Study On Melt Near-field Direct Writing Injection Monitoring And Its Control System

Melt near-field direct writing is a new high-resolution additive manufacturing technology improved by electrospinning technology.In the fields of tissue engineering,MEMS and so on,melt near-field direct writing technology has been developed and improved by many researchers because it can deposit ultrafine fibers in a predetermined path on micro and nano scale.At present,melt near-field direct writing technology is also advancing to industrial application like other additive manufacturing technologies,but in large-scale industrial production,continuous and uninterrupted jet is the core guarantee of the quality consistency of this technology.At the same time,due to the existence of jet lag effect,the lack of fiber deposition accuracy has also attracted scholars’ attention.Aiming at the problem that it is difficult to predict the continuous formation of jet and fiber deposition trajectory,this thesis will build a set of melt near-field direct writing jet monitoring system.（1）This thesis analyzes the requirements of direct writing injection monitoring system,constructs the on-line monitoring and control system of melt near-field direct writing effect,and puts forward the overall scheme of direct writing injection monitoring system.The hardware includes the design of high-precision mobile platform,the component selection and debugging of industrial high-speed camera,accelerating electric field and light source.The software includes trajectory prediction algorithm,voltage regulation control algorithm and corresponding operation interface layout.This chapter divides it into six modules,explains the functions of each module respectively,and finally describes the work flow of the software.（2）For the continuity requirements of direct writing process.Firstly,the forming law of Taylor cone is explored,then the shape,size and jet of Taylor cone are monitored in real time by industrial camera,and the YOLO V5 algorithm is used to realize the monitoring mechanism of whether Taylor cone is formed and whether jet is interrupted.After image processing,the area of Taylor cone is extracted and analyzed quantitatively.An intelligent voltage adjustment algorithm is proposed: the Taylor cone feature is calculated by using the image processing algorithm and compared with the standard feature,so as to determine how to adjust the voltage,realize the spray printing automation and solve the problem of continuous and uninterrupted jet in industrial production.（3）According to the deposition prediction requirements of direct writing process,the trajectory instability in the process of near-field melt direct writing deposition is analyzed.The key characteristic information such as jet angle is obtained through machine vision,and the deposition path prediction geometric model based on lag length and moving platform trajectory is established.The applicable scope of the trajectory prediction model is explored.When the speed of the collection device is greater than the stable speed of the jet,the jet has a lag effect,and the trajectory prediction model is applicable.It can solve the problem of uncertain deposition trajectory in direct writing to a certain extent.（4）The system is verified and analyzed.It is found that Taylor cone area is an important feature that can be observed by melt near-field direct writing.The Taylor cone area is quantitatively analyzed to explore the influence of various process parameters on Taylor cone area and the matching relationship between Taylor cone size and fiber diameter.Verify the voltage intelligent adjustment algorithm,meet the requirements of uniform spinning diameter,verify the feasibility of the deposition trajectory prediction algorithm,take the micro nano waveform structure as the prediction graph,observe the experimental results and compare with the prediction results,which proves that the prediction algorithm is feasible.This project aims to solve the problems of jet interruption and deposition that are difficult to predict with high response,low cost and high reliability.Through efficient image processing algorithm,the change of Taylor cone section area in direct writing is detected,the matching relationship between jet stability and Taylor cone shape is studied,and the early warning model of melt near-field direct writing jet interruption is established by deep learning,which provides theoretical basis and technical support for the industrialization of high reliable batch production of melt near-field direct writing technology.