Rigid-Flex PCB Design Guidelines: Material Selection, Stack-Up, and IPC Standards

There’s a specific moment in product development when rigid-flex PCB becomes the obvious answer: you’ve mapped out the assembly, the connector count is climbing, the harness routing is a disaster, and the mechanical envelope simply won’t accommodate two separate boards. A rigid-flex design solves all three problems simultaneously — and then creates a new set of problems for the engineer who doesn’t know the material system, the stack-up rules, or which IPC standards govern what.

This guide covers the essential rigid-flex PCB design guide material selection decisions, explains the stack-up architecture that determines bend life and signal integrity, and maps the IPC standard landscape so you know exactly what document governs each aspect of your design. Written from the production floor perspective — because the decisions that matter most are the ones your fabricator wishes you’d made correctly the first time.

What Rigid-Flex PCB Architecture Actually Means

A rigid-flex PCB is an integrated circuit board that combines rigid sections — where components are mounted and connectors are anchored — with flexible polyimide sections that allow the assembly to bend, fold, or flex either during installation or throughout its operational life. The rigid and flex sections are fabricated as a single laminated structure, not bonded together post-fabrication. This single-structure approach is what gives rigid-flex its reliability advantage over wire harness replacements and separate board-plus-connector assemblies.

The advantages are concrete: fewer interconnect points means fewer failure modes. In high-vibration environments — automotive, aerospace, industrial — every connector is a potential intermittent fault. Rigid-flex eliminates those connectors and replaces them with continuous copper, eliminating the failure mode entirely. Compared with separate rigid boards connected by wires or connectors, rigid-flex designs can reduce interconnect count, save space, lower weight, and eliminate common mechanical failure points.

The design complexity that comes with this architecture is real. The material system is more complex than FR-4 alone, the stack-up requires decisions that don’t arise in rigid-only work, and the IPC standard coverage spans multiple documents that must be understood together. Getting these right before tape-out is the difference between a first-article success and an expensive re-spin.

Rigid-Flex PCB Material Selection: Building the Stack from First Principles

The most important frame for rigid-flex PCB design guide material selection is that a rigid-flex board is not one material — it is a system of materials that must be mechanically compatible, thermally compatible, and electrically coordinated across the full stack. Each material choice in the flex section propagates consequences into the rigid section and vice versa.

Flexible Dielectric: Polyimide Is the Default — Know Why

Polyimide (PI) is the standard flexible dielectric material in rigid-flex designs and has been for decades. Its dominance comes from a combination of properties that no competing material fully matches: thermal stability to 260°C (surviving multiple lead-free reflow cycles), excellent chemical resistance, mechanical toughness under repeated bending, and a well-understood manufacturing process that every qualified flex fabricator supports.

Polyimide comes in two configurations for flex circuits. Adhesive-based construction bonds the copper foil to the polyimide core using acrylic or epoxy adhesive. This is the traditional approach and still widely used for cost-sensitive, lower-performance applications. Adhesiveless construction bonds copper directly to the polyimide through sputtering or casting processes, eliminating the adhesive layer entirely. IPC-2223 specifically calls out the use of adhesiveless flex cores for high-reliability designs, because the adhesive layer — with its high coefficient of thermal expansion — is a source of via hole reliability problems when it is present throughout the rigid sections of the build. 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.

For rigid-flex designs entering aerospace, medical, or mil-spec programs, adhesiveless polyimide is the baseline specification. For commercial or industrial designs where the flex section bends fewer than 100 times in its lifetime (static or bend-to-install applications), adhesive-based construction is cost-appropriate.

Polyester (PET) is sometimes cited as a lower-cost alternative to polyimide. It is not appropriate for any design involving soldering — PET’s maximum operating temperature of approximately 120°C rules it out for reflow assembly. It is limited to cold-bonded or mechanical-attachment applications. In a rigid-flex design where soldering is the assembly method, PET is not a viable flex dielectric.

