Technical Report

Mechanics and Metrology of Teflon^® Encapsulated O-Rings

Non-Linear Force Dynamics in Composite Sealing Systems

A design and quality-control reference for engineers specifying, testing and installing fluoropolymer-encapsulated seals. This report explains why the sealing force of an encapsulated O-ring does not scale linearly with diameter, sets out the dual classification the part actually requires, provides a corrected method for estimating installation load, and defines a repeatable laboratory protocol for measuring force to compress.

M-Cor TR-001 Rev. BJuly 2026Public – unrestrictedPrepared by M-Cor Engineering

The three findings that matter most

• Shore hardness is not a valid metric for an encapsulated seal. The indenter penetrates further than the jacket is thick, so the reading describes neither the jacket nor the core. Force to compress at a stated deflection is the only defensible specification.
• Force does not scale linearly with diameter. A small-bore encapsulated O-ring behaves as a structural arch. Per unit of circumference it can demand roughly four times the load of a large-diameter ring of identical cross-section.
• Most disputed test results are artefacts of the fixture, not the part. Temperature, pre-load, platen friction and weld position each move the measured number by more than typical manufacturing variation.

1. The composite paradox in high-performance sealing

The engineering of static and dynamic seals has historically relied on the predictable mechanics of homogeneous elastomers. A standard O-ring moulded from nitrile (NBR), fluorocarbon (FKM) or silicone (VMQ) behaves as a viscous solid with memory — effectively incompressible, transmitting applied pressure in all directions. That behaviour supports a simple linear extrapolation: if a seal requires a given force to compress one linear inch of material, a seal of twice the circumference requires twice the force. Young's modulus, compression set and hardness remain properties of the material, independent of the geometry into which it is formed.

The Teflon^® encapsulated O-ring (TEO) breaks that assumption. It was developed to combine the chemical inertness of polytetrafluoroethylene and its copolymers with the elastic recovery of rubber, and in doing so it stopped being a material and became a structure: a resilient elastomeric core — typically silicone or FKM — sealed inside a seamless, thin-walled jacket of fluorinated ethylene propylene (FEP) or perfluoroalkoxy (PFA).

That composite architecture creates a mechanical system in which the rules of rubber elasticity no longer hold. The interaction between a Shore A core and a Shore D jacket introduces structural rigidity, hoop stress and frictional hysteresis, which together render conventional durometer readings physically meaningless. More consequentially, it creates a non-linear relationship between sealing force and seal diameter. A small-diameter TEO is not simply a shorter version of a large one; it is a different structural entity, and it requires substantially higher compressive force to seal.

2. Material science of the constituent components

To understand composite behaviour, the two constituent materials must first be considered separately. The incompatibility of their moduli is the root cause of every force anomaly described later in this report.

2.1 The core: viscoelasticity and the Shore A scale

The core of a TEO is normally silicone (VMQ) or fluorocarbon (FKM). Within an encapsulated seal it performs a single primary function: it acts as the energiser. The jacket, a semi-crystalline thermoplastic, lacks the elastic memory to recover from sustained compression, particularly under thermal cycling. The core supplies the restoring force that maintains contact stress against the mating hardware.

Silicone (VMQ)

Selected for low compression set and a wide service range of roughly −60 °C to +200 °C, silicone is the standard core. It is a hyperelastic material with highly flexible polymer chains, deforming substantially at low hysteresis. Typical hardness is 70–75 Shore A; Young's modulus is very low, on the order of 1–5 MPa. Its Poisson's ratio approaches 0.49–0.50, so it behaves as an incompressible fluid — compressed vertically, it must expand laterally to conserve volume.

Fluorocarbon (FKM / Viton^®)

Used where the core must serve as a secondary chemical barrier or where higher temperature capability is required. FKM has a higher damping coefficient than silicone, so an FKM-cored TEO feels stiffer and rebounds more slowly. This reduces dynamic responsiveness but increases static load retention in steady-state service.

2.2 The jacket: crystallinity and the Shore D scale

The encapsulating jacket is extruded from FEP or PFA. Unlike the amorphous rubber core, these are semi-crystalline thermoplastics whose stiffness is orders of magnitude higher.

