After spending over a decade working with flex circuits across aerospace, medical devices, and consumer electronics, I’ve learned that flex PCB design isn’t just a variation of rigid board design—it’s a completely different discipline. The materials behave differently, the failure modes are unique, and the design rules that keep your boards reliable require careful attention to mechanical stress, bend performance, and layer construction.
This guide covers everything you need to know about flex circuit design and rigid flex PCB design, from calculating bend radius to avoiding the mistakes that cause field failures. Whether you’re designing your first flexible circuit or looking to improve your existing designs, you’ll find practical, tested guidance here.
Before diving into design rules, let’s establish what we’re actually building. Flexible printed circuits come in several configurations, each suited to different applications and design requirements.
Single-sided flex is the simplest and most cost-effective option. One conductive copper layer sits on a flexible polyimide substrate, covered by a protective coverlay. These work well for simple interconnects, LED strips, and applications where routing complexity is minimal.
Double-sided flex adds a second conductive layer, connected through plated through-holes (PTHs) or vias. This configuration handles moderately complex routing while maintaining good flexibility. Most bend-to-install applications use double-sided construction.
When routing density demands it, multilayer flex stacks multiple conductor layers with flexible dielectric between them. The trade-off is reduced flexibility—more layers mean a thicker cross-section and larger minimum bend radius. Dynamic applications rarely exceed four layers for this reason.
Rigid flex PCB design combines rigid FR4 sections with flexible polyimide ribbons in a single integrated assembly. The rigid areas carry components and connectors while the flex sections allow three-dimensional folding and dynamic movement. This eliminates connectors between separate boards, reducing failure points and improving reliability in applications like laptops, cameras, and medical implants.
Material selection directly impacts flexibility, thermal performance, and long-term reliability. Getting this wrong causes premature failures that are expensive to diagnose and fix.
Polyimide (PI) dominates flex circuit manufacturing for good reason. It handles temperatures up to 260°C, has excellent chemical resistance, and maintains flexibility over millions of bend cycles. Kapton from DuPont is the most recognized brand, but several manufacturers produce equivalent materials.
Polyester (PET) costs less than polyimide but can’t handle soldering temperatures. It’s limited to applications where components attach through pressure-sensitive adhesives or mechanical connectors rather than solder.
For rigid sections in rigid-flex boards, FR4 remains standard. Some high-frequency designs use specialized laminates like Rogers materials, but these add cost and complexity.
This is where many designers make their first mistake. There are two types of copper foil used in flex circuits:
Rolled Annealed (RA) copper has elongated grain structure aligned with the rolling direction. This gives it superior fatigue resistance—critical for dynamic flex applications where the circuit bends repeatedly throughout its service life.
Electrodeposited (ED) copper has a columnar grain structure that’s more prone to cracking under repeated flexing. It costs less and works fine for static bend-to-install applications, but specifying ED copper in a dynamic design invites failure.
Traditional flex laminates bond copper to polyimide using acrylic or epoxy adhesives. Adhesiveless constructions eliminate this layer, resulting in thinner overall thickness and better bend performance. For dynamic applications or designs requiring tight bend radii, adhesiveless materials are worth the cost premium.
Coverlay is a polyimide film with adhesive backing that protects flex circuits. Unlike rigid board solder mask, coverlay must be mechanically punched or laser-cut with windows for pad access before lamination. This requires planning coverlay openings during design—you can’t simply define them in Gerber data like solder mask.
Flexible solder mask (also called photoimageable coverlay) applies like liquid solder mask but remains flexible after curing. It’s easier to process than film coverlay but has thickness limitations and different electrical properties.
Nothing kills flex circuits faster than exceeding their bend radius limits. The math isn’t complicated, but understanding when and how to apply it prevents the cracked traces that cause field failures.
Static flex (also called “bend-to-install” or “Use A” per IPC-2223) refers to circuits that bend during assembly but remain fixed in the final product. Examples include folding a flex cable into position inside a laptop or bending a circuit to fit a curved enclosure.
Dynamic flex (“Use B”) describes circuits that bend repeatedly during normal operation. Think of the flex cable connecting a printer head, a laptop hinge, or the folding mechanism in a foldable smartphone.
