I've been making outdoor tables with curved steel legs. The need to bend heavy solid material in a uniform and controlled way resulted in a fairly simple machine that allows for innumerable rates of curvature and chord lengths.

The table you see in the image has four main legs that are made of 1/2" solid steel square stock. The legs are suited to the strength and stiffness required while also having the fluid quality that was desired.

This bending rig fit my needs and has changed how I bend things. I thought I might have to buy an expensive manufactured product, but this was both cheap and is highly adaptable.

Depending on your needs, a modification may be suitable. Whatever the case: get help checking that everything is strong enough to fully load the system....there are huge forces running through this type of device.

I hope this helps you save a few bucks and many hours when trying to make smooth bends in solid materials.

Step 1: The Basic Pieces: Front and Side View

The system consists of five main elements: An "I" beam base, a bottle jack, four pieces of threaded bar, two large angles and a range of bending templates. There are four bolts, various washers, rollers and a bunch of nuts.

The photos above show the machine without a bending template in place.

Step 2: Close-up Details

I'm not going to say this is complex.....it's just odd to write down what is simple. If you have questions, please send me a note and I can both answer the question and improve my notes here.

The first image of the front shows numbered holes in the pair of angles. One bolt without a nut is in hole #1. Adjusting the bolt positions (there are two bolts in two different holes at any one time) will adjust the chord length of the circle you are bending. Let's call those "chord length bolts". I number the holes such that I can record the criteria / parameters for each bend. The bolt on the upper left edge with a nut is a spacer bolt.

The second image is a side view and shows you a few things: The topmost bolt is the spacer bolt and the washers act to keep the angles apart to the right width. I'm bending 1/2" material I shimmed the distance to just over 1/2". I believe the spacer bolt nuts are best if slightly loose: this allows the angles to move a bit and keeps big loads out of the spacer bolts. When assembling, you need to make the angles all one distance from the bottom beam; this is done by leveling with the nuts on the threaded rod. As The chord length bolts have a simple short piece of steel pipe between the angle to act as a crude roller. As you place load on the system, these will roll, thus reducing friction on the bolt and piece you are bending. I've not found the need to use nuts on the chord length bolts, but I did select heavy bolts that were not threaded for the length that is within the angles. This put the load on a solid, cylindrical section of the bolt. (I don't think this is critical, but it makes things cleaner than smashing threads in the first few operations)

Step 3: The Bending Templates

As you can see in the first photo, I made four different templates to radii that suited my project. These are very simple: made of heavy 5/16" plate. The base has a hole in it, slightly bigger than the top of the piston of the bottle jack. The jack presses on the edge of the curved plate, not the base.

In the second photo you can see the bending template on the bottle jack, lifted to just into the space between the two angles.

In the third and fourth photos you can see the largest and smallest bending template in place. I note that the chord length bolts in the image with the smaller template are not in a useful position: i.e. they would be much closer so you create pressure between the bolt and template (more on this later).

Step 4: Bending the Bar

I mark the center of my bends on the top face of the bar, as that has proven to result in consistent bending. When you set a bar in the machine, you will be observing it's position through the open holes in the angles. Because of this, I mark the bend location with a magic marker and then a soap stone either side of that mark. Its simpler to see.

The second image has a bar loaded in the machine with a tail extending off to the left. The jack is centered between the chord length bolts and the bending template is centered between the angles. When you build yours consider this: the vertical position of your chord length holes is important. Ideally you would have the top of any bending template about 1/4" above the bottom of the angles, then space for your maximum stock thickness you want to bend, space for the thickness of the roller and then the bottom of the holes. I.e., the bending template must start the process by being between the angles, not below them, and you need space above that for a piece of straight stock to fit under the chord length bolts.

The third image is a completed bend. I note that in all bending operations, I drive the jack piston all the way to a tight fit between the bent bar, the bending template and the chord length bolts. This makes the bends consistent, rather than say, trying to measure travel of the piston somehow.

The fourth image has the bent bar placed on the machine.

Step 5: Multiple Bends, Multiple Legs and a Table

The first image has one bar bent three times to suit the design. I note that I cut and mark the position of center of all bends on all bar leg sock in one operation, followed by one type of bend on all bars. I move on to the next type of bend and pass all bars though the machine. For these legs, everything went through the machine in groups three times with three different templates.

The second image shows a stack of four legs all bent to the same fluid profile. (No this is not aircraft tolerances and a little tweaking may be required if you are bothered by slight differences. These were not tweaked. The ends at the last tight bend need to be trimmed, but that will happen after I've welded them to the table top.) All told, from cutting the bar to lengths and up to completing the bends, takes under an hour. Setting up the initial mock-up and defining the lengths, rates of curvature and center points of bends requires an amount of time determined by your own ability to get to understand what you want as a final piece. Running material is very simple. As an example: my list for this run looks like this: 44" pieces, first bend at 8 3/4" with the large template and bolts at holes #1 &16, second bend at 26 3/4" with the middle template and bolts at holes #7 & 16, third bend at 40 3/4: with the smallest template and bolts at holes #7 & 11.

