Improving the mechanical properties of Glass Reinforced Plastics by slight mechanical compression

Glass reinforced plastics (GRPs) are composite materials that have been used widely in engineering. Mechanical properties of fabricated GRP products depend on the type and arrangement of reinforcement, the type of the plastic as well as the reinforcement to plastic ratio. Some of these factors are interdepended in determining the quality of the final product. In the present study, the influence of applying mechanically a slight pressure during fabrication on the properties of GRP specimens, was investigated. Specimens were fabricated using the hand lay-up method and were further processed either by a vacuum bag compression or a mechanically applied pressure. The properties of the produced composite specimens were then determined experimentally. The application of mechanical compression was found to improve their properties. The effect of the mechanically applied compression or vacuum was evident as a reduction of the specimens’ internal void volume compared to the non-compressed ones, resulting in an increase in the density by 9–12% and a reduction in the thickness of the specimens by 15–20%. Additionally, the tensile and flexural strength of the specimens were increased by more than 15% and 5% respectively when pressure was applied mechanically, reflecting an improvement in GRPs’ mechanical properties.


Introduction
Glass fiber reinforced plastics (GFRPs) are composite materials used in diverse industrial applications [1].They are widely used in a variety of engineering constructions and manufacturing items produced in several industrial sectors such as the automotive, electronic, domestic goods industries, as well as the water sport industry and shipbuilding [2].They are also used for joining, repairing, maintenance and protection purposes [3].GFRPs are lightweight materials with high specific strength, stiffness and wear resistance.They also have the capability of absorbing vibrations and are not affected by corrosion.Although Carbon fiber reinforced plastics (CFRPs) have emerged as a potential successor of GFRPs [4], since they exhibit superior mechanical properties, their use in applications of moderate or low mechanical requirements (low end applications) is limited due to their high cost and environmental impact [1].
The manufacturing methods of glass and carbon fiber reinforced plastics, as well as their characteristic properties have been recently reviewed [5,6].The applied processes range from simple ones, such as the hand lay-up [7], to complicated ones, such as the advanced fiber placement [8] used for the production of high mechanical performance items.Although there are several different methods of manufacturing GFRP materials, the manually hand lay-up process is still extensively used in the marine industry.GFRP composites are the most popular materials for the construction of lightweight vessels [9].The properties and the quality of the final product are affected by the applied manufacturing process.The simplest method of manufacturing GRP composites involves the hand layup process which is a low cost, easily applied method for build up, maintenance and repair purposes.However, the quality of the outcome often depends on the experience and skills of the technicians involved [7].Although the core process of the method has remained practically unchanged over time, improvements can be introduced, aiming to reduce variation in the quality of the manufactured product caused by human intervention.The use of other contact molding methods for GFRP manufacturing, such as spray lay-up, injection and pultrusion, is limited by a higher cost and the necessity of specialized equipment [8].
Mechanical properties of the fabricated GFRP products depend on the type and arrangement of reinforcement, the type of the plastic, as well as the reinforcement to plastic ratio.Some of these factors are interdepended in determining the quality of the final product.A variety of other factors are also known to affect the properties of FRP composites.These include the fiber and polymer matrix type, the fiber arrangement, the volume fraction of each constituent phase, the porosity or the void volume content, the curing conditions and any post-curing treatment [5,6,8].Examination of the mechanical properties of fiber reinforced composites can be used to validate the applied manufacturing process.Several methods are used to optimize manufacturing procedures in order to minimize process-induced defects.These combine the use of numerical calculations and the experimental results from non-destructive testing [10] or the application of purely experimental techniques [11] to evaluate the impact of porosity on the mechanical properties of GFRPs.The presence of a large number of vacancies, as well as poor matrix adhesion to the reinforcing fibers, has been known as a critical factor that favors failure through matrix cracking [9].Modification of the fiber surface properties by using certain chemical compounds as coupling additives, has been found to improve the mechanical properties of glass fiber reinforced polyester composites [12].The tensile and flexural strength as well as the impact energy of GFR polyester composites fabricated by the hand layup method have been found to increase, when the fiber content of the composite increases from 15 to 60 %wt [13].Besides, GFRPs of higher density and inferior mechanical properties than the ones produced by hand lay-up, have been manufactured by the vacuum infusion method [14].The fiber to matrix volume ratio is a significant factor in determining the mechanical properties of GFR isopthalic polyester laminates that have been manufactured by hand lay-up and have undergone post curing treatment at 80 o C for 4 hours [15].An increase in the above-mentioned ratio has been reported to increase the tensile strength and decrease the flexural strength of the laminates.Reinforcement in the form of a fabric is known to be more efficient than the one of a mat consisting of dispersed chopped fibers in a random orientation.The latter exhibits quasi-isotropic in plane properties while the former exhibits inferior properties along specific directions [16].An increase of the tensile and bending strength of GFR epoxy composites fabricated by compression molding at 80 o C was observed when 0.8 %wt multi-wall carbon nanotubes were added and a pressure larger than 10 MPa was applied [17].
The purpose of the present work is to investigate specific modifications and adjustments of simple and low-cost fabrication methods, such as the simple wet hand lay-up, used for GFRP laminates in order to improve the quality and the properties of the products.In this context, a number of laminated specimens comprised of polyester resin reinforced with continuous glass fibers were prepared by the wet hand lay-up method and tested experimentally.A modification of the process was introduced by mechanically applying a slight compression at room temperature before curing of the resin.The effect of this modification on the properties of the fabricated specimens was assessed and evaluated.The outline of the present work is the following: the introduction is followed by an experimental section in which the materials, fabrication methods and the procedures used for the determination of the properties, are described.Subsequently, the results of the experimental tests are presented and discussed.Finally, the resulting conclusions are presented.

