Rheological values of the rigid PVC compound of ALFAVINIL code: RE97/1-773.
Abstract
The objective of this work is to verify the design of an existing die for the manufacture of an extruded profile using the simulation of the flow in the head using a simulation software that uses computational fluid dynamics and also the experimental design and construction of a calibrator by means of the extrusion the geometry and desired dimensions of the profile. The rheological behavior of rigid PVC in the extruded molten state was investigated, which in itself is a difficult target due to the intrinsic weakness of this polymer that degrades when heated above 140°C. By means of a special capillary rheometer, rheological data, k and n of the power law, were obtained to introduce them, together with the process input parameters and the flow channel geometry in the simulation software. The flow channel was drawn with the head and calibrator using CAD-3D software. The different parts of the calibrator were manufactured and assembled into the equipment. The extrusion was performed with the process parameters: screw speed and material temperature used in the simulation software. The results obtained by the extrusion, geometry and final dimensions of the profile, mass flow, pressure, and temperature in the head were compared with those delivered by the software, being the same satisfactory.
Keywords
- simulation
- extrusion
- PVC (polyvinyl chloride)
- process
- design
- rheology
1. Introduction
Extrusion molding simulation allows you to avoid or reduce the high costs of the manufacture of the die or head and the probability of avoiding problems during production. On the other hand, costly modifications are saved, eliminating the current trial-and-error process, which added to the loss of time and lost profits, making simulation a tool of great importance for the national industry, especially for small- and medium-sized companies, since computer tools are not used in the design of heads and in the extrusion process of thermoplastic profiles in our country. Computer simulation, using specific software, allows to accelerate these processes, making possible all the necessary considerations in the design before building the head, especially for processes and design of profiles with complicated geometries and even with the use of different types of materials to manufacture the same. In addition, it is also important to know how to design a caliper in order to obtain an extruded profile with the desired final geometry and dimensions [1].
As antecedents of the international research works that were carried out on the subject, some of them can be cited, although there are differences with respect to the present work, and in the materials, simulation software and extruder equipment, including the ones used in the final process of calibration and cooling of the profile, are the ones that most resemble it.
Therefore, we can cite the one carried out by Srinivasa et al. [2], in which the simulation is performed on an existing head with Polyflow software that uses computational fluid dynamics, with a double screw extruder and using polystyrene as a material, which is simpler to process than the composite of rigid PVC that uses the reverse extrusion simulation process with the objective that the results of the same are used to improve the design of a new head. In the conclusion, it is commented that discrepancies were found between the results of the computational simulation and the extrusion in the dimensions of the profile between 5 and 10%.
On the other hand, the work of Gupta [2] can be cited, who uses the finite element method for a three-dimensional simulation applied to the flow of the molten polymer and analyzes the shear and elongation viscosity in the head in a final rectangular way, which is represented by the truncated power law model. In this work, he obtains a recirculation of the molten material at the vertices and a loss of pressure, which causes an abrupt contraction that increases rapidly with the exponent
Finally, Bogale [3] in his thesis work on design and simulation of a head for a rectangular profile, through modeling and simulation using the COMSOL Multiphysic software and the CA-RREAU model [4], also uses the low-density polyethylene as a study material and with a temperature of 220°C in the head. With the characteristics of the extruder screw, the flow rate is calculated, with which the pressure in the extruder could be obtained, since when it is minimum the output flow is maximum and vice versa. Although the design of the head was optimized in order to obtain a homogeneous distribution of the shear stress in the cross section of the same, it was not possible to simulate the exact dimensions of the head for a rectangular profile with the Comsol software.
In addition, this research aims to evaluate the predictive capacity of the Polyflow simulation software that uses computational fluid dynamics. Zhao [5] applied to the case of a rigid PVC profile, which has certain particular characteristics such as the aforementioned thermal sensitivity.
