Final Report Summary - EXTRUWIN (Extruded window profiles based on an environmentally friendly wood-polymer composite material) Executive summary: The objective of the EXTRUWIN research project was to develop and produce a wood-plastic composite (WPC) window profile. The research and technological development (RTD) performer Fraunhofer WKI developed various WPC formulations and prepared compounds for this task. Various thermoplastic matrices and blends were tested, including polypropylene (PP), thermoplastic polyurethane (TPU), acrylonitrile-butadiene-styrene (ABS) and polymethyl methacrylate (PMMA). Coupling agents were selected based on the matrices used. Compounds were produced at Fraunhofer WKI using a ring die press (plast agglomerator / Palltruder), a co-rotating, 20 mm twin-screw extruder and a heating mixer. Profiles were extruded into various shapes at Fraunhofer WKI, Haka Gerodur and SAMP. Box and tape profiles were tested regarding their key mechanical and physical properties. A window profile could not be extruded due to a variety of technical challenges as outlined in detail in the deliverables and extrusion protocols. It was determined that WPC can be coated with waterborne paints. Waterborne 2K-polyurethane coatings and waterborne UV-curing coatings were developed in five different colours, namely RAL 5002, RAL 6002, RAL 7040, RAL 8011 and RAL 9016. Flame treatment was proven to be a suitable pre-treatment prior to coating to achieve a sufficient adhesion of the coating on the WPC substrate. The application of a pigmented layer followed by an ultraviolet (UV)-protected clear coat analogue to a car finish seems to have a positive effect, especially for the UV-curing coatings, in order to guarantee a sufficient through-hardening. The results from the outdoor weathering tests according to DIN EN 927-3 indicate that regarding the waterborne 2K-PU coatings, the series Pefalon X-400 performed best, followed by Pefalon X-500. Pefalon X-100 exhibits larger delta E for RAL 5002, RAL 6002 and RAL 9016 which exhibited a delta E of 3.2. Regarding the UV-curing coatings, Bayhydrol XP 2690 performed best. Bayhydrol XP 2736/2690 showed a significant discoloration with RAL 5002, 6002 and 9016. Alberdingk LUX 250, however, revealed a delta E of more than 3 for RAL 5002 and 6002. RAL 8011 displayed a delta E of 2.6. All other delta E-values were below 2. Summarising all weathering results, an indication was obtained that regarding the waterborne 2K-PU coatings, Pefalon X-400 performed best with the applied pigments. The organic pigments used in RAL 5002 and RAL 6002 must be viewed critically, as the light fastness of these systems was not optimal. This was also found for the UV-curing coatings. Utilising UV-curing coatings from the investigated systems, only Bayhydrol 2690 provided a sufficient stability against weathering. Some RAL colours could also be realised with LUX 250, such as RAL 7040 and RAL 9016. Bayhydrol 2736/2960 was also possible with RAL 8011. It could be found that the combination of binder and pigments had the most important influence on the discolouration of the coating. The influence of the resin on the durability of the coatings could not be determined within the project duration as all tested coatings were still intact after the exposure to weathering. All samples are kept on the natural weathering device and will be inspected every 6 months. The project partners will be kept informed about the results beyond the official end of the project. The newly applied inorganic pigment did not provide the required RAL colour but showed a significant improvement in light- fastness. Bonding of WPC profiles was investigated by RTD performer University of Applied Sciences in Biel. Preliminary results indicate that the values for maximum bond strength achieved with the WPC profiles produced during this project appeared comparatively low. Investigations are still on-going beyond the official end of the project. Project context and objectives: The objective of this research project was to develop and manufacture a WPC window profile. To achieve this objective, the consortium aimed to identify suitable raw materials for a WPC formulation (natural fillers or fibres, thermoplastics such as PP and polyethylene, coupling agents such as maleic-anhydride modified PP and polyethylene, lubricants, UV protection agents, etc.), to compound the raw materials, to extrude the compounds into profiles, to adapt extruder dies to the new compounds, and to develop