Materials
For the fabrication of the composite specimens pre-accelerated orthophthalic polyester resin (EXTRAPOL GP 2000, d=1.220 g/cm 3 ) was used as the matrix phase together with the appropriate amount of methyl-ethyl-keton peroxide (MEKP), 1.2-1.7 %wt., as a polymerization catalyst.The reinforcement consisted of commercially available E-Glass in the form of chopped strand mat (CSM-450, d=2.470 g/cm 3 ) or woven roving (WR-600, d=2.500 g/cm 3 ).

Fabrication process
Specimens with dimensions of 5 x 45 cm were fabricated by applying several layers (1, 2, 4 or 8 layers were applied) of reinforcement consisting of CSM-450 or WR-600 and adding the appropriate amount of pre-accelerated polyester resin containing the polymerization catalyst.The resin was not degassed prior to application.The techniques used for the fabrication were the following: (a) Hand lay-up [1,6,7,13], (b) Hand lay-up followed by a mechanical cold pressing of 0.6-0.7 MPa and (c) Hand layup followed by vacuum bag pressing inducing a pressure of 1atm (0.1 MPa) [1,6].Fabrication took place at a temperature of 25±2 o C and at relative humidity 60±5 % in all cases.

Determination of properties
Fiber content of the fabricated specimens was determined by the combustion method implementing the burn-off test according to ASTM D2584 standard.The density of the specimens was determined at 26 °C by the buoyancy method according to ASTM D3800 standard.The void volume of the specimens was determined according to ASTM D2734 standard using the data of the burn-off test and density of the specimens.Tensile and flexural strength were determined by a universal testing machine, according to ASTM D3039 and D790 standards respectively.Water absorption capability of the specimens was determined using an analytical balance and following the procedure described in ASTM D570 standard.

