As our team’s students take finals this week (& think about resuming robotics afterward), I’m diving into something I’ve been wanting to learn: 4-bar linkage intakes as explained by Nick Aarestad (@cadandcookies) in the excellent videos to which this post links.
I haven’t extruded anything yet - just want to make sure I’m on the right track here. One specific question: In the linkage sketch below that I overlaid on an early ball path sketch one of the students did, I added a crude sketch of a 4-inch-stroke pneumatic cylinder, but I don’t actually know how to best optimize placement, stroke length, etc. in this scenario. Any Do’s / Don’ts I need to be aware of when connecting to a linkage? Thanks.
Due to the lever effect, the closer the cylinder connects to the fulcrum the more force it will need to move your mechanism.
Also, one end will connect to your bar and that point will of course be the same point in both positions. But unlike your bars which are the same length in both positions, it will be its extended length in one position, and its unextended length in the other (and where those can coincide is where you will mount the other end of the cylinder.
Pretty much what John said - you have a lot of freedom when mounting pneumatic cylinders to mechanisms, but lever arm is probably the most important dimension to keep track of.
The RAMP video by Adam Heard does a great job of explaining / demonstrating this concept - highly recommend giving it a watch.
I use circles on the CAD centered on the cylinder picot point. One for fully extended and one for retracted. Then circles or curves that follow the motion of the linkage. Then I look for possible mounting point solutions visually. Then I swing/move the mounting point as constrained by the mechanism to check the other end of the stroke.
To further add to this: there is nothing saying the stroke of the cylinder needs to perfectly match the mounting point, as long as there is enough stroke. you can have your geometry for the cylinder mounting leave an inch of unused stroke.
OP:
Once you figure out the geometry make a cheap and dirty version with cardboard, and some of those brass cardboard “rivets” just to check everything
That’s a great tip; some leeway helps. We lost our ME mentor who knew this stuff, so I’m stepping up to help the design students.
Safe to assume there’s no reason the cylinder couldn’t stroke the other direction, with up being when the piston rod is out and down when it’s in? Seems like that might even be stronger if the intake took a hit while down.
For things that retract you can slip a stop tube on the cylinder rod to tune the position. You can also screw/unscrew the threaded end attachment to tune both ends some.
Yes, extend is always stronger. Sometimes the mechanism drives you one way or the other. Extended cylinders are MUCH easier to damage. However, as long as there isn’t a way to bend the rod, they will normally retract when overloaded, which is a nice safety feature for an intake.
To add (more) on to this; also be careful with cylinder angle relative to the link it’s connected to; perpendicular is usually most “efficient*”, and the smaller the angle gets, the lower your mechanical advantage.
*Obviously you won’t actually lose energy. However with small angles, you usually end up with a low mechanical advantage, so more energy is used for velocity than force. Also, depending on the specific linkage design, this may not always be as straightforward.
Yes! If the cylinder is exposed (or generally near the bumper) or has any other chance to be anything other than a 2 force member, having the retraction of the cylinder being the thing that deploys the intake is a good trick for protecting that point of failure (i.e. you don’t want a bent rod or sheared fittings)
Yup, ideally you want to size (and place) cylinders to minimize these cosine losses, but sometimes the mechanical design elsewhere means you don’t have much of a choice. You can always redirect the cylinder forces through another linkage if it is extreme (just make sure the cylinder remains a 2 force member)
As an aside: you can use these losses to your advantage if you have enough design freedom, you can make a mechanism “soft deploy” by having it operate on a non-ideal part of the sine curve when it is near hard-stops so it doesn’t slam into those stops with significant speed or potential energy acting in the cylinder.
Going back to this thread. Thanks all for the responses before. I think I’m getting closer. Here’s a picture of my current cylinder placement vs. one of the arms of the intake I’m working on:
The angle’s good for pushing the intake out, but the angle between cylinder and arm is around 12 degrees for bringing the intake back in… not very good, but should it still function? I’m using 4" cylinders because we have 3 matching ones in inventory unlike other sizes. It seems like it might start slow and jolt back when pulling the intake in… Unfortunately, there probably won’t be time to test it, so I’m looking for a little help from experienced eyes. Thanks.
This has the benefit of a much larger lever arm for your piston to push with (primarily when you’re putting it back up, which is when you need it) and keeping your piston more protected when the intake is out.
Second the reply above, though. If you need to operate with such a small lever arm, counterbalancing the intake will be a necessity.
Looking at reca.lc, I see the retract force of two pistons is 88 lb-ft, and 20% of that is 17.6. I’ll guess the intake will weigh 5 to 8 lbs, and most of that mass that needs to move is in the neighborhood of a foot away, so it seems like there should be enough strength.
What about stress on the arm? I’m thinking of making it thicker.
EDIT: Seems like reca.lc’s force calcs may not be right. I used this calculator instead and got a 2 cylinder force of around 47 lbs.
Yeah, I wanted to do that actually - I saw others doing similar. But because of other stuff on the robot, it didn’t work well. Maybe I’ll take another look at doing that.
I decided I’m going to elevate the cylinder mount to get better angle and accept the cost of more material & weight to accomplish it. I looked again at the option Nick mentioned, and can’t get it to work. I also looked at using a counterforce from a gas shock, and it could probably be made to work, but I’m running out of time, so I need to keep it simpler. Thank all.