 Advances in Technology Innovation , vol. 2, no. 2, 2017, pp. 34 - 39 34 Synchronized Injection Molding Machine with Servomotors Sheng-Liang Chen, Hoai-Nam Dinh * , Van-Thanh Nguyen Institute of Manufacturing Information and Systems, National Cheng Kung University, Tainan, Taiwan Received 30 March 2016; received in revised form 25 April 2016; accept ed 01 May 2016 Abstract Injection Mold ing Machine (IMM) is one of the most important equipment in plastic industry. As a cyclic process, injection mo lding can be divided into three steps includes filling process, packing-holding process and cooling process, among wh ich filling process and packing pro- cess are both most important phases for the quality of part, and the corresponding crucial pro- cess variables are injection velocity and packing pressure in filling and packing phases. Moreover the determining a suitable injection time, screw position and cavity pressure for transfer from injection v e- locity control to packing pressure control which is commonly called filling to packing switchover point is also critical for high quality part. This study is concerned with two research aspects: double ser- vomotors synchronization control for injection unit, and filling to packing switchover methods. The simulation result of switching method based on injection time and ball screw position those are similar, and the result of switching method based on the cavity pressure that is better. Keywor ds : s ervo motors, synchronization, v e- locity to pressure switchover, pres- sure control, injection molding machine. 1. Introduction Injection mold ing mach ine used servo motor to save energy and suit for prec ise products. Plastic In jection mo lding an e xtensive range of modern injection mo lding machines means that most custom mold ing require ments can be met. Plastic inject ion mo ldings will be manufactured using the most cost effective production methods . The injection process included heating process and injecting the material into the mold [1]. The process produces are very quickly with great accuracy. It is widely accepted that all-electric injection mo lding machines are the most energy efficient of the three technologies [2]. In [3], there analysis and give some solution about cla mping system. Hydrau lica lly driven injection translates the movement offering high force combined with high speed or electric servomotors has created the groundwork for realizing linear electromechanical screw movement with stronger force and more accurate. In this research, we focus only 2 phases, filling phase and packing phase, each of them has distinct control require ment and how to switchover fro m one to other. In the filling phase, the position and velocity of the in jection screw are controlled ensure the correct me lt front v e- locity. At the beginning of the packing phase, the cavity pressure is controlled at a higher boost pressure to ensure correct part weight. A fter that, the cavity pressure is controlled at a lower pressure to maintain the part quality and avoid over packing. It ends after the gate freezes off. 2. Methodologies and System Mod- elling Design 2.1. Methodologies Control of an in jection mold ing machine consists of many aspects. However, in this re- search we focus on designing controller for injection unit and techniques in filling phase and packing phase. The required injection force is so large but s ervo motor is e mployed that means designing control system for inject ion unit is e xtre me ly important. Even an injection mo lding mach ine with perfect in jection unit control sys- tem but that does not mean the product quality is assured. As discussed previously, to achieve high quality product, the process variable, cavity pressure is controlled. And the most difficult but most important in controlling cavity pressure is how to switchover fro m filling phase to packing phase. When the required inject ion force o r pres- sure is so large but because of energy saving, cost effective, environ ment pollution and fle xi- * Corresponding aut hor, Email: dinhhoainam52@gmail.com Advances in Technology Innovation , vol. 2, no. 2, 2017, pp. 34 - 39 35 Copyright © TAETI ble, the servo drive is a good choice for the injection unit drive system. However, the servo drive has power limitation, if the capacity of the injection mold ing machine is huge so one servo motor cannot supply enough power and it costs a lot. That is why double servo motor is emp loyed in the inject ion unit drive system. Ho w to ma ke them go smoothly with the same mot ion is one of the research objectives and the other is in - vestigat ing so me methods to switchove r be- tween filling phase and packing for our applica - tion. The specific objectives of the present work are:  Design a control system to control two servo motors move with the same motion. Double servo motors drive the in jection unit and they are controlled in order to achieve a high st a- ble large injection force or pressure.  Investigate some methods of switchover filling controlling and packing controller. 