• FEP (fluorinated ethylene propylene) — the standard jacket material. A copolymer of tetrafluoroethylene and hexafluoropropylene, its melt viscosity is low enough to extrude while retaining the chemical inertness of PTFE. Hardness registers around 55 on the Shore D scale. Flexural modulus at 23 °C is 550–650 MPa (80,000–94,000 psi) measured to ASTM D790, and maximum continuous service temperature is 204 °C.[7][11]
• PFA (perfluoroalkoxy) — a copolymer of tetrafluoroethylene and a perfluorinated vinyl ether, specified where service temperature reaches 260 °C or where semiconductor-grade purity is required. Published flexural modulus is 590–655 MPa and Shore D hardness around 60.[11]

It is worth stating plainly that the Shore D scale is not merely a continuation of Shore A. Under ASTM D2240 the two scales use different indenter geometries and different spring calibrations: Shore A a truncated cone with a rounded tip, Shore D a sharp 30° cone. They describe different mechanical regimes and cannot be interconverted by arithmetic.[5]

2.3 The modulus mismatch and the tube effect

The structural anomaly of the TEO arises from placing a material of 1–5 MPa modulus inside a shell of roughly 600 MPa — a mismatch of two orders of magnitude. When a solid O-ring is compressed, the material flows. When a TEO is compressed, the jacket first resists deformation through bending and hoop extension, before the core is meaningfully engaged at all.

This produces a structural threshold. Below a certain load the assembly behaves like a rigid plastic tube. Once the geometric yield point of the jacket — a function of shape, not of material yield strength — is passed, the jacket deflects and load transfers into the core. The transition is invisible to the eye but unmistakable in a force-deflection curve, appearing as a non-linear toe region followed by rapid stiffening. A solid rubber O-ring, by contrast, produces a smooth hyperelastic curve with no such inflection.

3. Classification: thermoplastic at the interface, elastomeric in behaviour

Encapsulated seals are routinely mis-classified in purchasing systems, material declarations and drawing notes, usually by being forced into a single category. The part is neither a thermoplastic component nor an elastomeric one. It is a composite whose two constituents are governed by two separate, non-overlapping nomenclature systems, and a specification naming only one of them is incomplete.

3.1 Two standards, two materials, one part

The jacket is a thermoplastic fluoropolymer. Its abbreviated designation — PFA, FEP, PTFE — comes from ISO 1043-1, which defines abbreviated terms for the basic polymers used in plastics.[1] That standard explicitly directs the reader elsewhere for rubber: nomenclature for rubbers and latices is given in ISO 1629.[1][2]

The core is an elastomer. Its designation comes from ASTM D1418, the nomenclature practice for rubber and rubber latices, which codes rubbers from the chemical composition of the polymer chain. Fluoroelastomer sits in the M class as FKM; silicone sits in the Q class as VMQ.[3] The ISO counterpart, ISO 1629, designates the same fluoroelastomer family FPM; the two terms refer to the same base material class.[2]

3.2 Function follows the core; exposure follows the shell

The practical logic is straightforward once the two layers are separated. The jacket presents a thermoplastic surface to the process media and therefore governs chemical compatibility, permeation and the upper service temperature. The core supplies elastic recovery — the memory that generates and retains contact stress — and therefore governs compression set, preload retention and low-load sealing. Remove the core and what remains is a fluoropolymer ring that behaves as a plastic gasket, not as an O-ring.

3.3 Why performance specifications attach to the core

ASTM D2000 classifies rubber by performance rather than by chemistry: types denote resistance to heat aging, classes denote resistance to swelling in oil, and a line call-out combines grade, type, class, hardness and tensile strength with any additional suffix requirements.[4] Those are precisely the properties the core controls, which is why a purchaser seeking a performance guarantee on an encapsulated seal writes the D2000 call-out against the core and handles the jacket through polymer identification plus separate chemical and thermal compatibility criteria. ASTM D2000 is technically aligned with SAE J200.[4]

3.4 Specification language

Table 1. Which layer governs which property, and where each is specified. No published standard classifies the assembled composite, which is why force to compress has to be specified explicitly.

