To investigate the progressive collapse mechanism of reinforced concrete (RC) sub-structure with double-span beams  under blast loading, three vertical impact tests on middle-joint of the RC substructures with double-span beams are carried out, which is aimed to consider the axial tension effect on the remaining sub-structure from the bearing column in failed process under blast loading. The results showed that, at the moment of loading, an obvious reverse arch action appeared in beam due to inertial effect, compressive arch action appeared with decrease of inertial effect and increase of middle joint displacement. Under the impact condition in test, no fracture occurred in the longitudinal bars. The damage mode of beam exhibited a flexural damage at the beam ends. Continuous longitudinal bars at the top of the beam can mitigate the crack development in beam span, and reduce the peak displacement of middle joint in dynamic response of substructure. 

A lot of previous low-velocity impact tests have shown that the failure mode of static flexural-dominant RC members designed by existing standards will transfer to punching shear failure under impact load. To make the mechanism and criterion of punching shear failure clear for RC beams under low-velocity impact, the reliability of the adopted constitutive models, parameters and numerical simulation method is verified by numerically recurring the flexural failure and shear failure modes, impact force-, reaction force-, and mid-span displacement-time histories in six drop hammer tests. The mechanism analysis of punching shear failure was conducted by the verified numerical simulation method to assess the influence on punching shear failure mode by drop hammer weight, sectional dimension, impact velocity, concrete strength, stirrup ratio, longitudinal reinforcement ratio, and shear span ratio. The result shows that the above parameters have a significant influence on the occurrence of punching shear failure for RC beams by changing the local velocity response or dynamic shear capacity of RC beams, except for longitudinal reinforcement ratio and shear span ratio. Further, based on the results of 330 numerical simulated scenarios, the empirical formulas of dynamic shear load V_d,sd and dynamic shear capacity V_d,sc in the key section were proposed, and the explicit criterion of punching shear failure for RC beams was established. The criterion was validated by 71 drop hammer tests (accuracy 90%), which can assist the impact-resistant assessment and design for RC structures.

Steel columns are the important load-carrying members in steel structures. However, there is still a lack of studies on the blast load distribution and dynamic response of steel columns subjected to close-in explosions, especially for the effect of angles of incidence on the blast load and dynamic response of steel columns. In this study, LS-DYNA software is used to investigate the blast load distribution and dynamic response of I-shaped steel columns subjected to close-in explosions under different angles of incidence. The following conclusions can be drawn when the strong axis of the column is loaded with shock wave, the clearing wave effect has a significant influence on the blast load distributed over the flange on the attack side. Compared with the case of an infinite reflecting surface, the reflected impulse on the flange is reduced by 10%-50%. The velocity of the clearing wave in close-in explosions is much higher than the ambient sound velocity assumed in the current blast-resistant design guidelines, which indicates that the ambient sound velocity assumption is not applicable to close-in explosions. When the shock wave impacts on the weak axis, the clearing effect is insignificant. Due to the multiple reflections of the shock wave in the region between the flanges and the web, the positive phase duration is enlarged. This gives a rise to a higher maximum impulse on the web than that in the case of an infinite reflecting surface. For α<α₀(α₀ is the critical angle of incidence which goes through the end of the flange on the attack side to the center of the I-shaped cross section), the main action zone of the shock wave is the flange-web-flange region. The blast load and local deformation of the web achieve the maximum at 30°. The two half-flanges open outwards, and the opening amplitude decreases with the angle of incidence. For α>α₀, the shock wave is mainly loaded on the flange. Its blast load and global deflection reach the maximum at 90°. The two half-flanges close inwards, and the closing amplitude increases with the angle of incidence.

During the testing of scaled models to simulate a nuclear explosion using a chemical explosion, it is necessary to implement specific measures such as restraint and sealing to prolong the duration of the chemical explosion. Although these measures ensure geometric similarity in structural deformation, they also amplify the testing costs and complexities, particularly when scaling down the model, resulting in significant inaccuracies. To address this limitation, this paper proposes a novel method for similarity analysis of the dynamic response of steel beams under blast loading. The proposed method’s accuracy was validated through numerical calculations, revealing that geometric similarity can be maintained during the elastic response stage, subject to certain conditions, including the use of the same material for both the scaled model and the prototype, similarity in geometric shape, and equivalence of the equivalent static load between the chemical explosion model and the prototype structure. However, the maximum mid-span displacement fails to satisfy geometric similarity during the plastic response stage. Under the condition that the duration of the scaled model’s explosion load is longer than the minimum action time, the similarity relationship between the model and the prototype can be expressed as the product of the geometric similarity coefficient and the ductility ratio coefficient, with the analytical expression for the latter provided.