The final image is a completed base using four legs from this process. Other bent elements in the base are in smaller stock and some have much tighter radii. I may outline that bending tool in another Instructable.

I have my tables galvanized, hence the light color.

I hope this helps in whatever project you may be considering. Call me cheap, but I think I ended up a superior tool when compared to anything available.

As before, get some help selecting the jack, bolts, the beam, template stock and angles. You want this to work the first time as well as be safe.

Ross, I've spent about 2 hours now, and redesigned your tool 3 times, and each time, I haven't improved anything. In fact, part count went up and ease of use went down. I really like that your design can be disassembled for storage. <br><br>It takes smarts to make something that works. It takes minor genius to make it simple.<br><br>The only thing I will change is to make the base from angle - a mirror image of the top - to decrease weight and make storage even easier. Other than the jack, it should all fit into a shoebox.<br><br>I need this in order to bend 3/4 iron pipe to make a special chair for my wife. I'll probably head over to the hardware store for parts next week.<br><br>If you felt like sharing the design for the smaller bender, I would very much like to see it. <br><br>I have adjustable bending pegs for freehand bending, but as you know, making curves that match is difficult. I have never commented before, but I just had to. Well done! <br><br>Katou
K,<br><br>Thanks!<br><br>I don't know if I rediscovered something old or discovered something new. It works though.<br><br>As you know, bending pipe has an interesting challenge in most curves: you have to induce even compression on one side of the pipe to create fairly even tension on the other side. It will be interesting to see how you do. I suspect that there is some sort of saddle shape you can make for each bending template face. <br><br>I'd love to see what you come up with.<br><br>R
My plan is to use a piece of the pipe itself as the edge of the template. Cut it in half lengthwise, open it up by beating on it. That should minimize the crushing. It is about 3/16&quot; (3-4 mm) wall thickness, which should also help.<br><br>It will take me a bit to build it, things are pretty busy here. Look for an update in a couple of weeks. <br><br>I think you formulated something new. I have been looking at designs for a couple weeks, and none have the exact same set of solutions.<br><br>Here are the unusual features I see in your design:<br><br>-Threaded rod acts as a spring, centering the piece directly over the ram at all times, for any thickness of material<br>-Side plates fit closely to work, which ensures the template is guided front to back<br>-Because the friction is greatest in the centre, the work itself holds the ram from deviating left to right. This is important because bottle jacks do not tolerate side loading .<br>-Can be broken down for storage or to make parts larger/smaller<br>-Requires very little welding<br>-Templates (dies) are cheap and fast to make<br>-Does not require special rollers for square/ rectangular work<br>-Uses standard parts available anywhere (if angle used for base)<br>-Does not require mounting to a bench<br>-Easily modified for different work - round, diamond, flat bar<br>-Can bend &quot;on edge&quot; about as easily as &quot;on the flat&quot;<br>-Adaptable to very large/odd shapes easily (you could easily bend a piece 3-5&quot; wide (75-120 mm) or a piece that was curvy by widening the template.<br><br>There may be others, but these are the things that attracted me.<br><br>Have you considered sharing the details of the smaller bender? <br><br>Now I'm wondering what other clever things you've built! <br><br>Katou
<p>By placing wooden cauls between the ram and upper anvil, you've also got a nice clamping press for glue- ups at no additional money or redesign, this is the sign of a very useful shop tool, well done on this project!</p>
​Man, who doesn't want your title! Thanks for the comment. I started out on the project thinking that I would bolt the jack to the I beam. While I assembled the bender it became apparent that it would be much more useful to have the ability to move the jack around and or use the jack the shop for other applications. As always, the simpler the better.
You could still attach the jack to the beam without losing the ability to adjust it laterally. It'll require more machining, but all you'll need to do is drill the base of the jack (2 holes will probably be enough if on opposite corners, but if you drill 4, make sure it's as square as possible), jack your largest template in place to center the jack, use those holes to drill into the beam as many times as you need making sure they're as parallel as possible, then find a way to use those holes to mill out a slot lengthwise as far as you need. Then it's just up to getting the right hardware, and there ya go, added stability and versatility (with a minor increase in fiddlines)! And if you mess up making things square and parallel on the beam, just mill the beam slots a little wider. You could also add a bit of measuring tape to the beam to help with reproducing fine adjustments, but it'll need to be countersunk to avoid getting marred by the jack. Hope that makes sense!
