4 bar linkage problem. 3

[RoarkS]

Youtube video

So I'm working on a 4 bar linkage for a fella.

The video is rev b.
You can see he is applying force to the input driver link. On the way down the coupler link goes down like its supposed to. But on the way back up, you can see it hits a dead spot and the mechanism doesn't return home.

As much as I would like to say its because of sloppy holes and bolts... the really nice billet version A does the same thing... so the customer has no confidence moving forward.

The first revision had the follower link and the coupler link get within 4 degrees of being in line. I thought this was an obvious small angle dead point. So rev b in the video here, I took from 4 to 15 degrees.

What am I missing guys? I'm supposed to be the pinch hitter and this is still messing up.

 

[3DDave]

If you draw a line between the two base pivots it looks to me that the middle link is going over that line. While the middle link hasn't gotten to an over center location it is very close so that even small amounts of clearance or deflection will yield variations. The usual way to deal with this is to add a spring to force the middle link and, in this case the extended finger, to close. This could be by using an extension spring or using a torsion spring at one pivot.

On the design side I would do a sensitivity analysis - put the linkage as is with the drive link in the closed position and then change the pivot center distances on each link, one at a time, by 0.001 inch or 0.01 mm and then recompute the positions and see where the output ends up. I expect it will move a large enough amount to be noticed. You can then decide how much variation each pivot center distance might vary and scale the output to match and then add all the contributions.

As I mentioned before - a lot of similar mechanisms end up with a spring to drive the linkage in the desired direction.

 

[DAVIDSTECKER]

With out digging deeper into the definition of a 4-bar link, I don't think that I would call this one.
I typically think of a 4 link to keep the planes parallel.
My first impression is that it is a hinge.
Is it possible to apply the input force to the other link?
My thinking is that upward force would be shared by both end points.

 

[3DDave]

Rarely do 4-bar linkages keep anything parallel. They can, it's just uninteresting to do so.

Typically they are 4 elements attached to each other to form a cycle such that A is attached to B which is attached to C which is attached to D which is attached to A, completing the cycle, and that these attachments remain through the operation of the mechanism.

Mostly the connections are pivots, but there is allowance for sliding contact as well.

What makes this more interesting is the extension of what would be the middle link, the short one in this case, with the long pointer or whatever that is to do.

 

[RoarkS]

Tried driving it with the front link. No change.
Gonna try moving the base pivots also to get away from the centerline crossing that line for over center. It does look like it has some over center action.

BTW these are courtesy steps on a lifted Blazer. Thus the kinda extreme vertical drop. At first they were using stock GM components but slowly moving towards full custom. Very high end Barrett Jackson build.

 

[3DDave]

There is a program called linkage which can help a little bit with figuring out various combinations of pivots and such.

It's free and not hard to learn. The guy who made it uses it to design linkages for his marble machines to lift marbles in interesting ways.

https://blog.rectorsquid.com/linkage-mechanism-designer-and-simulator/

I like this one; all 4-bar, all driven from one motor.

https://www.youtube.com/watch?v=rShllGfOrqM&t=7s

 

[MintJulep]

It seems like an over-center lock in the up position would be desirable. Prevent the step from deploying accidently.

That would require driving from the front link.

Consider driving the rear link to the closed position against a spring-stopper. That should force all of the unavoidable slop to do what you want - bring the thing to the proper closed position.

Linkage is a nice tool, but I've found it difficult to define links with precise dimensions. Looks like it's been updated since I last used it, so maybe it continues to improve.

 

[hydtools]

Look at AMP Research power steps.

Ted