Unknown Foundation Determination for Scour (0-6604-1)

Unknown foundations affect about 9,000 bridges in Texas. For bridges over rivers, this creates a problem regarding scour decisions as the calculated scour depth cannot be compared to the foundation depth, and a very conservative costly approach must be taken. The objective was to develop a global approach, which will reduce significantly the level of uncertainty associated with unknown foundations. This approach was developed in two parts: a data mining and inference approach where no testing at the site was necessary, and a testing approach where new tests for unknown foundations were used. The data mining and inference task made use of existing data such as soil type, known foundations on neighboring bridges, design practice, and the age of the bridge to infer the type and length of unknown foundation elements. The testing task consisted of developing two geophysical techniques, resistivity and induced polarization imaging, to obtain a picture of the soil and foundation below the surface level or river bottom. The outcome was a global framework in which one of the approaches or any combination therof, as well as the most useful current techniques (nondestructive testing methods if necessary), can be used to decrease dramatically the uncertainty associated with the unknown foundation. The inference process was trained by using bridges where the foundation was known and verified by comparison against case histories. The two testing techniques mentioned above were tested at the National Geotechnical Testing Site on Texas A&M's Riverside campus and then against full-scale bridges selected in cooperation with TxDOT.