Atomic motion behavior calculation and bonding mechanism analysis of explosive welding of high-strength and high-hardness titanium alloy Ti6Al4V/aluminum alloy 7075

The explosive welding technology of titanium alloy Ti6Al4V/aluminum alloy 7075 is the most valuable research problem in the field of Ti/Al explosive welding. In this paper, the smooth particle hydrodynamics and molecular dynamics algorithms are used to conduct multi-scale numerical calculations on the explosive welding of Ti6Al4V and 7075. The interface morphology, interface atomic movement, lattice changes and crystal defects are comprehensively analyzed. With the calculation as a reference, the Ti6Al4V/7075 composite with a flat interface is prepared. The calculation results indicate that with the transformation of titanium aluminum crystal structure, a large number of various dislocations, mainly 1/6〈110〉dislocations, are generated on the titanium side; During explosive welding, uniform and stable element diffusion occurred. Element diffusion mainly occurred in the unloading stage when the pressure was 0. The melting zone and ablation zone were composed of TiAl and TiAl3, respectively; The fundamental reason for the combination of Ti6Al4V/7075 is the diffusion of elements and the solid-phase binding of atoms.


Introduction
Explosive welding is an advanced technology utilized to create layered metal composites that employs the explosive energy of explosives to achieve high-strength bonding between two or more similar or dissimilar metal plates/pipes through high-speed oblique collision [1][2][3].Both titanium alloys and aluminum alloys are significant lightweight metal materials with distinct advantages such as high specific strength and excellent corrosion resistance.Titanium alloys possess exceptional comprehensive mechanical properties, while aluminum alloy smelting technology is well-established.The advantages of these two materials complement each other, enabling Ti/Al explosive welding materials to be highly valuable in various fields, including aerospace, biomedicine, and marine equipment [4,5].
Yan Li [6] successfully prepared a TA1/1060 explosive welding composite material and observed that the interface waveform gradually developed from a flat shape to a waveform morphology as the shock wave propagated.HB Xia [7] and Conghui Zhang [8] further investigated the microstructure and mechanical properties of Ti-Al composites and observed the presence of recrystallization and twinning structures at the material interfaces.R Chulist et al [9] used EBSD technology to study the microstructure evolution of the Al/Ti/ Al interface, investigating the material deformation mode and annealing process.In addition to the organizational structure, the intermetallic compounds generated at the Ti-Al interface have also become a focus of scholars' research [10].There are various Ti-Al intermetallic compounds at the interface of composite materials [11].D M Fronczek [12] conducted a comprehensive and careful study of Ti-Al intermetallic compounds, demonstrating that four intermetallic compounds, TiAl 3 , TiAl 2 , TiAl and Ti 3 Al, are produced during titanium-aluminum explosive welding.The annealing process mainly leads to the growth of the continuous intermetallic compound TiAl 3 , but also to the growth of anomalous grains on the aluminum side and the annihilation of deformation twins on the titanium side.M Fan [13] further explored the effect of temperature on the intermetallic compound of titanium-aluminum explosive welding and proposed that the composite can exhibit excellent bonding properties at 550 °C.
Regarding welding technology, D V Lazurenko [14] performed a single explosive welding of 40 layers of pure titanium and pure aluminum plates, focusing on the microscopic morphology of the multilayer composite plate interface.Xunzhong Guo et al [15] explored the crack sprouting and expansion mechanism of TA1/1060 composites during tension.However, the low strength of the base material, with pure aluminum having a tensile strength of only 110 MPa and pure titanium having a tensile strength of only about 350 MPa, seriously limits the upper limit of the strength of titanium-aluminum composites.Achieving higher strength titanium and aluminum alloy explosive welding has become a key issue of current research.
To improve the strength of titanium alloys, Lucjan Sniezek [16] and Dariusz Boroński [17] conducted a experiment with and without sandwich AA2519/Ti6Al4V explosive welding, respectively.The composite plate prepared by the experiment was loaded under different conditions to test the durability as well as the strength of the composite specimens.Wu [18] realized explosive welding of TA2 and aluminum alloy 5083 by using pure aluminum as an interlayer, improving the overall strength of the composite by enhancing the strength of aluminum alloy.The strength of Ti6Al4V can reach 1012 MPa, and the strength of 7075 is also much higher than that of ordinary steel.The combination of the two directly solves the problem of low strength in explosive welding of pure titanium and pure aluminum.Achieving a high-strength composite of Ti6Al4V/7075 explosive welding is the most effective means to improve the comprehensive mechanical properties of titanium aluminum explosive welding materials.However, Ti6Al4V/7075 explosive welding is extremely difficult.The welding difficulty of titanium alloy and aluminum alloy increases with the strength of aluminum alloy.7075 is the strongest metal in aluminum alloy, and the weldability window is extremely narrow [19].The abundant magnesium and zinc elements in aluminum alloy 7075 can effectively improve the tensile strength of the material.However, magnesium and zinc elements are prone to evaporation and burning at high temperatures, resulting in defects in the composite material and stress concentration, reducing the strength and fatigue resistance of the material [20].Compared to other aluminum alloys, 7075 exhibits poor plasticity and is prone to microcracks or brittle failure under the pressure generated by collisions.
In order to achieve high strength bonding of difficult to weld metals by explosive welding, numerical calculation has become an effective means.Distinguished by simulation size, numerical calculations can be divided into finite element methods above the micron level and molecular dynamics algorithms below the micron level.The finite element method is divided into four formations: Lagrange [21], Euler [22], ALE (Arbitrary Lagrange-Euler) [23], and SPH (Smooth Particle Hydrodynamics) [24].The Lagrange, Euler, ALE algorithms can calculate the deformation, collision pressure, temperature and other data of the base plate during the welding process, providing reference in terms of weldability window, process parameters, and other aspects [25,26].However, the SPH algorithm has been proved to be a more effective method for calculating the severe plastic deformation, peeling off and other behaviors of materials during explosive welding.RuifengLiu [27] and Yasir Mahmood [28] observed explosive detonation and metal jet phenomena that are difficult to observe in experiments when using the SPH algorithm for numerical calculation of explosive welding.IA Bataev [29] used the SPH method to study the formation and motion of interface re-entry jets, and proposed a new model for interface waveform growth.HLA B [30] discussed the JH-2 (Johnson-Holmquist-Ceramics) constitutive model using the SPH method and observed interface waveform structures consistent with the test results of composite materials.Recently, molecular dynamics (MD) algorithm has also shown great application value in the field of explosive welding.The paper [31] comprehensively analyzed the enormous application prospects of molecular dynamics in the field of connectivity.Yong Ma [32] used molecular dynamics to discover the atomic diffusion behavior in the explosive welding process, and SY Chen [33] and Feng [34] used molecular dynamics algorithm to discuss the distribution of the thickness of the element diffusion layer and the atomic concentration.Numerical calculation has become an important means to study explosive welding technology, which is of great significance to understand and analyze explosive welding process.However, there are still some problems in the numerical calculation of explosive welding, such as the numerical calculation does not correspond to the experiment parameters, and the numerical values on the two scales of SPH and MD algorithm have not been effectively combined.
In reference [27], although the authors obtained good interface waveform morphology during numerical calculations of titanium aluminum magnesium three-layer explosive welding, the explosive thickness set during simulation was only 2 mm.In actual experiments, this far falls short of the minimum detonation thickness of the explosive, and there is a significant difference between simulation and experiment.
Explosive welding is widely accepted as a solid phase metallurgical technology, relying on the atomic gravity of two materials at the interface to form a solid metallurgical bond.However, due to the physical performance of high temperature and high pressure at the interface in the process of explosive welding, in addition to the solid phase bonding of interface atoms, many scholars still have a big dispute about whether pressure welding and fusion welding occur in the process of explosive welding [33,35].
To address the aforementioned issues, this thesis integrated the SPH and MD molecular dynamics algorithms for the first time to comprehensively calculate the micro morphology, temperature and pressure changes, and atomic motion behavior of the Ti6Al4V/7075 explosive welding interface.The atomic diffusion behavior, crystal changes, and crystal defect growth during the welding process were analyzed using mean square displacement (MSD), common neighbor analysis (CNA), and defect analysis (DXA).The radial distribution function (RDF) was utilized to determine the phase state of the base plate and flyer plate at each stage of explosive welding, and the bonding mechanism of the Ti6Al4V/7075 explosive welding was analyzed based on the experiment results.The successful preparation of the Ti6Al4V/7075 composite plates with extremely high welding difficulty pioneers a new approach based on SPH and MD algorithms for analyzing, preparing, and experimenting the combination of difficult-to-weld metals.

