Fastener Selection for Aerospace and Automotive Hardware
A bolted joint fails in one of two ways. The bolt breaks, or the joint separates because the clamp load was never there. The second case is far more common and much harder to see, because the fastener looks fine right up until the bracket starts fretting and the hole goes oval. Fastener selection is therefore a question about preload, not about strength alone.
Clamp load is the design objective
A bolt tightened to 70 percent of its proof load holds the joint closed against vibration, thermal cycling and transverse load. Below that, the joint can slip, and once it slips the bolt sees shear and bending it was never designed for. Above about 90 percent, the fastener is close to yield and relaxes in service as the coating beds in. The window is narrower than most drawings suggest, which is why torque values should be calculated for the specific joint rather than copied from a table. Lubrication changes everything: the same torque can produce 30 to 40 percent different clamp load depending on whether the threads are dry, oiled, or coated with a zinc flake system.
Reading the property class
Under ISO 898-1, a class 8.8 bolt has a minimum tensile strength of 800 MPa and a yield strength at 80 percent of that, while 10.9 and 12.9 climb to 1,040 and 1,220 MPa. The higher classes are not automatically better: they are more sensitive to hydrogen embrittlement after plating, less forgiving of poor hole alignment, and they transfer more load into the joint rather than distributing it. Automotive chassis work commonly settles on 10.9. Aerospace structural joints often use fasteners with specified fatigue endurance and controlled interference fit rather than the highest available tensile number.
Material pairing and galvanic risk
Stainless fasteners in aluminum structure are a classic mistake in wet environments. The potential difference drives corrosion of the aluminum, and the joint seizes within a season. Cadmium and zinc plated steel fasteners in aluminum are less severe but still require a barrier. Titanium fasteners in carbon fibre structure need isolation to prevent galvanic attack of the aluminum fittings nearby, and stainless-on-stainless assemblies gall badly unless one side is coated or the alloy is chosen for galling resistance. Weight-conscious programs increasingly use titanium or aluminum fasteners in secondary structure, where the rule is to check both the strength and the thermal expansion mismatch.
Locking methods, ranked by what they actually do
Threadlocker adhesive fills the thread clearance and prevents relative motion, which is effective up to a temperature limit that is often lower than the label suggests. Prevailing torque nuts with a deformed thread work over a wide temperature range and survive reuse better than adhesive. Wedge-locking washers resist vibration by forcing the bolt to climb rather than rotate, and work well on hard surfaces. Split washers do very little, and the evidence for them has been thin for decades. Nylon insert locknuts lose their grip above roughly 120 degrees Celsius. None of these methods substitute for adequate preload. A locking feature on a loose joint only keeps the loose parts together.
Assembly tooling closes the loop. A torque wrench calibrated to plus or minus 4 percent, applied to a joint in its specified lubrication condition, produces a predictable clamp load. An impact driver on the same joint produces something else entirely, usually more scatter and sometimes a stretched shank. Where a joint is safety related, specify the tool, the torque, the verification method and the audit frequency, then record the results against the build serial number so a field question can be answered from the file rather than from memory.
Holes, burrs and fatigue life
Fatigue cracks in bolted assemblies usually start at the hole edge. A burr under the washer creates a stress concentration and a gap that lets the joint fret. A hole drilled 0.2 mm oversize reduces the bearing area and allows slip at lower load. Deburring both faces, controlling hole position to within 0.1 mm, and using washers with adequate thickness for the clamp load are cheap measures with large effects. In aerospace work, hole quality is inspected and recorded; in general industrial work it is often assumed, which is where the difference in service life comes from.
Counting fasteners as a cost line
On a vehicle or an airframe, fasteners are numerous enough to matter as a category. Standardising on a small set of lengths and diameters reduces inventory, reduces assembly error and improves purchasing leverage. It also makes torque tooling simpler. Design reviews that treat fasteners as an afterthought end up with 14 different part numbers where four would do, and a line-side picking error rate to match.
Verifying a fastener supplier
Ask for the mill certificate and the lot number on the carton, and confirm they match. Confirm the property class marking is legible and correct, since remarked fasteners are a persistent problem in some markets. For plated high-strength parts, ask for the baking record. Then ask for a torque-tension test report, per ISO 16047, which shows the clamp load achieved at the specified torque for that coating system. An automotive or aerospace fasteners manufacturer will have this data on file for every finish they supply.
References
ISO 898-1:2013, Mechanical properties of fasteners made of carbon steel and alloy steel, Part 1: Bolts, screws and studs with specified property classes. International Organization for Standardization, Geneva.
SAE J429, Mechanical and Material Requirements for Externally Threaded Fasteners. SAE International, Warrendale, PA.
VDI 2230 Part 1, Systematic calculation of highly stressed bolted joints, Joints with one cylindrical bolt. Verein Deutscher Ingenieure, Dusseldorf, 2015.
ISO 16047:2005, Fasteners, Torque/clamp force testing. ISO, Geneva.
Bickford, J. H., An Introduction to the Design and Behavior of Bolted Joints, 3rd edition. Marcel Dekker, New York, 1995.
ISO 4042:2018, Fasteners, Electroplated coating systems. ISO, Geneva.
