When a long bend comes out at 89 degrees on the ends and 92 degrees in the center, you may assume that sliding some brass shim stock under the bottom die is a permanent fix for the uneven angle. It is not. Shimming air-bending dies works only as a controlled, measured correction after the brake has been leveled, seated, mapped, and crowned. Apart from small, temporary adjustments needed to finish a run, blind shimming merely hides a structural deflection problem that maintenance or alignment needs to correct.

Related: How to Modify Press Brake Dies

Start With the Right Diagnosis: Uneven Bend Angles Are Usually a Symptom of Deflection

The 92° center / 89° ends pattern: what the ram and bed do under load

Watch a 200-ton brake reach bottom dead center on a heavy bend. If you place a magnetic dial indicator on the center of the lower bed and zero it before the bend, you will see the needle move as tonnage builds.

Hydraulic cylinders push down on the far ends of the ram. The heavy side frames push back up on the far ends of the bed. The center of the machine, however, has no direct support. Under massive pressure, the ram bows upward in the middle, and the bed bows downward. This increases the gap between the punch and die at the center of the machine. Because the punch does not penetrate as deeply into the V-opening, the bend angle remains open. The die itself has not changed shape; the massive steel frame holding it has temporarily stretched.

Why “fixing the die” feels logical when the frame, foundation, or crowning is the real variable

Looking down the V-opening of a 12-foot lower die, it appears to be a single, rigid piece of infrastructure. When the center bend comes out loose, slipping a piece of shim stock under that tooling can feel like a permanent fix for a localized problem.

You are treating a dynamic problem with a static solution. Think of the shim as a rigid bridge over a dynamic river. The shim remains exactly the same thickness, but the machine's deflection changes with every variation in tonnage, material thickness, and bend length. If the brake's foundation has settled slightly, or if the mechanical crowning wedges are packed with grit and binding, the frame will deflect unpredictably. Fixing the die assumes the tooling is at fault, masking the reality that the machine's geometry is shifting under your feet.

Separate machine deflection from material springback, grain direction, thermal drift, tooling wear, and punch penetration error

A sudden two-degree variation in the middle of a run does not automatically mean the frame is bowing.

Before blaming deflection, isolate the variables. Did the operator just load a sheet cut with the grain instead of across it? Springback can change dramatically. Has the hydraulic oil heated up over a long shift, causing thermal drift in the Y-axis positioning? Is the tooling visibly worn in the center from years of operators making short bends in the middle of the bed? If the punch penetration is consistent but the angle is wrong, you likely have a material or tooling issue. If the punch penetration physically changes from the ends to the center under load, you are looking at deflection.

For a broader refresher on how bend angle, punch penetration, material behavior, and tooling setup interact, ADH Machine Tool’s CNC-focused bending resources include this related guide to press brake bending basics.

Once you have confirmed that frame deflection is the real cause, leave the shim stock in the drawer—you need to establish a clean, measured physical baseline for the machine before attempting any correction.

Level, Seat, and Measure the Brake Before Any Shim Goes Under the Die

Why shimming an unlevel or twisted brake compounds misalignment instead of correcting it

Imagine a 150-ton brake sitting on a standard 6-inch shop floor slab. Over five years, the side facing the heavy-traffic aisle settles by just a quarter of an inch. The massive steel side frames twist slightly, causing the ram guides to bind and throwing the entire bending axis out of square. If you slide a .010" shim under the center die to fix a loose bend on this machine, you are not correcting deflection. You are building a ramp over a twisted foundation.

When you shim an unlevel machine, you lock that physical distortion into your tooling setup. The shim forces the punch and die to meet evenly for one specific bend, but it does so by working against the machine's natural geometry. Change the tonnage or the tooling, and that built-in twist responds differently, throwing your angles out again. You end up chasing your tail, adding and removing shims daily, because the machine's baseline is constantly fighting the correction.

What “level” means: foundation, frame, bed, ram, tooling seats, and ram-to-bed parallelism

Putting a carpenter's level on the front of the bed tells you nothing about the machine's geometry. In a press brake, "level" is a stack of precision tolerances built from the ground up.

A true level check requires a precision machinist level that reads to .0005 inches per foot. You must verify that the bed is level left to right to ensure gravity is not pulling the ram sideways, and front to back so the punch enters the die perfectly vertical. If the bed pitches forward, the punch will push the material horizontally, scraping the die shoulder and ruining the angle. Beyond the frame, level also means that the ram and bed are parallel at bottom dead center (BDC). If the Y1 cylinder bottoms out a fraction of a millimeter before the Y2 cylinder, the punch enters the die at a slant. No amount of shim stock in the center of the bed will fix a machine that is fundamentally out of parallel.

