Processing parameters and characterisation of flax fibre reinforced engineering plastic composites with flame retardant fillers

Composites: Part B 62 (2014) 12–18

Contents lists available at ScienceDirect

Composites: Part B journal homepage: www.elsevier.com/locate/compositesb

Processing parameters and characterisation of flax fibre reinforced engineering plastic composites with flame retardant fillers q Ahmed El-Sabbagh a,b,⇑, Leif Steuernagel a, Gerhard Ziegmann a, Dieter Meiners a, Oliver Toepfer c a

Institute for Polymer Materials and Plastic Engineering, Clausthal Univ. of Technology, Germany Design and Production Engineering Department, Faculty of Engineering, Ain Shams Univ., Egypt c Nabaltec AG, Alustraße 50 – 52, 92421 Schwandorf, Germany b

a r t i c l e

i n f o

Article history: Received 16 December 2013 Received in revised form 9 January 2014 Accepted 10 February 2014 Available online 18 February 2014 Keywords: A. Fibres A. Polymer-matrix composites (PMCs) B. Mechanical properties B. Thermal properties

a b s t r a c t This work studies the possibility of compounding natural fibres (flax) into engineering plastics (PA6 and PB6) and comparing the results with counterpart glass fibre composites. The problem in compounding is the difficulty to compound the fibres with such polymers of high melting temperatures without decomposing the natural fibre thermally. Preliminary experiments are tried to define the possible processing window using the kneader namely temperature, compounding time and shear rate. Fibre content is tried in range of 0–50 wt.% with 10% step. The mixing temperature covers the range around the melting temperature ‘Tm’ [Tm 20, Tm+20]°C. The use of pre-melting temperature in compounding would utilise the energy evolving by fibres mutual rubbing. Compounding time is optimised at the minimum level. Shearing rate is tried at 25, 50, 75 and 100 rpm. Optimum conditions are defined to be 210–230 °C and 200– 210 °C for PBT and PA6 respectively. Shearing rate is also defined to lie within 25–50 rpm. Two different additives of non-organic mineral and organic phosphate flame retardants are tried with the prepared composites either alone or in combination with each other. The loading of flame retardants is limited to 20 wt.% in order to leave a space for natural fibres as well as the polymer and to keep in turn the overall composite mechanical properties. A mix of 1:1 ratio between the both types of retardants is needed to reach V0 flame retardation level. Mechanical properties are even improved 30% in E-modulus and 4% in strength with respect to composites without flame retardants. However, the injection moulding is reported to be difficult because of the high viscosity and the parameters should be optimised regarding the desired flame retardance level and the required mechanical properties as well as keeping the fibres not damaged. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Application of natural fibres into the polyolefin plastics like polypropylene and polyethylene has been already deeply studied under different aspects; namely improving the fibre to matrix adhesion [1,2], flow and rheological behaviour [3,4] process optimisation [5,6] and even modelling of strengthening mechanisms [7,8]. These composites are even commercialised and still exploring fields of new applications [9,10]. However engineering plastics reinforced with natural fibres represent a more difficult challenge because of the higher melting point of these plastics. This requires q The results of this work have been presented at the 4th Conference of Natural Fibre Composites, Rome 17–18 October 2013. ⇑ Corresponding author at: Institute for Polymer Materials and Plastic Engineering, Clausthal Univ. of Technology, Germany. Tel.: +49 35323 722487; fax: +49 5323 722324. E-mail address: [email protected] (A. El-Sabbagh).

http://dx.doi.org/10.1016/j.compositesb.2014.02.009 1359-8368/Ó 2014 Elsevier Ltd. All rights reserved.

higher processing temperatures. For example, it is a common practice to inject PBT and PA6 at least by 250° and 230 °C respectively. These values are recommended to be increased by 20 °C in case of reinforced grades. This high temperature is supposed to result in the natural fibre decomposition firstly in hemicelluloses at 250 °C then in cellulose region starting from 300 °C [11]. That is why this topic of reinforcing engineering plastics with natural fibres rare in literature except trials [11–13]. These trials start with mixing Nylon polyamide PA6 with high purity cellulose fibres [12] as well as wood fibres [13,14]. These previous trials in 2001 lead to a patent later [15]. The patent [15] focuses on the design of extrusion process parameters where the polymer granulates are fed at the primary feeder of the extruder. Then the fibres are side fed to the extruder where low temperature profile along the extruder in Arcaya et al. [16] compounded flax, jute, pure cellulose and wood pulps with PA6 and PA6,6.

