10 Composites forming technologies 140 k=0.5 120 Yarn 100 80 k=2 60 40 Other direction 20 free 0 0 0.2 0.4 0.6 0.8 Strain (% 1.5 Results of biaxial tensile tests for a balanced glass plain weave fabric.17 The load is measured along one tow direction,with the constant k determining the wO3' ratio between strains in the loading and transverse directions. 是 uniaxially (denoted 'other direction free'),the non-linear force region extends to a strain of approximately 0.5%,but the force to completely straighten the tows is low.As the ratio between strains in the tested (warp)and transverse (weft) 豆 directions increases,the force curve tends towards the behaviour of an individual tow (yarn).The testing procedure and relevance of this behaviour to composite forming simulation are described in detail in Chapter 3. 豆 1.4 Ply/tool and ply/ply friction During an automated forming operation,friction between the material and forming tools governs the transfer of loads to the material.In multi-layer forming processes,friction between individual layers of material is also of importance.For example,when forming layers of prepreg at different orienta- tions to each other,compressive forces generated by intra-ply shear in one layer may be transferred into adjacent layers,causing compression along the fibre direction and hence wrinkling of the form.Hence to model forming processes accurately,measurement of friction at ply/tool and ply/ply interfaces is important. A number of test methods are possible for measurement of friction.The simplest approach is the so-called 'inclined plane method'.Here a block of tooling material is placed on a piece of fabric/prepreg mounted on a rigid plate. The plate is then inclined until the block starts to move,with the tangent of the
uniaxially (denoted `other direction free'), the non-linear force region extends to a strain of approximately 0.5%, but the force to completely straighten the tows is low. As the ratio between strains in the tested (warp) and transverse (weft) directions increases, the force curve tends towards the behaviour of an individual tow (yarn). The testing procedure and relevance of this behaviour to composite forming simulation are described in detail in Chapter 3. 1.4 Ply/tool and ply/ply friction During an automated forming operation, friction between the material and forming tools governs the transfer of loads to the material. In multi-layer forming processes, friction between individual layers of material is also of importance. For example, when forming layers of prepreg at different orientations to each other, compressive forces generated by intra-ply shear in one layer may be transferred into adjacent layers, causing compression along the fibre direction and hence wrinkling of the form. Hence to model forming processes accurately, measurement of friction at ply/tool and ply/ply interfaces is important. A number of test methods are possible for measurement of friction. The simplest approach is the so-called `inclined plane method'. Here a block of tooling material is placed on a piece of fabric/prepreg mounted on a rigid plate. The plate is then inclined until the block starts to move, with the tangent of the 1.5 Results of biaxial tensile tests for a balanced glass plain weave fabric. 17 The load is measured along one tow direction, with the constant k determining the ratio between strains in the loading and transverse directions. 10 Composites forming technologies Copyrighted Material downloaded from Woodhead Publishing Online Delivered by http://woodhead.metapress.com Hong Kong Polytechnic University (714-57-975) Hong Kong Polytechnic University (714-57-975) Saturday, January 22, 2011 6:57:44 PM IP Address: 158.132.122.4
Composite forming mechanisms and materials characterisation 11 angle of inclination defining the friction coefficient.More sophisticated variations on this approach exist,for example pulling the block along the surface of fabric/prepreg and measuring the force required to maintain a constant velocity.However,these techniques are best suited to materials that exhibit a constant coefficient of friction.Unfortunately this does not tend to be the case for fabrics or prepregs. Several alternatives to the above have been developed to allow variables such as normal pressure,testing rate and temperature to be varied.Ply pull out tests have been used to measure ply/ply friction.Data from such tests are discussed in Chapter 10.Murtagh developed a device whereby a layer of thermoplastic prepreg or tooling material was sandwiched between two prepreg layers,held together with a controlled normal pressure.The whole apparatus was heated to evaluate the effect of temperature on behaviour,and the tooling material was withdrawn at various rates with the required force measured using a load cell. Wilks20 designed apparatus based on a similar principle to Murtagh,although here a layer of fabric/prepreg was sandwiched between two layers of tooling material (Fig.1.6).For friction between dry fabric and tooling materials,some dependence on normal pressure is observed,with friction reducing marginally with increasing pressure as the fabric surface was flattened against the tool. Friction coefficients of between 0.2 and 0.4 have been measured between glass fabrics and tooling materials. Typical results are shown in Fig.1.7 for friction between a glass/ polypropylene thermoplastic composite and a steel tool.Shear stress increases Specimen 01 Connected to load cell Platens 店 Springs Springs 0 000 150 Start of test End of test 1.6 Schematic of ply/tool friction measurement apparatus
