Consolidation characteristics of soft sediments by seepage induced consolidation test

The thesis presented herein outlines the seepage-induced consolidation test (SICT) results of soft sediments dredged from Newark bay, NJ, for the determination of consolidation characteristics. The Seepage induced consolidation testing device was constructed based on the original version (SICT) presented by Dr. Prof. Dobroslav Znidarcic at the University of Colorado Boulder (CU). The author traveled to CU and consulted Dr. Znidarcic for recommendations on the device design and construction, preferred manufacturers, and sample testing procedures. A SICT device was then designed and constructed for Rutgers University in New Brunswick, with some modifications like deadweight loading system for easier troubleshooting. The SICT results for kaolinite clay, performed by Dr. Znidarcic at the CU laboratory, were compared with the author's test results from the Rutgers Soil and Soft Sediment Laboratory. The comparison served to assess the reliability and repeatability of the Rutgers University SICT device. Following the collection of dredged sediment samples from the Newark bay area, five different clays were purchased and their consolidation characteristics were tested. In addition, the tests sought to determine a correlation between sediment index properties and consolidation model parameters. Index property determination is less time consuming and cumbersome than consolidation testing under seepage force, which is more technical and can take up to two weeks for a single test. Therefore, there is inherent merit in gaining insight into consolidation behavior without running the seepage tests. A total of sixteen tests were performed. Eleven tests were conducted on the Newark soft sediments, and the remaining five on the clay samples. The results have been partially published via the 15th Asian Regional Conference in Japan, the International Conference on Civil Engineering in Iran, and the Journal of Geotechnical Engineering. The consolidation model parameters were correlated with the samples' index properties and the resulting equations satisfactorily approximated the reality.Ph.D.Includes bibliographical referencesby Masoud Janba