**4. Results**

#### *4.1. Surface Damage*

Figure 5a,b show the surface damage in the specimens under the action of components C and D, respectively.

**Figure 5.** Surface damage in the specimens after lightning strike: (**a**) component C; (**b**) component D.

For component C, serious ablation damage occurred on the top of the fastener, which was still tightly connected to the laminate. The damage in the laminate was located within a distance of 9 mm around the fastener. Resin discoloration was observed but very few fibers were exposed or warped.

For component D, no obvious damage was observed on the top of the fastener, but the fastener was loose and slightly sunken. Fiber tufts, fiber breakage, resin sublimation, and resin discoloration appeared on the surface of the laminate. These forms of damage extended 35 mm along the fiber direction and 19 mm orthogonal to the fiber direction. In addition, flaky fiber shedding was observed along the fiber direction.

#### *4.2. Damage in the Fastener Hole*

After the lightning strike tests, each specimen was cut, as shown in Figure 6c, and the cross-section of the hole was observed with a stereomicroscope (Leica/M125, Germany). Figure 6a,b are cross-sectional photos of the specimen after the action of components C and D, respectively. In these figures, the regularly spaced vertical lines are the shadows left by the synthesis of multiple cross-sectional photos, which do not represent the damage.

**Figure 6.** *Cont.*

**Figure 6.** Damage in the fastener hole of the specimens after lightning strikes: (**a**) component C; (**b**) component D. (**c**) Specimen cutting diagram for microscopic evaluation.

For component C, black particles cover the surface of the inner wall of the fastener hole and few fiber breakages or cracks are observed between the layers, which indicates that the temperature of the inner wall of the hole was not high when the specimen was subjected to a lightning strike and that the amount of resin vaporized by the lightning strike was not substantial.

For component D, Figure 6b shows extended delamination damage away from the hole. Resin matrix cracks and fiber breakages are observed around the hole, and black particles also appear in the hole. As the fastener penetrated the laminate, the delamination damage near the hole became extremely serious.

#### *4.3. Internal Damage*

The internal damage in the CFRP laminate can be non-destructively evaluated with an ultrasonic scanning device (KSI V400E, Germany). The frequency of the ultrasonic wave was 40MHz, and the pulse–echo mode was used for scanning. Ultrasound C-scan, which is sensitive to delamination, is based on the reflection of ultrasonic energy from the intermediate interface. When the ultrasonic wave encounters the damaged interface, the reflected energy in the form of pulse–echo amplitude is different from the undamaged condition [33]. The reflected ultrasonic signals are converted into image signals with different gray values.

Figure 7a,b show the internal damage morphology under the action of components C and D, respectively. These figures show that the damage expands from the fastener to the delamination boundary after the lightning strikes. The damage within the delamination boundary in Figure 7 consists of two zones. One is the thermal decomposition damage and ablation damage, where the material is pyrolyzed and vaporized (marked with dark red, hereinafter referred to as the decomposition damage). The other is speculated to be delamination (marked with blue), where the damage of the interlayer structure is caused by the internal pressure generated by the rapid expansion of pyrolysis gas [34].

**Figure 7.** Ultrasonic C-scan results of specimens after lightning strikes: (**a**) component C; (**b**) component D.

For component C, the damage area inside the specimen is small, the damage lengths in the 0◦ direction and 90◦ directions are approximately 30 mm, and the difference between the two directions is small (in Figure 7a, the damage length in the 0◦ direction is 30.4 mm, whereas the damage length in the 90◦ direction is 29.8 mm). This finding indicates that a small amount of resin in the specimen has undergone pyrolysis and gasification during the lightning strike.

For component D, a large area of damage is observed in the specimen, and the damage lengths in the 0◦ and 90◦ directions are greater than 100 mm, which indicates that during the lightning strike, component D induces more resin pyrolysis and gasification than component C. Moreover, an internal explosion occurs under component D. The rhomboid-shaped damage area might be caused by the orthogonal stacking structure of the specimen.

