Printed circuit boards sit underneath nearly every piece of electronics in daily life, and it is easy to take that foundation for granted. Within the PCB family there are three broad categories — rigid, rigid-flex, and flex — and all of them perform the same fundamental job of connecting components. The difference lies in construction. Rigid boards cannot bend, while a flexible circuit board can be shaped and routed to fit a product’s mechanical envelope, and a rigid-flex board combines both behaviors in a single unit. That flexibility is not simply a convenience. It changes the way a design must be planned, because a board that bends is subjected to mechanical forces that a flat rigid board never experiences.
What Separates a Flex Circuit From a Rigid Board
The idea of routing conductors onto a flexible substrate is older than many designers assume. Early in the twentieth century, researchers explored ways to simplify complex wiring by embedding conductors between insulating materials. Thomas Edison’s assistant, Frank Spange, wrestled with how to lay a circuit track onto an insulating surface, and Edison proposed mixing carbon powder into epoxy glue and spreading it over linen fabric paper — an approach that was never proven in practice but resembles the polyamide-based constructions used today. During the Second World War, German engineers applied flat conductor track technology to tank towers and the V2 rocket, and in the early 1950s American space researchers examining a recovered V2 discovered flexible circuitry, which helped drive adoption in the United States. Pat Bryan, working at Lockheed, carried samples back to California, and the technology spread into aerospace and beyond.
What that history shows is that flexibility is a materials and architecture decision, not an afterthought. Because a flex board bends, it has requirements that rigid circuits do not share. The design process has to account for the substrate and coverlay materials, the circuit architecture itself, where features are placed on the panel, and the number of layers in the stack. The most important early question is mechanical rather than electrical: how far will the board bend, how tight is that bend, and how many times will the flexing occur? A board that bends once during assembly and stays put behaves very differently from one that flexes continuously in a hinged product. Designers who understand this distinction, and who work within the specific constraints of flex technology, are the ones who capture its real advantages — compactness, light weight, and the ability to replace wire harnesses and connectors with an integrated circuit.
It is worth remembering that flex does not automatically mean fragile. A properly constructed flex circuit can be extremely robust, but robustness is engineered rather than assumed. The requirement for robustness is what drives the design rules described below, and it is also why early mechanical review matters more on a flex project than on a conventional rigid one.
Core Design Rules for a flexible circuit board
The single most useful concept in flex design is the neutral bend axis. When a circuit bends, the material on the outside of the curve is in tension and the material on the inside is in compression. Between them runs a plane that experiences neither. The most effective way to control that plane is through balanced construction — distributing copper and dielectric so the neutral bend axis lands near the center of the stack. The closer that axis sits to the middle of the material stack during flexing, the more evenly the resulting forces spread across the remaining layers, and the less stress each layer has to absorb.
Bend angle is the next lever. The less a circuit is flexed, the lower the possibility of damage. A maximum bend of 90° is a reasonable target for a design that must survive repeated movement. A correctly designed circuit can often be bent beyond 90° a single time during installation, but for dynamic applications — where the circuit is bent, flattened, and bent again many times — the bend should not exceed 90°. Beyond that threshold, copper fatigue and coverlay cracking become realistic failure modes rather than theoretical ones.
Thickness in the bend area matters just as much. Damage risk falls as thickness falls, so reducing base copper weight and dielectric thickness in the flexing region is a standard technique, and adhesive-based base materials can help designers bring overall thickness down further. Finally, radius should be treated as a design feature. A 90° bend can usually be achieved with either a sharp, small radius or a more gradual, large radius, and the gradual option is the safer one for a flexible circuit board. Note that bend radius is measured from the inner surface of the bend, so the number used in calculations should reflect that convention rather than the outer contour.
These rules interact. A thin stack with a generous radius and a well-centered neutral axis tolerates far more flexing than a thick stack with a tight corner, even when both use identical materials. Where a design cannot satisfy all four simultaneously, the bending requirement should drive the compromise.
Turning a Flex Design Into a Manufacturable Board
Design rules only deliver value if they survive contact with fabrication. That is why the practical step after laying out a flex circuit is a design review with the manufacturer that will build it — before tooling is committed, not after. A supplier that produces flex and rigid-flex boards in the same facility can speak to both sides of a hybrid design, which matters when a board transitions from a rigid section carrying fine-pitch components to a bending tail that routes to a connector.
Manufacturing capability determines which design choices are realistic. On the fine-feature side, minimum trace width and spacing of 0.002 inches and laser-drilled microvias open room for dense routing in the rigid portions of a rigid-flex assembly, while impedance control held to plus or minus five percent supports high-speed signal integrity across the flexible sections. For RF work, laminate families such as Rogers, Isola, Arlon, Taconic, and Panasonic MEGTRON grades are available for designs requiring low-loss or high-speed performance. Surface finish selection also deserves early attention: selective gold, hard gold, gold fingers, and ENEPIG each suit different contact and assembly requirements, and choosing late can force a layout change.
Real projects show why this collaboration pays off. A wearable device may need a flex tail that folds three times inside a housing smaller than a matchbox; a camera module may require a dynamic bend that cycles thousands of times; an instrument may use semi-flex construction to replace a connector without a full flex stack. In each case, the electrical design is only half the job. The other half is confirming the bend location, the layer stack, the copper weight, and the finish are all buildable together. Careful design, checked against a manufacturer’s real process limits, is what yields a circuit suited to its purpose and delivers the advantages of flex circuitry at the lowest overall cost.
