Backanalysis

Backanalysis is a process used to determine unknown parameters of a model by comparing the model's predictions to observed data. Essentially, it involves working backward from the results to find the causes or inputs that produced them. This technique is widely applied in various scientific and engineering fields.

In geotechnical engineering, backanalysis is commonly used to estimate the in-situ stress or material properties of soil and rock masses. For example, if a tunnel deforms in a certain way, backanalysis can be performed using a numerical model of the tunnel to infer the unknown rock properties that led to that observed deformation. Similarly, in structural engineering, observed deflections or strains in a structure can be used to backanalyze the material stiffness or boundary conditions.

The process typically involves an iterative approach. A model is created with initial estimates of the unknown parameters. The model is then run, and its output is compared to the measured field data. If there is a significant difference, the parameters are adjusted, and the model is run again. This cycle continues until the model's predictions closely match the observed data, thus providing an estimate of the unknown parameters. Backanalysis is a powerful tool for understanding complex systems where direct measurement of all parameters is impractical or impossible.