Experimental and Numerical Investigations on the Effect of MWCNT-COOH and Al₂O₃ Hybrid Nanofillers Dispersed CFRP Laminates Subjected to Projectile Impact

Abstract

Although carbon fiber-reinforced polymer (CFRP) composites have excellent intrinsic mechanical properties, they are vulnerable to impact loads because of their weak inter-laminar fracture toughness, which results in delamination damage. This study presents a novel hybrid nanofiller combination of multi-walled carbon nanotubes (MWCNT) and alumina nanoparticles (Al₂O₃) to determine improvements in the impact resistance of CFRP laminate. The projectile impact experimental study is carried out on 140 mm × 140 mm × 1.5 mm CFRP laminate with spherical and conical nose shape projectiles. The numerical study of the test plate consisting of six layers is performed, in which each layer is modeled using a shell element and connected through tiebreak contact. Using the Cowper–Symonds equation to determine the dynamic mechanical properties, the numerical validation is established considering the strain rate effect. The results, such as residual velocity, damage area, ballistic limit velocity and delamination obtained from numerical analysis, are compared with the experimental observations. In laminates with hybrid nanofillers, residual velocity decreased by 20% and 9% when spherical and conical projectiles were impacted, respectively. The study indicates that 0.1 wt% MWCNT + 1 wt% Al₂O₃ nanofiller concentration embedded CFRP offers better resistance against spherical and conical projectile impact.

1. Introduction

CFRPs (Carbon Fiber Reinforced Polymer) find wide applications as structural materials in modern aircraft and automobile industries due to their high specific strength and rigidity [1,2,3]. Composite materials are essential for preventing or slowing projectiles in military applications, particularly in armor technology. However, because of its weak out-of-plane mechanical characteristics, CFRP structures are vulnerable to damage when impacted by foreign objects, including birds, hail, bullets, and explosive fragments with low masses. The impact-related damage scenario in composites involves several damage mechanisms, like fiber breakage, matrix cracking, and delamination. Delamination is, in fact, regarded as the most significant damage mode resulting in the deterioration of composite stiffness and strength [4].

At both the macro and micro scales, researchers have made multiple attempts to enhance the resistance to impact and the ability to withstand the damage of CFRP. Buitrago et al. [5] investigated the energy absorption mechanism in thin woven laminates comprising carbon fibers and concluded that significant energy absorption for impacts below the perforation velocity is connected to elastic deformation of fibers and shear plugging. In contrast, impacts beyond the perforation velocity are related to the acceleration field of the laminate and shear plugging. Hosur et al. [6] studied the response of stitched and unstitched plain and satin weave laminates to projectile impact. They reported that satin weave-based panels exhibited flexural-tensile failure with increased damage area compared to shear failure mode in plain weave laminates. One method to improve fracture toughness is introducing nanofillers in the neat matrix. Dispersion of MWCNT in the epoxy resin increased CFRP laminae’s interlaminar toughness and impact strength [7,8,9]. Tehrani et al. [10] investigated the damping performance, impact resistance and impact damage progression of CFRP laminates by adding MWCNT to the epoxy matrix of the composites and determined that there was a 12% enhancement in failure strain. Kara et al. [11] studied the impact response of CFRP laminates with thin layers of MWCNT, aramid pulp, and graphene. The results indicated that compressive strength after impact improved by adding nanofillers.

Surface modification by attaching the functional groups onto the MWCNT surfaces can considerably improve the interfacial adhesion between MWCNT and the epoxy matrix. Cheon and Kim [12] investigated the impact resistance by introducing functional MWCNT and carbon fiber through acid and flame treatments and reported that functionalized MWCNT on the carbon fiber considerably increased the impact resistance of CFRPs. Rahman et al. [13] investigated the ballistic performance of E-glass/epoxy composites subjected to spherical projectile impact by incorporating amino-functionalized MWCNT (NH2-MWCNT) in the epoxy and reported a reduction in damage size with functionalized MWCNT-based laminates. Naghizadeh et al. [14] experimentally studied E-glass reinforced with MWCNT functionalized carboxyl (COOH-MWCNT) composites under high-velocity impact and observed reduced damage around the impact region for a conical projectile impact at different velocities.

In light of the promising improvement in impact properties obtained by adding nano-fillers to the neat matrix, further numerical studies were carried out to evaluate the damage mechanism under impact. Wang et al. [15] presented experimental and numerical investigations on CFRP panels subjected to projectile impact and compared the energy-absorbing capability of the composite for different thicknesses. Menna et al. [16] presented numerical studies on the low-velocity impact behavior of GFRP laminates using LS-DYNA and compared them with the experimental results, such as force, displacement and material damage. Additionally, Nunes et al. [17] investigated the ballistic impact of aramid composites and compared residual velocity and damage morphology of the impacted laminate. Morka and Jackowska [18] used a similar tool for numerical investigations of the ballistic impact resistance of CNT-reinforced composite plates. Sánchez et al. [19] presented a numerical approach for predicting impact damage and residual velocity of composite laminates subjected to spherical ball impact and concluded that the radius of a spherical ball influences the damage. Giannaros et al. [20] investigated the hypervelocity projectile impact effect on CFRP using smoothed particle hydrodynamics (SPH) approach, while Scazzosi et al. [21] used the same method and compared the numerical results of ballistic limit velocity and ballistic curve with experiments for a multi-layered armor system consisting of Kevlar and alumina. Shyamsunder et al. [22] implemented an orthotropic viscoelastic-viscoplastic material model with strain rate dependence to validate deformation, damage and failure of a flat unidirectional fiber-reinforced composite plate using a hollow 50 g Al-2024 projectile at different velocities and concluded that the developed material model and modeling techniques yielded good predictions.