Thus, from the simulation, the contour values of the profile at the head outlet, the distribution of speeds that influence the contractions and stresses in the final profile and also the deformations and shear stresses that determine the physical and mechanical properties of the final profile, are obtained. In addition, the extrusion of the profile is carried out, at a certain RPM of the screw, obtaining the values of the same from the pressure and temperature sensors in the head, and on the other hand, the mass flow at the outlet of the head is measured. Subsequently, the results of both processes are compared and the results that are described in detail below in the corresponding item are obtained.
2. Materials and methods
2.1 Material
2.1.1 Characteristics of material:
The material under investigation is a rigid PVC compound from the company ALFAVINIL RE97/1-773, which is currently used widely in the manufacture of all types of profiles for construction, in the refrigeration industry, among other applications, due to its excellent properties of resistance to corrosive media, good thermal, electrical and acoustic insulation, low weight, and mechanical properties suitable for each use. This is possible due to the possibility of designing its formulation with different types of materials (resin, plasticizer, additives, mineral fillers, pigments, thermal stabilizers, and others). It is also an ecological material, since it has the lowest carbon footprint and can be recycled.
On the other hand, due to its chemical composition, it is a thermally sensitive material, degrading from 140°C; therefore, for its processing, in addition to containing a thermal stabilizer in its composition, the use of special extruders is required, with screws with a low compression ratio (2:1), which perform the plasticization of the molten material with a low shear deformation or also known as shear rate, so that the heat provided for this process comes mainly from the heating bands of the cylinder and that the movement of the screw is mainly to improve the mixing and homogenization of the different components of the compound without generating heat by friction between the material, screw, and cylinder. Due mentoned reason, twin screw extruders, which have the best of these mentioned characteristics are recommended for use for this material, especially for medium and high productions.
2.2 Rheological characterization of the material
Most of the molten polymers behave as non-Newtonian fluids; that is to say, their shear viscosity (
On the other hand, thermoplastics, especially PVC, is classified as pseudoplastics; that is, the viscosity decreases with shear deformation (e.g., that is exerted by the extruder screw). To carry out computer simulation, precise information on the behavior of the material under the processing conditions must be possessed: temperature, flow, density, and the rheological characteristics of the material, especially these are very important because it is a composite. Regarding the rheological characteristics, it is necessary to know the
Material | Percent (%) | Details |
---|---|---|
Resin content | 66 | Supply: Solvay Argentina |
Plasticizer content | 2 | Vegetal origin, co-stabilizer |
Stabilizer content | 2 | Stabilizer pack—lead tribasic sulfate-based lubricant |
Contents of other elements | 24 | Micronized-coated calcium carbonate, average particle size 1.6 μm |
Titanium dioxide | 6 | High opacity |
A practical way to identify a PVC compound is through its
Measurement items | Shear rate (ɣ) 1/s | Shear tensile (T) bar | Viscosity (η) Pa | Mass flow (Qm) g/h | Volumetric flow (Qv) cm3/s |
---|---|---|---|---|---|
1 | 22.7 | 0.7 | 2933.7 | 1730.4 | 0.302 |
2 | 27.5 | 0.8 | 2725.4 | 2101.2 | 0.367 |
3 | 38.6 | 1.0 | 2511.2 | 2944.4 | 0.514 |
4 | 51.6 | 1.2 | 2329.8 | 3940.8 | 0.688 |
5 | 69.8 | 1.4 | 2069.3 | 5324.8 | 0.930 |
6 | 90.4 | 1.7 | 1881.5 | 6899.2 | 1.205 |
2.3 Methods
2.3.1 Drawing of the flow channel in the head
The existing head flow channel is drawn using 3D solid Computer Aided Design software (Figure 1).
2.3.2 Simulation
The flow channel geometry drawing is imported into the head using the ANSYS Polyflow simulation software.
Geometry meshing (Figure 2) is performed using the Polyflow default system, which performs the automatic subdivision into anisotropic tetrahedral elements for three-dimensional flows that have free surfaces such as the one presented in our case (flow of molten material at the head outlet).
In the same module, the names of each subdomain are incorporated, which are the areas of the melt inside the head: SD1 (Figure 3) and outside of it: SD2 (Figure 4) so that the simulation software recognizes which is each zone for the later calculations.