Theoretical calculations and experiments
2.1.Smooth particle hydrodynamics algorithm 2.1.1.Fundamentals of SPH algorithm During finite element calculations, the Smoothed Particle Hydrodynamics (SPH) algorithm does not require the division of any grid.Instead, the entire structure is discretized during the calculation process, without the need for particle connections.This approach effectively mitigates structural deformation distortion, convergence failure, operational distortion, and other issues commonly associated with grid-based methods [36].The SPH algorithm mainly involves two steps to achieve particle based calculation of macroscopic physical variables such as temperature, pressure, and density of the overall structure.Firstly, using the integration method to approximate the field function, the integral expression for f (x) is: Where x is the particle to be solved, x′ is the space coordinate of the particle interacting with the particle x to be solved, Ω is the solution decision region, and h corresponds to the smooth length in the finite element method.δ(x-x′) is the Dirac kernel function, when x = x′, the function value is 1; The function is 0 for x ≠ x′.
Using smooth function W to approximate kernel function δ, then f (x) can be approximated as: Taking the derivative of equation (2), then∇f (x) is: Where, the smooth function W must satisfy the normalization condition listed in the following equation: Where k is the scale factor.When the smooth length h approaches 0, the smooth function W is a kernel function δ, The kernel functions commonly employed in smoothed particle hydrodynamics (SPH) simulations include bell-shaped, Gaussian-shaped, and cubic spline functions.Among these options, cubic spline functions are the most frequently used.
Secondly, the particle approximation method is used to further approximate and optimize the kernel approximation equation.By weighting and summing the numerical values of a finite number of particles with independent masses and spatial coordinates, the integral and derivative results of the entire system are approximated.Equation (2) and (3) become: Where m j , ρ j is the mass and density of the jth particle, respectively.By combining the conservation of mass, momentum, and energy in continuum mechanics, the corresponding SPH algorithm form can be obtained:

. Strength equation and state equation
During explosive welding experiments of metal materials with high hardness and poor plasticity, adiabatic shear lines are often observed at the interface of composite materials.This phenomenon is characterized by dislocation defect accumulation when the strain rate exceeds 10 6 /s [37], indicating that the strain rate of materials during explosive welding can reach this level.Explosive welding is a dynamic process involving high temperature, high pressure, and high strain rate, resulting in significant plastic deformation of materials.Accurately describing the location, density, energy, stress, and strain of materials is crucial for numerical calculations.Currently, the Johnson-Cook and Steinberg-Guinan strength equations are widely used in numerical simulations of the impact behavior of metal materials.The Johnson-Cook equation is suitable for calculating the mechanical behavior of materials at low temperatures and high strain rates, while the Steinberg-Guinan equation is more appropriate for simulating the nonlinear mechanical behavior of materials at high temperatures and high strain rates.When the strain rate of the material exceeds 10 5 /s, the Steinberg-Guinan equation can more accurately predict stress, strain, and strain rate [38].Therefore, the Steinberg-Guinan strength equation was employed for the numerical simulation of explosive welding using smoothed particle hydrodynamics (SPH) of Ti6Al4V/7075.The expression for the Steinberg-Guinan intensity equation is given by: Equations (11), (12) applies to Y 0 [1+Bε]n Ymax, where G is the shear modulus, Y is the yield stress, ε is the effective plastic strain, β is the hardening constant, n is the hardening index, T is the temperature.η is the relative volume, Y 0 , G 0 , G´P , G´T and P are constants, when the temperature exceeds the material melting point, the material shear modulus and yield stress is defined as 0.
The Shock equation of state is a widely used equation that describes the dynamic behavior of materials under the influence of shock waves.This equation is particularly applicable in extreme conditions such as high-speed shocks and explosions, and is the most suitable equation of state for numerical calculations of explosive welding.The Shock equation of state defines the relationship between the density (ρ), pressure (p), energy (e), mass velocity (U P ), and impact velocity (U) of materials in the shock state.

