Threaded Fasteners vs Snap-Fits, Rivets and Adhesives: What the Evidence Can Settle Before You Specify

Can the joint deliver preload above the working load?

Before you commit to a threaded joint, confirm the design can actually build and control clamp load above the working load. Preload is the axial tension in the bolt before any external load is applied. It is what keeps the joint tight under vibration and operational forces ²⁰. When preload sits below the working load, the sources describe a predictable failure chain. Vibration and operational forces loosen the joint, the fastener fatigues, the joint separates, and shear slip follows ¹⁵. The same sources warn that more is not automatically better. Excess preload can yield or break the bolt, damage the clamped components, and facilitate stress corrosion cracking ¹⁵. Two assumptions are worth discarding early. The total load in a correctly preloaded joint is not the sum of working load plus preload. And fitting a stronger bolt at the same torque does not extend fatigue life. The stronger material is more brittle and can fail sooner ¹⁵. For high-reliability connections, ASME recommends stretch-controlled methods rather than torque alone ²⁰. Ask yourself whether your design can reach and verify preload above the working load. If it cannot, the threaded route is not yet specified, whatever the drawing says.

Which locking feature, and does it survive re-tightening?

Locking features are secondary aids to preload, not a substitute for it. They trade effectiveness against cost and reusability. Lock washers create tension between the bolt head and the surface. They are cost-effective and easy to install, but their effectiveness varies with the application and they may not prevent loosening entirely ¹⁴. Thread patches are pre-applied materials that raise friction between bolt and threads. They significantly reduce loosening and suit high-vibration applications. But they cost more, and most patched bolts cannot be tightened and re-tightened without the patch wearing off ¹⁴. That reusability limit is the deciding factor for any joint a technician will open more than once. The same source also points to bolt selection itself as the first step in preventing loosening. Match material and coating to the environment, grade to the stress, and use fine threads where vibration is expected ¹⁴. No source in this set quantifies vibration amplitude, frequency or duration, so you cannot size a locking feature against a stated vibration level. Ask the supplier what test data they hold for the feature you intend to specify.

Permanent or reusable — and does the joint need controlled access?

Reversibility is a choice you make inside the threaded family, not a reason to leave it. Security fasteners are supplied in both permanent variants, where removal is not intended, and reusable variants for applications where authorised maintenance access may be required ⁹. That distinction matters most where the joint is also an access point. An enclosure that a field technician opens for service needs a reusable drive. A joint that must resist unauthorised interference needs a permanent one. Decide which of the two your product actually needs before you fix the fastener family. The drive type, the tooling the service organisation must carry, and the tamper resistance of the joint all follow from that answer. If the joint must be opened repeatedly, the locking feature from the previous step has to be one that tolerates re-tightening.

Does the specification land in the hydrogen-embrittlement risk zone?

High-strength steel fasteners carry a hydrogen-embrittlement risk that has to be designed into the process, not inspected in at the end. Fasteners with a core hardness above 39 HRC (380 HV) or a nominal tensile strength over 1000 MPa sit in the high-risk zone. That includes property classes 10.9 and 12.9 and Grade 10 and 12 structural nuts ¹². Smaller diameters carry elevated risk because their higher surface-area-to-volume ratio accelerates hydrogen absorption during chemical processing ¹². The mechanism is a sequence. Hydrogen is picked up during cleaning and plating, and the coating then acts as a barrier that stops it diffusing back out at room temperature ⁵. For parts above 39 HRC (380 HV), non-acidic pre-treatment such as mechanical cleaning or alkaline de-rusting limits that pickup in the first place ⁵. ISO 898-1 carries a cautionary statement about stress corrosion cracking for property class 12.9 fasteners, which have a specified hardness range of 39-44 HRC ³. If your joint needs a 10.9 or 12.9 bolt, plan the pre-treatment and baking route with the supplier before the RFQ. Ask which controls they apply as standard.