A. El-Sabbagh et al. / Composites: Part B 62 (2014) 12–18

Mixing temperature is kept again 5 °C above the physical melting point measured by Differential scanning calorimetry DSC. Mechanical properties are reported to improve but unexpectedly, nothing about fibre structure damage specially or burning features with PA6,6 is mentioned. Wu et al. [17] prepared compoite made of polybutyleneterephthalate (PBT) and sisal fibres using blade rotor at mixing conditions of 220–230 °C, 50 rpm and 20 min. Mechanical properties are not reported [17]. As observed, the literature does not show intensive trials regarding the engineering plastics composites with natural fibres the next 10 years. This is attributed to the difficulty of the compounding process conditions which should be strictly controlled to avoid natural fibre decomposition. These conditions are namely temperature and time as well as the shearing rate. The Successful implementation of natural fibres within engineering plastics will open new market areas for the application of natural fibre composites due to their low cost and low density as well as their process-ability in the available injection and extrusion machines [7,8].

2. Preliminary thermal characterisation and material selection Compounding of natural fibre with an engineering plastic has utmost a narrow processing window which allows relatively safe

13

processing without natural fibres damage as well as keeping the polymer in the molten phase. So it is required to select natural fibres with high thermal stability and resistance to decomposition. Fig. 1 shows the thermal gravimetric analysis (TGA) of some commonly used natural fibres namely flax, hemp, sisal and Jute. The temperature range under consideration is within 100–300 °C. The change at 100° depends on the amount of water absorption level which is not important for this study because the fibres will be executed to drying step before compounding. On the other side, the processing temperature should not reach high temperatures as 300 °C. As shown the weight derivative curves in Fig. 1, jute has the higher thermal stability and resistance to degradation but its trend to decompose after the onset temperature is very fast. Thus suggests unstable situations after crossing the onset decomposition temperature which in other words means more gas evolution. From the other side, it is obvious that flax has the least change in weight which means more stable behaviour even at the expected high processing temperature. This means that the degradation features, where lignin decomposes, is kept at minimum level in case of flax natural fibres. Away from the quantitative TGA results, some qualitative measures of thermal degradation can be also noticed like the degree of fibres’ dark coloration and fibres’ smell like burned wood. This study deals with commercial flax fibres of relatively good thermal resistance. Compared to its weight at 100 °C, flax fibre

Fig. 1. TGA of flax, hemp, sisal and jute (air atmosphere, 10 k/min).

Fig. 2. Processing window of natural fibre with engineering plastic (example of PBT and flax natural fibre).

14

A. El-Sabbagh et al. / Composites: Part B 62 (2014) 12–18

loses less than 2% of its weight at 225 °C in air medium and 4% at 250 °C for both atmospheres respectively. PBT and PA6 are selected from the category of engineering plastics because they have relatively low melting points at 223 and 219 °C for PBT and PA6 respectively. Fig. 2 shows flax fibres which is characterised with its relatively good thermal stability (TGA is shown) and the differential scanning calorimetry DSC. As shown in Fig. 2, the processing window depends on the gap between the melting of the polymer and the degradation temperature of the natural fibre at a certain mass loss. At reaching 2% mass loss, the degradation temperature is 225 °C whereas 3% mass loss corresponds to 250 °C. The 3% mass loss will be considered the natural fibre degradation temperature in this study. This assumption is valid by considering previous literature [18] where 5% is reported to be the average value of the onset decomposition temperature taken from TGA. The expected reduction in degree of polymerisation and tenacity is supposed to be high in case of long exposure to high temperature. For example, Gassan et al. [19] reported that due to the chain scissions the tenacity of flax and jute fibres will be reduced to 30% of its original value in case they are exposed to 210 °C for 120 min. But in our case, it is exposed to the high temperature for 10 min maximum which will lead to a reduction of 10–15% in fibre tenacity according to the same Ref. [19]. This reduction can be neglected for shorter times of processing such as in extrusion where the residence time is assumed to be less than one minute [20]. In case of successful compounding of natural fibres with engineering plastics, other additives will be needed to ensure easy mixing; process stabilisation and flame retardation and thus an applicable product can be obtained to compete with the available glass fibre GF composites. Composites of PBT and PA6 which are reinforced with GF and mixed with flame retardants are already studied and commercialised. Literature about such composites shows the application of different flame retardant systems. Casu et al. [21] reported the results of implementing brominated organic compound-antimony trioxide in different loadings (0–20) weight% with PBT 30%GF. The oxygen index increased from 20 to 32 at room temperature testing condition. From the other side, PA6 reinforced with 30% GF is compounded with aluminum hypophosphite together with melamine pyrophosphate or melamine cyanurate to prepare halogen-free flame-retardant GFPA6 composites [22]. The flame retardant mix is limited to 20% weight. The oxygen index is found to increase to 29 as well as V0 level of UL94 test. But the mechanical tensile strength is affected negatively by introduc-