#### *4.4. Finite Element Simulation Results*

During a lightning strike, the Joule heat generated by the conduction current flowing through the laminate will cause a continuous rise in the CFRP temperature [35]. A temperature contour exceeding a specific threshold value is used to characterize the damage area in each layer [16]. The threshold value usually adopts the decomposition temperature of the resin. When the temperature exceeds this temperature, the area is considered to be the decomposition damage area. According to Table 1, the decomposition temperature of the resin is set to 300 ◦C [29]. Therefore, the area surrounded by the 300 ◦C temperature contour in the FEA represents the same effect as the dark-red area (decomposition damage) under the C-scan.

Figure 8(a1–a3) are the temperature profiles of the specimen under component C in the conduction current + heat flux, conduction current, and heat flux simulation models, respectively. Figure 8(a1,a2) indicate that the decomposition damage shape and size of the laminates are similar to each other and that the damage shape is circular. Such decomposition damage shapes, which are similar to the dark-red area in Figure 7a, do not reflect the anisotropy of the CFRP laminates because component C has a long duration, providing Joule heat in sufficient time to spread in all directions. The first layer (0◦ direction) and the second layer (90◦ direction) are two orthogonal layers, and their temperature distributions are shown in the upper-right corner and lower-right corner of Figure 8(a1), respectively. The decomposition damage shapes of the two layers are almost identical.

**Figure 8.** Temperature profiles of specimens under different simulation models and different lightning current components: component C in the (**a1**) conduction current + heat flux model, (**a2**) conduction current model, and (**a3**) heat flux model; component D in the (**b1**) conduction current + heat flux model, (**b2**) conduction current model, and (**b3**) heat flux model.

In addition, the temperature of the top of the fastener (in the center of Figure 8(a1)) exceeds 3000 ◦C after the combination of the conduction current and the heat flux. This temperature is much higher than the melting temperature of the fastener and causes severe ablation damage to the fastener. Figure 8(a3) shows the temperature distribution when using the heat flux model. In this model, the damage to the

laminate caused by heat flux is approximately 2.5mm, and the temperature of the fastener is similar to that in the conduction current + heat flux model, which means that the ablation damage of the fastener under the action of component C has an important relationship with the heat flux.

Figure 9(a1–a3) are the cross-sectional temperature profiles of the specimen under component C in three simulation models. The heat diffusion can be inferred from the temperature gradient of the fasteners in Figure 9(a1,a3); heat is transferred to the CFRP laminate along with the fastener, causing the material temperature around the fastener to increase. In Figure 9(a2), the temperature of the fastener is far less than 300 ◦C and the decomposition damage occurs in a small area around the fastener. Because the damage occurs in the conduction current model, it is completely caused by the Joule heat generated by the conduction current. According to the results of Figure 5a, Figure 8(a1–a3), and Figure 9(a1–a3), a small area of CFRP laminate damage and serious fastener damage will clearly occur under component C. Thus, arc heating (heat flux) is the main cause of fastener damage under component C.

**Figure 9.** Cross-sectional temperature profiles of the fastener hole under different simulation models and different lightning current components: component C in the (**a1**) conduction current + heat flux model, (**a2**) conduction current model, and (**a3**) heat flux model; component D in the (**b1**) conduction current + heat flux model, (**b2**) conduction current model, and (**b3**) heat flux model.

