Cause table
| Symptom | Cause | How to check | What to do |
|---|---|---|---|
| Angle is always short by a fixed amount | Normal springback, not yet compensated | Bend 5 parts, average them | Enter a fixed compensation in the controller |
| Springback varies sheet to sheet | Different lots, different yield strength | Check mill certs; measure thickness | Separate lots, re-calibrate per lot |
| Stainless springs back far more than carbon | Normal — stainless has higher yield strength | Compare both materials at identical settings | Keep a separate parameter set for stainless |
| Springback increases with a wider V | Normal — a wider V gives a larger radius | Test on two V widths | Select the V by the 8×t rule, then leave it alone |
| Fully bottomed and the angle still opens | Working near the machine's tonnage limit | Calculate the force actually required | Re-check tonnage against thickness and bend length |
Why stainless is always harder
Springback scales with the material's yield strength. 304 stainless has a higher yield strength than ordinary carbon steel, so at the same thickness and the same V it opens up more. Aluminium goes the other way.
This means one parameter set for all materials is the most common real cause behind "the press brake is inaccurate." It is not the machine — it is the parameter table.
The V opening changes everything
The common rule is a V opening of roughly 8 times sheet thickness. A wider V needs less force but produces a larger inside radius and more springback; a narrower V does the opposite and marks the sheet more easily.
If you change the V without changing the compensation value, the angle will be wrong. This is the number-one reason a part passes today and fails next week while "nobody changed anything."
The most commonly misjudged point
Many people treat springback as a defect to be fixed. It is not a defect — it only needs to be consistent.
A machine that is always 2° short is an easy machine: enter 2° and you are done. A machine that is 1° short sometimes and 3° short other times is the real problem — and the cause of that inconsistency is almost always the material, not the machine.
Controllers with angle measurement and automatic compensation shorten this process considerably. Even then, mixed lots will still wreck the result.
Calibrating for a new material
- Cut 5 samples from the same lot, same width
- Bend all 5 at identical settings, measure each angle
- Average them, compute the difference from target
- Enter the compensation, bend 3 more to confirm
- Record it: material grade, measured thickness, V opening, compensation value
Step 5 is the one most often skipped and the one that saves the most time later.