<p>Hamarito, Yep, lots of options for modifications. My first reaction to your idea is that in my set-up I have a 12 ton bottle jack in this puppy. 24,000 lbs and I use it all to complete the bends. All of that load has to be transferred across the lower beam, through the threaded rods, through the angles, through the chord length bolts, through the piece I'm bending, through the bending template and back into the bottle jack. (sorry to be overly pedantic here). In short, creating long gaps in the top of the beam as you describe is certainly possible, but would have to be looked at in terms of loads on the beam. After your note, I considered that I've used this thing more than 80 times and nothing has ever shifted under load, so it's unlikely there is a big safety issue lurking. What might be a suitable addition for me, from your suggestion is a simple rail on top of the I beam that acts as a &quot;fence&quot; (like a table saw) for the back of the base of the jack. This would center the jack under the angles above. I'm going to make that change.New ideas evolve out of others. Thanks!</p>
<p>Glad to help! I was donating blood while I wrote that, so I'm happy it translated. Now that I'm on a screen larger than my phone and not a pint low, I see details I neglected and some issues with my suggestion, but there are options. Now it's pretty obvious you prefer to keep the jack centered between the four support rods to split the pressure evenly, which also means the stability of the jack isn't much of an issue per se, but adding a rail like you mentioned would help ensure things line up properly. My big problem though is a background in stats and the medical field. It might actually be completely safe as is, but 80 uses out of however long a person plans to use something with this dangerous potential isn't all that many data points to pull from. It only takes one catastrophic failure to be, well, catastrophic. One of the most common comments coming through the ER from craftspeople and professionals was an over confidence in either their tools or their own ability. <br><br>Anyway, I'm largely thinking out loud for my own build, but the point of what I suggested is the ability to move the jack from left to right and lock it into place (which would be needed at that point) could be useful to create new, finer, and more interesting bends, corrections, and/or alterations, potentially without the need to reset the location of the spacers and/or swap templates. It also frees up about the full span of the bending area (rather than splinting it 50/50 with the jack always centered) allowing you more ability to bend material with one very short leg and one long, about twice as long as previously capable, rather than making a bend then cutting off excess material. I could be completely wrong, but it could give the ability to easily make complex, flowing bends close to things like the Nike swish. You're still able to do similar things with how it sits by just moving one spacer closer to the jack while leaving the other distant, but now we could be looking at stability issues again. Beyond all that, it could even be useful for unusual, off-center bend templates. <br><br>Then again, I don't have the tool in front of me, so you're better set to consider if that'd be useful, redundant if we're really just shifting the same geometry, or just overly eccentric because it doesn't add all that much additional function. Once I get around to making my own, I think might end up run with it, beefing up the threaded rods and going with a section of H-beam for the wider flange. I didn't notice originally, but your beam and jack are the same width. Passed all that, I'm also considering how to go about making the top angle iron easily adjustable to accommodate a range of material sizes. Could be as simple as using larger angle iron, doubling the threaded rod, milling a slot accordingly, and using larger, possibly grade 8, fasteners where necessary. And not to forget, adding some sort of measure markings. Making the changes for me shouldn't sacrifice the previous functionality, so nothing ventured, nothing gained. <br><br>All in all, thanks for the great design and inspiration!</p>
H,<br><br>Two rods can take 100% of the load. I have a factor of safety of about 1.2 when the Jack is hard to one side (which I avoid) and a factor of safety of 2 when centered. So do I think that is low? Maybe. I have no idea how I could ever double the load.<br><br>Typical factors of safety cover unusual abuse, use, wear, impact loads and user error. I think I'm good on all but the last one!<br><br>If you think of anything, please shout.
<p>IMO, you can go with a minimal safety factor, given the possible failure mechanisms. The situation in which two columns snap at once is highly unlikely, IMO, and if a single column snaps, there's zero risk for some piece of iron to take off at high speed in a random direction, since you don't have anything rotating or otherwise moving at high speed while working, and in case of a snap the remaining three columns will keep everything mostly in place. The jack will stop pushing after a few millimeters, not enough to rip some small metal piece apart and send it at high speed through the air - which is what you'd need for a bad accident, since your hands are nowhere near the jack's pushing head when operating the rig.</p>
