thermal ratcheting
Thermal ratcheting in bolted joints
By ISOKLAMP Engineering, Inc. Editorial Team · Updated
The loss that never stops
Embedment finishes. Coating creep slows to a crawl. Gaskets eventually approach an asymptote.
Thermal ratcheting does none of those things. Every cycle takes a little more, and the total is bounded only by how many cycles the machine sees. A joint that survives two years of duty can fail in the third for no reason the maintenance record will show.
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.
The mechanism
Consider an aluminium member clamped by a steel bolt.
| Material | Linear expansion coefficient |
|---|---|
| Aluminium alloys | 22 to 24 × 10⁻⁶ /K |
| Copper | 16,5 × 10⁻⁶ /K |
| Austenitic stainless (A4) | 16 to 17 × 10⁻⁶ /K |
| Carbon and alloy steel | 11 to 12 × 10⁻⁶ /K |
| CFRP, in-plane | 0 to 2 × 10⁻⁶ /K |
| CFRP, through-thickness | 25 to 35 × 10⁻⁶ /K |
Heating. The aluminium grows faster than the bolt. The stack tries to get longer than the bolt permits, so bolt tension rises sharply. For a 48 mm aluminium grip and a 60 K excursion, the differential is:
Δδ = (α_Al − α_steel) · L · ΔT = (23 − 12) × 10⁻⁶ × 48 mm × 60 K = 31,7 µm
At the joint's stiffness that is an additional 28,0 kN of bolt tension, on top of the 70,0 kN assembly preload. The bolt is now at 98 kN, and the aluminium bearing surface is carrying contact stress it was never designed for.
Yielding. Aluminium at 0,2 percent proof stress around 250 MPa yields locally under the washer. The bearing area indents. Permanently.
Cooling. Everything contracts back, but the indentation stays. The stack is now shorter than it was, and preload is lower than it started.
Repeat. Next cycle, same again, from a slightly lower baseline.
Per-cycle budgets
Analysis of the reference joint geometry with an aluminium member, VDI 2230 Sheet 1 stiffness model with an elastic-plastic bearing surface:
| Cycle | Permanent set added | Cumulative | Residual clamp force |
|---|---|---|---|
| 1 | 4,1 µm | 4,1 µm | 96,3 % |
| 5 | 2,2 µm | 12,8 µm | 88,5 % |
| 20 | 0,9 µm | 27,4 µm | 75,4 % |
| 50 | 0,6 µm | 41,1 µm | 63,1 % |
| 100 | 0,4 µm | 58,3 µm | 47,7 % |
| 250 | 0,3 µm | 96,0 µm | 13,7 % |
The per-cycle increment decays as the bearing area work-hardens and spreads, but it never reaches zero, and the cumulative total keeps climbing. A daily thermal cycle reaches 250 cycles inside a year.
Why the usual devices do not help
Locking washers. The nut never rotated. A wedge-lock washer holds a nut that was already held.
Belleville stacks. These genuinely help, and they are the current best practice for thermally cycled joints. A disc spring lowers effective stiffness, so the same differential expansion produces less tension rise on heating, which means less plastic indentation per cycle. The limit is that the spring still loses force along its curve as the stack shortens, and the accumulation still runs one way. Live loading slows the ratchet. It does not stop it.
Hardened washers. Raising bearing hardness reduces indentation at the two washer interfaces. Worth doing, and it does not touch the faying surfaces.
Larger bearing area. Reduces contact stress below the member's yield point, which can stop the ratchet outright where geometry allows. This is the correct fix when you have the room. Frequently you do not.
Arresting the accumulation
A ratchet is defeated by a counter-ratchet. Vibration-Actuated Take-up advances one way as the stack shortens, which is precisely the failure mode thermal ratcheting produces.