( ) U C S U 13
Where S′ is the Hugoniot slope, C 0 is the speed of sound of the material, Γρ = Γρ 0 is a constant, Γ 0 is the Gruneisen parameter, μ = ρ/ρ0 − 1, ρ is the material density and the initial density ρ 0 .Ti6Al4V and 7075 related parameters have been extensively studied and calculated, and all parameters have been embedded in the simulation calculation program as shown in table 1:

High-speed tilt collision model
During the explosive welding process, the explosive blast generates a significant blast pressure that bends and deforms the composite plate, causing it to collide with the lower base plate at a collision velocity (V p ) perpendicular to the angular bisector of the collision angle (β) (figure 1).To simulate the explosive welding behavior, a high-speed tilted collision model is utilized based on the aforementioned principle.Unlike direct modeling of the explosive welding process, the high-speed tilt collision model omits the explosives from the numerical calculation.Instead, the kinetic energy of the composite plate reflects the role of explosives in the system, reducing the complexity of the numerical calculation.This approach also helps to mitigate the impact of numerical calculation errors caused by inaccuracies in the explosives parameter settings.The thickness of the model-based composite plate is maintained consistent with the experiment, and the collision velocity (Vp) and collision angle (β) corresponding to the experiment parameters are determined using equations (17) and (18) [11]: 4 18 e e e e f f e e f f 2 Where, ρ e for the explosive density, δ e for the explosive thickness, ρ f for the composite plate density, δ f for the composite plate thickness, γ for the explosive effective multi-party index, V d for the explosive burst speed, thus establishing a direct link between the numerical calculation and the experiment process parameters.In the parallel method of explosive welding, the collision point moving speed Vc is equal to the burst speed V d , the process parameters are initially calculated by the explosive welding weldability window β-Vc, the weldability window within the selection of suitable process parameters, numerical calculations to obtain a good bonding effect after the experiment.
Deribas stated that the impact pressure of the base plate and compound plate collision should exceed the dynamic strength of the material, thus determining a lower limit for the velocity of movement of the collision point [39].
Where Hv is the Vickers hardness of the softer material in the base plate and compound plate, k1 is used to characterize the cleanliness of the material surface, with a value range of 0.6 (clean) to 1.2 (contaminated), and 0.85 in general.
Carpente [40] determined the upper limit of explosive welding using equation (20) in order to avoid the reflected back stress waves from affecting the interface bond.
Where T MP is the lower melting point, k is the thermal conductivity, C p is the constant pressure heat capacity, N is a constant, generally taking the value of 0.11.
The explosive blast velocity should be lower than the volumetric speed of sound of the welded material [2], which determines the right limit of the explosive welding weldability window, the volumetric speed of sound of the metal is calculated by Equation (21), the smaller value of the volumetric speed of sound of the base plate and replica plate material is the right limit of the explosive blast velocity.
where G is the Young's modulus of the metal.The left limit of the weldability window considers the generation of material jets [41], determines that the range of collision angles should be 2 to 31°, and gives a lower limit of the collision velocity (equation ( 22)).
Where R b is the tensile strength, and the larger value of the calculation results of the base plate and composite plate is selected as the left limit of the collision point movement speed.
By combining equations ( 19)-( 22), the weldability window is calculated as shown in figure 2. Only the shaded area in the figure 2 is the weldability area of Ti6Al4V/7075, which further illustrates the difficulty of combining the two materials.
It was pointed out above that the Mg-Zn elements in 7075 tend to ablate and evaporate at high temperatures, which affects the material bond strength, so it is more appropriate to select parameters near the lower limit of the Ti6Al4V/7075 weldability window.After several SPH numerical calculations, the ideal bonding interface can be obtained when the collision angle β is 11°and the velocity of movement of the collision point Vc is 2800 m s −1 , and the experiment parameters are further determined based on this result.

Molecular dynamics algorithm 2.2.1. Model and potential function
Interfacial behavior is the most central issue in the study of composite materials.The explosive welding process generates tens of thousands of megapascals of high pressure at the interface as well as thousands of degrees Celsius, and under the action of high temperature and pressure, the interfacial atoms exhibit behavior different from that at room temperature and pressure.Molecular dynamics calculations are one of the most effective The open source LAMMPS software [42] was used to build the molecular dynamics simulation model shown in figure 3: the upper layer of the model is the aluminum atomic region and the lower layer is the complex plate titanium atomic region.As a typical isomeric metal, titanium behaves as α-Ti with a dense hexagonal lattice (HCP) structure at low temperatures; when the temperature rises above 882 °C, titanium transforms into β-Ti with a body-centered cubic (BCC) structure.In the field experiment of explosive welding, the pre prepared material temperature is generally room temperature, so the close packed hexagonal structure of titanium at room temperature is selected to describe titanium in molecular dynamics modeling.The lattice constant of titanium is 2.95 Å, and aluminum has a face centered cubic FCC structure with a lattice constant of 4.05 Å. Considering the scale effect, in order to enable a finite number of atoms to exhibit the macroscopic physical and chemical state of the material.The model is set as a periodic boundary in the x and y directions, a shrinkable boundary in the z direction, and a transition zone with three atomic layer thicknesses is set at the upper edge of the Ti atomic region and the lower edge of the Al atom, respectively, to avoid adverse effects of shock waves on the calculation.A vacuum zone with a distance of 30 Å is set between the titanium aluminum atomic regions, and the width of the vacuum zone is greater than the cutoff potential of the interaction between titanium and aluminum atoms.Due to the disparity between the crystal structures and sizes of titanium and aluminum, when determining the modeling dimensions, the x,y periodic boundary dimensions are chosen to be integer multiples of the dimensions of the two atomic lattices in the corresponding directions to avoid serious structural defects such as atomic overlap and atomic missing at the periodic boundaries, which affect the stability of the system and the calculation accuracy.dimensions in the x,y direction are 64.8 × 61.3 Å.The system is constructed with a total of 51640 atoms, and the colliding crystal surfaces are aluminum (001) and titanium (0001) surfaces.
Molecular dynamics simulation follows the constraints of inter particle forces and Newton's second law, and selecting appropriate potential functions to describe the interactions between particles is crucial for molecular dynamics calculations.Among the numerous potential functions that describe inter particle forces, the embedded atomic potential function (EAM) developed by Daw and Baskes [43] in 1984, It has a significant accuracy advantage in molecular dynamics calculations of metal systems.The EAM expression is (equation ( 23)).
Where Φ ij is the repulsive energy between atoms i and j, F i ( ρ i ) Indicates that the atom i embedded in the free electron density is ρ The energy required in the air mass.To accurately describe the atomic forces between Ti-Al, the Ti-Al intercalation potential function developed by Zope [44] was selected for calculation.