Check for dirt, burrs, worn clamps, mismatched die segments, or poor seating before blaming deflection

A single human hair is roughly .003" thick. A piece of laser slag or a rolled burr on a die tang can easily measure .008". When a die sits on that debris, the tooling is raised locally, forcing the punch to bite deeper and overbend the material in that exact spot.

Operators often see this tight bend, assume the machine is deflecting strangely everywhere else, and start shimming the loose spots to match. Pull the tooling out. Wipe the tangs and the bed slot with a clean rag. Run a flat precision stone over the bed to catch and knock down high spots or burrs. Check the tooling clamps for broken springs or galled surfaces that prevent the die from seating flat. If you are mixing die segments from different manufacturers—or even from different batches from the same manufacturer—measure their overall heights with calipers. A .005" height difference between two adjacent die segments will look exactly like a deflection problem on the part.

Map a test bend at the ends, quarters, and center using consistent material and setup conditions

Take a piece of 10-gauge mild steel cut to the full length of the bed. Do not use drops, scrap from different heats, or material with inconsistent grain directions. Bend it.

Measure the resulting angle at five specific points: the far left end, the left quarter, dead center, the right quarter, and the far right end. Write those exact numbers directly on the steel with a marker. This is your physical map of what the machine is actually doing under load. If the numbers read 90-91-92-91-90, you are seeing classic, symmetrical machine deflection. The frame is bowing in the center exactly as physics predicts. But if the numbers read 89-90-92-93-94, you do not have a deflection problem. You have a parallelism issue, a twisted frame, or a damaged cylinder.

Use a dial indicator or parallelism check to confirm whether the error is symmetrical load deflection or a deeper alignment problem

Mount a magnetic dial indicator to the ram and sweep the bed from left to right. If the test bend showed a linear progression from 89 to 94 degrees, the indicator will likely confirm that the bed and ram are physically out of parallel at rest or at BDC.

This distinction determines your entire troubleshooting path. Shims cannot correct a parallelism error; that requires adjusting the Y1/Y2 cylinder parameters in the control or re-leveling the mechanical stops. If the test bend showed a symmetrical bell curve, open in the middle, but the dial indicator shows perfect parallelism at rest, you have isolated the issue to dynamic deflection under load. You now have a clean, seated, and measured baseline. The machine is true. The error is predictable. This is the point where you stop looking at manual shims and start looking at the machine's built-in systems designed to counter this exact curve.

Use Crowning First When the Brake Has It, Then Shim Only What Crowning Cannot Fix Quickly

You have verified that the machine is level, the bed is clean, the tooling is seated, and the cylinders are parallel. If the center of your test bend still measures wider than the ends, you are dealing with pure, unavoidable deflection. Physics dictates that under heavy tonnage, the side frames stretch and the unsupported center of the ram and bed bow away from each other. The solution to this structural reality is not sliding a piece of brass stock under the die. The solution is to use the machine’s built-in crowning system to actively counter the curve.

Manual wedge, hydraulic, and CNC crowning: what they correct better than static shims

Imagine bending a 10-foot piece of 1/4-inch plate. As the punch engages the material, the required tonnage forces the center of the ram upward and the center of the bed downward in a smooth, continuous parabolic arc. A crowning system is engineered to mirror this exact geometry. Whether it uses a series of precision-ground opposing wedges moved by a hand crank, hydraulic cylinders built into the lower beam, or a CNC-controlled motor tied to the machine's material library, crowning pushes the center of the bed upward in a sweeping curve; for shops that need this correction built into repeatable bending workflows, an ADH Machine Tool CNC press brake provides a relevant next step because its CNC-based bending platform is designed around programmed control rather than static compensation.

A shim cannot replicate this geometry. A shim is a rigid step. If you place a .015" piece of shim stock under the center of a die, you do not create a parabolic arc; you create a localized pivot point. The die bridges over the shim, leaving unsupported air gaps on both sides. When tonnage is applied, the tooling flexes over this hard spot, risking die fracture and creating a sharp transition in the bend angle rather than a smooth, continuous correction along the length of the bed.

Open in the center vs. closed in the center: which direction the crown must move

To adjust a crowning system effectively, you must convert the metal’s bend angle into the physical distance between the punch and die. When the machine deflects, the gap between the upper and lower beams increases at the center. The punch does not penetrate as deeply into the die cavity, leaving the bend angle “open,” or obtuse. If the ends measure 90 degrees and the center measures 92 degrees, the center is open. The crowning system must move upward to compensate, pushing the lower die closer to the punch so it penetrates deeper and closes that angle.