The failure modes of single-bucket labelling run in both directions. Calling the part a thermoplastic ignores the energiser that creates preload and resilience. Calling it an elastomer ignores the media-contacting surface whose polymer identity governs chemical compatibility and thermal limits. Either error can put the wrong part into a critical joint.

4. Theoretical framework: the non-linearity of force

The question that prompts this report is a practical one, and it is asked often: why is the force not linear across different inside diameters at the same cross-section?

For solid rubber O-rings, load per linear inch (LPLI) is effectively constant. If a 100 mm ID O-ring requires a given force per millimetre of circumference, a 20 mm ID ring of the same cross-section requires approximately the same, minor stretch effects aside. One sample can therefore be tested and the result extrapolated to any size in the family. For encapsulated seals that linearity collapses: a 15 mm ID TEO is measurably harder, per unit of circumference, than a 200 mm ID TEO of identical cross-section.

4.1 The geometry of hoop stress

The jacket is, in effect, a tube bent into a torus. Its compression mechanics are dominated by the ratio of cross-section to ring radius. When the ring is compressed axially, conservation of volume requires the cross-section to widen radially. For the width to increase, the circumference of the jacket tube must change: the outer surface of the torus must stretch in tension, while the inner surface must compress.

4.2 The arch effect in small diameters

Large diameters (ID » CS)

In a large ring — say 150 mm ID with a 5.33 mm cross-section — curvature is low and the jacket behaves essentially as a straight tube. The sidewalls flex outward readily, and resistance comes primarily from the flexural modulus of the jacket wall, typically 0.010–0.020 in thick. The hoop contribution is negligible, because the radius of curvature is large enough to absorb the deformation without appreciable circumferential stretching.

Small diameters (ID ≈ CS)

In a small ring — 15 mm ID at the same 5.33 mm cross-section — the jacket already carries significant internal stress simply from having been formed into a tight circle. It now acts as a structural arch. Widening the cross-section, which compression requires, forces the fluoropolymer to stretch circumferentially.

The decisive point is structural rather than material. For a thin shell, bending stiffness scales with the cube of wall thickness while membrane (stretching) stiffness scales linearly with it. At a wall of 0.010–0.020 in the shell is therefore orders of magnitude more compliant in bending than in stretching, even though both modes draw on the same resin modulus. A large-diameter ring can accommodate compression almost entirely by bending its walls; a small-diameter ring cannot, and the load path shifts into the far stiffer membrane mode. This is why the effect appears abruptly as diameter falls, and it is a more accurate account than the common shorthand that small rings “engage tensile rather than flexural modulus” — for a semi-crystalline fluoropolymer those two moduli are of similar magnitude.[11]

4.3 Mathematical representation

Total force may be approximated as the superposition of a core term and a jacket term, the latter dependent on ring radius:

F_total ≈ F_core + F_jacket(R)

While the core term scales linearly with seal length, the jacket term carries a non-linear geometric stiffness dependent on curvature. Normalising by circumference gives force per unit length:

F_linear = F_totalL ≈ k_rubber + k_FEP (1 + αtR )

• k_rubber — stiffness contribution of the elastomer core (constant)
• k_FEP — flexural stiffness of the jacket wall (constant)
• α — dimensionless shape factor
• t — jacket wall thickness
• R — mean radius of the O-ring

The interpretation is straightforward: as radius decreases, the term αt/R grows hyperbolically. This is the mathematical statement of why small rings feel disproportionately hard. Force to compress diverges from the linear baseline as the inside diameter approaches the cross-sectional dimension.

Table 2. Divergence of sealing force with decreasing diameter at constant cross-section. A 12 mm ID encapsulated O-ring can require close to four times the force per millimetre of circumference of a 200 mm ID ring of the same wall and cross-section. The ID:CS ratios shown are nominal, computed for a 5 mm cross-section; relative force values are M-Cor engineering estimates.[12]

5. Extrapolating force calculations for design

An engineer cannot look up a hardness value and calculate bolt load. The calculation must incorporate the geometric stiffness described in Section 4. This section provides a working method for estimating the required compressive load.

