Spall Plane Research Articles

Dynamic response of additively manufactured Ti–5Al–5V–5Mo–3Cr as a function of heat treatment

Both shock and shockless compression experiments were performed on laser powder bed fusion (LPBF) Ti–5Al–5V–5Mo–3Cr (Ti-5553) to peak compressive stresses near 15 GPa. Experiments were performed on the as-built material, containing a purely β (body centered cubic) microstructure, and two differing heat treatments resulting in a dual phase α (hexagonal close packed) and β microstructure. The Hugoniot, Hugoniot elastic limit (HEL), and spallation strength were measured and compared to wrought Ti-6Al-4V (Ti-64). The results indicate the LPBF Ti-5553 Hugoniot response is similar between heat treatments and to Ti-64. The HEL stress observed in the LPBF Ti-5553 was considerably higher than Ti-64, with the as-built, fully β alloy exhibiting the largest values. The spallation strength of the LPBF Ti-5553 was also similar to Ti-64. Clear evidence of initial porosity serving as initiation sites for spallation damage was observed when comparing computed tomography measurements before and after loading. Post-mortem scanning electron microscopy images of the recovered spallation samples showed no evidence of retained phase changes near the spall plane. The spall plane was found to have kinks aligned with the loading direction near areas with large concentrations of twin-like, crystallographic defects in the as-built condition. For the heat-treated samples, the concentrations of twin-like, crystallographic defects were absent, and no preference for failure at the interface between the α and β phases was observed.

Correlation between grain size and dynamic response of NiTi alloy during intense shock-induced multi-spallation

This work committed to performing non-equilibrium molecular dynamics simulations to explore the influence of grain size on multi-spall performance of nanocrystalline nickel-titanium (NiTi) alloy. By comparing the calculated Hugoniot relationship with present experimental results, the Finnis-Sinclair (FS) type potential used in this work was verified to be suited for reproducing the shock behavior of NiTi alloy at high strain rate. During shock wave evolution, a single plastic wave was clearly observed in the sample with grain size of 5 nm, while there was a more pronounced elastic-plastic double wave structure at larger grain size. According to the free surface velocity history, the predicted spall strength exhibited a decreasing tendency and subsequently increased when the grain size decreased from 50 to 5 nm, indicative of a critical grain size (10 nm). Correspondingly, there were multiple spall planes appearing in the sample, the grain size of which was not smaller than the critical one. Contrarily, a single spall plane was observed at 5 nm. By counting the void distribution statistic, we found that as the grain size increased, the void-nucleation behavior changed from the intergranular to intergranular/transgranular coexisting pattern. In consequence, more nucleation sites were provided to promote an increase in the number of voids. Interestingly, for intergranular, there were much more voids involved in spallation at smaller grain size, which was attributed to grain boundary effect. The results demonstrated that void evolutions rarely occurred at the grain boundaries parallel to the loading direction. Moreover, when the grain size was larger than 5 nm, the B19 phases were found to be produced around those potential void-nucleation sites. After void nucleation, these B19 phases decreased till disappearance, accompanied by the proliferation of a large number of [100] dislocations. Such dislocations accumulated in the region without voids could block the void growth in this region, which was expected to be the reason for the formation of multiple spall planes. On the contrary, the absence of dislocations at 5 nm was inclined to the fact that the space for dislocation emission and aggregation was preoccupied by nucleated voids.

An assessment of spall failure modes in laser powder bed fusion fabricated stainless steel 316L with low-volume intentional porosity

