|
Construction Equipment If it digs, pushes, hauls dirt "off road" post it here. |
|
Thread Tools | Display Modes |
#1
|
|||
|
|||
1:10ish Reach Stacker for "conex" cargo containers
(NB: if pictures aren't visible, you can change a setting in Chrome. See https://rctruckandconstruction.com/s...7&postcount=16 )
Bit of background: I've been building things of many sizes for years, ranging from scale models to RC vehicles to a 1.5ton ride-on backhoe with working hydraulics. I hate hydraulic fluid. I've done a number of 1:10 builds, starting with a Cross RC kit (nice, but kinda dry and boring cuz it's just assembling parts) to a partial scratch build HEMTT/PLS and then to full-on scratch builds. Historically my stuff isn't 100% accurate in terms of modelling, though I strive for a passable representation). I enjoy making the various moving parts move and the overall thing work as it should, so I take liberties with the design some to make arms move or cranes lift or whatever. This build is far more representative than it is accurate; some because it's not modelled after any specific 1:1 and some because it had to be overly strong. (And I might have overengineer certain bits, but it won't drop anything, that's for sure =)) ) I had the one big conex from the PLS build, and a coupla small ones from later. Sure, I can move them onto the PLS' flat rack by hand, but it would be nice to have a matching(ish) RC to do it for me, right? I mean, My usual Googling research brought me to doing something like this So here's the regulatory teaser, not the end result but somewhere towards the end It doesn't look so bad, but note it's about three feet long with the arm, ie the frame is like 28". Here's a set of the tires next to a Tamiya-type 1:14 tire: For once I didn't have to use pool noodles; these tires came with foams that are ridculously stiff even by my standards. Before the chassis, though, I actually started on the grabbing mechanism. I figured that if I couldn't make that work, there was no point in building anything to go under it. There's a mockup using some scrap wood and an early drive system with gear motors and limit switches. In the event, the lateral drive uses actuators on linear rods/bearings rather than the screw drives: turquoise thing in the middle is the top surface of a conex; since I'm doing top-loaded I didn't bother modelling the rest of it since that didn't affect the loading.
__________________
I mean, how hard can it be? Last edited by dremu; 09-06-2021 at 10:17 PM. |
#2
|
|||
|
|||
Re: 1:10ish Reach Stacker for "conex" cargo containers
Conex size (specifically height) does however affect stacking ability, and I really really wanted to be able to stack three containers. Which is possible, just requires a long (and strong) arm.
But the end result is not only impressive, it meets the stamp of approval: Two seconds, I tell ya. I turn away for TWO SECONDS to grab the camera and FOOP, up she jumps. Amazing it didn't all come tumbling down, but she is the most nimble of the the lot despite being the largest. Unfortunately, I couldn't sort a way to make the arm telescope, or have the grabber move up and down on the arm, and still retain enough strength to safely hold stuff. The big container is like 6 pounds empty, and more than a bit unwieldy. After a few months of being totally unable to store this machine properly, I had an epiphany that the end could be removable, which makes the thing infinitely easier to handle. Basically cut off the end and printed up a socket for it: with a set of plugs for power, lighting, and an Arduino up top. A further epiphany occurred shortly thereafter, which was the ability to have multiple heads, for instance a traditional crane: I was debating also doing a high-lift fork, but (1) the arm is really far off the ground, so the mechanism woulda been kludgy, and (2) I have a fork already, so designing one wasn't shiny or exciting. With some of Actuonix's longest/strongest actuators, the arm has tons of power and a good reach: and the crane has decent lifting power:
__________________
I mean, how hard can it be? Last edited by dremu; 09-06-2021 at 09:54 PM. |
#3
|
|||
|
|||
Re: 1:10ish Reach Stacker for "conex" cargo containers
The chassis is welded steel channel, along with a ten pound steel block on the back as counterweight. (Did I mention a bit of overengineering?)