Liquid Crystal Polymer (LCP) is the material of choice for high-frequency rigid-flex applications where signal loss at frequencies above 10 GHz is a design constraint. LCP offers a dielectric constant around 2.9 and exceptionally low moisture absorption (less than 0.04%), which keeps Dk stable across varying humidity conditions — a critical property for antenna and mmWave applications. For most rigid-flex designs below 10 GHz, standard polyimide handles signal integrity requirements adequately.

Rigid Section Materials: FR-4 and Beyond

The rigid sections of a rigid-flex design use laminate materials essentially identical to standalone rigid PCB work. Standard FR-4 is the default for most commercial and industrial applications. For designs that need halogen-free compliance, FR-4.1 or FR-15.1 materials are the equivalents under IEC 61249-2-21 compliance.

For high-speed digital applications above 5–10 Gbps, the rigid sections may require lower-loss laminates — Isola FR408HR, Panasonic Megtron 6, or similar. These should be selected using the same Dk/Df criteria as any high-speed rigid design, with the additional constraint that the material must laminate compatibly with the prepreg being used to bond the flex cores into the rigid section.

Ventec PCB materials cover both sides of the rigid-flex stack-up. The VT-47PP NF/LF and VT-47PP NF/LF Flex Rigid products are specifically designed for rigid-flex applications, combining lead-free compatible FR-4 prepreg with flex-compatible processing parameters. The VT-901PP NF/LF LCTE and VT-901PP NF/LF LCTE Flex Rigid polyimide prepregs serve the same function for polyimide-based rigid-flex constructions requiring the full Tg-250°C polyimide performance throughout the build.

Copper Foil: The Specification That Determines Bend Life

The copper foil specification in the flex section is not a procurement detail — it is a reliability determinant. The choice between rolled annealed (RA) copper and electrodeposited (ED) copper in the flex region determines whether the design survives its intended bend cycle life.

Rolled Annealed (RA) copper is manufactured by rolling copper ingots into thin foil and then annealing to relieve internal stress. This process aligns the grain structure of the copper longitudinally, producing a material with elongation before failure of 20–45% — the ductility that allows it to survive repeated bending without cracking. RA copper can withstand millions of flex cycles when the design is correct. For dynamic flex applications — any design where the flex section bends repeatedly during service life — RA copper is non-negotiable.

Electrodeposited (ED) copper is produced electrochemically with a columnar grain structure that is stiffer and more brittle under repeated stress. Its elongation before failure is 4–10% — adequate for rigid boards and for static bend-to-install flex applications, but insufficient for designs that require more than a few hundred bend cycles. Using ED copper in a dynamic flex design is not a cost saving — it is a scheduled field failure.

Coverlay vs. Flexible Solder Mask

In the flex sections, the standard solder mask used on rigid boards cannot be used. Standard liquid photoimageable solder mask cures to a rigid film that cracks immediately when the substrate bends. The two alternatives are coverlay and flexible solder mask.

Coverlay is a polyimide film with an acrylic or epoxy adhesive backing that is laminated over the copper traces. It maintains full flexibility, provides robust mechanical protection, and is the standard for any dynamic flex application. Coverlay openings are defined by a separate Gerber layer and are cut by mechanical routing or laser, not photolithographically like solder mask. Minimum coverlay to pad opening clearance is typically 0.5 mm; going tighter than this risks insulation breakdown or peeling.

Flexible solder mask is a photodefined alternative that costs less and allows finer pad opening tolerances than coverlay. It is acceptable for static flex sections and rigid sections of rigid-flex boards. For dynamic flex zones where the circuit bends repeatedly, flexible solder mask will eventually crack at high cycle counts. Standard practice: specify coverlay for all flex sections, flexible solder mask for rigid section coverage only.

