"Smart" Hook

This past weekend, I attended W0W (Week 0 Warmup) put by Team 2338 in Aurora, Il - great event.

They had the various field elements setup including the hangar. There wasn’t a lot of climbing but Team 111, Wildstang, did a couple of manual climbs and traversals with their latest robot. They may have only done a single traversal as they felt things were still a little risky (don’t blame them for a second - seeing it done was once was great, but scary).

Watching that, something that occured to me was the fact that a big part of the actions of a traverse (which our team is aiming for) were essentially blind. It would be nice to “know” that your hooks have engaged the appropriate bars and be able to confirm or abort a traverse attempt either automatically or even manually through a set of dashboard indications. Much like Frank Constanza inventing Festivus, I thought there might be a better way.

So I started messing around with the “Smart Hook”. It’s a 3D printed hook insert designed to be sandwiched by a pair of aluminum plates to carry a microswitch at the top center of the hook opening thereby providing a sensor to indicate hook to bar contact. Now you could indirectly count on an increase in lift motor current to tell that a mechanism was starting to take robot weight, but depending on the lifting mechanicals, that might not indicate multiple hooks are engaged whereas an actual physical hook to bar feedback signal might be a confidence booster for both automated and manual climbing efforts; especially when using multiple hooks like the typical paired setups.

The CAD can be found at Onshape

Some details on the design:

The CAD file as-is is intended to clamp onto a 1" square extrusion with the hook “insert” extending into the extrusion it is mounted to. There are a number of variables which control key aspects like widths, thicknesses, hook wrap angle, etc. There is also support for the locating the microswitch - the microswitch fits into a cavity in the 3D printer center insert which locates it in the center of the insert and there is a matching “plug” which fills the remaining space capturing the switch in the insert. The plug has about a 3 degree draft to it for easy insertion and the insert includes a hole centered on the inside side of the microswitch to allow the switch to be pushed out if needed. There is an opening on the back of the insert to provide wiring access to the microswitch sufficient to allowing wiring after the hook is fully assembled. The switch is located to ensure that when activated, it is reliably triggered but not pushed to the end of it’s travel so it should last essentially forever as it’s not being abused by taking any of the robot’s weight.

The entire hook “sandwich” is secured using 5 1/4" fasteners with 2 more fasteners securing the outer aluminum parts of the hook to the lifting extrusion. With the aluminum outer parts, the hook should be plenty strong enough for what we are doing. The 1/4 bolts are probably overkill and could be replaced with 10 or similar-sized hardware for weight savings.

The design can be parametrically scaled but I’m not adept enough with Onshape to handle the relocation of the microswitch in an automated manner (though I’ve got some ideas I might mess with); that means after you get the general hook geometry the way you like it, you’ll need to address the microswitch location but it’s not too bad. You’ll likely need to touch up the switch recess after doing that.

The basic hook geometry puts the center of the hook eye vertically in line with the center of the lifting extrusion for to keep the forces straight through the assembly vertical when in that orientation.

Here are a few pics to prove it happened:

For the last photo, I wired up the LED to the NO side of the switch to activate it when engaged - you’d probably want to connect it to the NC side of the switch so you could have a supervised circuit and detect a failure before grabbing a bar. I printed a short stand-in for 1 1/4" pipe to activate the switch.

Use it as you may or ignore it as you might - if you use the idea and someone digs it, tell 'em your using “Zog’s Smart Hook”.

– Chris Herzog, Lockport Porterbots #4292

25 Likes

Great idea and visually appealing. I might consider using shoulder screws or compression limiting sleeves to avoid crushing the plastic core.

1 Like

Good point and definitely an option - the prototype was printed using “burn it down” settings (15% infill with super fast travel) to be fast to prototype but for more typical use, I’d go with 4-6 outside layers and something on the order of a 40% infill. If our teams goes for it, I’d do an ABS center core but with the outside perimeters on the bolt holes, you’ve got something to snug against that goes totally solid through the insert (which is what you want at the minimum). The core shown happens to be ABS with PLA outer pieces because I put both my rigs to work at the same time to speed things up.

Even PLA could be OK here but I’m not sure about the “gooshiness” (technical term) of PETG which might cause some creep.

Something like ABS should avoid that and with some of the teams having more exotic options (no MarkForge here - bummer) having more choices.

With the current design dimensions, a 1 3/8" shoulder bolt like a https://www.mcmaster.com/94035A186/ with a pair of https://www.mcmaster.com/5360N127/ under the head and a #10 washer and nut on that end should get the length right to get things tight without any risk of crushing the insert.

In reality, sleeves might be more practical - and a whole lot cheaper as McMaster is clearly very proud of their shoulder bolts. Something like these might get the job done https://www.mcmaster.com/93441A427/. Switch to #10 bolts and a fender-style washer on each side and you’ll save enough money for a halfway decent lunch if you can stick to the value menu.

1 Like

Thinking of this microswitch?

2 Likes

That should do nicely - the 1.1" x 0.63" x 0.4" is a super common size.