Numerous researchers [23,24,25,26,27] studied the effect of nanofillers in composite laminates and reported that the mechanical, tribological and sliding wear properties improved with the addition of nanofillers. Kaybal et al. [28] presented the low-velocity impact behavior of nano-fillers dispersed CFRP composite laminate and reported that 2wt% of nanofiller dispersion in epoxy offers better impact resistance. Ávila et al. [29] studied the impact response of glass/epoxy laminates reinforced with nano clay and graphene nanosheets and reported that adding nanofillers improved the impact resistance and damage mechanism. Bandaru et al. [30] presented a numerical investigation on hybrid composites under projectile impact and reported that the stacking sequence plays an important role in impact resistance. Koricho et al. [31] investigated the impact behavior of hybrid nano clay and glass bubble dispersed GFRP laminates and reported that nano clay dispersed composite exhibited higher energy absorption compared to hybrid and neat. Gregori et al. [32] presented numerical and analytical investigations on the alumina/aramid composites and compared the residual velocity and deformation. Shibao et al. [33] also investigated the alumina/aramid-carbon hybrid composite laminates impacted by 7.62 MGI AP projectile numerically and concluded that the ballistic performance and integrity of the structure of the hybrid composite is better with constant CF proportion. It is found from the above literature that COOH-MWCNT and alumina offer better resistance to impact loading; hence, this paper attempts to study the projectile impact response of COOH-MWCNT and Al₂O₃ based hybrid nanofiller dispersed epoxy resin-based CFRP laminates.

This paper presents experimental and numerical investigations on the projectile impact behavior of COOH-MWCNT and dispersed CFRP laminate. The hybrid nanofiller/CFRP laminates of size (140 mm × 140 mm × 1.5 mm) were fabricated using the vacuum bagging method and tested using an in-house built gas gun-based projectile impact test setup using spherical and conical projectiles at four different impact velocities. Ultrasonic C-scan was used to assess the dispersion of nanoparticles and the damaged area within the laminate layers. The numerical analysis is performed with and without strain rate effects considering Chang-Chang’s composite damage material model. The numerically predicted residual velocity, energy absorbed, and damage morphology are compared with experimental results.

2. Experiment

2.1. Material and Specimen Preparation

A series of projectile impact experiments were carried out utilizing in-house developed test equipment to investigate the effect of hybrid nanofiller reinforcement on the failure behavior of CFRP composites. Three different types of laminates, namely; neat CFRP, CFRP with 0.1 wt% MWCNT-COOH + 1 wt% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ (C1A1) and CFRP with 0.1 wt% MWCNT-COOH + 2 wt% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ (C1A2) were fabricated and tested. Plain weave high-strength carbon fabric (200 GSM) with identical warp and fill yarns and a thickness of 0.25 mm were utilized to fabricate the specimens in this study. The matrix material is an aerospace-grade Bisphenol-A-based epoxy resin (LY556) mixed with a 10:1 weight ratio of hardener (HY951). The nanofillers used in this study are Alumina (${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$) nanopowder with an average particle size of 20–30 nm and MWCNT (-COOH functionalized) with 95% minimum purity having an outer diameter range of 30–50 nm, length ranging from 10–30 µm and 0.7wt% COOH content.

Multiple manufacturing processes exist for the effective fabrication of composite laminates, but the pultrusion manufacturing method is the most cost-effective of all known processes [34,35]. In the presented study, CFRP composite specimens were manufactured using a hand layup approach followed by a vacuum bagging procedure, as shown in Figure 1. Six 500 mm × 500 mm square pieces with [0/90] layup configuration were placed and cured with epoxy resin to make a composite laminate of 1.5 mm thickness. Nanofillers were impregnated into the epoxy matrix using an ultrasonic processor at 30 kHz for approximately 60 min to improve matrix-fiber bonding properties. Before fabrication, the composite consisting of pure epoxy, C1A1 and C1A2 nanofiller configuration was mechanically mixed at room temperature to avoid additional agglomerations. Following the addition of nanofillers, the composite laminate was vacuum bagged and cured at room temperature for 24 h. The fiber volume percentage of the specimen was determined using the burn-out method (ASTM D3171) and was found to be in the 53–55%. To cut composite plates into 140 mm × 140 mm, water jet cutting with a 1mm cutting edge margin was used. Further, 12 holes of 8 mm diameter were drilled into the specimen with a feed rate of 0.20 mm/rev and a spindle speed of 1320 rpm to secure the specimen in the impact fixture.

2.2. Material Testing

The basic material properties under tension, compression and shear were obtained using material samples cut from the base plate. Four samples for each test were cut out from the base plate, and the tests were conducted using a 100 KN MTS universal testing machine. According to ASTM D3039 [36], the tensile tests are carried out using a feed rate of 2 mm/min. The tensile strengths in longitudinal direction ($X_t$), transverse direction ($Y_t$), elastic modulus ($E_x$ and $E_y$) and poisons ratio (${\nu }_{xy}$) were obtained from the tensile stress–strain curve. The compression test was carried out according to ASTM D6641 [37] to determine the compressive strengths ($X_c,Y_c$) using a feed rate of 2 mm/min. A v-notched rail shear test was carried out at 2 mm/min, according to ASTM D7078 [38], to determine the in-plane shear strength ($S_c$) and modulus ($G_{xy}$) of the composite. The results obtained from the sample testing with the basic standard deviation are presented in

Table 1. Mechanical properties of woven carbon fiber laminated composite.