profile bonding techniques and suitable coatings. Specifically, the following objectives were identified for this project: 1. WPC material and coatings development: - Which formulation, i.e. composition of raw materials and additives (wood particles or fibres, thermoplastics (PP, PE, PVC), coupling agents, UV protection agents, etc.), should be chosen for a WPC window profile? - Which particle geometries should the fibres / particles possess to provide optimum stress transfer between the thermoplastic matrix material and the wood filler? - How can formulation components be compounded (mixed) and extruded? - Which WPC formulation(s) display optimum material properties (mechanical, physical, thermal properties, ageing, and coatability)? - Which coatings are suitable for WPC? Are any pretreatments (e.g. plasma treatment) required? Do coatings adhere directly on WPC? 2. Development of window profile prototypes: - Do compounders, extruders and dies need to be modified for processing of WPC profiles? - Can WPC formulations be equipped with a co-extruded cap made of PVC or an alternative thermoplastic material which may be coated? - Can WPC window profiles without co-extruded capstock be coated directly and are these profiles durable (i.e. resistant against UV light and moisture)? - How do coatings applied to our WPC material perform in comparison to commercially available WPC formulations? - Can window profiles be bonded or glued? - Do key properties of the extruded window profiles fulfil requirements of the specific standards? - How is the principal technical performance rated? Which are advantages and disadvantages compared to a PVC window profile? 3. Application of the extruded profiles for window prototype manufacturing: - Where are the strengths and weaknesses of the new window profiles? How can they be overcome with regard to the construction of the window? - How does the construction of a conventional PVC window need to be modified in order to accommodate the new raw material? - What is required in terms of marketing of the novel window profiles based on WPC? 4. Economic implications: - How does productivity of the WPC window manufacturing process likely compare to productivity of the PVC window production? - How can improvements in productivity be achieved? Project results: Results of work package (WP)1 (WPC formulation development): 1. Selection of raw materials (wood and other lignocellulosics, thermoplastic polymers, additives, coupling agents, UV stabilisers, lubricants): Commercial wood flour was selected as filler material due to price, availability, processability and performance. As stated in annex I (description of work), PP was chosen as the primary polymer for the matrix material of a WPC formulation intended for window profile extrusion. PP is a standard thermoplastic which is widely used in WPC extrusion in Europe. In addition, PP was blended with other polymers in order to improve mechanical, thermal and weathering properties of the WPC. Polymers with a melting point above 220 degrees of Celsius (PA 6, PA 66, PA 610, PC, PET, PBT, PEEK) are not suitable for extrusion with wood flour because wood will degrade at temperatures beyond 200 degrees of Celsius. Only polymers with a maximum processing temperature of 220 degrees of Celsius such as SAN, ABS, ASA and PMMA, PA 11, PA 12 and TPU can be used as formulation components in WPC extrusion. Typical coupling agents are maleic anhydride-modified PP (MAPP) for PP as a matrix material, and maleic anhydride modified polyethylene (MAPE) for PE as a matrix material. Both MAPP and MAPE were shown to reduce water uptake and improve mechanical properties of WPC. Other coupling agents used for WPC are isocyanates and silanes (Lu et al. 2000). According to Bledzki et al. (1998), the most effective coupling agent for WPC is polymethylenepolyphenylisocyanate. In this project, isocyanates were not used to due health and environmental concerns. Most coupling agents are proposed to facilitate chemical linkages and/or hydrogen bonds between the lignocellulosic material and thermoplastic (Felix and Gatenholm 1991, Lu et al. 2000). However, the wood-thermoplastic adhesion is believed to be dominated by physical adsorption and acid / base interaction if any (Wolcott 2001). Further additives used in WPC extrusion are, e.g. UV stabilisers, biocides and lubricants. UV stabilisers were not used in the WPC formulations in this project as they are part of the coatings or of the co-extruded capstock used. A lubricant was used to improve the processability of the WPC formulations. Lubricants are commonly separated into internal and external types. Internal lubricants affect viscosity and flow attributes because these additives are compatible with the resin of the melt, essentially lubricating resin molecules (Satov 2008). External lubricants affect anti-stick and slip attributes as these molecules will be incompatible with the melt, thereby separating and migrating to the surface of the melt, hence lubricating between the melt and the metal of process equipment (Satov 2008). 