KB58's Power Drawbar - Electric

Using printed parts as prototypes is nice at finding goofs early on, instead of in the aluminum parts. Some of the parts may be left as PLA, pieces that see little load. Somewhat related, there is a disappointing level of friction between the stainless rods and the bearing inserts. I'll try polishing the stainless, and if that doesn't work, new parts will be made with the stainless running directly on the PLA. The bottom blocks will be aluminum due to the torque imposed by the weight and operation of the wrench.

1722020399677.png
 
Two issues:

I'm coming up to speed on using printed parts. One is dimensional accuracy, both in the 3D printer itself, and material qualities. Case in point: I'm using 0.625" rod as guides for the impact wrench to slid up and down. Setting the CAD model to have 0.625" holes resulted in them being undersized. A test part was printed with holes ranging from what's supposed to be 0.625" to 0.635". Even the largest had too much friction to be useful. Also, running a reamer through the undersized holes did not help. Apparently, PLA has enough spring-back that even with a reamer, it just springs back to being undersized. They'll have to be remade sufficiently "oversized" to end up as needed.

Another issue is that the guides aren't thick ("tall") enough; it's way too easy for them to get even slightly off-square and then they jam. I don't want to use real bearings because... reasons. It's mostly cost, but also the challenge of keeping costs down. They'll be modified with taller guide sections, and we'll see how it goes.
 
Made changes to the guide plates, extending the bottom one all the way to the top piece. This way it can't (as easily) get jammed off-axis - I hope. Again, I'm trying to keep it simple and not descending into using a support system like a 3D printer or router!

The holes had to be printed waay over size, nearly 0.020 (0.5 mm), which seems nuts, but I have nothing to compare it to, maybe that's just how it is. I suppose if I had a lower end - or self-built printer - I could mess with it for hours to fine-tune the printing of holes. Oh, and another thing the printer does is create a bit of an "elephant's foot", where the first layer or two are slightly bigger. I know this because the reamer always has to cut more at the bottom of the print than the top. Oh well, it's not a huge deal.
 
Hmm, this thing might actually work. Next is fabricating both a pull handle and a method to select rotation direction - that last bit could be interesting.

1722020620149.png

1722020662288.png

1722020783653.png
 
Nearly asleep, a solution appeared that simplifies the design. Instead of having a linkage selecting rotation direction, the impact wrench itself pivots, contacting fixed points that depress the direction slider. The changed design is overall simpler and more elegant, always a desired goal. Time to order more filament...
 
The upper and lower cups holding the impact wrench can now rotate, while the horizontal supports remain static (no pic since it looks the same). This design makes depressing the trigger and setting rotation direction much easier. One reason the guides are extended the entire height of the assembly is to avoid jamming as much as possible by the off-axis force of the pull handle. If it doesn't get stuck, the design is set and the rest of the build is easy. If it does, it gets more challenging. Having the pull handle attached to the lower guide rather than the top should help minimize side-loading, but it's still a concern for the moment.

Linear bearings can certainly solve the issue but I'm trying to get this to work as is, as both a challenge and to keep costs down.
 
The last big hurdle is connecting the impact wrench to a handle. Shown below is the first rev, one half of a bracket that'll be mirrored for the other half - it's proving tough to get right. Lacking a $$$$ 3D scanner, it's a case of endless measuring and try-it-and-see because it's difficult to measure properly. Another issue is that the shape is clunky; it'll either be overly large to simplify printing, or require a lot of support, which has its own issues. The admittedly arbitrary design goal was to be able to call this a 3D-printed power drawbar. That's already not happening because the base blocks (that the guide rods sit in) are being made of steel for the increased stiffness.

Depending how well it fits, and how many revisions it takes to make it right - and my patience - it may be replaced with simple aluminum plate and spacers in order to "get on with things." At the end of the day - after half a dozen revisions - it's pretty close. Tomorrow it'll be refined a bit more and its mate printed as well.

1722020934739.png
 
The last iteration came out pretty well, so it was mirrored and the other side printed. In these pics, the vertical extensions on the bottom cross piece depress the direction button. The idea is that the (uninstalled) lever is first pushed or pulled to select direction, then pulled down to both engage the nut and to depress the speed button.

The next task is machining the spring seats down so that the assembly hovers just above the draw bar nut.

1722021076796.png

1722021118494.png
 
Development continues, with a couple issues. One is that the offset handle tries to rotate/twist the clamp and impact wrench. The clamp was modified to surround the upper cup, transferring twisting forces into the top bar and out to the steel rods. Yes, it'll block the ventilation holes, but overheating shouldn't be an issue since it runs for only seconds.

The other issue is trigger activation. If activated too high, the socket can start spinning too fast before engaging the draw bar nut. If too low, it can't get to full speed/torque. I don't know yet whether there's a happy middle ground. If not, wires can be run into the impact wrench to electrically activate the tool, but I'd like to keep the tool unmolested for simplicity - and the warranty.
 
Two steps forward, one (or two) steps back.

I was unable to make what would have been a real simple solution work: using a fixed point to depress the trigger. The problem is that as the draw bar unscrews, it backs out further than the device's idle position. End result, at about 90% of the way out, it stops unscrewing it. Doh...

Resigned to doing it the hard way, I took apart the impact wrench... foiled again. Modern assemblies are so integrated and modularized, the trigger/speed control is built-in such that it's basically a black box. Worse, instead the normal three wires coming out of it, there are five, so who knows what's going on there.

Fine, time to pull out Plan C: operating the trigger mechanically. I can see it in my head; it's just now a matter of making it real. More work for the printer.

1722021217349.png

1722021258527.png
 
You must log in or register to reply here.
Back
Top