Contact erosion and arc damage: life drivers and design responses

What contact erosion is

Contact erosion is the loss or transfer of contact material caused by electrical arcing. It is different from mechanical wear, which removes material by friction and impact, and from corrosion, which changes the surface chemistry. In most switching contacts, erosion is the life-limiting mechanism.

Erosion takes two forms. Net loss removes material from the face and spreads it over the surrounding surfaces or into the arc chamber. Transfer moves material from one contact to the other, building a pip on one side and a crater on the other. Both change the gap, the face shape and the contact resistance.

A contact can fail from erosion in several ways: the face wears through to the backing, the gap grows beyond the mechanism's travel, the pip bridges the gap, or the roughened face causes sticking. The failure mode decides how much erosion allowance the design needs.

Where the arc energy comes from

The energy in an arc is the integral of arc voltage times current over the arc duration. Voltage and current set the power; duration sets the energy. A high-current arc that lasts a fraction of a millisecond can do less damage than a lower-current arc that lasts several milliseconds.

At break, the arc starts when the contacts part and the current continues through the gap. The duration is set by the circuit and the opening speed. An inductive load holds the arc longer because the stored energy keeps the current flowing. A fast opening mechanism stretches the arc and raises its voltage, which helps to limit the current and shorten the burn.

At make, the arc starts as the contacts approach and the gap breaks down, or as a bounce separates them again. A single make with no bounce produces a short arc. A make with several bounces produces several short arcs, each of which erodes the face. That is why bounce is an erosion issue as well as a welding issue.

AC and DC are not equivalent

An AC arc is extinguished at the current zero, so the duration is limited to a fraction of a half cycle unless the circuit restrikes. At 50 Hz, a half cycle is 10 milliseconds, and most contact gaps clear well before that. The arc energy per operation is therefore bounded by the supply frequency.

A DC arc has no current zero. It continues until the gap is long enough or the circuit resistance is high enough to extinguish it. The time constant of the circuit, inductance divided by resistance, sets how long the current takes to fall. A DC contact that carries the same current as an AC contact can erode far faster.

This is why a contact rated for AC should not be applied to DC at the same current without a check. The voltage, the time constant and the opening speed all need to be part of the rating. A DC rating at one time constant does not cover a more inductive circuit.

Pip and crater formation

At low arc energy, material tends to transfer from the anode to the cathode, building a pip on one contact and a crater on the other. The transfer is driven by the arc root and the molten metal flow. As the energy rises, net loss usually dominates and the face becomes rough rather than pip-shaped.

A pip is dangerous because it changes the gap and the face geometry. It can bridge a small gap, hold the contacts apart, or concentrate the force on a small area. A crater weakens the face and can expose the backing material, which has different erosion and welding behaviour.

The direction of transfer depends on the material, the current and the polarity. A change from AC to DC, or a change in the circuit, can reverse the direction. If the design has a preferred polarity, state it, and test in that polarity.

Material and geometry

Materials with a refractory phase resist erosion because the phase stays solid while the silver melts. Silver tungsten, silver graphite and formulated silver tin oxide are the usual choices for high-energy duty. Fine silver and silver nickel erode faster because the silver matrix melts and transfers readily.

Geometry matters as much as material. A domed face concentrates the arc root at first, then spreads it as the face wears. A flat face spreads the arc over a larger area but can trap material. An edge radius avoids a sharp rim that erodes early and distorts the field.

The contact thickness is the erosion allowance. A thin contact can wear through to the backing before the mechanism reaches its rated life, even if the material is good. The allowance should be calculated from the expected loss per operation and the required life, with a margin for the scatter in the data.

Sizing the erosion allowance

The calculation is simple in principle: allowable volume loss divided by volume loss per operation gives the expected life. The difficult part is the loss per operation, which depends on the load, the material and the mechanism. Use supplier data where it matches the application, and a life test where it does not.

As an illustration, a contact with a usable volume of 30 cubic millimetres and an assumed loss of 0.0015 cubic millimetres per operation would give about 20,000 operations. That is a made-up figure for the arithmetic, not a material property. A real estimate needs measured loss at the actual current, voltage and arc duration.

Add margin for scatter and for the end-of-life condition. The contact does not need to disappear to fail; it needs to reach a state where the resistance, the gap or the welding risk is unacceptable. Define that state first, then calculate the allowance to stay above it.

Measuring erosion

Mass loss is the simplest measure. Weigh the contacts before and after a run on a balance with enough resolution to see the change, and convert to volume using the density. The limitation is that transfer between contacts can cancel out in the total, so weigh each contact separately.

Profile measurement shows where the material went. A contact profilometer or a focus variation microscope can map the pip and the crater and give a volume change. Photographs at intervals are cheaper and often enough to show the trend, provided the lighting and the magnification are consistent.

Electrical measurements complete the picture. Contact resistance and voltage drop show the effect of the face change on performance. Arc duration and arc voltage, captured on an oscilloscope, show whether the mechanism is opening as intended. A life test that records only pass or fail misses the trend that predicts the next failure.

Design measures that extend life

The first measure is to reduce the arc energy. Faster opening, arc suppression and zero-cross switching all shorten the arc. The second is to reduce the number of arcs. Damping the armature and setting the force to avoid bounce removes the extra make arcs that quietly consume life.

The third is to choose a material and a thickness that match the duty. A refractory material with a thin face is still a thin face. The fourth is to protect the face from contamination. A sealed cavity or a controlled atmosphere reduces film growth and keeps the arc root stable.

Finally, test the actual load. A life test at the rated current with a resistive load is a starting point, but it will not show the effect of inrush, inductance or a cold filament. The useful test reproduces the circuit, the mechanism and the environment.

• Shorten the arc with faster opening or suppression.
• Remove bounce to cut the number of make arcs.
• Match the material to the arc energy, not to the current alone.
• Size the contact thickness from the life target.
• Seal or filter the contact chamber where contamination is a risk.
• Record mass, profile and resistance through the life test.