Property	Values			ASTM Standards
	Neat CFRP	C1A1	C1A2
Density ρ (kg/m3)	1480 ± 5	1500 ± 5	1500 ± 5	ASTM D3171
Elastic Modulus $E_{11}=E_{22}$ (GPa)	18.27 ± 0.66	24.74 ± 0.82	22.42 ± 0.64	ASTM D3039
Shear Modulus $G_{12}=G_{13}$ (GPa)	4.80 ± 0.01	5.10 ± 0.09	4.80 ± 0.38	ASTM D7078
Major Poisson’s ratio ${\nu }_{21}$ (mm/mm)	0.07	0.07	0.07	DIC-MATLAB
Tensile Strength, $X_t$ = $Y_t$ (MPa)	463.45 ± 18.43	583.41 ± 40.34	546.69 ± 65.92	ASTM D3039
Compressive Strength, $X_c=Y_c$ (MPa)	324.41 ± 1.66	414.21 ± 4.33	382.68 ± 0.86	ASTM D6641
Shear strength $S_c$(MPa)	53.42 ± 0.13	82.39 ± 0.87	59.93 ± 2.07	ASTM D7078
Normal failure stress, NFLS (MPa)	52 ± 6	88 ± 4	52 ± 2	ASTM D638
Shear failure stress, SFLS (MPa)	31 ± 5	61 ± 5	41 ± 4	ASTM D2344

. The stress–strain curves obtained from the tests from tensile, compression and shear are presented in Figure 2, Figure 3 and Figure 4, respectively. Similar ASTM standards were referred to by several authors to determine the mechanical properties of CFRP laminates [39,40,41].

2.3. Projectile Impact Testing

A single-stage gas gun system was used for the projectile impact test, as illustrated in Figure 5. A compressor, reservoir with a 10 bar pressure capacity, regulating valve, barrel, and a rebound protection casing to prevent the reflecting bullet are all part of the test setup. During projectile impact testing, the air is pressurized to a certain pressure in the reservoir using a compressor, and the pressurized air is discharged into the barrel to launch the projectile at the required muzzle velocity. As shown in Figure 6, projectiles of two shapes, spherical (12 mm diameter and 7.12 g mass) and conical (12 mm diameter with an apex angle of 60°, 14.5 mm length, and 7.62 g mass), were used to investigate the projectile impact response of CFRP composite laminates. The composite plate experiments were carried out by clamping the test plate between two fixture plates made of mild steel plates of 20 mm thickness with the same dimensions as the test plates (140 mm × 140 mm) with an exposed area of 80 mm × 80 mm. The test plate is secured to the fixture using 12 M8 bolts to simulate clamped conditions. Figure 7 depicts the entire test setup.

Figure 8 represents the impact response of CFRP laminates at impact velocity levels ranging from 38.2 m/s to 96.8 m/s. In the case of spherical and conical projectile impact, the plate exhibited only denting when the impact velocity was less than 45 m/s. Denting is found to be the least in the case of C1A1 when compared to the other configurations. The front face of all specimens had a circular opening, and the back face had a four-sided pyramidal opening, which can be attributed to the bi-directional layup of the woven carbon fabric composite plate. Samples with ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ witnessed a more ductile mode of failure and absorbed more kinetic energy through the formation of new surfaces in the surroundings of the impacted area. In the case of complete perforation, penetration occurs largely on the rear face of composites, leading to cone formation involving tensile failure of primary yarns, elastic deformation of secondary yarns, matrix cracking, delamination, and in some cases, shear plugging of the laminate was observed when impacted by a conical projectile.

3. Theoretical Background

The numerical study of the projectile impact on CFRP panels is studied using LS-DYNA commercial non-linear finite element code using the Lagrangian approach considering rigid projectile structure interaction.

3.1. Material Model

For the numerical simulation of plain-woven carbon fiber laminated, continuum damage mechanics is utilized to predict intra-laminar damage in laminate during projectile impact. Failure criteria like Hasin [42,43] and Chang-Chang [44,45] developed to predict the failure initiation and evolution are utilized in numerical simulation to determine the failure behavior. Many examples can be determined in literature [17] showing the efficiency of the Chang-Chang material model’s inefficient prediction of composite failure under projectile impact. Thus, to study the failure behavior, an orthotropic elastic behavior-based Chang-Change failure criterion was utilized.

The Chang-Chang failure criterion is a brittle failure model that takes an in-plane stress-induced matrix and fiber failures into consideration. In the elastic zone, the material’s stress–strain behavior is given by

$${\epsilon }_{11}=\frac{1}{E_{11}}\left({\sigma }_{11}-{\upsilon }_{12}{\sigma }_{22}\right)$$

(1)

$${\epsilon }_{22}=\frac{1}{E_{22}}\left({\sigma }_{22}-{\upsilon }_{21}{\sigma }_{11}\right)$$

(2)

$$2{\epsilon }_{12}=\frac{1}{G_{12}}\left({\sigma }_{12}+\alpha {\sigma }_{12}^3\right)$$

(3)

where ε₁₁ (mm/mm), ε₂₂ (mm/mm) are the normal stain_, in direction 1(longitudinal) and direction 2 (lateral) ε₁₂ (mm/mm) shear strain, σ₁₁ (N/mm), σ₂₂ (N/mm) the normal stress_, in directions 1 and 2, σ₁₂ (N/mm) shear stress, E₁₁ (N/mm), E₂₂ (N/mm) the elastic modulus in direction 1 and 2, G₁₂ (N/mm) shear modulus and α a unitless weighting factor for the nonlinear shear stress.

The following equations give the failure flags for the history variables e_f, e_c, e_m and e_d, which represent tension and compression for the fiber direction and tension and compression for the matrix direction.