This paper reports on spall failure and damage modes in Laser Powder Bed Fusion fabricated Stainless Steel 316L (SS316L) with intentional levels of low-volume (1–5 vol. %) porosity and pore sizes of 200, 350, and 500 μm. The fabricated specimens were subjected to uniaxial-strain plate-impact loading at ∼4.5 GPa, to initiate incipient spall failure. Analysis of velocimetry profiles measured using multi-probe photon-Doppler velocimetry coupled with post-mortem analysis of soft-recovered samples reveals local suppression of spall failure (termed as spall-dominated) as a function of porosity, as the failure mechanism transitions from spall-centered tensile stress dominated to a pore-centered microstructure-dominated damage mode involving void/crack nucleation and growth at pre-existing pores. The critical porosity level where the suppression of spall failure is first observed, as well as the spall location, is dependent on both the volume fraction and the size of the initially fabricated pores. In samples of 500 μm pore size, the suppression of spall failure is observed with as little as 1 vol. % porosity, while samples with smaller pores (200 μm) still experience spall-centered tensile stress dominated failure with higher levels (5 vol. %) of porosity. In the case of pore-centered microstructure-dominated failure, spall damage can occur but the spall plane is shifted toward the rear free surface, or more generally in areas further away from the region with pores. Highly heterogeneous deformation twinning, shear banding, grain rotation, and cracking are observed in the vicinity of pre-existing pores and expected spall failure sites.

Anisotropic spall failure of additively manufactured 316L stainless steel

The dynamic high strain rate tensile (spall) failure of additively manufactured (AM) materials is an area of great interest but much less understood than material failure under other loading regimes. The research presented here examines the effects of microstructure anisotropy in 316L stainless steel fabricated by laser powder bed fusion (LPBF) on its spall failure and properties. Plate-on-plate impact experiments were performed using an 80-mm gas gun to investigate the spall behavior of AM fabricated sample discs with impact along the perpendicular and parallel orientations relative to build directions. Experiments were also performed on wrought 316L stainless steel as baseline material with uniform equiaxed grain microstructure for comparison with the AM material of same composition. The spall experiments involved free-surface velocity profiles measured using Photon Doppler Velocimetry (PDV) to capture the spall pull-back signals, combined with post-mortem analysis of the microstructure of soft-recovered impacted samples. In the case of AM samples, spall failure initiated preferentially at melt pool boundaries in both impact directions. The spall plane location and spall strength, however, were both heavily influenced by the orientation of the impact direction relative to the build direction in the AM samples. The spall behavior is also observed to be more complex in the samples impacted perpendicular to the build (through-thickness) direction, due in large part to the orientation of the microstructural features such as melt pool boundaries, laser scan tracks, and general grain morphology.

Dynamic properties of 316l stainless steel repaired using electron beam additive manufacturing

Additive manufacturing has the potential to be used for repair of high-value parts, but the strength and integrity of such repaired parts, especially under dynamic loading, in still unknown. This work presents the results of high strain-rate dynamic testing of additively repaired stainless steel samples. The 316L samples were intentionally damaged and subsequently repaired with 308L stainless steel wire using electron beam additive manufacturing. The repaired samples were subjected to gas-gun impact testing to induce incipient damage and determine the role played by the repaired region in dictating damage and failure. The results show the following: 1) the equation of state of the repaired region is comparable to the original material as shown by the unchanged spall plane location, 2) the deformation behavior is similar to the original material as shown by the similar Hugoniot Elastic limit, 3) the spall strength of the repaired region is slightly higher than the original material due to varying grain size, and 4) the damage morphology in the repaired region is different and involves a high rate of void coalescence. Despite these minor differences in the spall strength and the damage morphology, the overall dynamic response of the original material and repaired samples is similar suggesting that AM is a promising approach for repairing high-value components.

Subgrain geometrically necessary dislocation density mapping in spalled Ta in three dimensions

In this work, the intragranular distributions of the densities of geometrically necessary dislocations (GNDs) generated in a polycrystalline tantalum (Ta) sample loaded in shock compression to spall are investigated. The sample studied is a highly resolved three-dimensional (3D) map of the orientations, grain boundary morphology, and voids across a 0.276mm3 (with sample dimensions of 688.5µm × 669µm × 600µm) region of the microstructure about the spall plane obtained by TriBeam tomography. We present a formulation for calculating the crystallographic screw and edge character GND densities at a material point within body centered cubic (BCC) crystals when provided any number of potential slip systems or modes. Combining the 3D characterization, GND formulation, and a sample containing ≈ 6000 grains, we sought statistically relevant correlations between GND density per grain with grain size, grain orientation, void formation, and proximity from the spall plane. We show that GND density per grain increases with decreasing distance from the spall plane. Most grains had heterogeneously distributed GND densities but were free of any detectable voids, at least within the ≈ 1.5µm resolution of the characterization technique. It is found that the GND density for most grains is insensitive to grain size and grain orientation. Only the fraction of grains with the highest GND densities tended to have a [001] parallel to shock-direction orientation. Intragranular 3D maps of GNDs reveals that grains that contain voids have high GND density concentrations in the intragranular region immediately surrounding the void. The local intragranular GND mapping also showed voids that are connected to grain boundaries preferred to grow into the grain with the larger size, regardless of orientation.