2. Preparation of wood and lignocellulosic raw materials (milling, screening, refining, sieving, etc.): Commercial softwood flour (Lignocel BK 40-90, J. Rettenmaier & Söhne GMBH & Co., Rosenberg, Germany) was used as received without further processing. Refiner wood fibers (thermomechanically produced wood fibers) and mechanically processed hemp fibers were previously investigated regarding their reinforcement potential for thermoplastic matrices (Schirp and Stender 2010). It was determined that compounding in a thermokinetic mixer is a useful step for processing of WPC with refiner and hemp fibres as little fibre damage occurred. However, during extrusion in a conical, counter-rotating twin-screw extruder, both natural fibre types were severely shortened due to strong shear forces, and homogeneous dispersion of fibres in the matrix was not achieved. Despite that, overall, WPC based on hemp fibres displayed the best strength properties and lowest water uptake of the formulations tested. Especially charpy impact bending strength was improved when hemp fibres were used instead of wood flour. One drawback of hemp fibres is their higher price compared to wood flour or wood refiner fibres. It was decided to use wood flour as filler for the thermoplastic matrix in this project. 3. Characterisation of wood and lignocellulosic materials (particle geometry, particle analysis): According to information supplied by the manufacturer, Lignocel BK 40-90 displays the following properties: - Particle structure: cubic - Particle size: 300-500 micron - Bulk density: 170 g / l - 230 g / l - Glow residue (850 degrees of Celsius, 4 h): ca. 0.5 % pH value: 5,5 +/- 1 Sieve analysis according to modified DIN 53734 provided the following results (data from Rettenmaier): - 500 um sieve opening diameter: 10 % max. residue, - 300 um sieve opening diameter: 65 % max. residue, - 150 um sieve opening diameter: 95 % max. residue. 4. Development of WPC formulations: Preparation and characterisation of PP-PA12-blends and PP-PA6-blends: During discussions between HakaGerodur and Fraunhofer WKI, it was decided to start with blends based on PP and PA. Four blends based on PP and PA12 were prepared in which MAPP was used as coupling agent. The blends were compounded in a 20 mm, co-rotating twin-screw extruder (Theysohn EKS-TSK 020) at 315-340 rpm, 180 - 200 degrees of Celsius and 12 bar, and subsequently pelletised using a strand granulator (Pelltec type SP50 pure). The granules were analysed using a differential scanning calorimeter (DSC; Mettler Toledo DSC822e). With DSC, the heat flow of a sample in mW is determined which is required to keep the sample at the same temperature as a reference sample. When the sample undergoes a physical transformation such as phase transitions, more or less heat will need to flow to it than the reference to maintain both at the same temperature. During the test, a small test sample (some mg) and a thermally inert reference sample are heated in crucibles at a specific heating rate in nitrogen or oxygen atmosphere. First, melting points of individual blend components were determined. The samples were heating up under a nitrogen atmosphere from - 20 to 190 degrees of Celsius with a heating rate of 10 K / min, thereafter cooling down from 190 to - 20 degrees of Celsius with a cooling rate of 10 K / min and once again heating up from - 20 to 220 degrees of Celsius with a heating rate of 10 K / min. The measured properties included heat of fusion (?Hm) and melting temperature (Tm). The melting temperature was taken of the second heating-up period. The heat of fusion was calculated from the melting peak area. The melting points of PA12 and PP were 178.2 and 163.44 degrees of Celsius, respectively. The melting point of MAPP was slightly lower than the melting point of PP (162.52 degrees of Celsius). Figures are provided in D4. The melting points of the four blends were also determined with the same method