Molecular dynamics calculation process
After the model is built, the structure is energy minimized to reduce the energy of the system to a lower level, which initially eliminates the problems of overlapping atoms and too close distances caused by the modeling.In order to make the molecular dynamics calculation correspond well with the field experiment of explosive welding, it is necessary to design the molecular dynamics calculation process reasonably.Research [33] divides explosive welding molecular dynamics simulation into loading stage, unloading stage and cooling stage.In this paper, combined with the results of SPH algorithm, explosive welding molecular dynamics calculation is further divided into the following six stages: Initial stage: After model construction and minimization, the system relaxes in the NPT ensemble, initializing the model temperature to room temperature of 300 K, further eliminating modeling errors and making the system more reasonable and stable.
Loading collision stage: After relaxation is completed, the system ensemble is replaced with a microcanonical ensemble (NVE), and the number of atoms, total volume, and total energy of the system during the loading stage are controlled to remain unchanged.Fix the Al atom transition zone, apply a z-axis velocity of 1600 m s −1 and an x-axis velocity of 1000 m s −1 to the Ti atom zone, and make the Ti atom zone hit the Al atom zone until the system volume shrinks to the minimum value.This stage corresponds to the moment when the flyer plate hits the base plate in the field experiment of explosive welding.
Post collision relaxation stage.Maintain the NVE ensemble, fix the atoms at both ends of the system under the high temperature and pressure generated by collision, and perform a 200 ps relaxation of the system.This stage corresponds to the state that the system maintains extremely short high pressure under the continuous action of detonation pressure and detonation wave after the flyer plate hits the base plate during the explosive welding experiment.
Unloading stage: Under the NPT ensemble, maintain the end temperature of the loading time, and relax the system for 200 ps under 0 external pressure.This stage corresponds to the release of the high-pressure state after the detonation wave during the explosive welding experiment, and the continued maintenance of the hightemperature state.
Cooling stage: under the NPT ensemble, the system temperature is lowered to room temperature at a certain rate.Relevant research [34] indicates that the physical correctness is still maintained when the cooling rate is between 1012 and 1014k/s.This article selects a speed of 1012k/s for cooling.During explosive welding experiment, due to factors such as metal heat conduction and air heat dissipation, the temperature generated by collision will drop at an extremely high speed at the interface, which corresponds to this stage.
Room temperature relaxation stage: After falling to room temperature, under the initial condition of 0 external pressure, the system relaxes for 200 ps until the crystal structure of the system is stable.
The above six stages can be used for complete molecular dynamics calculation of explosive welding experiment.By analyzing and processing the calculation results, the atomic motion behavior in the explosive welding process can be effectively analyzed.

Interface MSD analysis and RDF analysis
The diffusion coefficient reflects the ability and degree of mutual penetration between atoms, and is an important reference for analyzing the diffusion behavior of atoms in composite materials.In his study of Brownian motion, Einstein concluded that the average square of the moving distance of randomly moving particles is proportional to time.The diffusion coefficient D is: Where d is the dimension of the moving system, and when the material is a cubic lattice, d is equal to 3 due to the equivalence of the diffusion of atoms in all directions.ri denotes the spatial position of the particle at moment i, and the symbol '〈 〉' is averaged over all atoms in the selected range.And the diffusion coefficient is closely related to the calculation of the mean square displacement of the particle, which is calculated as shown below: Comparing equations ( 24) and (25), it can be seen that in the MSD curve graph with time as the horizontal axis, 1/6 of the slope of the MSD curve of cubic lattice materials is the diffusion coefficient of atoms.Molecular dynamics algorithms can easily obtain the coordinate values of particles at any moment of motion, calculate the mean square displacements of various categories of atoms in the system, and further obtain the diffusion coefficients of corresponding atomic types.In addition to the cubic crystal, even if the atoms of materials with other crystal structures have different diffusivity in all directions, the atomic diffusion coefficient in a certain direction can be calculated by substituting the coordinate value in a specific direction, so as to judge the diffusion degree of the interface atoms in the explosive welding process.
The interface bonding mechanism has always been the focus of controversy in explosive welding technology.By calculating the atomic radial distribution function, the interface phase state in the welding process can be determined.By determining the crystal structure of materials at different times, it provides a reference for the determination of explosive welding bonding mechanism.The radial distribution function g (r) is used to describe the distribution probability of other particles in the local time away from the specific particle r, and the expression of the radial distribution function is equation (26): Where ρ 0 is the distribution density of all atoms in the system and the size of ρ 0 is ρ(r) is the distribution density of particles at a distance r from the reference particle.If the total number of particles in the spherical shell at a distance from the specified atom r to r + Dr is dN, then g(r) can be calculated by equation (27): The radial distribution function can be used to describe the correlation of electrons and the order of particles, so it can be used to judge the arrangement state of atoms during explosive welding.Figure 4 shows the structure of the ordered solid metal atoms.Because the metal atoms in the solid state are arranged in a long range order, with the increase of the atomic distance r, g (r) will continue to appear an obvious peak; In contrast, the material in the liquid state has the characteristics of short-range order and long-range disorder (see figure 5), so after the radial distribution function has an obvious peak at the first nearest neighbor, the peak value decays rapidly with the increase of r, and finally tends to 1 (remotely close to the ideal gas distribution) The physical interpretation of radial distribution function is shown in figures 4 and 5.

Ti6Al4V/7075 explosive welding experiment 2.3.1. Experiment materials
The composition and physical properties of the materials used in the experiment are shown in tables 2 and 3.Both Ti6Al4V and 7075 have high strength and hardness, which has been proved to be the main reason for the difficulty of explosive welding of the two.In the explosive welding experiment, the design of rectangular plate is obviously better than that of square plate under the same area.On the one hand, the short side of rectangular plate is more conducive to the emission of interface gas in the explosive welding process, improving the interface bonding quality; On the other hand, compared to square plates, rectangular plates have longer long edges, providing longer distances for stable detonation of explosives and making it easier to obtain composite plates with stable bonding quality.The base plate and flyer plate used in the experiment are both 200 mm × A 300 mm rectangular plate with a flying plate thickness of 3 mm and a base plate thickness of 5 mm.According to the weldability window and SPH numerical calculation results, in order to increase the controllability of explosive welding and improve the welding quality, 46 # low-speed powdery emulsion ammonium nitrate explosive mixed with quartz sand and a small amount of wood chips was finally used for welding.The explosive density was about 0.9 g cm −3 , the detonation speed was about 2800 m s −1 , and the thickness was 35 mm.