Results and Discussion
The v/v composition of the fabricated specimens was determined experimentally implementing the combustion method to obtain the reinforcement and matrix weight.The buoyancy method was implemented for determination of the density.The % v/v content of the specimens are cited in Table 1.Typical values for the density are given for the 8-layer specimens in Table 2.The average density values of the specimens consisting of 1, 2, 4 and 8 layers are cited in Table 3.It is noted that the application of a slight mechanical compression results in an increase of the %Vfbr content and a decrease of the void volume of the specimens.This is associated with an increase in the density of the specimens as can be seen from the data given in Tables 2 and 3.The application of vacuum pressing also results in decreased void volume and increased density, as expected.As can be seen in figure 1, the application of an external pressure, either mechanically or by vacuum, has a significant effect on the fabricated specimens causing a decrease in the void volume content.In addition, higher density values were found for the specimens that had been pressed mechanically, regardless of the type of reinforcement (see figure 2).An average increase in density of 12.0 % and 9.0 % was found for the CSM and the WR reinforced specimens respectively.The resin/fiber mass and volume ratios of the specimens fabricated by different methods are cited in Table 4.In the case of mechanically applied compression, specimens have a higher resin/fiber mass ratio (values of 1.43 for CSM and 2.90 for WR) than the ones fabricated by the simple hand lay-up or vacuum bag methods.On the other hand, the corresponding resin/fiber volume ratios decrease (values 1.53 for CSM and 0.73 for WR).
The thickness of the specimens fabricated by different methods are shown in figure 3 for the CSM reinforced and in figure 4 for the WR reinforced ones.Applying a slight mechanical compression was found to reduce the thickness of the fabricated specimens comprised of 2, 4 or 8 layers regardless of the type of reinforcement.This is related to the higher density, the increase in resin/fiber mass ratio and the decrease in resin/fiber volume ratio mentioned above.It indicates that mechanical compression is an effective way to fill in the internal vacancies and reduce the void volume of the specimens.Tensile and flexural strength of the specimens are cited in Table 5.As can be seen from the data in Table 5 and figure 5 a significant increase in the tensile strength was found when hand lay-up fabrication was followed by mechanical compression.Flexural strength was found to increase significantly in the case of specimens with CSM reinforcement, but only slightly increased (about 5%) in the case of WR reinforcement (see figure 6).Typical curves of water absorption in a 30-days period are shown in figures 7 and 8 for the 4-layer CSM reinforced and 2-layer WR reinforced specimens respectively.In Table 6 the amount of water absorbed (range and average values) by the specimens consisting of 2, 4 and 8 layers are cited and in Table 7 the corresponding values per unit volume of the specimen are given.

Figure 1 .
Figure 1.Void volume content (average values are shown) of the GFRP laminated, CSM (red columns) and WR (blue columns), reinforced specimens fabricated by different methods.PE refers to poly-ester resin as the matrix polymer.

Figure 2 .
Figure 2. Density (average values are shown) of the GFRP laminated, CSM (red columns) and WR (blue columns), reinforced specimens fabricated by different methods.PE refers to poly-ester resin as the matrix polymer.

Figure 5 .
Figure 5. Tensile strength of the GFRP laminated, CSM (red columns) and WR (blue columns), reinforced specimens fabricated by different methods.

Figure 6 .
Figure 6.Flexural strength of the GFRP laminated, CSM (red columns) and WR (blue columns), reinforced specimens fabricated by different methods.

Table 1 .
Volume content of the fabricated specimens.Standard deviation values are given in parentheses.

Table 2 .
Density of the 8-layer fabricated specimens.Standard deviation values are given in parentheses.

Table 3 .
Average density values of the fabricated specimens consisting of 1, 2, 4 and 8 layers.Standard deviation values are given in parentheses.

Table 4 .
Matrix to fiber ratio (average values) of the fabricated specimens.Standard deviation values are given in parentheses.
Figure 3. Thickness of the CSM/polyester laminates fabricated by Hand lay-up (dark red), Hand lay-up followed by compression (red) and Hand lay-up followed by pressing under vacuum (orange).Figure 4. Thickness of the WR/polyester laminates fabricated by Hand lay-up (purple), Hand lay-up followed by compression (blue) and Hand lay-up followed by pressing under vacuum (light blue).

Table 5 .
Tensile and flexural strength of the fabricated specimens (4-layer specimens were used for the determination of σUTS and 8-layer specimens for the determination of σflex).