Interesting arm and Help

So, This year we have attempted a very unique arm design. It is a double reverse 4 bar-6 bar. It has been very challenging to build, and we still haven’t been able to work out all the problems.
In theory the arm goes straight up and down at the front, but in reality it arcs backwards and forwards, which makes it difficult to lift with. Also, the back bar is tilted backwards, out of frame perimeter as seen in the first attachment. If anyone knows what could be causing these problems I would really appreciate the help.
Here is a picture with it down:

Here is a video of it lifting, around week4:

Here is a video of the problem, on a prototype:

Forgot to mention the most major problem, which is that the chain skips on one side or both on any section of the chain, which seems to be random.

At the center of it’s range of travel, the bars should all be horizontal. If they are not, then it will do interesting things like move forward or backward as it goes up and down.

The chain skipping is probably due to having a lot of load, and the chain having slack, and the parts not being as rigidly anchored as they need to be. It might be possible to add a chain tensioner of some sort, which could be as simple as adding an extra, larger sprocket to “float” between the two existing sprockets. Or it could require a more complicated fix. Larger pictures would help folks see how is built.

The gif in the second attachment appears to show a lot of flex in the arm assembly. There are a lot of pivots in the whole arm assembly so this flex will be difficult to avoid. That flex is probably what leads to the back bar being tilted backwards, out of frame perimeter. You might be able to add a piece of surgical tubing to pull the top of the back bar forward to keep it within the frame perimeter. It will have to be light since it will be stretched as your arm extends upward, making your motors work harder.

Is the arm “sticking” as it starts to lift? It seems to move reasonably smoothly in the two gifs. Does it still work the same way when you load it and put a cube on the end of the arm?

Assuming that the arm does not stick, does the forward and back motion of the end of the arm matter and prevent you from scoring? As long as the arm goes up far enough, can you live with this motion?

It might help to know which chain(s) are skipping.

also, we use gas springs to hold up arms, and the motor/chain thing only has to make them move up and down, it does not have to support all the weight. We’ve done this several times on our arm robots, including this year.

Regarding the chain skipping, check to see if you have flexing in the parts that support the chain sprockets.

In 2015 we built an elevator that was fine, however when we started to add more totes the top support bar would flex just enough to cause the chain to skip. Luckily we saw this on the practice robot and prepared a fix to carry into the regional. If we hadn’t seen this before the regional, it would have been really hard to find at the regional.

If you can’t resolve the problem with the chain skipping through tensioners, a very easy alternative would be to put #35 chain on the sprockets. Even if the C-C isn’t exactly right, I’d be shocked if it skipped. The main drawbacks are weight and the price of purchasing new sprockets, chain.

An additional option would be to simply increase the size of the sprockets. A larger sprocket would have more teeth on the interface and would therefore be less likely to skip.

Thanks for all the replies. First of all, the bars are very very close to all horizontal when they are at the center range of travel(The GIF does not show this). All Chains are highly tensioned, with tensioners on each. The arm is sticking as it lifts very slightly, as visible in the first gif. The forwards backward motion makes us unable to drive the robot for fear of chopping the RoboRIO in half. We have designed slides to circumvent this, but they might be causing the twist in the back bar. It flexes a lot because the material choice is extruded polycarbonate. I don’t think that the problem is a lack of motor power, as the entire arm weighs maybe 7 lbs, and it is driven by a CIM on a 500:1 reduction. (Actually, there is a small but nonzero chance that we swapped the gearboxes and actually have a 50:1 reduction, which would explain a lot). The chains that are skipping are the ones on the vertical bars, so I don’t think that the bar is flexing as those are aluminum. We are already AT 120 lbs, so switching to #35 chain is probably not an option for us. I’m still not sure what is causing the back to bend out of FP, but I think adding the surgical tubing like many vex teams use on similar lifts may help out with that. AS far as I know, it acts the same way under load.

Please post a photo showing the RoboRio in relation to the arms. The gifs don’t show enough detail though in the first gif, there is something that looks like the RoboRio in the middle of your chassis and it doesn’t look like any part of the arm gets close to it.

Did you remove your arm from your competition robot before Bag & Tag or is it in the bag?

Oh boy. You certainly weren’t lying about that arm being interesting. There’s been some good suggestions on here already but I’ll talk about the two main words that come to my head when I see those gifs, compliance and backlash.

Compliance is essentially the “springiness” of materials. Every material behaves like a spring and deforms under force, even relatively stiff ones like steel or aluminium. This deformation is normally hardly noticeable because most materials we use are quite stiff and don’t stretch very far before breaking. However when you have long slender pieces of metal like the links in your arm the deformation can end up being significant. I won’t go into the math here but if you take a 20 inch piece of 1x1x/1/16" wall aluminium tubing, hold it at one end and apply 15 lbs of force to the other you’ll see almost an 1/8" of deflection. For a single joint arm or four bar this deflection isn’t ideal but can be worked with. However looking at your gif I see four different member loaded like this on each side. That could easily add up to over half an inch of compliance. This is the source of the bounciness and swaying in your gifs and probably a good amount of the sag you’re having problems with.

The other word that comes to mind when you’re dealing with this many links is backlash. If you’ve ever taken your robot drive-train and rolled it back and forth you’ve probably felt backlash. The whole robot will roll very easily for a small part of a revolution and then you’ll feel the resistance from the motors. This happens because of all the gaps in the system, like in the bearings, gears, and shafts. All of these gaps need to be closed before the force is transferred from the motors to the ground. When you have a small number of joints or stages the backlash is normally not a problem. However your system has a lot of joints and it looks like the overall backlash is very high.

Now here’s the tough part. You might be able to apply some surgical tubing and chain tensioners to fix your current issues. You might be able to make it legal. But without a total rebuild of the system that bounce and sway will still be there. I don’t see any way you’re going to operate that mechanism quickly without it breaking from wobbling all over the place. If it’s moving this much stationary just imagine what will happen during driving or when another robot hits you.

So you’ve got a tough pill to swallow. You can spend the whole season fighting this mechanism and probably do very poorly. Or you can adapt your strategy. It looks like you’ve got a decent cube manipulator. Maybe you should consider chopping the top stages off the arm and giving up the ability to score on the scale. There’s a lot a good switch and exchange robot can accomplish this year and I haven’t seen many teams filling that role. You’ve still got time before your first regional to come up with a plan and fabricate it. As someone who’s been in that position before I know it’s a daunting proposition to radically alter your strategy after bag day. But, ultimately I think you would have a better season for it. Feel free to shoot me a DM if you have any questions or if there’s anything I can do to help. Good luck!

I noticed that there is no cross bracing. If you can afford the weight, adding some thin angle extrusion in the form of an ‘X’ pattern would help the side to side swaying.

Dave