Joggle and Hemmed Stamped Parts: Design Guide for Offset Bends

Two EMI can halves arrive at the assembly line with a 0.15 mm gap at the flush overlap joint, even though the drawing called out the mating surfaces as "flush." The root cause was not the press, not the coil, and not the operator. A designer had modeled a joggle offset equal to one material thickness on 0.4 mm SPCC with an inside radius of 0.1 mm, placed the bend 1.2 mm from the free edge, and assumed that because the stamping house holds +/-0.005 mm on positional features, the formed seam would inherit that number. It did not. Air-formed bent dimensions on that alloy land in the +/-0.05 mm band before anyone argues about springback, and the joggle walked off the web and tore the edge. The assembly shipped with a seam gap that the customer's RFI test flagged, and the tooling was already cut. Joggles and hems look like the cheapest features on a stamped part. They are also the features most often specified from memory, and memory is where flush lap joints go to die.

The Snapshot

• A joggle is two opposed bends that offset the sheet midline on purpose, so overlapping flanges close into a flush, zero-step surface instead of a double-thickness stack.
• Practical inside-radius floor for a joggle on mild steel is about 1.0 t on the tight leg and 1.5 t on the relaxed leg; below that you are coining, not bending, and tonnage roughly triples.
• Minimum web between a joggle line and a free edge is roughly 2.0 t plus one inside radius; at 0.5 mm material that is 1.5 mm of strip you cannot cheat with a faster press.
• Bend-angle tolerance on formed features runs +/-0.5 to 2 degrees by material, while the same press holds +/-0.005 mm on pierce positions; never merge the two numbers on one drawing.
• Hems fold the sharp cut edge into itself, removing the burr hazard from fingers and gloves, and raise edge bending stiffness by up to an order of magnitude depending on hem type.
• On a 25 to 110 ton press band, joggle and hem loads scale with tensile strength, thickness squared, and strip length of bend, which is why one 60 mm hem on 304 stainless can eat half a 63 T line's budget.
• The reliable way to mate two joggled halves is one datum-controlled partner: the joggle lives on one part, the other stays flat, or the stack doubles.

What A Joggle Actually Is

Strip a joggle down to its geometry and it is a Z made of two bends sharing one short web, and that one sentence is the whole of joggle sheet metal design. Bend one goes down, bend two goes back up, and the segment between them steps the sheet over by the offset you designed. The reason engineers draw a joggle instead of a lap is mass and interface cleanliness. If two 0.6 mm flanges simply overlap, the seam is 1.2 mm proud of the surrounding surface, and everything downstream, a gasket, a lid, a paint line, a human hand sliding across an access panel, has to deal with that step. Joggle one of the parts and the overlap closes flush: the offset dimension equals one thickness, the high side of the joggle receives the flat side of its partner, and the combined stack at the seam equals the combined stack anywhere else. This is the flush overlap joint that aerospace skin laps, automotive hood and decklid edges, and electronics closure seams all chase.

The offset dimension is the number that drives every other constraint. On a single-thickness joggle the offset is 1.0 t. On a double lap, where a doubler or a second flange enters the same pocket, the offset becomes 2.0 t or more, and the bend legs stretch accordingly. Because the two bends oppose each other, springback on one leg partly cancels springback on the other, which is why a well-designed joggle is often more angle-stable than a single 90 degree bend. Partly, not entirely: the legs must spring equally, which needs matching inside radii on both punch shoulders and a die-wear schedule behind them. The same logic makes the joggle a natural springback sensor: when the flush seam shows a visible step, the two leg radii have drifted apart. For can-style closures this dovetails with the mating-surface logic in the EMI shield can design guide.

Joggle stamping also differs from an ordinary offset bend in tooling effort. A shallow offset of a few tenths on thin gauge forms in one station with a simple punch and pad. An offset above about 2 t needs a multi-hit form stack or a deep die pocket, because the punch must clear the first bend line to make the second without stretching the web past what the material forgives. Push the return leg tall enough and the profile shades into a gooseneck bend, where the flange curves back over itself and the same offset logic governs whether the tool can reach.