angle of inclination defining the friction coefficient. More sophisticated variations on this approach exist, for example pulling the block along the surface of fabric/prepreg and measuring the force required to maintain a constant velocity. However, these techniques are best suited to materials that exhibit a constant coefficient of friction. Unfortunately this does not tend to be the case for fabrics or prepregs. Several alternatives to the above have been developed to allow variables such as normal pressure, testing rate and temperature to be varied. Ply pull out tests have been used to measure ply/ply friction. 18 Data from such tests are discussed in Chapter 10. Murtagh19 developed a device whereby a layer of thermoplastic prepreg or tooling material was sandwiched between two prepreg layers, held together with a controlled normal pressure. The whole apparatus was heated to evaluate the effect of temperature on behaviour, and the tooling material was withdrawn at various rates with the required force measured using a load cell. Wilks20 designed apparatus based on a similar principle to Murtagh, although here a layer of fabric/prepreg was sandwiched between two layers of tooling material (Fig. 1.6). For friction between dry fabric and tooling materials, some dependence on normal pressure is observed, with friction reducing marginally with increasing pressure as the fabric surface was flattened against the tool. Friction coefficients of between 0.2 and 0.4 have been measured between glass fabrics and tooling materials. Typical results are shown in Fig. 1.7 for friction between a glass/ polypropylene thermoplastic composite and a steel tool. Shear stress increases 1.6 Schematic of ply/tool friction measurement apparatus. Composite forming mechanisms and materials characterisation 11 Copyrighted Material downloaded from Woodhead Publishing Online Delivered by http://woodhead.metapress.com Hong Kong Polytechnic University (714-57-975) Hong Kong Polytechnic University (714-57-975) Saturday, January 22, 2011 6:57:44 PM IP Address: 158.132.122.4
12 Composites forming technologies 0.040 ±0.5mm/s180Ce-0.5mm/s200C 0.035 8-0.5mm/s220C --0.8mm/s180C -◆-0.8mm/s200C--0.8mm/s220C 0.030 ±-1.2mm/s180Cc ◆-1.2mm/s200C 量-1.2mm/s220C 0.025 今 0.020 0.015 0.010 0.005 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Normal pressure(MPa) 1.7 Shear stress at the interface between steel tools and glass/polypropylene under applied normal pressure.20 Results are for velocities of 0.5,0.8 and 1.2mm/s at 180C,200C and 220C. with normal pressure,although the relationship is not linear.At a given normal pressure,the shear stress increases with increasing rate and with decreasing temperature. Along with other published studies,the results in Fig.1.7 suggest that for prepreg,the friction coefficient (ply/tool and ply/ply)should be defined as a function of rate,temperature and pressure.Wilks20 suggested a phenomeno- logical model for shear stress at the interface of the form: T=m+uP 1.5 0 where the first term represents shearing of a polymer film at the prepreg surface (with viscosity n and shear strain rate and the second term representing Coulomb friction caused by fibre reinforcement penetrating the polymer film (with friction coefficient u and normal pressure P).In practice the values of u and the polymer film thickness used to define must be determined empirically. 1.5 Ply bending The ability of fabric or prepreg to bend out of plane is of course critical for forming of curved components.It is perhaps surprising then that this topic has received relatively little attention,at least for fabric reinforcements and composites.One reason may be that the bending resistance is usually orders of magnitude lower than intra-ply shear resistance,which in turn is significantly lower than tensile stiffness in the fibre direction(s).However,in terms of modelling,this in fact presents a problem as traditional shell element formula- tions will have a bending stiffness related to the in-plane stiffness of the
with normal pressure, although the relationship is not linear. At a given normal pressure, the shear stress increases with increasing rate and with decreasing temperature. Along with other published studies, the results in Fig. 1.7 suggest that for prepreg, the friction coefficient (ply/tool and ply/ply) should be defined as a function of rate, temperature and pressure. Wilks20 suggested a phenomenological model for shear stress at the interface of the form: _ P 1:5 where the first term represents shearing of a polymer film at the prepreg surface (with viscosity and shear strain rate _) and the second term representing Coulomb friction caused by fibre reinforcement penetrating the polymer film (with friction coefficient and normal pressure P). In practice the values of and the polymer film thickness used to define _ must be determined empirically. 1.5 Ply bending The ability of fabric or prepreg to bend out of plane is of course critical for forming of curved components. It is perhaps surprising then that this topic has received relatively little attention, at least for fabric reinforcements and composites. One reason may be that the bending resistance is usually orders of magnitude lower than intra-ply shear resistance, which in turn is significantly lower than tensile stiffness in the fibre direction(s). However, in terms of modelling, this in fact presents a problem as traditional shell element formulations will have a bending stiffness related to the in-plane stiffness of the 1.7 Shear stress at the interface between steel tools and glass/polypropylene under applied normal pressure. 20 Results are for velocities of 0.5, 0.8 and 1.2 mm/s at 180ëC, 200ëC and 220ëC. 12 Composites forming technologies Copyrighted Material downloaded from Woodhead Publishing Online Delivered by http://woodhead.metapress.com Hong Kong Polytechnic University (714-57-975) Hong Kong Polytechnic University (714-57-975) Saturday, January 22, 2011 6:57:44 PM IP Address: 158.132.122.4