Experiment process
During the collision process between the base plate and the flyer plate, a significant amount of metal jet is generated, which washes the bonding surface and plays a crucial role in cleaning it.However, to achieve higher quality bonding, it is necessary to polish the bonding interface between the base plate and the flyer plate during explosive welding to reduce the impact of impurities and unevenness on the welding quality.The high brittleness of 7075 makes it susceptible to damage due to excessive stress during the collision process caused by explosions.Therefore, soft sand was used as the foundation during the experiment, and non-woven fabrics were placed to cushion it.The flatness of the material is also an essential factor affecting the bonding quality.To support the flyer plate and the explosives laid on it, a 'V' shaped pure copper sheet is placed on the upper layer of the base plate, ensuring that the flyer plate does not deform due to the weight of the explosives.Most copper sheets will be blown away from the interface by the shock wave generated by the explosion without affecting material bonding.A small amount of residual copper sheets at the interface will also not affect the overall bonding quality due to their good metal compatibility.After arranging the explosives, detonators are placed at the midpoint of the short edge of the composite board, and they are detonated (figure 6).

Microscopic testing and characterization
Cut the composite board into 10 pieces × A 10 mm sized test block is polished on the side and corroded with 5% HNO 3 +10% HF+85% H 2 O. Conduct SEM micro morphology testing and EDS element scanning on the sample, analyze the micro morphology and element distribution at the interface, and analyze the bonding quality and mechanism of Ti6AL4V/7075.When molecular dynamics calculations were performed, after entering the unloading phase, the system temperature was set to maintain the high temperature conditions generated in the loading phase and the system pressure was set to 0 external pressure.Figures 7 and 8 show that the interface remains at the high temperature generated by the collision for a period of time after the collision between the base plate and the flyer plate, while the interface pressure decreases rapidly until it reaches 0. The results of the SPH temperature and pressure calculations indicate to some extent that the temperature and pressure settings of the system at each stage during the molecular dynamics calculations of the paper are also reasonable.
Figure 9 shows the numerical calculation results of SPH in explosive welding at different times.The results indicate that the interface does not generate an obvious waveform morphology, and the interface is flat and straight as a whole.However, metal materials with similar physical and chemical properties are more likely to  produce waveform interfaces during explosive welding, and the waveform size is related to parameters such as explosive parameters and gap for explosives.In general, in metal explosive welding experiments, the straight and wavy interface has fewer interface defects, and many scholars consider it to be a better interface.Metal jet is an important guarantee for high-quality combination of explosive welding [3].Further observation shows that only a small amount of metal jet can be observed at the interface when t = 2.4us.As the calculation continues, the number of metal jets at the interface gradually increases.
The interface temperature nephogram shows (figure 10) that during explosive welding, there is a strip of high temperature zone at the entire interface where the impact point passes.Compared with Ti6Al4V, aluminum alloy 7075 has higher thermal diffusion coefficient and thermal conductivity, so the width of the high temperature zone on the aluminum side is significantly wider than that on the titanium side.
Figure 11 shows the cloud diagram of the interface pressure distribution at different times.During explosive welding, an approximately circular high-pressure action zone was generated at the interface.Consistent with the temperature distribution rule, the area of the high pressure zone on the aluminum plate side is significantly   larger than that on the titanium plate side.Impact pressure is transmitted in the form of stress waves in metals.In an ideal elastic state, the transmission speed of stress waves in metals is similar to the volume sound velocity.Although the volume sound velocity of aluminum alloy 7075 is slightly higher than Ti6Al4V, it cannot produce the comparative effect shown in figure 12.The reason for this is that although the pressure at the interface is much lower than the elastic modulus of the two materials at room temperature, the interface material has undergone plastic deformation under the influence of collision temperature.Under the influence of temperature and pressure, the density and crystal structure of the interface material undergo significant changes, resulting in a significantly faster transfer speed of stress waves on the aluminum side than on the titanium side.

Molecular dynamics calculation results
The temperature and pressure changes of the system during the loading and loading relaxation stages are shown in figure 12, and the original temperature of the system is the room temperature finally controlled by relaxation.When conducting molecular dynamics simulations, the initial velocities applied to all titanium atoms are macroscopically represented as the system's temperature.Therefore, the initial value of the temperature curve shown in figure 12 is 3500 °C.After the collision, the system gradually stabilized, and the average temperature remained at around 1250 °C.The initial pressure of the system was 0, but after the collision, the system pressure rapidly increased and eventually stabilized at 22 GPa after a relaxation time of 200 ps.The temperature and pressure values of the system obtained by molecular dynamics simulations are consistent with the values obtained by finite element calculations for explosive welding, and the two calculations have mutually confirmed each other.Figure 13 shows the common neighbor analysis (CNA) results obtained from molecular dynamics simulations.After the collision (figure 13(a)), the atoms at the interface exhibit a clearly disordered fluid state, while the titanium atoms away from the interface maintain their original hexagonal close-packed (HCP) structure.The stable face-centered cubic structure of aluminum partially transformed into a body-centered cubic structure, which has been reported to occur only under extremely high-pressure conditions [45]. Figure 13(b) shows that after a period of relaxation with the titanium and aluminum positions fixed at the end of the collision, the crystal structure of the entire system undergoes a significant change.Only a small amount of HCP structure remains on the titanium side, replaced by the body-centered cubic and face-centered cubic structures that appear at high temperatures.The temperature increase makes the HCP structure of titanium unstable, and the crystal structure transforms into a more stable face-centered cubic structure.A similar  phenomenon was observed by A. Aguayo [46] in the study of titanium crystal structure.The transition from close-packed hexagonal to face-centered cubic structure causes the titanium volume to expand by about 4%, and the change in crystal structure also alters the physical and chemical properties of titanium.The more stable face-centered cubic structure makes titanium perform well at high temperatures, with higher toughness and ductility.Figure 13(c) shows that after unloading, aluminum completely transforms into a disordered fluid structure, and titanium begins to precipitate HCP and face-centered cubic crystal structures.During the cooling phase, as the temperature decreases, titanium almost completely transforms into the more stable HCP and facecentered cubic structures, while aluminum still maintains a disordered fluid state due to the limitations of the simulation time (figure 13(d)).
Dislocations are common defects in crystalline materials, and their form, density, and relationship with the mechanical properties and failure mechanisms of materials are closely related.Therefore, investigating the types of dislocations, their generation mechanisms, and their ability to slide or climb during explosive welding is of great significance for the study of materials used in explosive welding.Figure 14 shows the results of dislocation analysis for each stage of the system.The initial stage of the system is an ideal crystal structure, and after sufficient relaxation, there are no internal crystal defects such as dislocations.During explosive welding, due to severe plastic deformation, the system has a large number of 1/6〈112〉Shockley incomplete dislocations.Even in the final cooling stage, there are still a large number of 1/6〈112〉dislocations in the titanium that has basically recovered its solid crystal structure, indicating that serious plastic deformation has occurred on the titanium side, and a large amount of internal stress still exists after the material is hardened (figure 14(e)).Figure 14(a) shows that during the loading stage, 1/2〈110〉 and 1/6〈110〉 dislocations appeared.The 1/2 〈110〉 dislocation is an effective dislocation for transmitting and absorbing stress, which can cause plastic deformation and structural damage in the material.The 1/6〈110〉 dislocation is a typical shear dislocation, which is generated due to the orderly atomic displacement caused by plastic deformation of the metal.During the relaxation stage after collision (figure 14(b)), the huge internal stress generated by the collision caused severe atomic displacement in the system.However, restricted by the high-pressure environment, the stress in the system cannot be effectively released, and the movement of atomic displacement to the correct position is also limited.Therefore, a large number of dislocations with short length and high quantity were generated.Due to the melting of aluminum into an amorphous structure, no dislocations were observed on the aluminum side during the unloading and cooling stages (figures 14(c)-(e)).During the cooling stage (figure 14(d)), the number of dislocation lines on the titanium side decreased and the length became longer, indicating that the plastic deformation of the material was more severe than in the previous stage.As the temperature decreases, the stress generated by plastic deformation increases, and the activity of dislocations increases.Compared to short dislocations, the displacement slip ability of long dislocations increases, which directly leads to a decrease in material strength and cracks or failure.In addition, the long dislocations shown in figure 14 are also the fundamental cause of adiabatic shear line [37].During the analysis of dislocations in the system, the presence of 1/3〈110〉 dislocations was also observed.It is worth mentioning that these dislocations often occur with the formation of intermetallic compounds.