Joggle Geometry Rules That Hold

The first rule: inside radius before anything else. The tight leg of a joggle wants a minimum inside radius of roughly 1.0 times material thickness on dead-soft and mild steels like SPCC, and 1.5 to 2.0 times thickness on work-hardening alloys like 304 stainless, where the bend region cold-sets and microcracks at the outer fiber if you force it tighter. The relaxed leg, the one that is allowed to open up toward the flat surface of the part, can run at 2.0 t because its springback is partly absorbed into a cosmetic fillet rather than the mating geometry.

The second rule: the web between the two bend lines. That web becomes the top face of the offset, and it also becomes the seating surface your mating flange lands on. If it is shorter than the material will allow to flow, the two bend radii interact and the web domes or twists instead of sitting flat. The safe floor is about 2.0 t plus one inside radius; below that the joggle stops behaving as two independent bends and starts behaving as a crushed feature.

The third rule: the web between the joggle and a free edge. This is the one CAD catches least, because in the model the edge is infinite and in the strip it is a blank outline or a cut station a millimeter away. Formed metal needs backstop. If the outer bend leg of a joggle runs out of material, the edge curls, the burr lifts, and the finished flange is a fraction of a millimeter short at the very line the seam was designed around. Keep the joggle-to-edge distance at 2.0 t plus one radius minimum, and where the design absolutely cannot afford it, move the cut line in the progressive die downstream of the form so the flange is sized after forming, not before.

Here is the table suppliers actually want to see quoted on RFQs, computed for the 0.2 to 2.0 mm band that covers most precision progressive-die work:

Material thickness t (mm)	Min inside radius, tight leg (mm)	Min web between bends (mm)	Min web to free edge (mm)	Practical max offset
0.2	0.20 (1.0 t)	0.60	0.60	1.5 t
0.3	0.30 (1.0 t)	0.90	0.90	2.0 t
0.5	0.50 (1.0 t)	1.50	1.50	2.0 t
0.8	0.80 (1.0 t)	2.40	2.40	3.0 t
1.0	1.00 (1.0 t)	3.00	3.00	4.0 t
1.5	1.50 (1.0 t)	4.50	4.50	5.0 t
2.0	2.00 (1.0 t)	6.00	6.00	6.0 t

Those minima assume mild steel. Multiply the radius columns by 1.5 to 2.0 for 304 stainless and springy bronzes, by about 1.2 for hard tempers like 5052-H32 aluminum, and drop the offset ceiling when the joggle sits near a hole or another stress riser. The table is a floor, not a target: designing at 1.0 t minimum on every joggle buys you the least amount of springback compensation range in the die and the shortest radius life on the tooling.

Springback, Air Bend Versus Coining

Every bend springs back. The elastic strain that was stored during forming releases when the punch retracts against the tool. On an air bend, the punch pushes the strip into an open V-die and the final angle is whatever is left after that release. Air bending is forgiving on radii, wear, and lubricant variation, and it lets one die set make small angle corrections by adjusting shut height, which is why it is the default for prototype and low-volume joggle work. The trade is that springback is fully present: 0.5 to 1.0 degrees on SPCC, 2 to 4 degrees on 304 stainless, and sometimes more on hard aluminum tempers. A joggle with 3 degrees of uncompensated springback on each leg opens its offset by a visible amount over a 25 mm flange, and that is the step you see at the seam. Compensation in the die comes in three flavors. Overbend: cut the punch so the formed part springs toward nominal, the cheapest fix, and why first-article angle corrections on production tooling routinely run 0.5 to 1.5 degrees off the drawing. Radus change: open up the inside radius on the tight leg to shed springback, which costs corner sharpness. And coining: at bottom dead center, the punch squeezes the bend zone past yield so the plastically deformed hinge zone dominates and the angle holds, with the joggle legs effectively set to a machined flat-to-flat dimension. Coining a joggle also sets the offset dimension itself, which is the whole point on a flush overlap joint that must mate blind. The penalty is force. Air-bending tonnage per unit length scales roughly with tensile strength times thickness squared divided by die opening; coining needs three to ten times that, on top of die wear that shortens regrind intervals from hundreds of thousands of hits to tens of thousands. Most production joggles on thin gauge land in the middle: slight overbend with a light bottom strike, not a full coin. Run the arithmetic once so it stops being folklore. A 40 mm joggle line in 0.5 mm 304 stainless, 400 MPa tensile, in a 4 mm die opening air-bends at roughly 33 kN per meter, about 1.3 kN for that line, trivial on any press. Coining multiplies it five to eight times, add pilot and strip-push loads riding the same slide, and a 120 mm hemmed perimeter on 1.0 mm stainless can crowd a 25 ton line's budget while a 63 ton press barely notices. On a 25 to 110 ton press band, treat every forming station as a budget line. Piercing and bending loads add, 300 strokes per minute is not 60 strokes per minute for dynamic allowance, and the joggle or hem station usually decides whether the job runs on a 45 ton or needs a 110 ton press. Specify strip width alongside the bend detail: a wide stainless strip drags a strip-bending load the CAD file never mentions.

