Optimised pilot line: Modular design for use as a blown film (above) or flat film system (below). The 3-layer coextrusion system with increased take-off speed (blown film system: up to 50 m/min, flat film system: up to 120 m/min), achieves near-production conditions (photo: Dr. Collin GmbH) ANDREAS MAYER GÜNTER MÜLLERFERLI S uccessful introduction of a new film to the market is always preceded by a thorough application-oriented and application-relevant evaluation of the new raw materials and film formulations. The objective is to optimise the efficiency of each individual process step while taking the respective technical and economic boundary conditions into consideration. More and more often, however, the time Translated from Kunststoffe 9/2005, pp. 164 169 Rapid Analysis of on Pilot Lines Film Properties Product Development. Blown and cast film must not necessarily be tested on large-scale systems: The extrapolation of tests conducted on pilot lines to production systems that are more than ten times the size provides very satisfactory results. Careful use of pilot lines thus offers considerable time and cost benefits when developing products and applications. and cost factors are more decisive than the absolute quality of the results. High quality is generally equated with a large investment of time and money.nor is it possible to bypass a so-called practical test under real-world conditions on the usual production and downstream converting equipment when manufacturing and testing blown and cast films either. The number of production parameters and application criteria to be observed is simply too high. Up to this point, however, quality, time and cost aspects should be optimised equally. That this statement does not have to be contradictory will be illustrated here by way of examples of selected blown and cast film trials. Focus on Film Properties It is not only the resins and formulations that determine the properties of a film. Production method, machine type and even process conditions have an almost equal effect, e.g. the blow-up ratio and frost line for blown film or the line speed and air gap for cast film. Examples in the literature illustrate the relationships. These insights and data can be applied to other cases to only a limited extent, how- 1
EXTRUSION ever. Sometimes even qualitative application of this information is questionable. Other (model-based) theoretical approaches are intended to overcome this shortcoming by providing generally valid rules for application or even process models. Above all, the development of process models requires the particularly timeconsuming determination of the necessary resin characteristics an almost hopeless undertaking for multi-component systems. Comparison Trials as a Solution? A practical approach to evaluating the necessary resin and film properties can be obtained from everyday plant practice. For instance, while film properties are determined in the laboratory in the form of objective absolute values, a subjective evaluation generally takes place through direct comparison with a selected reference. Corresponding results can then be presented as star diagrams and quickly provide information in this way on the relative advantages and disadvantages of other approaches or solutions. This approach also promises success when manufacturing and analysing film samples. Another objective of this same project was to establish a reference quality regardless of whether individual resins or complete formulations are involved. With appropriate process knowledge and application experience, the quality of such a reference can also be produced on smaller pilot lines with sufficiently high reliability as to the information provided. Comparison with the resins and film formulations to be analysed should be possible, if in all cases identical system configurations and operating parameters are used. It goes without saying that selection of the correct operating point is essential. In this regard, a number of important aspects must be observed, for instance, correct melting and homogenising of the melts as well as careful consideration of orientation and relaxation effects during film formation. In the film laboratory at Dow Chemical Iberica in Tarragona, selected blown and cast film trials were used to determine whether the described procedure provides suitable results. For this purpose, film characteristics determined from a series of experiments conducted on