How to Choose a Press Brake: A Fifteen-Step Path to Acceptance

1. Start from the part, not from the tonnage

Most purchasing mistakes do not come from a poor machine. They come from describing the work badly. Two machines with the same stroke and throat depth can behave completely differently on your floor. A clear description of the work is what lets a supplier recommend the right machine.

Before comparing quotations or starting research, answer seven questions: what material and thickness do you bend? what bending force does the part need? what output and part complexity do you have? what accuracy and tooling do you require? how much automation do you actually need? what is the budget, and what does the learning curve look like?

This guide follows the order of a real purchase: define the requirement, size the force, then compare machine type, control, accuracy, tooling, output, safety and cost.

2. Material grade matters more than thickness

A higher tensile strength means more force at the same thickness, and aluminium and stainless steel bend very differently. Record your typical thickness and your maximum thickness separately — they may point to different machines and different tooling. Inside radius and minimum flange also matter, because the die opening and the material determine them together.

3. Bend length sizes the machine, not the blank

A sheet 3 m wide that is bent across 1 m does not need a 3 m capacity. Size the working length from the longest bend you actually run, and record long parts separately so occasional work does not force you to oversize the whole machine.

4. Output drives automation; accuracy drives quality

Match the machine to your production profile. High-mix, low-volume work rewards fast changeover, flexible tooling and offline programming. Repeat, high-volume work rewards stable cycle time, reliable handling and automation potential. Record annual volume, average batch size, shift pattern, changeover frequency and operator availability — these decide whether quick clamping, tool changing or robotics will pay off later.

5. Same material, different geometry, different machine

Two jobs in the same material and thickness can require different machines once geometry enters. Consider box depth and return flanges (which set opening height and die profile), offset and conical bends, angle and flange-length tolerances, springback between material batches, and visible surfaces that cannot be marked. These details decide tooling, rear axes, daylight, and whether front support is needed.

6. The punch presses into the die. That is the whole idea.

A press brake forces a punch (upper tooling) into a die (lower tooling) clamped around the workpiece, bending sheet and plate. The basic cycle: bring the blank against the backgauge, close the ram so the punch forms metal around the die, then retract and remove or reposition the part.

Every press brake shares the same core parts: frame, slides and bed, punch and die tooling, backgauge, control system, drive system and safety system.

One point is more useful than any other: the drive and the control are two separate decisions. "Hydraulic" or "electric" describes where the force comes from; "NC" or "CNC" describes how the machine is controlled. A machine can be hydraulic and CNC. Keeping the two dimensions separate prevents mismatched comparisons.

7. Decide how you will bend, then size the force

Before sizing a machine, decide how you will bend. The method changes force, tooling, radius and flexibility.

Air bending — the punch presses the sheet into a V-die without touching the bottom. One die can make many angles; it needs the least force and offers the most flexibility, but the inside radius depends on the die opening and the material.

Bottoming — the sheet is pressed fully into the die, so the die sets the angle. It needs more force and a dedicated die per angle, but it improves angle consistency.

Coining — higher force imprints the punch geometry for the tightest, most repeatable radius. It needs more tonnage and dedicated tooling, so it is never the default route to accuracy.

Your tonnage calculation must be based on the method you will actually use.

• Air bending: least force, one die many angles, radius follows the opening
• Bottoming: the die sets the angle, better consistency
• Coining: tightest radius, highest tonnage and tooling cost

8. Bending force depends on five variables, not just thickness

Bending force depends on tensile strength, thickness, bend length, die opening and method — not just thickness. A widely used air-bending estimate for mild steel is:

P = 650 × S² × L / V

Where (metric): P = bending force (kN); S = thickness (mm); L = bending length (m); V = die opening (V width) in mm. A coefficient of 575 corresponds to low-carbon steel (~450 MPa); adjust for other materials.

Worked example — 3 mm mild steel, 2.5 m long, 20 mm V-die: P = 575 × 3² × 2.5 / 20 ≈ 646 kN ≈ 66 tonnes.

Then apply the important corrections: material factor — stainless steel is roughly 1.5× mild steel, aluminium roughly 0.5×; use your actual material strength. Method — bottoming needs several times the force of air bending; coining more. Load limits — check the rated load per machine and tool section, and the limits for concentrated and offset loads. A 100-tonne machine is not designed to run to full capacity on a short bend in one pass. Capacity headroom — leave sensible margin for material variation and future parts.

Never plan to exceed the rated limit. Check the final figure against the manufacturer data.

