This study presents a non-explosive method for simulating blast loading on reinforced concrete (RC) slabs. The method involves using a newly developed blast simulator to apply a speedy impact load on the slab, which generates a pressure wave similar to that of an actual blast. Both experimental and numerical simulations were carried out to evaluate the effectiveness of the method. The experimental results showed that the non-explosive method can produce a pressure wave with a peak pressure and duration analogous to those of an actual blast. The numerical simulations also showed good agreement with the experimental results. Additionally, parameter studies were conducted to evaluate the effects of the rubber shape, the impact velocity, the bottom thickness, and the upper thickness on the impact loading. The results indicate that pyramidal rubber is more suitable as an impact cushion for simulating blast loading than planar rubber. The impact velocity has the widest range of regulation for peak pressure and impulse. As the velocity increases from 12.76 to 23.41 m/s, the corresponding range of values for peak pressure is 6.457 to 17.108 MPa, and for impulse, it is 8.573 to 14.151 MPa∙ms. The variation in the upper thickness of the pyramidal rubber has a more positive effect on the impact load than the bottom thickness. With the upper thickness increasing from 30 mm to 130 mm, the peak pressure decreased by 59.01%, and the impulse increased by 16.64%. Meanwhile, when the bottom part’s thickness increased from 30 mm to 130 mm, the peak pressure decreased by 44.59%, and the impulse increased by 11.01%. The proposed method provides a safe and cost-effective alternative to traditional explosive methods for simulating blast loading on RC slabs.

Blast simulators are capable of applying blast-like loading to components in a safe and controlled laboratory environment, overcoming the inherent shortcomings of blast testing in terms of data acquisition, test cycle time, and cost. In this paper, reasonable assumptions and refinements are made to the components and shape of the impact module, a key component of the blast simulator, to achieve diversity in simulated blast loading. By designing four rubber shapes, the importance of ellipsoid rubber as an elastic cushion for simulating blast loading was determined. In order to assess the effectiveness of this optimization, numerical calculations based on a calibrated finite element model were performed around four factors: flat rubber thickness, ellipsoid rubber thickness, impact velocity, and impact modulus mass. Additionally, a grey correlation sensitivity analysis was carried out to evaluate the effect of these factors on the impact loading on the reinforced concrete (RC) slab. The results indicate that peak pressure and impulse had opposite sensitivities to velocity and mass. Changes in ellipsoid rubber thickness had a more positive effect on the impact loading than flat rubber thickness. An in-depth study of the role of these influencing factors is important for the design and improvement of impact modules.

Word spelling is a fundamental skill at the center of K-12 English education. While there is a lot of academic research and commercial software on handwriting recognition, little effort has been made to recognize misspelled words, which is an important step toward AI grading. One significant challenge is the infinite possibility of wrong spelling but extremely limited training data compared to correct words. We observe that the current popular OCR algorithms such as CRNN and CNN-Ngram suffer prediction shift phenomenon. To this end, we propose a spelling checking framework consisting of self-supervised pre-training, supervised fine-tuning, and post-processing. To address the long-tailed balancing problem, we propose a novel data generation and a novel dataloader. Experimental results on the misspelling dataset demonstrate the effectiveness of the proposed framework, and the generated and augmented data both contribute to improving the performance.

This study employs a drop-hammer blast simulator to conduct eight impact tests on reinforced concrete (RC) slabs. Experimental results demonstrate that the peak pressure and duration exhibit waveform characteristics similar to those of actual blast loads. Numerical simulations further validate strong agreement with experimental data. Through dimensional analysis, a functional expression correlating parameters with impact load is derived. Orthogonal design methodology optimizes parameter combinations, while experimental and numerical results jointly determine unknown coefficients to establish a predictive model for impact load. Key findings indicate that impact velocity exerts the most extensive influence on peak pressure and impulse adjustment. Variations in pyramid rubber height demonstrate more significant effects on impact load than flat-layer thickness. The proposed method offers a safe and cost-effective alternative to conventional explosive testing for simulating blast loads on RC slabs.