# O-rings

Elastomer sealing, compound identity, gland geometry and failure limits, illustrated by Parker 2-214 in N0674-70.

Reviewed: 2026-11-01

Tags: components, seals, elastomers

Parker's **2-214** has **24.99 mm** nominal ID and **3.53 mm** cross-section. An ideal **2.80 mm-high × 4.80 mm-wide** rectangular gland gives **20.68% squeeze** and **72.82% fill** with equal mean circumferences, no stretch, corner radii or swelling. Check actual tolerances, pressure direction, motion and compound compatibility before using the geometry. [European handbook, p. 10, Table 2.3](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=10).

## The ring and gland form the seal

Initial axial or radial squeeze creates sealing contact. Pressure redistributes the nearly incompressible elastomer against the supporting gland and mating surfaces. [Parker, *O-Ring Handbook*, European edition PTD 5705 EN, 09/2025, p. 7](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=7).

Several classifications describe different physical relationships:

| Classification | Meaning |
| --- | --- |
| Static or dynamic | Whether the sealed surfaces move relative to each other. Static service can still include small movements caused by pressure. Dynamic service includes reciprocating, oscillating and rotary motion. |
| Radial or axial compression | The direction in which the O-ring cross-section is squeezed. A piston or rod gland commonly applies radial compression; a face seal applies axial compression. |
| Piston or rod gland | A piston gland is in the inner machine element; a rod gland is in the outer element. The location changes the mating diameters and tolerance stack. |

These distinctions follow [Parker's European handbook, pp. 7–9](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=9). A static face seal and a reciprocating piston seal can use the same nominal ring cross-section while requiring different gland geometry and operating conditions.

## Polymer family is not compound identity

NBR, EPDM and FKM identify elastomer families. A formulated compound also contains fillers, curing ingredients, ageing retardants and other additives that modify its properties. Consequently, a family name or a nominal hardness does not uniquely identify fluid resistance, low-temperature behaviour, compression set or friction. [Parker European handbook, pp. 7 and 75](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=7).

Parker's product examples include NBR **N0674-70** and **N0552-90**, and FKM **V0747-75**. Even the two NBR compounds have different nominal hardnesses. The compound number and size designation describe different things: changing the compound does not change the meaning of the size number, but a shared size does not establish equivalent service performance. [Parker European handbook, p. 61, Table 5.3](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=61).

Temperature and medium act together. Absorption can swell an elastomer; extraction of soluble ingredients can produce shrinkage. These processes can occur concurrently and change mechanical behaviour. A temperature range established in a non-aggressive medium is not automatically valid in a different fluid, cleaning agent or lubricant. Parker also notes that stated pressure, temperature and surface-speed maxima are interrelated. [European handbook, pp. 4 and 75](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=75).

The compatibility chart illustrates the distinction between screening information and a service qualification. Its legend uses 1 for satisfactory, 2 for fair, 3 for conditional and 4 for unsatisfactory; a blank means insufficient data. The chart also separates families listed for static and dynamic service from columns restricted to static service. For example, its acetone row rates NBR 4 and EPDM 1. That family-level comparison does not establish suitability at an unspecified temperature, concentration or duty cycle. [European handbook, p. 123, compatibility table and footnote 1](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=123).

## Squeeze, stretch and gland fill

These quantities measure different deformations or volumes:

| Quantity | Geometric relationship | Consequence |
| --- | --- | --- |
| Cross-section squeeze | Reduction of the section against the effective gland height or depth | Creates initial sealing contact; affected by section and hardware tolerances |
| Installed stretch | Increase of the ring's inside diameter relative to its free inside diameter | Changes the cross-section as well as the ring circumference |
| Gland fill | O-ring volume divided by available gland volume | Determines how much space remains for dimensional variation, thermal expansion and fluid-induced swelling |

Parker's European handbook gives a worst-case squeeze calculation using minimum cross-section and maximum gland depth, and explains that stretch reduces the effective cross-section. Its US handbook defines gland fill separately. [European handbook, pp. 12 and 101](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=101); [US handbook ORD 5700, p. 3-9, physical PDF p. 55](https://www.parker.com/content/dam/Parker-com/Literature/O-Ring-Division-Literature/ORD-5700.pdf#page=55).

