This paper presents the results of sound velocity measurements performed in natural uranium during the high-rate strain of sound in  $$\alpha$$ - and  $$\gamma$$ -phase regions, as well as in a melting region at a shock wave front. The processes are recorded using the manganin sensor method (range 20–110 GPa) and the optical analyzer technique (range 70–260 GPa). In the region of the solid uranium phase, changes in Poisson’s ratio, Young’s modulus, bulk modulus, and shear modulus along the uranium Hugoniot are determined.

It is proposed to combine the optical analyzer technique (OAT) and multiply photon Doppler velocimetry (MPDV) in every explosive experiment in order to overcome the reported discrepancies in the sound velocities measured in shock-compressed metals using different techniques. Such a combination is demonstrated to be efficient in explosive experiments with stepped samples of 12Kh18N10Т austenitic stainless steel and high-purity Mg95 magnesium under their shock-wave loading in ranges of 60–120 and 20–30 GPa, respectively. The OAT ensures classical recording of the longitudinal and bulk sound velocities. Depending on the loading intensity, the MPDV technique recorded temporal changes in the velocity of the “sample–indicator” interface or temporal changes in the shock-front velocity in the indicator. These data were used to monitor the parameters of the shock-compressed sample and to determine the instant when the first characteristic of the rarefaction-wave fan overtakes the shock-wave front in the indicator. Within the range of relatively low loads where the indicating fluid remains transparent and acts as a window material, the MPDV registers temporal variations in the velocity of the “sample–indicator material” interface. The OAT ensures reliable registration in the range of high-intensity loads, while the MPDV ensures time-resolved registration of the steady-shock-wave front velocity in the indicator up to the instant of its overtaking by a rarefaction wave. Both techniques are observed to work well in the intermediate range of loads. Combining these techniques enhanced the reliability of the obtained consistent data on the sound velocities in shock-compressed structural materials and also allowed a decreased number of explosive experiments with samples of toxic materials.

The Kitagawa-Takahashi (K-T) diagram, implemented by the El Haddad equation, relates the conventional fatigue limit to the crack size, enabling a boundary for the cyclic stress range to be established, below which an infinite life of the structural component may be, theoretically, ensured for any crack size due to the non-propagation of microand macrocracks. In order to account for the inherent random character of the fatigue phenomenon in real materials and the need of extending the K-T applicability to any prefixed number of cycles, advanced probabilistic S-N models should be considered to define the fatigue limit. In this way, a new basis towards a probabilistic Kitagawa-Takahashi-El Haddad approach is provided in agreement with the asymptotic matching proposed by Ciavarella-Monno. INTRODUCTION AND MOTIVATION The Kitagawa-Takahashi (KT) diagram [1] represents a boundary in terms of crack size and stress range for which infinite fatigue lifetime of structural or mechanical components can be safely ensured due to non-propagating microand macrocracks. Such fatigue life assessment can be related to both the classical fatigue limit concept, resulting from the experimental-based S-N approach, and the threshold stress intensity factor range, as defined by the crack propagation law. Even after the transcendent improvement provided by the intrinsic crack concept of El Haddad (EH) [2], two issues need to be dealt with: a) the extension of the KT-EH diagram to a fatigue limit for finite number of cycles, which is not necessarily identified with the endurance limit for N=∞, and b) a stochastic definition of the KT-EH diagram as a consequence of the variability of the basic fatigue functions being considered (S-N and crack growth rate curves). Both represent practical requirements related to structural integrity design. In this work, a new approach to the problem is supplied by considering the probabilistic S-N developed by Castillo and Fernandez-Canteli [3], which provides a sound basis in the definition of the KT-EH line, permitting also the model to be

Introduction. Cerium is an element with a single 4f-electron. Nevertheless, under the action of pressure and temperature, it fairly simulates features in the behavior of an actinide having five 5felectrons. More than 50 years ago, the process of isostructural (with preservation of the crystal lattice type) –-electronic phase transformation with a high jump (20%) in the specific volume was experimentally observed in cerium [1,2]. Cerium is only one element of the Mendeleev's Periodic table, which has the critical point in the solid-phase range both in the case of positive (compressive), and negative (tension) stresses. An unusually complex phase diagram for cerium [3] and the location of its critical points in the region of relatively low pressures and temperatures have made cerium the object of numerous experimental researches though mainly static ones. Dynamic experiments, i.e. shock-wave and explosive experiments with samples made of metallic and moreover high-purity (99.99 %) cerium, are few in number [2]. We can mention only the following: ­ electrocontact measurements performed at VNIIEF in 1968-69 (Altshuler, Bakanova et.al. [4,5]); ­ discrete diagnostics by LANL in 1973 (Gust, Royce [6]); ­ analog diagnostics by LLNL in 1975 (Carter et al. [7]); ­ high-resistance manganin gauge at VNIIEF in 1999 (M.N. Pavlovsky et.al. [8]); ­ PVDF-gauge at VNIIEF in 2005 (Borisenok et.al. [9]); ­ optical analyzer technique and the manganine gauge technique to measure sound velocity in cerium being shock-compressed within 30-140 GPa (Zhernokletov, Kovalev et.al. [10]); ­ laser VISAR-diagnostics at VNIIEF in 2007 (Pushkov, Ogorodnikov, Erunov [11]); ­ laser-interferometric VISAR-diagnostics on high-purity cerium samples (Hixson et al. 2002, Cherne et al. 2005-2007 [12]). The purpose of this work is to obtain new data on spall strength of high-purity cerium being loaded in the region of  and –liquid phase transformations. Material, samples, and conditions of their explosive loading. Wedge samples (30406 mm 3 , 1200 angle) of cerium having high-purity (99.99 wt %) and the 6.75 g/cm 3 initial density were manufactured in LANL and delivered to RFNC-VNIITF for investigations. These samples were loaded through the 12Kh18N10T steel base-plate with the 5-mm thickness by sliding and normal detonation of an HE layer having different type and thickness. We used the PETN-based plastic explosive with the 0.7-, 1-, and 5-mm thick layers, as well as the HMX-based explosive composition, the layers being 10-, and 20-mm thick.