Figure 8(b1–b3) are the temperature profiles of the specimen under component D in the conduction current + heat flux, conduction current, and heat flux simulation models, respectively. Figure 8(b1,b2) show that the decomposition damage shape and size of the specimen were similar, and that the decomposition damage was centered on the fastener and fanned out along the 0◦ and 90◦ directions. The decomposition damage in both directions exceeded 50 mm. This decomposition damage shape, which resembles the dark-red area in Figure 7b, reflects the anisotropy of the CFRP laminates because component D has a short duration, which does not provide sufficient time for the Joule heat to spread evenly in all directions. The first layer (0◦ direction) and the second layer (90◦ direction) are the two orthogonal layers, and their temperature distributions are shown in the upper-right corner and lower-right corner of Figure 8(a1), respectively. Both layers showed that the decomposition damage in the fiber direction was substantially greater than that perpendicular to the fiber direction. In addition, Figure 8(b1,b3) show that no noticeable temperature rise occurred at the top of the fasteners from the heat flux. Figure 8(b3) is the temperature distribution when the heat flux model is used. In this model, the highest temperature at the top of the fastener is approximately 550 ◦C, which does not ablate the fastener. There are no visible signs of damage around the fastener, which means that the heat flux has little contribution to the damage of the CFRP with fasteners when component D is applied.

Figure 9(b1–b3) are the cross-sectional temperature profiles of the specimen under component D in three simulation models. The cross-sections of the specimens in Figure 9(b1,b2) show that the Joule heat generated by the conduction current caused the massive decomposition damage area in the laminate, whereas the fasteners remained at a lower temperature. According to the results of Figure 5b, Figure 8(b1–b3), and Figure 9(b1–b3), large CFRP laminate decomposition damage and minor fastener damage will occur under component D. Thus, the Joule heat generated by the conduction current is the leading factor for CFRP laminate decomposition damage under component D.

Figure 10 compares the damage profiles in the conduction current + heat flux model (marked with the solid red line) with the thermal decomposition damage and ablation damage profiles in the ultrasonic C-scans (marked with the black dotted line). The damage in the specimen caused by component D is much greater than that caused by component C. Figure 10a,b show that the FEA results are in good agreement with the experimental results. Therefore, it is reliable to study CFRP laminates with fasteners by FEA.

**Figure 10.** Damage comparison between simulated lightning strikes and finite element simulations: (**a**) component C; (**b**) component D.

#### **5. Discussion**

Dong et al. [17] considered the Joule heat effect and arc heat effect when investigating the damage of CFRP laminates without fasteners under components D and C. Comparing their results with ours, it can be found that there is a significant difference in damage between CFRP laminates without fasteners and CFRP laminates with fasteners with the same lightning current components.

For CFRP laminates without fasteners, the results showed that the Joule heat effect and the arc heat effect could cause damage. Component D controls the area of in-plane damage, while the sequential injection of component C after D aggravates the in-plane damage and tends to increase the damage depth. The Joule heat effect plays a leading role in component D, while the arc heat effect plays a dominant role in component C.

For CFRP laminates with fasteners, when the lightning strike acts on the fastener, because the conductivity of the fastener is approximately 1000 times greater than that of the carbon fibers [36], the lightning current will be discharged through the fastener first and then redistributed in each layer of the laminate in the form of a conduction current. The amplitude of the conduction current in each layer is related to the stacking sequence, the grounding mode, and the electrical conductivity. During the test, the specimen is grounded on four sides. In this grounding method, the conduction current can be scattered along each layer to the ground boundary. According to the surface damage (Figure 5), the fastener hole damage (Figure 6), and the internal damage (Figure 7), the damages induced by components C and D differ markedly. Laminates subjected to component C sustained less damage than those subjected to component D. However, the fasteners were severely ablated under component C and not under component D.

When the gap between the discharge electrode and the laminate is broken down, an arc will be generated, forming a high-temperature ionization region at 30,000 K [15]. For component C, the high-temperature arc, which has a long duration, acts on the fastener. The accumulation of heat makes the fastener gasify. Figure 5a shows that the fastener has sustained serious ablation damage. For component D, because of the short duration, the heat of the high-temperature arc seldom accumulates in the fastener; thus, Figure 5b shows that the fastener has minor ablation damage under component D.