9. Size working length from your typical part, not your largest

Base the working length on the longest bend you run regularly, plus operating margin.

Then go beyond the nominal figure: check the distance between the uprights (which limits how wide a part can enter the machine and matters when flanges exceed the housing), and weigh whether a longer machine adds real capability or only cost. When long parts are rare, a tandem or an outside partner can be cheaper than a larger main machine.

10. Beam, daylight, throat depth and opening height

These specifications are often left out of basic guides, yet they decide whether tall parts and deep boxes are possible:

Stroke — ram travel, which affects deep boxes and output.

Daylight / open height — space to load tall tooling and remove the finished box.

Throat depth — how deep the flange can reach into the machine.

Match each email part against your tallest tooling, deepest box and largest return flange, and ask the supplier to prove the difficult parts by simulation or sample bends.

11. Let the work define the machine type

Comparing machine types only makes sense once you know the work and the required capacity. Remember: the drive type by itself does not determine the control level, the achievable accuracy or the maximum thickness.

12. Hydraulic and electro-hydraulic brakes

Hydraulic machines drive the ram with cylinders and are the general-purpose mainstay: the widest tolerance range, good fit for heavy work, and a mature service and tooling ecosystem. Modern versions are usually electro-hydraulic (servo-valve controlled), giving closed-loop synchronization and strong repeatability. Trade-offs: hydraulic maintenance, heat, and energy use during long low-visibility runs. Best for: general and heavy fabrication, and broad capability from one machine.

13. Servo-electric and hybrid brakes

Servo-electric machines drive the ram with servo motors and ball screws, with no hydraulic system. They are fast, energy-efficient, quiet, clean and precise, with low maintenance, but usually have a smaller force range — they suit fine precision work. "Electric" describes the drive, not the control. Hybrid machines combine servo drive with hydraulics for higher force at better efficiency and response. Best for: precise small and medium parts, and fast, clean, energy-conscious production.

14. Tandem and automated press brake systems

These problems solve for specific situations, not every buyer. Tandem systems synchronize two or more machines to bend very long or wide parts (cab panels, wind-tower sections). Robotic cells automate loading, bending and unloading for stable, repeatable part families — high output, but added programming, guarding and maintenance cost. Automation level and machine class are separate decisions.

15. Press brake types compared

No machine is best at everything. The table below maps the drive arrangement to the kind of output you usually want.

16. Buy the control capability your parts need

Buy the control capability your parts require — not the highest axis count on the brochure.

17. Manual, NC or CNC control

Manual / basic — simple, occasional, low-precision work and prototyping.

NC — automates basic angles and backgauge positioning through stored jobs; well suited to standard repeat parts.

CNC — multi-axis, multi-step programs, graphical (usually 3D) setup, simulation and correction; the choice for complex, high-precision or high-mix work.

Compare control systems on interface availability and language, program storage, CAD/CAM import, offline programming, and one often overlooked item — software licensing and update cost. Modern 3D control also helps less experienced operators, although trained people remain essential.

• Manual: lowest cost, lowest accuracy, good for prototypes
• NC: stored jobs, standard repeat parts
• CNC: multi-axis, simulation, complex mixed work
• Check: program storage, CAD/CAM import, offline programming, licensing

18. What the main press brake axes do

Axis names only matter once you know which part problems they solve:

Y (Y1/Y2) — ram movement and left/right synchronization.

X (X1/X2) — backgauge depth (flange length); conical parts have independent fingers.

R — finger front and back stops.

Z1/Z2 — transverse fingers supporting the bed.

Crowning axis — deflection compensation along the bed.

19. Do not compare accuracy figures before deciding which number you mean

Do not compare accuracy numbers before establishing which figure is being quoted. Positioning accuracy, repeatability and finished-part bend angle accuracy are not the same thing. A machine can hold excellent axis figures and still produce inconsistent parts when material, tooling, springback, temperature or frame deflection is uncontrolled. Ask the supplier to state the test conditions, and whether the figure describes the machine or the measured part.

20. Crowning and deflection compensation

Under load the bed and ram deflect slightly, opening the angle in the middle of a long part. Crowning compensation keeps the angle consistent; CNC crowning adjusts for each job automatically. The longer the bend and the tighter the tolerance, the more this matters. Verify by measuring several points across the length.

21. Angle measurement and material compensation

First let the machine measure the actual bend angle during forming and correct on the second hit, or sense the thickness and adjust automatically. This is valuable when material varies by batch, such as cutting off the first part of a sheet.