Table 1: Rigid-Flex Material Selection Summary

Material Layer	Static Flex / Bend-to-Install	Dynamic Flex	Rigid Sections	Key Spec to Verify
Flex Dielectric	Polyimide (adhesive-based acceptable)	Polyimide adhesiveless	N/A — rigid laminate	IPC-4204 (flex substrate spec)
Rigid Laminate	FR-4 or high-Tg FR-4	Same	FR-4, FR-4.1, high-speed laminates	IPC-4101 slash sheet
Copper — Flex Zones	RA preferred; ED acceptable	RA copper — mandatory	ED copper acceptable	Elongation spec: RA ≥ 20%
Copper — Rigid Zones	ED or RA	ED or RA	Standard ED copper	Weight per design current
Flex Protection	Coverlay or flexible soldermask	Coverlay — mandatory	Standard solder mask	IPC-4203 (coverlay materials)
Bonding/Adhesive	Acrylic or epoxy prepreg	No-flow prepreg / adhesiveless	Standard FR-4 prepreg	No adhesive in rigid core zones
High Frequency	Standard PI adequate to 5–10 GHz	LCP for >10 GHz	Low-loss laminate as required	Dk/Df at operating frequency

Stack-Up Design: Where Rigid-Flex Gets Complicated

The stack-up of a rigid-flex PCB is more complex than a rigid multilayer because you are effectively designing two different cross-sections — one for the rigid zones and one for the flex zones — that must be laminated together in a single manufacturing cycle.

Core Stack-Up Principles

Center the flex layers in the stack-up. Positioning the flex cores symmetrically about the neutral axis of the rigid stack-up minimizes the bending moment that the flex copper must absorb. An asymmetric stack where flex cores are positioned toward one outer layer will cause the board to curl during lamination and impose residual stress on the flex conductors in service.

Maintain symmetry across the midplane. The rigid stack-up should be symmetric about its midplane — layer 1 to layer N should mirror the dielectric thickness, copper weight, and material type. Asymmetric rigid stack-ups warp during lamination. In rigid-only boards this is a known problem; in rigid-flex it is worse because the flex section cannot provide the same mechanical restraint as a symmetric rigid core.

Use no-flow prepreg at the flex-to-rigid transition. Standard FR-4 prepreg flows significantly during lamination press cycles. In the transition zone where flex and rigid sections meet, this flowing resin can penetrate the flex area and stiffen it — reducing bend capability and concentrating stress exactly where the design is most vulnerable. No-flow or low-flow prepreg (sometimes called “bondply” in flex circuit manufacturing) is specified for the lamination layers adjacent to flex cores to prevent this resin invasion.

Air gap construction for tight bend zones. When multiple flex layers must navigate a very tight bend radius, stacking them bonded together multiplies the effective thickness and dramatically reduces bend capability. Air gap (also called bookbinder or loose-leaf) construction leaves adjacent flex layers unbonded in the bend zone, allowing each layer to slide independently during bending. This allows the flex section to perform bends of 180° or more. However, bookbinding construction costs approximately 30% more than standard rigid-flex and requires explicit definition on the fabrication drawing.

The Transition Zone: Most Failure-Prone Location in Any Rigid-Flex Build

The rigid-to-flex transition zone is the most mechanically stressed location in a rigid-flex assembly and the area where the most manufacturing challenges arise. Several specific requirements apply here that don’t exist anywhere else in the stack-up.

Via keepout in the transition zone. Plated through-holes (PTHs) and vias located at or near the transition from rigid to flex are subjected to thermal expansion stresses during assembly and operation. The coverlay adhesive that terminates in the transition zone has a very high CTE — significantly higher than the surrounding FR-4 and copper. If vias are drilled through this adhesive zone, they are subjected to stresses from adhesive expansion and contraction that crack the barrel plating. IPC-2223 defines a minimum keepout distance for vias from the transition zone. This is not a suggestion — it is the boundary between via reliability and via cracking.

Anchor copper features in the transition. The copper traces exiting the flex zone and entering the rigid zone should fan out into pads or teardrops at the transition line to distribute stress over a wider area. Narrow traces running straight into the transition without pads or width changes concentrate stress precisely at the point where the material stiffness changes abruptly.

