Current tide of industry 4.0 brings engineers and scholars a wave of sensing various process parameters, storing huge data in cloud server, and engaging data analysis. These developments enable the information of manufacturing process to be reused and facilities the monitoring and diagnosis of machines and devices, predicting machine life time, and then optimizing process parameters. Thereby, the goals of promoting product quality and production rate both can be further achieved. Although current sensing technology and the cost of sensors require more effects to fulfill industrial needs, the realization of broadly using sensors in injection molding machines and molds is definitely approaching soon. Consequently, understanding of injection molding process will be gradually far from a difficult situation of black or grey boxes. However, data mining to the big data heavily relies on the knowledge of polymer processing, lacked to engineers and technicians. Since the injection molding quality is complicatedly influenced with the variations of processed materials properties, mold design and manufacturing, process parameters setting, the auxiliary devices such as mold temperature controller, the realization of intelligent manufacturing for injection molding 4.0 requires the participation of polymer processing scholars to overcome the bottleneck of domain knowledge application. This study uses tie-bar elongation signals on quality control and parameter optimization. It consists of the comparison of tie bar elongation and cavity pressure signals during adequate switchover point; the development of velocity-to-pressure switchover control method; the experimental comparison of part weight consistence engaged with various switchover methods. Furtherly we apply the tie-bar elongation based velocity-to-pressure switchover control method on quality control and parameter optimization. It consists of the comparison of the proposed approach on thin-walled and thick-walled molding par