Yao Yao

To reduce the dependence of supersulfated cement (SSC) on high-quality ground granulated blast-furnace slag (GGBS), metakaolin (MK) was used to replace 0–40% of GGBS by mass. The effects of MK content on rheology, hydration, microstructure, mechanical properties, and chloride penetration resistance were investigated. Increasing MK replacement increased the dynamic yield stress but reduced the cumulative heat release at 72 h. Ettringite and C-(A)-S-H gel remained the main hydration products, whereas residual gypsum increased at higher MK contents. Among all mixtures, the sample containing 20% MK exhibited the highest chemically bound water content, a relatively dense pore structure, and the lowest electrical flux at 28 d. However, MK replacement of 30% or more coarsened the pore structure, reduced strength, and increased chloride penetration. Therefore, replacing 20% of GGBS with MK reduced slag consumption while maintaining a favorable balance among bound-water content, compressive strength, and electrical flux.

Void evolution driven by electromigration in metallic interconnects is a mechanically constrained moving-boundary problem in which capillarity, electron-wind forcing and elastic energy jointly determine morphological stability. Here we formulate a reduced inverse morphology framework based on Suo-type spectral coordinates, replacing repeated contour remeshing by a low-dimensional modal description of the void boundary. Mechanical coupling is introduced through a finite-element elastic-energy projection onto the Suo basis. This projection defines the stress generalized force and the associated local stress tangent entering the reduced morphology operator. The resulting closure problem is handled by two neural networks with distinct roles: one learns the elastic-energy projection quantities required by the stress-renormalized operator, and the other embeds these quantities in a Suo-PINN inverse model to infer branch identity, proximity to the critical state and uncertainty bounds. In the zero-stress limit, the framework recovers the circular-void benchmark of Suo, Wang and Hao, with a reproduced critical value of(\xi_c=10.649316) and a PINN-estimated value of (\xi_c=10.650000). Stress-coupled predictions are interpreted as trends on a reduced stability manifold rather than as mesh-converged critical loads. The main significance of the method is that it converts a complex void shape into interpretable stability coordinates, providing branch identity, distance to criticality, uncertainty bounds for the inferred driving parameter, and stress-shifted risk as diagnostic outputs. It therefore offers a physically interpretable route for connecting high-fidelity simulations or experimental void images with stress-assisted electromigration failure assessment.

In this study, corner non-functional solder joints were designed as solder joint canaries, and their cycles to failure were compared under four-point bending and thermal cycling tests. A more efficient design and verification method for solder joint canaries was developed. By changing the PCB pad area of the corner solder joints, they were used as canaries for internal functional solder joints, allowing in-situ prediction of internal solder joint failures. To obtain the cycles to failure of solder joint canaries of different sizes, both thermal cycling tests and four-point bending cyclic tests were performed. The four-point bending test provides a more controllable and faster stress loading process. The results show that, compared with the four-point bending test, the thermal cycling test has a longer testing period and shows less distinct regulation effects among different-sized solder joint canaries. To verify the experimental results, a unified creep–plasticity constitutive model coupled with damage was proposed based on a previous model. Using the four-point bending simulation model, the damage evolution during solder joint fatigue was simulated, and the failure behaviors of different-sized solder joint canaries were validated. Further analysis shows that the initial R50 solder joints contain more low-angle grain boundaries and smaller orientation differences compared with the R100 solder joints, resulting in a longer fatigue life. These findings provide important theoretical and methodological support for the efficient design and verification of solder joint canaries.

We report a temperature compensation method for surface acoustic wave (SAW) strain sensors that accounts for the thermal expansion effects of the measured object. This approach effectively reduces the cross-sensitivity error of the SAW sensors at elevated temperatures caused by the thermal expansion of the target substrate. Experiments were conducted using SAW sensors fabricated on 122°X-cut lithium tantalate substrates. The temperature and strain response characteristics of the LT SAW sensor were experimentally analyzed over a temperature range of 25°C to 140°C and the results showed that the resonant frequency decreased linearly with increasing temperature. In addition to the intrinsic temperature response of the LT SAW sensor, the cross-sensitivity error induced by the thermal expansion in the measured object was identified as a major source of measurement error. A dual-sensor differential structure was implemented to quantify the strain interference caused by the thermal expansion of the measured object. A temperature-strain coupling model of the measured object was established using finite element simulation. A resonant frequency correction formula incorporating temperature compensation was proposed. The experimental results show that the measurement error of the sensor using the proposed temperature compensation formula was only 9.13% at 120℃, demonstrating the effectiveness of the method. This study provides theoretical and technical support for high-precision measurement of SAW strain sensors in high temperature environments.