A simple method for tuning the temperature control parameters for the die and cylinder of a plastics extruder is investigated. This method is based on a fuzzy control system. Recently, an increasing demand for small lot size manufacturing has led to frequent and rapid changes in the manufacturing process. This requires temperature controllers to find the appropriate control parameters quickly with an auto-tuning method. Unfortunately, most commercially based PID controllers apply the so called “Limit cycle method” to tune the appropriate control parameters. This method takes a relatively long time to find the parameters, because it requires several on/off cycles of heater output to form a temperature wave in order to idenify the process characteristics. Furthermore, this method discards the first peak of the temperataure wave, thus wasting time. The first peak is usually higher in magnitude than succeeding waves, but this characteristic can not be explained as long as we assume that the process is a first-order system plus a transportation lag. In this paper, assuming that the process is a second order system, we can explain the characteristics of the wave form theoretically. We find that the magnitude of the peak at an appropriate set temperature can be estimated using the magnitude of the peak of the tuning temperature, regardless of the value of the time constant and gain of the process. We also generate a simple fuzzy rule that can utilize the magnitude of the first peak as a control parameter. Experimental results show that this method is practical enough for its accuracy and robustness, and can decrease the tuning time by one third compared to that of conventional PID controllers.

View full abstract

A devolatilization model for polymer melt in a multi-vent screw extruder has been proposed. In this model an overall film-interface geometry factor (OFIGF) comprised of local film-interface geometry factors (LFIGFs) is used as a measure of devloatilizing performance of a screw extruder. Moreover, the OFIGF is related to screw geometry, and opening angles of the open areas in vent sections and dependent on degree of filling or flowrate and screw speed. A solution was given how to calculate the film-interface areas and their exposure time in a venthole unfitted zone (VUfZ) without an open area and a venthole fitted zone (VFZ) with an open area for intermeshing co-rotation twin-screw extruders, and the OFIGFs for the two cases mentioned above were predicted under the conditions found in screw geometry and the opening angle of open area. The results suggested that devolatilizing performance of VFZ is lower than that of VUfZ due to the decrease in surface renewal effect, and a plot of ln{(w_out-w*)/(w_in-w*)} against N^1/2/Q does not strictly satisfy with a linear relationship in the general range of flowrate and screw speed owing to the dependencies of the OFIGF on them and those dependencies should be further considered for predicting the devolatilizing performance, and designing and scaling-up screw extruders.

View full abstract

The devolatilizing performance of a long vent section in an intermeshing co-rotation twin-screw extruder manufactured by The Japan Steel Works was measured by using an ethylbezene/polystyrene mixture. The experimental data were fitted by using the previous devolatilization models and a new model proposed by the authors. As a result, a plot of ln {(w_out-w*)/(w_in-w*)} against N^1/2/Q based on the previous models shows the flow rate dependence. In contrast, a plot of ln {(w_out-w*)/(w_in-w*)}/(La_bN^1/2/Q) against N based on the new model can be approximated by a linear relationship. This suggests that the effect of the changes in degree of filling, length of filled zone in a vent section as well as the change in viscosity of fluid induced by shear heating on the devolatilizing performance should be taken into account. Hence, the new model is more applicable for analyzing the devolatilizing performance of a twin-screw extruder.

View full abstract

Co-rotating twin screw extruders are extensively used in polymer compounding. In co-rotating twin screw extruders, the kneading disk region is the most influential for compounding efficiency. However, only a few studies have so far been made in the flow analysis in the kneading disk region, because this flow is too complicated. In this paper the flow in the kneading disk region was studied with a closed co-rotating twin screw single stage mixer made for the purpose of the model analysis. We measured the torque, the pressure and the resin temperature with polypropylene under various compounding conditions. Consequently, it was found that the flow property measured by this experiment was equal to the flow property measured by a Capillary Rheometer.

