Electrical arcs are dangerous and can cause great damage such as electrical fire, therefore, detecting the arc faults for power electronic systems on aircraft, ship and automobile is now receiving considerable concern. To develop suitable detection and location schemes for arc faults, modelling arc faults to represent and predict arc characteristic and transient response within the power distribution system is essential. Series arc faults which are the majority cause of arc faults on aircrafts and vehicles are initiated by the degradation of contacts due to vibration . In this paper, DC series arc was generated by two different rigs: 1) an arc fault demonstrator driven by a stepper motor, 2) shaking table which emulates the vibration caused by aircrafts, vehicles or ships. The experimental results are presented and analysed to obtain the arc fault characteristics. The DC series arc model is extracted based on the experimental results performed on arc demonstrator and can predict arc behaviour with a wide range of supply voltage and load condition. The DC arc generated on the shaking table is more similar to practical arc faults as occurring on aircraft or ship electronic systems. The experimental results show the similarity between the shaking table arc behavior and arc model which contributes in the investigation, design and verification of arc fault detection and location strategies.