as described above. It was determined that the melting points of PP and PA12 did not change significantly with varying concentrations of the coupling agent. In addition to the PP-PA12 blends, two commercial PP-PA6 blends (KE 243/09 and KE 244/09, Poloplast GMBH & Co. KG, Leonding, Austria) were tested as formulation components in WPC. The PP-PA6 blends contained 20 % (KE 244/09) or 40 % (KE 243/09) PA6. Melting points of the blends were determined using DSC. The samples were heating up under a nitrogen atmosphere from -20 to 190 degrees of Celsius with a heating rate of 20 K / min, thereafter cooling down from 190 to - 20 degrees of Celsius with a cooling rate of 20 K / min and once again heating up from -20 to 250 degrees of Celsius with a heating rate of 20 K / min. It could be seen that the PP-PA6 and PP-PA12 blends each displayed two melting points. A larger peak at ca. 165 degrees of Celsius represents the melting point of PP. The smaller peaks of the two PP-PA6 blends at ca. 220 degrees of Celsius represent the melting point of PA6. From the melting peak area of PA6 the different contents of PA6 are visible. The peaks at approximately 175 degrees of Celsius represent the melting point of PA12. Preparation and characterisation of PP-TPU blend: A PP-TPU-blend was prepared in a 20 mm, co-rotating twin-screw extruder (Theysohn EKS-TSK 020) at 450 rpm, 180 - 200 degrees of Celsius and 9 bar, and subsequently pelletised using a strand granulator (Pelltec type SP50 pure). The blend consisted of 89.5 % PP (Sabic 510A), 9.5% TPU (Elastollan C80A, BASF) and 1 % MAPP (Scona TPPP 8112, Kometra). The PP-TPU-blend as well as the blend components were analysed using DSC. The samples were heating up under a nitrogen atmosphere from -60 to 210 degrees of Celsius with a heating rate of 10 K / min, thereafter cooling down from 210 to - 60 degrees of Celsius with a cooling rate of 5 K / min and once again heating up from - 60 to 210 degrees of Celsius with a heating rate of 5 K / min. TPU does not display any melting point due to its amorphous structure but the glass transition temperature at approximately - 40 degrees of Celsius can be observed. Melting points of the PP and the blend were both determined to be at 165 degrees of Celsius so under the conditions of this experiment, there was no change regarding the melting point of PP and the blend. The melting point of the MAPP was at approximately 164 degrees of Celsius. Preparation and characterisation of PP-ABS blends: The blends were prepared in a 20 mm, co-rotating twin-screw extruder (Theysohn EKS-TSK 020) at 450 rpm, 176 - 202 degrees of Celsius and 13 bar, and subsequently pelletised using a strand granulator (Pelltec type SP50 pure). The blend components and blend ST 100210-A were analysed using DSC. The samples were heating up under a nitrogen atmosphere from 0 to 210 degrees of Celsius with a heating rate of 10 K / min, thereafter cooling down from 210 to 0 degrees of Celsius with a cooling rate of 5 K / min and once again heating up from 0 to 210 degrees of Celsius with a heating rate of 5 K / min. ABS (amorphous) did not display a melting point, however, the glass transition point can be seen at 103 degrees of Celsius. The two small melting peaks at approximately 125 degrees of Celsius in the curve of the coupling agent (Admer SF 730E) represent the PE-backbone of the coupling agent. The melting points of the PP (Sabic 510A) and the blend were both at 165 degrees of Celsius. A small peak in the blend curve at approximately 150 degrees of Celsius likely represents the shifted melting point of the PE-backbone of the coupling agent; this shift was probably due to the reaction of the PE backbone with the PP. Dynamic mechanical analysis (DMA): DMA can potentially provide valuable molecular and morphological information about a material in the solid state by subjecting it to dynamic loads over a broad range of temperature and frequency. During measurement, a sinusoidal strain is applied to the sample, while measuring the sinusoidal stress response. A portion of the response output is in phase with the strain input and represents the energy stored in the material or the elastic component. The remaining response is out of phase with the strain and represents the energy dissipated by the material or the viscous component. DMA was conducted in dual cantilever mode in a Tritec 2000 analyser (PerkinElmer; formerly Triton Technology Ltd, Keyworth, United Kingdom). Dynamic temperature scans specific to the several