Evaluation of copper spall response using Mie-Grüneisen equation-of-state parameters determined from molecular dynamics

In this study, the formation and application of an across-scale modeling approach is presented to study the influence of shock loading conditions on copper spallation performance. Molecular dynamics simulations coupled with multi-scale shock technology are conducted to determine characteristic parameters of the Mie-Grüneisen equation of state involved in the Slater and Dugdale-MacDonald models. Afterwards, the resultant parameters are embedded into the finite element method to explore the dependence of spall strength, strain rate, and damage evolution rate. It is found that an increase in loading velocity promotes the spallation occur in advance. Combined with free surface velocity and spall plane pressure, it is also discovered that the material experiences an entire spallation prior to ''pullback signal''. As expected, the spall strength determined as the maximum stress is higher than that estimated from the free surface velocity. Besides, there is a temperature-rise effect caused by void-induced thermal dissipation during the spallation. Moreover, the position of spall plane is strongly dependent upon flyer thickness, thereby exhibiting a strain rate effect on spallation behavior. A comparison between numerical results and published experimental data highlights the adequacy of such a hybrid method to capture the copper spall response.

Spall damage mechanisms in laser powder bed fabricated stainless steel 316L

This paper reports on spall damage mechanisms in laser powder bed fusion (LPBF) fabricated stainless steel 316L (SS316L) subjected to uniaxial, strain plate-impact loading and explores failure evolution with increasing impact velocity and peak pressure. Analysis of velocimetry profiles reveals a heterogeneous failure response with several different failure mechanisms activated during impact and resulting spall. The analysis is supported by optical microscopy and electron backscattered diffraction (EBSD) observation of soft recovered impacted samples, which show evidence of localized nano-twinning, twin growth, severe grain rotation, and grain refinement along the spall plane, in addition to pore nucleation, coalescence, and growth. The observed failure mechanisms, including crack formation and propagation along high angle grain boundaries (HAGBs), are not consistently indicative of ductile fracture typical of wrought stainless steel.

Spallation of polycrystalline tungsten using laser-induced shock at ultrahigh strain rates

Laser-induced shock technology was used to achieve the shock loading of polycrystalline tungsten (W) at an ultrahigh strain rate. The shock response and fracture behavior of W were investigated at different tensile strain rates. Unlike the intergranular fracture mode under quasistatic conditions, the transgranular–intergranular-mixed fracture mode of W was detected at high strain rates and the proportion of transgranular cleavage mode increased. The spall strength of W was positively correlated to the strain rate, reaching 5.86 GPa when the tensile strain rate reached 106 s−1 in the spall plane. The mobility of screw dislocations increased at high strain rates, and the simultaneous emission of screw dislocations improved the spall strength while accelerating the expansion of crack nucleation positions, inducing numerous transgranular fractures.

Role of void shape on shock responses of nanoporous metallic glasses via molecular dynamics simulation

The void shape significantly affects the mechanical properties of nanoporous metallic glasses (NPMGs) under conventional loading. However, how the shock responses of NPMGs depend on the void shape has been rarely studied. In this work, shock simulations of Cu50Zr50 NPMGs containing voids with different shapes are performed by molecular dynamics, aiming to investigate the roles of the void shape on plastic deformation, void collapse, and spallation. The results reveal that the void shape has a significant effect on the peak value of the atomic strain curve and the fraction of atoms participating in the plastic deformation. There is a very weak dependence of the void collapse mode on the void shape under the weak shocks, where the voids collapse perpendicularly to the shock direction. As the shock strength increases, the void collapse mechanism gradually transforms into an internal jet, and the void shape plays an important role in the jet formation. During spallation, the initial damages tend to concentrate in the void collapse regions and then coalesce into large damages. The void shape influences the coalescence degree of spall damage. The spall strength is positively correlated with strain rate and local five-fold symmetry, but negatively correlated with temperature on the spall plane. Additionally, the void influences the spall strength due to the different distribution of the above factors after voids collapse in nanoporous samples.