2.2. System Modelling Design Fig. 1 The schematic of the injection molding control system [4] The present work uses PC-based control system. The servo drives communicate with the controller via Ethernet. The controller sends command to the servo drives and receives the encoder feedback signals from the servo drives. The cavity pressure is measured by a load ce ll and it co mmunicates with the controlle r v ia RS232. The data are monitored on the screen and can be collected by input output programming. The control algorith ms and the human machine interface are written in visual C # 2005, as shown the overall in jection molding control system in Fig. 1. Servo motor system identification and characterizat ion: the first step in the system identification stage is collecting data that are input and output in operat ing system. Then sys- tem identification is done by using the system identification toolbox in Matlab. The transfer functions in z do main a re obtained. The systems here consist of servo drive and servo motor, and the velocity controller is integrated in the driver. So, the transfer functions are the velocity closed loops. System identificat ion verificat ion: verify the system identification result by com- paring simu lation and e xperiment. That means with the same co mmands one passes by the p hysical system the others passes by the transfer function, and then compare the outputs . Control system design: the transfer functions of two servo motors are obtained; control theory is applied to design the control system. More specific, first, design individual motor controlle r and then design controller for double servomotors control system. All steps are simu lated in Matlab and Simu lin k. After that, the controlle r a lgo- rith ms are written in co mputer progra ms, the performance of the controllers are tested, and the controller para meters can be tuned and adjusted. With acceptable synchronous error, two servomo- tors are connected to a timing belt and this sys- tem will drive a ball screw. 3. Double Servomotors Control System 3.1. Control System Design According to the different para mete rs of system are input, are calculated and converted fro m control signals to drive motors, then ap- plied to machine. Four servo motors control moving p laten slip on ma in t ie bar. Servo motor as drive in jection machine, servo motor speed stability and high prec ision position control, in addition to control effect of a llowing high re - producibility can also fine-tune the speed control and position control. Fig. 2 Structure of motors block control in injection machine Advances in Technology Innovation , vol. 2, no. 2, 2017, pp. 34 - 39 36 Copyright © TAETI The purpose of motion control system is used to calculate the random variable consists of slide motion, velocity. Deve loping state-space form, the governing equations can be written by using the rotational speed and electric current are the state variables. u L i L R L K I K J B idt d m mm m                                   1 0 (1) 𝑦 = [1 0] [ 𝜔𝑚 𝑖 ] (2) Where L is the winding inductance, R is the winding resistance, B is damp ing coeffic ient, i is the current, u is the armature voltage, K is the torque gain, 𝜔𝑚 is the rotational ve locity, J is the armature (rotating part of motor) mo ment of inertia. Some e xperiences have done, we find the transfer function of two servo motors with 2000rp m, and position outputs are measured. The transfer functions from veloc ity command to velocity output of two A C servo motors can be expressed in Laplace variable s. 243.002812.0 06412.062.14 21    ss s G (3) 4279.06198.0 708.123.14 22    ss s G (4) The cross coupled control system min imizes synchronous errors in mu lt i-a xis mot ion control system [5]. It is co mmon used in CNC motion control system. Technology of cross couple control is developed for mu lti a xis [6]. The par- alle l technique control in mult i paralle l system was proposed [7]. The technique used feedback signal as positional signal and veloc ity signal to modify the command signal. This control has simp le structure, and easy to imp le ment. Ho w- ever, some possible effects can decrease the performance of this control method, for e xa mp le, unmatched sinusoidal disturbances; reduc- tion degrades the stability, unmatched model. Utilize the advantages of the control methods above, combination between them was made. The new synchronous controller is mentioned in [8]. Th is controller can overcome the difficulty of the master slave control method and utilize the advantages of the synchronous controller and the relative dynamic stiffness motion control. We have realized that cross -coupled control is used to increase the coupling relationship between axes. It a lso ma kes the capacity of disturbance better, decrease the error model. However, the measuring performance cannot improve in each axis. Enconder 1/s Controller1 C1(s) Filter1 Motor 1 G1(s) Angular velocity Angular position+ - Enconder 1/s Controller1 C2(s) Filter2 Motor 2 G2(s) Angular velocity Angular position + - Position average Fig. 3 Block diagram of cross -coupling controller of two motor The errors of coupling axis are determined by the difference between the angular position and the average angular position. To trac king the same tra jectory input, where appropriate filter s . In this case the control la