This does not replace correct tooling, process control and trained operators.

22. Tooling is part-specific, so it belongs in your specification and budget

Tooling is part-specific and belongs in your specification and budget — a new machine almost always needs new or additional tooling. Consider tool standards and compatibility, punch profiles and die openings, section blades, tool heights and load ratings, and whether existing tooling can be reused.

23. Choose the V opening and tooling profile

The V opening sets force, inside radius and minimum flange at the same time: a larger opening lowers force and increases the radius; a smaller opening does the opposite and raises the minimum flange. The punch profile must clear the box, the return flange and earlier bends. Avoid treating an "rule of thumb" opening as universal. Choose the V opening deliberately.

24. Manual, hydraulic and automatic tool clamping

Clamping affects setup time, alignment, safety and automation readiness. Manual bars are cheapest; fast-changing hydraulic clamps switch tools almost at a fingertip, removing most installation time — provided the change frequency, small batches, tool standardization and labour conditions suit it. For shops with frequent changeover, this can lift brake utilization from 40% to 60% or higher.

25. Faster cycles do not mean more parts

On a real shop floor, the actual bending time often takes only 30% to 40% of the operator's day — the rest goes to setup, handling, correction and shift change. Compare the whole cycle, not the main-axis speed.

26. Measure the complete bending cycle

Run the same representative part on each candidate and record programming/offline time, tool loading and alignment, material handling, first-piece correction and inspection, actual bending time, part removal and repositioning, and scrap/rework. The figure that matters is good parts per hour under the operator assumptions you actually face — not cycles per minute on an empty machine.

27. When automation pays

Match each option to a real constraint: offline programming where changeover is frequent; front support or sheet followers for large or soft blanks; automatic tool changing where changeover frequency is high; robots for stable, repeatable part families; MES or network integration when traceability matters.

Justify each one with labour availability, utilization, batch size, part stability and payback — not with theoretical speed.

28. Treat safety as a core requirement, not a feature list

Start with a documented risk assessment for your application. Modern machines provide operating-point guards (light curtains, laser guards at the tool tip) that stop the ram when a hand breaks the beam, plus emergency stop, interlocks, safe operating speed and lockout/tagout. Confirm each of these for the machine you are considering.

The critical point is to understand what the system does not cover: secondary pinch points between flanges, or between part and machine, are usually outside the guard envelope. No system makes a brake one hundred percent safe, so training and a documented risk assessment remain essential.

Requirements depend on your jurisdiction and the standards currently in force — verify against the official standard and the machine documentation.

• Confirm safeguards: which protection systems are included and which are optional?
• How do setting, tool change and special modes behave?
• Which hazards fall inside the guard envelope and which do not?
• What training, documentation and local compliance support is provided?
• If a robot or special tooling is added, how is safety re-verified?

29. Press brake prices span a wide range depending on size, control, axes, tooling and automation

Press brake prices span a wide range depending on size, control, axes, tooling and automation. Treat the following as rough market bands, not quotations:

30. What drives the price

The biggest cost drivers are tonnage and bending length, followed by CNC control, axes, tooling, crowning and angle measurement systems, and any automation (tool changers, robots, plate handling).

Two machines of the same tonnage can differ significantly once control, axes and tooling are specified — which is exactly why comparing price alone is misleading.

• Tier one: tonnage, bending length
• Tier two: control, axes, compensation
• Tier three: tooling and automation
• Equal price does not mean equal cost

31. Compute the real five-year cost, not the purchase price

Over the defined period, compare total machine cost under the same assumptions: annualized capital or financing cost; freight, rigging, installation and disposal; tooling and clamping; software, licensing and updates; training and labour; energy and consumables (oil, filters); preventive maintenance, repairs and spares; expected downtime, scrap and rework; and residual value.

Then compare the cost per good part, or the total annual cost — not the purchase price.

32. At this stage you compare production solutions, not just machines

At this stage you compare production solutions, not just machines. A consistent process is your best defence against misleading quotations.

33. What an RFQ should contain

Send every supplier the same pack: representative drawings and material specification, required tonnage, bending length and geometry (or ask them to calculate and state it), tolerance and production requirements; proposed and existing tooling; controller, axes, compensation, safety and software requirements; installation, training, warranty and service scope; and delivery, commissioning, FAT/SAT and acceptance criteria.

Review the exclusion clauses as carefully as anything else.