smaller pilot and larger production systems were analysed and compared. Feature The pilot lines employed for these experiments used the same extruders. Only the downstream equipment differed. The modular concept of the pilot line provided by Dr. Collin GmbH, Ebersberg/Germany, permits rapid assembly of extrusion lines in a large variety of configurations, and is exceptionally well designed for flexible investigation of small quantities of material. Rapid screening of individual parameters and performance of multiple experiments at a low cost are additional important fields of application for this equipment. The production systems available included a 3-layer coextrusion system for blown film from Hosokawa Alpine AG, Augsburg/Germany, and a 3-layer coextrusion system for cast film from Primplast, Melzo/Italy. The characteristic system sizes are compared in Tables 1 and 2. All trials were performed using identical system parameters and resins from the same lot. In this way, material- and processrelated effects could be largely excluded. Feature (Dr. Collin GmbH) (Primplast s.r.l. ) Number of layers Up to 5 layers 3 layers Extruder types employed Throughput % Program D % Program P Film D1 D2 D3 P1 P2 Layer A 25 LDPE 400L LDPE 400L LDPE 400L 20 Layer B 50 LDPE 400L Affinity* PL 188G Elite* 5400G 2 Carl Hanser Verlag, München Dowlex* 2045S Inspire* 112 60 Layer C 25 LDPE 400L LDPE 400L LDPE 400L 20 * Trade mark The Dow Chemical Company Table 3. Blown film structures investigated (Dr. Collin GmbH) Number of layers Up to 5 layers 3 layers Extruder Maximum throughput Die Cooling A. 25 mm x 25 D B. 30 mm x 25 D A. 6 kg/h B. 10 kg/h 60 mm X 1.8 mm Radial spiral mandrel distributor Monocooling ring A. 25 mm x 25 D B. 30 mm x 25 D A. 6 kg/h B. 10 kg/h A. 55 mm x 35 D B. 75 mm x 35 D A. 150 kg/h B. 300 kg/h Multi-layer adapter Dr. Collin Davis Standard Cast film die 300 mm 712 mm Mono automatic die Melt stabilisation Air knife Vacuum box Roll diameter (primary roll) 144 mm 1000 mm Maximum line speed 60 m/min 450 m/min Hopper Gravimetric feeding Gravimetric multi-component feeding Table 2. Comparison of the cast film systems employed (Hosokawa Alpine AG) A. 50 mm x 30 D B. 65 mm x 30 D A. 120 kg/h B. 220 kg/h Maximum line speed 30 m/min 130 m/min Hopper Gravimetric feeding *manual operation during the trials Table 1. Comparison of the blown film systems employed 200 mm x 1.8 mm Axial spiral mandrel distributor with horizontal predistribution Automatic cooling ring* with Inner cooling Gravimetric multicomponent feeding
Comparison of Blown Films Blown film 50 µm thick with the structures summarised in Table 3 were produced on both systems with a blow-up ratio of 2.5. Material type MFI type MFI value [g/10min] Density [g/cm 3 ] Comments Dow LDPE 400L 190/2.16 1.0 0.925 Standard PE-LD Dowlex 2045S 190/2.16 1.0 0.920 Basic stabilised PE-LLD 190/2.16 1.05 0.919 Affinity PL 1880G 190/2.16 1.0 0.902 Elite 5400G 190/2.16 1.0 0.916 Inspire* 112 230/2.16 0.4 0.900 * Trade mark The Dow Chemical Company Table 4. Characterisation of the materials used for the blown films Higly stabilised PE-LLD with slip and antiblock additives Highly stabilised polyolefin plastomer Program D P System (Collin) (Alpine) Highly stabilised high-performance polymer Propylene-based polymer designed for blown film applications (Collin) (Alpine) Output [kg/h] 8.4 160 7.3 106 Bubble diameter [mm] 235 785 235 785 Circumferential tolerances [%] 3.6 9.4 5.6 6.1 7.7 9.1 6.3 8.5 Quantity per sample [kg] 10 400 10 400 Manpower per sample [h] 0.5 1.5 3 0.5 1.5 3 Table 5. System parameters for the blown film investigations Measurement Correlation coefficient R 2 Slope value Tensile values in machine direction Elongation at yield 0.565 0.89 Tensile stress at yield 0.963 0.81 Elongation at break 0.896 1.01 Tensile stress at break 0.934 0.93 Toughness 0.836 0.9 Tensile values in crosswise direction Elongation at yield 0.729 0.93 Tensile stress at yield 0.99 0.93 Elongation at break 0.936 0.98 Tensile stress at break 0.9 1.04 Toughness 0.939 1.01 Puncture resistance 0.974 1.36 Elmendorf machine direction 0.893 0.99 Elmendorf cross direction 0.912 0.82 Table 6. Comparison of data from the blown film investigations The resins employed for this (Table 4) were selected to provide the wide possible spectrum of different polyolefin polymers. The most important system parameters are presented in Table 5. An evaluation of the extruded films through offline thickness measurements