ing the flame retardant whatever its type. For instance, Zhang et al. [23] reported that only a mixture of 10.3% brominated phenylethane and 3.7% antimony trioxide resulted in the reduction of PBT-30%GF strength from 109.3 to 96.1 MPa. The impact strength is as well is reduced 20%. In this work, it is aimed to add flame retardants of organic and inorganic bases with preference for inorganic metal hydrates for their non-harmful impact on the environment as well as their availability and low cost burden. Le Bras et al. [24] used the ammonium polyphosphate APP alone as well as synergised with pentaerythritol (PER) and melamine (MEL) for the Flax/PP 40:60 system. Schartel et al. [25] used the ammonium polyphosphate (APP) and the expandable graphite EG. He reached HB level with 25% APP and V-1 level with 25% EG. El-Sabbagh et al. [26] used the hydrated inorganic metals in percentages (less than recommended for pure polymer) to keep the mechanical properties. All the previously mentioned literature concerning the flame retardants were handling polyolefin thermoplastic polymers (poly propylene, polyethylene, etc.). Therefore in this study, the use of flame retardants in the engineering plastics reinforced with natural fibres and the behaviour to flame and smoke is necessary and it is supposed to have research novelty as no literature is found in the available literature sources. It will be aimed that the flame retardants not to exceed the 20% of the composite content in order to keep enough polymer for binding with the natural fibres.

3. Experimental work The material used in this study as follows. PBT ULTRADUR  B 2550 NATUR is supplied by BASF. PA6 is supplied by SCHULAMIDÒ 6 MV 14 is supplied by Schulman GmbH. Melt temperatures are 223 and 219 °C for PBT and PA6 respectively. Flax is supplied by Sachsenleinen- Germany and then internally treated by 5% NaOH for one hour and then overnight in the oven at 100 °C [26]. Alkalinisation helps in surface cleaning as well as to increase the available sites candidates for coupling the OH groups in the cellulosic fibre with the host polymer. Additives for flame retardants are nonhygroscopic Exolite 766 with organic phosphate base supplied by Clariant-Germany in the form of white powder. Bulk density is 400–600 kg/m3 and median particle size is 20–40 lm. On the other side, the mineral Actilox 200SM AlOOH supplied by Nabaltec-Germany. The median particle size is 0.35 lm. Use of minerals as flame retardants is environmentally accepted but the problem that

Fig. 3. Weight loss of flax fibre at isothermal condition for 15 min at different temperatures.

15

A. El-Sabbagh et al. / Composites: Part B 62 (2014) 12–18 Table 1 Experimental plan of PBT samples. Effect of: Speed (rpm) Temperature (°C) Fibre content (wt.%)

Values

Notes

25

50

75

100

210

220

230

240

0

10

20

30

25

50

75

100

200

205

210

220

0

10

20

30

40

50

–Fibre content is 30% –Temperature is 220 and 230 °C –Speed is 50 rpm –Fibre content is 30% –Temperature is 220 °C –Speed is 50 rpm

Table 2 Experimental plan of PA6 samples. Parameter Speed (rpm) Temperature (°C) Fibre content (wt.%)

Values

Notes

40

50

–Temperature is 205 and 210 °C –Fibre content is 30% –Speed is 50 rpm -Fibre content is 30% –Temperature is 205 °C –Speed is 50 rpm

Fig. 4. Temperature-speed for feasible compounding of (a) PBT/30% flax and (b) PA6/30% flax.

