Abnormal strain and structural cracks are common precursors to engineering accidents in concrete structures. However, traditional strain and crack detection methods exhibit a lag, and discovering structural abnormalities in a timely manner incurs high labor costs. Consequently, strain self-sensing cement-based materials have garnered widespread attention from researchers due to their intelligent and automated characteristics. Nevertheless, current cement-based strain self-sensing materials lack exploration in conduction mechanisms, often failing to establish a good correlation between changes in electrical resistance and system strain. Therefore, in this paper, nickel powder, which exhibits excellent durability and conductivity, is selected as the conductive component for the preparation of cement-based strain self-sensing materials. The study explores its conduction mechanism and various electromechanical constitutive models under different states, providing an important reference for correlating the electromechanical relationship of strain self-sensing materials.

The cement used is Swan Brand P·O 42.5 ordinary Portland cement, and the water reducer is polycarboxylate superplasticizer with a water reduction rate of 18%-29%. The nickel powder adopts micrometer-grade high-purity ultrafine conductive nickel powder, with a Ni purity of 99.999% and a loose bulk density of 1.40-1.68 g/cm³. The execution standard is GB/T 7160—2008.

Mix cement, nickel powder, water, and admixtures, then stir and disperse them before pouring the mixture into molds. Place the molds into a magnetic field coil and let them sit for a day before demolding and curing. During preparation, mix rapidly for 5 min to form Mixture 1. Then add nickel powder (with volume fractions of 0%, 5%, 10%, 15%, 17%, 20%, and 23%) to the cement and mix rapidly until the color of the mixture is uniform and no longer changes. This indicates that the nickel powder is fully dispersed and evenly distributed, forming Mixture 2. Pour Mixture 1 into Mixture 2 and mix rapidly for another 5 min. In this step, the water reducer increases fluidity while allowing re-dispersion of agglomerated cement and nickel powder particles, forming a nickel powder cement slurry called Mixture 3. Place Mixture 3 in a vacuum drying oven, adjust the temperature to 20 degrees Celsius, and vacuum for 15 min to eliminate bubbles and fully compact the mixture. Pour the defoamed Mixture 3 into molds with dimensions of 20 mm×20 mm×40 mm and 20 mm×20 mm×80 mm, and seal with plastic wrap to reduce water loss. Prepare two sets of test pieces for each volume fraction, allowing them to form under magnetic field strengths of 0 Gs (blank control) and 200 Gs for about 24 h. Then let them sit at room temperature for 2 d before demolding. Cure them in a 60 ℃ steaming box for 3 d, then cure under standard conditions for 4 d. Finally, dry them in a vacuum drying oven at 60 ℃ for about 24 h until the weight no longer decreases.

With the increase of the volume fraction of nickel powder, the resistivity changes can be roughly divided into three stages, namely, the stage of slow resistivity decrease, the stage of rapid resistivity decrease, and the stage of resistivity decrease again but at a slow rate. The rapid resistivity decrease stage of nickel powder cement-based materials conforms to the percolation theory, and the percolation threshold is between 15% and 17%. In this paper, the percolation threshold is taken as 15%. Near the percolation threshold, the conduction mechanism accords with the tunneling effect theory. For specimens formed under the action of a 200 Gs magnetic field, their resistivity in the direction along the magnetic field will decrease, while their resistivity in the direction against the magnetic field will increase. The measured maximum ratio between resistivity in the direction against the magnetic field and resistivity in the direction along the magnetic field is close to 4. Based on the tunneling effect and percolation theory, this paper derives an equation representing the change of resistance with strain, namely, the electromechanical constitutive model, and expresses this relationship as a polynomial function of strain. The parameters of the polynomial are determined using piezoresistive experiments. A four-point bending test is used to obtain piezoresistive curves of specimens with a 15% doping level formed under no magnetic field and under a 200 Gs magnetic field when subjected to pure bending loads. Based on the bending resistance variation of specimens with and without a magnetic field, this paper proposes an electromechanical constitutive equation representing the bending resistance variation rate and strain value. Subsequently, a supplementary equation for resistance variation after cracking of pure bending specimens is proposed, forming the basic theory for monitoring strain and cracks in bending members.

The specific conclusions are as follows: The optimal mixing ratio for nickel powder cement is approximately 15%. Around this ratio, the conductive behavior of the material conforms to the tunneling effect theory. Through magnetic field treatment, the resistivity of the material decreases in the direction of the magnetic field and increases in the perpendicular direction. When this material is compressed, its resistance first decreases and then increases. During the elastic stage, the relationship between resistance and strain is nearly linear, with a maximum decrease rate exceeding 60%. The material treated with a magnetic field exhibits a smoother performance under compression. When the material is bent, its resistance also decreases initially and then increases, but the decrease rate is very small, only about 1%. The material treated with a magnetic field demonstrates a more stable resistance change when bent. Based on these experimental results, we have derived the electromechanical constitutive relationship of the material under uniaxial compression and pure bending loads.