View full abstract

The fiber degradation in the kneading disk region of co-rotating twin screw extruder was studied, with a closed single stage mixer of co-rotating twin screw extruder as a model for the analysis.
In the previous paper, we explained that the stress in the kneading zone was proportional to the screw torque. In this experiment, we measured the fiber length distribution and the average fiber length, under various compounding conditions.
Consequently, it was found that the fiber length distribution was determined by the shear stress and the total number of rotations, and if the average fiber length was equal, the fiber length distribution was also equal.

View full abstract

On the basis of the flow analysis network (FAN) method, a two-dimensional simulation for the non-isothermal flow of non-Newtonian fluid was developed and applied to the flow of polypropylene (PP) melt in a non-intermeshing counter-rotating twin screw extruder. The constitutive model for the PP melt we used was the Cross model with an Arrhenius-type function for temperature dependence. Calculating the movement of fluid particles permitted us to represent not only streamlines them-selves but internal residence time profiles in the flow field. Then, the energy equation was able to be successfully solved along the streamlines after replacing it in the convected coordinates.
Taking into accunt back flow and leakage flow, flux fields and streamlines in the flow field clearly showed the difference of conditions between staggered and matched screw flight configurations. The profiles of pressure, shear rate, shear viscosity, temperature and residence time also told us the mechanism of this complicating flow in the extruder. The non-isothermal flow of non-Newtonian fluid in the extruder was differentiated from the isothermal flow of non-Newtonian fluid and the isothermal flow of Newtonian fluid in terms of screw characteristics. The effects of several geometrical and operating conditions, including staggering angle, helix angle of screw flight, screw rotational speed and set temperature, on the screw characteristics were also investigated. Furthermore, some simulated results were compared with experiments using a model twin screw extruder and they demonstrated good agreements in pressure profiles and screw characteristics.

View full abstract

Along with an increase in the applications of thermoforming to products with more complicated configurations and higher precision, the simulation of this process has become strongly required. In this research, the thermoforming process which consists of blowing, stretching and vacuuming was employed using acrylic-modified PVC sheet. For the achievement of a profitable computer simulation, the employed process was observed in detail. The variation of extension ratio in each position of a specimen sheet in successive stages was investigated and the mean strain rate was obtained as nearly 0.1 (1/s). The variation of the temperature of the specimen sheet was measured. Thickness distributions of the sheet after each forming stage was also investigated. Then, an examination of the high-temperature and large-strain behavior of acrylic-modified PVC sheet as a function of the forming conditions was performed by uniaxial tensile testing. According to the obtained results by the tensile test, the stress-strain curves depend on the extension rate and especially on the temperature of the specimen. Furthermore, the loading curve is different from the unloading curve and also the deformation generated in the loading can not be recovered completely after the unloading. In conclusion, the visco-plastic behavior and the influence of the sheet temperature must be considered to model the material behavior in the employed thermoforming process for the computer simulation.

View full abstract

The deformation patterns and the thickness distribution of the products in a thermoforming process with blowing, stretching and vacuuming was simulated. A simple material model (Mooney-Rivlin model) was employed in this simulation. The visco-plastic behavior and the effect of the temperatare distribution of the specimen sheet in the observed thermoforming process were taken into consideration in the value of the Mooney constant for each element of the modeled sheet in the finite element analysis. For axisymmetric forming, the calculated results with uniform temperature in the specimen sheet had a fairly good agreement with the experimental results. On the other hand, for non-axisymmetric forming to make a 3-D rectangular box, the calculated results with uniform temperature in the specimen sheet had some differences from the experimental result, but those with distributed temperatures were in good agreement with the experimental results. In conclusion, it is confirmed that consideration of the distributed temperature in a specimen sheet for the simulation of the thermoforming process is most important. The strain rate of the specimen sheet in the thermoforming process or the anistropy of the specimen sheet had little effect on the precision of the calculated results.

View full abstract

We have developed a method of directly observing the blow molding process from inside the parison, which enabled us to ascertain some important aspects of parison deformation. This method consists of observing the displacement and deformation of a parison through a fiberscope inserted inside the parison.
During parison inflation, under various friction conditions, we confirmed that the parison does not slip against the mold wall and that the deformation takes place only in the “free” parison. We also observed that the “blow out” takes place at the highly deformed area of the “free” parison. This also supports the above observation.
We estimated the heat transfer coefficient from the surface temperature change of the parison before mold closing and the molded part after mold opening under various conditions. The heat transfer coefficient between parison and mold is estimated to be about 100 to 260W/m²K.
On the basis of the above observations, we tried to simulate the inflation process using a general purpose FEM program (ABAQUS). The thickness distribution of a molded part, thus computed, was in good agreement with experimental results.

View full abstract

In the blow molding process, pellets are melted and pumped by a screw device into an accumulator. By plunger extrusion, the polymer melt flows into an annular die and emerges vertically downward to form a parison.
The shape of the parison (diameter, thickness, length) should be controlled by setting the molding conditions (shape of annular die, flow rate, temperature, polymer properties, etc.) to make a successful blow molding process. Therefore, a method for predicting dimensions of the parison is necessary to research the best molding conditions.
We performed numerical studies of polymer melt flows in several types of real annular dies using the finite element method. The constitutive model for polymer melts that we used was the Giesekus model.
It was found that the relationship between the measured swelling ratio of parisons in these annular dies and the normal stress difference at the exit dies from analysis results could be expressed by a function. The function was independent of the geometry of the die in this study. This function may play an important role for the prediction of shape of parisons.

View full abstract

We have investigated the chemical improvement of plastic materials by crosslinking reactions but not the physical one by blending of other polymers or reinforcing with glass fiber. The present article is concerned with the improvement of poly (ethylene telephthalate) (PET) by the crosslinking reactions with various types of multifunctional epoxy compounds in the absence of any catalyst. The reactions of PET with small amounts of epoxy compounds, such as 4, 4′-diglycidyletherbisphenylmethane (DGEBF), tetraglycidyl-4, 4′-diaminodiphenylmethane (TGDDM) and 4, 4′-diglycidyletherbisphenylsulfone (DGEBS) were carried out in a single-screw extruder and cast into films. The PET film obtained after reactive extrusion in the presence of TGDDM (0.3∼0.4phr) (crosslinking agents) showed good tensile strength (83MPa, cf. PET=72MPa) and good tensile impact strength (30kJ/m², cf. PET=15kJ/m²). The thermal stability and chemical resistance were also improved.

View full abstract

This study investigates effects of mixer type and operating conditions on the graft reaction of maleic anhydride (MAH) on polypropylene using a 2-stage type (NEX-T 50; combination of continuous mixer NCM 50 with single screw extruder, KE 65) and 1-stage type (corotating twin screw extruder KTX 44). The graft reaction by the NEX-T 50 achieved a higher reaction level than that achieved by the KTX 44. The high grafting efficiency of the NEX-T 50 is attributed to microscopic dispersive and distributive mixing of MAH and peroxide in polypropylene in the NCM at a relatively low resin temperature. The MAH grafted PP was then blended with PA 6, and the impact strength of the alloy was measured. The alloy of the PP grafted using the NEX-T 50 showed higher impact strength than that produced in the KTX 44.

View full abstract

PPE/PA polymer alloy is industrially manufactured by blending the three components of PPE, PA and elastomer in an extruder. It is well known that the properties of the alloy are greatly affected by not only by the type of polymers, but also their morphology.
In this study, we report morphological changes in the mixing process, and we also report a relationship between the final morphology and mixing conditions. The domain size of the PPE phase and the morphology of the elastomer strongly depend upon the degree of mixing and the styrene content of the elastomer.
The impact strength of the alloy can be closely correlated with the domain size.

View full abstract