Still Extruding Thin, Soft TPU? Build the Wall Instead.
Where the die runs out
Soft thermoplastic polyurethane (TPU) can be extruded. In the 60A to 80A range, though, extrusion loses its precision tools, and its working window narrows to small, moderately thick, straight tubes. Of more than 400 extruded polyurethane tubes listed on Chamfr, the best 80A tube has an outside diameter (OD) only 64 times its wall. No polyurethane tube of any hardness is listed above 0.43 in (10.9 mm) OD [5].
Microlayer works outside that window. Medibrane already lists 40A tubing with a 0.001 in (0.025 mm) wall on a 0.111 in (2.82 mm) OD, a ratio of 111. The same process builds covers on frames, braids and shaped mandrels that no die can make.
This article sets out what the resin makers and the market say about soft TPU extrusion, the five places where microlayer wins, real device requirements on each side of the line, and the parts that still belong to extruders.
Why soft TPU extrusion runs out
The limit is not a single wall number. It is a chain of process restrictions, and the resin makers describe them themselves.
1. No sizing tools. One resin maker’s TPU extrusion guide states: “Do not use sizing rings to control O.D. of TPU tubing on materials softer than 50-55D” [1]. Another describes contact calibration as not applicable to TPU because of its friction [2]. A 60A to 80A tube is therefore free-extruded, with its roundness set by internal air, water level and the gap to the quench.
2. Little drawdown. The same guide notes that TPUs “do not draw down well compared to other thermoplastics like polyethylene” and cuts tubing dies only 20 to 30% oversize [1]. A very thin soft wall needs a die gap close to the finished wall, so any pin-to-die offset appears directly as wall variation.
3. Low melt strength. Freshly extruded TPU tubing is prone to distortion and needs supporting air to avoid collapse [2]. The larger the diameter and the thinner the wall, the harder this becomes.
4. Tack and shrinkage. Plastics Technology reports that low-durometer medical polyurethanes around 80A are very tacky and shrink at variable rates, which makes inline cutting to length “extremely difficult”. The usual workaround is to rack the tubes for 24 hours and cut them by hand [3].
One medical tubing extruder has described the corner plainly: as diameters increase and wall thickness and durometer decrease, “the extrusion process can become unstable” [4].
What the market actually sells
On 15 October 2026 we analysed 473 polyurethane tubing listings on Chamfr, 409 of them extruded single-lumen or balloon tubes, and computed OD divided by wall for each [5]. The best ratio rises with hardness, which is what the loss of sizing tools predicts.
Durometer | Best extruded OD/wall | That tube (OD × wall) | Median extruded | Microlayer listed |
60A and softer | 34 | 0.092 × 0.00275 in (2.34 × 0.070 mm), 30A | 13 | 111 at 40A: 0.111 × 0.001 in (2.82 × 0.025 mm) |
61A to 70A | 56 | 0.0985 × 0.00175 in (2.50 × 0.044 mm) | 29 | Not listed |
71A to 80A | 64 | 0.191 × 0.003 in (4.85 × 0.076 mm) | 11 | 96 at 80A: 0.096 × 0.001 in (2.44 × 0.025 mm) |
81A to 90A | about 100 | 0.196 × 0.002 in (4.98 × 0.051 mm), 85A | 13 | Not listed |
55D to 75D | 149 | 0.149 × 0.001 in (3.78 × 0.025 mm) | 20 | 112 at 50D |
Three further limits show in the data:
Thin soft walls stop at small diameters. Walls of 0.003 in (0.076 mm) or less stop at about 0.10 in (2.5 mm) OD for 70A and softer, and at 0.19 in (4.9 mm) OD for 80A.
Large soft tubes have thick walls. Above 0.25 in (6.35 mm) OD, the thinnest soft wall listed is 0.006 in (0.15 mm).
There is a size ceiling. Nothing is listed above 0.43 in (10.9 mm) OD, in any hardness.
Microlayer is not ahead everywhere. At 50D and harder, extruders reach 0.001 in walls on larger diameters than Medibrane lists, and those parts belong to them.
How microlayer works
A TPU solution is atomised and sprayed onto a rotating mandrel in many thin passes, each drying before the next. The wall is built, not drawn, and three things follow:
The mandrel sets the inside. The inner diameter and shape copy a ground, measured tool, not a die gap, an air pressure and a cooling bath.
The pass count sets the wall. Thickness is programmed, and it can change along the part.
Anything can sit under the spray. A nitinol frame, a braid, a coil, ePTFE, a textile or an earlier polymer layer becomes part of the wall, bonded without sutures.
Because the polymer goes on from solution, there is no melt-strength or sizing limit on how soft the grade can be.
Five places microlayer beats extrusion
1. Thin walls on large diameters in soft grades. The mandrel supports the wall throughout, so there is no sag, collapse or ovality to fight. Ratios of 100 and above are routine, against 34 to 64 for the best soft extrusions.
2. Very soft grades. From solution, 40A is as easy to build as 80A. Below 60A, extrusion offers almost nothing at thin walls.
3. Shapes that are not tubes. Tapers, flares, funnels, cones, cups, closed ends and leaflets are all mandrel shapes. Thickness can also vary along the part: thicker at a landing zone, thinner across a hinge.
