Structural instability caused by the fracture and coalescence of rock bridges within rock masses under external loads is a common issue in slope and mining engineering. Quantifying and revealing the precursory characteristics of rock mass fracture is crucial for ensuring the stability of surrounding rock and preventing fracture instability. In order to investigate the precursory characteristics and failure mechanism of fractured sandstone during deformation and failure

a series of uniaxial compression tests was conducted on fractured sandstone specimens with six different ligament angles. The mechanical strength parameters

fracture modes

and multifractal characteristics of acoustic emission (AE) parameters were analyzed

and the time intervals between different AE parameter variances and key fracture points were discussed. The precursor identification and early - warning time based on the time function of AE hit rates and the variance of AE parameters were obtained. The results show that when the ligament angle increases from 0° to 150°

the elastic modulus and peak stress of fractured sandstone exhibit an inverted Gaussian - like trend

decreasing first and then increasing

and reaching the minimum values at a ligament angle of 60°. The failure mode transitions from indirect coalescence dominated by compression - shear cracks to direct coalescence dominated by tension - shear cracks. The spectrum width (∆) exhibits an evolutionary trend of decreasing first and then increasing. The time function of AE hit rates and the variance of AE parameters can effectively identify early - warning points

sub - critical fracture points

unstable fracture points

and final failure points during the deformation and failure process. By comparing the warning times for the three key fracture feature points

it was found that the warning time of the method based on AE energy variance is the shortest

while that of the method based on AE rise time variance is the longest. Further comparisons indicate that the variance of AE energy

as a precursor factor

is more sensitive than the variances of AE count and AE rise time in identifying critical fracture points of rocks.