For an ideal section with diameter `c` and effective gland height `h`, the geometric squeeze is `100 × (c − h)/c`. An installed ring may have a different effective section from its free nominal section. The largest section with the smallest gland height produces a different squeeze from the smallest section with the largest height; nominal arithmetic does not describe that full tolerance stack. [European handbook, pp. 12 and 101](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=12).

The two cited handbook editions give different installed-stretch guidance. The European edition describes up to **6%** stretch as a general guideline. The US edition discourages assembled stretch above **5%**, explaining that greater internal stress can accelerate ageing while allowing that some exceptions sacrifice useful life. These are source-specific recommendations, not a single universal threshold. Temporary assembly elongation is another condition: the European edition's separate discussion gives a maximum of 50% of inside diameter during assembly and also requires elongation below the material's breaking elongation. That is not an installed-stretch allowance or a target for fitting a seal. [European handbook, pp. 8 and 19](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=19); [US handbook, p. 3-9, physical p. 55](https://www.parker.com/content/dam/Parker-com/Literature/O-Ring-Division-Literature/ORD-5700.pdf#page=55).

For most O-ring applications, the US handbook describes **60–85% gland fill**, with 75% as its preferred value, to accommodate tolerance variation, swelling and thermal expansion. It separately calls for at least 10% void. The preferred band and minimum void are distinct statements; they do not make 90% fill an optimum for every application. [US handbook, §3.7, p. 3-9, physical p. 55](https://www.parker.com/content/dam/Parker-com/Literature/O-Ring-Division-Literature/ORD-5700.pdf#page=55).

## Static and moving service require different conditions

Moving seals balance contact against friction and wear. Parker's European hydraulic discussion describes average cross-section deformation of 10–15% and a minimum of 8% after its stated tolerance calculation. Its pneumatic discussion uses lower compression and describes a special floating-piston arrangement in which the section is not initially compressed. Those are different designs; neither is a universal squeeze rule for all dynamic O-rings. [European handbook, pp. 12–13](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=13).

Pressure can push elastomer into the clearance between machine parts. Eccentricity, component flexure, bolt stretch, temperature and compound swelling can make the working gap more severe than the nominal assembly gap. Repeated opening and closing of that gap can nibble or shear material from the ring. A back-up ring provides support on the low-pressure side; pressure direction and reversal affect the support arrangement. [European handbook, pp. 114–115, §10.2](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=114).

Surface texture is also specific to the contacting location and motion. The following are Parker's recommended maximum roughness values for the **contact surface**, with the tables' additional bearing-area condition `tp > 50%`:

| Contact condition | Ra | Rmax |
| --- | --- | --- |
| Static, non-pulsating pressure | 1.60 µm | 6.30 µm |
| Static, pulsating pressure | 0.80 µm | 3.20 µm |
| Reciprocating hydraulic seal | 0.40 µm | 1.60 µm |

Groove bases, sides and entry chamfers have their own rows. Ra and Rmax describe different surface measurements, so the two columns are not interchangeable. [European handbook, p. 10, Table 2.2; p. 13, Table 2.8; p. 101, §8.16](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=10).

Installation can damage a suitable compound before it reaches service. Sharp threads or edges can cut it; rolling during fitting can contribute to spiral failure; cross-drilled ports can shear a ring as it traverses the opening. Chamfers, protective fitting sleeves, clean glands and compatible assembly lubricants address different parts of that problem. [European handbook, pp. 18–19, §2.5](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=18).

## Exact example: Parker 2-214 in N0674-70

Parker's European size table identifies **2-214** with nominal inside diameter **24.99 mm** and cross-section **3.53 mm**. Its Kit No. 4 description specifies N0674-70 and includes that size. The compound table identifies **N0674-70 as black NBR, 70 ±5 Shore A**, with **−35 to +100°C in the static-temperature column**. These are catalog values, not measurements of an individual ring or a fluid-independent service envelope. [European handbook, p. 44, Table 4.3; p. 61, Kit No. 4; p. 77, compound table](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=44).

The nominal free outside diameter derived from these dimensions is `24.99 + 2 × 3.53 = 32.05 mm`. It is a geometric result, not an additional inspected dimension.