The two mechanisms run against each other. The thermal ratchet removes length at a decaying per-cycle rate. The take-up restores length at a rate set by the joint's micro-slip activity, which in a thermally cycled joint is substantial because differential expansion produces radial slip at every interface on every cycle.
| Cycle | ISOKLAMP CFR | Belleville stack | Plain washer |
|---|---|---|---|
| 1 | 98,9 % | 97,8 % | 96,3 % |
| 20 | 98,2 % | 91,4 % | 75,4 % |
| 50 | 97,9 % | 86,0 % | 63,1 % |
| 100 | 97,6 % | 79,7 % | 47,7 % |
| 250 | 97,1 % | 68,2 % | 13,7 % |
| Reserve used at 250 | 19,2 % | n/a | n/a |
ISOKLAMP figures are predicted by finite-element analysis of the ISK-16 geometry coupled to VDI 2230 Sheet 1 mechanics with an elastic-plastic bearing model. Baselines are calibrated against published thermal cycling data, sourced on Full residual clamp force dataset.
The reserve consumption line is the one to design against. At 250 cycles the ISK-16 has spent 19,2 percent of its 0,50 mm, which projects to roughly 1 300 cycles of aluminium thermal ratcheting before replacement, and that number is readable off the indicator window rather than inferred. See Preload reserve indicator.
Where thermal ratcheting dominates
- Aluminium and copper busbars bolted with steel fasteners. See Busbar and battery joints.
- EV battery pack module-to-busbar joints. See Battery pack preload.
- Hygienic stainless plant under CIP and SIP cycling. See CIP and SIP thermal cycling.
- Engine and exhaust mounting into aluminium castings.
- Heat exchanger flanges and header bolting.
- CFRP-to-metal joints, where through-thickness expansion is high and matrix creep compounds it.
Design check
- Compute the differential expansion, Δδ = (α₁ − α₂) · L · ΔT, using your worst-case excursion, not the nominal.
- Convert to peak bolt tension using your joint stiffnesses.
- Compare peak bearing stress against the softer member's proof stress at temperature, not at ambient. Aluminium loses substantial strength by 150 °C.
- If bearing stress exceeds proof stress, the joint will ratchet. Either increase bearing area, reduce stiffness, or provide take-up.
Step 3 is where most dissimilar-metal joints are quietly already failing. More on Non-rotational preload loss.
Frequently asked questions
What is thermal ratcheting in a bolted joint?
Thermal ratcheting is progressive, permanent preload loss in a joint whose members and fastener expand at different rates. Heating raises bolt tension because the softer member grows faster, the bearing surface yields locally, and cooling leaves the stack permanently shorter. Each cycle repeats from a lower baseline, and unlike embedment the loss does not self-limit.
How much preload does thermal ratcheting cost per cycle?
For a 48 mm aluminium grip clamped by a steel bolt through a 60 K excursion, analysis gives 4,1 micrometres of permanent set on the first cycle, decaying to about 0,3 micrometres by cycle 250 as the bearing area work-hardens. Cumulative loss reaches 96 micrometres by 250 cycles, which is 86 percent of the assembly preload gone.
Why does aluminium against steel ratchet so badly?
Aluminium expands at 22 to 24 × 10⁻⁶ per kelvin against 11 to 12 for steel, roughly double. A 48 mm grip through a 60 K excursion produces 31,7 micrometres of differential expansion, which adds about 28 kN of bolt tension on top of assembly preload. Aluminium yields at around 250 MPa proof stress, so the bearing surface indents permanently.
Do Belleville washers stop thermal ratcheting?
They slow it substantially and are current best practice, but they do not stop it. A disc spring lowers effective joint stiffness so each thermal excursion produces less tension rise and less plastic indentation. The accumulation still runs one way, and the spring force falls along its load-deflection curve as the stack shortens.
How do you stop thermal ratcheting completely?
Either remove the cause or counter it. Increasing bearing area until contact stress falls below the softer member's proof stress at temperature stops the ratchet outright where geometry allows. Where it does not, a one-way take-up mechanism restores length at the same rate the ratchet removes it. Analysis predicts 97,1 percent residual clamp force at 250 aluminium thermal cycles with 19,2 percent of take-up reserve consumed.
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.
- Design partner programme is open. Eight slots, two per sector. 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.