Study of the process of impregnation of basalt fibre filler by phenol formaldehyde binder

Transcription

1 Plasticheskie Massy, No. 6, 2009, pp Study of the process of impregnation of basalt fibre filler by phenol formaldehyde binder I.D. Simonov-Emel yanov, N.L. Shembel, A.N. Trofimov, P.V. Surikov, and A.S. Kuklin M.V. Lomonosov State Academy of Fine Chemical Technology, Moscow Selected from International Polymer Science and Technology, 37, No. 3, 2009, reference PM 09/06/32; transl. serial no Translated by P. Curtis Abstract The optimum parameters of impregnation of reinforcing basalt fibre filler by phenol formaldehyde resins of grades BZh-3, LBS-20, and SFZh-302, and also by binders modified with epoxy resin ED-20, have been determined. The pultrusion process is one of the main ways to produce reinforced profiles of different cross-section from polymer composites with a continuous filler [1]. It includes a number of individual process stages combined into a single continuous production cycle: application, impregnation, squeezing out, drying, and curing of the binder. The process is conducted in a continuous dynamic regime, which requires the coordination of individual process stages both in terms of the values of the parameters and in terms of time. Mathematical modelling of the process in a dynamic impregnation, drying, and curing regime presents considerable difficulties and is a separate problem that will not be considered in the present article. The main stage of the pultrusion process is impregnation. The reinforcing fibre filler is delivered at a certain speed into the impregnation bath, which is filled with polymer binder, and is coated with the polymer binder. In the process of application, the binder begins to fill the internal pores (interfibre space) of the fibre filler, and also coats the filler with a layer of certain thickness. A more detailed examination of the process of binder application to reinforcing filler was given in reference [2]. Known equations [1] for solving the impregnation problem the Darcy and Washburn equations require knowledge of values of the corresponding constants, which, for the given form of reinforcing filler, can be determined only experimentally. The main parameters for solving the impregnation problem are the characteristics of the fibre filler (the density, porosity, and structural parameters of the fibre), the surface tension and the rheological characteristics of the polymer binder, and the parameters of the pultrusion process (temperature, pressure, speed, and time). As reinforcing filler, we used ground roving of complex basalt strands with an elementary strand diameter of 13 µm and a linear density of 2520 tex, treated with oil KV-02, with 60 twists per 1 mm of roving, produced according to the TU specifications (notation NRB KV-02(0-60)). The specific tensile load for roving with an elementary strand diameter of 13 µm is at least 500 mn/tex (gf/tex). Roving treated with oil KV-02 is compatible with epoxy, phenolic, epoxyphenolic, melamine, and acrylate resins. As binders, we used liquid resol phenol formaldehyde resins (PFRs) of three grades: LBS-20, SFZh-01, and BZh-3. Table 1 presents the main characteristics of the PFRs used. In order to select and substantiate the temperature regime for impregnation, an investigation was made of the dependence of PFR viscosity on temperature and content of modifying additive, in the present case epoxy resin ED-20. Binder viscosity was determined by the standard method on a Brookfield rotation viscometer [3]. Figures 1 to 4 show the dependences of PFR viscosity 2010 Smithers Rapra Technology T/47

2 Table 1. Characteristics of PFRs ( Karbolit OJSC, Orekhovo-Zuevo) Resin grade SFZh-301 LBS-20 BZh-3 Batch number Solvent Ethyl alcohol Ethyl alcohol Water Nominal viscosity, s 7/(2 12) Dynamic viscosity, mpa s 1460/( ) 1013/( ) Mass fraction of dry residue, % 84.9/ /(71 78) Analysis conditions: temperature 100 ± 3 C 100 ± 3 C time 45 min 2 h 20 min Mass fraction of free phenol, % <9.0/ 9.0 <10.0/ /(8 16) Mass fraction of free formaldehyde, % 0.7/ 2.0 Gelation time at 150 ± 2 C, s 77/(80 110) 123/( ) Mass fraction of water, % 12.3/ 19 Bakelisation losses, % (T = 180 ± 5 C, 1 h) 27.2/ 30 Alkali content, % 0.3 Note. Numerators actual quality indices of obtained resins; denominators quality index standards according to existing specifications. For LBS-20, the quality index standards for higher-grade lacquer are shown Figure 1. Temperature dependence of the viscosity of liquid phenol formaldehyde resins of grades LBS-20, SFZh-301, and BZh-3 Figure 3. Dependence of the viscosity of SFZh-301 resin on temperature and content of ED-20 resin Figure 2. Dependence of the viscosity of LBS-20 resin on temperature and content of ED-20 resin Figure 4. Dependence of the viscosity of BZh-3 resin on temperature and content of ED-20 resin T/48 International Polymer Science and Technology, Vol. 37, No. 9, 2010

