1、 crystallization; differential scanning calorimetry （DSC）; injection molding; poly（propylene） （PP）2 INTRODUCTIONOver a period of recent years, fluid-assisted injection molding technologies, including gas-assisted16 and water-assisted718 injection molding, have attracted more and more attention in re
2、searches and applications, due to that they offer potential to the production of complex, highly integrated parts in virtually one process step. Water- and gas-assisted injection molding （WAIM and GAIM） are similar processes and they both utilize high-pressure fluids with viscosities much lower than
3、 the polymer melt viscosity to displace the polymer melt within thick sections of the cavity. Just before demolding, the fluid is released from the core to provide a hollow but fully formed geometry. The GAIM is a well-known technology for producing hollow parts like media ducts, handles or clutch p
4、edals. The advantages of the WAIM have made this technology an interesting alternative to the GAIM over the last few years. In WAIM, water is used as the actuating medium. The main benefit of using water instead of gas is the much better cooling ability of water. The considerably more efficient cool
5、ing effect and high heat capacity of water result in a very efficient double sided cooling of the parts. Since the water is in direct contact with the molten polymer, cooling from the inside is even better than cooling from the outside by metal. A reduction in cycle times of about 5070% was achieved
6、 with WAIM, when similar plastics and part geometries were compared. This potential cycle time reduction shows the productivity improvement possibilities. Moreover, WAIM has been reported to form a thinner residual wall thickness than standard GAIM and to form channels around changes in direction th
7、at are more concentric than standard GAIM; the greater inertia of the water is thought to contribute to this effect.18 The incompressibility, low cost, and ease of recycling the water makes it an ideal medium for the process. So WAIM opens new possibilities for new and exciting applications. For exa
8、mple, automotive manufacturers have the potential for a broad range of cost savings with the production of fluid media piping and car door handles Despite the advantages associated with the WAIM process, the molding window and process control are more critical and difficult since additional processi
9、ng parameters are involved. These new waterrelated processing parameters include the amount of melt injection, water pressure, water temperature, and water injection delay time.13 Since this technique is relatively new and still in its infancy, in-depth research on the product design, machine design
10、, mold design, the control of process parameters, and the quality of the product has not yet been conducted. Certain processing parameters in WAIM play an important role in achieving a high quality hollow part. Moreover, to the best knowledge of the authors, no research paper has ever reported the c
11、rystallization behavior of water-assisted injection molded part.Figure 1 Schematic of experimental setup for waterassisted injection molding.For the above-mentioned reasons, in this work, four processing variables were investigated in terms of their effects on the water penetration length and residu
12、al wall thickness in water-assisted injection molded polypropylene （PP） curved pipe. Then Taguchi method was employed to optimize the processing parameters to maximize the penetration length. Finally, the samples, taken from two positions, near the beginning and the end of the water channel, respect
13、ively, were used to analyze the crystallization behavior of molded part. 3 EXPERIMENTALExperimental set-up and material A water injection unit, which mainly consists of a high pressure water forming and conveying unit （including its temperature controlling）, a water injector, and a control circuit,
14、was recently developed in this laboratory. The water injector used was an orifice-type. WAIM experimental investigations were carried out using the newly developed water injection unit combining with an 80-ton conventional injection molding machine, which has a highest injection rate of 84 cm3/s. Fi
15、gure 1 schematically shows the experimental setup for water-assisted injection molding. The mold used for experiments is schematically shown in Figure 2. The mold cavity is a curved one with a circular crosssection of / 18mm. The curved mold cavity has an unwound length of 387.5 mm. The temperature
16、of the mold could be regulated by a water-circulating mold temperature controller. In this work, a short-shot process was used, in which the mold cavity is firstly partially filled with the polymer melt followed by injecting water into the core of the polymer melt through the water injector. The wat
17、er penetrates into the core of the melt and pushes the melt again to the mold wall. Finally, the polymer cools down and solidifies and a hollow product is obtained. An injection molding grade PP, grade CJS700 from Petrochina Guangzhou Petrochemical, was used in this work. Its melt flow index is abou
18、t 8.35 g/10 min （at 2308C and 2.16 kg）.31 Experimental methodology：The present experiments were composed of two stages: a preliminary experiment and an orthogonal experiment. First, four processing parameters, including short-shot size, melt temperature, water injection delay time, and water pressur
19、e, were selected and a preliminary experiment was conducted by changing one of the parameters in each experiment. In doing so, it was possible to better understand the effect of every processing parameterFigure 2 Schematic of mold used for water-assisted injection molding. All dimensions are in mm.J
20、ournal of Applied Polymer Science DOI 10.1002/appTABLE IProcessing Parameters Used in Preliminary ExperimentsProcessingParametersShort-shotsize （%）Melttemp （C）Water pressure （MPa）Delaytime （s）68.51906373.22007477.621085-91011 on the water penetration length and residual wall thickness of the water-a
21、ssisted injection molded curved pipe. Moreover, the ranges of processing parameters could be determined by conducting a few experiments. Table I lists the values of these processing parameters. Preliminary experiment was performed by changing one parameter at a time and keeping the others constant （
22、highlighted in Table I）. That is, when changing one parameter in each experiment, the values highlighted were used for other processing conditions.Second, using the same four factors as those in the preliminary experiment, an orthogonal experiment was conducted to determine the best set of processin
23、g condition that maximizes the water penetration length. Three levels for each factor were chosen using Taguchi method as shown in Table II. So nine experimental runs based on the orthogonal array L9（34） as shown in Table III were conducted. The signal-to-noise （S/N） ratio was used to measure the qu
24、ality characteristic deviating from the desired value in Taguchi method. The water penetration lengths obtained in experiments were then analyzed to calculate the S/N ratio to optimize the water penetration. The maximization of the S/N ratio leads to the minimization of any property that is sensitiv
25、e to noise. A larger water penetration length is normally required in most industry cases because that the longer the water penetrates, the more the water pressure can pack the polymer melt against the mold wall. By maximizing the water penetration length, the shrinkage as well as the sink mark at t
26、he far end of molded parts can be minimized. Therefore, the larger-the-better characteristic of S/N ratio h was employed for the optimization of penetration length: where n is the number of experiments and yi is the measured property （water penetration length in this work） for the ith experiment.The
27、n the response table and figure of the S/N ratio were obtained, from which the significant factors could be identified. The combination of each factor at the level that produces the highest S/N ratio would be the optimum processing condition. Using this optimum set of factors could give the maximum
28、penetration length value in this design of experiment.To further examine the relative significance of the processing parameters on the water penetration length, a standard analysis of variance （ANOVA） was performed.During experiments, the control circuit in the water injection unit received a signal from the injection molding machine and controlled the time and pressure of the injected water. Five parts were molded for each set of processing conditions.32 Characterizations of molded partsThe length of water penetration in molded curved pipe was measured. The residual wal