There's no suspension, so the axles are just mounted on brackets directly to the frame. Initially I'd used some Chinesium portal axles for an Axial Wraith or something, but later discovered that neither of them were equipped with differentials. That combined with the crazy wide tires made turning unmanageable, so eventually I switched to a different set of axles which could be set up with diffs. Seems like diffs are most common in the steering axle and less so in the straight, but for this setup remember it's the rear axle that steers, and since the front has the dual tires, any improvement in turn is good! Both axles are driven, shafts are solid brass rod with flats ground on the ends for the U-joints, to a dual-output gear motor. The motor was centered on the first set of axles so I was just using couplers and didn't need U-joints, all lined up nicely. When I switched axles I didn't want to re-engineer the whole thing, so just changed out to the U-joints. At least since there's no suspension the shafts don't need to slip. Steering is via hopefully-sturdy metal gear servo with aluminum arm. Left side is rear, you can see the steel block counterweight, and the silver square sticking down in the middle is the bottom of the gear motor. The top protrudes up into the cab. There's a regular ESC to drive the motor and provide BEC power, and then the electronics go a little crazy. There's a PDB by the battery, because the main lift actuators need their own motor ESC (just behind the cab, yellow and blue wires going in.) Left side with all the orange and white wires is the lighting Arduino. I have a thing for LED's and blinking and get too clever for my own good, so I'll often do a separate 'dweeno for this. As an example https://youtu.be/w3ayw7PV4zY When backing up, the white lights strobe blink-blink-off-off while the orange ones just wig-wag. When moving, the orange ones blink, and if the arm is moving or whatever, the front ones come one as appropriate to warn of possible impact. There's also a switch on the TX to pop on all the lights should you wish to induce blindness or seizures. Second Arduino controls the two actuators and reads their position feedback, so that they stay aligned. Wouldn't want one pushing harder than the other.
__________________
I mean, how hard can it be? Last edited by dremu; 09-06-2021 at 10:20 PM. |
#4
|
|||
|
|||
Re: 1:10ish Reach Stacker for "conex" cargo containers
The grabber mechanism is made to adjust to the various cargo containers I've built, which makes it wider than I would have liked.
I also couldn't sort a way to make it entirely collapsible and still retain strength (again, my habit of over-engineering.) The end plates ride on 8mm rods with linear bearings and each is moved by one of the big Actuonix actuators with location feedback. These two are NOT synchronized, the way the lift arms are, so that you can shift the end plates to one side or the other. This makes it easier to pick up a container if it's off center, or unevenly loaded, you can line up the grabber on its center of gravity and not the physical center. Too many wires? Never! Rather than try and run wires up and down the arm for all the various functions, it was easier to run just power and data from the RX. (There is one extra wire from the lighting Arduino for the rotating beacon light up there.) This worked out really nicely when I had the idea to make the end removable, just put plugs on the wires and pop. The crane head we saw up above, just has an Arduino to drive the coupla LED's and the winch servo and then some interconnect wiring. For this one, the grabber head, it's more complicated. There's an Arduino that handles the various movement functions, and then the two red boards with the heatsinks are motor controllers. Two run the actuators that slide the grabber end plates, and two drive the center knuckle motors. Those are the silver gear motors on the right and forward sides. These are geared WAAAAY down, to like .5RPM (yeah, like 30 revs per HOUR), so they can directly drive the knuckle mechanism. That knuckle is M8 rod welded in a cross, riding in 608 skateboard bearings, and then coupled to the motor output (left side, blue aluminum behind the wires.) The side motor pitches the whole thing fore and aft, and the front motor rolls side-to-side. The entire mechanism also yaws, rotates around a vertical axis, using a smaller actuator in the back (upper right of that last pic.) Also, note the little blue circuit board with four wires between the arm end (aluminum tube, top center) and the beacon. We'll get to him in a minute. The net result of all of this is that the container under load can be shifted in all three axes: Now, obviously you wouldn't actually want to tilt it that much, it goes to like 30*, but the pitch is relative to the arm angle. That means that as the arm rises, the pitch angle changes. You can just keep shifting the one control up and the other down to level it yourself ... or the Arduino can do it for you. I had some digital gyroscope/accelerometer chips left over from my wildly unsuccessful attempts to do a self-levelling quadcopter; that's the little blue circuit board above. I've no idea how it works inside with no moving parts, but the chip spits out a signal as to how far off level it is in degrees. The Arduino can read that, and then carefully adjust the pitch and roll motors to bring the load back to level. This way you're level when you pick up the load, lift the arm, and it stays level through the whole process. Now there's no guarantee that the load will be truly level when you go to pick it up, so one switch on the TX turns the auto-level on and off so you still have manual control. Also, while in motion the auto-level wants to over-correct, so it should be turned off while driving. Finally, I coded t\he auto-level so it won't kick in if it reads more than so many degrees off, under the assumption that either it's getting bogus data (which I never found to happen, but, you know, Just In Case), or that you intentionally are way off level. Once you get less than about 10 degrees of level, you can enable the auto-level and it will wiggle the load into place.