in the laboratory shows that, in spite of the smaller and thus more difficult to adjust die ring, thickness tolerances quite close to those encountered in actual practice can be achieved even with the pilot line. With trial runs longer than those usually selected for laboratory requirements, better values can be achieved in both cases. The following characteristics were used when analysing the film qualities: short-term-tensile test in accordance with ISO 527-3, impact resistance according to ISO 7765-1, Elmendorf tear resistance according to ASTM 1922-94A. For the values determined by these standardised methods, it must always be borne in mind that individual mean values can easily exhibit statistical standard deviations on the order of 5 to 10 %. Starting with the hypothesis that pilot and productions systems basically behave similarly and thus produce films with similar properties, a directly measurable quality criterion results for evaluating the two series of experiments. When comparing corresponding film data, the extent to which a linear correlation with appropriately qualified coefficients exists must be investigated. A comparison of the tear resistance values for test films in the lengthwise and crosswise directions is shown in Fig. 1. The correlation coefficients determined are 93.4 and 89.9 %. Considering the boundary conditions discussed previously, this result can be judged as quite good. A more exact analysis of the two diagrams shows further areas of agreement: the individual values deviate from the straight-line correlation by only 5 to about 10 % this is within the range of standard deviation for the respective mean values, the slope values deviate only slightly from 1 and thus indicate direct comparability of all individual correlated films over a rather wide range of measurements. The tear force measurements point to a similar orientation of all films that were compared with respect to lengthwise and crosswise directions. This statement is confirmed by the comparison of dynamic impact resistance values in Fig. 2. Statements similar to those above for the tear resistance can be made regarding evaluation of the correlation and deviation of individual values. 3
EXTRUSION A large difference, however, is evident in the level of the two series of data. In this regard, the significantly lower specific throughput of the pilot line must be taken into account. The consequence is faster cooling with a correspondingly lower frost line. This may explain the rather clear difference. Analysis of the other data that were measured leads to statements similar to those for the two characteristics discussed above; these will not be discussed further here. The results from this data comparison are summarised in Table 6. Comparison of Cast Films As part of a stretch film project, 17 µm thin monofilms produced from commercially available PE-LLD resins from various suppliers were investigated (Table 7). The selected line speeds were 30 and Resin Ladene/TM 318B Flexirene/TM CL 10 Dow LLDPE 1221 Dowlex SC 2216 Dowlex SC 2107 Manufacturer Sabic Polimeri Dow Pacific Dow Europe Dow Europe Melt index [g/10min] Fig. 1. Comparison of the tear resistance values of blown film in the machine and crosswise directions shows good agreement between the pilot line (Collin) and production-scale system (Alpine) Shrinkage in the machine direction 2.8 2.6 2 3.3 2.3 Density [g/cm 3 ] 0.918 0.917 0.918 0.919 0.917 Table 7. PE-LLD resins used for the cast films Fig. 3. Comparison of the orientation and shrinkage behaviours of blown films: The pilot line (Collin) stretches the melts somewhat less than the production-scale system (Primplast) Impact resistance 250 m/min. In addition to the measurements made in the blown film investigations, puncture resistance values according to ASTM 5748 were determined. The extremely different operating modes of the two systems must lead to corresponding differences in the orientation behaviour in the machine direction. On the basis of the smaller shrinkage values, Fig. 3 confirms that the pilot line does not stretch the melts as much. As a consequence, these differences in film orientation are noticeable in the different values for the elongation at rupture in the machine direction (Fig. 4). Compared to the blown film investigations discussed above, the correlation coefficients for the cast film comparisons are always at a lower level. This is also confirmed in the summary of all of the results (Table 8). Formation of the melt web between the outlet from the die and cooling on the chill roll is responsible for the majority of Fig. 2. Comparison of the impact resistance values of blown films produced on the pilot line (Collin) and on the production-scale system (Alpine) shows a good correlation Elongation at break in the machine direction Fig. 4. Comparison of the values for elongation at break in the machine direction of blown films produced on the pilot line (Collin) and on the production-scale system (Primplast): Differences in the orientation of the films become noticeable here 4 Carl Hanser Verlag, München