Fig. 5. Effect of a-kneading speed b-kneading temperature on composite strength.

the required quantity of mineral flame retardant is heavily needed which in turn affects the process-ability as well as the mechanical strength negatively. The compounding takes place in a kneader machine where the effect of processing parameters such as the mixing speed, temperature and the fibre content is studied. Fig. 3 shows the thermal gravimetric analysis TGA carried out in air on flax fibres at isother-

mal conditions. Loss in weight at the 200–230 °C did not exceed 3.5%. This is attributed to the rich amorphous share of hemicellulose share [6]. In order to avoid excessive thermal decomposition and fibre damage, the kneading time is kept minimum as required for good compounding at 10 min. Fibre content varies from 0 to 50 wt.% with 10% step. Kneading speed is tried at 25, 50, 75 and 100 rpm. The mixing temperature covers not only the melting

16

A. El-Sabbagh et al. / Composites: Part B 62 (2014) 12–18

Fig. 6. Effect of flax content on composite properties (a) tensile strength and (b) impact strength.

Table 3 UL 94 and LOI results of PBT/30% Flax at different mix ratios of Actilox:Exolite. Actilox (%) Exolite (%) UL94 LOI

0 0 HB 21.6

20 0 HB 23.6

10 10 HB 24.4

5 15 V0 25.6

0 20 V0 27.4

least 88 h according to ISO 291 for test room conditions. Tension tests were made using Zwick 0.25 ton tensile machine. Test is conducted and evaluated according to ISO 527-1. After optimisation of the processing parameters, single fibre content is selected for the further tests of the composites with the flame retardation properties. The flame retardants are added to the polymer matrix before the insertion of the natural fibres. The flame retardants mix should not exceed 20% of the composite weight. The Actilox and Exolite contents changes in 5% step namely (0/20%, 5/15%, 10/10%, 15/5% and 20/0%). The compounded materials are also injection moulded to the tensile specimens as well as the flame resistance samples namely the cone calorimetry test of 100  100  3 mm3 and UL94 samples 125  12.5  3 mm3, UL94 (Flame Chamber made by Atlas Material Testing Solutions) 130  13  4 mm3 specimens according to ASTM D 635. LOI (Stanton Redcroft) test was carried out according to ASTM D 2863 with specimen dimensions 120  10  4 mm3.

4. Results and discussion Fig. 7. Effect of flame retardants on tensile & impact strengths for composites of (a) PBT and (b) PA6.

4.1. Visual inspection

phase range but also the pre-melting temperature. Low temperature mixing depends on the energy generated during friction. Tables 1 and 2 present the parameters studied in the plan of the experiments. During kneading the temperature is measured to be compared with the set temperature value. This helps in defining if there is a rise in temperature due to fibre or polymer granulate friction in case of sub-melting compounding. The colour and the smell of the compounded material are observed to check if the natural fibre is visually well distributed and if that they have not been burned. The compounded material is then shredded. Granulates are injection moulded to 1BB form (Width  thickness  gauge length = 2  2  10 mm3) according to ISO 527–2 using Allrounder 220C 600–250, Arburg, Lossburg, Germany. Impact samples are injected according to ISO 291. sample cross section is 2  4 mm2. 10– 15 samples are tested according to ISO 179–1 with 1 J striker. The temperature pattern and the injection shear rate are selected to match the kneading temperature and speed. Samples of mechanical testing are conditioned at 23 °C/50% relative humidity for at

The visual test of the compounded samples regarding the fibre homogeneity and non-burning features are summarised in Fig. 4. The samples with proper aspects is called ‘OK’, otherwise it is called ‘Burn’. Fig. 4 shows the feasible temperature/speed conditions without getting symptoms of burned (thermally damaged) natural fibres as well as keeping suitable compounding conditions. The filled and non-filled markers are corresponding to the theoretical and the actual temperatures respectively. The difference between the theoretical and the actual temperatures is 10 °C maximum in case of PBT composites while that of PA6 is 13 up to 25 °C for 25 and 50 rpm speeds. At higher speed of 100 rpm, the elevation in temperature remains almost constant at 22 °C difference between the theoretical and the actual temperatures, while in case of PA6, the elevation in temperature increases to 45 °C. Therefore, the elevation in temperature in dependence on speed is prominent for PA6 composites. This elevation in temperature of PA6 composites more than PBT composites is attributed to the heat agitation which accumulated in the aliphatic polyamide

A. El-Sabbagh et al. / Composites: Part B 62 (2014) 12–18

17

Fig. 8. Morphology of tested samples after cone calorimetry tests.