For the fiber tension mode,

$${e_f}^2={\left(\frac{{\sigma }_{aa}}{X_t}\right)}^2+\beta \left(\frac{{\sigma }_{ab}}{S_c}\right)-1\left\{\begin{array}{c}\ge 0\hspace{1em}failed\\ <0\hspace{1em}elastic\end{array}\right.$$

(4)

$$E_a=E_b=G_{ab}=v_{ba}=v_{ab}=0$$

For the fiber compression mode,

$${e_c}^2={\left(\frac{{\sigma }_{aa}}{X_c}\right)}^2-1\left\{\begin{array}{c}\ge 0\hspace{1em}failed\\ <0\hspace{1em}elastic\end{array}\right.$$

(5)

$$E_a=v_{ba}=v_{ab}=0$$

For the matrix tension mode,

$${e_m}^2={\left(\frac{{\sigma }_{bb}}{Y_t}\right)}^2+{\left(\frac{{\sigma }_{ab}}{S_c}\right)}^2-1\left\{\begin{array}{c}\ge 0\hspace{1em}failed\\ <0\hspace{1em}elastic\end{array}\right.$$

(6)

$$E_b=v_{ba}=0\Rightarrow G_{ab}=0$$

For the matrix compression mode,

$${e_d}^2={\left(\frac{{\sigma }_{bb}}{2S_c}\right)}^2+\left[{\left(\frac{Y_c}{{2S}_c}\right)}^2-1\right]\left(\frac{{\sigma }_{bb}}{Y_c}\right)+{\left(\frac{{\sigma }_{ab}}{S_c}\right)}^2-1\left\{\begin{array}{c}\ge 0\hspace{1em}failed\\ <0\hspace{1em}elastic\end{array}\right.$$

(7)

$$E_a=v_{ba}=v_{ab}=0\Rightarrow G_{ab}=0$$

Once the failure condition is satisfied, the defined elastic constants are considered to be zero and the element is deleted from the simulation.

${\sigma }_{aa} (\mathrm{N}/\mathrm{m}\mathrm{m}),{\sigma }_{bb} (\mathrm{N}/\mathrm{m}\mathrm{m})$ are the stresses in fiber and matrix direction, ${\sigma }_{ab} (\mathrm{N}/\mathrm{m}\mathrm{m})$ represent the in-plane shear stress $X_t (\mathrm{N}/\mathrm{m}\mathrm{m}),X_c (\mathrm{N}/\mathrm{m}\mathrm{m})$ the longitudinal tensile and compressive strength, $Y_t (\mathrm{N}/\mathrm{m}\mathrm{m}),Y_c$ (N/mm) the transverse tensile and compressive strength whereas S_c (N/mm) represents shear strength and $\beta$ is the unitless weighting factor for the shear term in tensile fiber mode.

Equations (4)–(7) are used to calculate each failure as given above. Whenever the failure criteria are satisfied at the integration points of an element, it distorts rapidly, thus preventing the failed ply from carrying additional stress. Four critical strain values were used to introduce element deletion, which can delete an element and reduce the stresses to zero, namely, maximum strain for fiber tension (DFAILT) (mm/mm), maximum strain for fiber compression (DFAILC) (mm/mm), maximum strain for matrix-straining in tension and compression (DFAILM) (mm/mm), and maximum tensorial shear strain (DFAILS) (mm/mm). At each time increment, the longitudinal strain in each direction of the ply axis is evaluated; if one of them reaches an ultimate value, the element is removed from the mesh. The material properties used for modeling each ply of the laminate are obtained from the specimen’s tensile, compression and shear testing as per the respective ASTM standard. The average value of the material properties for four samples with standard deviation is considered in the analysis for all three types of composites and is presented in Table 1.

3.2. Failure Model for Delamination

Delamination damage was modeled in the simulation using a surface-to-surface tiebreak contact algorithm based on interlaminar properties in terms of normal and shear strengths. Between shell elements of each ply, the penalty contact formulation (i.e., automatic-surface-to-surface-tiebreak) was chosen from the various formulations available in LS-DYNA. Each ply is modeled as a shell layer of elements in this method, but the nodes between plies initially in contact are bound together, preventing sliding motions until a failure criterion is reached, contributing to delamination onset. The nodal tension, in particular, is tracked during the study and incorporated into the interface strength-based failure criterion as expressed in Equations (8).

$${\left(\frac{\left|{\sigma }_n\right|}{NFLS}\right)}^2+{\left(\frac{\left|{\sigma }_s\right|}{SFLS}\right)}^2\ge 1$$

(8)

where NFLS (N/mm) and SFLS (N/mm) are the normal and shear interlaminar strength, ${\sigma }_n$ (N/mm) and ${\sigma }_s$ (N/mm) are the normal and shear stresses. Both NFLS (matrix tensile test) and SFLS (interlaminar shear strength) are determined experimentally (Table 1), and the test findings showed improved normal and shear failure strength in laminates reinforced with MWCNTs, which is consistent with the available literature [33].

3.3. Strain Rate Effect on CFRP Composite

To study the effect of dynamic loading on CFRP laminates, the Cowper–Symonds equation (Equations (9)) is used to relate quasi-static and dynamic strengths to strain rate, as given by [46]

$${\sigma }_{11}\left(\dot{\epsilon }\right)={\sigma }_d\left(\dot{\epsilon }\right)=\left\{\begin{array}{cc}{\sigma }_0,& \dot{\epsilon }\le 1.6X{10}^{-3}{s}^{-1}\\ {\sigma }_0\left[1+{\left(\frac{\dot{\epsilon }}{A}\right)}^{m_{\sigma }}\right],& \dot{\epsilon }>1.6X{10}^{-3}{s}^{-1}\end{array}\right.$$

(9)

where ${\sigma }_0$ (N/mm) and ${\sigma }_d$ (N/mm) are the quasi-static and dynamic strengths at given strain rate. ‘A (s⁻¹)’ represents a material parameter for strength, and ‘$m_{\sigma }$’ is a unitless parameter representing strain rate sensitivity estimating parameters for strength. The strain rate parameters for CFRP composites’ strength (A, $m_{\sigma }$) are taken from the literature [46]. The dynamic strength of the material is calculated using the average strain rate, which is obtained from the strain time history at each layer. The final average strain rate is obtained by averaging the strain rate of all six layers, and the dynamic strength parameters are calculated using Cowper–Symonds Equation.