Composite forming mechanisms and materials characterisation 13 material.Hence it is necessary to understand the relative magnitude of bending stiffness so that the forces associated with out of plane bending can be scaled appropriately. Bending stiffness has long been measured for apparel fabrics.121 A standard- ised test can be performed for bending resistance of fabric under its own weight.22 This involves sliding a strip of fabric off the edge of a platform until the self-weight causes it to bend to a specified angle (41.5 is specified in the standard).The length of strip necessary to reach this threshold value is recorded, from which the bending rigidity is approximated.Results for woven apparel fabrics indicate that the bending rigidity in the fibre directions is significantly higher than in the bias(45)direction.Young et al.3 have applied this tech- nique to calibrate a finite element model for composite bending.The procedure involved simulating the experiment and adjusting a bending scale factor so that the predicted bending behaviour matched experimental observations More informative techniques measure the mechanical resistance to bending using,for example,cantilever,three-point bending or axial buckling experi- ments,and such tests have been applied recently to prepreg.For example Martin (SL6-LS et al.24 measured the three-point bending behaviour of unidirectional glass/ polypropylene composites using a V-shaped punch.At elevated temperature this resulted in an increase in bending force with increasing displacement rate.All tests exhibited an initial increase up to a peak value,after which the force ssaudmaur'peaypoowy/:dg 19.759 plateaued or reduced gradually.One issue with this approach is that the material has to be supported during bending to stop it from deforming under its own weight.To avoid this problem a simple buckling test can be used.25 Wang et 14 Applied load 12 Specimen 10 Clamps 三 8 6 100mm/min 30mm/min 2 10mm/min 0 0 6 9 12 15 Displacement(mm) 1.8 Bending/buckling behaviour of unidirectional carbon/epoxy thermoset prepreg under axial compressive loading at different rates
material. Hence it is necessary to understand the relative magnitude of bending stiffness so that the forces associated with out of plane bending can be scaled appropriately. Bending stiffness has long been measured for apparel fabrics.1,21 A standardised test can be performed for bending resistance of fabric under its own weight.22 This involves sliding a strip of fabric off the edge of a platform until the self-weight causes it to bend to a specified angle (41.5ë is specified in the standard). The length of strip necessary to reach this threshold value is recorded, from which the bending rigidity is approximated. Results for woven apparel fabrics indicate that the bending rigidity in the fibre directions is significantly higher than in the bias (45ë) direction. Young et al.23 have applied this technique to calibrate a finite element model for composite bending. The procedure involved simulating the experiment and adjusting a bending scale factor so that the predicted bending behaviour matched experimental observations. More informative techniques measure the mechanical resistance to bending using, for example, cantilever, three-point bending or axial buckling experiments, and such tests have been applied recently to prepreg. For example Martin et al.24 measured the three-point bending behaviour of unidirectional glass/ polypropylene composites using a V-shaped punch. At elevated temperature this resulted in an increase in bending force with increasing displacement rate. All tests exhibited an initial increase up to a peak value, after which the force plateaued or reduced gradually. One issue with this approach is that the material has to be supported during bending to stop it from deforming under its own weight. To avoid this problem a simple buckling test can be used.25 Wang et 1.8 Bending/buckling behaviour of unidirectional carbon/epoxy thermoset prepreg under axial compressive loading at different rates. Composite forming mechanisms and materials characterisation 13 Copyrighted Material downloaded from Woodhead Publishing Online Delivered by http://woodhead.metapress.com Hong Kong Polytechnic University (714-57-975) Hong Kong Polytechnic University (714-57-975) Saturday, January 22, 2011 6:57:44 PM IP Address: 158.132.122.4
