Intelligent monitoring and control system for molten metal drop-on-demand jetting by water-hammer effect

The molten metal drop-on-demand (DoD) jetting technology holds great potential for applications in fields such as electronics manufacturing and metal additive manufacturing. To ensure the stability and reliability of droplet quality, monitoring the jetting system state and controlling the jetting process are essential. However, existing studies primarily focus on conventional ink rather than molten metal, while monitoring the jetting system state and controlling the jetting process remain unintegrated. This study develops an integrated intelligent monitoring and control system for a self-developed water-hammer-based molten metal DoD jetting system. Firstly, the geometric features of droplets are collected and analyzed under varying liquid level heights and different driving waveform parameters. Subsequently, a jetting simulation model is established, and then it is used for sampling to train a prediction model for droplet volume reachable range, which can be used to monitor the state of the jetting system. Next, a multi-objective molten droplet control system is designed based on deep reinforcement learning, enabling the control system to simultaneously regulate the droplet volume and shape. Finally, simulations and experiments are conducted on the developed monitoring and control system. The results demonstrate that the monitoring system can accurately determine the current state of the liquid level. Furthermore, under normal liquid level height, the control system can achieve precise and rapid control over both the droplet volume and shape. This study contributes to improving the production quality of molten metal DoD jetting systems.