polymers were conducted with selected specimens at sequential frequencies of 1, 5 and 10 Hz and with an amplitude of 50 µm. Prior to beginning of measurements, specimens were cooled using liquid nitrogen. When the required temperature was reached, measurements were started using a constant heating rate of 2 K / min. Three specimens with 2 % coupling agent and two specimens with 4 % coupling agent were analysed using DMA. The peak of the loss modulus curve at approximately 5 degrees of Celsius corresponds to the glass transition temperature of the PP. The glass transition temperature of the ABS is less clearly visible in the loss modulus curve. Activation energies for the glass transition temperatures of the blend components PP and ABS were calculated using the Arrhenius equation (Turi 1997). Activation energy is described as the energy required to facilitate a reaction between two molecules or the energy required to cause a molecule of a liquid or chain segment of a polymer to jump from its present position to a nearby hole (Son et al. 2003). For composites, high activation energies are associated with large degrees of interactions between polymer matrix and filler. It could be seen that when the amount of coupling agent was doubled, the activation energy for the glass transition of the PP remains unchanged whereas the activation energy for the glass transition of the ABS was doubled from 602 kJ / mol to 1212 kJ / mol. This result indicates that interfacial adhesion of the blend is improved when the amount of coupling agent is doubled. The results obtained using DMA show that this method is well suited to determine the quality of the interfacial adhesion in blends and WPC compounds and is useful for further investigations. Preparation and characterisation of PP-PMMA blend: A PP-PMMA-blend was prepared using a 20 mm, co-rotating twin-screw extruder (Theysohn EKS-TSK 020) at 444 rpm, 176 - 200 degrees of Celsius and 12 bar, and subsequently pelletised using a strand granulator (Pelltec type SP50 pure). The blend consisted of 79% PP (Sabic 510A), 19 % PMMA (Plexiglas Formmasse 7H, Evonik) and 2 % coupling agent (Scona TPPP 2507FA, Kometra; MMA-grafted PP). The blend was analysed using DSC. The samples were heating up under a nitrogen atmosphere from - 60 to 210 degrees of Celsius with a heating rate of 10 K / min, thereafter cooling down from 210 to -60 degrees of Celsius with a cooling rate of 5 K / min and once again heating up from - 60 to 210 degrees of Celsius with a heating rate of 5 K / min. Since PMMA is amorphous, it does not display a melting point; however, the Tg at approximately 110 degrees of Celsius can be seen. The melting point of the blend was at 163.1 degrees of Celsius whereas the melting point of PP was at 165.4 degrees of Celsius. This decrease of the melting point indicates that a reaction between PP and PMMA occurred. Preparation and characterisation of PP-PS blends: PP-PS blends were prepared using a 20 mm, co-rotating twin-screw extruder at 205 rpm, 160 - 198 degrees of Celsius and 12 bar, and subsequently pelletised using a strand granulator. The blends and blend components were analysed using DSC. The samples were heating up under a nitrogen atmosphere from 0 to 210 degrees of Celsius with a heating rate of 10 K / min, thereafter cooling down from 210 to 0 degrees of Celsius with a cooling rate of 5 K / min and once again heating up from 0 to 210 degrees of Celsius with a heating rate of 5 K/min. It was determined that melting temperature and crystallinity decreased with decreasing amount of PS in the blend. Processing of formulations: WPC formulations with PP, PP-PA12-blend and PP-PA6-blend: PP-PA12 blend no. 3 and PP-PA6 blend KE 243/09 were selected for further processing, i.e. compounding with wood flour and MAPP followed by extrusion. The WPC compounds consisted of 70 % wood flour (Rettenmaier BK 40/90), 27 % blend, 2 % MAPP (Scona TP 8112 FA, Kometra) and 1 % lubricant (Clariant Licolub H12). A reference formulation was also prepared which consisted of the same components, except that the blend was substituted with PP (Sabic 510A). A thermokinetic mixer (type TSHK 100, Papenmeier / Lödige, Paderborn) was used for preparation of WPC compound. A plough-blade mixer (type FM 130 DS, Lödige, Paderborn) was used for cooling and further granulating of the compounds. The moisture content of the prepared compounds was approx. 