A global medical device company building a braided catheter shaft ran into a reflow problem that its heat shrink supplier, Zeus, later wrote up as a case study. After lamination, the braid was showing through the thermoplastic jacket and leaving blisters along the shaft. According to the case study, the heat shrink met its incoming-inspection specifications but recovered too tightly over the assembly during reflow, and the part was running at roughly 65% yield loss.
Zeus engineers worked through the recovery process with the customer using Six Sigma tools. The analysis pointed at two dimensions of the heat shrink itself, its recovered inside diameter and its expanded inside diameter. Zeus and the OEM settled on a larger recovered ID and a smaller expanded ID, which lowered the shrink ratio, and Zeus tightened the tolerance by half. Per the case study, yield loss on that part fell below 3%. Those are Zeus's figures for one customer's part, and a different stack would need its own numbers.
The useful lesson sits in where the fix landed. Zeus's account describes no change to the oven setpoint. The defect was set by the dimensions of a consumable sleeve that gets cut off and discarded after every shaft. A braided shaft's reflow window belongs to the whole stack, and teams that tune only temperature and speed are adjusting two of several inputs.
The stack, from the mandrel out
Reflow lamination turns separate layers into one shaft. A mandrel sets the lumen. Over it go a liner (often etched PTFE, sometimes with a thin tie layer), the braid or coil, and a thermoplastic jacket, which may be several tubes of different durometers butted end to end. A heat shrink sleeve, commonly FEP, goes over everything. Zeus's reflow and assembly page describes what happens next: as heat is applied, the sleeve shrinks and applies radial compression while the jacket melts through the braid and bonds to the etched liner. The sleeve comes off afterward, and the mandrel is pulled.
Quasar Medical's primer on reflow lamination frames the requirement plainly. The jacket has to get hot enough to flow fully through the braid without degrading the polymer, damaging the liner or changing the braid, and travel speed sets how long each section sees that heat. Quasar recommends developing the two together with designed experiments rather than adjusting one at a time. A Nordson MEDICAL white paper on PET heat shrink for reflow names the same pair, temperature and line speed, as the two most important process parameters, and describes hanging a weight on the end of the sleeve during reflow to get full shrinkage and lamination.
The geometry leaves very little room. Teleflex Medical OEM's design notes for braid-reinforced shafts build the wall from the liner, twice the braid wire diameter (because the wire crosses itself) and the jacket, with liner and jacket each typically limited to about 0.002 in. (0.05 mm). On a shaft like that, the jacket over a braid crossover is only a couple of thousandths thick before reflow pushes any of it into the gaps between wires.
Temperature, travel speed and pressure
The thermal targets overlap more than the data sheets make obvious. Zeus's recovery guidance gives FEP heat shrink a recovery temperature of 420 °F ± 50 °F (215 °C ± 10 °C), and notes that the part underneath can act as a heat sink and that FEP may change length during recovery. Jacket grades melt well apart from one another. In a distributor's copy of Arkema's Pebax MED data sheet, the softer 3533 grade melts at 144 °C and the 7233 grade at 174 °C. A shaft that steps from a soft distal segment to a stiff proximal one carries several melt points under one sleeve, so a single setpoint is a compromise, and some reflow machines are built to change travel speed along the length for that reason.
The braid can narrow the window too. In a 2019 paper in the Journal of Cardiovascular Magnetic Resonance, researchers built 6 Fr shafts with PFA-insulated stainless braid and reflowed Pebax layers over it at up to 190 °C on a Beahm Designs laminator. They report that the insulation was undisturbed, since PFA melts near 315 °C. Coated or insulated wire, polymer filaments and conductors embedded in the wall all carry their own temperature limits into the reflow recipe.
Pressure deserves as much attention as heat, and the Zeus case shows why. The sleeve's shrink ratio, wall thickness and recovered ID decide how hard it squeezes the melt. Zeus lists FEP and peelable FEP sleeves in ratios from 1.3:1 to 2.0:1, with wall tolerances of ±0.002 in. Quasar's primer puts the trade-off in one line: too little pressure can leave voids, and too much can push the braid through the jacket or thin the wall unevenly. An AP Technologies article on FEP heat shrink adds that the sleeve acts as a temporary mold, so wrinkles or wall-thickness variation in the sleeve transfer straight to the catheter surface, and that sleeves shorten during recovery and can pull away from joints and transitions.
What a hotter recipe did to one shaft
A detailed public look at temperature sensitivity comes from a 2021 master's thesis by Cathal Merz at Munster Technological University. The device, identified only as Cat. A, was the proximal shaft of a neurovascular catheter: a PTFE liner film-cast on a silver-plated copper mandrel, a single-helix nitinol ribbon reinforcement, multi-segment outer polymer layers and an FEP sleeve, laminated first horizontally and then on a vertical laminator with a weight hanging from the proximal end.
The manufacturer's root-cause work on proximal-shaft fractures found that raising the lamination temperature increased the fracture rate, and repeating that part of the study gave the same result. The thesis built its experiments around low, nominal and high settings of 220.56, 232.22 and 243.89 °C (round numbers in Fahrenheit, 429, 450 and 471 °F), with the low setting chosen at the outer layer's melting point to keep adhesion to the reinforcement. A plastics engineering consultant examined fractured high-temperature samples under SEM and found cracks starting at the interface between the reinforcement and the outer layer. The examination found no evidence of thermal degradation, which had been the first suspicion, so the mechanism stayed open. Micro-CT of a sample tested to kink showed the layers delaminated from one another.