Table 2: Common Rigid-Flex Stack-Up Configurations

Configuration	Layer Count	Flex Layers	Rigid Zones	Typical Application	Approximate Cost Factor
1F-2R (simple)	2 rigid + 1 flex	1 flex core (2 Cu layers)	1 rigid section each end	Simple interconnects, wearables	1.5× rigid
2F-2R	4 rigid + 2 flex	2 flex cores	1 rigid each end	Compact camera modules, medical	2–2.5× rigid
4R-2F multilayer	6–8 total	2 flex cores (center)	Multiple rigid zones	Aerospace, server backplane	3–5× rigid
Bookbinder / air-gap	6–10 total	3+ flex cores unbonded	Multiple rigid zones	High-cycle dynamic, military	+30% vs. standard RF
Hybrid (RF)	4–8 total	LCP flex cores	Low-loss rigid laminate	5G, radar, high-frequency RF	4–6× rigid

Bend Radius Rules: The Design Rule That Determines Reliability

The minimum bend radius calculation is the most critical dimensional constraint in any flex or rigid-flex design. Violating it doesn’t cause visible defects on first article — it causes fatigue cracking in the field after a product has been in service for months.

IPC-2223 is the authoritative standard for bend radius requirements and distinguishes sharply between two usage categories that have completely different design requirements:

Use A — Static Flex (bend-to-install): The circuit bends fewer than 100 times in its lifetime. The flex is used to route around a corner during assembly and stays there. Minimum bend radius is typically 6× total flex thickness for single-layer, 12× for double-layer, increasing further with layer count.

Use B — Dynamic Flex (repeated bending): The circuit bends repeatedly during service — opening and closing of a hinge, rotation of a joint, movement of a printer head, wrist motion in a wearable. Minimum bend radius is dramatically larger: 100× total thickness for single-layer dynamic flex, 150× for double-layer. This is not a typo — a dynamic flex design has a minimum bend radius 15–25× larger than a static design of the same thickness. Dynamic flex applications must use RA copper, may require adhesiveless construction, and should be validated with physical bend cycle testing on prototypes before production commitment.

The consequence of getting this wrong is not a design that fails inspection — it is a design that passes all incoming inspection, ships to customers, and fails in the field after sufficient accumulated flex cycles. That failure mode is expensive, hard to diagnose, and entirely preventable.

Conductor Design in Bend Zones

Trace routing through flex zones follows rules that do not apply in rigid PCB work:

All traces in bend zones must run perpendicular to the bend axis — not parallel to it. A trace running parallel to the bend axis is being repeatedly stretched and compressed along its length every flex cycle. A trace running perpendicular to the bend axis bends with the substrate rather than being axially stressed.

No vias in bend zones. The via barrel creates a stress concentration — the stiff plated cylinder in a flexible substrate is the weakest point for crack initiation. Map bend zones at the start of design and enforce a strict via exclusion zone. Move vias to rigid sections or stiffened regions.

Stagger traces on opposite layers. In a double-layer flex zone, traces on the top copper and traces on the bottom copper should not be directly aligned over each other. Overlapping traces create local stiff regions that concentrate bending stress. Offsetting them distributes the stiffness more evenly.

No right-angle bends in trace routing. Traces should use curved transitions (minimum 3× trace width) rather than sharp corners. Sharp corners in the copper create stress concentrations at bend points.

IPC Standards Map for Rigid-Flex PCB Design

Understanding which IPC document covers which aspect of your design prevents the common mistake of applying rigid-board standards to flex-specific requirements.

Table 3: IPC Standards Reference for Rigid-Flex Design and Production

IPC Standard	Current Rev.	Scope	When You Need It
IPC-2223	Rev. E (2025)	Sectional design standard for flexible and rigid-flex circuits	Primary design reference — bend radius, conductor design, material specs, coverlay, stiffeners
IPC-2221	Rev. B	Generic PCB design standard	Board-level design rules applicable to rigid sections
IPC-6013	Rev. E (2021)	Qualification and performance for flex and rigid-flex PCBs	Acceptance criteria, performance classes, transition zone requirements
IPC-4101	Rev. E	Base laminate materials for rigid sections	Specifying rigid section laminates by slash sheet number
IPC-4204	Latest	Flexible base dielectric materials	Specifying polyimide flex substrates
IPC-4203	Latest	Coverlay materials	Specifying coverlay film and adhesive
IPC-A-600	Rev. J	Acceptability criteria for printed circuit boards	Visual inspection reference for flex and rigid sections
IPC-7711/7721	Latest	Rework and repair	Rework procedures specific to flex circuits
IPC-6012	Rev. E	Qualification for rigid boards	Applies to rigid sections of the build independently

IPC-6013 Performance Classes

IPC-6013 uses the same three-class performance structure as IPC-6011 and IPC-6012, but with acceptance criteria tailored to flexible substrates:

Class 1 (General electronics): Minimum performance, limited visual acceptance requirements. Very rarely specified for flex circuits — the cost difference between Class 1 and Class 2 is minimal in flex manufacturing, and most customers require at least Class 2.