Exceptionally high spallation strength for a high-entropy alloy demonstrated by experiments and simulations

High-entropy alloys are materials with an increasing number of technological applications. Amongst them, the Cantor alloy, FeMnCoCrNi, shows desirable mechanical properties at normal loading conditions. In this study we focus on the performance of the Cantor alloy at the ultra-high deformation rates of shock waves. We study shock-induced spallation using both experiments and atomistic simulations. Experimental loading is achieved using high power laser, with VISAR to obtain velocity profiles and spall strength, followed by transmission electron microscopy of the recovered samples. Molecular Dynamics (MD) simulations of shock-induced spallation are compared with experiments. Both experiments and simulations show a high spall strength which would be beneficial for certain applications, with experiments giving ~8 GPa at ~107s−1 and MD giving almost ~30 GPa at ~109s−1. The difference between experiments and simulations can be explained by the difference in strain rate. Post-mortem analysis of the experimental samples shows nanotwins near the spall plane, while MD simulations show a highly disordered region giving rise to void nucleation and spall during loading.

Effects of shock-induced phase transition on spallation of a mild carbon steel

Effects of shock-induced phase transition (PT) on spallation of a mild carbon steel are investigated via plate impact experiments at impact velocities of 0.433–1.163 kms−1. The free surface velocity history is recorded to deduce spall strength and wave propagation/interaction within the sample. Postmortem samples are characterized with scanning electron microscopy and electron backscatter diffraction. Detailed wave analysis indicates that the PT affects wave propagation and interaction. Such wave propagation and interaction finally lead to the change of spall strength and spall plane position for different stress ranges. In the peak shock stress range of 13–17 GPa (slightly above the PT stress), the release fan of the PT wave accelerates the free surface but is not involved in the major spallation, giving rise to an overestimation of spall strength by the conventional acoustic method. However, at higher shock stress above PT (>17 GPa), the release fans of the PT wave interact with each other and induce secondary spallation, and the acoustic method is applicable in such cases; the major spall plane moves to the sample central region, and multiple secondary spall planes appear. In addition, the reverse PT upon release facilitates nucleation and growth of deformation twins which subsequently act as nucleation sits of voids during spallation. With increasing shock stress, the spall damage exhibits mode transitions from brittle damage (transgranular cleavage cracks), to mixed brittle and ductile damage, and to ductile damage (void nucleation and growth).

Role of pre-existing dislocations on the shock compression and spall behavior in single-crystal copper at atomic scales

Large-scale molecular dynamics simulations are carried out to investigate the role of pre-existing dislocation loops on the shock-induced deformation and spall behavior of single-crystal Cu microstructures. This study investigates the role of loading orientation and initial density of pre-existing dislocations on the decay behavior of the Hugoniot elastic limit (HEL) as well as the damage nucleation and growth behavior during spall failure of single-crystal Cu systems. The results suggest that the presence of pre-existing dislocation loops results in a decrease of the shock wave velocity and a substantial decay of the HEL values. The increased decay behavior is attributed to the decrease in the density of Shockley partials at the shock front as the shock wave travels through the metal as compared to defect-free initial single-crystal microstructures. Similarly, the presence of pre-existing dislocations is observed to result in increased values for the spall strength as compared to defect-free initial single-crystal microstructures wherein a higher density of dislocations results in the nucleation of a larger number of smaller voids. The decay behavior of the HEL values is observed to have a power–law dependence on the shock propagation distance with the initial dislocation density as a parameter. Similarly, a power–law dependence is also proposed for the number of voids nucleated at the spall plane with a dependence on the size of the voids as well as the initial density of dislocations. The evolution of microstructure (dislocation densities and voids) for the various loading orientations and initial densities of dislocations is discussed.