• Part data: drawings, material certificate, tonnage, bending length and geometry
• Tooling: proposed and existing tooling compatibility
• Control and axes: controller, axes, compensation, safety, software
• Delivery and acceptance: commissioning, FAT/SAT, criteria and timing

34. Service, warranty and spares

Downtime is one of the most expensive hidden costs in manufacturing, so evaluate support at purchase time: local and remote service coverage, guaranteed or typical response time, spare-part locations and delivery time, and warranty duration (including travel, labour and consumables).

Confirm training, software support and controller life cycle, and ask for customer references running similar machines and similar work.

35. Sample bends, FAT and final acceptance

Convert claims into evidence before final payment: check representative easy, difficult and long parts — not only the samples the supplier chooses. Record installation time, tool-change time, cycle time, first-piece correction and good output; measure bend angles, flanges and consistency; verify control, axes, compensation, safety, software transfer and tooling; and state in writing the FAT/SAT responsibilities, tolerances and remedies if the machine fails.

36. The most expensive mistakes rarely come from the machine — they come from how it was chosen

The most expensive mistakes rarely come from the machine itself — they come from how the machine was chosen. These are the patterns we see most often.

• Sizing tonnage from thickness alone
• Letting one rare part size the whole machine
• Confusing drive type with control level
• Comparing cycle speed instead of total cycle time
• Comparing spec accuracy instead of finished parts
• Deferring tooling, safety, installation and training
• Comparing purchase price instead of installed and total cost
• Buying without sample testing and written acceptance criteria

37. Sizing tonnage from thickness alone

A customer once sized a machine from thickness and ignored the material tensile strength and die opening. In air bending, stainless steel needs roughly 1.5× the force of mild steel, so the "big enough" machine ran to its limit and accuracy suffered. Size from material strength, bending length, die opening and method — not thickness alone.

38. Comparing cycle speed instead of total cycle time

The fastest ram rarely wins; setup, handling and correction dominate the day.

39. Comparing spec accuracy instead of finished parts

Positioning figures can look excellent and still fail to deliver consistent bend angles.

40. Mixed shop floors and high-mix production

Prioritize flexibility: graphical CNC with offline programming, flexible tooling with a suitable opening height, 4-6 axes depending on geometry, quick clamping (possibly with automatic tool changing), and broad tooling choice.

41. Precision small and medium parts

Prioritize consistency: a hydraulic, hybrid or servo-electric machine within your force range, repeatable positioning, stable tooling, angle correction where needed, and programs that are easy to reuse. Judge by measured finished-part capability, not axis specifications.

42. Long or heavy plate bending

Prioritize strength and rigidity: tonnage per length of tolerance, a rigid frame that holds across the working length, sufficient working length and side-frame clearance (or a tandem arrangement), heavy tooling, a support system, and careful foundation and safety planning.

43. Repeat or automated production

Prioritize stable output: repeatable part families, offline programming, automated handling, a connected tooling strategy, appropriate guarding, and a payback based on good parts and labour — not theoretical speed.

44. How much tonnage do I need?

Size = tonnage + bending length + material. Calculate the force for your most common material at the longest regular width (using your intended die opening and method), then set the bed length to the longest regular width plus machining allowance. Confirm both stay within the machine and tooling load limits.

45. How much tonnage for a 1/4 inch (6 mm) part?

For 6 mm material, apply a factor of about 1.5 relative to mild steel and repeat the calculation for your die opening and bending length. Run the estimate in our bending force calculator, then have GWMT review the actual part.

46. How many backgauge axes do I need?

Answer with the parts you actually run: one X covers parallel flanges; two X help on tapers or widely varying flange lengths; Z axes support transverse fingers on long parts; R gives independent finger positioning. Let the parts, not the axis count, drive the decision.

47. Can I reuse existing press brake tooling?

Confirm the tool standard, upper tool height, die opening type, clamping interface and load rating. Two punch heights of 88 mm and 96 mm are not interchangeable; the system path, return-flange height and tool height must match.

48. Conclusion: follow this order, not the price tag

Choosing a press brake is not about maximum or minimum price. It is about matching the machine, control, tooling and support to the parts you actually make. Work in this order: define the parts and process, size the tonnage and geometry, choose the drive and control, specify axes, accuracy and tooling, evaluate real output, plan the cost, and verify the supplier through the RFQ and sample tests.

The right brake for your shop is the one that reliably produces good parts at a reasonable total cost — and you can prove that before you sign.