Class 2 (Dedicated service electronics): Standard commercial and industrial requirement. Covers the majority of consumer electronics, industrial, and telecommunications rigid-flex applications.

Class 3 (High-performance electronics): Aerospace, military, medical implant, and life-safety applications. Tighter visual acceptance, 100% electrical testing, additional coupon requirements, and more stringent PTH plating specifications.

Rigid-Flex Design Checklist: What to Define Before Sending to Fabrication

Before releasing a rigid-flex design to fabrication, the following items must be explicitly defined on the fabrication drawing and stack-up notes — not assumed or left to the fabricator’s discretion:

The application usage type: Use A (static/bend-to-install) or Use B (dynamic). The target bend cycles must be stated for Use B designs. The minimum bend radius at each flex zone, confirmed against IPC-2223 for the specific layer count and copper type. The copper foil type in each flex zone: RA or ED. For dynamic applications, RA must be explicitly called out. Whether construction is adhesive-based or adhesiveless in flex cores. Coverlay specification on all flex section outer copper layers. No-flow prepreg at all flex-to-rigid transition lamination interfaces. Via keepout dimensions from all transition zones. Air gap or bookbinder construction where multiple flex layers navigate the same tight bend.

Useful Resources and Standards Downloads

Resource	URL / Source
IPC-2223E Sectional Design Standard for Flex/Rigid-Flex	ipc.org (purchase required)
IPC-6013E Qualification for Flex and Rigid-Flex Boards	ipc.org (purchase required)
IPC-4101E Base Laminate Materials Standard	ipc.org
IPC-4204 Flexible Base Dielectric Materials	ipc.org
IPC-4203 Coverlay Materials Standard	ipc.org
Ventec Rigid-Flex Compatible Laminate Range	ventec-group.com/products
Ventec VT-47PP NF/LF Flex Rigid Datasheet	ventec-group.com/products/lead-free-assembly/vt-47
Ventec VT-901PP Polyimide Flex Rigid Datasheet	ventec-group.com/products/polyimide/vt-901
Sierra Circuits Flex PCB Design Guidelines	protoexpress.com/blog/flex-pcb-design-guidelines
PCBSync IPC-2223 Explained	pcbsync.com/ipc-2223
PCBSync IPC-6013 Complete Guide	pcbsync.com/ipc-6013
PCBSync Flex PCB Design Guide	pcbsync.com/flex-pcb-design
Würth Elektronik Flex Solutions Design Guide (PDF)	we-online.com (free download)
Altium Rigid-Flex PCB Design Guidelines (PDF)	resources.altium.com (free download)

5 FAQs About Rigid-Flex PCB Design

Q1. When should I use rolled annealed copper instead of electrodeposited copper, and is there a cost impact?

RA copper is mandatory for dynamic flex applications — any design where the flex section bends more than a few hundred times in service. The grain structure difference is fundamental: RA copper has 20–45% elongation before failure versus 4–10% for ED copper. In dynamic use, ED copper will develop fatigue cracks at the copper grain boundaries under repeated stress and eventually open. The cost premium for RA copper over ED copper is modest — typically 10–20% at the material level — and trivially small compared to the cost of a field failure investigation and product recall. For static bend-to-install applications, ED copper is acceptable and the cost savings are real. When in doubt about whether your application is static or dynamic, classify it as dynamic and specify RA.

Q2. What’s the practical difference between a static and dynamic flex application, and why does it change the design so dramatically?

