gasket creep compensation
Gasket creep compensation in bolted flanges
By ISOKLAMP Engineering, Inc. Editorial Team · Updated
The gasket is usually the largest single loss in the stack
Put a polymer gasket in a bolted flange and it becomes the dominant source of preload loss, usually by a wide margin. Embedment at a steel interface runs 3 to 8 micrometres. A 3 mm PTFE gasket runs 20 to 60. It also keeps going long after embedment has stopped.
| Gasket material | Thickness | Creep over first 6 months | Continues? |
|---|---|---|---|
| Virgin PTFE | 3 mm | 40 to 60 µm | Yes, slowly |
| Filled / expanded PTFE | 3 mm | 20 to 35 µm | Yes, slowly |
| EPDM, 70 shore A | 3 mm | 25 to 45 µm | Largely stabilises |
| Silicone, 60 shore A | 3 mm | 30 to 55 µm | Largely stabilises |
| Compressed fibre | 1,5 mm | 15 to 30 µm | Stabilises |
| Graphite | 3 mm | 10 to 20 µm | Stabilises |
Values are representative ranges from published manufacturer compression-set data at ambient temperature. Elevated service temperature increases all of them substantially, and PTFE in particular creeps far faster above 60 °C.
Reference joint used throughout
Every figure in this article refers to the same reference joint, so numbers are comparable across articles and against your own calculations.
| Parameter | Value |
|---|---|
| Bolt | M16 × 2,0, property class 10.9, to ISO 898-1 |
| Assembly preload F_V | 70,0 kN |
| Clamp length | 48 mm, steel on steel |
| Bolt stiffness k_S | 1,04 × 10⁹ N/m |
| Member stiffness k_P | 5,71 × 10⁹ N/m |
| Load factor Φ | 0,154 |
| Transverse test | DIN 25201-4:2010-03 Annex B, 2 000 cycles, ±0,45 mm slip |
Stiffnesses are calculated to VDI 2230 Sheet 1 using the standard cone-of-compression method.
What the creep costs
At 0,884 kN per micrometre for the reference joint:
| Gasket creep | Clamp force lost | Residual | Seating stress retained |
|---|---|---|---|
| 20 µm | 17,7 kN | 52,3 kN | 74,7 % |
| 30 µm | 26,5 kN | 43,5 kN | 62,1 % |
| 40 µm | 35,4 kN | 34,6 kN | 49,4 % |
| 60 µm | 53,0 kN | 17,0 kN | 24,3 % |
The right-hand column is the one that matters for a sealed joint. Gasket seating stress is what makes the seal. Lose 60 micrometres of PTFE and you are holding a quarter of the seating stress you designed for, with every bolt perfectly tight.
This is why flanges weep without anyone having done anything wrong.
Live loading, and where it runs out
The established answer is live loading: a Belleville stack under each nut, adding elastic travel so a given creep costs less clamp force.
It works, and it is a genuine improvement over a plain washer. It has three limits.
Force falls along the curve. A disc spring stores energy on a load-deflection curve. As it extends to follow the creeping gasket, the force it delivers falls. Recover 40 micrometres and you have moved down the curve by a proportional amount. Live loading reduces the slope of the loss, it does not flatten it.
Travel is limited. A single Belleville gives 0,15 to 0,25 mm of usable travel. Stacking in series multiplies travel but also multiplies stack height, and flange geometry rarely has the room.
No rotation control. A disc spring does nothing about Junker slip. Flanges on vibrating plant frequently need both, which means a Belleville stack plus a locking device, plus the stack height for both.
| Belleville live loading | Full creep compensation | |
|---|---|---|
| Mechanism | Elastic compliance | One-way helical take-up |
| Force across travel | Falls along spring curve | Flat, quasi-zero stiffness |
| Usable travel | 0,15 to 0,40 mm | 0,50 mm |
| Restores lost length | No | Yes |
| Arrests rotation | No | Yes |
| Stack height | 2 to 6 mm | 3,40 mm |
| Reserve readable | No | Yes |
Disc spring figures are reconstructed from published manufacturer load-deflection curves, sourced on the test data page.
What full compensation requires
To hold seating stress through gasket creep you have to put the thickness back, not merely charge less for losing it. That needs:
- A reserve of clamp length larger than the creep budget with margin. 0,50 mm against a 60 µm worst case is a factor of 8,3.
