in Injection Moulding

Transcription

1 Microcellular Foaming with Supercritical CO 2 in Injection Moulding Microcellular Foaming with Supercritical CO 2 in Injection Moulding V. Goodship, R.L. Stewart, R. Hansell, E.O. Ogur and G.F. Smith WMG, IARC, Department of Engineering, University of Warwick, Coventry, CV4 7AL Received: 8 January 2004 Accepted: 27 January 2004 ABSTRACT This paper discusses the systems for microcellular injection moulding developed and patented by Trexel and IKV. It then introduces a third design with initial findings and recommendations for modifications. The results suggest that microcellular foaming is possible introducing supercritical carbon dioxide in the nozzle providing certain design criteria are met such as a sufficient pressure drop from the plastication unit to the mould to enable rapid nucleation. Results from experiments with polystyrene are presented with consideration for pressure drop, weight reduction and injection speed. INTRODUCTION A supercritical fluid (SCF) is any gaseous fluid that is compressed above its critical pressure and temperature. The properties of supercritical fluids are different to their non-supercritical counterparts for example density and viscosity can alter drastically in regions close to the critical transition. One such commonly used material is supercritical CO 2 (scco 2 ). CO 2 becomes supercritical at a critical temperature (Tc) of 31 C and a critical pressure (Pc) of 7.4 MPa making it relatively easy to obtain. This material is abundant, inexpensive, nonflammable and environmentally friendly. It also has a high solubility for nonpolar organic compounds. The properties of scco 2 have been widely utilised in a diverse field of applications ranging from reaction media for chemistry, extraction and separation processes such as chromatography, solvents for processing foods, and in the manufacture of powder coatings. (1) It has also become attractive to a wide range of polymer based applications, for example, as a solvent in polymerisation techniques (2), as an extraction agent, and for the removal of low molecular weight residues. This is because despite being a good solvent scco 2 does not dissolve polyolefins unless the MW is very low. It can also be used for improving the performance of polymer blend Cellular Polymers, Vol. 23, No. 1,

2 V. Goodship, R.L. Stewart, R. Hansell, E.O. Ogur and G.F. Smith properties (3). scco 2 acts as a plasticizer for many polymers and can be used to lower the melt and solution viscosities (4). However, it is its use as a foaming agent that is of relevance to this paper. Amorphous polymers swell under the influence of scco 2 and can absorb carbon dioxide to a greater extent than crystalline polymers, and therefore amorphous polymers have an increased potential for both plasticization and foaming (5). It was also found that there was a link between a polymers diffusivity and the cell sizes that are produced (6). Therefore, different polymers may require slightly different processing conditions if optimum design for microcellular foaming is to be achieved. According to Suh (7), a microcellular foam has a cell structure of less than 30 µm. The advantages of producing microcellular foams over standard cell sizes are numerous including better physical properties, the ability to produce thin walled sections and improved thermal insulation. Another important benefit of this process is the material savings that can be achieved for applications which to not require the full mechanical properties achieved with unfoamed materials. Three systems for producing foams with injection moulding using scco 2 will now be considered. These systems are the Mucell system by Trexel, the IKV system and an in-house system designed by University of Warwick. It should be noted a fourth system has been announced by Sulzer Chemtech, Switzerland. This appears to have many features similar to the IKV system however no further details are available at this time (8). All three systems represented here assume the creation of a single phase of polymer and gas with a solubility gradient. Much work has been done on the extrusion of microcellular materials. In extrusion it has been found that the gas diffusion process depends on solubility, diffusion rate and extrusion parameters such as foaming temperature and saturation pressure (in order to promote the development of a microcellular structure) (9). The supercritical fluid and polymer create a single-phase solution. For this, The CO 2 needs to be dissolved with the polymer at high pressure, the solubility being pressure dependent. Therefore lowering the pressure quickly to below the saturation point the CO 2 comes out of solution. Trexel: the MuCell system An injection moulding system using scco 2 is commercially available by licence from Trexel, Inc., Woburn, MA. The technology is known as MuCell. 26 Cellular Polymers, Vol. 23, No. 1, 2004