A static (Use A) flex circuit bends once or a handful of times — typically during installation — and remains in that bent position for its service life. The minimum bend radius for a single-layer static flex is 6× the flex thickness. A dynamic (Use B) flex circuit bends repeatedly during operation — the hinge of a laptop, the head carriage of a printer, the wrist joint of a robot, the fold mechanism of a foldable phone. The minimum bend radius for a single-layer dynamic flex jumps to 100× the flex thickness. That’s more than 16× larger for the same material construction. Beyond bend radius, dynamic designs must use RA copper, adhesiveless construction is strongly preferred, vias must be completely absent from bend zones, and the design should be validated with accelerated bend cycle testing. Under-specifying dynamic bend requirements is the most common cause of rigid-flex field failures.

Q3. Why can’t I just use standard solder mask on flex sections?

Standard liquid photoimageable solder mask cures to a rigid polymer film that has very little elongation before cracking — typically less than 5%. When the substrate bends, the solder mask cracks at or near the flex radius, creating exposed copper that is vulnerable to oxidation, moisture ingress, and shorting. Coverlay — a polyimide film with adhesive backing — is the correct protection for flex sections because the polyimide film maintains flexibility through the same bend range as the substrate. Flexible solder mask formulations exist and can handle a limited number of bend cycles on static applications, but their long-term performance in dynamic flex is significantly worse than coverlay. The rule is simple: coverlay in flex zones, standard solder mask in rigid zones.

Q4. My rigid-flex has a 0201 component very close to the transition zone. What are the risks and how do I mitigate them?

The transition zone is the highest-stress location in any rigid-flex assembly. Thermal cycling, vibration, and the mechanical mismatch between rigid and flex materials all concentrate stress at this boundary. A small SMD component placed immediately adjacent to the transition zone will be exposed to higher shear stress on its solder joints than the same component placed in the center of a rigid section. For non-critical components, maintaining a minimum clearance of 2–3 mm from the transition zone is good practice. For BGAs, fine-pitch connectors, or any component where solder joint failure is safety-critical, increase the clearance to 5 mm or more, or consider routing signals through the transition and placing the component farther into the rigid section. Your fabricator’s DFM review should flag components too close to transition zones — this is exactly the kind of issue that is cheaply fixed at layout but expensive to discover after first article testing.

Q5. Can I use rigid-flex for a high-speed digital design with SerDes lanes above 10 Gbps?

Yes, with the right material selection and routing discipline. The flex sections introduce discontinuities that a rigid-only design doesn’t have — the dielectric changes at the transition zones, and the reference plane geometry may differ between rigid and flex zones. For channels below 10 Gbps on standard polyimide flex, the signal integrity impact is manageable with careful impedance control and continuous reference planes through flex zones. Above 10 Gbps, the challenges compound: standard polyimide’s moisture absorption shifts Dk in humidity-varying environments, and the glass weave effects from the rigid section can generate skew on differential pairs. For 10–28 Gbps channels in rigid-flex designs, specify controlled-impedance flex construction, maintain continuous ground planes through the flex section, use LCP for the flex dielectric if frequencies exceed 10 GHz, and run full channel simulation with material-accurate flex section models before committing to fabrication.

Conclusion: Rigid-Flex Success Is Front-Loaded

The decisions that determine whether a rigid-flex PCB succeeds or fails are almost entirely made before fabrication starts: material type, copper foil specification, bend radius calculation, via keepout enforcement, stack-up symmetry, transition zone design, and standards classification. None of these are complex calculations, but all of them are non-obvious to engineers coming from a rigid-board background, and all of them have consequences that don’t show up until the product is in the field.

Rigid-flex PCB design guide material selection in practice means building the stack-up as a complete material system — not choosing a rigid laminate and a flex substrate independently. The polyimide grade, the copper foil type, the prepreg at the transition, the coverlay, the adhesiveless construction in dynamic zones — these choices interact, and they must all be right simultaneously for the design to meet its reliability target.

Work with your fabricator during stack-up development, not after layout is locked. The fabricator’s DFM review on a rigid-flex design covers constraints that simulation tools don’t model. Get the stack-up confirmed before routing, calculate bend radii with IPC-2223 margin before placing components near flex zones, and document the application usage class on the fabrication drawing explicitly. The board that makes it through qualification is the one where every one of these decisions was made deliberately.