Every combination of tube and tooling has a radius below which you stop getting a bend and start getting a kink. Knowing roughly where that sits for your material saves you finding it the other way, with a full-length stick of DOM already in the machine.
What's happening inside the bend
Three things are happening at once inside the bend, and they're pulling against each other.
The outside of the bend is in tension. The material stretches, so the wall gets thinner there — and that's where an inspector's ultrasonic probe goes.
The inside is in compression. Compressed material has to go somewhere, and if it can't thicken smoothly it buckles. That's a wrinkle.
And the cross-section wants to collapse toward an oval, because the tension on the outside and the compression on the inside both pull the walls toward the neutral axis. Ovality.
Every bending problem you'll have is one of those three winning.
The two numbers that predict trouble
Ratio of CLR to diameter — the "D" number. 3D is comfortable. 2D is demanding. Below 2D you need real tooling and you'll see significant thinning regardless.
Ratio of diameter to wall — D/t. This is the one people ignore. A 1.75" tube at 0.120" wall has a D/t around 14.6 and is reasonably robust. The same tube at 0.049" wall has a D/t near 36 and will wrinkle and collapse at radii the thick one handles without complaint.
Thin wall and tight radius together is the combination that needs a mandrel. Either one alone is usually survivable.
What a mandrel actually does
A mandrel is a plug — often a ball mandrel, a series of linked balls — that sits inside the tube through the bend, supporting the wall from the inside so it can't collapse inward or buckle.
It doesn't stop the outside wall thinning. Nothing stops that; it's a consequence of stretching material. What it stops is ovality and wrinkling, which is what makes tight radii in thin wall possible at all.
The supporting cast matters as much as the mandrel:
- The pressure die holds the straight section against the bend die so the tube can't pull away.
- The wiper die sits right behind the point of bend on the inside, filling the gap where wrinkles would otherwise start. Tight-radius bends in thin wall fail here more than anywhere.
- Lubrication, because the mandrel is sliding inside a tube that's gripping it.
This is why mandrel benders cost what they cost. It's not one tool, it's four tools that have to be set up in relation to each other.
Non-mandrel, honestly
Most garage rotary-draw benders — a JD2 Model 3, a Pro-Tools 105, a Baileigh RDB — are non-mandrel, and they're fine for the job most people are doing.
On typical cage material at 3D and above, a well-set-up non-mandrel bend gives acceptable ovality and no wrinkling. The tooling is a fraction of the cost and setup is minutes rather than an afternoon.
Where it gives up: thin wall, tight radius, or anything where appearance on the inside of the bend matters. Exhaust primaries in 16-gauge at 2D is exactly the brief that eats non-mandrel bends.
If you only ever bend 1.75" × 0.120" DOM at 5.25" CLR, you will never need a mandrel and you shouldn't buy one.
Reading a bad bend
Each failure tells you what to change.
Wrinkles on the inside. Not enough support behind the point of bend. Wiper die missing, worn, or set back too far. Also happens if the pressure die is loose and the tube is drawing back.
Flat spot on the outside. Insufficient internal support, or you've simply exceeded what that wall can do at that radius. If there's no mandrel in the tube, this is the tube telling you it wanted one.
Oval section throughout. Normal in small amounts, with 5% a common ceiling. Excessive and uniform usually means the bend die groove doesn't match the tube, so it isn't being supported around its circumference.
Hump or dent at the start. Pressure die setup. The tube is being pinched as it enters the bend.
Cracking on the outside. You've exceeded the material's ductility. Different material, bigger radius, or in the case of an unusually hard heat, different stock.
What this means at the design stage
Two things worth fixing before you draw rather than after you bend.
First, if your rulebook has a minimum wall and you're bending near the tooling limit, order above minimum. The thinning happens on the outside of the bend and that's where the probe goes. A member bought exactly at minimum can measure under it afterwards and fail, and the paperwork from the mill won't help you.
Second, pick radii you can actually bend with what you own. Not the radius that looks right in the model — the radius on the die sitting on the shelf. A frame full of 2D bends in thin wall is a frame that needs equipment you don't have, and it'll be beautiful on screen right up to the moment you try to make it.
Comments
Loading comments…