Figure 5b shows that resin damage occurs on the surface of the laminate far from the fastener after component D is applied. This damage should be caused by dielectric surface discharge [12]. The dielectric breakdown model considers that the discharge direction is determined by the local electrical potential gradient. During the lightning strike, a large amount of induced charge will accumulate on the surface of the specimen. When the electric field intensity generated by the accumulated charge exceeds the critical value, discharge will occur on the surface of the material, causing damage to the resin. For component C (Figure 5a), due to the low amplitude of the discharge voltage, dielectric breakdown is less likely to occur; thus, the resin damage on the specimen surface should be caused by the metal droplets splashing and attaching to the surface after the fastener has melted.

During a lightning strike, the resin on the inner wall of the fastener hole is pyrolyzed and gasified, which generates high-temperature, high-pressure gas [37]. The heated gas expands rapidly, producing a high-speed gas flow doped with black particles, which may be the product of carbon fiber sublimation and pyrolysis carbonization at high temperatures [38]. Some of the black particles entered the fastener nut through the gap, whereas the rest remained on the inner wall of the fastener hole (Figure 6).

Figure 7 shows that there is a great difference in the internal damage of the laminates under different lightning components.

For component C, the current amplitude is 200 A. Due to the high current density around the fastener, a large amount of Joule heat can be generated. Moreover, the arc heat will spread to the laminate through the fastener (see Figure 9(a1,a3)). The combined effect of the Joule heat and arc heat results in resin pyrolysis. In the zone far from the fastener, the current density and the arc heat decay rapidly, and no internal damage in this zone can be observed under the C-scan. The internal damage caused by component C is concentrated in a small area around the fastener, as shown in Figures 7a and 8(a1).

For component D, the fastener spreads the conduction current to all the layers, and this current is approximately 6.7 kA per layer. These conduction currents can generate large amounts of Joule heat, allowing the resin to pyrolyze over large areas and release gas. According to Figure 6b, Figure 7b, and Figure 9b, it can be speculated that these high-temperature, high-pressure gases trapped inside the CFRP laminate may cause an internal explosion when the gas pressure reaches a critical level, which will result in serious delamination damage.

#### **6. Conclusions**

In this work, components C and D were used in simulated lightning strike tests. Ultrasonic C-scans and stereomicroscopy were used to evaluate the damage sustained by the specimens during the lightning strike tests. Moreover, the electrothermal coupling theory was adopted to model the different effects of the arc heat and Joule heat. The following conclusions are drawn.

(1) The damage to the laminate was concentrated around the fasteners. The conduction current flowed through the fastener to all layers and caused damage in each layer. With increasing distance from the fastener, the current density and the arc heat decayed.

(2) Component D, which had a high current amplitude and a short duration, led to serious surface and internal damage in the CFRP laminate and little damage to the fastener. The damage was mainly caused by the Joule heat generated by the conduction current. Component C, which had a low current amplitude and a long duration, ablated and gasified the fastener and caused less damage to the CFRP laminate. In this process, the arc heating produced by the arc played a leading role.

(3) The temperature profiles in the conduction current + heat flux model were analogous to the thermal decomposition damage and ablation damage profiles from the C-scan. Therefore, the conduction current + heat flux model is reasonable in FEA of CFRPs with fasteners subjected to lightning strikes. The simulation results show an obvious anisotropy in Component D but not in Component C, because component C has sufficient time for the Joule heat to spread in all directions, whereas component D lacks sufficient time.

This work evaluated the damage in CFRP laminates with fasteners subjected to lightning current components C and D and found that the damage under different lightning current components presented unique characteristics. Due to the variety of lightning strikes in nature, the damage induced by other lightning current components and multi-components needs further study on the basis of the research results of the single lightning current component.

**Author Contributions:** Conceptualization, J.C. and Z.F.; data curation, J.C.; formal analysis, J.C., X.B., and J.L.; funding acquisition, Z.F.; investigation, J.C., X.B., and J.L.; methodology, J.C.; project administration, Z.F.; resources, Z.F.; supervision, Z.F.; validation, J.C.; visualization, J.C., X.B., and J.L.; writing—original draft, J.C.; writing—review and editing, Z.F. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by National Key R&D Program of China, grant number 2017YFC1501506.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