ws adopted for a ll system 4. Pressure Control and Velocity to Pressure Switch Control 4.1. Velocity to Pressure After two servomotors are synchronized , we connect them to a ba ll screw by a timing belt in Fig. 4 [9]. Assume that, the gear ratio is unity, the pitch radius of the ball screw is equal to the shaft radius of the motor, and the thread lead of the ball screw is 5mm. So now, the motion command of the ball screw can be computed fro m the motion co mmand of the motor. The ball screw position x and the ball screw linear ve loc ity v are expressed by the following equations . Fig. 4 Schematic of the injection unit drive system [9]   360 h x mm (5) Advances in Technology Innovation , vol. 2, no. 2, 2017, pp. 34 - 39 37 Copyright © TAETI   / dx v mm s dt  (6) In which, θ is the motor position in degree, and h is the thread lead of the ball screw in mm. When ball screw move, the cavity pressure will be changed when the mold is closed. Assume we can measure the cavity press by a load ce ll. But the difficulty is how to ca ll ca lculate the pressure command. In this study we don’t jump into the theoretical way, for e xa mple , with a given ma - terial, mo ld, and para meters of the mechanisms we have to calculate the pressure command via mathe matics way. But we will use a practica l approach. Because the servomotors are stet-upped in the velocity control mode that means the final co mmand to the driver is ve lo c- ity co mmand, so the pressure command has to be transferred into the velocity co mmand. The relationship between them is found by finding the relationship between the pressure feed back and the velocity feedback. The velocity of the ball screw is computed fro m its position by taking a discrete derivative. The procedure to connect pressure to velocity is shown in Fig. 5. Fig. 5 Conversion of pressure into velocity procedure 4.2. Pressure Control Cavity pressure measurement is crucia l for process parameter. The flo w o f the plastics is increase in high pressure injection. In lo w pressure case, the flo w of the p lastic also reduces . And the cavity pressure during the packing phase is the key variab le to ensure the co mplete filling and the we ight of the final product. But we can see the performance of the pressure controller in a whole cycle. A pressure controller is designed to make sure the output cavity pressure follo w the co mmand cavity pressure command. The closed loop pressure is show in Fig. 6. The pressure controller is a simp le p ro- portional controller and the proportional gain is tuned by try-and-error method. Fig. 6 The closed loop pressure 4.3. Switchover from Velocity to Pressure Filling to pac king switchover is known as velocity to pressure switchover, where velocity refers to injection velocity and pressure to packing pressure. In the filling phase, the v e- locity of the inject ion screw is controlled ensure the correct me lt front velocity. At the beginning of the packing phase, the cavity pressure is controlled at a higher boost pressure to ensure correct part we ight. After that, the cavity pres- sure is controlled at a lowe r pressure to maintain the part quality and avoid over packing. It ends after the gate free zes off. The ve locity to pre s- sure switchover structure is shown in Fig. 7. The most important in velocity to pressure switchover is how to locate the switching point. If switchover occurs too late, cause an over-packed cavity, characterized by a pressure peak in the packing phase. The pressure peak will not reduce to the lower holding pressure until the switchover because the high injection pressure is still applied after volu met ric filling. Switching over early may g enerate an un- der-packed cavity, characterized by a pressure drop in the packing phase. Load cell Function change unit from pressure to velocity Kp Driver position velocity pressure Fig. 7 Flow chart of inject processing machine As we know that the reference input co m- mand is velocity, but output is pressure. There- for we must find the relation between velocity and pressure. Following [6], re lation between displacement, a xia l motor input differentia l pressure and flow rate is founded as 1 2 1 ( ) [ ( ( )) ( ) 2 ( ( )) ( ) ] Q t g y t Ps P t g y t Ps P t     (7) where Q(t ), P(t) a re flow rate and d ifferentia l pressure of motor input, Ps is source pressure, g1(y(t)), g2(y(t)) are conductance of all edges of value calculates by Advances in Technology Innovation , vol. 2, no. 2, 2017, pp. 34 - 39 38 Copyright © TAETI   1 2 ( ( )) ( ( )) 0 ( ) g K y t B if y t B g y t if y t B         (8) 2 2 ( ( )) ( ) ( ( )) 0 ( ) g K y t B if y t B g y t if y t B        (9) 5. Simulation and Experiment Re- sults The present work uses PC-based control system. The servo drives communicate with the controller via Ethernet. The controller sends command to the servo drives and receives the encoder feedback signals from the servo drives. Hardwa re is use in this study is provided by Fo xnum Co mpany. It includes controller, ser- vomotors and servo drivers . Servo driver can connect to controller via a co mmun ication card COMM3. The simu lation stage follows the control system design stage. We show the s imulation results step-by-step, the