Experiment results
Samples were taken at equidistant intervals along the detonation direction, and the SEM results of the interface are shown in figure 15.The numerical calculation results show that the interface temperature and pressure gradually increase as the detonation reaction progresses.Consistent with the numerical simulation results, the interface waveform becomes more pronounced as the detonation reaction progresses.The SEM test results are consistent with the simulated interface morphology results: the interface is mainly composed of a flat shape and a small wave shape, and a typical vortex structure is generated at the interface in the explosive welding experiment.In addition, a large amount of β-Ti was detected in the sample.This is because a large amount of aluminum elements at the interface promote the generation of β-Ti during collision at high temperatures, and the extremely high cooling rate of explosive welding preserves the β-Ti.
Further observation of the interface morphology of the three samples (figure 16) revealed three typical morphologies of Ti6AL4V/7075 explosive welding: near the explosion point (figure 16(a)), the interface exhibits a flat morphology with good bonding quality and no defects such as pores or casting structures.Partial circular structures are present on the aluminum side, which is caused by local disturbances during bonding, as pointed out in reference [47].As the explosion distance increases, continuous melting regions appear on the interface, mainly on the aluminum side (figure 16(b)).Numerical simulation results indicate that this is due to prolonged high-temperature regions with large widths on the aluminum side (figure 16(d)).At positions far from the explosion point, the interface has higher wave heights and more obvious waveforms, but also shows signs of erosion (figure 16(c)), which to some extent affects the bonding quality.As mentioned earlier, this is caused by evaporation erosion of magnesium and zinc elements in aluminum alloy 7075 at high temperatures.
Element point scanning tests were conducted on the continuous melting zone (L1, L2 positions) and burn loss zone (L3 position) using EDS to further analyze the corresponding element distribution and phase composition.The element testing results of the melting zone are shown in figure 17.The atomic weight of aluminum in the melting zone is slightly higher than that of titanium, and the ratio of the two elements is close to 1:1.Based on the atomic content, it can be inferred that the main intermetallic compound between titanium and aluminum in the melting zone at the interface is TiAl.Figure 18 shows that the aluminum atomic content in the burning loss zone is about three times that of titanium.Based on the content distribution, it can be inferred that the main intermetallic compound in this area is TiAl 3 .Both TiAl and TiAl 3 are typical brittle intermetallic compounds.The continuous intermetallic compound layer provides a certain bonding energy for the composite material, while the intermittent brittle intermetallic compound destroys the uniformity of the composite material, forming defects in the entire composite material, making it prone to crack and failure at room temperature.
The EDS surface scanning results of the composite plate (figures 19(c) and (d) show that titanium is only distributed on the flyer plate side, and almost no titanium elements are detected on the base plate side.Similarly, the element line scanning results show that there is diffusion depth of aluminum elements above 35um on the flyer plate side, with an atomic quantity ratio of over 15%, and no titanium elements are present on the base plate side (figure19(b)).The combined results of line scanning and surface scanning confirm that the diffusion of aluminum elements on the titanium side is uniform and continuous, and atomic diffusion during explosive welding mainly occurs in the form of aluminum atoms spreading towards the titanium side.

Analysis of interface atomic diffusion behavior 3.2.1. MSD results and analysis
The calculated results of atomic MSD at each stage are shown in figure 20.During the loading collision stage, the MSD values of both types of atoms increase rapidly.The reason for the increase in MSD values during this stage can be explained by the fact that when the titanium atom region collides with the titanium atom region at a certain initial velocity, both types of atoms undergo a large displacement, resulting in a rapid increase in MSD values.During the relaxation stage after the collision, the MSD values of the atoms remained at a high level as at the end of the loading collision stage.However, due to the fixed boundary conditions at both ends of the system,  the overall volume of the system cannot change, resulting in a horizontal trend in the MSD curve during this stage.The slope of the curve is close to zero, indicating that the diffusion coefficients of both types of atoms during this stage are also close to zero.The slightly higher MSD values of titanium atoms compared to aluminum atoms are due to the fact that as the provider of the overall system velocity, titanium atoms vibrate more violently during this stage.During the unloading stage, the MSD values of both types of atoms continue to increase, and the diffusion coefficients of both types of atoms reach their maximum values during the entire calculation period.Further observation of the graph shows that the growth rate of the titanium atom MSD curve slows down, indicating that the diffusion coefficient of titanium is gradually decreasing, whereas the growth rate of the aluminum atom MSD curve continues to increase, indicating that the diffusion coefficient of aluminum is still increasing during the unloading stage.After unloading, the MSD values of titanium atoms remain close to zero  in the following two stages, indicating no further diffusion.The MSD values of aluminum atoms during the cooling stage first increase and then stabilize, indicating that aluminum atoms continue to diffuse during the initial stage of cooling.The different behaviors of the two types of atoms during the cooling stage may be related to the huge difference in their melting points.The initial temperature during cooling is still higher than the melting point of aluminum, which gives aluminum atoms some diffusion capability.
The MSD calculation analyzed the diffusion coefficients of two types of atoms at each stage, providing a reference for studying the diffusion behavior of interface atoms.However, due to the presence of a vacuum region in the early stage of the system, atoms can diffuse towards the vacuum region.In addition, the presence of factors such as changes in system volume means that the diffusion coefficients determined by MSD calculations can only prove that atoms have undergone some diffusion behavior in space.Further analysis is required to determine whether diffusion occurs between atoms and the degree of diffusion, which can be achieved by combining atomic distribution images and atomic positions.