Fig. 9. Effect of flame retardants on cone calorimetry aspects for composites of (a) PBT and (b) PA6.

and finally released in heat form while the aromatic benzyl ring of the PBT allows the internal energy dissipation through the delocalisation. It is also noticeable that the elevation in temperature decreases with the increase of the theoretical compounding temperature. For instance, PA6 at 100 rpm: the temperature rises from the theoretical values of 210 and 220 °C to 250 and 255 °C respectively namely 40 and 35 °C elevation in temperature. Hence the safety zone for processing the PBT natural fibre composites is limited by the solid line oval shape surrounding 210–230 °C and 25–50 rpm, while in case of PA6, the oval skewed shape is surrounding the 200–210 °C and 25–50 rpm. The oval shapes of dashed lines in Fig. 4 shows the same safety zones regarding the actual measured temperatures. 4.2. Mechanical testing results Speed and temperature effects on the of PBT and PA6 composite tensile strength are shown in Fig. 5. Effect of speed shows a slight direct proportion with the composite strength. However the stan-

dard deviations points to a non-significance. Complete insignificance of temperature effect is found concerning the tensile strength. The effect of the fibre content on the mechanical properties of the compounded composites is also studied. Fig. 6 shows the tensile modulus and strength. The tensile strength results are correlated with similar polymers reinforced with 22.5% glass fibres (commercial product). The obtained strength results are close to the glass fibre composites bearing in mind the lower density of the natural fibres. The optimum parameters of compounding flax with both PBT and PA6 are followed by the addition of flame retardant fillers which targets the production of attractive industrial products. Exolit 766 and Actilox percent are manipulated in 20% combination as shown in Fig. 7, namely 0/20, 5/15, 10/10, 15/5 and 20/0 weight percent of both flame retardants. Tensile strength does not show significant reduction with the addition of the flame retardant fillers while the impact strength in contrary is decreased to 15% and 40% for PBT and PA6 composites respectively. The combination of both

18

A. El-Sabbagh et al. / Composites: Part B 62 (2014) 12–18

flame retardants should not exceed 20% of the composite. V0 rating according to UL94 is reached with 20/0 and 15/5 of Exolit/Actilox while 10/10 system is limited to V1 rating. The results of UL94 and LOI tests are listed in Table 3 showing the PBT/Flax at different Actilox/Exolite percentages. The morphology of the cone calorimetry tested samples after being burned are shown in Fig. 8. The (10/ 10%) of Actilox/Exolite show coherent feature which is a good sign for the sample and its corresponding UL94 classification is ‘HB’. On the other side, the 5:15% has the best flame retardance level ‘V0’ but the surface is not coherent. This means that more study is needed to find an optimum value in between the both ratios where coherent surface and ‘V0’ classification. The LOI results show a matching trend as listed in Table 3. The quantitative results of the cone calorimetry tests are shown in Fig. 9 where the flame resistance in terms of heat release rate heat release rate HRR and its peak PHRR as well as the faster distinguishing time ‘tt flame out’ are seen. 5. Conclusion The study shows the possibility of producing high added value natural fibre engineering plastic composites. Processing parameters are defined. Flame retardant fillers are added and V0 rating is obtained by optimising the flame retardants with the presence of ammonium polyphosphate in 10% at least. The study opens the way for the application of additives like flame retardants to have applicable products. This can be summarised in the following points:  Possibility of producing PBT and PA6 composites with flax natural fibres and the processing parameters are defined where the kneading temperature is set at a temperature lower than the melting point taking into consideration the rise in temperature due to the internal friction.  Mechanical tensile strength is improved to 20% at 40–50 wt.% of fibre whereas the E-modulus is almost tripled.  Flame retardant fillers are added and HB rating is obtained by optimising the flame retardants with the presence of Exolite:Actilox in 10:10.  Flame retardant fillers are added and V0 rating is obtained by optimising the flame retardants with the presence of Exolite:Actilox in 15:5.  Mechanical properties are also optimised at the flame retardant ratio of 10:10.

