*2.1. Influence of Interface Modification on In-Plane Properties*

To determine the influence of the interlaminar interface modification on the in-plane properties of the composite, tensile tests according to DIN EN ISO 527-4 are performed. The rectangular specimens were 200 mm long, 25 mm wide, and 2 mm thick and were tested with a loading velocity of 0.5 mm/s. Figure 2a shows the fracture patterns of exemplary tensile specimens with the respective contact surface proportions *κ*. The fracture surfaces of specimens with high contact area fractions show smooth fracture surfaces, while specimens with low contact area fractions show more jagged fracture patterns. This behaviour is a consequence of different load redistribution processes during damage propagation. When a single layer of a specimen with high interlaminar properties fails, the load is redistributed to the other layers. These are subsequently overloaded in the immediate vicinity of the first point of damage and subsequently fail there. In contrast to test specimens with a high contact area, test specimens with a smaller contact area show a different failure mechanism. If a single layer in the laminate with low interlaminar properties fails, the crack does not run into the adjacent layers, but separates and decouples the individual layers from each other. This causes the adjacent individual layers to fail at their respective weak points, which do not necessarily have to be close to the position of the original initial failure.

**Figure 2.** Failure behaviour of interface-modified textile-reinforced CFRP; (**a**) tensile specimens after tests and (**b**) corresponding mechanical in-plane properties in weft direction.

The Young's modulus remains almost constant with the increasing contact area *κ*. The strength, however, increases with increasing contact area *κ*, which is due to an influence on the local stress state as a result of the perforation as an interference point. Additionally, the load transfer between the individual layers is limited by the low interlaminar property. The investigations carried out here prove that the in-plane properties under tensile load are not significantly influenced by the selected type of interface modification.

The in-plane failure behaviour is described by an adaptation of Hashin's failure theory [15]. Here, no interaction between tensile *σ*<sup>11</sup> and shear stresses *σ*<sup>12</sup> and *σ*<sup>13</sup> in tensile fibre mode is assumed: *<sup>σ</sup>*<sup>11</sup>

$$\frac{\sigma\_{11}}{\chi t} = 1 \text{, for } \sigma\_{11} > 0. \tag{2}$$

The fibre compressive mode is modelled analogously:

$$\frac{|\sigma\_{11}|}{\mathbf{x}\mathbf{c}} = 1, \text{ for } \sigma\_{11} < 0,\tag{3}$$

where *xt* and *xc* are the tensile and compressive strengths in the fibre direction. Failure in the 2-direction is also modelled as fibre failure for woven fabrics due to symmetry. Since out-of-plane failure is modelled with the cohesive element approach, the out-of-plane failure mode is supressed by setting strengths to very high values for the single-ply material model. The LS-DYNA material model 58 (\*MAT\_058) [16] is used for modelling. MAT\_058 also enables the modelling of damage evolution using the Matzenmiller model [17]. The material's stiffness is gradually degraded until a residual level of stress (*slimt1, slimc1*) is reached. This is kept constant until the final failure strain (*fail1, fail2*) is reached. Since the material can still transfer loads in compression failure, higher values are assumed for *slimc1* and *fail2* than for the corresponding values for tensile failure *slimt1* and *fail1*. Figure 3 shows the modelled stress–strain curve in the fibre direction. Table 1 shows the key material parameters. The full LS-DYNA material cards are shown in Tables A2 and A3 in Appendix A.

**Figure 3.** Investigated interface modification designs.



### *2.2. Influence of Interface Modification on Out-of-Plane Properties*

The influence of the interlaminar contact area proportion on the delamination behaviour is investigated using five pre-cracked double cantilever beam (DCB) test specimens. The specimens have an interlaminar contact area *κ* of approx. 0.15, 0.30, 0.45, and 1.0, respectively, and are tested according to ISO15024. The results are presented in Figure 4. The relationship between through-thickness strength, the strain energy release rate, and *κ* is reasonably well-represented by a linear fit.

**Figure 4.** (**a**) Through-thickness tensile test results and (**b**) DCB test results for different interlaminar contact areas.

Cohesive zone approaches are used for modelling delamination between adjacent single plies. They allow an evaluation of the delamination initiation and a description of the delamination growth. LS-DYNA offers cohesive zone models as part of contact formulations or through special cohesive elements.

The accuracy of cohesive elements is higher in comparison to contact formulations [18]. Therefore, cohesive elements with a bilinear material model are used here for modelling the delamination behaviour of interface-modified multilayer composites.

Table 2 shows the key parameters for the cohesive model without interface modification (*κ* = 1.0). The full LS-DYNA material card is shown in Table A4 in Appendix A. The parameters of the strengths (*t* and *s*) and the critical energy release rates (*gic* and *giic*) are adjusted to account for the interface modification in the material model. Since PTFE does not adhere to the composite material, the interlaminar properties of the strengths and strain energy release rates are also zero at *κ* = 0 (no perforation, intact release foil). A linear relationship between the interlaminar contact area and strain energy release rate in Mode I applies to both the strengths and the characteristic values in the Mode II load case (Figure 4).