4. Encapsulation and sutureless lamination. A frame, braid, coil, ePTFE layer or textile on the mandrel becomes part of the wall, and one structure can stack different materials and hardnesses. The interface is what has to be qualified, so peel and cyclic-fatigue testing is part of every development.
5. Even thickness on 3D shapes. Dip coating, the other seamless solvent route, drains under gravity and pools at concave corners and frame junctions [7]. A published spray study of polyurethane heart-valve leaflets held the thickness spread below 12.5 µm on a 150 µm wall, and the most durable valves came from the tightest spread [6].
Delivery systems: where the polymer has to move with a frame
Catheter shafts are well served by extrusion. The opportunity is at the working end, where the polymer has to expand, fold and resheath with a metal structure.
Feature | Published requirement | Why a die cannot make it | Microlayer capability used |
Funnel tip on an aspiration catheter [8] | Non-permeable elastomeric PU or ePTFE cover on a nitinol braid funnel that opens from 2 mm to about 15 mm and can be resheathed | A cone on a braid, at very high strain | Braid encapsulation, shaped mandrel, soft grade |
Embolic protection filter [9, 10] | PU membrane with 110 to 140 µm pores, as a cone or closed bag on a nitinol loop, for vessels of 3.5 to 15 mm | A cone or closed bag on a frame | Solid membrane on a shaped mandrel, then laser-drilled pores |
Recapture or loading sleeve on a valve delivery system [11] | Compliant film folded over the end of a funnel frame | Film bonded to a 3D frame | Sutureless lamination onto the frame |
Encapsulated pump cannula [12] | 22 Fr (7.3 mm) nitinol cannula covered in PU, pre-curved, with windows | Reinforcement inside the wall, plus a curve and windows | Frame encapsulation on a curved mandrel |
Ultra-soft thin sleeves for tips and balloons | 0.001 in (0.025 mm) wall at 40A | Below the soft extruded envelope (best 34:1) | Listed 40A microlayer tubing at 111:1 |
Implants: thin covers at ratios in the hundreds
Implant covers run from about 20 to 160 µm on diameters of 8 mm and up. That puts OD/wall in the hundreds, three to six times the best soft extruded tube.
Component | Published requirement | OD/wall | Why a die cannot make it | Microlayer capability used |
Covered biliary stent [13] | PU film of 20 to 40 µm (0.0008 to 0.0016 in) within a PU and silicone cover on an 8 mm stent; PU above 60 µm enlarges the stent | 200 to 400 | Three to six times the 80A record, bonded to a frame | Thin wall on a frame, multi-material stack |
Covered esophageal stent [14] | PU membrane on a 12 to 22 mm stent with flared ends and a silicone-coated lumen | Not published | Larger than any PU tube listed; flares; encapsulation | Large shaped mandrel, encapsulation |
Polymer heart-valve leaflets [6, 15] | 80 to 160 µm on a 23 mm frame; durability tracked a thickness spread under 10 to 12.5 µm | About 145 to 290 | A 3D leaflet bonded to a frame | Even thickness on a 3D mandrel, frame bonding |
Covered stents and stent grafts | Often fabric sewn to the frame, with a needle hole at every stitch | Not applicable | An attachment problem, not a tube problem | Sutureless lamination along every strut |
The last row is our engineering view rather than a published polyurethane precedent. Many covers on the market are still sewn fabric, and lamination replaces the stitches with a continuous bond.
The next frontier: steerable large-bore sheaths
Published steerable guides and sheaths at 12 to 24F carry walls of about 0.5 to 1.3 mm, an OD/wall of only 6 to 10 [18 to 22]. The deflectable section is a stack of PTFE liner, steel braid or coil, pull wires in their own lumens and segmented polyether block amide (PEBA), reflowed under FEP heat-shrink at about 190 to 220 °C [23]. One published mitral guide stacks two braids and two PEBA layers, the outer one 0.0125 in thick, over a nylon layer carrying the pull-wire lumens [24].
That thickness is there for a reason. Bending strain is large (roughly 27% at the surface of a 24F catheter bent to a 15 mm radius), the polymer carries compression and resists kinking [26], and the inner and outer arcs of a full curl can differ by up to 0.50 in [27]. Device reports for large-bore steerable guides list breaks, torn material, kinks and liner delamination [30].
Microlayer could change that trade-off where a metal spine carries the load:
Thin jacket on a spine. Patents already pair slotted nitinol with a PEBA cover only 0.002 in thick [25]. Microlayer can build 0.002 to 0.003 in of 60A to 80A TPU directly on the spine, sealing the slots and bonding to the metal, with no reflow heat on the liner.
Bore or profile. Replacing a 0.0125 in outer jacket with 0.003 in saves about 0.019 in (0.48 mm) on the diameter, roughly 1.4F, provided the spine takes over the jacket’s structural role. This is our engineering estimate, not a published result.
Pull-wire lumens inside the wall. The lumen tube is laid on the first layers and sprayed over instead of being buried by reflow.