The names of the contours (surfaces) of each subdomain SD1_BS2 (Figure 5) and SD2_BS2 (Figure 6) are defined; the flow input surfaces to the head: SD1_BS1 (Figure 7) and the flow output surfaces of the same: SD2_BS1 (Figure 8) are also defined.
Subsequently, the setup module is entered and the input variables are entered using the Polydata program: flow (8.5 kg/h), material density (1.42 g/cm3), and the rheological values
Next, in solution, the computational run is made to calculate the results for the model defined in the previous steps.
Finally, the post-processing is carried out with which the results of the calculations performed are obtained, such as the values of the distribution of speeds, shear deformations, pressures, contours of the profile at the exit of the head, and the graphs of each of them for analysis.
2.3.3 Experimental design and construction of the calibrator
The caliper is experimentally designed and drawn with 3D solid computer-aided design software (Figure 10).
The different parts that make up the calibrator are built and assembled (Figure 11), placing the assembly in the chamber or calibration-cooling pan and leaving the fixing screws of the upper part unadjusted, just positioning them, leaving the necessary play to enter the profile for calibration.
The gauge defines the final external dimensions of the profile [1], so that the profile in the molten state at the outlet of the head, at temperatures close to 200°C, enters the gauge, where vacuum is applied and the material copies the internal dimension calibrator.
On the other hand, when the calibrator is immersed in a pan with water at 20°C, the profile cools, solidifying.
In order to define the internal dimensions of the caliper, it was taken into account that the molten material, when it cools, contracts and in the case of rigid PVC compounds it can reach high values of up to 4%.
In addition, another decrease in the cross-sectional area must be taken into account due to the stretching caused to the material by the dragging carried out by the draft equipment, which can be between 5 and 10% for low-thickness rigid PVC profiles [8].
3. Results
3.1 Rheological values
The rheological values were obtained from the test carried out by technical personnel of the company ALFAVINIL S.A., manufacturers of PVC compounds, using a RheoDrive 7 capillary rheometer, model: Rheomex 19/25 OS, in line with an extruder, whose screw has a ratio of compression 2:1, L/D: 25 and a capillary matrix of slit H: 2.0 mm, W: 20 mm and with a profile of temperatures of the material of 150, 170, and 190°C in the extruder and 200°C in the head.
3.2 Simulation
3.2.1 Speed distribution in different sections
In the post-processing stage of the simulation, cross sections are defined at certain distances in the flow channel, to analyze the distribution of the velocity vectors in the same section and in each of them (Figure 15).
It can be seen that in the central regions of the sections in the feeding zone (conical zone), these velocity vectors have a higher value (celestial zone) than in the contours since, as the flow of the material is close to the inner wall of the flow channel, the velocity is zero (blue zone), this is consistent with what happens in reality when a fluid circulates through a conduit, and the velocity profile will be maximum in the center and tending to zero in regions close to the inside walls of the same.
As the flow of the molten material progresses, velocity vectors can be observed in the central areas of each cross section of greater magnitude (more intense green color), and this is due to the abrupt change of the geometry in the flow channel going from a conical rectangular; in particular, if we analyze the fourth cross section from the entrance of the flow to the head, it is observed that in the center there is a zone of velocity vectors of even greater magnitudes (red color).
Variations in the velocity vectors at each point of the same section cause the flow channel not to comply with the design principle of “minimum volume” of the heads; that is, the melt must reach all the points at the same time of the cross section or with the same speed.
These differences between the sliding of the fluid in the central zones with respect to those that are closer to the inner wall of the head cause internal shear stresses in the molten material that could later lead to the appearance of failures in service; this becomes more critical in profiles with complex shapes and in the particular case that between width and thickness, there is an important dimensional difference.
3.2.2 Distribution of shear deformations
In the graph shown below (Figure 16), it is observed that in the area where the geometry changes, shear deformations are generated (all colors are very high except blue), and this indicates that in these places, the molten material can get to degrade, especially in our case in which rigid PVC is used, due to the generation of high localized temperatures due to friction and shear.