Hem Types And What A Hem Actually Buys You

A hem folds the cut edge of a flange back into itself. It starts with a 90 degree bend, then closes the edge over, either partially or fully. The first thing it buys is safety and cleanliness: the sheared edge, the one that carries a burr from the blanking station no matter how sharp the punch is, disappears inside the fold. Nobody slices a glove closing an enclosure lid, and no loose burr particle stands at the seam as an RFI threat. The second thing it buys is stiffness. A flat 0.6 mm flange is a hinge line; the same flange with a rolled hem behaves like a deep edge member, and panel-edge drumming on automotive hoods and trunk lids is tuned with hem geometry, not thickness, because hemming adds bending stiffness at constant mass. The third thing is cosmetic: the visible edge becomes a radius instead of a line of bare sheared steel.

Four families dominate design conversations. The tear-drop hem is the round one: the edge curls into a closed loop, the most forgiving to form and the most common on sheet metal boxes and covers. The rolled hem, sometimes called the grooved or teardrop variant depending on shop tradition, flattens that loop into a tight radius for a cleaner look at higher forming load. The fishtail hem opens the curl into a V-shaped flare that springs back less and reads sharper on a cosmetic edge. The arrowhead hem closes the edge into a tight double fold with a pointed cross-section, the tightest radius you can ask for, used when the design needs the hemmed edge to tuck into a channel or meet a near-zero reveal. Arrowheads demand the most tonnage and the most accurate shut-height control, and they crack on stainless if you ask for a radius below what the alloy's elongation will survive.

Hem type	Cross section	Forming difficulty	Edge stiffness gain	Typical use on stamped parts
Tear-drop	Closed circular curl	Low, most forgiving	High	Enclosure covers, brackets, general de-burr
Rolled	Flattened tight radius	Medium	High	Cosmetic panel edges, visible perimeters
Fishtail	Open V flare	Medium	Medium	Springback-prone alloys, hidden edges
Arrowhead	Tight double fold, point	High, needs coining pressure	Highest	Channel tucks, zero-reveal automotive edges

Hem direction matters more than people expect. A die-side hem curls toward the die face and needs pocket clearance; a punch-side hem curls the other way. On a two-sided part like an EMI can wall butting a mating frame, decide which face carries the radius before the die is detailed. And when a flange is both joggled and hemmed, the hem station must come last: hemming work-hardens and locks the edge, so forming a joggle against a locked hem leaves a crack waiting at the fold intersection.

The Tolerance Stack You Should Draw

Precision progressive-die stamping genuinely holds +/-0.005 mm on hole-to-datum and hole-to-hole positions, and +/-0.02 mm on many outside-profile dimensions, on a press with a rigid feed and a coil that knows its own mind about gauge. Bent dimensions do not inherit those numbers. A 90 degree air-formed flange angle tolerance of +/-0.5 to 2 degrees, applied at the tip of a 15 mm flange, translates to 0.13 to 0.5 mm of position. An offset joggle dimension on thin material, formed rather than coined, realistically lands at +/-0.05 to +/-0.10 mm before anyone argues, and +/-0.02 to 0.03 mm only after deliberate overbend tuning and a coining hit. Two independent joggled halves, each carrying that band, stack: worst case your flush seam has a 0.10 to 0.20 mm mismatch between opposing offsets, and a "flush" joint is flush only within the tolerance band you actually drew.

The fix is to never make the assembly pay for two formed dimensions. Put the joggle on one part and let the mate stay flat and pierce-position-controlled. If both must carry it, tie them to a common datum and write seam flatness as an assembly-level requirement verified with a gauge, not two identical part-level numbers that add in quadrature and surprise everyone at pilot build.

