Performance of Methods for Analysis of Relative Floor Elevation Measurements in Residential Structures

Abstract

Forensic evaluations of residential structures often make use of relative elevation surveys of the floor slab. The objective for the analysis of these data is to estimate the postconstruction distortions which have occurred to the floor slab, in order to determine if such distortions might result from excessive movements of the supporting soils. However, such estimates are complicated by the findings of several recent studies that residential slab-on-grade floors exhibit substantial deviations from level at the time of construction, and by the expectation of some postconstruction movements in response to structural loads. In this study, commonly used methods were used to back-analyze relative elevation data from slabs with known distortions as a means to check the performance of these methods. The results strongly suggest that relative elevation surveys may be overemphasized in current practice, as (1) maximum elevation differences significantly overestimate distortions; (2) maximum slopes are quite insensitive to distortions; and (3) pattern recognition approaches are unreliable as an indicator that distortions have occurred.