3 Microcellular Foaming with Supercritical CO 2 in Injection Moulding There are numerous claims made for this process (10) for example it is claimed that this technology can lower the viscosity of the melt by up to 78 C, (especially useful if processing heat sensitive materials such as PVC leading to both reduced energy consumption and less chance of material degradation). The reduction in injection pressure that is produced can allow increases in the number of cavities and hence improved manufacturing efficiency. The MuCell process configuration includes a specifically configured screw and feed system designed to optimise the thermodynamic instability necessary to achieve rapid foaming. CO 2 is introduced into the injection barrel to form a single-phase solution with the polymer melt. A microcellular structure with injection moulding can be achieved. The following equipment is required to run Mucell: A metering system for the supercritical fluid (SCF) to ensure the fluid is at the appropriate pressure and volume. A Trexel designed screw and possibly new barrel Software and system modifications to create and maintain the uniformity required of the single-phase solution through the injection moulding cycle. This system introduces the SCF into the barrel and requires the purchase of both a license for the technology and machine modifications. The next two systems attempt to introduce the material into the nozzle rather than the barrel, making the change to foaming much simpler and cheaper. They are much newer and therefore less established than the Trexel system but enable further understanding on the basic requirements for microcellular moulding requirements. IKV Process An alternative technique is being developed by IKV, Aachen, Germany (11). In this system gas is introduced in an especially designed injection nozzle rather than the barrel as in the Mucell system. This nozzle is shown schematically in Figure 1 and is mounted between the plasticizing unit and the shut-off nozzle of a conventional machine. In order to produce a homogeneous distribution of gas in the melt a ring shape die design was used using a torpedo at the centre of the melt flow channel. This is manufactured using sintered metal which is permeable to the gas. Cellular Polymers, Vol. 23, No. 1,

4 V. Goodship, R.L. Stewart, R. Hansell, E.O. Ogur and G.F. Smith Figure 1. Gas injection nozzle schematic The results reported so far relate to the importance of the relative injection moulding parameter. Injection speeds, melt temperature and the concentration of the blowing agent are considered the key machine settings with this system. With a 8 mm wall thickness a density reduction of up to 66% was achieved using this system with PS. Generally larger weight reductions can be achieved using thicker cavities. This correlation seemed independent of the material types used in the experiments as illustrated in Figure 2. Figure 2. IKV results for various materials 28 Cellular Polymers, Vol. 23, No. 1, 2004

5 Microcellular Foaming with Supercritical CO 2 in Injection Moulding University of Warwick System A further simple bolt-on nozzle-type design was developed by University of Warwick to introduce scco 2 into the back of the nozzle through a sintered injection port. The nozzle also incorporated a series of Sultzer static mixers through which the gas and polymer mixture passed before injection into the tool. This design is shown in Figures 3 and 4. The mixers are relatively low shear, as a rough guide giving 55 s -1 for each g/s of injection speed for PS and 70s -1 for each g/s of PP injection speed. Processing trials were carried out using this system. Like the IKV design, the advantages of this design are: a simple bolt on design the machine can still be used for conventional injection moulding The trials were carried out on a Battenfeld BKT1500 machine using a 3 mm thick square plaque tool. The scco 2 was at 5 C and a pressure of 20.7 MPa. Materials Initial testing was focused on exploring the potential of the new nozzle system. The materials considered here are: Polystyrene (PS). BASF POLYSTY 143E GR2, Polypropylene (PP), Bassell K % talc filled PP, DSM Degradation The use of the static mixers to replace the conventional nozzle gave no evidence of any increased degradative effect of the polymer even using conventional injection moulding. This is due to the relatively low shear rates they generate. The Problem of Residence Time In initial trials it was found that the residence time of the gas in the nozzle was insufficient for foaming due to the fast injection times common in injection moulding processes. It was considered that the injection time of approximately 1 second did not allow for sufficient dosage/diffusion of gas in the melt. Only on using very low injection speeds of 25 ccm, to generate slower filling and Cellular Polymers, Vol. 23, No. 1,

6 V. Goodship, R.L. Stewart, R. Hansell, E.O. Ogur and G.F. Smith Figure 3. New nozzle design 30 Cellular Polymers, Vol. 23, No. 1, 2004

7 Microcellular Foaming with Supercritical CO 2 in Injection Moulding Figure 4. Manufactured nozzle. Gas is injected between the screw tip and the Sultzer static mixer hence a slower passage of material through the nozzle was any sign of a foaming effect evident. This enabled gas injection blow times of up to 8 seconds, however full filling of the tool was not possible as the gate was freezing off. Therefore a way to increase the residence time of the gas in the melt was required. Blow times were increased from just the injection time up to 15 seconds. In order to enable this, screw recovery was kept at a relatively slow level to enable the polymer to be moving until the blow time was complete. This had the added benefit of increasing the surface area of the polymer available to the gas. The injection carriage was kept in its forward position during moulding to reduce any drooling effects and keep the pressure in the melt. A low back pressure of 0.7 MPa was used to encourage migration of the gas from regions of high pressure (injection point) to regions of low pressure (in front and behind injection point). An improved design would be to incorporate a shut-off nozzle after the static mixers, this would improve the systems ability to maintain and control the internal melt pressure. Increasing the blow times had the desired effect of feeding gas more effectively in the polymer to enable further mixing/diffusion. There appears to be sufficient melt pressure (profiled temperatures with a colder feed have been used) to keep the gas in solution in the injection cylinder. If any gas was feeding back down the screw in a melt fingering type effect, such as seen in gas assisted moulding the additional mixing of the plasticizing screw would also aid the homogenising. The total diffusion time can thus be increased to incorporate both screw recovery and cooling time. Blow times incorporating regions when the melt was stationary i.e. blowing after screw recovery was complete, produced very poor mouldings, with heavily pitted surfaces caused by regions of gas and polymer stratification. Cellular Polymers, Vol. 23, No. 1,