In order to reduce or prevent this detrimental effect from being generated, it is advisable to modify the geometry of the head flow channel so that the transition between the two zones is smooth.
The simulation results predict in this case the detrimental effects that occur with the current head design, so that if it is necessary to build a new head, the geometry must be modified, so that the shear deformations are low and if possible, their values are between 50 and 100 1/s, which is advisable to process this type of material and as described in the rheological test, they are the values where the Power or Ostwald Law is valid.
Another analysis that is carried out is on the contour areas at the entrance and exit of the molten material from the head, in which the shear deformation is zero, especially if a final parallel area of the matrix is left sufficient to have a swelling of the die, minimum fade, and within the recommended parameters [6].
3.2.3 Comparison of original profile contours vs. cast profile outlet head
Another important result of this simulation is obtained: the contour of the profile in the molten state at the outlet of the head at a distance of 60 mm.
To do this, the orientation of the axes of the channel geometry is modified in the post-processing stage, eliminating all contours and surfaces except for the contours of the output profile of the head (original) and the profile of the molten material at the output, from the head at a distance of 60 mm.
By superimposing both contours, the geometric and dimensional differences that exist between them can be seen and therefore, this image is exported and said differences are dimensioned (Figure 17).
They are due to the aforementioned incidence of the swelling effect of the melt that causes an increase in the cross-sectional area of the flow of the melt at the outlet of the head.
In a previous simulation, a channel geometry model was used with an outer zone at the head outlet of 10 mm, with which the deformations were greater, there were important geometric and dimensional differences, and the profile could not have been calibrated or manufactured, because it would have stuck in the caliper, thus optimizing the design of the model by increasing the length of the area at the exit of the cast profile with a new simulation at a distance of 60 mm.
The dimensional variation at a distance of 60 mm from the head can be seen in Table 3.
Characteristics | Original profile | Profile molten at heat out | |
---|---|---|---|
mm | mm | Dimensional differences (%) | |
Thickness | 3.0 | 3.08 | 2.7 |
Wide | 30.0 | 31.8 | 6.0 |
Therefore, maximum deformation values were achieved due to the swelling effect of the melt within the recommended parameters, which are 3–6% for thicknesses of rigid PVC profiles between 3 and 4 mm thick [6].
3.2.4 Pressure distribution
The graph with the pressure distribution was obtained, since at the entrance of the flow of the molten material to the head the pressure is maximum and is the one generated in the extruder at the exit of the dosing zone, with values that depend on flow rate and material temperature, as well as screw and die design (back pressure).
The pressure decreases as the flow advances inside the head, as can be seen in Figure 18, so that at the exit of the matrix (blue area), it is zero, as it should be since the flow of the material cast is stress free.
It is observed that the pressure value obtained from the simulation in the area where the pressure sensor is located in the head (light blue color) has values between 142 and 68 bar (in Figure 19 the pressures are expressed in Pascal (Pa) and the The value measured with said pressure sensor is 51 bar for a flow rate of 8.5 kg/h and a temperature measured with another sensor in the head of 200°C.
This difference between the simulated and real pressures of the process may be due to the high shear deformations that occur inside the head and that when using the Power Law or Ostwald model, the same is for deformations, much lower and on the order of 50–100 1/s.
The large difference between the shear deformations obtained as a result of the simulation (between 5.2107 and 5.6107 1/s in the highest values of red color) and that used by the Power Law affects the results of the pressure distribution that delivers the software.