3 on temperature and content of modifying additive. At room temperature, PFRs have a fairly high viscosity (20 40 Pa s), which hinders the process of impregnation of the basalt fibre roving. With increase in temperature to 80 C, the viscosity decreases in all cases, with the exception of BZh-3 with 20 and 30% ED-20 (Figure 4), to 200 mpa s. Thus, there is a reasonable range of viscosity values, which makes it possible to control the impregnation process through the binder temperature in the bath. In accordance with our calculations in reference [2], Figures 5 to 7 show the interrelation between the amount of binder applied and its viscosity, the surface tension, and the speed at which the filler is pulled through the bath. Obtained data [2] on compaction of elementary fibres, basalt strands, and roving made it possible to establish that the minimum binder content in finished articles should be at least 38 vol% in order to fill the cavities and produce high-quality articles. With account taken of the content of solvents and the weight losses (Table 1) during drying and curing (bakelisation, i.e. conversion to resite) of liquid binders based on PFR, the volume content of binder applied to the surface of the fibre filler should lie in the range 43 50%. The speed at which the fibre filler is pulled through the bath and the entire unit depends on the design (length) of the curing zone, the temperature, and the kinetics of curing of the binder. We showed [4] that, for rods of mm diameter, the real speed of pultrusion amounts to m/min, and consequently, for application of the required amount of binder at such a speed, the binder viscosity should lie in the range mpa s (Figure 5). Analysing Figures 1 to 4, we can see that the investigated binders must be heated to C. However, account must be taken of the fact that, with increase in temperature, the inert solvent begins to evaporate from the binder, which leads to an increase in its viscosity. The rate of curing in resol PFRs increases with temperature, which again leads to an increase in binder viscosity. The introduction into PFRs of epoxy resins in small quantities as modifiers leads to a small increase in viscosity. During prolonged storage at T = 20 C, for a binder based on PFR of grades LBS-20, SFZh-301, and BZh-3 there is a change in viscosity (Figure 8). For practically all PFRs, during storage in air for days, the viscosity changes little and then increases sharply, which is connected with the gradual evaporation of the inert solvent and with the occurrence of a chemical reaction leading to the formation of a three-dimensional structure of the PFR. In order to prolong the life of binders based on PFR, it is recommended that Figure 5. Dependence of the required quantity of applied binder with a specified viscosity and a surface tension of 40 mn/m on the pull-through speed of the reinforcing fibre filler Figure 6. Dependence of the pull-through speed of the reinforcing fibre filler on the surface tension and viscosity of the binder, given the application of a layer with a thickness equal to 0.4 of the strand radius, which corresponds to a content of liquid binder in the polymer composite of 50 vol% Figure 7. Dependence of the pull-through speed of the reinforcing fibre filler on the surface tension, viscosity, and required quantity of applied binder: viscosity 100 mpa s (curves 1 and 2), 200 mpa s (curves 3 and 4), and 400 mpa s (curves 5 and 6); content of liquid binder in the polymer composite: 45 vol% (curves 1, 3, and 5) and 55 vol% (curves 2, 4, and 6) 2010 Smithers Rapra Technology T/49

4 layer in the impregnation bath of 100 mm, with account taken of the density of the resin, equal to 1150 kg/m 3, amounts to P a = Pa. The capillary pressure P c as the binder runs into the interfibre slit of thickness H was calculated by means of the formula P c = 2s/H (1) where s is the surface tension (mn/m) and H is the size of the pores between elementary fibres (m). The impregnation pressure was calculated by means of the formula Figure 8. Dependence of the viscosity of PFRs of different grades on the storage time P impreg = P a P c (2) The pore size between elementary fibres was calculated for cubic packing of basalt fibres with account taken of monostrip porosity, H = m. Calculation of the capillary pressure by means of formula (1) (s = 30 mn/m, P c = Pa) and comparison of this pressure with P a showed that impregnation proceeds under the action of capillary forces: P c > P a. The radius of the bunch of fibres R 0 in the basalt fibre monostrip was calculated by means of the formula Figure 9. Model for calculation R 0 = H D T ρ f (3) they be stored at temperatures no higher than C in a cold chamber. To calculate the parameters of impregnation of reinforcing basalt filler by the investigated polymer binder compositions, use was made of the model and mathematical formulae proposed in reference [5]. Model approaches to studying the impregnation of complex fibre systems help us to analyse and correctly organize the pultrusion process, to calculate the overall size of the impregnation bath, and to optimize the speed of movement of the basalt roving. By way of example, calculations will be given concerning the determination of the temperature time and the force speed parameters of impregnation of basalt roving by resin BZh-3. The model for calculating binder filling of the gap between fibres, H, under external pressure, P a, is shown in Figure 9. The necessary data for calculation are the linear density T = m/h, where m is the mass of a basalt fibre monostrip and h is the length of a basalt monostrip segment. For monostrip, the linear density T = kg/m. The density of the basalt fibre P f = 2800 kg/m 3. The external pressure with a height of the binder where H is the pore size (m), D is the diameter of an elementary fibre (m), T is the linear density (kg/m), and r f is the density of the fibre (kg/m 3 ). The dimensionless interfibre gap H * was calculated by means of the formula H * =H/R f (4) where H is the pore size (m) and R f is the radius of the fibre (m). The dimensionless time of impregnation of the interfibre space, t * impreg if, was calculated by means of the formula t * impreg if = 0.75/H* (1 + H * ) + (0.75/H *3/2 ) arctg 1/H *1/2 where H * is the interfibre gap, a dimensionless quantity. The coefficient of permeability of the interfibre space K x is determined by means of the formula K x = R 2 * f /6t impreg if (5) (6) The total impregnation time of the bunch of fibres was calculated by means of the formula T/50 International Polymer Science and Technology, Vol. 37, No. 9, 2010