3.4. Ballistic Limit Velocity

Projectile impact behavior on composite laminates is a complex process involving projectile properties, orientation and the target material. A projectile impact can lead to three distinct conditions, namely: (a) projectile bounce back, indicating that the energy transferred is less than the laminate’s energy absorption capability. (b) ballistic limit velocity at which the projectile perforates the target while exhibiting zero residual velocity. (c) There is a residual velocity after perforation, indicating the energy transferred is higher than the energy absorption capability of the laminate. The ballistic limit velocity is an important criterion in the design of any structure under impact loadings. The ballistic limit velocity and energy absorbed can be computed using Equations (10) and (11) [47].

$${\mathrm{V}}_{\mathrm{b}}=\sqrt{{\mathrm{V}}_{\mathrm{i}}^2-{\mathrm{V}}_{\mathrm{r}}^2}$$

(10)

$${\mathrm{E}}_{\mathrm{a}\mathrm{b}\mathrm{s}}=\frac{1}{2}\mathrm{m}{\mathrm{V}}_{\mathrm{i}}^2-\frac{1}{2}\mathrm{m}{\mathrm{V}}_{\mathrm{r}}^2$$

(11)

where m represents mass in kg, and ${\mathrm{V}}_{\mathrm{b}},{\mathrm{V}}_{\mathrm{i}}{,\mathrm{V}}_{\mathrm{r}}$ are ballistic, initial and residual velocities $(\mathrm{m}/\mathrm{s})$, respectively.

4. Finite Element Modeling

The finite element modeling of the projectile impact test fixture, along with the test plate subjected to projectile impact, is carried out using LS-DYNA nonlinear finite element code. Because of symmetry, only one-fourth of the geometry was considered for modeling to save computational time. In the experimental setup, the test plate is clamped between two fixture plates made of mild steel of 20 mm using 12 M8 bolts connecting the top cover plate, test plate, and bottom cover plate. For the analysis, the projectile, impact test fixture, bolt head, bolt shank, and nut are modeled using Lagrangian solid elements with elastic properties and the standard material properties of steel having an Elastic Modulus of 210 GPa, the density of 7860 kg/m³ and Poisson’s ratio of 0.3. In the finite element modeling of a test plate consisting of six layers, each layer is modeled using shell elements and connected through tiebreak contact. The contact between the bolt and bolt hole, bolt head, nut and the test plate, the projectile impact test fixture, is defined using automatic surface-to-surface contact.

A detailed convergence study with different sets of elements for the test plate is performed. The element size of 0.25 mm and 0.20 mm have shown nearly the same value for the residual velocity. Hence, in all subsequent analyses for computational efficiency, a 0.25 mm mesh size is used. The complete finite element model of the projectile impact test fixture, along with the test plate, bolt, and projectile, is shown in Figure 9.

5. Results and Discussion

5.1. Experimental

To investigate the effect of projectile geometry on the impact response of neat CFRP, C1A1 and C1A2 samples, four impact energies were maintained for each projectile geometry, totaling 24 experiments, with eight experiments for each laminate variant. After each impact test, the test panels were analyzed using ImageJ 1.53e (free software) to assess the back-face damage. The visible damage on all of the panels was divided into three categories, namely: front-face damage, back-face damage, and interior delamination. Figure 10 depicts the rear face damage of a typical panel caused by a spherical bullet impact at a velocity of 79.6 m/s. For the majority of the test samples, after projectile penetration, the CFRP laminate nearly closed the holes, sustaining a rhombus-shaped projection. In the case of a projectile rebound, the rear face damage showed a pyramid shape in denting. In all the experimental studies, the rear face damage followed a similar pattern for both spherical and conical projectiles. This behavior is in conformity with the literature [48]. The panel nomenclature, projectile geometry, impact velocity and energy, and damage area for each test specimen are presented in

Table 2. Panel nomenclature, velocity and energy.

Test No	Panel Designation	Impactor Nose Shape	Incident Velocity	Incident Energy	Residual Velocity	Residual Energy
			Vi (m/s)	Ei (J)	Vr (m/s)	Er (J)
1	N_S1	Spherical	42.80	6.52	Rebound	-
2	N_S2		62.20	13.77	37.30	4.95
3	N_S3		72.50	18.71	51.20	9.33
4	N_S4		79.60	22.56	60.20	12.90
5	C1A1_S1		42.00	6.28	Rebound	-
6	C1A1_S2		60.20	12.90	24.70	2.17
7	C1A1_S3		72.00	18.46	44.90	7.18
8	C1A1_S4		79.50	22.50	53.70	10.27
9	C1A2_S1		42.20	6.34	Rebound	-
10	C1A2_S2		61.00	13.25	32.90	3.85
11	C1A2_S3		71.80	18.35	46.80	7.80
12	C1A2_S4		79.80	22.67	54.20	10.46
13	N_C1	Conical	40.20	6.16	Rebound	-
14	N_C2		63.70	15.46	48.60	9.00
15	N_C3		82.40	25.87	71.60	19.53
16	N_C4		95.40	34.68	85.30	27.72
17	C1A1_C1		38.20	5.56	Rebound	-
18	C1A1_C2		64.40	15.80	43.80	7.31
19	C1A1_C3		84.30	27.08	68.00	17.62
20	C1A1_C4		96.80	35.70	81.20	25.12
21	C1A2_C1		41.00	6.40	Rebound	-
22	C1A2_C2		63.40	15.31	46.50	8.24
23	C1A2_C3		84.60	27.27	71.30	19.37
24	C1A2_C4		96.10	35.19	84.60	27.27

.