Conversely, if you apply too much crown, you overcompensate. The bed bows upward so strongly that the punch penetrates deeper at the center than it does at the rigidly supported side frames. The resulting bend will be “closed,” or acute, in the center—measuring 88 degrees while the ends remain at 90. Reading this open-to-closed relationship tells you exactly which direction to drive the wedges or how to adjust the CNC parameters. You are raising the bed to close the angle and lowering it to open the angle.

Why random center shims can fight the crowning system and lock one correction into every tonnage and material

Crowning systems are dynamic; shims are static lumps. Suppose an operator slides a strip of shim stock under the center die to correct an open bend on heavy plate, then forgets to remove it when the next shift switches to light-gauge aluminum. The CNC crowning system automatically calculates the new job, reducing its upward bow because the lighter material requires less tonnage and causes less deflection.

The machine control assumes the bed is flat. But the forgotten shim is still there, rigidly raising the center die. You have quietly introduced a permanent high spot into a dynamic system. The aluminum part bends severely closed in the center, and the operator, trusting the CNC, assumes the crowning motor is broken or the machine is out of parallel. Mixing static shims with dynamic crowning corrupts the machine’s calculations, locking a single rigid correction for one specific tonnage into every later job you run.

When a non-crowning brake or production emergency makes controlled shimming reasonable

Not every shop floor runs late-model brakes with CNC crowning. If you are using an older mechanical brake or a basic hydraulic machine with a flat bed, deflection remains a physical reality, and you must compensate manually.

For shops weighing whether repeated manual compensation is still acceptable, ADH Machine Tool’s CNC-based bending equipment can be a useful comparison point; readers who want concrete model and capability details can review the downloadable brochures.

Even on modern machines, crowning creates a global curve that cannot correct localized tooling defects. If you run thousands of short, heavy brackets in the exact same spot on the bed, you will eventually wear a physical divot into the shoulders of that specific die section. The crowning system cannot push up only one localized two-inch section of the bed to save a worn die. In these specific scenarios—where the machine entirely lacks the hardware to create a curve, or the tooling itself is locally compromised—measured, controlled shimming changes from a lazy shortcut into a necessary mechanical intervention.

Build the shim pack from the bend map, not from the worst angle alone

Imagine bending a 12-foot piece of 10-gauge steel and measuring a 92-degree open angle exactly at the center, with perfect 90-degree ends on both sides. The immediate shop-floor temptation is to grab a thick piece of shim stock, slide it directly under the die at the 92-degree mark, and press the pedal again. Machine deflection is not a cliff; it is a sweeping curve. A shim, however, is a rigid step. If you address only the worst angle, you are building a sharp rock in the middle of a dynamic river, point-loading the die under extreme tonnage. To shim safely, you must build a correction that mimics the machine’s natural curve; for shops that repeatedly fight long-part deflection, ADH Machine Tool’s CNC bending portfolio also makes a tandem press brake a practical equipment path to evaluate instead of relying on shims as a recurring fix.

Locate the peak of the deflection curve and determine how far the correction must taper

You cannot correct a curve if you do not know its shape. Before cutting any shim stock, map the bend along the full length of the part. Use your protractor and measure the angle every 12 inches. You might see a progression like: 90, 90.5, 91, 92, 91, 90.5, 90.

This sequence shows exactly what the bed is doing. The peak of the deflection is the 92-degree measurement, but the curve actually starts three feet away on each side. If you put a shim only under the center foot of the die, the neighboring sections are left unsupported. When the ram comes down, the die will flex over your center shim like a seesaw. To work correctly, your correction must cover the entire six-foot affected area, tapering to zero exactly where the bend naturally returns to 90 degrees.

Choose consistent shim stock with known thickness tolerance; avoid scrap strips, foil slivers, and point loading

Precision depends on predictable materials. Digging through the scrap bin for sheared sheet metal drops, torn banding strap, or crumpled aluminum foil brings large thickness variations into a process that requires accuracy to thousandths of an inch. Scrap metal often has burrs or sheared edges that create high spots, while soft foils compress unpredictably under high tonnage.

Use graded brass or stainless steel shim stock. These materials are made to tight thickness tolerances and resist compression under load. When you use a piece of .005" brass, you know exactly how much you are raising the die. If you place uneven or unknown materials under a lower die, the tonnage concentrates on the thickest, hardest points. This point loading acts like a wedge against the bottom of the die, and during a heavy bottoming or coining operation, it can split a hardened tool straight down the middle.