Graded transitions. Wall and hardness can change gradually along the deflectable section instead of at butt joints. This too is our engineering view, not yet proven on a steerable shaft.
There is precedent at the edges. One deflectable-sheath patent assembles its coils by dipping or spraying before moulding in urethane [28], and a hypotube patent dip-coats 65A to 75D polyurethane to encapsulate laser-cut slots [29]. We found no commercial large-bore steerable sheath with a sprayed TPU jacket. Proving the route will require data on deflection force against bend radius, ovality at full deflection, kink resistance, cyclic-deflection fatigue and bond strength to the shaft.
Where extrusion stays the right answer
Plain catheter shaft jackets sit well inside the soft extrusion window, and Medibrane refers that work to extruders.
Microcatheter distal jackets run about 0.0032 to 0.0037 in (0.08 to 0.095 mm) of 80A to 90A polyurethane on a 0.031 to 0.035 in (0.8 to 0.9 mm) OD, an OD/wall of about 10. They are extruded and reflowed over the braid [16].
Expandable sheath jackets are published at about 0.005 in (0.13 mm) on a 0.260 in (6.6 mm) OD, a ratio of about 52, inside the 80A envelope [17].
Simple tubes, such as a recent request for a 0.092 in OD polyurethane tube with a 0.014 in wall (a ratio of about 7), are cheaper and faster to extrude. We told the customer so.
Long continuous lengths and high-volume commodity tubing also favour extrusion. Microlayer builds part by part on a mandrel, so it earns its cost where the part needs something a die cannot give.
Quick screen: is your part a microlayer part?
A soft TPU part is a microlayer candidate if any one of these is true:
It is not a constant-section tube: a funnel, cone, cup, flare, closed end or leaflet, or a polymer that has to encapsulate a frame, braid, coil, ePTFE or textile.
Its OD is above 0.43 in (11 mm), or above 0.25 in (6.35 mm) with a wall under 0.006 in (0.15 mm).
Its wall is under 0.002 in (0.05 mm) at 80A or softer.
Its OD/wall is above the best extruded tube for its hardness: 34 at 60A and softer, 56 at 61A to 70A, 64 at 71A to 80A, about 100 at 81A to 90A.
It stacks materials or hardnesses that must bond without sutures.
If none applies, and especially for a long straight tube at 55D and harder or a soft tube under about 25:1, an extruder is the better partner. These thresholds come from what is sold today, not from hard physical limits, so a part near the line is worth a conversation.
Send us a drawing or a sketch of your cover, sleeve, membrane or tip. We will tell you plainly whether microlayer is the right process, and if it is not, which one is.
Medibrane develops and manufactures microlayer polymer walls, covers and encapsulations for catheters and implants as an FDA-registered, ISO 13485:2016 certified CDMO. To discuss your part, book a consultation.
References
Lubrizol. TPU Extrusion Processing Guide. https://lubrizol.com/-/media/Lubrizol/Engineered-Polymers/Documents/EP-Literature/24-0003580-EP-ExtrusionGuide-LTR-SPREADS.pdf
BASF. Elastollan processing recommendations.
Plastics Technology. Inline cutting and curing of TPU tubing. 2020.
PlasticsToday. Pipe and profile: Putnam pushes urethane tubing. https://www.plasticstoday.com/plastics-processing/pipe-and-profile-putnam-pushes-urethane-tubing
Chamfr. Polyurethane tubing category. Accessed 15 October 2026.
Coetzee. Spray-fabricated polymeric TAVI leaflets. University of Cape Town, 2019.
AorTech. US Patent 7,833,565 B2 (dip-coating thickness control).
Funnel aspiration catheter. US Patent 11,013,523 B2.
Embolic protection filter. FDA De Novo DEN160043.
Boston Scientific. Filter bag embolic protection system.
Recapture sleeve. European Patent EP 2 558 030 B1.
Abiomed. Impella RP Flex polyurethane-covered pump cannula.
Covered biliary stent. US Patent 10,080,640 B2.
Merit Medical. ALIMAXX-ES polyurethane-covered esophageal stent.
Leo HL, et al. Polymeric heart valve leaflets. Ann Biomed Eng. 2005.
Microcatheter jacket. US Patent 11,471,646 B2.
Expandable sheath. US Patent 10,327,896.
Abbott. MitraClip steerable guide catheter IFU (EL2123340).
FARADRIVE steerable sheath. FDA 510(k) K233248 and technical specifications.
Boston Scientific. WATCHMAN TruSteer access system brochure.
Medtronic. FlexCath Advance steerable sheath.
Integer. Destino Twist steerable introducer.
Steerable sheath reflow build. US Patent 10,493,708.
Mitral guide catheter layer stack. US Patent 7,666,204 B2.
Slotted nitinol spine with thin Pebax cover. US Patent 7,402,151.
Polymer matrix carrying compression in conventional shafts. US Patent 10,716,916.
Arc-length mismatch in deflecting sheaths. US Patent 8,376,991.
Dipped or sprayed coils in a deflectable sheath. US Patent 7,803,130.
Dip-coated polyurethane over laser-cut hypotube. US Patent 10,561,820.
US FDA. Medical device reports (openFDA).
Comments