Atomic distribution image analysis
The atomic distribution of the system at each stage is shown in figure 21.Since the model is set with periodic boundaries in both the x and y directions, the x direction is displayed with twice the periodic length to enhance observation.Figure 21(a) shows the atomic distribution of the system when titanium and aluminum collide and the system volume is at its minimum.The collision generates high temperature and pressure, causing the system length to be significantly compressed in the z direction.The two types of atoms are interlaced at the titaniumaluminum interface.Figure 21(b) shows the atomic distribution after a 200 ps relaxation period under the high temperature and pressure generated by the collision.After a period of relaxation, the interface atoms become more irregular in shape, indicating a tendency for atomic diffusion.However, the diffusion phenomenon between atoms cannot be clearly observed from figure 21(b) alone.Figures 21(a) and (b) indicate that the high temperature of the system during the loading and relaxation stages increases the intensity of atomic motion, providing energy for atoms to leave their original positions.However, the high pressure generated by the collision reduces the distance between atoms in the entire system, especially at the interface, making it more difficult for atoms to leave their original positions and increasing the difficulty of diffusion.Combined with the MSD results, this stage can be understood as the collision causing all atoms in the system to gain greater motion speed.However, due to the high pressure, atoms can only oscillate in limited positions and cannot undergo significant displacement diffusion or vacancy diffusion.Therefore, although the interlaced morphology of interface atoms can be observed, the mutual diffusion of atoms cannot be clearly observed.Figure 21(c) shows the atomic distribution of the system at the end of the unloading stage.Compared with the previous two stages, the high pressure that constrains atomic motion disappears, and the system volume recovers and stabilizes.The distance between atoms increases, and the energy barrier required for atoms to leave their original positions decreases, making it possible to clearly observe the mutual diffusion of interface atoms.Figure 21(d) shows the atomic distribution of the system during the cooling stage.Compared with the unloading stage, the concentration of atomic diffusion slightly increases when the system temperature drops to room temperature, but the maximum diffusion distance of atoms does not change significantly.The atomic morphology after the cooling and relaxation stage is shown in figure 21(e), which is almost unchanged compared to the previous stage.

Atomic coordinate analysis
To further analyze the diffusion behavior at the explosive welding interface, this paper further processed the atomic coordinate data obtained from molecular dynamics calculations and quantitatively analyzed the diffusion thickness of interface atoms and the change in the median plane position of atoms.Since the mutual diffusion of titanium and aluminum atoms mainly occurs in the z-axis direction, the z-coordinate values of all atoms in the five-stage system were exported.Referring to the EDS element line scanning method used in the experiment analysis, Python was used to statistically analyze tens of thousands of atomic coordinates and calculate the proportion of titanium and aluminum atoms under each z-coordinate value.This method achieved element line scanning along the z-axis direction of the molecular dynamics calculation results, and the processing results are shown in figure 22.When the proportion of titanium and aluminum atoms is both higher than 5%, it is determined that element diffusion has occurred at that position, and the median plane of atomic diffusion is the plane where the number of the two types of atoms is closest.Figure 22(a) shows the results of atomic coordinate analysis during the loading and collision stage.Contrary to the conclusion drawn from directly observing the atomic distribution, it was found through data analysis that atomic diffusion also occurred during the loading and collision stage.The diffusion occurred from z = 173 Å to z = 178 Å, with a diffusion thickness of 5 Å and a diffusion median plane at z = 175 Å. Figure 22(b) shows the atomic positions after a 200 ps relaxation under high temperature and pressure.The diffusion occurred from z = 165 to z = 178 Å, with an increased diffusion thickness of 13 Å and a diffusion median plane shifted to z = 171 Å. Due to the significant compression of the system volume along the positive z-axis direction during the collision stage, the system volume exhibited a certain rebound tendency during the relaxation after the collision.Therefore, the atomic diffusion in this stage was along the direction of volume rebound, i.e., diffusion in the negative z-axis direction.After a 200 ps relaxation in the unloading stage, the atomic diffusion thickness increased dramatically to 33 Å, with a diffusion position of z = 157 ∼ 190 Å.At this time, the system volume rebounded more severely, and with the increase of interatomic distance, atomic diffusion occurred in both positive and negative z-axis directions, with a diffusion median plane at the atomic plane of z = 163, which continued to move downward compared to the previous stage (figure 22(c)).
During the cooling stage, the atomic diffusion position changed to z = 157 ∼ 194 Å, with an increase in atomic diffusion thickness of only 4 Å.The diffusion median plane was at z = 164 Å.However, compared with figure 22(c), it can be seen that the change curve of the two types of atoms at the diffusion position was smoother, indicating that the two types of atoms were still in mutual motion at the diffusion position.Figure 22(e) shows the atomic diffusion thickness image after cooling and relaxation, which is not significantly different from the cooling stage.
Based on the comprehensive analysis of the MSD curves, atomic distribution images, and atomic coordinate data, it can be concluded that element diffusion occurred in all stages of titanium-aluminum explosive welding during molecular dynamics simulation, with the most significant diffusion behavior occurring during the unloading stage at zero external pressure.

