In Russia, AI Will Prevent Metal Fatigue
Russian researchers are combining Soviet-era strength-of-materials science with modern AI to forecast structural defects long before they appear
Russian engineers from South Ural State University have developed a new predictive tool that estimates the lifespan of automotive and aerospace components without destructive physical testing. The methodology merges classical mathematical models of the Soviet scientific school with modern artificial intelligence algorithms, allowing researchers to simulate how micro‑damage evolves into structural cracks under real operational loads.
According to project lead Aleksei Erpalov, the breakthrough lies in accuracy. He noted that the new model calculates durability based on the full spectrum of stresses parts experience in motion, a precision that was unattainable with older empirical methods. The university created a digital test bench—a virtual transport assembly—that simulates vibration and dynamic loads to analyze how micro‑defects grow and determine safe service life.
The system already demonstrates high predictive reliability, aligning closely with physical test results. Researchers plan to adapt the technology for railway equipment, aviation, and industrial machinery. Developed under the Priority 2030 program, the project could accelerate the transition to predictive maintenance, where equipment autonomously signals approaching critical wear, reducing risks and operational costs.