8 V. Goodship, R.L. Stewart, R. Hansell, E.O. Ogur and G.F. Smith One major drawback with this very simple dosing system is that it is impossible to judge how much gas is in the system and only trial and error can be used to decide a sufficient blow time. However, this would be easily rectified with a gas metering system Nucleation It was necessary to use low levels of chemical blowing agent (CBA) to initiate the foam effect. Talc filled PP was of especial interest as the talc should act as a passive nucleation agent or kicker. However, with gas alone, foams could not be induced except at very low injection speeds. With just CBA alone (2%) the cell sizes produced appear uncontrolled and overlarge. The addition of low levels (>0.2% by wt) of active nucleation agents and CO 2 gas together have also been assessed, the chemical blowing agent used in the work reported here was Supercell C, azodicarbonamide from Americhem. 0.1% in combination with the CO 2 gas appears to be the optimum loading. Lower levels are difficult to dose or disperse accurately unless available as a masterbatch producing mouldings of irregular properties. The foaming characteristics appear to be linked to the processing parameter of injection speed and the concentration of blowing agents, both physical and chemical. Injection temperature is also likely to have an effect however the trials carried out so far have been limited by the activation energy of the CBA (220 C). As an example, for a moulding of 3mm thickness, a combination of 0.2% CBA and gas gave a weight reduction of 15% with a blow time of 10 seconds for talc filled PP. The total residence time of the gas in the polymer before injection is estimated to have been around 40 seconds which incorporated the cooling time of the previous moulding. Therefore no increase in total cycle time was required as a slow screw recovery speed was used to dose/diffuse the gas. Pressure Drop One reason why CBA were still required may be in the design of the nozzle system itself. The critical pressure drop required for microcellular foaming of HIPS is reported as being around 10 9 Pa/s (12) however such critical pressures are likely to be highly material dependant due to factors such as crystallinity, branching and additives which can all affect the number of available nucleation sites. This pressure drop can be calculated as follows: 32 Cellular Polymers, Vol. 23, No. 1, 2004

9 Microcellular Foaming with Supercritical CO 2 in Injection Moulding dp dt 32µ DV = 6 d 4 2 Where D= screw diameter (0.06 m), µ- material viscosity (50 pa.s), d = nozzle orifice diameter (0.003 m), V= injection speed The figures in brackets relate to the machine/material/nozzle specification used in these experiments. For the injection speeds used in these experiments for PS, the calculated pressure drop is shown in Table 1. It can be seen from Table 1, that as the screw speed is increased the pressure drop also increases. The critical pressure required to create microcellular foams is however also related to the gate size, the material, the gas percentage and the melt temperature. The relationship between pressure drop and weight reduction can be seen in Table 2. For these experiments polystyrene material with 0.1% CBA and CO 2 gas was moulded at a variety of injection speeds, all other parameters were kept the same. Material dosage was increased until a full mould filling was achieved. Mould weights per run were found to be consistent to +/- 0.1 g. Table 1. Pressure drop for polystyrene mouldings I njection speed (ccm/s) Pressure drop (Pa/s) Table 2. Effect of injection speed on weight reduction injection (ccm/s) standard (100) speed moulding plaque thickness moulding weight (g) weight reduction (%) 3 mm Cellular Polymers, Vol. 23, No. 1,

10 V. Goodship, R.L. Stewart, R. Hansell, E.O. Ogur and G.F. Smith It can be seen from Table 2 that there is a link between injection speed and potential weight reduction. Higher speeds lead to more foaming as can be seen from comparing the SEM pictures shown in Figures 5-8. Figures 5 and 6 show the level of foaming at 20 ccms and 250 ccms respectively. The sizes of the cells produced however are quite similar as shown in Figures 7 and 8. The elliptical shape is due to shear distortion. It should also be noted that the surface finish was unblemished. There are considerably more cells produced at a higher injection speed. This is due to an increased pressure drop in the tool at a faster speed. At a further increased injection speed the pressure drop would continue to rise and more cells would be achieved until a saturation point was reached. In these experiments, a sufficiently high pressure drop was not generated to achieve Figure 5. SEM of polystyrene sample (injection speed 20 ccms) Figure 6. SEM of polystyrene sample (injection speed 250 ccm) 34 Cellular Polymers, Vol. 23, No. 1, 2004