Analysis of Ti6Al4V/7075 binding mechanism
Figure 23 shows the results of the radial distribution function (RDF) calculation for titanium-aluminum atoms.The figure indicates that during the initial stage of system establishment, both titanium and aluminum exhibited an ordered solid metal crystal structure, with clear peaks appearing even at a far distance, and g values between peaks were 0. The first peak of the RDF represents the nearest neighbor distance of atoms.Compared with the initial state, the nearest neighbor distance of atoms decreased significantly during loading, and after relaxation in the following stages, the nearest neighbor distance gradually returned to the initial level.After collision relaxation, the average temperature of the system increased above the melting points of titanium and aluminum, and as the interatomic distance increased, the radial distribution function of aluminum gradually became a straight line, indicating that aluminum had become a lattice-disordered molten state.In contrast, the radial distribution function of titanium still showed clear peaks with increasing interatomic distance, indicating that titanium still maintained some characteristics of a solid metal.The high pressure generated during collision restricted the displacement of atoms and increased the melting point of the system metal to some extent, so that the melting point of titanium was higher than the system temperature at this time, and no melting occurred.During the loading stage, titanium and aluminum existed in a solid+fluid combination form.As concluded from the previous analysis of atomic distribution images and atomic coordinate statistics, the diffusion of interface atoms was most severe during the unloading stage.At this stage, the pressure of the system disappeared, and the melting points of both materials decreased to the melting point values at normal pressure.The radial distribution function of atoms at this stage also showed that both titanium and aluminum were in a short-range ordered, long-range disordered molten metal state, which belonged to a fluid+fluid combination.Further analysis of the figure shows that titanium briefly melted during the unloading stage and quickly returned to a long-range ordered solid metal state, while aluminum remained in a molten state in the following stages.
The effective criterion for distinguishing pressure welding from fusion welding is the occurrence of plastic deformation without material melting.The initial height of the titanium-aluminum model system was 242 Å, and the model height stabilized at 209 Å in the final relaxation stage, indicating that the system underwent plastic deformation.However, the radial distribution function calculation (figure 24) shows that both materials became fluid during the unloading stage and underwent melting, which does not meet the criteria for pressure welding.Some theories [48] suggest that under pressure, the diffusion of the two metals can occur, resulting in diffusion welding, which is also a manifestation of pressure welding.However, the analysis of atomic diffusion behavior shows that the ultra-high pressure during the explosive welding process hindered the movement of atoms, and atomic diffusion occurred only during the unloading stage at zero pressure.Therefore, pressure welding cannot be used as the reason for the bonding in this explosive welding process.Throughout the welding process, the two metals underwent various phase combinations, including solid +solid, solid+fluid, semi-molten+semi-molten, fluid+fluid, and.Although the fluid+fluid material state exhibited by the materials during the unloading stage meets the phase criteria for fusion welding, testing of the materials (figure 16) revealed the generation of a continuous melting zone at the titanium-aluminum composite interface that is unfavorable for bonding, as well as obvious ablation defects.Therefore, it cannot be concluded whether high-temperature melting is conducive to bonding or whether fusion welding occurred during the explosive welding process.
The interface atomic diffusion behavior and material EDS element analysis results show that continuous and stable element diffusion occurred at the titanium-aluminum interface, generating a certain thickness of titanium-aluminum intermetallic compound layer, which enhanced the bonding strength of the materials.Therefore, diffusion welding is a reason for the interface bonding in this explosive welding process.

Conclusion
The present study employed the SPH smoothed particle hydrodynamics algorithm and MD molecular dynamics algorithm to conduct numerical calculations of Ti6Al4V/7075 explosive welding at two scales, and experimental verification was performed based on the calculation results, producing the highly challenging high-strength and high-hardness Ti6Al4V/7075 composite material.
The weldability window of Ti6Al4V/7075 is extremely narrow, and both the SPH and experimental results indicate that the Ti6Al4V/7075 interface is mainly flat.Local continuous melting zones dominated by TiAl and ablation zones dominated by TiAl3 can be observed in the composite material samples.
The molecular dynamics results show that during the explosive welding process, a large number of 1/6〈110〉 dislocations were generated due to plastic deformation on the titanium side, and the aluminum lattice structure underwent a transformation from FCC to BCC.The crystal structure of titanium also underwent a transformation from HCP to FCC and BCC during the explosive welding process, and a large amount of β-Ti was still observed on the titanium side of the composite material obtained from the experiment.
The interface exhibited obvious element diffusion, which mainly occurred during the unloading stage.Diffusion welding and atomic solid-phase bonding were the main reasons for the bonding in this explosive welding process

Figure 3 .
Figure 3. Molecular dynamics model of titanium aluminum explosive welding.

3. Results and analysis 3 . 1 .
Analysis of interface microscopic morphology 3.1.1.SPH calculation results Figures7 and 8show the calculation results of interface temperature and pressure, respectively.Three positions along the detonation direction are sequentially recorded as Gauge1 ∼ Gauge3.The interface temperature calculation results show that the interface temperature and pressure immediately increase at the moment of collision.The interface temperatures at the three test positions along the detonation direction are 1269, 1962, and 2362 °C, respectively, and the peak pressure at the interface is 13.6, 19.8, and 25.6 GPa, respectively.The temperature and pressure show a gradually increasing trend.The temperature and pressure values calculated by the finite element numerical calculation results are basically consistent with the system temperature and pressure results calculated by molecular dynamics.

Figure 10 .
Figure 10.Cloud chart of temperature distribution for SPH numerical calculation.

Figure 11 .
Figure 11.Cloud chart of pressure distribution for SPH numerical calculation.

Figure 12 .
Figure 12.Temperature and pressure results of molecular dynamics system.

Figure 13 .
Figure 13.Molecular Dynamics CNA Calculation Results: (a) the end of the load collision stage, (b) the end of the post collision relaxation stage, (c) the end of the unloading stage, (d) the end of the cooling stage, and (e) the end of the room temperature relaxation stage.

Figure 14 .
Figure 14.Calculation Results of Dislocations: (a) the end of the load collision stage, (b) the end of the post collision relaxation stage, (c) the end of the unloading stage, (d) the end of the cooling stage, and (e) the end of the room temperature relaxation stage.

Figure 16 .
Figure 16.Typical Interface Morphology: (a) location of the detonation point, (c) location far from the detonation point, (b) location halfway between (a) and (c) location, (d) local result of SPH numerical simulation.

Figure 17 .
Figure 17.EDS point scanning results of melting zone.

Figure 18 .
Figure 18.EDS point scanning results in the burning loss area.

Figure 19 .
Figure 19.EDS Surface Scanning Results of the Interface: (a) element scanning position, (b) element line scanning result, (c) titanium element surface scanning result, (d) aluminum element surface scanning result.

Figure 21 .
Figure 21.Interface Atom Distribution Image: (a) the end of the load collision stage, (b) the end of the post collision relaxation stage, (c) the end of the unloading stage, (d) the end of the cooling stage, and (e) the end of the room temperature relaxation stage.

Figure 22 .
Figure 22.Atomic Coordinate Analysis Results: (a) the end of the load collision stage, (b) the end of the post collision relaxation stage, (c) the end of the unloading stage, (d) the end of the cooling stage, and (e) the end of the room temperature relaxation stage.

Figure 23 .
Figure 23.Calculation results of radial distribution function of titanium and aluminum atoms.

Table 2 .
Composition of base plate and flyer plate materials.

Table 3 .
Mechanical properties of base plate and flyer plate materials.