How do you break the galvanic circuit in a mixed-material joint?

Where the fastener and the joined part are dissimilar metals, galvanic corrosion needs an electrolyte and electrical continuity. Break either and the couple weakens. An insulated joint stack of non-metallic washers, sleeves, spacers and bonded sealing washers breaks continuity and keeps moisture off the interface. The source describes this as potentially more effective than changing the fastener material alone ⁸. Isolation is not free of consequences. The components must withstand compression, temperature, UV exposure and chemical contact. If a bushing cracks or a washer cold-flows under preload, metal-to-metal contact returns ⁸. Stack-up changes from added isolation also shift clamp load and torque response, so the torque specification has to be revisited ⁸. Material matching is the other lever. Fasteners of the same material as the joined component, or close to it on the galvanic scale, are described as the most effective approach in most industrial applications. Aluminium fasteners in aluminium joints may not meet strength, wear or installation requirements ¹⁹. Coatings help but are not permanent barriers. Threads cut, bearing surfaces rub and drive recesses deform during installation, so the coating system belongs in the joint design rather than standing alone ¹⁷. Check that whatever isolation you specify survives the joint's compression and service environment. Confirm the torque figure still holds with the new stack-up.

How much of the joining cost is still open at design freeze?

Most of the assembly cost is decided before the drawing is released, so the part count and the joining method need to be settled early. A design-for-assembly review asks of every BOM item whether it needs to move relative to its neighbours, be made from a different material, or stay separate for service or assembly. If all three answers are no, the part can be consolidated ²³. Most fasteners, brackets, stiffeners and split housings fail that test. Consolidating them is reported to yield assemblies with 30 to 60% fewer parts ²³. The timing matters because roughly 80% of assembly cost is already committed by design freeze ²³. Removing parts after tooling is cut means reopening molds and revalidating fits ²³. Both figures come from a contract manufacturer describing the Boothroyd Dewhurst method, so treat them as that source's reported experience rather than an industry benchmark. For your own programme, the practical question is whether the joining method is fixed before the tooling order goes out. If it is not, the cost of changing it later is tooling and revalidation, not just a new part number.

Does the qualification schedule fit the programme?

Hydrogen-embrittlement qualification is slow, and the durations are long enough to move a launch date if they are discovered late. Check these against your programme before the specification is frozen.

• Sustained load testing runs a minimum of 200 hours at 75% of specified strength, and is used for qualification and lot acceptance ¹³.
• Incremental step loading is the accelerated option at 24 hours, used for development and process optimisation rather than lot acceptance ¹³.
• Analytical methods (ASTM G129, ISO 7539-7, ASTM F1624) are not suitable for production-scale part testing because of time and cost ⁷.
• For susceptible zinc-electroplated parts above 39 HRC (380 HV), baking runs 8-10 hours at 190-220 °C, and can require up to 24 hours depending on size and strength ⁷.
• A 4-hour bake at approximately 190 °C is described as insufficient for extracting hydrogen and possibly detrimental, so confirm the bake window your supplier actually runs ⁷.
• Confirm whether your specification requires baking at all — standard Grade 8 and property class 10.9 fasteners are stated not to need it, yet specifications such as ASTM B633 often require it unnecessarily ⁷.

Where the sources disagree on the same threaded-fastener decision

Two points in this set are stated inconsistently or in tension, and both are worth resolving before the specification is frozen. The first is whether baking applies to your grade at all. The second is how much inspection burden a galvanizing specification buys. Each row names the item, both positions, and what to verify.

FAQ

Related pages

• Rivets product page
• Surface finishes product page
• Cold Heading vs Machining for Fasteners: How to Choose a Route on Strength Class, Coating and Qualification Evidence
• Thread Locking for Vibration: Nylon Patch, Threadlocker, Safety Wire or Lock Washer — What the Evidence Can Rank
• Thread Pitch and Engagement Length for Steel, Aluminium and Plastic Joints: What the Evidence Settles