Two other details carry over to many programs. The manufacturer's original tests were manual, "kink by hand" and a wrap around a mandrel, and produced pass/fail data that could not show how much a process change moved the shaft. The thesis also points out that ISO 10555-1 sets tensile requirements but no standard method for other shaft mechanical properties, which leaves teams to build their own bend and kink fixtures.
Defects, and where they usually start
Pulling the sources together, the common reflow defects trace back to fairly specific parts of the stack:
- Voids and bubbles. Quasar attributes them to trapped air or moisture in the jacket, and a patent on dual-braided shafts originally assigned to St. Jude Medical notes that reflow may leave small air pockets between a coil and the polymer. Arkema's processing guidelines call for drying Pebax pellets left open more than two hours, and its 4533 SA 01 data sheet lists 0.4% water absorption at 23 °C and 50% RH. That guidance is written for pellets before extrusion; whether finished jacket tubing needs a bake before reflow is a question a short sectioning study can settle for a given grade and storage condition.
- Braid show-through and protruding wire. Too much sleeve pressure, as in the Zeus case, or wire ends pushed outward, per Quasar.
- Wall thinning and OD variation. Quasar notes that wall variation changes both outer diameter and local stiffness. With the thin jacket allowances in Teleflex's arithmetic, the crossovers are the place to measure.
- Delamination between liner and jacket. Quasar links it to liner preparation or insufficient heat. Zeus has described separation of the etched liner from the jacket as a recurring issue across the industry, one typically not detected until final testing, and it reports that its Tie Layer coating gave up to 2.5 times higher bond strength in its own tests. That is a vendor figure, best checked on your own liner, jacket and recipe.
- Weak or uneven joints at durometer transitions, also on Quasar's list, and the place where sleeve shortening and multiple melt points meet.
Catching them before verification does
Cross-sectioning remains the development workhorse, but it only shows the planes you cut. An abstract from Baxter in Microscopy and Microanalysis makes the case for micro-CT on device work: sectioning can alter or destroy small internal geometry, while CT reconstructs a volume that can be sliced anywhere, with a scan and reconstruction taking about an hour in many cases. For a reflow study, that means voids, wire position and wall thickness over crossovers can be mapped along a length instead of at a few cuts.
Mechanical testing then has to be sensitive to the process, not only to the design. FDA's PTCA catheter special controls guidance is a useful template even outside coronary devices. It recommends bond-strength testing wherever thermal fusion or other joining methods are used, after preconditioning the catheter through a tortuous-path fixture, along with kink testing around successively smaller mandrels and torque testing with the distal tip held. ISO 10555-1:2023 adds tensile testing of each section and junction, with minimum peak forces that rise with outside diameter, up to 15 N at 1.85 mm and above. The Merz thesis supports the preconditioning step from the other direction: bending samples around a mandrel before testing produced significant delamination in Cat. A, damage an unconditioned sample would not show. The thesis also compared shafts from an R&D laminator with shafts from a production laminator, a comparison worth writing into any transfer plan.
Put together, a workable order looks like this. Fix the sleeve dimensions and tolerance as part of the design, not as a purchasing detail. Run temperature and speed as a designed experiment with the edges of the window built and tested, not just the center. Section or scan at the crossovers and transitions. Precondition before pulling, bending and twisting. A reflow recipe that has only ever been run at its nominal setpoint hasn't been characterized, only demonstrated, and the Cat. A history shows how long a gap like that can stay hidden.
Sources
- Zeus, "Cost Saving Catheter Construction Solutions" (case study)
- Zeus, "Catheter Reflow and Assembly Aids"
- Zeus, "Heat Shrink Recovery Process"
- Matt Allen, "Tie Layer: A breakthrough solution to catheter failure," Zeus, March 28, 2019
- Quasar Medical, "Catheter Reflow Lamination: How Liners, Braid, and Jackets Become One Shaft"
- Nordson MEDICAL, "PET Heat Shrink for Reflow Applications" (white paper)
- AP Technologies, "The Role of FEP Heat Shrink Tubing in Catheter Assembly"
- Teleflex Medical OEM, "Design Information for Braid-Reinforced Catheter Shafts"
- Arkema, Pebax processing guidelines
- Arkema, Pebax 4533 SA 01 technical data sheet
- Arkema Pebax MED data sheet (distributor copy, NCC)
- Cathal Tadhg Merz, "Neurovascular Catheter Measurement System Development for Process and Design Effect Evaluation," master's thesis, Munster Technological University, 2021
- "A cardiovascular magnetic resonance (CMR) safe metal braided catheter design for interventional CMR at 1.5 T," Journal of Cardiovascular Magnetic Resonance, 2019
- US Patent 8,182,466 B2, "Dual braided catheter shaft"
- "Digital Radiography/Computed Tomography of Medical Devices," Microscopy and Microanalysis, 2017
- FDA, "Certain Percutaneous Transluminal Coronary Angioplasty (PTCA) Catheters — Class II Special Controls Guidance," Sept. 8, 2010
- ISO 10555-1:2023, "Intravascular catheters — Sterile and single-use catheters — Part 1: General requirements" (sample pages)
This article is put together on a best-effort basis from publicly available information, including company websites, press releases, and trade publications, as of the date of publication. Specifications, certifications, products, and company details can change, so confirm anything important directly with the supplier. Opinions are the author’s own. Mentions of a company, product, or marketplace are not endorsements or sponsorships, and nothing here is professional, engineering, regulatory, or purchasing advice.