The optical lever method is used to obtain new data characterizing the dynamic properties of quenched 30KhGSA steel in the loading range up to 30 GPa. These data include the shock compressibility (below and above σ xx α−ε = 15 GPa, which corresponds to the α’-ɛ transformation in the steel), the shear and spall strength, and the kinetic laws of the interrelated processes of high-rate deformation and fracture. These systematic experimental data are of interest and can be used to verify modern interrelated kinetic multilevel strength models for the steel under study and to certify two-dimensional software packages.

Explosive experiments are performed to determine the conditions of incipient and developed spall fracture in wedge samples of a U-1.5% Mo alloy. All experimental data are described by one empirical dependence, which relates the tensile stresses in a spall plane to the stress gradient in a tension wave and the amplitude of shock-wave compression preceding tension.

It is shown that substantial changes in the average grain size (by two orders of magnitude) and a twofold increase in the quasi-static yield stress and strength for uranium and 1.3-fold increase for the U-0.3% Mo alloy did not lead to a change in their shear strength upon a shock-wave loading. There is no correlation between the change in the shear and spall strength and the decrease in the average grain size. A tendency toward a regular increase (other conditions, i.e., amplitudes and the durations of the loading pulse, being equal) in the spall strength of materials in the row “extruded U, extruded U-0.3% Mo alloy, cast U-1.5% Mo alloy” was noted. The increase in the spall strength is connected with alloying and the real content of molybdenum in the alloy rather than with the effect of extrusion.

Results of explosive experiments are presented with time-resolved and space-resolved diagnostics of free surface velocity profiles of wedge samples by optical lever method. The objective of this work is to record free-surface velocity dispersion of samples made of high purity iron with fine and coarse grains, as well as 30KhGSA steel, in delivery state and hardening up to HRc 50. It is demonstrated, that the value of velocity dispersion of free surface of the samples under identical loading modes is varied according to initial microstructure and grain size of investigated metals.

Possible experimental set‐up for investigation of the stage of spalls closing in plates before impact on base plate are considered. Multi‐wave configurations in indicator‐matter situated on the base plate are observed by optical analyzer technique. Oscillograms and results of their processing are presented. Thickness and average density of the spall layer in the plate immediately before its impact on the base plate are estimated.

The new data are presented on relaxation of an elastic precursor in unalloyed depleted uranium and two its alloys. Results were obtained under low-intense explosive loading. Statistic thermofluctuational model was used for approximation of experimental data. The inversion of strength properties of the tested U-Mo and U-Fe-Ge alloys at their quasi-static and high-rate loading was revealed.

Registration results of longitudinal C-L (sigma(XX)) and volume C-B (sigma(XX)) sound velocities in shock-compressed aluminum alloy are presented. Experimental data were obtained in wide range of longitudinal stress, including the stress, corresponding to solid-liquid shock-induced transformation. By using experimentally measured values of sound velocities, the changes of Poisson ratio and shear modulus were calculated along the shock adiabat. These data are needed for calibration of resent elastic-viscous-plastic models.

Data are given that have been obtained by studying five wedge-like samples of armco iron loaded using sliding detonation of layers of explosives of various composition and thickness. While carrying out explosive experiments, information on the structure and parameters of compression and rarefaction waves and on their variation depending on the distance to the loaded surface has been obtained. The data obtained directly during loading are compared with the results of comprehensive investigations of samples that remained intact after shock loading. Good agreement has been revealed between the boundary of transition from three-wave to two-wave configuration upon the loading of armco iron using an HMX-based explosive composition and the position of the boundary of transition from a plateau with high values of microhardness H-mu(x) and Vickers hardness H-V(x) to a region where they decrease sharply, which is related to the transition from the region of the complete alpha-epsilon transformation to a region in which the transformation occurred only partly. Some features of highstrain-rate deformation of armco iron in the region of the initial a phase and in the region of occurrence of the alpha-epsilon transformation have been determined.