Start thin, test often, and avoid jumping straight to a thick shim

Overcorrecting is easier than it may seem. If your center angle is open by two degrees, you might assume you need a .020" shim to close it. However, placing a thick shim under a die often changes how the tool seats in the holder, sometimes pivoting the die slightly or causing it to bridge in unexpected ways.

Start with half the thickness you think you need. Slide a .005" or .010" shim into the peak area and run a test piece. Measure the new angle progression to see how the machine responds. Does the center drop to 91 degrees? Does the taper move outward? By approaching the final dimension gradually, you keep the machine from suddenly bending acute in the center, which would only force you to tear the setup down and start over.

Feather the shim outward so you correct the bend without creating a new high spot

To match the machine's parabolic deflection, your shim pack should resemble a shallow staircase. Using your bend map, build the correction in layers.

Suppose your mapped deflection covers six feet. Your base layer is a six-foot-long piece of .005" stock, centered under the die. Next, cut a four-foot piece of .005" stock and center it on top of the first layer. Finally, cut a two-foot piece of .005" stock and center it at the very peak. You have now built a smooth, three-tiered ramp that provides .015" of lift at dead center, tapering down to .005" at the edges before smoothly returning to the flat bed. This feathered approach matches the machine's curve so the tooling can safely absorb the tonnage.

Keep total correction under 0.5 mm unless maintenance has approved a specific temporary workaround

There is a firm mechanical limit to what shimming can actually correct. If you find yourself building a feathered shim pack with a total thickness greater than 0.5 mm (roughly .020"), stop.

A gap larger than 0.5 mm indicates a mechanical failure that shimming will only conceal. The ram may be permanently bowed, the bed may be warped, or the die shoulders may be worn out. Covering a half-millimeter gap with brass stock places immense, unnatural stress on the machine frame and tooling. At this point, the issue is no longer an operator adjustment; it is a maintenance problem. Unless the maintenance department has explicitly mapped the machine and authorized a temporary shim pack to complete a critical run, going beyond 0.5 mm is simply inviting broken tooling.

Install and Validate Shims Without Sacrificing Die Support or Repeatability

You have mapped the curve and cut your feathered layers. But placing them under the tooling is where theory meets actual steel. Imagine sliding a .015" brass shim under a standard 4-way V-die on a 200-ton brake, only to find that the die now chatters when clamped. That chatter is the sound of lost bearing contact. A perfectly feathered shim pack is useless—and dangerous—if it makes the die rock, shift, or lose its structural footing, turning a localized angle correction into a mechanical hazard.

Decide whether the shim belongs under the die, the die holder, or an individual segment based on how the tooling seats

Tooling geometry determines shim placement. If you are running a solid, full-length American-style die directly on the bed, the shim goes between the bed and the die. But if you use European-style segmented tooling in a die holder, placing a shim directly under a single 100 mm segment creates a step. When the ram engages, that raised segment takes the brunt of the tonnage before the adjacent segments engage, creating a risk of shattered tooling.

Instead, place the shim under the die holder itself. By shimming the holder, the segmented dies remain perfectly flush with one another along the top edge. The holder absorbs the localized lift and distributes it across the segments. The goal is to raise the bending surface without interrupting the continuous plane of the tooling.

Lock out the machine, support the tooling, clean and deburr contact surfaces, and clamp evenly

Never put your hands between the punch and die without locking out the ram. Once the machine is safe, lift the die or holder only enough to slide your feathered pack into position. Before the brass goes in, wipe the bed and the bottom of the tooling with a clean rag and solvent. A single curled steel chip trapped under your shim pack adds an unplanned .010" point load to your carefully measured correction.

Inspect your brass stock for sheared burrs. If you cut the shim yourself, run a flat stone over its edges. When lowering the tooling back down, seat it firmly. If you are using hydraulic clamping, engage it and watch the die pull down flat and tight. If you are using manual set screws, tighten them evenly from the center outward to avoid bowing the die holder against the shim pack.

Maintain full-bearing contact so the die does not rock, shift laterally, or damage the holder under tonnage

Press brake dies are designed to transfer tonnage straight down through their base and into the bed. When you add a shim, you change that load path. A die that sits high on a shim pack can act like a seesaw. Try to rock the die by hand. If it clicks or tips forward and backward, the shim is too narrow or positioned off-center.

Under load, a rocking die will try to kick sideways. This lateral movement puts heavy shear force on the clamping tangs or set screws, which are meant to hold the die in position, not absorb bending tonnage. If the die does not have full bearing contact across its entire base, the concentrated pressure can coin a permanent indentation into the die holder or machine bed. You must confirm solid metal-to-metal contact before cycling the ram.