14 Composites forming technologies al have applied this technique to thermoset prepreg,including unidirectional and woven carbon/epoxy composites.Typical data are included in Fig.1.8 for unidirectional prepreg,illustrating the effect of rate on bending/buckling behaviour.The force response increases linearly with increasing rate,suggesting that the phenomenon is dominated by flow of the polymer between fibre layers. The curve shape illustrates a peak load corresponding to buckling,followed by a reduction in force as displacement is increased.Similar tests for woven fabrics show an initial peak followed by a small reduction and then a plateau in the force,illustrating that the fibre architecture has a clear effect on the bending behaviour.Both material response curves fall within the range observed for dry textiles,25 although for such materials no clear rate effect is observed. 1.6 Compaction/consolidation Sunysijqnd At the end of forming,the material must be compacted or consolidated to increase the fibre volume fraction and(for prepreg)eliminate voids.An under- standing of compressibility,typically in terms of compaction pressure versus thickness or fibre volume fraction,allows the required pressure to be determined for the target fibre content.For multi-layer,multi-material reinforcement preforms,the compressibility of each material type is likely to be different,so that each layer attains a different fibre volume fraction under the imposed compaction pressure (or at the desired laminate thickness).Compaction has received a great deal of attention,particularly for dry fibre mats and fabrics. Robitaille27 has published an extensive review of both experimental methods and modelling approaches for reinforcement compaction,whilst Garciazs has reviewed similar techniques for thermoplastic and thermoset prepreg.Testing procedures appear relatively simple,with material compacted between two parallel platens within a universal testing machine.The platens are moved together usually at constant rate to either a pre-determined load or thickness.At this point either the thickness or force can be held constant to measure relaxation or compaction creep.Care must be taken to ensure that the platens are as flat and as parallel as possible,and their relative displacement must be measured carefully-best practice here is to attach an LVDT(linear variable displacement transducer)between the platens. 1.6.1 Compaction behaviour of reinforcements A great deal of experimental data exists for fabric reinforcements,and it is beyond the scope of this chapter to provide a comprehensive review.Large data sets have published by Robitaille27 and Correia,2 and these form the basis for the present discussion.Typical data are included in Fig.1.9.The graph shows the evolution of average fibre volume fraction (V)as a function of compaction pressure (P)for multiple layer stacks of a range of materials.All results show
al.26 have applied this technique to thermoset prepreg, including unidirectional and woven carbon/epoxy composites. Typical data are included in Fig. 1.8 for unidirectional prepreg, illustrating the effect of rate on bending/buckling behaviour. The force response increases linearly with increasing rate, suggesting that the phenomenon is dominated by flow of the polymer between fibre layers. The curve shape illustrates a peak load corresponding to buckling, followed by a reduction in force as displacement is increased. Similar tests for woven fabrics show an initial peak followed by a small reduction and then a plateau in the force, illustrating that the fibre architecture has a clear effect on the bending behaviour. Both material response curves fall within the range observed for dry textiles,25 although for such materials no clear rate effect is observed. 1.6 Compaction/consolidation At the end of forming, the material must be compacted or consolidated to increase the fibre volume fraction and (for prepreg) eliminate voids. An understanding of compressibility, typically in terms of compaction pressure versus thickness or fibre volume fraction, allows the required pressure to be determined for the target fibre content. For multi-layer, multi-material reinforcement preforms, the compressibility of each material type is likely to be different, so that each layer attains a different fibre volume fraction under the imposed compaction pressure (or at the desired laminate thickness). Compaction has received a great deal of attention, particularly for dry fibre mats and fabrics. Robitaille27 has published an extensive review of both experimental methods and modelling approaches for reinforcement compaction, whilst Garcia28 has reviewed similar techniques for thermoplastic and thermoset prepreg. Testing procedures appear relatively simple, with material compacted between two parallel platens within a universal testing machine. The platens are moved together usually at constant rate to either a pre-determined load or thickness. At this point either the thickness or force can be held constant to measure relaxation or compaction creep. Care must be taken to ensure that the platens are as flat and as parallel as possible, and their relative displacement must be measured carefully ± best practice here is to attach an LVDT (linear variable displacement transducer) between the platens. 1.6.1 Compaction behaviour of reinforcements A great deal of experimental data exists for fabric reinforcements, and it is beyond the scope of this chapter to provide a comprehensive review. Large data sets have published by Robitaille27 and Correia,29 and these form the basis for the present discussion. Typical data are included in Fig. 1.9. The graph shows the evolution of average fibre volume fraction (Vf ) as a function of compaction pressure (P) for multiple layer stacks of a range of materials. All results show 14 Composites forming technologies Copyrighted Material downloaded from Woodhead Publishing Online Delivered by http://woodhead.metapress.com Hong Kong Polytechnic University (714-57-975) Hong Kong Polytechnic University (714-57-975) Saturday, January 22, 2011 6:57:44 PM IP Address: 158.132.122.4