I pulled the CAD from https://www.mcmaster.com/7779K51/ but worked the travel from actual switches I had on had. The travel on that switch is better than the one I pulled the physical CAD from - it matches what I I actually used for travel measurements. It’s Normally Open if you’re good with that.

The thing I really like there is the overtravel - great feature to keep the switch alive and keep the mount position a little more flexible. The key is to have protrusion to activate reliably, get a little into the over travel area for insurance, and still have some travel left for mechanical protection against damaging the internals. The switch I pulled the CAD from is a little, actually more than a little, too on the razors edge: 0.05" travel with .004" over travel - that’s going to be a demanding switch to mount.

Something like https://www.mcmaster.com/7779K41 is forgiving along the lines of the travel for the switch you referenced if you want Normally Closed behavior.

That’s a good call, I’ll add some comments pointing people to both of those switches. That’s pretty much right on with the switch I physically used.

Added a comment to the CAD document capturing this discussion covering travel and over travel and pointing to a few options including that one you posted.

Good call and worthwhile information if someone is going to try this.

That’s a cool design, and a nice compact package.

We are using two pairs of these proximity sensors to detect the bar since we wanted to avoid having any mechanical moving parts (i.e. the microswitch). It works really well, but they are much bigger than the switch and don’t provide nearly the same compact design.

5 Likes

Nice. That’s a cool implementation.

I’ve got a bunch of inductive proximity probes that I bought for a 3D printer build I just finished but I just imagined it with the microswitch in my minds eye and went that way.

These are the ones that I’ve got around (TL-Q5MC2-Z - you’ll find them all over the place for about $10). They’ll run on 10-30VDC so they are a great match for our application. About a 5mm sense distance. Repeatable enough to do 3D printer bed meshing measurements.

Clearly a few different ways to slice this problem.

Climb on!

1 Like

4004 did this in 2016 for climbing, as we could properly see from the other end of the field, and would cause us to climb slowly to know if we were engaged. Once we hit the switch the robot code took over and our climb time was cut in half.
Were looking to implement the same thing again this year.

1 Like

That’s exactly the thinking here.

I try to impress on the students that “the robot has the best vantage point” for a lot of things it has to do. A few sensors can make a lot of difference in either relaying information to the human drivers or getting the robot to handle things in an automated way.

1 Like

Would you happen to have an STL of the individual parts?

Sure - here are the STLs for a hook designed to fit on a 1" square extrusion with with 1/16" walls (the accommodation for the extrusion thickness is 0.070" as opposed to 0.063" to allow for a slip fit - the inside corner radius is “typical” for that size extrusion but a little sanding might be needed depending on your exact extrusions.

The “outside” part is an STL but my intent was that it would be made of aluminum - you can print the part as a full size template or follow the drawing I’ve attached as well. I’d probably go the template route as the drawing is a bit hairy.

Let me know if you end up doing something with it!

Sensing Hook Outside Drawing.pdf (118.8 KB)

Sensing Hook Parts - Hook Outside.stl.txt (251.3 KB)
Sensing Hook Parts - Hook Insert.stl.txt (337.1 KB)
Sensing Hook Parts - Switch Plug.stl.txt (34.1 KB)

To be able to attach the STLs, I added a .txt extension which you’ll want to remove. DM me if you have any problems.

While more complex, something I think a lot of teams and designers have forgotten is that hooks do not have to be passive. It’s entirely possible to make a latching hook, even one that releases via a mechanism, and these may be easier to detect if latched, either by sensor or by mechanical action. I’m surprised we haven’t seen more push-on “gate latch” type mechanisms for hanging this year. There are mechanical solutions to this problem too!

That said, this mechanism is very cool and a great use of printing to replace simple spacers and to accurately locate a useful sensor.

1 Like

If that plastic core is printed with compression in mind, I would feel very confident in a PLA core withstanding the compression of being tightened by hand with a combination wrench. Adding perimeters until the hole perimeters connect with the outer perimeters is one way way to do that, or other holes can be modeled into the part to cause the slicer to add more perimeters to strengthen the part. Solid PLA is very difficult to crush.

The arrangement of aluminum plates on the outside with PLA in the middle is a great way to do this, in my opinion.


All hail king Zog!

in all seriousness though, neat concept. I’ve always been a little hesitant to use microswitches, as I’m not a huge fan of their sensitivity to vibration and other environmental factors. maybe something like an M12 inductive proximity sensor would be a neater package?

1 Like

@bscharles had a setup using prox sensors and the there are some other potential form factors as well for sure - makes for a nice contactless setup. The prox sensors tend to push some specific packaging challenges due to their length but you might be able to do something by rotating them down towards the lift arm if that doesn’t impact how well they detect the bar.

The plus here is that since the robot weight is activating the switch, you can use a higher actuation force switch to forestall false activation without too much impact on functionality.

1 Like

Just built the cores with the switches and this thing is nice. Machining the plates and full assembly tomorrow. Thanks for the files. We have included, “Zog’s Smart Hook” in our technical documentation for the bot.

1 Like

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.