The Short Answer
If you're sourcing a tube bending machine and you've opened with “CNC or hydraulic?” — you've started in the wrong place. Roughly 70–80% of whether a bend comes out right is decided by tooling geometry and springback compensation, not by tonnage, brand, or control system. I've watched buyers spend three weeks comparing servo axes and then accept a die package that had never been run on their actual material.
Run the sequence backwards instead. Tube spec and tooling first, bending method second (draw, roll, press), machine type third, supplier fourth. That order shortens your RFQ and cuts the number of trial loops you'll pay for.
Why You Can Trust This
I'm the quality and brand compliance manager at a metal fabrication shop in Ohio. Anything that comes into the shop gets bent on our material before it passes me. In 2024, I ran 214 first-article inspections — anything from a two-axis hydraulic bender up to a six-axis CNC machine. About 31% of them bounced on tube geometry. Not late delivery. Not surface finish. Not the wrong voltage.
When I pulled the rejection log and looked at the reasons together, they all pointed at the same thing: a die set that didn't match the material.
It took me about 18 months and somewhere around 60 quotes to understand that the machine spec sheet is the least predictive document in a tube bending machine sourcing file. It tells you what the builder is capable of. It tells you nothing about what your tube will do in their die.
Where It Actually Breaks Down
The “standard tooling” trap
“Standard” is the word that costs the most money. On round tube, a supplier's “standard die” usually means a 1.5D bend die of some series, 50 mm spindle bore or thereabouts, clearance set for nominal wall thickness. On 25 mm OD, 2 mm wall 1018 tubing that's fine — the radius is generous and springback behaves predictably.
Square tube is a different animal. A square tube bending machine die has to match the square section with a supporting mandrel, or the inner wall collapses at the bend point. I've seen a 40×40×2 mm square section go through twelve test pieces on a machine where the die clearance was off by what looked like a rounding error. The machine was fine. The die set was not.
So the question to ask during sourcing isn't “will your machine bend 40 square?”
The question is: “Have you run 40×40×2? With which die set? What's the die material? When was the last time you ran that size, and can I see the setup sheet?”
The answers tell you more than the datasheet ever will.
CNC vs hydraulic — the part that matters
A CNC tube bending machine earns its price through repeatability, not capability. It can hit the same bend angle on part 900 that it hit on part 1. If your volume is 50 pieces and the batch dies, CNC gives you speed but not much else on that job.
A hydraulic tube bending machine with good tooling and a careful operator can hold tolerance on a small run and cost meaningfully less up front. What you give up is the ability to run unattended and the consistent springback compensation that CNC will calculate for you.
When I compared two runs side by side last year — same 50-piece job, same tube, two different machines, only the die set changed between them — the angle variance on the CNC machine was not the deciding factor. The die clearance was. We scrapped fewer parts on the cheaper machine because the tooling on it actually matched the tube.
That's the part nobody puts in a spec comparison.
What to actually ask CNC tube bending machine suppliers
Not every supplier will have good answers. That's useful information too. Here's the list I use, in order:
- Which die series do you stock, and what's your lead time on a custom die for my section?
- Do you run a first-article submission before shipment, and who signs the report?
- What risk assessment file comes with the machine? (For EU-bound equipment, this is the ISO 12100 file and CE marking under Machinery Directive 2006/42/EC — if they can't name the standard, that's a flag.)
- What's your policy on die swaps if the first article fails?
- Can I speak to a customer running my material grade and section shape?
Spec confirmed, die matched, payment terms clear. In that order — and only that order.
I should add: item three surprises people. A machine without a proper risk assessment file can be a problem at your own insurance review, not just at the border.
When This Order Doesn't Apply
Tooling-first works best for shops with a settled product line — one or two tube families, stable volumes, tolerances that don't move. If your shop runs a new section every week, you're going to live on custom dies either way, and the machine platform might reasonably come first because you need the flexibility of a CNC bender with an open die architecture.
There's also the case where you already own a die library. If your existing tooling fits a hydraulic tube bending machine with the right spindle bore, chasing the cheapest possible machine around that constraint is rational. I've done it. It worked.
And if you're buying used — whether from a dealer or directly — the die situation is almost always murkier. Some used machines come with a full die library and full documentation. Some come with three mismatched dies and a folder of setup sheets from a product line that hasn't existed since 2019. Nobody in that category is automatically bad. Just go in expecting to verify more and get less in writing.
One honest limitation: I work on tube and pipe. If you're bending flat bar or profiled extrusions, the tooling-versus-machine balance shifts and some of the above won't transfer cleanly.
Where This Leaves You
Tube bending machine sourcing isn't a spec comparison exercise, it's a tooling match exercise with a machine attached. The buyers who get burned are usually the ones with the cleanest spreadsheets, because the spreadsheet never asked what die they were going to use.
Ask about the die first. Everything else gets easier from there.