The maximum lifetime of a spark plug is limited by electrode erosion. Over the course of millions of repeated sparking events, the electrode material ablates and the electrode gap increases which degrades performance. Once a critical gap is reached, the spark plug is no longer operational and must be replaced. Due to the long relevant time scales over which erosion occurs, and the difficulty of analyzing the spark plug environment during operation, determining spark plug lifetime typically requires extensive field testing. The objective of this work is use a computational model that can accurately simulate the electrode erosion process and make predictions on the effective lifetime. The problem is challenging in that there are a vast range of time scales, all of which must be resolved to model the erosion. Time scales range from milliseconds needed to resolve arc physics up to days and weeks for time timescales of electrode deformation due to ablation. As a first step, dynamic coupling between arc physics and an ablating, eroding electrode is developed. An existing commercial arc solver code-VizSpark, capable of modelling the spark event with high fidelity, is used to model the arc physics which determine the net heat fluxes to the electrodes. The electrodes are modelled using an immersed object method, which allows them to dynamically change in shape as the simulation progresses. Mass flux from the electrodes due to ablation is modelled using temperature dependent metal vapor pressure curves. As mass is removed from the electrode-gas interface, the immersed object dynamically deforms, which in turn modifies the gap voltage and the arc physics. Overall, this approach provides predictive capability for the arc-induced erosion over the entire life-time of the spark-plug.