Modeling of shock-wave processes in aluminum under the action of a short laser pulse

A hydrodynamic model of shock-wave processes in a material under the action of a short high-intensity laser pulse is considered. The simulation is carried out for the case of an aluminum target 90 μm thick, irradiated by a laser pulse with a duration of 70 ps and a maximum intensity of 14.7 TW/cm2. In the corresponding laboratory experiment, on the rear side of the target after irradiation, a spall of a part of the material is recorded at a depth of 10±1 μm. Calculation of the time dependence of the pressure and density of aluminum in the spall plane makes it possible to determine the tensile strength of aluminum at a high strain rate.

Spall and subsequent recompaction of copper under shock loading

Research was undertaken to study spall and subsequent recompaction of oxygen-free high-thermal-conductivity copper using a single-stage large-bore light gas gun capable of planar impacts. Gun experiments were conducted that produced an initial spall in the target with a subsequent recompaction of the spall damage/layer by the use of layered impactors. Symmetric spall experiments at similar conditions were also conducted as a control to the recompaction experiments. Photonic Doppler velocimetry was used to obtain the velocity history of the target's back surface; these velocity histories were analyzed, and the results were compared to numerical simulations. After recovery of selected samples, the microstructure was analyzed using optical imaging microscopy and electron backscatter diffraction. Recompaction waves were clearly observed in the time-resolved data obtained at the back surface of the targets. Two different series of experiments were performed to understand the effect of damage morphology on the stress required for spall recompaction. In the initial series, the peak pressure of the first shock was systematically increased to alter the amount of damage in the material. While in the second series, the peak pressure of the second shock was increased, keeping the magnitude of the first shock relatively constant to understand the effect of peak stress on the recompaction process. Findings clearly show that recompaction results in a perturbed band of the microstructure at the location of the expected spall plane in all cases except where the peak stress of the second shock is lower than the critical value.

Method of determining parameters of void growth damage model and its application to simulation of spall test

Spallation of ductile metal is of great importance in many scientific and engineering fields, which is due to the interaction between the incident shock waves and the reflected waves. Physically, the spallation is caused by nucleation, growth and coalescence of microvoids for ductile material. Therefore, numerical simulation of spall process usually involves theoretical model of void growth. However, due to the limited knowledge of microvoid properties, many empirical parameters are included in the void growth model, which are usually determined by comparing numerical results with experimental data. Therefore, a key problem arises in the numerical modeling of damage and spall fracture, that how the parameters of the void growth damage model can be determined. In this work, we present a theoretical method to determine the parameters based on the free surface velocity (FSV) profile. Firstly, the critical state of damage is discussed based on the relationship between characteristics of FSV and change of physical quantity in spall plane. Then, the propagation and interaction of shock waves during the evolution of spall damage are considered. Lastly, the physical meanings of the parameters of the void growth damage model are further discussed. So, based on the relation among spall strength, damage and pull-back of FSV, a physics-based method to determine the parameters of the model is given. The applicability of this method is verified by the simulation of the spall experimental data on typical ductile metals OFHC copper and tantalum. The parameter-determining method given in this paper can not only expand the scope of application of the damage model and effectively improve the reliability of the calculation results, but also provide a good reference for the determination of parameters of other spall damage model.

Effect of peak stress and tensile strain-rate on spall in tantalum

Materials subjected to dynamic environments experience a complex and wide range of stress, strain, and strain-rate conditions. To have confidence in material models, an accurate, predictive capability is required. In this study, we present a series of flyer-plate impact tests on well characterized, high purity tantalum. The shock-waves generated at impact release from the free-surfaces, reflect, and interact to produce incipient spall fracture. By varying the flyer-plate material and impact velocity, both the peak stress and the strain-rate in the samples were controlled independently. Velocimetry was used on the rear free-surface of the samples to measure the shock-response and the spall strength. While this measurement provided the same spall strength for all cases, at approximately 5.1 GPa, when the samples were sectioned during post-mortem, the quantity and distribution of internal damage was markedly different. For the high-strain rate cases, voids remained small and isolated, whereas in the lower strain-rate experiments, the spall damage was far more localized, with a well-defined continuous spall plane. With the use of hydrocode simulations, this was discovered to result from how the different release rates affect the interaction volume inside the sample. These results highlight the importance of careful sample recovery, and the risks of relying solely on free-surface velocity measurements.

