Cyclic Large Strain and Induced Pore Pressure Models for Saturated Clean Sands

Semiempirical probabilistic models are described to assess cyclic large strain and induced excess pore-water pressure responses of fully saturated clean sands. For this purpose, available cyclic simple shear and triaxial tests were compiled and studied. The resulting ru versus γ, and γ versus N data...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of geotechnical and geoenvironmental engineering 2012-03, Vol.138 (3), p.309-323
Hauptverfasser: Cetin, K. Onder, Bilge, H. Tolga
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Semiempirical probabilistic models are described to assess cyclic large strain and induced excess pore-water pressure responses of fully saturated clean sands. For this purpose, available cyclic simple shear and triaxial tests were compiled and studied. The resulting ru versus γ, and γ versus N databases are composed of 101 and 84 cyclic test data, respectively. Key parameters of the proposed ru and γ models are defined as critical shear strain, relative density, effective confining stress, and equivalent number of loading cycles. Consistent with the maximum likelihood methodology, model coefficients were estimated by maximizing the likelihood function. For comparison purposes, the compiled database was again used to evaluate the performance of existing ru models. Both for comparison and calibration purposes, for each framework, two separate sets of limit-state models were used: model implemented with (1) the original and (2) the updated model coefficients. The model performances are assessed by simple statistics (i.e., mean and standard deviation) of residuals. It is concluded that existing models produce inconsistently biased predictions that vary in the range of 2.5 to 70%. The successes of the proposed and existing models are also assessed for the validation database composed of additional 10 cyclic test results. In addition to (1) repeated improved predictions, (2) differentiating contractive or dilative cyclic soil responses, and (3) incorporation of strain-dependent modulus degradation effects, the main advantage of the proposed methodology is the probabilistic nature of model predictions, which enables the incorporation of the model uncertainty into pore pressure generation predictions.
ISSN:1090-0241
1943-5606
DOI:10.1061/(ASCE)GT.1943-5606.0000631