In the neat CFRP laminate, the spherical projectile is impacted at four different incident velocities (42.8 m/s, 62.2 m/s, 72.5 m/s, 79.6 m/s), the incident and residual velocity measured with the help of a velocity sensor (chronograph). The projectile fired at 42.8 m/s rebounded after hitting the target plate, and for the remaining three velocities, the projectile penetrated the test plate. The residual energy and residual velocity were found to increase with an increase in incident velocity. The damaged area measured from the test specimen was found to increase with the increase in velocity as the projectile completely perforates through the panel. The experiments with spherical projectile are repeated with nearly the same velocities (54.7 m/s, 66.1 m/s, 73.3 m/s, 79.3 m/s and 50.9 m/s, 67.1 m/s, 74.6 m/s, 80.3 m/s) for the CFRP with C1A1 and C1A2, and a similar trend is noticed for residual velocity and residual energy. The residual velocity and residual energy for C1A1 samples were less than the other configurations, indicating the increase in energy absorbing capability of 0.1 wt% MWCNT-COOH + wt% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ dispersed CFRP laminates. Similarly, a conical projectile was fired at four impact velocities (40.2 m/s, 63.7 m/s, 82.4 m/s, and 95.4 m/s). The projectile fired below 40.2 m/s and rebounded after hitting the target plate, and for the remaining three velocities, the projectile penetrated the test plate. The incident and residual velocity are measured with the help of a velocity sensor, and the residual velocity is found to increase with an increase in incident velocity.

5.2. Damage Area Representation

In the morphological analysis, the damaged area corresponding to the sample configuration for all the test cases is represented in Figure 11. A centralized indentation was observed in the front face of the laminates after impact with velocities below the ballistic limit by spherical and conical projectiles, followed by a bulge in the back face. At velocities above the ballistic limit, the front face of the laminate showed a circular puncture with a diameter similar to a projectile, as well as a more visible rhombus-shaped rupture on the back face. On the front faces, the collapse was deeper as the incident velocity increased, and transverse cracks were evident on the back face of the laminate. In a comparison of spherical and conical projectiles, it was observed that spherical projectiles produced greater damage than conical projectiles for most of the impact scenarios. This is to be expected since a spherical projectile has a larger contact surface during the impact event than a conical projectile at the same impact velocity. Also, a conical projectile tends to cut through the plies, whereas a spherical projectile induces early delamination, resulting in comparatively higher damaged areas [49].

In general, the C1A1 panels displayed greater ballistic resistance, evidenced by a lower damaged area compared to other sample configurations for all projectile velocities. It is interesting to note that samples with C1A1 configuration deformed less under impact than C1A2 configuration, signifying deterioration in ballistic resistance properties on the addition of MWCNT-COOH beyond 0.1 wt% and ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ beyond 1 wt%, respectively. However, the damage pattern was similar for all the laminate configurations.

5.3. Delamination

The specimen was removed from the fixture after the test, and the panels were scanned with an immersion UT C-scan to obtain a quantifiable mapping of the delamination area to determine the extent of the damage. In order to quantitatively evaluate the influence of hybrid nanofillers on the internal delamination, the damaged area of C-scan images was measured using an image analyzer software (14.1V) hence represent, the damaged dimensions were obtained up to two decimal places for both the spherical and conical projectiles. A typical C-scan image of spherical and conical projectile impact is shown in Figure 12, in which the different color gradient in the impacted zone, distinct from the surrounding color gradient, defines the boundary of the damaged region due to projectile impact. The damaged area obtained using C-Scan for spherical projectile impact for neat at 79.6 m/s, C1A1 at 79.5 m/s and C1A2 at 79.8 m/s is found to be 28.06 mm × 27.10 mm, 22.60 mm × 22.03 mm, and 25.95 × 25.29 mm, respectively. The damaged area obtained using C-Scan for conical projectile impact for neat at 95.4 m/s, C1A1 at 96.8 m/s and C1A2 at 96.1 m/s is found to be 21.65 mm × 21.87 mm, 17.21 mm × 23.66 mm and 20.60 mm × 19.75 mm, respectively. The delaminated area obtained from C-scan is found to be more than the visibly damaged area. The delaminated area for C1A1 and C1A2 samples was lower compared to neat CFRP composites, indicating higher energy absorption by the nanofillers in the matrix without inducing damage to the laminate. Comparing the impact damage qualitatively, it can be concluded that the delamination area is more for spherical projectile impact compared to conical projectile impact. Similar observations were made in the literature determining larger damage areas produced by hemispherical projectiles [50].

5.4. Numerical

5.4.1. Strain Rate and Dynamic Strength Calculation

While composites are generally strong, they become weak when subjected to impact loads. In many of the situations where fiber-reinforced polymer composites are used as potential materials, high strain rate loading is likely. Studies on the strain rate effect of CFRP composites show that fibers play an essential role during tensile loading, but compressive loading is dependent on both matrix and interface parameters [51]. Therefore, it is important to implement the strain rate sensitivity of the CFRP.

To estimate the residual velocity considering the strain rate effect under dynamic loading, the average strain rate was calculated using the strain—time history for each impacting velocity. Firstly, strain-time history was plotted for each layer, as shown in Figure 13, and then the average value of the 6 plies was considered as the strain rate at a particular velocity. The average strain rate obtained for each impacting velocity was used to calculate the dynamic strength using Equation (9), and the respective values are represented in

Table 3. Strain Rate effect on the strength of material.