Material	Typical air-bend angle tolerance (90 deg bend)	Realistic bent-dimension tolerance (flange length or offset)	Springback tendency	Die-side countermeasure
SPCC / DC01 mild steel	+/-0.5 to 1.0 deg	+/-0.05 mm	Low, 0.5 to 1.0 deg	Overbend, optional light bottom strike
304 stainless (annealed)	+/-2.0 to 4.0 deg	+/-0.08 to 0.15 mm	High, work-hardens	Overbend + coining hit, radius 1.5 to 2.0 t
5052-H32 aluminum	+/-1.0 to 2.0 deg	+/-0.05 to 0.10 mm	Medium-high, elastic	Overbend, avoid coining (surface pickup)
SPCC hard (SPCD/SPCE band)	+/-1.0 to 1.5 deg	+/-0.05 mm	Medium	Overbend, watch inside-leg cracking
C2680 brass / C5191 phosphor bronze	+/-1.0 to 2.0 deg	+/-0.02 to 0.05 mm	Medium, springy	Overbend, tight radii not available

Those numbers assume a sharp die, first-article tuning, and a supplier who measures bent angles instead of eyeballing them.

Progressive Die Sequencing And Bend Allowance Math

On a progressive die, a joggle or hem is a station, and a station has an order problem; the progressive die stamping feed-and-form logic behind that is worth reading alongside this section. The universal rule: form before cut-off, and form before any free edge is created if the form needs material backstop. In practice that means the bend block and the joggle punch ride on strip that is still connected to the carrier, so the flange has something to push against, while the trim that releases the flange profile comes after the form. This is inverted from what many solid-model designers assume, because in CAD every feature is Boolean and orderless. A blank that is already cut to shape cannot hold a crisp joggle near its edge on a high-speed press; the carrier web is the fixture. If the design cannot route the joggle before the trim station, the alternative is an in-die robot or a separate secondary forming operation, and both cost more per piece than moving the trim line.

Bend allowance math decides whether the part fits in the strip at all. The neutral axis of a bent zone shifts toward the inside radius, and the developed bend length is the angle times that neutral radius. The standard bookkeeping: bend allowance minus twice the setback gives the bend deduction, the amount the flat pattern shrinks relative to the sum of the flange legs. K-factor, the neutral-axis shift as a fraction of thickness, runs about 0.33 for sharp bends and 0.4 to 0.5 for radii at or above one thickness on mild steel. On a 0.5 mm SPCC joggle leg at 1.0 t radius and 90 degrees, setback is roughly 1.5 mm and bend deduction about 1.3 mm per bend; miss it across a four-bend cluster and every hole downstream lands a half millimeter off, which is how a part claiming +/-0.005 mm positional holes ships beautiful and out of tolerance.

On thin material with tight radii, the neutral axis shifts further and the bend allowance shrinks; that is the regime where suppliers ask for radius callouts on every bend in the drawing, not a blanket "break edges" note. Put the inside radii and the offset dimension directly on the part model, show the developed strip layout in your first-article data if you can, and let the tooling cost go to the station that needs precision, which is almost never the cosmetic one.

Material Suitability Notes

SPCC and its mild-steel family remain the default for joggle and hem work: low springback, generous elongation, inside radii to 1.0 t or sharper without the edge cracking, and the widest window of predictable die behavior. When a customer asks for the tightest joggle offset the geometry table will carry, mild steel is the answer.

304 stainless is the troublemaker. It work-hardens as it forms, so the bend zone stiffens mid-stroke, springback climbs with every forming hit, and the edge of an arrowhead hem can crack along the fold intersection if the radius falls below about 2.0 t. Coin the offset, budget for a coining station and regrind intervals, and accept that first-article angle tuning on stainless dies may take two or three shim iterations. The payoff: the stainless joggle holds its flush profile far longer in service than a mild-steel one that relaxes under thermal cycling.

5052-H32 aluminum sits between. It springs back more than mild steel because of its higher strength-to-modulus ratio, and coining is less attractive because the aluminum picks up on tool steel under high bottom-dead-center pressure, marring the cosmetic face. Overbend with generous radii, and if the design needs a coin for the offset, call it out so the die shop adds chromium or titanium-nitride coating to the coining surfaces instead of discovering galling at pilot.

How You Inspect What You Specified

A joggle seam is a dimensional question, a flushness question, and sometimes a functional question, and each wants a different tool. For production-level go/no-go, a checker gauge or dedicated seam fixture is the workhorse: a hardened block with the nominal pocket geometry, drop the part in, feel for rock, or read the gap at the seam with a feeler or a light gap. Checker gauges cost nothing per part and catch the drift that a CMM running five pieces a day never sees, which is the slow march of die wear through the run.