Acknowledgments Thanks are due to the personnel in the institute of Polymer Materials and Plastics Engineering. Based on these results, a project proposal is approved and financed by ZIM-Kooperationsprojecte (AIF Projekt GmbH) under the code number of 205609SL2 References [1] Kozlowski R. Handbook of natural fibres: processing and applications, vol. 2. Woodhead Publishing; 2012. [2] Reich S, El Sabbagh A, Steuernagel L. Macromol Symp 2008;262(1):170–81. [3] Le Moigne N, Van den Oever M, Budtova T. Polym Eng Sci 2013. http:// dx.doi.org/10.1002/pen.23521. [4] Ramzy A, El-Sabbagh A, Steuernagel L, Ziegmann G, Meiners D. J Appl Polym Sci 2013. doi: 10.1002/app.39861. [5] Joseph P, Joseph K, Thomasb S. Compos Sci Technol 1999;59:1625–40. [6] Yan Z, Zhang J, Lin G, Zhang H, Ding Y, Wang H. J Reinf Plast Compos 2013. http://dx.doi.org/10.1177/0731684413501925. [7] EI-Sabbagh A, Steuernagel L, Ziegmann G. Polym Compos 2009;30(4):510–9. [8] Taha I, El-Sabbagh A, Ziegmann G. Polym Polym Compos 2008;16(5):295–302. [9] Trade and market division of Food and Agricurtural Organisation of the united nation FAO. 2012, report,http://www.fao.org/fileadmin/user_upload/ futurefibres/docs/Publications/UnlockingFibrePotentialFinalDocument.docx. [10] Studie zur Markt- und Konkurrenz -situation bei Naturfasern und Naturfaser Werkstoffen (Deutschland und EU), nova-Institut. Gülzower Fachgespräche, Band 26, FNR Finanzierung durch Verbraucherschutz; 2008. (FKZ 22020005) [11] Shimazu F, Sterling C. J Food Sci 1966;31(4):548–51. [12] Tajvidi M, Takemura A. J Thermoplast Compos Mater 2010;23:281–98. [13] Sears K, Jacobson R, Caulfield D. Underwood J. In: The sixth international conference on woodfiber – plastic composites, Montreal, Canada; 2001. P. 27– 4. [14] Jacobson R, Caulfield D, Sears K, Underwood J. In: The sixth international conference on woodfiber – plastic composites, Montreal, Canada; 2001. p. 127–33. [15] Sears K, Jacobson R, Caulfield D, Underwood J. Patent number Patent 6730249; 2004. [16] Arcaya P, Retegi A, Arbelaiz A, Kenny J, Mondragon I. Polym Compos 2009:257–64. [17] Wu C, Yen F, Wang C. Polym Bull 2011;67:1605–19. [18] Yao F, Wu Q, Lei Y, Guo W, Xu Y. Polym Degrad Stabil 2008;93:90–8. [19] Gassan J, Bledzki A. J Appl Polym Sci 2001;82:1417–22. [20] Ruyck H. Modelling of the residence time distribution in a twin screw extruder. J Food Engin 1997;32(4):375–90. [21] Casu A, Camino G, De Giorgi M, Flath D, Laudi A, Morone V. Fire Mater 1998;22:7–14. [22] Zhao B, Chen L, Wang Y. Fire and polymers VI: new advances in flame retardant chemistry and science. ACS Publications; 2012 [chapter 12]. [23] Zhang S, Ban Q, Huang H, Li Y. In: The 5th ISFR; 2009. p. 178–83 [October 11– 14, Chengdu, China] [24] Le Bras M, Duquesnea S, Foisb M, Griselb M, Poutch F. Polym Degrad Stabil 2005;88:80–4. [25] Schartel B, Brauna U, Schwarzb U, Reinemann S. Polymer 2003;44:6241–50. [26] El-Sabbagh A, Steuernagel L, Ziegmann G. J Reinf Plast Compos 2013;32(14):1030–43.