Panel Designation	Incident Velocity (m/s)	Average Strain Rate (s−1)	Quasi-Static Strenght ${\mathit{\sigma }}_0$ (MPa)	Dynamic Strenghts ${\mathit{\sigma }}_{\mathit{d}}$ (MPa)
N_S1	42.8	916.65	463.45	538.27
N_S2	62.2	1002.82	463.45	545.93
N_S3	72.5	1054.57	463.45	550.55
N_S4	79.6	1217.43	463.45	565.20
C1A1_S1	42	820.36	583.41	666.94
C1A1_S2	60.2	1040.88	583.41	691.51
C1A1_S3	72	1161.55	583.41	705.14
C1A1_S4	79.5	1269.68	583.41	717.46
C1A2_S1	42.2	924.59	546.69	635.79
C1A2_S2	61	1007.93	546.69	644.52
C1A2_S3	71.8	1244.10	546.69	669.57
C1A2_S4	79.8	1350.34	546.69	680.97
N_C1	40.2	895.91	463.45	536.45
N_C2	63.7	1175.25	463.45	561.39
N_C3	82.4	1306.89	463.45	573.32
N_C4	95.4	1423.25	463.45	583.96
C1A1_C1	38.2	827.53	583.41	667.73
C1A1_C2	64.4	1078.82	583.41	695.78
C1A1_C3	84.3	1325.11	583.41	723.81
C1A1_C4	96.8	1443.77	583.41	737.48
C1A2_C1	41	957.26	546.69	639.20
C1A2_C2	63.4	1111.20	546.69	655.42
C1A2_C3	84.6	1197.03	546.69	664.54
C1A2_C4	96.1	1446.44	546.69	691.35

. Further, in the numerical simulation, the dynamic strength was updated, and the corresponding results were obtained.

5.4.2. Rate Dependence Effect on Residual Velocity

When the impact energy is equal to or larger than the perforation energy, the impactor perforates the panel absorbing a part of the impact energy; the other part remains in the impactor as a residual velocity enabling it to continue traveling in the same impact direction, before the velocity was completely diminished. In all the cases for both spherical and conical projectile impact, the residual velocity is found to be considerably low in C1A1 samples compared to the CFRP neat and C1A2 laminates, and the numerically predicted residual velocity shows a very close correlation with experimental values. It is observed that samples with 0.1 wt% MWCNT-COOH + 1 wt% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ nanofillers showed better resistance to projectile impact as compared to samples with 0.1 wt% MWCNT-COOH + 2 wt% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ nanofillers. In the case of conical projectiles, the effect of hybrid nanofillers was less significant when compared with spherical projectiles. This can be attributed to the fact that; the conical projectile impact has been performed at a comparatively greater velocity, and the contact area is lower for a conical projectile when compared with a spherical projectile. The residual velocity obtained from the numerical analysis considering the strain rate effect for all the test cases is represented in Figure 14 [49,50].

5.4.3. Ballistic Limit Velocity

The ballistic limit velocity obtained using both experimental and numerical studies based on the incident and residual velocity is presented in Figure 15. The ballistic limit velocity is calculated using Equation (10). The numerical velocity-based ballistic velocity predictions show a very close correlation with that of the experimental one, within an 8% deviation. The ballistic limit velocity is found to be enhanced in the case of C1A1 CFRP laminate by 10.7% for spherical projectile impact and 19.9 % for conical projectile impact. Also, for the neat CFRP, the ballistic limit velocity is found to be 18.5% less for the conical projectile compared to the spherical projectile impact. A similar comparison revealed that the ballistic limit velocity is 11.8% and 18.4% lower in conical projectile impact compared to spherical projectile impact for C1A1 and C1A2 CFRP laminates. This implies that the 0.1 wt% MWCNT-COOH + 1 wt% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ nanofillers enhances the energy absorbing capability of CFRP composite panel and thus increases the ballistic limit velocity.

5.4.4. Damage Contour

In all the 24 experiments conducted, the projectile penetrated in 18 cases and rebounded in 6 cases. In this section, the numerical results obtained for perforated cases in the form of panel damage are presented along with counterpart experimental comparison. All panels showed visual damages grouped into three main categories: front-face damage, back-face damage and internal delamination. During the projectile impact phenomenon, a significant amount of axial stress is induced on the top layer of the laminate due to contact force. As the projectile impact continues, tensile failure along the fiber direction is dominant, resulting in a cruciate damage pattern that spreads radially. The damage pattern is not evident in the top layer. Instead, the top layer exhibits circular damage with the size of the projectile diameter, whereas, in the bottom ply, the cruciate damage pattern is more evident. All the forms of damage form are represented in Figure 16 and Figure 17 for a spherical and conical projectile, respectively.

The extent of the back face damage area due to spherical and conical projectiles obtained from the experimental and numerical studies is presented in Figure 18. The damaged area predicted from the numerical analysis is found to be within 10% accuracy compared with experiments for most of the cases validating the numerical approach. The numerical simulation shows a deviation of more than 10% in one case, which can be due to experimental variations. The damaged area is found to be the least for C1A1 samples subjected to spherical and conical projectile impact compared to neat and C1A2 CFRP panels for the same projectile indicating the enhanced energy absorbing capability and more ductile deformation to absorb impact energy by 0.1 wt% MWCNT-COOH + 1 wt% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ dispersed nanofillers laminates.

5.4.5. Damage Progressive Failure Modes

The progressive failure mode predicted for the spherical projectile and conical projectile impact at a velocity of 79.5 m/s and 96.8m/s for C1A1 samples is shown in Figure 19 for different time steps. From the Figure, it can be seen that the progressive failure is accelerated when impacted by the conical projectile, whereas in spherical projectile impact, the damage progress was comparatively slower, and the damaged area was larger based on the diameter. For a spherical projectile, the effect in velocity linearizes at about 0.12 ms, whereas for a conical projectile, the velocity linearizes at about 0.16ms, mainly due to the cylindrical extension at the end of the projectile.

During the impact event, the tensile and compressive waves due to the impact propagate across the thickness of the laminate. Since CFRP shows brittle behavior, highly localized deformation is observed for all the impact cases. The localized deformation progresses as the projectile penetrates the laminate, making the contacting plies take the impacting geometry’s shape. This is evident from Figure 20, where the bottom plies open up more for spherical projectiles when compared to conical projectiles. Finally, the bottom ply moves outward from the point of contact and reaches the diameter of the projectile.

