The position of an axis can be displayed to many decimal places with a CNC screen.
This doesn’t necessarily mean that the end result is precise to the same number of decimal points.
If metal has been introduced, the thickness of the material, springback, tooling, contact with the backgauge and machine deflection, even the way the operator measures the part, will come into play.
This is why it is best to evaluate a precision metal bending machine starting from the finished product and working backwards.
APEX’s Press Brake platform is all about rigid construction, synchronized hydraulic control and CNC positioning for repeatable sheet-metal bending. However, in the case of a precision shop, machine features are only significant when they have an impact on the drawing.
Begin by asking what do they mean by “Accurate” on their part?
A part may be at the correct angle and yet not pass inspection.
Let’s take a look at an enclosure panel that is bent. The angle of the bend is 90° which may be acceptable and the flange is 0.8 mm too short. A cover can be too long or short at either end but straight down the middle. When the part is multi-bend, the hole on the end might not be in the correct position because of the small error it has taken on along the way.
Precision must consequently be broken down into a number of outputs:
- bend angle;
- flange dimension;
- bend-line position;
- straightness;
- left-to-right consistency;
- repeatability between parts.
The key distinction that APEX’s Press Brake Bending Accuracy guide makes is that: Angle, Dimensions, Straightness, Parallelism and Repeatability must be measured individually.
Before comparing machines, determine which of these is really the one that will allow your customer to accept or reject the part.
| Precision Item | What to Measure | Suggested Check |
| Bend Angle | Actual vs. drawing angle | Digital angle gauge |
| Flange Dimension | Edge to bend position | Caliper / height gauge |
| Straightness | Bend-line deviation | Straightedge / laser |
| Left-Right Consistency | Same feature at both ends | Compare identical points |
| Bend Position | Bend to hole / reference edge | Caliper / fixture |
| Repeatability | Variation between parts | Multi-sample measurement |
The CNC Position is NOT the Finished-Part-Measurement.
The Y-axis is an indication of the location of the ram’s movement.
The X-axis indicates the location the backgauge was directed to move.
Neither number alone is proof that the finished component is in accordance with the drawing.
Sheet thickness, tensile strength, rolling direction, V-die opening, punch radius and springback are also factors that affect the actual bending results. APEX thus suggests that recorded CNC data should be compared to the actual part measurements, and the positions of the axis should not be considered as a complete quality result.
Here’s a great precision workflow:
Program, bend, measure, correct and record.
It’s the final step that is important.
When a correction is made, retain with material batch, tooling information and program… do not ask the next person to find it!
Backgauge accuracy and Bend-angle accuracy are two different types of accuracy.
The two problems tend to be confused.
If the flange is incorrect, but the angle is right, then endless changes of ram depth will probably not address the underlying issue. Always check for sheet referencing and positioning of the backgauge first.
If flange dimension is correct, but the angle is open, consider ram penetration, springback, tooling and material behavior.
This difference is more relevant for the more complex parts, where multiple bends are required with varying reference edges.
The positioning of the X, R, Z1 and Z2 backgauge axes is controlled by a servo axis to achieve more complex positioning in a High Accuracy CNC Press Brake, with CNC crowning and Y-axis control dealing with other aspects of the bending process.
The more geometry you need, the more axes you need, but no more.
Precision is not the most extensive option list – it’s the proper variable controlled.
Parts that take longer to do are more accurate tests than Easy Coupons.
There’s a lot that can be concealed in a short test strip.
However, frame deflection, crowning setup, left-right synchronization and condition of tooling are much more evident in long workpieces. APEX advises to check long bends at the left side, in the center and on the right side instead of at a convenient spot.
Crowning is a first variable to check if the center is different than either end. If the two ends differ from one another, Y1/Y2 synchronization or tooling variation and/or off center loading may be a consideration.
Independent cylinder axes and programmable V Axis crowning and servo backgauges on appropriate configurations, such as APEX’s Hydraulic Press Brake Machine.
For the buyers, who are to make long precision panels, take a long component to the acceptance test.
Never allow a 300 mm sample to be cause for a 3 m production problem.
| Measurement Result | First Area to Investigate |
| Angle correct, flange wrong | Backgauge / sheet reference |
| Flange correct, angle wrong | Ram depth / springback / tooling |
| Center angle differs from ends | Crowning |
| Left and right angles differ | Y1/Y2 / tooling / loading |
| Bend line is not straight | Deflection / tooling condition |
| Parts vary within one batch | Material / setup / positioning consistency |
The one good part doesn’t say much about repeatability.
It is always tempting to be tweaking, tweaking, tweaking till the one perfect component arises.
Everyone then takes a picture of it.
There’s a different test for precision production.
Repeat the bending process with the same batch of materials, tooling and CNC program used to bend several samples. Take measurements of angle and flange location on all parts. According to APEX’s accuracy guidance, it is advisable to take 5–10 samples when checking the accuracy range and repeatability instead of one.
This distinguishes two concepts which are easy to confuse:
Accuracy: How close is the student’s response to the desired outcome?
Repeatability – How close will repeat results be to each other?
A machine may repeat but be off target. That can many times be fixed.
The machine with results that vary is a more difficult production problem.
Closed-Loop Control is Helped But Measurement Still Closes the Real Loop
Today, CNC Press Brake Machines for Sheet Metal Bending are equipped with linear-scale feedback and electro-hydraulic control to track the actual position of the machine and modify the movement of the machine during working.
This enhances the machine control.
The second loop in the factory is still to be completed: verifications of the workpiece.
A calibrated angle gauge, caliper, height gauge or appropriate inspection device may be used as specified in the drawing. Use the same place to take measurements. Don’t test one sample on the left one day, and one in the middle the next.
It is possible to have apparent variation as a result of poor inspection practice even when no change has been made in the bending process.
As much as a repeatable machine is important in a precision process, so is a repeatable measurement method.
Purchase Precision Around the Drawing today!
Send a number of actual drawings before enquiring a supplier for the “most accurate” machine.
Draw a line beside those dimensions that are of real importance.
Look for the longest bend, smallest tolerance, most challenging flange and combinations of multiple bends in a single assembly dimension.
Then identify the machine function requirements that are satisfied.
For some shops multi-axis positioning is required. For others, it’s the crowning or the measurement of the angle or process records that make the difference. Most require a combination.
The best precision metal bending machine is not the machine that has the best accuracy advert on a brochure.
It’s the one that can deliver your critical dimensions time after time – and provide your quality team a viable means to demonstrate it.
FAQ
Q1. So, what factors affect the accuracy of a precision metal bending machine?
There are more factors than machine positioning. Other factors that affect the final product are material variation, tooling, springback, backgauge position and crowning and measurement method.
Q2. But is bend angle accuracy equal to part accuracy?
No. A part can have the proper angle but wrong flange size or possibly the position/bend or straightness can be wrong. Precision inspection shall be carried out according to the drawing, and not be based on a single measurement.
Q3. What is the correct way of testing the repeatability of the press brake?
Take a number of samples of the same material and use the same tooling and program to bend and then measure the same location on each sample. Look at the differences between results and not just the best sample.
Q4. What are the values of long parts in precision testing?
Long bends will make it easier to spot the deflection, crowning and left/right variation. Measuring 3 points (left, center and right) gives more information than measuring 1 point.
Q5. Is the precision of CNC bending always highest with more CNC axes?
Not automatically. Any extra axes are useful if the geometry of the work piece requires more position control, but can’t replace proper tooling, material control and inspection.