What locking washers do, and what they don't

Bolt-locking methods are usually compared on one axis: how well they stop the nut rotating. That hides the more important question. A joint loses clamp force two ways, and every product on the market addresses only one of them. Below, ten methods are compared on both: does it arrest rotation, and does it recover preload lost without rotation. The second column is empty for every incumbent, including the good ones.

The capability matrix

Ten bolt-securing methods, technical report §3. ISOKLAMP CFR figures are model-predicted; baselines are calibrated reconstructions of published data.
Method	Principle	Stops rotation	Recovers preload	Temp. ceiling	Re-use	Damages mating face	Principal limitation
Plain washer, unsecured	Friction only	No	No	Material limit	Yes	No	0 %
Split spring washer	Elastic spring back	No	No	Material limit	No	Yes — edge bite	DIN standards withdrawn 2004
Serrated flange nut	Serration bite	Partly	No	Material limit	No	Yes — severe	Destroys plating and coatings
Nylon-insert nut	Prevailing torque	Partly	No	120 °C	Limited	No	4,0 % residual
Double nut / jam nut	Thread preload	Partly	No	Material limit	Yes	No	23,4 % ± 14,0
Anaerobic threadlocker	Adhesive bond	Yes	No	150–230 °C	No	No	62,3 % ± 18,0
All-metal locknut	Deformed thread	Yes	No	Material limit	Limited	No	Galling risk on stainless
Thread-form nut	Proprietary thread	Yes	No	Material limit	Yes	No	85,3 %
Wedge-locking washer pair	Cam angle > thread lead	Yes	No	Material limit	Yes	Yes — radial teeth	92,1 % ± 1,4
ISOKLAMP CFR	Self-locking helical take-up	Yes	Yes	−196 to +650 °C	Yes	No	99,4 % ± 0,6

Every ISOKLAMP CFR figure above is model-predicted from closed-form bolted-joint mechanics and finite-element analysis. No physical testing of ISOKLAMP hardware has been performed. Baseline figures are reconstructions calibrated against published data; sources are listed on the test-data page.

Do wedge-locking washers prevent preload loss from creep and embedment?

No, and they do not claim to. A wedge-locking washer pair is the best rotational locking device in general industrial use: 92,1 % ± 1,4 residual under transverse vibration, n = 12. It recovers 0 % of relaxation, because the cam angle arrests rotation and relaxation does not involve rotation. NASA-STD-5020 defines the whole category this way: “safety devices principally intended to resist rotational loosening and prevent loss of fasteners”.

Does threadlocker prevent preload relaxation?

No. 62,3 % ± 18,0 under vibration — note the standard deviation, which is the real story. Cure is inhibited on passivated stainless, full cure takes 24 h, it has a shelf life, and it is single use. Where it cures properly it is a genuinely strong rotational lock.

Do nylon-insert nuts maintain clamp force?

4,0 %. They are a fall-off-prevention device and they do that job well and cheaply. They are not a preload-retention device and were never designed as one. Nyloc alternatives →

Do spring washers work?

0 %. The relevant DIN standards were withdrawn in 2004. The spring-back is small compared with the elastic stretch of the bolt it is meant to protect.

Do double nuts work?

23,4 % ± 14,0, and entirely dependent on assembly sequence. Assembled exactly to procedure — thin nut first, held while the thick nut is tightened against it — a double nut is effective. Assembled the way it usually is, it is not.

Do thread-form nuts help?

85,3 %, genuinely good on re-use, and a clean solution where you control the thread specification. It requires a proprietary thread and offers no relaxation compensation.

Every section above gives the incumbent full credit for what it does well. A comparison that concedes is a comparison that gets cited.

Questions

Do wedge-locking washers prevent preload loss from creep and embedment?: No, and they do not claim to. A wedge-locking washer pair holds 92,1 % residual clamp force under transverse vibration, which is excellent, and recovers 0 % of relaxation. NASA-STD-5020 defines the whole category as devices intended to resist rotational loosening.
Does threadlocker prevent preload relaxation?: No. Anaerobic threadlocker gives 62,3 % ± 18,0 under vibration; note the standard deviation. Cure is inhibited on passivated stainless, full cure takes 24 h, it has a shelf life and it is single use.
Do nylon-insert nuts maintain clamp force?: 4,0 %. They are a fall-off-prevention device, not a preload-retention device.
Do spring washers work?: 0 %. The DIN standards for them were withdrawn in 2004.
Do double nuts work?: 23,4 % ± 14,0, and entirely dependent on assembly sequence.
Do thread-form nuts help?: 85,3 %, genuinely good on re-use, requires a proprietary thread, and offers no relaxation compensation.