The distinction matters enormously for design rules. Dynamic applications need much larger bend radii and more conservative trace routing to survive millions of flex cycles.
The IPC-2223 standard provides minimum bend radius ratios based on circuit construction and application type:
Where r is the minimum bend radius and h is the total thickness of the flexible portion.
Let’s calculate the minimum bend radius for a double-layer dynamic flex circuit with a total thickness of 0.2mm:
Minimum bend radius = (r/h ratio) × thickness Minimum bend radius = 150 × 0.2mm = 30mm
For a static application with the same construction: Minimum bend radius = 12 × 0.2mm = 2.4mm
This illustrates why understanding your application is critical—the dynamic requirement is over 12 times more demanding.
From experience, I recommend adding margin to calculated minimums:
Trace routing in flex circuits follows different rules than rigid boards. The copper must survive repeated stress without cracking, which changes how you approach layout.
Route traces perpendicular to the bend axis whenever possible. Traces running parallel to the bend experience alternating tension and compression with each flex cycle, accelerating fatigue failure. Perpendicular traces flex more evenly and last longer.
When you must route parallel to the bend, use curved transitions rather than running traces straight into the bend zone.
Sharp 90-degree corners and acute angles create stress concentration points where cracks initiate. Use curved traces with the largest radius your design allows. If you need direction changes in the bend zone, use 45-degree angles or smooth arcs rather than right angles.
I-beaming occurs when traces on opposite layers of a double-sided flex align directly over each other, creating a stiff beam that resists bending. The flex naturally wants to bend around this stiff section, concentrating stress at the edges.
Stagger traces on opposite layers so they don’t overlap in bend zones. This distributes stress more evenly across the cross-section and allows smoother flexing.
Wider traces resist cracking better than narrow traces, but they also reduce flexibility. In bend zones:
Add teardrops where traces connect to pads and vias. This gradual transition reduces stress concentration at the pad-to-trace interface. Most PCB design software includes teardrop generation tools—use them.
Vias create particular challenges in flex circuits. The plated barrel that creates electrical connectivity between layers also creates a rigid structure that can crack under flexing.
This is the most important via rule: never place vias in areas that will bend. The rigid via barrel can’t flex with the surrounding material, so stress concentrates at the via edges until the barrel cracks.
Place vias at least 20 mils (0.5mm) away from bend zones. If your design absolutely requires vias near bends, move them onto stiffeners where they’re mechanically supported.
For vias that must be placed in flexible areas (even static flex):
Unlike rigid boards, flex circuits cannot use resin-filled via-in-pad. The via barrel remains open, which causes solder wicking during assembly—solder flows down through the via instead of forming a proper joint.
If component density demands via-in-pad, either:
Rigid flex PCB design adds complexity but offers significant advantages for three-dimensional assemblies. The rigid sections support components while flexible sections enable folding, bending, and dynamic movement.
The layer structure differs between rigid and flex regions of the same board. Key principles:
Place flex layers in the stack-up center. This protects flexible materials during outer-layer processing and places flex layers near the neutral bend axis where stress is lowest.
Use even layer counts. Asymmetric constructions warp during lamination and create uneven stress distribution.
Maintain continuous flex layers. The flexible layers run continuously through the rigid sections—they’re not separate boards connected together.
Specify no-flow prepreg for rigid regions. Standard prepreg can flow onto flex areas during lamination, stiffening sections that need to remain flexible.
The boundary between rigid and flex sections requires careful attention. This transition zone experiences stress concentration as the stiff rigid section meets the flexible ribbon.
Keep components, traces, and vias away from transitions. Maintain at least 50-60 mils (1.27-1.52mm) clearance from the rigid-to-flex boundary.
Avoid sharp corners in the flex outline near transitions. Use radii at corners to prevent tear initiation.
Consider transition zone stiffeners that taper gradually rather than ending abruptly.
When designs require bend radii tighter than standard formulas allow, the bookbinder technique uses progressively longer flex layers. Like pages in a book spine, each layer extends slightly more than the one beneath it, allowing tight bends without overstressing any single layer.
This technique adds manufacturing complexity and cost but enables bend radius to thickness ratios below 6:1 when necessary.
Stiffeners add rigidity to specific areas of a flex circuit, supporting components, enabling ZIF connector insertion, or reinforcing attachment points.