Re-bend the same test strip and recheck the angles at the ends, quarters, and center

Do not use a fresh piece of steel to verify your correction. Take the exact test strip you used to map the initial deflection. Put it back in the brake, align it exactly as before, and re-strike the bend.

Using the same strip removes material thickness and grain direction as variables. Measure the ends, the quarter marks, and the center again. You want the center angle to close up until it matches the ends. If the center is over-bent, your shim pack is too thick. If the quarter marks have opened up, your feathering taper is too steep. The physical bend is the only feedback loop that verifies your math.

Reset the overall ram depth or programmed angle only after the local correction is stable

Once your test strip measures a consistent angle across its full length—for example, a uniform 91 degrees from end to end—your local deflection problem is solved. The shim pack has successfully leveled the playing field. Only at this point should you adjust the machine's global controls.

Adjust the Y1/Y2 ram depth or change the programmed angle in the controller to bring that uniform 91 degrees down to the target 90 degrees. Never adjust the global ram depth to chase a local deflection curve; you will only over-bend the ends in an attempt to save the center. Stabilizing the local correction first isolates the variables and restores predictability. However, this stabilization depends entirely on the required shim thickness remaining small enough to be safe.

Treat the 0.5 mm Rule as a Stop Sign, Not a Setup Preference

A half millimeter may look like only a few sheets of paper, but under heavy tonnage it becomes a major structural wedge. When a localized correction exceeds this limit, you are no longer compensating for natural material variation or minor tooling wear. You are forcing the tooling to bridge a physical gap that the machine's frame cannot close. This is where a temporary setup trick becomes a way of masking a mechanical failure.

What excessive shim thickness suggests: bed deflection, ram sag, gib wear, hydraulic imbalance, foundation movement, or worn tooling

If you need .030" of brass to get a straight bend, the tooling is rarely the culprit. That much missing steel means something in the press brake's geometry has given way. The bed may be permanently bowed from years of overloading short parts in the center. Ram sag or worn gibs can cause the upper beam to tilt or twist under tonnage. Hydraulic cylinders may be out of sync, fighting each other instead of driving the punch evenly into the die. Even a settling foundation can twist the machine frame enough to require massive shims just to find parallel. Stacking brass under the die does not fix any of these root causes; it only treats the symptom while the underlying problem gets worse.

When thick shim stacks keep returning on heavy or long bends, the better next step may be to reassess whether the machine is still matched to the work. ADH Machine Tool’s CNC bending portfolio includes large press brake solutions for operations that need more reliable geometry, capacity, and control than setup corrections alone can provide.

Why permanent shim stacks hide damage and make future angle problems harder to diagnose

Leaving a large shim pack under the die after the job is finished creates a trap for the next operator. A static shim is a rigid fix for a dynamic problem. When the next job requires lower tonnage or a shorter bend length, that permanent shim becomes a high spot and forces the center to over-bend. The operator then starts chasing the issue, adjusting Y1/Y2 parameters or adding counter-shims to fight the original shim. Before long, the machine’s baseline is completely lost. You can no longer trust the crowning system, the ram encoders, or the tooling, because a hidden lump of brass is distorting every measurement on the floor.

Record shim thickness, location, material, job number, and date so the next operator does not inherit a mystery

If you must shim to get a critical job completed, treat that shim as documented tooling, not shop-floor debris. Write down exactly what you did. Record the shim thickness, the precise location on the bed, the material used, the job number, and the date. Tape this record to the setup sheet or enter it into the controller’s notes. When the shift changes, the next technician needs to know exactly why the die is sitting proud. This transparency prevents the next operator from tearing down the machine looking for a hydraulic fault when the real cause is a forgotten piece of brass left over from last week’s heavy plate run.

Decision path: emergency shim for production, crowning adjustment for repeat work, and professional alignment or service when the shim stack keeps growing

Every time you reach for shim stock, you are making a maintenance decision. If you are mid-run on a critical order and the center angle drops by a degree, an emergency shim is a valid tactic to finish the shift. If the same part requires the same correction every time it reaches the floor, your crowning system needs recalibration or adjustment to handle that specific tonnage curve. But if you find yourself cutting thicker shims every month just to hit 90 degrees on standard-gauge steel, stop. The machine is telling you it is compromised. Pull the tooling, check the bed for parallelism, and inspect the gibs. If your shim pack keeps growing beyond that half-millimeter mark, put the brass down and schedule a service call to correct the deflection at its source; if the issue points to a broader bending-capability or equipment evaluation, ADH Machine Tool’s CNC-based bending and sheet metal solutions make it a practical supplier to contact for a technical discussion.