3.3 Tests with the extruder equipment
An optimization of the process and the design of the calibrator is carried out through three tests carried out with the extruder, its auxiliary equipment, and the experimental calibrator installed. The conclusion of these machine tests showed that small modifications should be made in the caliper design, such as grooves to improve the application of the vacuum, and also the machine parameters were adjusted, in particular the screw revolutions and the drag speed. The said tests are carried out at a temperature of the molten material in the head of 200°C, obtained by means of a sensor located therein (Figure 12) and modifying the flow rate by varying the revolutions of the extruder screw. The pressure values in the head are recorded by means of another sensor (Figure 13) that the equipment has and which is also located in the head, observing its value on the extruder control panel. Each of these tests is detailed below: 3.3.1. First test: The screw speed is adjusted for a mass flow rate of 5 kg/h, measuring with the sensor a pressure = 55 bar, being the dimensions of the extruded profile at the exit of the gauge of width = 26 mm and thicknesses: 2.6 mm on one side and 2.9 mm on the other side of the width. It is determined that the cause of not reaching the required dimensions (width = 30 mm and thickness = 3 mm) is a vacuum deficit in the calibration pan by vacuum and cooling (Figure 14). The vacuum is generated by a pump located in the water tank that is located at the bottom of the vacuum and cooling calibration pan and this vacuum is measured with a pressure measuring instrument or manometer (Figure 19).
3.3.1 Second test
The screw revolutions were adjusted to 48 rpm so as to obtain a mass flow rate of 5.7 kg/h and a pressure of 47 bar was measured.
To increase the vacuum, a calibrator will be made with internal grooves that communicate the place (hole) where the final geometry of the profile is defined, with the room where the mentioned vacuum is applied.
The dimensions of the profile were measured at the exit of the caliper and the following values were obtained: width = 29.6 mm and thicknesses in the central area of the profile = 3 mm and at each end of the width = 2.7 mm.
Therefore, although the dimensional value of the width is improved, the geometry is still not copied well on the sides and also the geometry is not rectangular, since the thickness in the central part is 3 mm and decreases toward the edges reaching a value at both ends of 2.7 mm; therefore, the process is not yet suitable to correct this defect.
3.3.2 Third test
The screw revolutions are again modified to 65 rpm, obtaining a mass flow = 8.5 kg/h, and it is noteworthy that this flow value is the same that is used as one of the input variables in the simulation software.
The measured pressure was 51 bar (in this case, there is a difference with the values obtained in the simulation, since for the area where the pressure sensor is located, the results were between 142 and 68 bar.
The increase in the mass flow, accompanied by an increase in the drag speed and using the calibrator with the grooves to improve the application of vacuum on the flow of molten material that comes from the head and that is calibrating and cooling, achieves the desired geometry of the outlines, which is the output of the caliper.
On the other hand, the required dimensions are achieved with an error of less than 5% at the ends of the width, since it was measured with a coliza foot that the width = 30 mm and at the ends = 2.9 mm, that is, the error was 3.3%.
The flow measurement was performed by taking the amount of material that was extruded per unit of time and it coincides with the flow rate that was used as an input variable in the simulation software, that is, 8.5 kg/h, working with the screw at 65 rpm.
The sensed pressure was 51 bar and that obtained from the simulation software ranged between 142 and 68 bar, and the explanation of these differences is due to the fact that the Power Law model used is for low deformation speeds and in the simulation, said deformation shear was very high, especially in the area where the geometry of the channel was modified from conical to rectangular (transition zone).
4. Discussion
The result of the first simulation using a geometry of the flow channel with an area at the outlet of the head of 10 mm (Figure 20 area in green color) generated high deformations due to the swelling effect of the melt, which means that there was no possiblity to calibrate and manufacture the profile to the required dimensions, as explained above.
Therefore, the model is modified by increasing the length of said area to 60 mm and it is simulated again, obtaining deformation values that are within the recommended range. This length of the area of the molten profile outside the head is the one that is recommended to use for the extrusion of the same; that is, the distance between the head and the gauge should be of that magnitude to ensure proper calibration and not having clogging when the profile in the molten state enters the calibrator. When a vacuum is applied in the calibration-cooling pan, the molten material in the calibrator increases its cross section by copying the internal dimensions of the same and advancing with a continuous flow by means of the drag force, while it solidifies when immersed in water at 20°C. In addition, the software uses overlock optimization in the said exit zone, which is the one shown in this work, since by increasing the length of the free-flow zone at the exit of the head, the deformation due to the swelling effect of the fade decreases.
