but then a couple of weeks ago that rolling cheap tools outfit Cummins showed up nearby for a day. They are a company that sends trucks full of cheap Chinese tools (think Harbor Freight) around the country for one-day stops at hotel ballrooms, and hawk their wares. Most of what they sell is total garbage. Dont even think of buying power tools there, or anything you need to be sharp, or precise. But if what you want is a cheap set of welding clamps, or a hydraulic jack, or whatnot, you can outfit yourself with something that will do the job but doesn't look very pretty. Such as this hydraulic pipe bender. 12 tons of bendy goodness; came with dies for up to to 2 inch pipe. I looked at that, and it's $99 price tag, and decided to purchase it, and build my own hoop.
For this first one, i used plain old 3/4 I.D. black pipe from the plumbing dept of Home Despot. It's generic mild steel, with a nominal wall thickness of .113. Thats plenty good enough, strength wise, for a 36 inch diameter circle.
The pipe bender works by using the built-in hydraulic jack to push the pipe against 2 fixed pinions, using curved dies to avoid crushing the pipe walls. Unlike a ring roller, it is not a continuous feed device. You have to bend a bit, back off, change position, bend a bit, etc. To get a nice consistent bend, and have it come out to the size i wanted, i'd need a few more tools.
These are inside and outside radius guages. I made them out of plywood, with a jigsaw.
I started out with 72 inch pieces of pipe, each one to be bent into a half-circle. I marked each one with a set of numbered marks along its length, which i could then use to mark where i needed to bend
stuff next. At first, i just passed the pipe through the bender, applying a slight bend at every other numbered mark. eventually, the shape started getting close to the right radius, and then i used the radius gauges to apply more carefully located bending. It was actually very easy to use the gauges to figure out where to center changes about, and use the numbers to apply force at the right places, almost exactly like truing a bicycle wheel. This is what it looked like from above:
The first half circle i overshot on, and had to figure out how to unbend it. It ended up useable, but not as consistent as i would have liked. No matter, i would need some small pieces for testing later anyway. The next two half circles took only about 1.5 hours to make, alternating between bending at the machine and testing with the radius gauges:
Once i had two pipes bent, i cut off the ends with a pipe cutter, and welded them together with the TIG.
This is where i decided to not use black pipe for the next one. It's a bloody PITA to remove all of the protective coating from the pipe, and if you leave any behind, it boils off in your weld pool and you don't get good welds.
Hey! I've got a circle!
Next, i used the CNC plasma table to make a few of these tabs out of 1/4 inch plate:
Went outside in the driveway, set up some work lights, and got the MIG welder out. I cut off two small pieces from the first half-circle section, and welded a tab onto each one, giving me these:
Then i stuck two onto the actual hoop. Let it cool down. Wrapped it with athletic tape. Here's a shot:
So at this point the hoop is essentially done. The only question is, would i actually trust it? That's where the other project i've been working on comes in. It's "The Machine That Breaks Things". Or, more specifically. a tensile stress testing machine. Essentially, it is a beefy frame, with a couple of hydraulic jacks and cables, that can pull on something REALLY hard. In addition, an S-beam load cell sensor and indicator displays the precise amount of tensile force in the system. Here's a picture of it, in it's current incarnation. Standing to the left of it is the cnc plasma table, which cut out the tabs.
In the photo, the object being tested is one of the two test tab-and-curved-pipe sections produced in Step 5. Here's
a closer shot; taken at about 1550 lbs tensile force:
I took each test piece and tested them up to 2000 lbs. Given that we will rig the hoop off of a loop of dynamic climbing rope, thats well over 8X the maximum force we ever expect to generate on the hoop (which has 2 tabs).
No signs of fatigue or deformation after repeated cycling. So yes, i'll trust it.
Here's klingonlandlady on the final result: