My short-term research goal is to contribute to the field of structural engineering by focusing on two specific areas:
- Experimental and computational methods for assessment and retrofitting of existing infrastructure
- Performance based design of structural systems against natural hazards
Large Scale Experiments of Structures
In 2017, American Society of Civil Engineers published a comprehensive infrastructure report card for the United States that gave an overall grade of D+. Proposed solutions include smarter investments in research to utilize new approaches, materials and technologies to achieve better infrastructure resilience. A complete revamp of existing infrastructure in the US is economically infeasible. Therefore, a need for smart retrofitting options, aided by innovative structural health monitoring techniques, has risen. I want to focus on developing research in this area because the potential impact of breakthroughs is immediate. My efforts in this area, discussed later, will be aided by my research experience at the PhD level.
As part of my PhD dissertation, I investigated the effects of cold climate exposure on a specific steel bridge substructure column to cap-beam composite connection designed for good seismic performance. This connection was named the grouted shear stud (GSS) connection and is fabricated as a socket type connection. A short steel stub is welded onto the bottom flange of the cap-beam. The inside of the pipe stub is welded with shear studs in a row-column arrangement. This acts as a socket into which an erect pile/column is inserted. The top of this pile/column is also welded with shear studs. The annular region created by this action is then grouted utilizing a high strength grout. The function of this connection is to eliminate the brittle failure mode of weld fracture by relocating the plastic hinge onto the virgin steel section of the pile/column. Fulmer et al. (2015) successfully showed proof of this concept. Alaska Department of Transportation (AKDOT) sponsored this project. My project is a follow up to this work by Fulmer et al. (2015) sponsored by AKDOT.
Cementitious grout materials are not conventionally used for large-scale applications such as the GSS connection. Hence, its behavior in such an arrangement is not well studied. Before implementing the GSS connection in practice, AKDOT wanted to make sure that the connection is durable in cold climate. I tried to approach this problem from two different perspectives. First, at the material level, I investigated the vulnerability of the commercial cementitious grouts to cold climate exposure through a series of standard durability tests. Cold climate durability of concrete material has been studied extensively in the past but no such studies on grouts (where unlike concrete, no coarse aggregate is used) were revealed during the literature search. I performed a series of durability tests as per the ASTM standards on five different commercially available cementitious grouts to assess their vulnerability in cold climate. A general conclusion from the durability studies that followed was that commercial cementitious grouts are notoriously inconsistent in their performance. However, before diving deeper into the material level explanation, I decided to take a step back and look at the bigger picture.
The second perspective therefore was from the structural standpoint. If a damaged connection could still serve the function of transferring the required forces through the connection, one could put the GSS connection to practical use sooner. To answer this, I utilized a novel approach in experimental testing. Multiple large-scale specimens were tested with the independent variable being the degree of damage. Effective states of connection damage due to years of cold climate exposure were simulated artificially by mixing different amounts of expanded polystyrene (EPS) beads with the grout in each test. Addition of EPS beads mimic years of cold climate exposure only in a global sense. For structural performance, the two most important properties of the grout material are its elastic modulus (E) and compressive strength (f’c). EPS addition reduced E and f’c of the grout thereby effectively produced the same structural behavior as material damaged by durability issues. The difference in performance at each level of connection damage was documented, analyzed and quantified. I used the finite element software OpenSees to model my tests computationally. I used this calibrated model to perform a parametric study to obtain fragility functions for the structural system at various levels of damage. Finally, I provided recommendations to AKDOT to aid in informed decision making regarding bridge maintenance.