5. Conclusions
Through simulation, a 60-mm-long profile geometry model is obtained at the outlet of the head, thus minimizing the swelling of the melt and obtaining a deformation with dimensional differences that are within the recommended percentages, as described in Section 3.2.3.
The 60-mm length obtained from the simulation is recommended to use in the extrusion process to overlap the distance that must exist between the head and the caliper in order to avoid problems of clogging of the molten material that enters the caliper and apply vacuum.
From the results of the simulation, it is determined that the distribution of the velocity vectors in each of the cross sections is not homogeneous and also the shear deformations had very high values, which cause internal stresses to be generated in the profile that causes their failure in service.
Therefore, the geometry of the head flow channel should be redesigned to improve these parameters that influence the final properties of the profile, such as the generation of high internal stresses that affect the quality of the product.
An optimal parallel zone can also be defined, because its length influences the swelling effect of the melt; that is to say, by increasing it, a greater relaxation of internal stresses is obtained and consequently a less deformation of the material at the exit of the caliper [7]. It was possible to carry out the rheological test and obtain the k and n values of the Power Law for the rigid PVC compound of the present work using a special capillary rheometer connected in line with an extruder and they could be used in a simulation software that uses fluid dynamics. Computational and the direct extrusion method [9] is used to optimize the design of a head of a rectangular profile in rigid PVC.
Therefore, the predictive capacity of the ANSYS Polyflow simulation software is verified for this work and with it, it can be applied for the simulation of more complex profile geometries.
Finally, a specific calibrator is designed and built for this profile and in only three tests with the extruder equipment, the operation of the same and of the process is optimized, verifying that the flow used in the simulation is correct, obtaining a profile with the geometry and final dimensions required and with a dimensional error of less than 5%.
Acknowledgments
Special thanks to the University Technician Gustavo Pierson, alIng. Federico Tirapelli, and Technician Javier Ríos Zabala, staff of the Higher Polytechnic Institute-U.N.R. for their collaboration in the design, construction of the gauge, and participation in the extrusion process.
To Dr. Patricia Frontini from INTEMA, UndMP and Dr. Javier Signorelli from IFIR, CONICET Rosario for their contributions and suggestions.
Finally, to Dr. Fernando Angiolini and Mr. Leonardo Tort, technical and managerial staff of the company ALFAVINIL S.A., who selflessly lent their collaboration to carry out the rheological tests and provided the material to be able to carry out the extrusion tests.
References
- 1.
Szarvasy L et al. Computer aided optimization of profile extrusion dies. International Polymer Processing, the journal of the Polymer Processing Society. 2000; 15 (1):28-39 - 2.
Srinivasa R et al. Extrusion simulation and experimental validation to optimize precision die design. In: ANTEC 2004. IL: Northern Illinois University; 2004 - 3.
Vlachopoulos J, Fattamann G. The Role of Rheology in Polymer Extrusion. Hamilton, ON, Canada: Department of Chemical Engineering, McMaster University, Polydynamics, Inc.; 2015 - 4.
Titow MV. PVC Technology, 4th ed. England: Elsevier Applied Science Publishers; 2012 - 5.
Zhao W. Finite element simulation of 3D unsteady viscoelastic free surface flow with level set method. In: ANTEC, Society of Plastics Engineers Annual. 2010 - 6.
Michaeli W. Extrusion Dies for Plastic and Rubber. 3rd ed. Munich, Germany: Hanser Publishers; 2003 - 7.
Goncalves ND et al. Design of complex profile extrusion dies through numerical modeling. Journal of Non-Newtonian Fluid Mechanics. 2013; 200 :103-110 - 8.
Limper A, Fattmann G. Application of an Online Rheometer to Evaluate the Melt Properties of PVC. Germany: Institut für Kunststofftechnik, University of Paderborn; - 9.
ANSYS. ANSYS Polyflow Tutorial Guide. Release 17.0. ANSYS, Inc.; 2016