of the word detent is more appropriate.

The idea of thinking of the boss on the non-rotating part as a spring makes a lot of sense. However, I am not sure I follow you on "the spring is probably too short." The screen shots I provided were, more or less, side shots. And "too short" to me references the vertical direction. I was concerned I actually may have too much engagement (too tall) and once the parts were engaged a person would not be able to rotate them.

Perhaps, you would offer more clarification?

While I do agree, making a physical representation of a problem is sometimes simpler and offers more information, I do not follow on how using paperboard in this situation would help. I am intrigued by the idea, but I unable to see how to construct a model to help me decide on the correct geometry. Unless, I am to glue a "stack" of boards together to create the detent?

 

[MikeHalloran]

No, you don't need to laminate a stack of posterboard profiles. A single thickness of manila folder stock would do nicely in seeing the sliding friction and pressure angles involved, especially if you make the 2D model much larger than life.

WRT to the length of the spring, I'd expect it to be ~10x the depth of the detent cog, and maybe tapered in depth a little.

The 'feel' of which you speak is indeed tangible, as a torque, or more specifically as a torque vs. angle function. It would probably be helpful to find a physical detent, anywhere, that has the right 'feel', and to measure its characteristics with a torque watch and some sort of angle indicator. Once you have that set of numbers, you can work through your mechanism a stage at a time to arrive at a design. It's the sort of thing I'd do in Excel. Assume no friction at first, then add it as you gain confidence in the math model.





Mike Halloran
Pembroke Pines, FL, USA

 

[mbuis]

Very cool, Mike. Thank you for the information.