**6. Mechanical performance of re-used composite**

The principle of structural re-use of EoL thermoset composite products is based on the possibility to partially benefit from the strength that is still present in the material. The so-called L/D-ratio of the parts that are re-used (strips or flakes) as reinforcing elements is an important parameter. In this ratio L represents the longest dimension of the re-used composite part and D represents the smallest dimension. Although the role of the L/D-ratio on elastic properties of fibre reinforced materials is described already long ago with elastic models [13], the influence of the L/D-ratio of reinforcing elements on strength of the complete reinforced material or product is more difficult to model.

The influence of the L/D-ratio of reinforcing elements on the strength of the new product that is made with it has been further investigated with a well-defined shape: a rectangular strip. To investigate the effect of the length of strips on reinforcement, a series of panels was produced and tested. To allow for a good comparison, instead of EoL-material, virgin glass mat reinforced polyester laminate with a constant quality was used as a base material to produce the strips.

The base material for making the strips for the investigation was a glass mat reinforced polyester laminate with a thickness of 5 mm. For the glass mat reinforcement 6 layers of continuous glass fibre mat (CFM) of 450 g/m2 were used (Unifilo U813, OCV). As resin a low-viscous DCPD polyester resin (Synolite 1967 – G 6, AOC) was used. This laminate was made by vacuum infusion resulting in a glass content of 37 wt%. The laminate was made with peel-ply on both sides that was removed after curing for improved adhesion lateron.

**Figure 6.** *Layers of GRP strips as core in a test panel in preparation for infusion [4].*

From the laminate, strips were cut with a diamond-tipped blade to a width of 20 mm, that were used as reinforcing elements in a test panel with the dimensions L x W x T = 1000 x 300 x 24 mm. Incorporating four layers of these strips positioned flatwise resulted in a reinforcing core of 20 mm. A surrounding shell of about 2 mm thickness was created by wrapping the core in an infusion glass mat (Polymat HI-FLOW M03P Core, Scott & Fyfe, consisting of 2x 450 g/m2 glass mat reinforcement layers). The final panels were produced by means of infusion with the same DCPD polyester resin as for making the 5 mm laminates. The reinforcing strips were used in the core with lengths of 40 mm, 80 mm, 200 mm and 1000 mm, respectively. The photo in **Figure 6** shows the flat-wise incorporation of the fourth layer of 40 mm strips just before it was covered with the infusion glass mat.

To investigate the effect of adhesion between strips and embedment resin, a test panel was produced with 40 mm strips with flat surfaces (by omitting peel-ply layers, resulting in smooth surfaces with relatively bad adhesion). Moreover, a test panel was produced using 1000 mm strips, that was cut from a polyester boat hull without any surface treatment. Finally, a test panel was produced with 200 mm strips placed vertically on their sides instead of flatwise. In all test panels strips were oriented in length direction of the test panel and the strips were placed staggered with respect to the neighbouring strips (both horizontally and vertically). **Table 1** gives the overview of tested configurations.


#### **Table 1.**

*Combinations of materials and strip lengths tested.*

**Figure 7.** *Bending test of a specimen made with 80 mm strips [4].*

**Figure 8.** *Mechanical strength as function of strip length for different test materials [4] (markers indicate the mean value of 5 tests).*

From each test panel samples were cut in length direction (i.e. in de direction in which the strips are oriented) with the dimensions L x W x T = 360 x 50 x 24 mm. Tests were repeated 5 times. The samples were tested in three-point bending in accordance with ISO 178. The photo in **Figure 7** shows the testing of a sample with four layers of strips of 80 mm in length, placed flat-wise.

The graph in **Figure 8** shows a clear correlation between the bending strength and the length of the strips in the core for flat-wise laid strips with good adhesion. With a 1000 mm strip length the bending strength reaches 204 MPa. The negative effect of bad adhesion on bending strength is seen in the case of the 40 mm strips where the peel-ply has been omitted: the bending strength is as low as 36 MPa, less than half of the strength with 40 mm strips with good adhesion (88 MPa). Also a relatively low strength of 99 MPa is found when using strips cut from a polyester boat-hull, which can be attributed also to bad adhesion. During the test these strips that were obtained from abolished boats, delaminated at the gelcoat side, showing

#### *Industrial Re-Use of Composites DOI: http://dx.doi.org/10.5772/intechopen.99452*

a smooth delamination surface, which is an indication of a relatively bad adhesion. By placing the 200 mm strips vertically, a higher bending strength is observed (163 MPa) as compared to the flat-wise placement (141 MPa).

With the set-up of strips laid flat-wise and a good adhesion with the embedding resin, the effect of the L/D-ratio is considered. It is assumed that the panel strength with strip lengths of 1000 is the maximum attainable panel strength in this set-up (204 MPa). At a strip length of 40 mm (L/D-ratio of 8) only 43% of the maximum attainable strength of the panel is found. With increasing strip length the bending strength of the panel increases in a degressive manner. At a strip length of 200 mm (L/D-ratio of 40) a panel bending strength is found that is 69% of the maximum attainable panel strength. From these considerations it can be concluded that a significant part of the possible maximum panel strength is obtained for a L/D-ratio which is of the order of 50 or higher.

In the graph, also at the location of 80 mm length, the strength of a panel made from flakes is shown. That strength was found to be 81 MPa. The flakes were made by shredding EoL thermoset composite in such a way that a mean flake length of roughly 80 mm was obtained. It must be remarked that the lower strength as compared to the strength of 80 mm strips (116 MPa) can partly be attributed to the lower content of reinforcement (higher resin content).

Shredding may be a more economical way of machining EoL thermoset composites into reinforcing elements than sawing or water cutting, although this has to be investigated further. Shredding is a very promising method because a large quantity of EoL composite products can be machined at relatively low cost. However, the flakes that result from the shredding process must have a quality level that makes them suitable for the re-use in new products, e.g. sufficiently high L/D-ratio, dry, dust free and good adhesion properties with the embedment resin.