1 % (wt). WPC formulations with PP, PP-TPU blend, PP-ABS blend and PP-PMMA blend: The polymer blends were prepared using a co-rotating 20 mm compounder as described previously. Compounding of the blends with wood flour, MAPP and lubricant was performed using a heating mixer and plough-blade mixer. Manufacturing of samples for testing: WPC formulations with PP, PP-PA12-blend and PP-PA6-blend: Extrusion at Fraunhofer WKI: Compounds were extruded into three-box profiles (70 mm x 17 mm, wall thickness 4 mm; flat surfaces) on a 54 mm, conical, counter-rotating twin-screw extruder (Battenfeld miniBEX 2-54C). A crammer feeder (type KSW, Kreyenborg GmbH, Münster, Germany) was used to add the compounds into the extruder. Specimens for mechanical and physical property testing were machined by conventional techniques. The largest sample dimension was coincident with the extrusion direction. After cutting, specimens were conditioned to constant weight (less than 1% change during 24 hours) at 20 degrees of Celsius and 65 % relative humidity and weighed to the nearest 0.001 g. WPC formulations with PP and PP-TPU blend: Extrusion at HakaGerodur: It was determined if the Fraunhofer WKI standard WPC compound could be extruded using processing equipment available at HakaGerodur in Gossau (Switzerland). 10 kg each of two compounds were prepared using a thermokinetic mixer. Compound 1 (standard Fraunhofer WKI compound) consisted of 70 % softwood flour (Rettenmaier BK 40/90), 27 % PP (Sabic 510A), 2% MAPP (Scona TPPP 8112, Kometra) and 1 % lubricant (Licolub H12, Clariant). Compound 2 consisted of the same components, except for PP which was substituted with 27 % PP-TPU blend. This PP-TPU-blend was prepared in a 20 mm, co-rotating twin-screw extruder and consisted of 89.5% PP (Sabic 510A), 9.5 % TPU (Elastollan C80A, BASF) and 1 % MAPP (Scona TPPP 8112, Kometra). Extrusion trials were performed in February 2010 at HakaGerodur using a 45 mm single-screw extruder and a tubular die with a 10 mm diameter and 1 mm wall thickness. This configuration was used to determine if small diameter, extremely thin profiles can be extruded using our wood-plastic compounds. This in turn served as an indicator if the materials can be extruded into more complex window profiles. Compound 1 could not be extruded using this configuration. Compound 2 could be extruded into profiles, however, calibration was insufficient so that the profiles were irregularly shaped. Extrusion temperature was between 165 degrees of Celsius (zone 1) and 180 degrees of Celsius (zone 4), and die temperature was 150 degrees of Celsius. A second extrusion trial was performed in March 2010 using both compounds and a 30 mm single-screw extruder. Oval profiles with 13.8 mm width, 5.6 mm thickness and 1 mm wall thickness were successfully extruded. Feeding of compound 1 was difficult due to the presence of some larger compound particles which had to be sieved out. Extrusion of both compounds was successful, however, processing of compound 2 was easier and conditions were more stable. It appears that the TPU component in the polymer blend has a beneficial influence on processing and material properties. Surfaces of profiles extruded with compound 2 displayed a smooth, glassy surface. It is expected that profiles extruded with compound 2 can be easily recycled due to the TPU component. Material properties will likely not be compromised following reprocessing. These advantages justify the slightly higher material costs associated with the incorporation of TPU (approximately 4 - 7 EUR per kg of TPU). WPC formulations with PP, PP-TPU, PP-ABS, PP-PMMA: Extrusion at Fraunhofer WKI: Extrusion was performed as described previously. In addition, profiles were extruded using Fibrex-WPC-compound which consisted of 50 % PVC and 50 % wood flour. These profiles were extruded to serve as references. Extrusion of the two formulations with PP alone as well as of the formulation with the PP-TPU blend worked well. However, during extrusion of the WPC compounds with the PP-ABS blends and the PP-PMMA blends, small particles were released from the profile surface after exiting the die. These particles were analysed using attenuated total reflectance - Fourier transform infrared spectroscopy (ATR - FTIR), and it was determined that the particles consisted of ABS and PMMA. Although the quality of the extruded profiles was not fully satisfactorily, specimens for mechanical and physical tests were prepared. WPC formulations with PP, PP-TPU, PP-ABS, PP-PMMA: Extrusion at HakaGerodur: 20 kg each of WPC compounds with PP-TPU blend, PP-ABS blend and PP-PMMA blend