5.1 The inadequacy of Shore hardness

Shore hardness cannot be converted mathematically into sealing force for an encapsulated seal, for reasons that are geometric rather than theoretical:

• A durometer indenter penetrates up to 2.5 mm at the zero-hardness end of the scale.[5] A jacket wall is 0.010–0.020 in — 0.25 to 0.51 mm. The indenter therefore passes through the jacket and reads a blend of jacket, core and support, weighted unpredictably.
• ASTM D2240 practice calls for a specimen at least 6 mm thick, so the supporting surface cannot influence the reading.[5] A jacket is more than an order of magnitude below that, and the curved surface of an O-ring does not satisfy the flat-specimen requirement either.
• A Shore D indenter is a sharp 30° cone. On a thin fluoropolymer jacket it pierces or permanently deforms the surface rather than measuring elastic recovery.
• The accepted alternative is load per linear inch (or per linear millimetre) at a stated compression percentage, conventionally 20 %. ASTM D1414 sets out test methods applied to whole O-rings rather than to moulded slabs.[6]

5.2 Calculating the sealing load — the geometric correction method

To predict the force required to seat a flange or compress an encapsulated seal, M-Cor applies a base elastomer force modified by a material multiplier and a geometric correction factor:

F_install = L_circ × F_base × M_jacket × K_geo

• F_install — total force required, in newtons
• L_circ — mean circumference of the seal, π (OD + ID) / 2
• F_base — force to compress a solid silicone O-ring of the same cross-section to 20 %, taken from standard elastomer tables
• M_jacket — material multiplier accounting for jacket stiffness; typically 2.0–2.5 for FEP and higher for PFA
• K_geo — geometric correction factor accounting for small-diameter hoop stress, from Table 3

Table 3. Geometric correction factors. These are M-Cor engineering estimates derived from internal force-to-compress data and are offered for first-pass sizing; they are not a published industry standard. Verify against measured values for any critical joint.[12]

Worked example — 20 mm ID × 3.53 mm cross-section port

The 3.53 mm cross-section is the AS568 / ISO 3601-1 standard 0.139 in section.[8][10] Mean circumference is π (27.06 + 20) / 2 = 73.9 mm, taken below as 74 mm. The baseline solid-silicone force of 3.5 N/mm at 20 % deflection is illustrative, used to demonstrate the method; substitute the measured value for the specific compound in any real calculation.

F = 74 × 3.5 × 2.5 × 1.65 ≈ 1 068 N

The equivalent solid silicone O-ring would require only 74 × 3.5 ≈ 259 N. The encapsulated seal therefore demands roughly four times the bolt load. Designing the joint on the rubber figure produces stand-off or leakage; designing the hardware on the rubber figure may also under-size the fasteners.

6. Metrology: measuring force to compress correctly

Manufacturing variability in jacket wall thickness and core fill, combined with the non-linearity described above, makes accurate measurement genuinely difficult. Handheld durometers produce false data on encapsulated seals under any circumstances. The only valid metric is force to compress (FTC), measured on a universal testing machine. This section is written as a working procedure.

6.1 Equipment and fixture requirements

Table 4. Minimum fixture specification for encapsulated-seal force testing.

6.2 Measurement procedure

Step 1 — Sample conditioning

Encapsulated seals carry thermal memory. A part stored in a cold warehouse presents a rigid jacket; one left in a hot vehicle presents a soft one. Condition samples at 23 °C and 50 % relative humidity for at least 24 hours before testing.

Step 2 — Geometric verification

Do not assume the nominal cross-section. Measure the actual cross-section at four equidistant points with a low-force micrometer and calculate the mean, then compute the deflection target from the measured value. For example, a measured mean of 5.40 mm gives a 20 % target of 1.08 mm compression, or a target height of 4.32 mm.

Step 3 — Machine zeroing

Bring the platens into contact, zero both force and displacement channels, then separate to admit the sample.

Step 4 — Placement and toe compensation

Centre the seal on the lower platen. Encapsulated rings are frequently slightly ovalised, so the upper platen contacts the high points first. Beginning displacement measurement at first touch counts the flattening of that ovality as cross-section compression and yields a falsely shallow curve. Apply a pre-load — 1–2 N for small rings, 5–10 N for large — to seat the sample, zero displacement at that point, then begin the test.

