MnROAD | NRRA | Structure & Teams | Geotechnical Team
Overpredicted Resistances of Non-Displacement Piles in Sands Using Static Analysis Methods
Status: In development
Project overview
On several bridge projects in Minnesota, the non-displacement piles were installed much longer than their estimated lengths. This resulted in costly change orders due to the costs and delays associated with procuring and fabricating additional piling during construction. The non-displacement piles were primarily H piles and open-ended pipe piles. The subsurface conditions at the projects primarily consisted of loose to medium dense sands without a definitive hard bearing layer, so the piles derived most of their resistance from side resistance (skin friction). The pile resistance and desired lengths were estimated using FHWA’s recommended Nordlund Method. Some possible reasons for the overpredicted resistances include limitations with the Nordlund method in the project-specific subsurface conditions, difficulty in determining soil-pile interface friction, effects of H pile points, open-ended pipe pile cutting shoes or rings, pile plugging, and/or driving strength loss (pile setup), among others.
The goal of this research is to determine the cause of the overpredicted resistances of the non-displacement piles in the project-specific subsurface conditions and either recommend modification factors that can be applied to the Nordlund method or recommend another static analysis method (such as the Beta or API method) to accurately estimate pile lengths. To achieve these goals, the existing data (where the resistance of non-displacement piles in sands was overpredicted) as well as data from two new projects (where MnDOT plans to drive non-displacement piles) will be systematically analyzed in this study. Regression methods and sensitivity analyses, as well as numerical simulations, will be used to better understand the reason for the overprediction of the pile resistance, and modify the existing static analysis models to improve the pile resistance predictions in sands.
Tasks
Task 1: Literature review
A comprehensive literature review will be conducted at this stage to better understand the problem and the potential causes of the resistance overprediction. The advantages and disadvantages of other static analysis methods will also be systematically reviewed. In addition to scientific journal/conference publications, the literature review will include any available technical reports from other DOTs with similar issues and/or cases where the resistance of non-displacement piles in sands was accurately predicted.
- Deliverable: A comprehensive report on the literature review
- Date due: 1/31/2026
Task 2: Collecting and analyzing the existing data with overpredicted pile resistance
The existing data from previous MnDOT projects (and other DOTs if possible) where the resistance of non-displacement piles in sands was overpredicted will be collected and thoroughly analyzed in this task. First, the Nordlund Method will be used to estimate the resistance of each pile and the required length of the pile based on this calculation. The calculated resistances will be compared to the actual resistance determined after the construction. Regression analyses will be used to determine if the Nordlund method can be modified to achieve more reliable pile resistance estimations. H pile points, open ended pipe pile cutting shoes or rings, pile plugging, and/or driving strength loss (pile setup) will also be analyzed to determine their effect on the actual resistance after construction.
In the Nordlund method, the resistance of a pile with a uniform cross-section is a function of the friction angle of the soil, the friction angle of the soil/pile interface, the effective unit weight of the soil, the pile length, the pile perimeter, and the volume of soil displaced. Sensitivity analyses will be conducted to determine which factors have more influence on the accuracy of the predictions. The results of the sensitivity analyses will help improve the required modifications by focusing the modifications on the factors with the highest influence on the predictions. The sensitivity analyses could also help better explain the reason for the overprediction of the pile resistance in the first place.
Next, the pile resistance of all the cases examined above will be calculated using other available static analysis methods, e.g., Brown Method, Eslami and Fellenius Method (if CPT data is available), Nottingham and Schmertmann Method (if CPT data is available), Beta method, LCPC Method (if CPT data is available), and API method. The results of these methods along with the modified Nordlund method proposed above will be compared to the actual pile resistance calculated after the construction in the field to determine the best method for designing non-displacement piles in sands. If needed, modifications to these methods will also be proposed similar to the ones proposed for the Nordlund method. All the calculations will be conducted using Excel and APile software.
- Deliverable: A comprehensive report will be prepared including all the pile resistance calculations based on all the typical statics methods and proposed modifications (the modifications might not be final at this stage and they will be fine-tuned at the end of task 4).
- Date due: 6/30/2026
Task 3: Driving new non-displacement piles at two new projects in Minnesota
To ensure the methods proposed in task 2 for designing non-displacement piles in sands are generalized and not biased due to any hidden biases in the limited available data, two new projects where MnDOT plans to drive non-displacement piles in Minnesota will be chosen by the research team and the TAP members. The information obtained from the previous tasks, especially the sensitivity analyses, will be used to guide the remaining site investigations, design the pile, and drive the piles. The accuracy of the proposed methods (the modified Nordlund method or the best method determined in task 2) will be evaluated in these two projects. If needed, additional modifications will be proposed based on the results of these projects.
- Deliverable: A comprehensive report will be prepared including the results of the two new pile-driving projects and any changes in the modifications proposed for Nordlund and/or other methods.
- Date due: 12/31/2026
Task 4: Three-dimensional (3D) numerical simulations
In this stage of the project, a 3D finite volume model of non-displacement piles in sands will be developed using the commercial FLAC3D software (Michigan Tech has the latest version of the software, i.e., V9). The model will be calibrated (validated) based on the results of field data used in tasks 2 and 3. The calibrated model will then be used to simulate the pile resistance under different conditions that have not been considered in previous tasks due to the lack of field data. These conditions could include more complicated subsurface soil profiles not encountered in previous projects, or cases where any anticipated future changes in the conditions of previous projects (e.g., change in groundwater table) are taken into account. The conditions to be investigated will be determined based on the findings of the previous tasks and after consulting the TAP members. The best method proposed after tasks 2 and 3 will be used to predict the resistance of the piles in each of these conditions and the results will be compared to the resistance calculated from the 3D numerical simulation. The calibrated model will also be used to conduct more comprehensive sensitivity analyses which helps to gain a better understanding of the effects of different conditions on the accuracy of predicted pile resistances. If needed, additional modifications (to Nordlund and/or other methods) will be proposed based on the results of these simulations and sensitivity analyses.
- Deliverable: A comprehensive report will be prepared including the results of all the simulations, sensitivity analyses, and any changes in the modifications proposed for Nordlund and/or other methods.
- Date due: 1/31/2027
Task 5: Editorial review and publication of Final Report
All the findings of the research project including the proposed best method to accurately determine the pile resistance of non-displacement piles in sands will be synthesized in a final report. The final report along with a project summary report and an implementation report will be prepared and submitted to the NRRA and MnDOT.
- Deliverable: Final Publishable Report that meets MnDOT’s editorial guidelines and standards, a project summary report, and an implementation report will be submitted..
- Date due: 2/28/2027
Project team
Email the Project Team
Principal Investigator: Mohammadhossein Sadeghiamirshahidi, Ph.D., Michigan Technical University, msadeghi@mtu.edu
Technical Liaison: Joseph Nietfeld, MnDOT, joe.nietfeld@state.mn.us
Technical Advisory Panel (TAP): Contact us to join this TAP
- Ceren Aydin, MnDOT
- Ali Khosravi, NCAT
- Deepak Maskey, Caltrans
- Joseph Nietfeld, MnDOT (TL)
- Supraja Reddy, Illinois Tollway
- Ryan Snook, Michigan DOT
- Colter Schwagler, ND DOT
- Tri Tran, Wisconsin DOT