the film s physical properties. Small variations in this narrow region result in some quite significant changes in the respective properties. Considering this aspect, the quality of the comparison performed here is rather acceptable. Comparison of Costs A generally valid cost comparison between pilot lines and production systems is not possible because of the great variety of possible system configurations. On the basis of the cast film investigations described here, at least an attempt should be made to compare several important costs (some estimated) incurred during production of the samples (Table 9). This rather simplified comparison shows that by using a pilot line in the present case costs of about 7300 EUR could be saved over the six hours required to produce each sample. With a purchase price of about 290 000 EUR for a 3-layer Measurement Correlation coefficient R 2 Slope value Tensile values in machine direction Elongation at break 0.817 1.2 Tensile stress at break 0.695 0.86 Impact behaviour Force 0.88 1.07 Elongation 0.82 1.05 Puncture resistance 0.9 1.35 Elmendorf crosswise 0.88 0.89 Table 8. Comparison of data from all cast film investigations Item Production line Time the system was used for the trials [h] 20 6 Investment Purchase price [EUR) 500 000 290 000 Depreciation time [a] 5 5 Production hours/year [h] 7200 **/ 2000* 2000 Depreciation costs/hour (10 % int. on capital) [Euro/h] 16.67** / 62.50* 36.25 Depreciation costs for sample production [EUR] 1250**/ 333.35* 217.50 Personnel Personnel required [individuals] 2 1 Personnel hours required [h] 40 6 Personnel costs/hour [EUR/h] 30 30 Total personnel costs for sample production [EUR] 1200 180 Energy Energy costs [EUR/kWh] 0.1 0.1 Required power [kw] 100 3.5 Required energy [kwh] 2000 27 Total energy costs [EUR] 200 2.7 Material costs Material required for sample production [kg/h] 300 10 Material price/kg [EUR/kg] 1 1 Material costs [EUR] 6000 60 Total expenses [EUR] 8650**/7733.35* 460.20 Savings from pilot line [Euro] 8189.80**/7273.15* * for 24-h operation as a production system ** for 8-h operation as a laboratory system Table 9. Cost comparison for producing cast film samples pilot line, such a system has already been amortised after about 239 hours of operation (<30 workdays). Items such as building rental, water costs and the not inconsequential lost profit attributable to use of the production system to provide samples were not even taken into consideration in this estimation. In the example considered here, the costs of a relatively small production system with a die width of 712 mm were compared to those for the pilot line. The larger the production system involved, the more favourable is the cost comparison for the pilot line. Moreover, relatively lowcost test materials (about 1 EUR/kg) were employed. When working with expensive, newly developed materials, the figures are even more clearly in favour of a pilot line. Often the amount of material available for a new product is too small to even produce samples on a production system, but enough to provide important information from a pilot line. In contrast to a production system that can be used as a development platform only in exceptional cases and after lengthy waiting periods, a major benefit of a pilot line, in addition to the not inconsiderable cost savings, is that, it is always available for development work. As a result, the pilot line contributes to a noticeable acceleration in development a competitive advantage that cannot be overlooked in our fast-paced times. THE AUTHORS DR.-ING. ANDREAS MAYER, born in 1954, is a project manager in the Plastics Division at Dow Europe GmbH, Horgen/Switzerland. DR.-ING. GÜNTER MÜLLERFERLI, born in 1960, is employed at Dr. Collin GmbH, Ebersberg, in research and development. Contact: pilot-lines@drcollin.de 5