__________________
I mean, how hard can it be? Last edited by dremu; 09-06-2021 at 10:23 PM. |
#5
|
|||
|
|||
Re: 1:10ish Reach Stacker for "conex" cargo containers
The grabbing mechanism proper is M4 bolts turned down to a point on the ends. All of my containers have a standardized spacing for the top holes. Trying to make the pins motorized would have made an already complex mechanism excessively so, and reduced strength.
I've found it easier to get one side lined up and then combine shifting sideways with closing the pins to grab the load. Once loaded, it can be lifted and moved easy peasy. Same holds true for the smaller sized containers
__________________
I mean, how hard can it be? |
#6
|
|||
|
|||
Re: 1:10ish Reach Stacker for "conex" cargo containers
The complexity of controlling the thing made for complex needs at the TX. I love me the FlySky FSi6's, cheap and hackable, but this build took a bit more hacking than others:
Both sticks return to center. Right stick is traditional move/steer. Left stick is pitch and roll for the knuckle mechanism. Note to self: I really should print little plates for the switches, done that on other builds so I don't forget which does what. From left to right: SwA. auto-level on-off. SwB. yaw (rotate), changed to side-side. VrA. open-close grabber. VrB. shift grabber, changed to side-side SwC. raise/lower arm. SwD. turns on ALL the lights, for testing or just for laughs. SwE is the only one I didn't end up using, call that a pretty good maximization of the TX and the custom firmware. You can't really tell here, but some move side-to-side vs up-down. Also, some were originally pots; dropping in a switch meant also soldering resistors across: (yes, the yellow and white wires are hanging loose, that's mid-work.)
__________________
I mean, how hard can it be? Last edited by dremu; 09-06-2021 at 09:52 PM. |
#7
|
|||
|
|||
Re: 1:10ish Reach Stacker for "conex" cargo containers
So that's it for pix. I did take some videos, some of which came out appalling awful and others only pretty bad.
https://youtu.be/w3ayw7PV4zY Strobes (shown above) https://youtu.be/7cMXQjp9Sdo Arm lifting, with the crane head https://youtu.be/MqMlcEXy6zE Crane lifting a load. The speed is set so that lifting is slower than dropping, so, umm, like watching paint dry, sorry. https://youtu.be/X3aSlXKV8QE Yaw (side-to-side rotation) of the grabber. This is from behind, front of the machine to the top; actuator on left makes the magic happen. https://youtu.be/t1RPzq_JTCI https://youtu.be/CeFtVl67hFk First video is opening and closing the grabber ends but also shifting left/right (sorry, hard to operate the camera and the TX and the same time.) Second one is just the lateral shift. https://www.youtube.com/watch?v=5FcrUck-W24 This one shows the autolevel function. I know I committed the cardinal sin of taking the video sideways, but wanted to show the range of the arm and how the grabber moves accordingly. When unladen it does wiggle a little, overcorrects slightly. https://youtu.be/vEsK3pcwzVg And when laden (and the camera properly oriented, albeit for a shorter lift), you can see the auto-level do its thing. I'm only lifting the arm there; the auto level is adjusting botm pitch and roll to keep the thing level. (The red LED's strobe for auto-level, vs just orange for manual control.) And HRH Duke sticks his head in, unsure if he's curious and should make friends, beat the thing senseless, or run away at warp speed. https://youtu.be/vDs547fVBHU Full speed ahead! Yeah, that's top speed for the machine. The gearing was intentionally chosen to given lots of torque and ease of control vs speed. https://youtu.be/zNb9ccB4vJQ Were this a fork, you'd want the load as low as possible while moving to avoid tip over. However, this thing has the insane counter weight in the back and wide stance, and you need to see around the conex, so maybe moving with it up like that makes sense. Anyway, again in the vein of "I can't operate the camera and the TX simultaneously", that's the big conex being shifted around. https://youtu.be/1rgeLK44IZc I think I hit stop way early on that one, but gives you an idea of loading with the little conex. https://youtu.be/bEb_0MIic0g This one is longer and prolly more informative. I notice I broke my own rule and have auto-level turned on while moving; was really focused on lining the thing up.