5 t total = 4 η R 2 * 0 t impreg if /9 K x P impreg (7) where h is the viscosity of the binder (Pa s), R 0 is the radius of the bunch of fibres (m), t * impreg if is the dimensionless impregnation time of the interfibre space, K x is the coefficient of permeability of the interfibre space (m 2 ), and P impreg is the force exerted during impregnation (Pa). Figures 10 and 11 give data on the parameters of radial impregnation of basalt roving by resin BZh-3 as a function of the interfibre distance H and viscosity, which were calculated in accordance with the model concepts described above. Analysis indicates that the impregnation of basalt roving is not a limiting stage of the pultrusion process. The use of an impregnation bath of 500 mm length with a reserve ensures impregnation at a pull-though speed of m/min, and here the time for which the roving is in the impregnation bath amounts to no more than 25 s. Even with an interfibre gap in the basalt monostrip of µm (Figure 11), for complete impregnation a time of 6 18 s is necessary, which is ensured with the given bath dimensions. According to data of dynamic impregnation, for all the polymer binders used it was established that, with the specified physicochemical and rheological parameters of these liquids, with impregnation rates of no more than 1 m/min, a high impregnation quality is achieved, and the porosity of the system does not exceed 2%. The structure of the basalt roving makes it possible to carry out the impregnation process, i.e. filling of the interfibre space with liquid polymer binder, and also to apply the binder to the external surface of the roving, which promotes monolithisation of the basalt-fibre-reinforced plastic. The ratio of the components and basalt fibre in rods should be controlled using squeezing devices of different design. In the pultrusion process, it is most expedient to squeeze out excess binder by controlling the geometric dimensions of the rod being formed with the aid of spinnerets of specified diameter, taking into account that the shrinkage of a polymer composite of reinforced structure during curing exceeds fractions of a percent, while the entire volume of the solvent will be removed at the drying stage. The diameter of the squeezing spinneret should not be lower than a certain critical value at which failure of the basalt fibre occurs [6]. Thus, the optimum temperature time and force speed parameters of impregnation of reinforcing basalt fibre filler by phenol formaldehyde resins of grades BZh-3, LBS-20, and SFZh-302, and also by binders modified with epoxy resin ED-20, have been determined. REFERENCES Figure 10. Dependence of the coefficient of permeability of the interfibre gap K x, the impregnation pressure P impreg, and the dimensionless impregnation time of the interfibre space t * impreg if of a basalt fibre monostrip on the interfibre gap H Figure 11. Dependence of the total time of radial impregnation of a basalt fibre monostrip on the interfibre gap H and the viscosity of the resin 1. M.I. Terent eva and V.V. Il in, Manufacture of products from polymer composites by pultrusion. Tekhnika, Ekonomika, Informatsiya. Ser. Tekhnika i Tekhnologiya, No. 1, 1985, pp I.D. Simonov-Emel yanov et al., Investigation of the process of applying polymer binder to the surface of fibre reinforcing filler. Konstruktsii iz Kompozitsionnykh Materialov, No. 3, I.M. Belkin et al., Rotation Instruments. Measurement of the Viscosity and Physicomechanical Characteristics of Materials. Mashinostroenie, Moscow, 1967, 272 pp. 4. I.D. Simonov-Emel yanov et al., Investigation of the process of curing basalt-fibre-reinforced plastic based on a phenol formaldehyde binder. Plast. Massy, No. 5, 2009, pp V.A. Goncharenko, Radial impregnation of a bunch of fibres under external pressure. Plast. Massy, No. 5, 2005, pp Smithers Rapra Technology T/51

6 6. I.D. Simonov-Emel yanov et al., Nomogram for determining the content of fibre filler in the production of articles from reinforced polymeric materials by pultrusion. Plast. Massy, No. 5, 2007, pp T/52 International Polymer Science and Technology, Vol. 37, No. 9, 2010