Step 5 — Compression profile

Run at 10 mm/min, the accepted quasi-static rate. Faster rates inflate the reading, because the viscoelastic core damps and the jacket is given no time to relax. Compress to 25 % deflection so that data beyond the 20 % reporting point is captured.

Step 6 — Data extraction

Record two values. Peak force is the force at the instant the 20 % target is reached; use it for installation and bolt-load calculations. Relaxed force is taken after holding the crosshead at 20 % for 60 seconds, during which the force decays exponentially through stress relaxation; use it for long-term sealing pressure estimates.

7. Variables that produce false readings

The following effects generate what is best described as ghost data: readings that look entirely plausible but are artefacts of the test setup rather than properties of the part. Each is capable of moving the result by more than normal manufacturing variation, which is why disputes over encapsulated-seal hardness so often prove unresolvable until the fixtures are compared.

7.1 Barrelling and platen friction

As the seal is compressed it expands radially, and friction between the fluoropolymer jacket and the steel platen opposes that expansion. On dry steel the ring cannot slide, so the cross-section barrels outward and the apparent stiffness rises. With oil on the platens the ring slides freely and the measured force falls. Most sealing applications are dry or process-fluid lubricated, so dry testing is generally preferred as the worst-case condition — but the test report must state which was used. A report that omits it cannot be compared with another laboratory's.

7.2 The splice and weld anomaly

Encapsulated seals are not moulded; the jacket tube is welded at a single point. The weld zone is usually slightly stiffer and may carry a dimensional bulge on the order of 0.05 mm. If the upper platen contacts that high spot first, the machine registers contact early and the displacement record is skewed, so the calculated 20 % point is reached before the remainder of the ring is fully compressed and the reported force reads low. Inspect the weld before testing; where the load path allows, orient it away from the primary contact zone, or measure the cross-section at the weld and calculate a target specific to that zone.

7.3 The air gap — soft start

Where manufacturing tolerances have run loose, the core may be marginally smaller than the jacket bore, leaving a microscopic air gap. The force-deflection curve then shows an extended flat region at the start, during which the machine is merely flattening a hollow tube before reaching the core (the dashed trace in Figure 2). A pronounced lag of this kind indicates an under-filled part. Sealing performance suffers because the core cannot energise the jacket immediately on installation.

7.4 Temperature hysteresis

Fluoropolymers have a high coefficient of thermal expansion and a modulus that falls sharply with temperature; FEP rigidity, as measured by flexural modulus, decreases significantly as temperature rises toward its maximum continuous-use limit.[11] Consider a quality laboratory sited next to a moulding press at 30 °C and an engineering laboratory held at 20 °C: the first will consistently pass parts the second measures as too hard. A ten-degree gap of this kind can move the measured force by roughly eight to ten per cent — comfortably larger than the tolerance band usually under dispute. Every disagreement about encapsulated-seal hardness should be resolved in a temperature-controlled environment before any other variable is investigated.

Table 5. Quality-control troubleshooting reference.

8. Implications for design and application

Non-linear force behaviour and the metrology described above translate into design rules that differ materially from those for solid rubber O-rings.

8.1 Groove design

Standard O-ring sizes and housing dimensions are given in ISO 3601-1 and ISO 3601-2, and in SAE AS568.[8][9][10] ISO 3601-1 states plainly that its dimensions and tolerances are suitable for any elastomeric material — the gland system was drawn around rubber, which flows into groove corners under compression.[8] A fluoropolymer jacket does not flow; it buckles. Placing an encapsulated seal in a standard groove at high volumetric fill risks crushing or pinching the jacket and cracking it. M-Cor recommends widening the groove by approximately 10–15 % so the stiffer jacket can expand laterally without impinging on the groove walls.[12]

8.2 Surface finish

Rubber will seal against a surface finish of about 32 µin (0.8 µm) Ra, because it flows into the peaks and valleys of the machined surface. A fluoropolymer jacket bridges over those valleys instead. For a gas-tight seal, improve the sealing-face finish to 10–16 µin (0.25–0.4 µm) Ra.

