Mixed-material assemblies: thermal expansion and corrosion

Why mixed-material joints are hard

Most electrical assemblies are mixed by necessity. Aluminium is light and cheap, copper conducts better, steel is strong and cheap, and plated fasteners are available off the shelf. The combination solves one problem and creates two: the metals form a galvanic couple when an electrolyte joins them, and they expand by different amounts when the temperature changes.

The two problems interact. Corrosion can seize a fastener, which then cannot slide when the joint moves. Movement can crack a coating or a sealant, which lets the electrolyte in. A design that addresses only one of them tends to fail through the other.

The galvanic couple

When two different metals touch in the presence of an electrolyte, the less noble one corrodes. The driving force is the potential difference between them. Aluminium against copper is a large difference; aluminium against stainless steel is also significant; steel against zinc is the basis of galvanising and is used deliberately.

Three factors decide how fast it happens. The first is the potential difference, which comes from the pair. The second is the electrolyte: distilled water is nearly harmless, salt water is aggressive, and a humid atmosphere with chloride or acid contamination sits in between. The third is the area ratio. A small anode against a large cathode is the worst case, because the corrosion current concentrates on a small area.

That area ratio rule explains a common failure. A small aluminium rivet or a thin aluminium washer against a large steel plate corrodes quickly, while a large aluminium plate with a small stainless screw lasts much longer. The design should aim for a large anode and a small cathode, or break the electrical path.

Plating helps but is not a guarantee. A zinc or nickel coating on a steel fastener protects the steel until the coating is damaged, then the exposed steel can become the cathode and the aluminium the anode. A damaged coating can be worse than no coating if it creates a small exposed area.

Differential expansion in numbers

Thermal expansion is the change in length per degree of temperature change, expressed as a coefficient. Typical values are about 23 parts per million per kelvin for aluminium, 17 for copper, 12 for steel and 70 or more for an unfilled polyamide. These are indicative, and the exact value for an alloy or a filled polymer should come from the material data sheet.

The absolute movement is small, but the differential is what loads the joint. For illustration, take a 300 mm aluminium bar bolted to a steel bar and a temperature rise of 60 K. The aluminium grows by about 0.41 mm and the steel by about 0.22 mm, so the joint sees roughly 0.19 mm of relative movement. Over a 40 mm bolt spacing that is a significant shear strain, and over a short rigid bracket it can be enough to yield the fastener or crack the housing.

The numbers are illustrative, but the conclusion is general. Long parts and large temperature swings produce more movement. Stiff joints resist it, so the load goes into the bolts, the welds or the plastic. A joint that is stiff and brittle in the wrong place will eventually crack.

Polymers add a second effect: their expansion is larger and their modulus falls as they warm. A metal insert in a plastic housing can loosen when the plastic grows away from it, then grip again when it cools. That cycle can wear the interface and reduce the pull-out strength.

Joint details that survive the cycle

Give the joint a direction. A slotted hole, a floating nut, a shouldered screw or a compliant washer lets the parts move without loading the fastener in shear. The slot should be sized for the calculated movement plus the tolerance, and the washer should be large enough to bridge it without digging in.

Keep the clamp force stable. A joint that relies on friction will lose clamp force as the materials creep or relax, and then it will start to move. A disc spring or a Belleville washer can hold the force over a wider range, provided the joint is designed for the washer and the torque is set for the spring, not for a plain washer.

Separate the functions. The fastener can locate, clamp or conduct, but asking it to do all three in a mixed-material joint is difficult. An insulating sleeve can take the location, a washer can take the clamp load and a separate bonding strap can take the current. That separation makes each part easier to design and inspect.

Allow for the polymer. A metal insert in a plastic housing needs a wall section that can take the expansion, and the insert should have a mechanical grip such as a knurl or a groove. If the insert is moulded in, the plastic shrinks onto it; if it is pressed in after moulding, the interference has to survive the temperature range.

Two common pairs and how they behave

The table compares the two pairs that cause most of the problems in electrical assemblies. Both can be made to work, but they need different details.

Isolation, coatings and drainage

The most reliable way to stop galvanic corrosion is to break the electrical path. An isolating washer, a plastic sleeve, a coated fastener or a sealant can do it, provided the isolation survives assembly and service. A painted interface is not isolation if the paint is scratched by a washer.

Where isolation is not possible, control the environment. A drain hole stops water collecting in a pocket. A sealant keeps the electrolyte out of the faying surface. A generous fillet or a gasket can keep the joint dry. The aim is to remove the electrolyte, because without it the couple is inactive.

Coatings need a defined thickness and a defined damage tolerance. A plated fastener with a thin coating may corrode at a scratch. A conformal coating over a joint may crack when the joint moves. Specify the coating, the thickness, the adhesion and the test, and inspect the coating at the edges where it is thinnest.

Choose compatible fasteners. A stainless fastener in an aluminium plate is common, but it creates a strong couple if the joint gets wet. A zinc plated steel fastener is less noble than the stainless but protects the aluminium sacrificially for a while. Neither is universally right; the environment decides.

Validating the assembly

Test the assembly as a system. A thermal cycle between the service extremes, with the joint under its working load, shows whether the clamp force holds and whether the seal or coating cracks. A salt spray test to a recognised method shows the corrosion pattern, but it compresses years into weeks, so use it for ranking and not as a life prediction.

Measure the things that matter. Torque before and after cycling, resistance across the joint, and a visual check of the interface. A joint that passes a pull test but loses torque may still fail in service, because the failure is movement, not strength.

Section the joint after the test. A cross-section shows the corrosion at the interface, the condition of the plating and whether the isolation survived. It is often more informative than a pass or fail result.