A TIME DOMAIN FATIGUE PROBLEM TAYLORED INTO FREQUENCY DOMAIN THROUGH AN OPTIMIZATION APPROACH

Resumo

Time domain is the most broadly choice in fatigue testing to completely represent random events,
happening in numerous synchronous input channels. The primary disadvantage of this approach is the broad testing
span and equipment cost. Time domain-based equipment tests are made from an intricate equipment, which
requires servo engines be working to instigate an amount of load at a particular time window. These tests spend a
huge time cost, since they require a similar length span of the event that they are repeating, times the necessary
reiterations. The frequency domain strategy for fatigue testing requires less intricate equipment, since there is no
requirement for servomotors. Moreover, the test length is reduced, since there is no compelling reason to run the
full event times the necessary reiterations. The current disadvantage is the limitation to represent random events
with different synchronous information channels. Thus, frequency domain tests are mainly applied for basic binary
tests, unfit to represent random events. This work aims to introduce another approach that utilizes the fatigue result
as input, reversing it to discover the loads that would be required to replicate the result in frequency domain. This
can help future developments reducing costs and time required for fatigue testing. Optimization approaches and
computations simulations are applied to solve this inverse problem.