In addition to establishing a new bridge connection detail for seismic areas, the above research opens up other areas of inquiry. The GSS connection can be easily modified to be utilized as a retrofit option of vulnerable steel bridges. However, before utilizing it in retrofitting applications, some large-scale tests are required to be performed. Although its genesis was in bridge research, the GSS connection can also be implemented in buildings. The plastic hinge relocation concept utilized in the conceptual design of the GSS connection has been gaining traction in the past decade. Apart from conventional strengthening techniques, new materials like fiber reinforced polymer (FRP) are also being investigated to achieve similar results. Having focused on structural testing of steel structures under the guidance of Dr. James Nau, I want to branch out to do research in other materials like reinforced concrete (RC) and FRP. Consistent interactions with my fellow research group members at NC State University under the guidance of Dr. Mervyn Kowalsky and Dr. Rudi Seracino have provided me with sound understanding of the state-of-the-art in these areas. Potential funding sources for this area includes the Transportation Research Board, Federal Highways Association (FHWA), State Departments of Transportations and the American Society of Civil Engineers.
Performance-based Design
The 2018 United Nations report on world urbanization prospects estimates the proportion of world population living in urban areas to go from 55% today to 68% in 2050. Close to 2.5 billion people could be added to world cities. In addition to tackling the needs of their urban population such as housing, energy systems and infrastructure, countries will have to content with the rising stakes of facing natural hazards. A natural hazard can cause larger socio-economic problems in the future compared to the same event in the past. The movement towards performance-based design of structures have drastically improved our capability of estimating both structural and economic consequences of natural hazards. Characterizing hazards within a probabilistic framework and classification of structural damage into performance limit states have allowed engineers better control in design. However, this movement is still in its infancy considering the vast amount of unexplored possibilities. One such possibility for example is the following. Practicing engineers characterize earthquake hazards utilizing peak parameters such as spectral accelerations, spectral displacements etc. while ignoring other ground motion properties such as number of cycles and out of balance nature, which have been proved to affect structural response. As new performance limit states emerge that could potentially capture effects of these new parameters, we need more research to address these issues.
Under the overarching theme of performance-based design, my research will focus on improving both hazard characterization as well as performance limit states for structural systems. My interest in performance-based design sprouted during my masters courses under Dr. Mervyn Kowalsky. As part of an independent study I did under his guidance unrelated to my PhD dissertation, I explored the out of balance nature of earthquake ground motions. During an earthquake, a structure being pushed towards one direction to a significantly higher displacements compared to the opposite direction can have unintended consequences. The following example illustrates this. In reinforced concrete (RC) columns, longitudinal bar buckling is an important performance limit state as the monetary value of repairing it may far outweigh that of replacing it. Structurally, exceeding this limit state is a function of not only the RC section capacity but also the loading history of the ground motion. The state-of-the-practice for design of RC columns for this limit state assumes that the loading is balanced, i.e., the ground motion pushes the structure equally in both direction. However, my research revealed that this is not the case. Extreme values of out of balance ratios of even 5 to 1 were observed. In a journal paper that is currently under review, I proposed a new characterization of ground motion hazard to complement the existing peak parameters that accounts for the out of balance nature of ground motions.
Within the first few years as a faculty, I want to extend natural hazard characterization research in other directions as well. Machine learning techniques such as supervised and unsupervised learning can help improving our understanding of ground motions. The large uncertainty of the processes that cause hazards such as earthquakes make it a suitable addition to the class of problems that machine learning algorithms are trying to solve. Ground motion data collection has improved drastically in the last three to four decades. I believe that producing efficient algorithms will motivate even more investment in obtaining ground motion data, which will further improve solution accuracies. The National Science Foundation (NSF) is currently funding explorations in this area specific to natural hazards.
My long-term goal as of now is to set up a sustainable research group that can tackle multi-facetted problems faced by the nation as well as the world. Specific areas of interest for this group will depend on the outcomes of my short-term research goals. In recent years, barriers between different fields of engineering are collapsing as technologies in diverse areas have started to complement each other. In a few years, I envisage multi-disciplinary collaborations to be the norm and hence it is vital to keep up with emerging technologies in other fields that may be useful in Civil Engineering. This will be the core philosophy of my research group.