error of master and slave motors in Fig. 8. W ithout synchronous controller, the error will increase. In fact, even we select the same motors, but it will happen. With synchro- nous controller, the error is co mpensated. That makes the error reduce so much. The performance of the closed loop pressure with the proportional gain Kp = 2.0 is show in Fig. 8. At the beginning of the cycle , the output pressure cannot track the pressure command but when time goes on nearly 2seconds, t he output can track the co mmand. It is easy to recognize that in the injection phase the system is a non-linear system that is why only a simp le proportional controller cannot deal with. Ho w- ever, our purpose is to control the cavity pres- sure during the packing phase, and at this in- terval, time is la rger than 2 second, the out pressure can follow the pressure command. If the system is non-linear in any phase, especially in packing phase, we must take its properties into account. Until now, we can just only ma ke sure two motors can tracking the command. How is about the synchronous position error between the m? We cannot predict it in the physical system, because when some controlle rs are e mployed, they will take action every t ime to co mpensate the errors. Even the feedback controlle r can eliminate the steady sate error, the feed forward controller can reduce the tracking error, but at the transient state, two motor response very diffe rently. Using synchronous controller to decrease the synchronous error of motors Fig. 8 Performance of the closed loop pressure Fig. 9 Synchronous error with synchronous controller and without synchronous controller Fig. 10 Synchronous position error, without synchronous controller and with syn- chronous controller Fro m Fig. 10 we get the co mparison the Maximu m absolute of synchronous error between using with synchronous controller and using without synchronous controller. Fig. 11 Trac king position error of motor 1 and motor 2 Advances in Technology Innovation , vol. 2, no. 2, 2017, pp. 34 - 39 39 Copyright © TAETI We may consider now, when the synchronous controller is added into the system, it has effect on the performance of the overall control system or not? So, we have to study about some pro p- erties of the overall system. Two important properties of a control system are time response and tracking erro r. So, we can see time response of motor 1 and motor 2 in Fig . 10 and Fig. 11. The position errors of two motors have a little suddenly values. The overall control is better than using open loop controller. Table 1 The value of synchronous position error System Without synchronous controller With syn- chronous controller Maximum ab- solute of syn- chronous error 2.0304 0 1.2300 0 6. Conclusions Th is research focuses on two subjects: (1) design and imple ment control a lgorith m for double servo motors and their application to the injection unit control system. In particu lar, the present work utilizes these advanced algorithms on the injection unit that is driven by double servo motors; and (2) investigate some methods to switchover fro m filling phase to packing phase for our application. The acco mplishments in this research involve modelling, control a l- gorith m i mp le ment at ion , s imu la t ing , and e xperimental imp le mentation. Control system can be made more intelligently with artific ia l intelligence techniques. The advent of low cost very large scale integrated microprocessor and the proficiency of hard ware, control algorithms can be implemented in the hardware to get high response and performance. References [1] “Injection Molding” http://www.custompar tnet.com, 2009. [2] A. Kanungo and E. Swan, “A ll electric injection mo lding machines: How much energy can you save?” Proceedings fro m the Thirtieth Industrial Energy Technology Conference, Ne w Orleans, LA, USA, May 6-9, 2008. [3] F. Johannaber, “Injection mo lding machines: A user’s guide,” 3rd ed. Hanser Munich Vienna: New York, 1994. [4] “F70i injection mo lding mach ine control sy stem” http://www.foxnum.com/Product021_en.aspx [5] Y. Ko ren,, “Cross -coupled bia xia l co mputer control for manufacturing systems ,” Journal of Dyna mic Systems, Measurement, and Control, vol. 102, no. 4, pp. 265-272, 1980. [6] MarVin H.Cheng, Cheng-Yi Chen and Aniruddha Mitra, “Synchronization controller synthesis of mult i-a xis mot ion system,” In- ternational Journal of Innovative Co mputing Information and Control, vol.7, pp. 918-921, July 2011. [7] O. S. Kwon, S. H. Choe, and H. Heo, “A study on the dual-servo system using im- proved cross -coupling control method,” 10th International Confe rence on Environ- ment and Electrical Engineering (EEEIC’11), May 2011, pp. 1-4. [8] M. F. Hsieh, C. J. Tung, W. S. Yao, M. C. Wu and Y. S. Liao, “Servo design of a ver- tical a xis drive using dual linear motors for high speed electric discharge machining,” In- ternational Journal of Machine Tools and Manufacture, vol. 47, no. 3-4, pp. 546-554, 2007. [9] N. A kasaka, “A synchronous position co n- trol method at pressure control between mu lti-ac servomotors driven in inject ion mo lding machine,” in SICE 2003 Annual Conference, vol. 3, pp. 2712 -2719, Aug. 2003.