5.4.6. Delamination Area

After the projectile impacts the specimen, it is examined using the C-scan ultrasonic method to determine the extent of the damage.

Table 4. Comparison of delamination area.

Panel Designation	Impact Velocity	Delamination Area (mm × mm)
		Experimental	Numerical
N_S1	79.6 m/s	28.06 × 27.10	27.52 × 26.07
C1A1_S1	79.5 m/s	22.60 × 22.03	21.77 × 21.03
C1A2_S2	79.8 m/s	25.95 × 25.29	24.71 × 24.06
N_C1	95.4 m/s	21.65 × 21.87	21.27 × 20.37
C1A1_S1	96.8 m/s	17.21 × 23.66	18.76 × 22.10
C1A2_S2	96.1 m/s	20.60 × 19.75	20.57 × 20.86

presents the delamination damage area obtained from experimental and numerical simulation for both spherical and conical projectile impacts at a specified velocity. In the numerical analysis, the delamination area is predicted based on the separation of the tiebreak contact. The delamination area is found to be less for C1A1 and C1A2 laminates compared to neat CFRP for spherical projectile impact. The decrease in delamination area for C1A1 and C1A2 samples when compared to neat CFRP laminates signifies higher energy absorption ability of the laminate due to the nanofillers in the matrix. Thus, we can conclude that with the addition of 0.1 wt% MWCNT-COOH + 1 wt%, the laminates absorbed more impact energy by crack bridging by the nanofillers, resulting in a reduced delaminated state of the laminate. A similar trend was observed for conical projectile impact, where the neat sample shows higher delamination, whereas the delamination area dimension is comparatively lower for both C1A1 and C1A2 samples. Thus, the nanofillers in the samples offer better resistance to projectiles piercing through the panel, preventing extended interlaminar failure compared to neat specimens.

5.4.7. Denting

In three scenarios, the impact energy was less than the energy needed to perforate the laminate, so the bullet rebounded during the experiment and dented the test specimen. In this impact event, some of the projectile kinetic energy is lost as a result of wave propagation, resulting in permanent denting, while the remaining energy facilitates the rebounding of the projectile. Figure 21 shows a comparison of experimental and numerical denting, with the spherical projectile impacting at 42.8 m/s for N_S1, 42.0 m/s for C1A1_S1, and 42.2 m/s for C1A2_S1, and the conical projectile impacting at 40.2 m/s for N_C1, 38.2 m/s for C1A1_C1, and 41.0 m/s for C1A2_C1. The denting of the plate in the numerical simulation is obtained from the total displacement by subtracting the elastic displacement. From the obtained data, it is observed that the addition of nanofillers resulted in a maximum reduction of 16% and 40% in dent depth when compared to neat samples for spherical and conical projectiles, respectively. The variation for the conical projectile is higher as the projectile impact velocity is comparatively lower to neat samples. Thus, it can be concluded that the addition of nanofillers had a positive effect on CFRP damage.

6. Conclusions

In this paper, experimental studies were carried out on 140 mm × 140 mm × 1.5 mm (thick) neat, 0.1 wt% MWCNT-COOH with 1 wt% and 2 wt% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ hybrid nanofiller CFRP panels subjected to spherical and conical projectile impact are presented. The numerical simulation of the same problem is carried out using LS-DYNA, a commercial non-linear finite element code. Based on the projectile impact experiments and numerical predictions, the following conclusions can be drawn:

• In comparison to neat CFRP laminates, hybridization of MWCNT and ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ nanofillers have a significant impact on energy absorption capability. Furthermore, the laminate with 0.1 wt% MWCNT-COOH + 1 wt% ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ has better performance.
• According to the experimental results, the panel has a rhombus-shaped failure at the back face and a pyramidal-shaped indentation after the projectile perforates the panel.
• In the case of the C1A1 CFRP laminate, the ballistic limit velocity is increased by 10.7% for spherical projectile impact and 19.9% for conical projectile impact.
• C-scan shows that the damaged area is larger than the visibly damaged area. When compared to neat CFRP, the damaged area for C1A1 and C1A2 samples was found to be lower, indicating that the nanofillers in the matrix provided additional reinforcement to absorb more energy.
• The FE model produced good results in terms of residual velocity, ballistic limit velocity, damaged area, and delamination as compared to the experimental data. The simulation method was able to capture the nanofiller hybridization influence on the impact behavior of CFRP laminate. The developed model can be further used for different nanofiller hybridized composite laminates under projectile impact loading.

7. Future Scopes

Future studies to better understand the behavior of CFRP laminate under projectile impact can focus on improving the following experimental and numerical approaches:

• Utilizing better experimental methods with improved data collection methodologies to understand the mechanism that governs the impact behavior of CFRP laminates. For example: utilizing a high-speed camera with 3D imaging capability to study the deformation in the laminate during the impact event.
• Another potential improvement can be introduced by using the meso-heterogeneous approach in the numerical simulation. Using such complex modeling mythologies with strain rate effect can effectively determine the failure sequence and methods to prevent such failure modes.

Possible practical applications based on enhancement in strength CFRP when reinforced with nanofillers are: CFRPs with CNT are commonly used in the wingtip fairings in the aerospace industry. Understanding the influence of nanofillers in these materials’ behavior under impact can help improve the design of aircraft structures to ensure safety during potential collisions. Also, CFRP is increasingly used in the automotive industry to make lighter and fuel-efficient vehicles. There is a need to understand how these materials behave under impact to improve safety during collisions. Using CFRP structures reinforced with nanofillers will eventually increase the impact load-bearing capability for these structures.