For FAI and PPAP, an optical CMM or vision system beats contact probing on formed features: it captures bend profile, hem radius, and offset in one non-contact pass, without dragging a stylus across a soft aluminum hem and reporting the scratch as geometry. Hold the part in an assembly-posture fixture, because free-state bent parts rock and free-state angle data is fiction.

For the seam itself, run a profilometer across the joint line on the mated assembly, the honest answer to whether a flush overlap joint is flush. Name which of the three methods the control plan pays for, and keep die-tuning and inspection language identical.

DFM Checklist For Joggled And Hemmed Parts

This list overlaps the wider DFM rules to check before committing stamping tooling; run it before you release the die, and treat every unchecked item as a change order arriving at pilot build:

• Offset dimension stated as a multiple of t, with tight leg and relaxed leg inside radii called out separately.
• Joggle-to-free-edge and joggle-to-hole distances at or above the geometry table minima for the chosen alloy.
• Sequencing check: does the form run before the trim that creates its free edge? If not, is a secondary form budgeted?
• Bend allowance and K-factor assumptions confirmed with the die shop, flat pattern checked against strip layout.
• Only one mated partner carries the joggle offset, or the seam tolerance is written at assembly level with a gauge.
• Bent-dimension tolerances drawn as bent-dimension tolerances, never merged into the +/-0.005 mm positional callout.
• Hem type chosen against edge stiffness, cosmetic face, and tooling load on the target press class (25 to 110 ton).
• Springback countermeasure agreed: overbend, radius change, or coining, with regrind plan if coining.
• Material suitability reviewed: 304 stainless gets a coining budget, hard aluminum tempers get overbend only, mild steel gets the tight geometries.
• Inspection method named in the control plan: checker gauge for drift, optical CMM for FAI, seam profilometer for flushness if the drawing claims flush.
• Deburr requirement stated even on hemmed edges, because the hem hides the edge, it does not remove the root.

Frequently Asked Questions

What is the difference between a joggle and a regular offset bend?

An offset bend steps the entire part line up or down and is used for clearance and attachment geometry. A joggle is a local, shallow offset whose purpose is to close a lap joint flush, so the two bends share one short web and the feature sits inside a flange rather than running across the whole part. Practically, joggles live at smaller offsets, tighter bend radii, and higher die-side control than general offset bends.

Can I hold +/-0.005 mm on a joggle offset dimension?

Positionally, yes, because the form station sits at a known distance from datum holes and pierce positions hold that band. As a bent dimension, no, not without a coining station and a gauge: an air-formed offset starts at +/-0.05 to +/-0.10 mm by alloy, tightening to +/-0.02 to 0.03 mm after tuning with a light coin. If that band is unacceptable, move the precision to the flat, datum-controlled partner and let the joggle float.

Why does my hemmed edge crack on stainless but not on mild steel?

304 work-hardens as it forms, so the tight outer radius of an arrowhead or rolled hem exceeds the alloy's elongation during the fold, especially if the die was cut for a 1.0 t radius. Specify 1.5 to 2.0 t on the tight leg, keep the hem clear of joggle lines, and take a radius shim iteration at prototype instead of chasing cracks in production.

Does a hem really replace edge deburring?

It hides the sheared edge from contact and view, which is usually the point. It does not remove the burr root, and a heavy burr can lift the hem or show through the closed fold as a bump on the cosmetic face. Keep punch maintenance on the blanking station; a hem is an edge treatment, not a substitute for burr control.

How do I stop two joggled halves from showing a visible seam step?

Make only one part carry the offset and keep the mating part flat, or tie both joggles to the same assembly datum and verify with a seam checker gauge at build rather than with identical part-level tolerance callouts. If both halves must carry the feature, budget for coined offsets and accept a shim-iteration tuning phase at first article, because two independent air-formed bands add, and the assembly inherits the sum.

The Bottom Line

The dominant risk is tolerance fiction, drawing formed features as if they were pierce features and stacking two formed bands into a visible seam gap. Progressive-die stamping wins when the offset lives on one datum-controlled partner, geometry respects the radius and web minima, and flushness is gauge-verified, not assumed.