Ok! Stream of consciousness, love it.<br>I've worked big construction all my life, been an EMT as Search and Rescue. I know horrific accidents.<br>I ran numbers on the elements and gave myself for an amount of asymmetrical loading, but not 100%. I should check if two rods will take the load. That said, the obvious here is that my description avoided sizing members for people and drove the point home about big loads. <br><br>One form of insurance I'm going to get is a full face mask.<br><br>I've thought about asymmetrical bends &quot;Nike Swoosh&quot;, but this rig is based on balancing loads on top of a jack. Everything is pretty much centered. I think rotation of the bending temple would be a very real problem with a swoosh. There might be a variant that works. Jacks are very bad news with side loads.<br><br>The comment on varying the width of the top angles for other material: it will work with this rig. I could probably go thinner to zero and wider to about 3/4&quot; by just adding washers. Note the length of the threaded bars is so great, they would just spread a bit. At some point, you'd drill four new holes in the I beam further out to allow for greater spacing of the angles.....after you checked the loads.<br><br>Thanks, and tell me how yours goes! Stay safe.<br><br>
<p>You can build a table from two flat pieces of 10 mm sheet metal, with ribs of the same material welded onto one of them in a grid pattern, and the second plate welded just on the edges. The grid pattern will prevent the ribs from tilting even if not welded on both edges/faces of the table, and the ribs will provide excellent rigidity. You could then use a few plates of decreasing size stacked atop each other, centered on the face where the jack touches this table. The table could have pieces of pipe fitting tightly on the four legs, to prevent it from rotating itself. The bending templates you could then simply screw onto the table, and shift them in any direction you like. Plus, building two such tables, sliding one onto the legs to be pushed up by the jack and screwing one in place at the top, instead of the bending rig, gives you sort of a home-made hydraulic press which you can use with other molds too, not just bending templates.</p><p>I hope I make sense.</p>
<p>I'm trying to come up with a comical way to put this, but I haven't figured out the mechanics (ba-dum-tss). </p><p>I figured you'd be able to simply add or subtract a few washers to allow a<br> decent range of materials, but I've some 1&quot; and 1 1/2&quot; bars that would <br>be nice to form without need of a forge or heavy machinery. On that note, heating up the stock before bending should help with the tensile stresses and might even allow the ability to bend multiple bars at the same time if something calls for concentrically bent parts/designs or you just want to bend thicker or rectangular stocks. And if snaking is an issue, simply leaving it to cool in the form should be enough to help it stay how you want it, and clamping the material that passes above the work area would also help.</p><p> Despite somewhat valid contentions from some engineer friends (which mostly boil down to gripes about cost), I'm more than happy to over-engineer, but an FoS of 2 is probably totally fine so long as you don't try anything too flashy. However, there is an easy modification that could double the load, reopen options for asymmetrical templates, and solves any tipping issues: 2 bottle jacks. Assuming they're spaced properly, you'd also need to rig it up so they'd extend at about the same rate to ensure the load balances as much as possible as the bend forms, but without running the numbers, it's hard to say if making a simple adjustable lever (so you could change the spacing of the jacks) in the vague shape of a field goal post would be good enough. But that's one reason why I like to over-engineer strategic points - if the tool material's failure load is high, it means there's plenty of MoS, so fewer requisite calculations, unless you want to be absolutely certain or just want to know the maximum amount of stress you can throw at it. Plus it really increases durability, which could be important if stress fractures start to form after many repeated loads due to metal fatigue. Anyways, it's understandable if you don't think it's a good option for the existing set up, unless you want to get smaller bottle jacks.<br><br>And now I'm onto thinking about any needed modifications if you also wanted to make multiple bends at the same time, like if you wanted to quickly make an &quot;M&quot; shape. Probably should stop there before we end up with a manually operated industrial grade stock bender capable of producing the frame of an '83 Corvette in one shot. </p>
<p>Nice design. Simple, elegant, robust, adaptable. Well done.</p>
<p>Great solution. As I need a tube bender at a reasonable price, some days ago I thought a jack and some hardware would be the solution, but could not figure how putting it together. You hace such a great design here!! </p>
<p>Excellent Idea Sir! Keep it up and I hope you share more of your great ideas with us. </p><p>Thank you</p>
<p>Really impressive! A BIG thanks for sharing with us all.</p>
<p>Great instructable on a clever idea.</p><p> As for the question about asymmetric bends the answer may lie in using two or more smaller jacks to press the template against the material. </p><p>Dont forget that you can always preheat the metal to a greater or lesser degree to make the bends easier and less stressful. </p>
<p>good thinking works great going to try pipe and adjust it to accommodate it thanks for the simple way to make an end did it old way this is so much easier Bruce Ellington</p>
Bruce, you are a braver man than I. When you bend pipe, I'd love to see how it goes and what you built. I'm relatively new to steel.
<p>What a way to do it thank you so simple and effective </p><p>Thanks for your ideas effort </p>
<p>A verry handy tool tu use athome,</p><p>nice idear, gonna copy it, thanks.</p>

About This Instructable



More by ross asselstine:A Great Bending Tool for Solid Material How to Build a Firewood Fence 
Add instructable to: