Press Brake Outer Mold Line (OML): Setup, Tooling, and Inspection Guide
When a work instruction directs an operator to locate a flange from the “outer mold line,” the natural reaction may…
You’ve got a piece of 16-gauge stainless steel in front of you. You want a perfect 90-degree bend, but every time the ram comes back up, the metal relaxes to 92 degrees. Springback. It’s the first enemy every new press brake operator meets. So you do what seems logical: you drive the punch deeper into the V-die and crush the sheet metal until it has no choice but to hold its shape. The part comes out perfect. You feel like a genius. But while you’re admiring that flawless angle, you don’t hear the hydraulic cylinders groaning or feel the microscopic stress fractures spreading through your tooling. You didn’t just bend metal. You used a 10-pound sledgehammer to swat a fly on a glass table.
Look at a standard 90-degree punch positioned above a matching 90-degree V-die. To an untrained eye, the tooling looks exactly like a mold. The assumption is purely geometric: if you have a 90-degree cavity and push a 90-degree wedge all the way to the bottom, the metal trapped between them must become exactly 90 degrees. You aren’t floating the metal in the air and guessing when to stop the ram. You’re stamping it into submission.
It feels incredibly safe. It removes the frustrating guesswork caused by material variation, grain direction, and yield strength. If the metal tries to fight back, you simply add more tonnage until it stops. The machine does the thinking, the tooling sets the limit, and the operator just steps on the pedal. Why wouldn’t you want a process that guarantees the angle every single time?
Bend a piece of steel, and its outer fibers stretch in tension while the inner fibers crush in compression. When the ram releases, those elastic fibers try to return to their original state. To eliminate that memory, you have to do more than fold the metal. You have to permanently deform the neutral axis—the invisible line through the center of the sheet that is neither stretching nor compressing.
You are literally coining the material, much like the U.S. Mint stamps a penny. You force the punch tip to bite deeply into the steel, flowing the material outward until it completely fills the bottom of the die. You aren’t just bending anymore; you’re restructuring the steel at a molecular level. That requires driving the punch and die into near-full contact, trapping the material with nowhere to go. What does that kind of violent molecular restructuring actually cost?
A sheet of mild steel arrives from the mill measuring just two-thousandths of an inch thicker than nominal. In a modern air bending setup, the machine barely notices. But in coining, because the punch and die mate with virtually zero clearance, that extra two-thousandths of an inch of steel has nowhere to go. The tonnage spikes exponentially.
The closer you get to true coining, the less forgiving the setup becomes. Tiny errors in material thickness, tooling angle, or bed deflection stop being minor corrections. They instantly become scrap parts, shattered tooling, or a permanently bowed ram. You traded the nuisance of springback for a process with absolutely zero margin for error. If a method turns every minor material inconsistency into a machine-killing event, is it really a shortcut at all?
Look at a standard Amada tonnage chart for mild steel. To bend one foot of material with air bending, the machine needs roughly 20 to 30 tons of force. The metal rests only on the two upper shoulders of the V-die. As the punch descends, it pushes the center of the sheet downward, creating a simple three-point contact system. The metal is literally suspended in the open space of the die cavity.
Because the material never touches the bottom of the V-die, the tooling angle does not determine the final part angle. Punch depth controls everything. If you want a sharper angle, you drive the ram a few thousandths of an inch deeper. This is exactly why modern CNC brakes rely on air bending as the universal standard. Control systems, angle-measurement lasers, and adaptive crowning allow the machine to adjust punch depth on the fly. You achieve precise angles through calculated geometry, letting the machine work smarter rather than harder.
Move to bottoming, and the force requirement immediately doubles to 30 to 60 tons per foot. The punch drives the material down until the outside of the bend makes full contact with the angled faces of the V-die. The sheet metal meets the steel mold. Beginners often confuse this with coining, but the physical mechanics are fundamentally different.
In bottoming, the punch tip radius closely matches the natural inside radius of the folded metal. You force the sheet to wrap tightly against the die’s geometry, but you do not change the sheet’s actual thickness. The neutral axis—the core of the metal—survives the process intact. Because that axis survives, its elastic memory survives with it. You have spent twice the tonnage of air bending, added unnecessary wear to your tooling, and still must overbend the part to compensate for springback.
Coining demands 80 to 120 tons per foot, and often much more depending on the alloy and die geometry. The punch does not merely push the metal against the die walls. It bites into it. To coin properly, you use a punch with a sharper angle and a smaller tip radius than the die cavity. When the ram reaches the bottom of its stroke, that sharp tip physically penetrates the inside radius of the sheet metal.
The material thins at the bend line and is forced to flow outward, completely filling the bottom of the die. This is severe plastic deformation. You are permanently crushing the neutral axis so the steel forgets it was ever flat. At this point, you are no longer folding sheet metal. You have turned your press brake into a stamping press.
That shift from bending to stamping rewrites the survival equation for your equipment. At 120 tons per foot, the bottleneck is rarely the machine’s total hydraulic capacity. The real limit is the physical durability of your tooling. Coining requires very small inside radii and sharp punch noses to penetrate the metal, concentrating all that extreme force on a microscopic sliver of hardened steel.
Tooling chips. Punch tips flatten. Dies split down the middle. When the sheet metal can no longer compress, the kinetic energy generated by the hydraulic cylinders does not simply disappear. It acts like a shockwave, traveling upward through the punch body into the ram and downward through the V-die into the machine bed. In effect, you have sent a hundred tons of dead-head pressure directly into the structural skeleton of your machine.
Look at a standard tonnage chart for 10-gauge mild steel. It tells you that air bending requires roughly 15 tons of force per foot. When an operator struggles with springback on that 10-gauge part, the temptation is to switch methods, drive the ram deeper, and stamp the angle permanently into place. But the moment you decide to coin that same piece of steel, the chart you trust becomes a trap.
Press brake force requirements rise exponentially, not linearly. To force the punch tip to penetrate the neutral axis of the sheet, you do not merely add a few tons of pressure. You multiply the air-bend tonnage by five, and sometimes by as much as ten, depending on the alloy and the die opening. Suddenly, a routine 15-ton bend demands 100 to 150 tons per foot. It is like using a 10-pound sledgehammer to swat a fly on a glass table: the fly is springback, the sledgehammer is coining tonnage, and the glass table is your $100,000 machine. How much abuse can the equipment actually take before the glass shatters?
A 200-ton press brake sounds invincible until you read the fine print on the manufacturer’s spec sheet. Machine makers rate that 200 tons across the entire length of the bed, usually 10 or 12 feet. In practice, that means the machine is designed to handle roughly 16 to 20 tons per foot safely, which is why shops considering higher-capacity bending should compare the load distribution and CNC capabilities of equipment such as ADH Machine Tool’s large press brake before assuming more nominal tonnage solves the problem.
For readers comparing machine capacity against real bending loads, ADH Machine Tool’s CNC press brake and sheet metal equipment materials can be a useful next reference; you can download the technical documents to review concrete product information before pushing a machine anywhere near its localized limits.
Coining rarely occurs across the full bed. Operators usually coin small, stubborn brackets right in the middle of the machine. If you apply 120 tons of coining force to a 12-inch part dead center, you are technically within the global 200-ton limit. However, you have vastly exceeded the machine’s localized centerline load limit. The force is no longer distributed; it becomes a concentrated shockwave. Driving that much pressure into one spot forces the heavy steel ram to bow upward in the middle, creating a permanent deformation known as ram upset. Once the ram deflects, you will never bend a perfectly straight, accurate part on that machine again. So what happens to the delicate components caught between the ram and the bed when that pressure peaks?
Modern CNC punches are precision-ground to tolerances of a ten-thousandth of an inch and hardened to withstand years of standard friction. They are not designed to act as anvils. When you coin, the sharp tip of the punch must physically bite into the sheet metal, concentrating all that multiplied tonnage in a contact patch less than a millimeter wide.
Standard air-bending V-dies are just as vulnerable. Because coining requires a narrower V-opening to force the metal to flow, the outward horizontal pressure on the die shoulders is immense. The die does not simply wear out; it splits straight down the middle, sending hardened steel shrapnel across the shop floor. You end up replacing $2,000 worth of shattered precision tooling just to avoid programming a proper springback calculation. If the tooling explodes and the ram permanently bows, how does this extreme brute force affect the actual skeleton of the press brake?
Step back and watch the side frames of a press brake during a heavy coining operation. On older mechanical brakes, the danger is immediate and catastrophic. Because these machines rely on a massive flywheel and a heavy clutch, the ram must complete its cycle once engaged. If you miscalculate the coining tonnage, the machine cannot stop; it can literally lock up at the bottom of the stroke, trapping thousands of pounds of kinetic energy and forcing you to cut the tooling out with a torch just to free the ram.
Modern hydraulic brakes are smarter, using relief valves to dump pressure before the frame snaps. But that safety net can create a dangerous false sense of security. Continually maxing out the hydraulic system to coin parts microscopically stretches the side frames over time, fatigues the massive steel welds, and blows out internal cylinder seals. You are slowly trading your machine’s structural integrity for a slightly sharper inside radius. If coining is this mathematically destructive to every part of the press brake, why do some veterans still insist on doing it?
If coining shatters dies and bows rams, you might wonder why any operator in their right mind would still do it. Here is the absolute rule: coining is strictly a last-resort rescue operation, reserved exclusively for paper-thin materials where the massive tonnage needed to crush the metal still remains safely below the machine’s breaking point. We do not do it because we like brute force. We do it because, sometimes, the blueprint leaves us no choice. Imagine holding a piece of titanium so thin it feels like heavy-duty aluminum foil. You are making a bracket for a satellite. The tolerance is plus or minus a quarter of a degree. Air bending that kind of material is like trying to fold a wet tissue with chopsticks. The material lacks the internal structure to hold a consistent radius in the air. You have to stamp it. Coining survives in modern shops not as a general-purpose tool, but as a highly specialized rescue operation for parts that refuse to behave. But what exactly makes these specific industries the lone holdouts?
Medical implants and aerospace chassis share a frustrating trait: they require absolute, unforgiving repeatability in materials that resist being bent. When you form a 0.015-inch stainless steel surgical tray, even a half-degree of springback variation means the tray will not lock into the sterilizer rack. Air bending depends on perfectly uniform material thickness and yield strength to predict springback. Paper-thin sheet metal is never perfectly uniform.
The only way to guarantee the angle is to physically crush the material’s structural memory. You bottom out the punch, bite into the neutral axis, and force the metal to take the exact shape of the die. It works. You get a perfect, dead-flat angle with zero springback. But never forget the constraint that makes this safe: this extreme measure is justified only when the material is so exceptionally thin that multiplying the bending force will not destroy a $100,000 machine. At what point does that math turn against you?
Physics sets a hard boundary on this technique.
If you try to coin 20-gauge steel, the required tonnage spikes, but your machine can probably absorb it. Try the same move on a piece of quarter-inch plate, and the tonnage multiplier scales with the thickness. Multiply the force needed to bend quarter-inch steel by ten, and you will immediately max out the hydraulic system, stall the ram, or blow apart your V-die like a hand grenade. The metal simply has too much volume. It cannot compress enough to let the punch tip penetrate the neutral axis without catastrophic resistance. At that point, you hit a hard limit: the material will crack, split, or destroy the tooling long before it yields to the punch. The rule is unchanged: coining is strictly forbidden unless the gauge is so light that multiplying its bending force by ten still leaves your ram and dies completely out of danger. If the material is thin enough to survive the tonnage, what else can force an operator into this corner?
Sometimes the material thickness is acceptable, but the shape of the part itself leaves you no room to maneuver. Consider a complex electronic enclosure with four tight bends placed close together. Every time you bend sheet metal, the radius stretches slightly. In air bending, that stretch can vary by a few thousandths of an inch depending on the steel’s grain direction. A few thousandths of an inch means little on a single bracket. On a box with multiple consecutive bends, however, those tiny variations accumulate. By the time you reach the fourth bend, the mounting holes no longer line up.
Coining eliminates this tolerance stack-up. By stamping a sharp, permanent inside radius, you lock the stretch into a predictable, repeatable dimension. You sacrifice the tooling to save the part. It is a desperate trade-off, used only when the geometry demands absolute precision across multiple planes. But do not let complex geometry tempt you into breaking the rule: even with impossible tolerance stack-ups, coining is permitted only if the material is thin enough that the required tonnage will not crack your ram. If you are not building a titanium satellite chassis or a paper-thin surgical tray, you have no business coining the other 99 percent of the steel sitting on your pallet.
For the remaining 99 percent of everyday fabrication, the only acceptable method is air bending. For decades, operators treated coining as the ultimate trick for defeating springback. It is like using a 10-pound sledgehammer to swat a fly on a glass table. The fly is your springback, the sledgehammer is the massive coining tonnage, and the glass table is your $100,000 press brake. You might kill the fly, but you shatter the table in the process. Today, we have better ways to deal with the fly.
Look closely at the title block of your blueprint. Beginners often see a 90-degree bend callout and assume it means exactly 90.000 degrees. It rarely does. Standard commercial fabrication tolerances usually allow plus or minus one degree, sometimes more. If the part is a bracket for a tractor or a standard electrical enclosure, a fraction of a degree of springback will not scrap the job. The final assembly process—welding, bolting, or riveting—will pull that slight variation flush.
Coining a part that requires only standard tolerances is pure ego. You are risking premature wear on your dies and ram just to hit a number on a protractor that the customer did not ask for and will not pay for. Before you even touch the controller, let the print dictate the method. If the tolerance is standard, the method is air bending.
In the past, coining was a safeguard against inconsistent machines. You stamped the metal because you could not trust the brake to stop in exactly the same position twice, and you could not prevent the bed from bowing in the middle. That justification disappeared with modern CNC controls and precision-ground tooling.
Today’s controllers actively compensate for springback, adjusting ram depth on the fly based on material thickness and yield strength. Automatic crowning systems map the bed’s exact deflection under load and counteract it, ensuring the angle at the center of a ten-foot sheet matches the edges precisely. We no longer need to crush the metal to lock in an angle because we can predict and control the air space inside the V-die to within thousandths of an inch. The bottleneck has shifted from brute force to setup discipline. With precision-ground punches and a CNC controller doing the calculations, air bending can produce dead-on angles without ever bottoming out, which is why a modern CNC press brake from ADH Machine Tool fits naturally into this workflow: its CNC-based bending focus supports the move from tonnage-heavy coining to controlled, repeatable forming.
For shops pushing that same precision across longer parts or higher-throughput work, a CNC-based tandem setup can extend the logic of controlled air bending rather than returning to brute-force coining. ADH Machine Tool’s tandem press brake fits naturally into that workflow, giving operators a more scalable way to manage long bends, repeatability, and production efficiency.
Before you commit to a bending method, do the mental math on your machine’s limits. Air bending requires only a fraction of the force—typically around 10 to 15 tons per foot for standard mild steel. Coining that same foot of steel can demand more than 100 tons. If you have a 120-ton brake, coining a single two-foot bracket can max out your entire hydraulic system.
If those limits are close on your own floor, it may be worth discussing the application before committing tooling or tonnage. ADH Machine Tool’s R&D work across press brakes, automation, and intelligent equipment makes a practical next step for matching bend method, material thickness, and production requirements; you can contact the team to review the setup.
What is the payoff for redlining your equipment? A half-degree tighter radius? If you run an air-bending test piece and measure the angle, you will almost always find that a quick depth adjustment on the CNC controller brings the second hit within tolerance. You save your tooling, protect your side frames, and still ship a good part. The test shows that the massive tonnage spike required for coining offers no practical return on investment for standard fabrication.
The hardest habit to break is the belief that pushing harder makes the part better. When you coin, you are fighting the metal. When you air bend, you are working with the machine’s geometry. True precision does not come from concentrating thousands of pounds of kinetic energy into a localized strike. It comes from understanding how punch radius, die opening, and material properties interact in open space.
Once you trust your tooling and controller, you realize that springback is not an enemy to be crushed but a variable to be calculated. On your next setup, before you even consider changing the tonnage setting to force a stubborn angle, run a single test bend in the air, measure the actual result, and adjust the ram depth.