Spall response of single-crystal copper

We performed a series of systematic spall experiments on single-crystal copper in an effort to determine and isolate the effects of crystal orientation, peak stress, and unloading strain rate on the tensile spall strength. Strain rates ranging from 0.62 to 2.2 × 106 s−1 and peak shock stresses in the 5–14 GPa range, with one additional experiment near 50 GPa, were explored as part of this work. Gun-driven impactors, called flyer plates, generated flat top shocks followed by spall. This work highlights the effect of crystal anisotropy on the spall strength by showing that the spall strength decreases in the following order: [100], [110], and [111]. Over the range of stresses and strain rates explored, the spall strength of [100] copper depends strongly on both the strain rate and shock stress. Except at the very highest shock stress, the results for the [100] orientation show linear relationships between the spall strength and both the applied compressive stress and the strain rate. In addition, hydrodynamic computer code simulations of the spall experiments were performed to calculate the relationship between the strain rate near the spall plane in the target and the rate of free surface velocity release during the pullback. As expected, strain rates at the spall plane are much higher than the strain rates estimated from the free surface velocity release rate. We have begun soft recovery experiments and molecular dynamics calculations to understand the unusual recompression observed in the spall signature for [100] crystals.

Defect and damage evolution during spallation of single crystal Al: Comparison between molecular dynamics and quasi-coarse-grained dynamics simulations

The links between loading orientation of single crystal Al and the dynamic evolution of defects (dislocations, twins, stacking faults etc.) during spallation are investigated using molecular dynamics (MD) simulations. The microstructural evolution during the shock compression of single crystal Al is observed to be primarily guided by the nucleation and evolution of Shockley partials and twin partials. The shock response and spallation behavior of single crystal Al is observed to be anisotropic and is influenced by the density of various types of dislocations during shock compression and void nucleation at the spall plane. The capability of the “quasi-coarse-grained dynamics” (QCGD) method to reproduce the MD predicted nucleation, interaction and evolution of dislocations during shock compression and spallation of single crystal Al is discussed. The QCGD method retains the relative contribution of different types of dislocations during propagation of shock compression wave, interactions of the wave and spallation of single crystal Al as predicted by MD using a fraction of representative atoms allowing the modeling of of larger sized systems for larger times.

Influence of initial texture on the shock property and spall behavior of magnesium alloy AZ31B

In this work, flyer plate impact experiments are conducted on a commercially available magnesium alloy, AZ31B. Equal channel angular pressing (ECAP) process is an established procedure for fabrication of ultra-refined grain structures in metals and alloys. The microstructure of commercially available AZ31B is refined through ECAP process and two sets of specimens with different initial textures and microstructure are considered here. One set of the specimens is machined from a hot rolled AZ31B Mg alloy plate and the others are processed by ECAP of four pass after samples are cut from the initial material. The Hugoniot elastic limit (HEL) and spall strength are determined by analyzing the measured free surface velocity profiles of the specimens. Scanning electron microscope (SEM) is used to observe the microvoids produced along the spall plane and electron back scatter diffraction (EBSD) is employed to examine the twinning distribution on the cross-section of the recovered specimens cutting along the thickness direction. It is found that elastic precursor decay appears in the as-received (without ECAP process) specimens but it is absent in ECAPed specimens. The HEL of the as-received specimens is higher than that of ECAPed specimens over various impact velocities and showed strain rate sensitivity. The spall strength of both sets of specimens is almost identical and enhanced by increasing the impact velocity. The SEM results showed that in the as-received specimens the spall is dominated by coalescence of microvoids emerged along the grain boundaries in a brittle manner. However, in the ECAPed specimens spall was dominated by coalescence of microvoids characterized by inner dimples formation by the pull out of grains in a ductile mode.