- Flat force delivery. A quasi-zero-stiffness element delivers the same force at 0,45 mm of consumed travel as at 0,05 mm. That is what keeps seating stress constant rather than merely slowing its decline.
- One-way action. tan α_c < μ_r, so the take-up cannot be pushed back by the next thermal cycle.
- Rotation control on the same part, since flange bolts on rotating plant see transverse load.
Performance through a creep cycle
Representative flange joint, M16 class 10.9, 3 mm virgin PTFE gasket, 180 days at 60 °C:
| Elapsed | ISOKLAMP CFR | Belleville stack | Plain washer |
|---|---|---|---|
| Day 0 | 100,0 % | 100,0 % | 100,0 % |
| Day 7 | 98,4 % | 93,7 % | 82,1 % |
| Day 30 | 98,0 % | 88,2 % | 68,4 % |
| Day 90 | 97,6 % | 83,1 % | 55,9 % |
| Day 180 | 97,3 % | 79,4 % | 47,2 % |
Predicted by finite-element analysis of the ISK-16 geometry coupled to VDI 2230 Sheet 1 mechanics with a PTFE creep model fitted to published compression-set data. Belleville and plain washer baselines are calibrated against published gasket relaxation data, sourced on the test data page.
The Belleville curve is doing real work. It holds nearly 80 percent where a plain washer holds
- It still declines continuously, because every micrometre it gives up costs force along the
spring curve. The CFR curve is flat because the take-up refills the thickness and the constant torque element does not care how much travel it has already spent.
Where this applies
- Food, dairy and beverage flanges with PTFE or EPDM. See hygienic bolting.
- Pharmaceutical process piping under repeated SIP. See CIP and SIP thermal cycling.
- Chemical service with expanded PTFE at elevated temperature.
- Composite and GRP flanges, where the member itself creeps as well as the gasket.
- Any flange with a retorque interval you would like to extend or eliminate.
Compare against disc springs in detail on the Belleville alternative. Reserve monitoring on the preload indicator.
Frequently asked questions
How much does a gasket creep in a bolted flange?
A 3 mm virgin PTFE gasket typically creeps 40 to 60 micrometres over its first six months at ambient temperature, and considerably more above 60 degrees. Filled PTFE runs 20 to 35, EPDM 25 to 45, silicone 30 to 55, compressed fibre 15 to 30, and graphite 10 to 20. PTFE and elastomers continue creeping slowly rather than stabilising.
What does gasket creep cost in clamp force?
For an M16 class 10.9 joint preloaded to 70 kN, the loss is 0,884 kN per micrometre. Twenty micrometres of creep costs 17,7 kN and leaves 74,7 percent of the seating stress. Sixty micrometres costs 53,0 kN and leaves 24,3 percent, which is why flanges weep with every bolt still perfectly tight.
Does live loading with Belleville washers solve gasket creep?
It substantially improves matters without solving them. A disc spring adds elastic travel so a given creep costs less clamp force, but the force it delivers falls along its load-deflection curve as it extends, so seating stress declines continuously rather than holding. Usable travel is also limited to 0,15 to 0,40 mm, and a disc spring provides no rotation control.
What is full gasket creep compensation?
Restoring the lost gasket thickness rather than reducing the penalty for losing it. That requires a reserve of clamp length larger than the creep budget, flat force delivery across the full travel so seating stress does not decline, one-way action so thermal cycling cannot push the take-up back, and rotation control on the same part for flanges on vibrating plant.
How does compensation performance compare over six months?
For a representative M16 flange with a 3 mm virgin PTFE gasket at 60 degrees, analysis predicts 97,3 percent residual clamp force at 180 days with a CFR washer, against 79,4 percent for a Belleville stack and 47,2 percent for a plain washer. The Belleville curve declines continuously because every micrometre it gives up costs force along the spring curve.
Take it further
- The ISOKLAMP technical report covers the full derivation, the geometry, and every dataset behind these figures.
- Full residual clamp force dataset gives the residual clamp force numbers for ten securing methods, with sources.
- Contact sales puts you in touch with the engineering team. Bring us a joint that keeps failing and we will run a VDI 2230 Sheet 1 analysis on it.
Engineering questions go to engineering@isoklamp.com. An engineer answers, not a form.