were prepared as described above at Fraunhofer WKI and sent to HakaGerodur for extrusion trials. These extrusion trials are described under WP3. Testing of key mechanical and physical properties of all extruded formulations was done at Fraunhofer WKI. Details were provided in Deliverable 4. In summary, the principal differences between the Fibrex-based and the PP and blend-based formulations are the following: (a) fibrex-based WPC display significantly lower water uptake and dimensional changes after water uptake (boiling water test) due to the lower content of wood in Fibrex; (b) flexural strength of WPC with Fibrex-based WPC and PP-based WPC appear to be similar; (c) flexural modulus of elasticity is significantly higher for WPC with PP or PP-based blends and 70 % wood flour; (d) charpy impact bending strength is significantly higher for WPC with Fibrex due to the higher polymer content. Results of WP3 (Development and extrusion of window profile prototypes): Fraunhofer WKI prepared the following three WPC compound samples using a plast agglomerator (Palltruder): (a) 50 % wood flour, 47 % Sabic 505P (MFI: 2,1), 2 % MAPP, 1 % lubricant; (b) 60 % wood flour, 37 % Sabic 513A (MFI: 5,5), 2 % MAPP, 1% lubricant; (c) 70 % wood flour, 27 % Sabic 510A (MFI: 10.5) 2 % MAPP, 1 % lubricant. These samples were sent to SAMP in Varese (Italy). SAMP decided that the 50 / 50 compounds should be used for an initial trial. Therefore, Fraunhofer WKI produced 300 kg of the 50 / 50 - compound using the Palltruder. A trial was performed at SAMP on 15 December 2010. A protocol (in German) was provided by Thierry Bauer of HakaGerodur (dated 17-12-2010). A Bausano 75 mm parallel twin screw extruder with screw temperature control was used. A hollow-core profile of 30 mm x 30 mm x 2 mm was extruded. Details of the first trial: - Zone 1: 150 degrees of Celsius, zone 2: 160 degrees of Celsius, zone 3: 165 degrees of Celsius, zone 4: 175 degrees of Celsius; adapter 175 degrees of Celsius - Die: 175 degrees of Celsius - Run without screw temperature control - Revolutions per minute: 8 - Material gets stuck in cooling system Details of the second trial: - Zone 1: 155 degrees of Celsius, zone 2: 165 degrees of Celsius, zone 3: 170 degrees of Celsius, zone 4: 180 degrees of Celsius; adapter 180 degrees of Celsius - Die: 180 / 190 degrees of Celsius - Run with screw temperature control - Revolutions per minute: 8 - Extrusion works but profile appears not homogeneous enough. Details of the third trial: - Zone 1: 155 degrees of Celsius, zone 2: 165 degrees of Celsius, zone 3: 170 degrees of Celsius, zone 4: 175 degrees of Celsius; adapter 175 degrees of Celsius - Die: 175 degrees of Celsius - Run without screw temperature control - Revolutions per minute: 4 - Extrusion works but due to high amount of polymer (50 %) vacuum calibration is required - As vacuum calibration is not available at SAMP, the trials were interrupted. 21 samples of extruded profiles (300 mm length) were sent to Fraunhofer WKI for testing. Due to the unsatisfying shape of the profiles, only water uptake of the samples was determined to obtain an indication about the profile performance. Water uptake was 7.4 %. It was determined using the boiling water test according to the quality and testing specifications for terrace decking made from wood-polymer composites (see http://www.qg-holzwerkstoffe.de/WPCTerrassendielen.cfm online for further details). The water uptake of the profiles was deemed too high considering that the water uptake of WPC profiles with 60 % PVC and 40 % wood was only 3.2 %. The remaining WPC compound (approximately 200 kg) was sent from SAMP to HakaGerodur for further trials. Additional WPC compound (200 kg) of the same composition (50 % wood flour, 47 % PP) was prepared at Fraunhofer WKI using the Palltruder and sent to HakaGerodur. From 15 to 17 February 2011, trials were run at HakaGerodur. Thierry Bauer documented the trials in two protocols (dated 18 February 2011). In the first trial, a window profile tool (120 mm x 50 mm x 2 mm) was used on a Weber DS9.25 twin-screw extruder, however, the quality of the profiles was not satisfactory. Therefore, HakaGerodur switched to a square profile tool (30 mm x 16 mm x 2 mm) on a 60 mm Müller single-screw extruder. Again, no profiles could be extruded due to technical difficulties as explained in the protocol by Thierry Bauer. The next trial was performed on 23 February 2011. A protocol was provided by Thierry Bauer (dated 23 February 2011). A flat