coating creep
Coating creep and bolted joint preload
By ISOKLAMP Engineering, Inc. Editorial Team · Updated
The trade you did not know you made
Corrosion protection is not optional on most fasteners, so the surface finish gets specified early and rarely revisited. What the specification does not say is that most coatings are softer than the substrate, and everything softer than the substrate flows under contact stress.
A zinc flake coating at 8 to 12 micrometres nominal thickness is carrying the full bearing stress of the joint at both washer faces and, in many designs, at the faying surfaces too.
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.
Budgets by coating
| Coating | Nominal thickness | Creep at 20 °C | Creep at 100 °C |
|---|---|---|---|
| Zinc flake (Delta-Tone, Geomet class) | 8 to 12 µm | 4 to 9 µm | 7 to 15 µm |
| Zinc flake with integral lubricant topcoat | 10 to 16 µm | 6 to 12 µm | 10 to 20 µm |
| PTFE-impregnated | 15 to 25 µm | 8 to 18 µm | 15 to 30 µm |
| Electroless nickel, as plated | 15 to 25 µm | 1 to 3 µm | 2 to 5 µm |
| Electroless nickel, heat treated | 15 to 25 µm | Below 1 µm | 1 to 2 µm |
| Hot dip galvanised | 45 to 85 µm | 8 to 20 µm | 15 to 35 µm |
| Anodised aluminium (member side) | 15 to 25 µm | 2 to 6 µm | 4 to 10 µm |
| Phosphate and oil | 3 to 8 µm | 2 to 5 µm | 4 to 8 µm |
Values are representative ranges at typical underhead bearing stress for a class 10.9 fastener at nominal preload. Creep is stress-dependent, so a design at reduced bearing area moves to the upper end of each range.
Two observations engineers usually find useful. First, hot dip galvanising is the largest single coating contributor, purely because there is so much of it. Second, heat-treated electroless nickel is effectively creep-free, which is why it appears in precision joint design despite the cost.
What it costs
At 0,884 kN per micrometre for the reference joint, and counting a coated fastener bearing on coated members, the load path may include four to six coated interfaces.
Worked example. M16 class 10.9, zinc flake with lubricant topcoat, through two hot dip galvanised plates, service at 90 °C:
| Interface | Coating | Creep |
|---|---|---|
| Under head | Zinc flake + topcoat | 9 µm |
| Washer to plate 1 | HDG | 14 µm |
| Faying, plate 1 to 2 | HDG both sides | 22 µm |
| Plate 2 to washer | HDG | 14 µm |
| Nut bearing | Zinc flake + topcoat | 9 µm |
| Total | 68 µm |
Sixty-eight micrometres is 60,1 kN of clamp force, from an assembly preload of 70,0 kN. The joint is left holding 9,9 kN, 14 percent of design, with the nut exactly where you put it.
That is an extreme but entirely realistic specification. It is the reason galvanised structural connections have retorque requirements.
Why it interacts badly with torque control
Coating creep does something worse than shortening the stack. It changes the friction coefficient during the tightening event itself.
A lubricated zinc flake coating gives an underhead friction coefficient around 0,10 to 0,14. As the coating flows and thins during tightening, that coefficient shifts. Torque-controlled tightening converts torque to preload through the friction coefficient, so preload scatter on coated fasteners is characteristically wide, typically ±23 to ±30 percent against ±10 to ±15 for angle-controlled tightening on uncoated threads.
You therefore start from an uncertain preload and lose a large, poorly bounded fraction of it to creep. DIN EN ISO 16047 exists to characterise the first half of that problem. The second half is a design allowance.
Design responses
Specify electroless nickel where it matters. Heat-treated EN is effectively creep-free. It is the right answer for precision joints and the wrong answer for a thousand structural bolts on cost.
Keep coating out of the load path. Use uncoated hardened washers under the head and nut so the bearing interfaces are steel on steel, and accept coating only on the exposed surfaces. This is cheap and removes two to four of the interfaces in the table.
Reduce bearing stress. Creep rate is strongly stress-dependent. Larger washers, or a flanged head, can move a coating well down its creep curve.
Angle-control the tightening. Removes the friction coefficient from the preload equation and cuts scatter roughly in half. Does nothing about creep, but it means you know where you started.
Provide take-up. A reserve of clamp length spent against the creep budget holds clamp force where the other measures only reduce the loss.
| Method | Residual at 90 °C, 180 days, coated stack |
|---|---|
| ISOKLAMP CFR | 96,8 % |
| Uncoated hardened washers + wedge-lock pair | 74,1 % |
| Belleville stack | 68,3 % |
| Zinc flake throughout, plain washers | 22,6 % |
Predicted by finite-element analysis of the ISK-16 geometry coupled to VDI 2230 Sheet 1 mechanics with coating creep models fitted to published thickness-loss data. Baselines are calibrated against published data, sourced on Full residual clamp force dataset.
Sixty-eight micrometres against a 0,50 mm reserve is 13,6 percent consumed, leaving a comfortable margin and a readable indicator. See Preload reserve indicator.
Related loss mechanisms in Non-rotational preload loss and Locking methods compared.
Frequently asked questions
What is coating creep in a bolted joint?
Coating creep is the plastic flow of a surface coating under sustained contact stress. Zinc flake, PTFE-impregnated, hot dip galvanised and polymer coatings are all softer than the substrate and flow under the bearing stress of a preloaded joint, thinning by 5 to 20 micrometres and shortening the clamped stack. The nut does not rotate.
Which fastener coatings creep the most?
Hot dip galvanising is the largest single contributor at 8 to 20 micrometres at ambient and 15 to 35 at 100 degrees, mainly because the layer is 45 to 85 micrometres thick. PTFE-impregnated coatings run 8 to 18 micrometres. Zinc flake with a lubricant topcoat runs 6 to 12. Heat-treated electroless nickel is effectively creep-free at below 1 micrometre.
How much preload can coating creep cost?
A realistic worst case, an M16 class 10.9 zinc flake fastener through two hot dip galvanised plates at 90 degrees, budgets 68 micrometres across five coated interfaces. At 0,884 kN per micrometre that is 60,1 kN from a 70 kN assembly preload, leaving 14 percent of design clamp force with the nut untouched.
Why does coating affect preload accuracy as well as retention?
Torque-controlled tightening converts torque to preload through the friction coefficient, and a coating's friction coefficient shifts as the layer flows during the tightening event itself. Preload scatter on coated fasteners is characteristically ±23 to ±30 percent, against ±10 to ±15 for angle-controlled tightening on uncoated threads.
How do you design around coating creep?
Specify heat-treated electroless nickel where precision justifies the cost. Keep coating out of the load path by using uncoated hardened washers so bearing interfaces are steel on steel. Reduce bearing stress with larger washers or flanged heads, since creep rate is strongly stress-dependent. Use angle control so you know your starting preload. Where the budget is still large, provide a reserve of clamp length to spend against it.
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.