### Assumed geometry, not a qualified gland

For a simple cross-section exercise, take the nominal 3.53 mm circular section and assume a rectangular gland section **2.80 mm high and 4.80 mm wide**, without back-up rings. These height and width values appear in Parker's axial-compression table for a 3.53 mm section. The exercise additionally assumes no circumferential stretch or compression, no corner radii, no thermal expansion or fluid swelling, and equal mean circumferences for ring and gland so that their volume ratio reduces to a section-area ratio. [European handbook, p. 10, Table 2.3](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=10).

```text
Section squeeze = 100 × (3.53 − 2.80)/3.53 = 20.68%
Ring section area = π × 3.53²/4 = 9.787 mm²
Gland section area = 2.80 × 4.80 = 13.44 mm²
Idealized fill = 100 × 9.787/13.44 = 72.82%
```

The squeeze relationship and volume-fill definition are described in the [European handbook, p. 12](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=12), and [US handbook, p. 3-9, physical p. 55](https://www.parker.com/content/dam/Parker-com/Literature/O-Ring-Division-Literature/ORD-5700.pdf#page=55). The results above are derived nominal geometry. They do not instantiate the complete flange arrangement: pressure direction, actual annular dimensions, ring and hardware tolerances, radii and operating deformation are absent. Parker's axial table prints `h +0.10 mm` and `b +0.20 mm`; those tolerances alone already make the nominal values incomplete.

A separate hypothetical installation with ring inside diameter stretched from 24.99 to 25.40 mm gives `100 × (25.40 − 24.99)/24.99 = 1.64%` installed stretch. That exercise is not combined with the unstretched section calculation. The resulting section change would affect a subsequent squeeze and fill calculation. [European handbook, p. 101, §8.15 and Fig. 8.13](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=101).

## Compression set and ageing are not installed squeeze

Squeeze is the deformation applied by the hardware. **Compression set** measures residual deformation after unloading under specified test conditions. **Stress relaxation** concerns the decline of force retained while an elastomer remains deformed. Both depend on formulation, time, temperature, deformation and contact medium, and neither is a direct service-life measurement. [European handbook, pp. 93–95 and 107](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=94).

Parker's separately dated N0674 data sheet, printed **17 May 2024**, reports the following laboratory distinctions:

| Property | Reported result and conditions |
| --- | --- |
| Hardness | Reference 70 ±5 Shore A; typical 72, on a specimen thicker than 6 mm |
| Compression set at 100°C | Typical 14% after 70 h, with 25% deformation, test specimen B, method A, DIN ISO 815-1 |
| Compression set at 125°C | Typical 34% after 70 h, with the same listed deformation, specimen and method |
| Volume change in ASTM-1 test medium | Typical −2.8% after 70 h at 100°C |
| Volume change in ASTM-3 test medium | Typical +12.5% after 70 h at 100°C |

The data-sheet footer identifies the results as typical laboratory material properties, not specification values or necessarily results on finished goods. Thus 72 Shore A is not the measured hardness of the 3.53 mm-section ring, 14% set is not 14% installed squeeze, and a 125°C test is not a 125°C continuous-service rating. The different volume changes also illustrate why a generic “oil-resistant” description cannot establish compatibility with every oil. [Parker, *Compound N0674, NBR, black*, print date 17.05.2024, single sheet, physical p. 1](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Compound_Data_Sheets/EMG_PTD_Compound-Data-Sheet_NBR_N0674_EN.pdf#page=1).

## Source scope and remaining limits

This reference concerns conventional solid elastomer O-rings. The principal geometry and example source is Parker's **152-page European PTD 5705 EN handbook, dated 09/2025**. US handbook references identify printed section pages as well as PDF pages; the separately dated **N0674 sheet is printed 17.05.2024**. Manufacturer guidance and typical laboratory results describe their stated conditions, rather than a qualified assembly. [European handbook, back cover, PDF p. 152](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Catalogs/Catalog_O-Ring-Handbook_PTD5705-EN.pdf#page=152); [N0674 data sheet, p. 1](https://www.parker.com/content/dam/Parker-com/Literature/Praedifa/Compound_Data_Sheets/EMG_PTD_Compound-Data-Sheet_NBR_N0674_EN.pdf#page=1).

The cited nominal size, compound descriptions and laboratory results do not establish current-production or current-lot properties, interchangeability, an actual gland's clearance, regulatory approval or a customer's service qualification. Full AS568, ISO, DIN and ASTM compliance is not established by their appearance in manufacturer tables. Actual fluid composition, pressure peaks, gap under load, surface condition, temperature history and motion remain necessary inputs to an application-specific performance assessment. No universal pressure, speed, leak rate, maintenance interval or seal life is derived here.