8.3 Installation stretch

Rubber tolerates 20–50 % stretch during installation without damage. FEP has a low elongation yield: stretching an encapsulated seal more than 5–10 % — snapping it over a piston, for instance — can yield the jacket, producing stress whitening and permanent deformation; the 5–10 % figure is M-Cor installation guidance.[12] A yielded jacket does not recover, and the seal will leak. Where installation stretch cannot be avoided, specify a design that allows assembly over a chamfered lead-in, or consult M-Cor about a split-gland or alternative configuration.

Table 6. Design rules at a glance.

9. Conclusion

The Teflon^® encapsulated O-ring is a composite system in which a thin-walled, high-modulus pressure vessel interacts with a hyperelastic core. That interaction produces a force response which is non-linear with respect to diameter and which invalidates the extrapolation methods engineers reasonably apply to solid rubber.

Three points carry the practical weight of this report:

• Discard Shore hardness. It is not a valid metric for a composite seal. Specify and report force to compress at 20 % deflection.
• Respect hoop stress. Small-diameter encapsulated seals are structurally rigid arches and can require up to four times the bolt load of a rubber equivalent.
• Control the metrology. Accurate measurement demands a universal testing machine, temperature control, defined pre-load and stated friction conditions. Handheld gauges produce numbers, not data.

Understanding the physics of the jacket — its flexural modulus, its hoop stiffness and its thermal sensitivity — allows these seals to be modelled, measured and deployed with confidence, without falling into the familiar traps of linear elastic assumption.

10. References

Numbered markers in the text refer to the sources below. Property values quoted for fluoropolymer jacket materials are taken from resin-manufacturer literature; test methods and nomenclature are cited to the governing standards. Where a value originates in M-Cor's own records it is identified as such rather than attributed to a published source.

1. ISO 1043-1:2011 — Plastics: Symbols and abbreviated terms, Part 1: Basic polymers and their special characteristics. Cited for: abbreviated designations for jacket polymers (PFA, FEP, PTFE); cross-reference directing rubber nomenclature to ISO 1629.
2. ISO 1629 — Rubber and latices: Nomenclature. Cited for: ISO designation of the core elastomer; FPM as the ISO equivalent of FKM.
3. ASTM D1418 — Standard Practice for Rubber and Rubber Latices: Nomenclature. Cited for: classification of the core by polymer-chain chemistry; FKM in the M class, VMQ in the Q class.
4. ASTM D2000 — Standard Classification System for Rubber Products in Automotive Applications (aligned with SAE J200). Cited for: type and class definitions, line call-out structure for core performance specification.
5. ASTM D2240 — Standard Test Method for Rubber Property: Durometer Hardness. Cited for: indenter geometry, penetration depth and minimum specimen thickness, which establish the limits of durometer testing on thin-jacketed composites.
6. ASTM D1414 — Standard Test Methods for Rubber O-Rings. Cited for: test protocols applied to whole O-rings rather than moulded slabs.
7. ASTM D790 — Standard Test Methods for Flexural Properties of Unreinforced and Reinforced Plastics. Cited for: method underlying the quoted jacket flexural modulus values.
8. ISO 3601-1 — Fluid power systems, O-rings, Part 1: Inside diameters, cross-sections, tolerances and designation codes. Cited for: standard cross-sections; statement that the dimensions suit any elastomeric material.
9. ISO 3601-2 — Fluid power systems, O-rings, Part 2: Housing dimensions for general applications. Cited for: gland and groove dimensions drawn for elastomeric seals.
10. SAE AS568 — Aerospace Size Standard for O-Rings. Cited for: dash-number size system, including the 0.139 in cross-section used in the worked example.
11. Chemours (formerly DuPont) — Teflon™ FEP and PFA fluoropolymer resin properties literature. Cited for: flexural modulus, Shore D hardness, maximum continuous service temperature, and the decrease in rigidity with rising temperature.
12. M-Cor Inc. — internal force-to-compress and installation records. Cited for: K_geo correction factors, relative force values in Table 2, groove-widening and installation-stretch guidance.

Note on sources. Standards are cited by designation without a year where the content referenced is stable across recent editions; verify against the current edition before use in a controlled document. Competitor publications have deliberately not been used as references in this report.