The dependence of arcing duration and energy in break operations for automotive relays was analysed with breaking current/voltage waveforms. Endurance tests were conducted with AgSnO2, AgNi10 and AgNi0.15 contacts under the loads of resistance, lamp and inductance, respectively, at 14 VDC. The experimental results shows breaking current (or break arc) duration is usually low and stable before welding occurs. The welding may appear suddenly or randomly without any preceding cumulated increases in the arc duration, the quantity of electric charges, nor the arcing energy. The welded contacts may be re-opened in the later break operation and, thereafter, can keep on working normally for many times.

In order to realize better understanding of influential order sequences of surrounding atmospheres on break arc durations of electrical contacts in DC load conditions, a quantitative mathematical model, which aims to indicate dependences of break arc durations on several gas parameters such as molecular mass, viscosity, specific heat capacity, thermal conductivity, electro-negativity, and ionization potential, was analyzed. Break arc durations of AgCdO contact pairs were measured in several kinds of surrounding atmospheres (N2, Ar, He, air, O2 and CO2) under different DC voltage and current conditions, and data fitting processes were conducted. As a result, a candidate mathematical model was established, which could indicate possible influential order sequences of surrounding atmospheres on break arc durations in the range of the tested conditions.

Application of transverse magnetic field (TMF) is one of the most important ways to improve electric life and breaking capacity of DC relays. For better understanding of dependence of arc durations on transverse magnetic field, a series of experiments were conducted under an external transverse magnetic field with 12 pairs of AgSnO2 contacts in a DC 28 V 60 A/30 A/5 A circuit, respectively. By using permanent magnets, the transverse magnetic field was obtained and the magnetic flux density at the gap center was varied from 13 to 94 mT. The results show that breaking arc duration is decreased monotonically with increases in the magnetic flux density, but making arc duration isn't decreased monotonically with increases in the magnetic flux density. In addition, both the magnetic flux density and the breaking arc duration have threshold values Bl and Tbmin, respectively, which means the breaking arc duration is almost stable with the value Tbmin even if the magnetic flux density is higher than Bl.

To clarify the dependence of arc duration on atmosphere, experiments were conducted under conditions of air, N$_{2}$, Ar, He and CO$_{2}$ with the pressure of 0.1,MPa in a 14,V/28,V/42,V circuit respectively. A quantitative relationship between arc duration and gas parameters such as ionization potential, thermal conductivity was obtained from the experimental data. Besides, the inherent mechanism of influence of atmosphere on arc duration was discussed.

The sputter erosion of arcing contacts is a very complex phenomenon, which is determined by the interaction between electromagnetic force, heat conduction and surface tension of liquid metal. A new model for evaluating the sputter erosion of electrodes is described in this paper, which is based on the electromagnetic forces against the molten pool, flowing velocity, kinetic energy and the surface tension of the molten pool. Erosion tests on AgSnO2, AgNi10 and AgNi0.15 contacts under the loads of resistance, lamp and inductance respectively at 14 VDC have been carried out. Experimental results indicate good agreement with the model's simulation. The model shows how the current and density, specific heat and other parameters of material affect the erosion rate.

To clarify how the occurrence of contact welding is related to the series of arc duration characteristics in consecutive make and break operations, electrical endurance tests were conducted on commercially available automotive relays, and the voltage waveforms of make and break arcs between the electrodes were recorded with LabVIEW. Experimental results indicate that welding may occur suddenly or randomly with increasing number of operations. A single arc or a group of make or break arcs with a long arc duration does not necessarily result in contact welding, but a group of longer make or break arcs within a narrow range of operation numbers can cause imminent contact welding (such an effect can be called the “group of longer arcing duration effect”). It is confirmed that contact welding may occur in both make and break operations, but the welding probability during make operations is much higher than that during break operations.