__________________
I mean, how hard can it be? |
#8
|
|||
|
|||
Re: 1:10ish Reach Stacker for "conex" cargo containers
End result is not as pretty as, well, the majority of builds, not visually accurate to much of anything, but certainly functional and I find it fun. Just the grabbing of the containers requires patience and care to catch each corner Just So.
-- A
__________________
I mean, how hard can it be? |
#9
|
|||
|
|||
Re: 1:10ish Reach Stacker for "conex" cargo containers
That thing is a monster, I love it. Amazing amount of engineering you've put in. I really like the amount of flexibility you have built into it so that it can handle loads on most any kind of surface from almost any angle of approach. Really cool is the use of the gyroscope/accelerometer board for leveling. I never thought of them for use in our kind of builds before. Got me thinking now of an auto leveling bucket system for a front end loader, hmmm. And for something like this who cares that much for appearance. As you said it's the fun of figuring everything out to make it do what you want it to do. For something this complicated that's major satisfaction. Good work.
|
#10
|
|||
|
|||
Re: 1:10ish Reach Stacker for "conex" cargo containers
dremu:
Couple of questions for you: 1 - What sensor did you use for the pitch and roll detection? I've got an idea rattling around in my head that I can use an Arduino to auto-level a Manitou 2150 crane project that I've been working on, and possibly on another crane project that's in the future work queue. I've never done anything like auto-leveling before, any suggestions where to look for some example code? 2 - You've got your guide rods fixed in the center of the grabber and the bearings on the movable outer ends. Did you try it the other way around, and let the guide rods move with the outer ends? Linear bearings for 3D printers are dirt cheap - about $0.50 apiece for REALLY cheap ones and about a buck apiece for better quality. You could get sc8luu 40mm-ish long bearing blocks and only have to have one set on each side of the grabber. A decent sc8luu bearing block is about $10 each. That would allow you to take a several inches off the guide rods for each side. It looks like you could wind up with a unit that probably wouldn't be much wider than your smallest container. Don |
#11
|
|||
|
|||
Re: 1:10ish Reach Stacker for "conex" cargo containers
That is some awesome engineering! Nicely done!
|
#12
|
|||
|
|||
Re: 1:10ish Reach Stacker for "conex" cargo containers
Thanks all for the good words!
The gyro boards are sold as both GY-521 and MPU-6050, Amazon, Fleabay, Aliexpress would all have 'em. From that you can Google the code, easy enough, or in a pinch I can dig up mine and post the relevant sections. Wrt the stupidly wide sliders, ddmckee, your idea sounds eminently sensible. I'll admit I started this build like a year ago and so my memory is a bit fuzzy ... I might have had a good reason to do it the backwards way I did, or I might not Maybe I couldn't get enough reach with the hardware I had to grab both the large and small conexes, or maybe when I tested it there wasn't enough strength in the center and they drooped. Or maybe I just didn't think of it (slaps forehead.) I'll see if I kept notes about it, which I do sometimes, and look at the earlier CAD work. Otherwise that might a great way to reduce the size of that head. -- A
__________________
I mean, how hard can it be? |
#13
|
|||
|
|||
Re: 1:10ish Reach Stacker for "conex" cargo containers
dremu:
I know what yopu mean about long-term projects, I started my Manitou build in November, 2019. I know what you mean about fuzzy memories. I started posting that project on another forum, so I can go back and get some idea of "What the **** was I thinking?" I've done a lot of 3D design work on the Manitou and printed out some prototype parts - testing to be sure they'd work. If I printed out enough parts, and did enough cutting and hacking on the lower chassis half, I could make a rolling chassis at this point. I've got the Arduino Nano clone controlled lights working on a breadboard. About that time I had a brain-fart about a servo controlled switch that I could use to switch 1 ESC between multiple motors, instead of using 8 channels for just the leveling feet. It would've worked... but I never really liked that idea anyway - too clunky. Then I discovered "Mick thebass" and his youtube channel, that guy can really pack 10 pounds into a 5 pound sack! He got me wondering how much of the work I could make the Arduino handle. Then the Devil put the thought in my head that I could make the Arduino auto-level the crane on command. This RC conversion stuff is a REALLY, REALLY, deep rabbit hole isn't it? Don Last edited by ddmckee54; 09-10-2021 at 03:23 PM. |
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
|
|