11 Microcellular Foaming with Supercritical CO 2 in Injection Moulding Figure 7. SEM of polystyrene moulding (injection speed 20 ccms) Figure 8. SEM of polystyrene sample (moulded at 250 ccms) full saturation. Making the nozzle orifice smaller could easily rectify this. The maximum pressure drop here of Pa/s is insufficient for microcellular moulding of polystyrene. It is the rate of pressure reduction which controls the time period for nucleation and growth (13). In this case a higher pressure drop is required to increase the nucleation rate. CONCLUSIONS 1. The new nozzle system enables foams to be produced using scco 2. However Cellular Polymers, Vol. 23, No. 1,

12 V. Goodship, R.L. Stewart, R. Hansell, E.O. Ogur and G.F. Smith There must be a sufficient residence time of the gas for diffusion to occur. This can be achieved by using blow times beyond the injection time e.g. With PS approx. 10 seconds. The residence time can be increased to incorporate screw recovery and cooling time offering a larger surface area of polymer for gas diffusion. Low levels of CBA (0.1%) are required to initiate foaming. This may not be necessary when further design modifications are carried out to give a greater pressure drop. 2. There is a clear link between injection speed and weight saving. 3. There must be a sufficient pressure drop to enable rapid nucleation and cell growth at a saturation level. This can be achieved with the design of the nozzle and with consideration of factors such as tooling. 4. It is necessary to maintain pressure in the plastication unit. In this case this was achieved by using the counter-pressure of the moulding in progress. This can be more effectively achieved by incorporating a shut-off nozzle in the design. 5. The critical pressure drop for microcellular moulding of PS exceeds Pa/s in this system. ACKNOWLEDGEMENTS This research was supported by the EPRSC, Faraday: Fluid-assisted injection moulding, grant number GR/R77483/01 Thanks also to Ian Lawson at Salzer Chemtech (UK) Ltd REFERENCES 1. Y.P. Sun (Ed), Supercritical Fluid Technology in Materials Science and Engineering: Synthesis, Properties and Applications, Marcel Dekker, New York, J.L. Kendall, D.A. Canelas, J.L.Young, J.M. DeSimone, Chem. Rev., 99, 2, (1999), M. Lee, C. Tzoganakis, C.B. Park, Adv in Polym technol, 14, (2000), Cellular Polymers, Vol. 23, No. 1, 2004

13 Microcellular Foaming with Supercritical CO 2 in Injection Moulding 4. M. Lee, C. Tzoganakis, C.B. Park, Annu Tech Conf ANTEC Conf Proc Volume 2, pp , Soc Plast Eng, Brookfield, CT, USA, Shieh, Yeong-Tarng, Su, Ian-Hon et al, J Appl Polym Sci., 59, 4, (1996), J. Wang, X. Cheng, X. Zheng, M. Yuan, H. Mingjun, J. He., J Polym Sci Part B, 41, 4, (2003), N.P.Suh, Innovations in Polymer Processing, Chapter 3, (J.F. Stevenson (ed)), Hanser/Gardner Publications, C. Smith, Licence-free alternative to Trexel s Mucell system in development, PRW, 12/12/ accessed 14/ 12/03 9. B.A. Rodheaver and J.S. Colton, Polym. Eng. Sci., 41, 3, (2001), D. Pierick and R. Janisch, Conference Proceedings of the Third International Blowing Agents and Foaming Processes Conference, 2001, 13 th -14 th March 2001, Frankfurt, Germany, Paper 19, Rapra Technology, Michaeli, W and S. Habibi-Naini, Conference Proceedings of the Third International Blowing Agents and Foaming Processes Conference, 13 th -14 th March 2001, Frankfurt, Germany, Paper 9, Rapra Technology, J. Xu and D. Pierick, Microcellular Foam Processing in Reciprocating-screw Injection Moulding Machines. Technical report from Trexel Inc., accessed 28/11/ S.K. Goel and E.J. Beckman, Polym Eng Sci., 34, 14, (1994), Cellular Polymers, Vol. 23, No. 1,

14 V. Goodship, R.L. Stewart, R. Hansell, E.O. Ogur and G.F. Smith 38 Cellular Polymers, Vol. 23, No. 1, 2004