profile tool (83.5 mm x 5 mm x 2.6 mm) was used on a Weber DS9.25 twin-screw extruder. A successful extrusion was not possible as was outlined in the protocol. The next trial was performed on February 24, 2011. A protocol was provided by Thierry Bauer (dated 25 February 2011). An oval profile tool (17.7 mm x 8 mm x 3 mm) was used on a Müller 60 mm single-screw extruder. However, using the Palltruder compound (50 % wood, 47 % PP), oval profiles could not be successfully extruded due to the difficulties explained in the protocol. Despite the difficulties experienced during the trials, HakaGerodur was interested in obtaining further WPC compound for extrusion trials. Fraunhofer WKI provided a list with possible compound formulations. Thierry Bauer from HakaGerodur chose the following composition: 50 % wood flour, 47 % PP (Sabic 510A with MFI of 10.5) 2 % MAPP and 1 % lubricant. Fraunhofer WKI prepared 70 kg of this WPC compound using a heating mixer in April 2011. The compound was processed at HakaGerodur on a 45 mm Müller single-screw extruder using a tool with the dimensions 30 mm x 16 mm x 2 mm. The trial was performed on 23 June 2011 and documented by Thierry Bauer in a protocol (dated 24 June 2011). The compound could not be extruded into a profile. In addition, the compound was extruded using a tool of 83.5 mm x 5 mm x 2.6 mm (flat profile) in a trial on 25 August 2011. Thierry Bauer provided a protocol for this trial (dated 29 August 2011). Results of WP4 (Testing of window profile prototypes and development of profile bonding techniques): Since no window profile prototypes could be extruded, no testing was performed. However, other profiles which were extruded at HakaGerodur and Fraunhofer WKI were tested. Please see results of the previous WP for additional information. Bonding of profiles was investigated at FH Biel. Results were reported in D7. RESULTS OF WP 5 (Estimation of potential economic impact): Potential economic impact of a WPC window is expected to be high. A strengths, weaknesses, opportunities, and threats (SWOT) analysis for a WPC window profile was performed: Strengths: (a) substitution of a significant amount of PVC with wood, a renewable, ecologically friendly resource; (b) high stiffness (modulus of elasticity) of WPC; (c) low thermal expansion of WPC; (d) low thermal conductivity of WPC. Weaknesses: (a) low impact strength of WPC; (b) higher water uptake and lower dimensional stability compared to PVC; (c) bonding, gluing needs to be improved; (d) difficulty to produce thin-walled profiles. Opportunities: (a) customers are interested in obtaining ecological, sustainable products; (b) customers are willing to pay more if solutions offered are technically or ecologically superior; (c) possibilities for coatings and co-extrusion exist, i.e. coloured profiles. Threats: (a) changes in processing required (dies, screw design, processing parameters); (b) investments required; (c) scepticism regarding new technologies; (d) lack of long-term experience with WPC products. A cost comparison for a PVC and a WPC window profile was provided by Hanspeter Kuster in deliverable 9. Results of WP6 (Training, dissemination and exploitation): Two project meetings were held at Fraunhofer WKI where project participants were shown how wood particles for WPC compounds were prepared, how WPC compounds were prepared using a heating mixer, co-rotating twin-screw extruder or plast agglomerator (Palltruder) and how profiles were extruded into tape and box profiles using a 54 mm, counter-rotating twin-screw extruder. Regarding dissemination, it is planned that the most important project results will be published in a scientific journal. Potential impact: The consortium learned that extrusion of a WPC profile is extremely complex and challenging and that the effort in accomplishing each WP was underestimated. Not every WPC compound could be run with any tool and extruder. The development of a tool for a specific compound formulation and application is necessary and should be considered in future projects. A positive aspect was that various WPC compounds could be tried on different extruders and therefore, we were able to gather important experience for future projects. Regarding dissemination, it is expected that some of the project results will be published in a scientific journal. List of websites: http://www.extruwin.eu/ Related documents Final Report - EXTRUWIN (Extruded window profiles based on an environmentally friendly wood-polymer composite material)
