**4. Discussion**

The blunt striker did not cause any visible damage in the reinforcing layers, however, the area of delamination was largest between analyzed cases. As mentioned, the delamination process absorbs part of the impact energy. The main part of the energy di fference between the initial kinetic energy of blunt striker and the kinetic energy after its rebound is connected with the energy absorbing delamination process. Including the phenomenon of delamination in numerical studies allows the scientist to estimating the extent of the structure damage. It should be remembered that not every

damage that was caused is immediately visible, as was the case of a panel impacted by the blunt striker. The reinforcing material was not damaged, however, the composite structure was damaged, which a ffects its strength. The same fact applies to cases in which other strikers were used. As shown by numerical research, the real areas of damage were much larger than that estimated on the diameter of the cavities formed after impact in experimental research. Experimental research presented in another paper [68] confirms that the delamination area grows with each next reinforcing layer from the impact point. The shape of the delamination areas was also similar to the shape that could be observed in experimental research [22]. It is worth to noting that delamination areas can be reduced by adding, for example, nanofibers. Research presented in another paper [62] showed that the incorporation of polycaprolactone nanofibers reduces the delaminated area by about 27% in glass fiber reinforced polymer composites. Research presented in another paper [62] also confirms that the epoxy glass composites reinforced with polycaprolactone nano fibers are less susceptible to impact damage than the same composites without polycaprolactone nano fibers.

The comparison of rebounded energy value in the case of conical and ogival strikers with damages caused by those strikers in composites suggests that the energy was absorbed in di fferent ways, depending on the strikers' geometry. Rebounded energy was the same, but damaged was di fferent. In the case of the conical striker, damage in reinforced layers was bigger. In the case of the ogival striker, the delamination areas was bigger, which suggests that the delamination absorbed a higher part of the impact energy than in the case of the conical striker, which causes the smallest damage in the reinforcement layers.

In the case of the composite made from glass fibers and polyester resin [69], it was noted that the largest areas of damage were obtained when the conical and hemispherical strikers were used. In the conducted research, the largest diameter of damage in the first reinforcing layer was also noted for these two strikers. In the case of the research presented in a further paper [69], the composite material absorbed di fferent amounts of the impact energy depending on the striker geometry. The amount of absorbed energy in descending order was as follows: conical striker, ogival striker, hemispherical striker, and blunt striker. The same results were obtained for carbon epoxy composite (impact energy up to 6 J) in another paper [70]. In the conducted numerical research, the same order was observed (Figure 9). Carbon fiber reinforced composite in the case of the blunt striker impact (50 J of the impact energy) absorb almost the same amount of impact energy as in the case of the conical striker [45]. In the conducted research, the hybrid composite absorbs di fferent amounts of impact energy in the case of those strikers. In the case of the conical striker, the composite absorbs a few times more impact energy than in the case of the blunt striker. In the case of the analyzed composites, the plastic behavior of aramid fibers causing the damage caused by the strikers with di fferent geometries is much more diverse than in the case of, for example, the composite made from carbon fibers, the damage of which is brittle [69,70].

The conducted experimental and numerical research indicates that the process of energy absorption in composite panels is complex. In the case of the experimental research, energy was absorbed by all factors listed in the introduction. Owing to the adopted modeling scale, the impact energy in the case of numerical research could not be absorbed by friction between the fibers (inside the reinforcing layer). The energy was absorbed by the friction between the striker and the reinforcing layers and by the friction between the adjacent reinforcing layers. The numerical results present that the delamination absorbed some of the impact energy, which is best seen in the case of the blunt striker. The presented Tresca equivalent stress (Figure 8) indicates that the primary yarns were subjected to the highest stress; however, in the adopted modeling scale, it is di fficult to clearly divide the energy absorbed by the reinforcing layer into the energy absorbed by the destruction of primary yarns, deformation of secondary yarns, and shear plugging. The di fference between energy absorption in the case of experimental research and absorption in the numerical research (which results from the adopted modeling scale) obviously has an impact on the result. The lack of consideration of some energy absorbing phenomena increases the damage in the composite panel.

It was stated that carrying out numerical research simultaneously with experimental research can significantly a ffect the quality of the obtained conclusions. Appropriate model preparation and consideration of physical phenomena occurring in the analyzed material can provide a lot of important information. The calibration of the numerical model and its verification based on the performed experimental research is also important. If the model is considered as a satisfactory, it can be used to observe phenomena that are hard or even impossible to observe during the experiment. In the conducted research, such a phenomenon was delamination areas between reinforcing layers. In the case of multilayered composites, it is very important to use the right modeling scale, because observing some phenomena on the model prepared in the wrong scale is not possible. During the interpretation of the numerical research results, the simplifications used should be taken into account, particularly simplifications resulting from the used modeling scale. A compromise between which phenomena can be observed and the time of calculations should be found. Preparing an increasingly detailed model, one should take into account that the calculation time may increase several times.
