Announcements

Thesis Defense: Sriram Sankaranarayanan, ““Optimization with mixed-integer, complementarity, and bilevel constraints with applications to energy and food markets”

The Department of Civil Engineering

and

Advisors Sauleh Siddiqui, Asst. Professor
and Amitabh Basu, Assoc. Professor

Announce the Thesis Defense of
Doctoral Candidate

Sriram Sankaranarayanan

Monday, October 15, 2018

9:30am

Latrobe 106

“Optimization with mixed-integer, complementarity, and bilevel constraints with applications to energy and food markets”

In this dissertation, we discuss three classes of nonconvex optimization problems, namely, mixed-integer programming, nonlinear complementarity problems, and mixed-integer bilevel programming.

For mixed-integer programming, we identify a class of cutting planes, namely the class of cutting planes derived from lattice-free cross-polytopes, which are proven to provide good approximations to the problem while being efficient to compute. We show that the closure of these cuts gives an approximation that depends only on the ambient dimension and that the cuts can be computed efficiently by explicitly providing an algorithm to compute the cut coefficients in $O(n2^n)$ time, as opposed to solving a nearest lattice-vector problem, which could be much harder.

For complementarity problems, we develop a first-order approximation algorithm to efficiently approximate the covariance of the decision in a stochastic complementarity problem. The method can be used to approximate the covariance for large-scale problems by solving a system of linear equations. We also provide bounds to the error incurred in this technique. We then use the technique to analyze policies related to the North American natural gas market.

Further, we use this branch of nonconvex problems in the Ethiopian food market to analyze the regional effects of exogenous shocks on the market. We develop a detailed model of the food production, transportation, trade, storage, and consumption in Ethiopia, and test it against exogenous shocks. These shocks are motivated by the prediction that teff, a food grain whose export is banned now, could become a super grain. We present the regional effects of different government policies in response to this shock.

For mixed-integer bilevel programming, we develop algorithms that run in polynomial time, provided a subset of the input parameters are fixed. Besides the $\Sigma^p_2$-hardness of the general version of the problem, we show polynomial solvability and NP-completeness of certain restricted versions of this problem.

Finally, we completely characterize the feasible regions represented by each of these different types of nonconvex optimization problems. We show that the representability of linear complementarity problems, continuous bilevel programs, and polyhedral reverse-convex programs are the same, and they coincide with that of mixed-integer programs if the feasible region is bounded. We also show that the feasible region of any mixed-integer bilevel program is a union of the feasible regions of finitely many mixed-integer programs up to projections and closures.

Undergraduate Research Positions on the Mechanical Properties of 3D Printed Lattices

Undergraduate Research Positions on the Mechanical Properties of 3D Printed Lattices

Positions are currently available for undergraduate research assistants to work on a collaborative project between the Extreme Mechanics of Architected Materials group (Prof. Stavros Gaitanaros), the Topology Optimization group (Prof. James Guest) and the Applied Physics Laboratory (Dr. Elizabeth Congdon).

The project focuses on the mechanical properties of brittle cellular materials synthesized through additive manufacturing. Our goal is to combine experiments, modeling and design algorithms to create stronger architected materials that can sustain extremely high temperatures and compressive loads. These lightweight materials are ideal for thermal shielding of space structures and could be used in future space missions.

Interested students should contact Prof. Stavros Gaitanaros (stavrosg@jhu.edu) for more information.

 

The Johns Hopkins University is an affirmative action/equal opportunity employer and welcomes applications from women and members of underrepresented groups.

Post-Doctoral Fellowship in Bio-secure Mobility: Network Modeling of Infectious Diseases

Post-Doctoral Fellowship in Bio-secure Mobility: Network Modeling of Infectious Diseases

Applications are invited for a full-time postdoctoral position at the Department of Civil Engineering and the Center for Systems Science and Engineering at Johns Hopkins University under the supervision of Associate Professor Lauren Gardner. The applicant will be expected to undertake and develop research on the topic of biological incident risk, with a focus on the development of models for i) outbreak prediction, ii) inferring outbreak risk factors, and iii) optimizing resource allocation for outbreak mitigation. Candidates should have expertise in one or more of the following areas: network modelling, optimization, machine learning, statistical modelling, and data visualization, with previous experience working on epidemiological applications.

The postdoc will work closely with an international multidisciplinary team of faculty and Ph.D. students from a number of institutions. In addition, the PI will make every effort to mentor the postdoc for transition into a faculty position. This includes guidance on grant-writing, teaching opportunities, and translation of research. Women and Underrepresented Minorities are highly encouraged to apply. This is a year-long postdoc that can potentially be extended up to two years upon satisfactory performance and availability of funding.

SELECTION CRITERIA

The candidate will be expected to:

  • Possess a PhD degree in computer science, engineering, applied or computational mathematics, or a closely related field.
  • Have expertise in one or more of the following areas: network modelling, machine learning, statistical modelling, data visualization.
  • Previous experience working on epidemiological applications.
  • Experience with GAMS, MATLAB, Python, or other programming languages.
  • Demonstrated experience in analyzing large scale data sets.
  • Demonstrated experience in working on large-scale multi-disciplinary projects
  • The ability to work effectively as part of a multi-disciplinary, regionally dispersed research team
  • Illustrate the motivation and discipline to carry out autonomous research.
  • High level interpersonal, written and oral communication skills in English.
  • A record of research accomplishment as reflected in publications in peer-reviewed journals and conferences and presentations at scientific meetings.

Anticipated start date is March 1, 2019. Review of applications will begin immediately and continue until the position is filled. Complete applications should include the following (in a single pdf file) to l.gardner@jhu.edu:

(1) A cover letter

(2) A full curriculum vitae

(3) Up to two research publications and/or preprints

(4) The names and contact information for three references

(5) (Optional) A one-page original research proposal on the topic of your choosing, with the following headings: Motivation, Research Questions, Research Approach, Methods, Data Sources, Timeline. Proposals will be judged on creativity and originality, so think big!

The Johns Hopkins University is an affirmative action/equal opportunity employer and welcomes applications from women and members of underrepresented groups.

Announcing the Fall 2018 Graduate Seminar Speakers

Below is a listing for the speakers being showcased during the Fall 2018 Civil Engineering Graduate Seminar Series. All civil engineering graduate seminars are FREE and open to the public. Attendance is required for all enrolled Civil Engineering graduate students. For information on individual seminars, please refer to the Events Calendar.

For directions and information on parking please see Maps & Directions link at www.jhu.edu and select information on Homewood Campus.

Thesis Defense: Ahmad Shahba, “Crystal Plasticity Finite Element Simulation of Deformation and Fracture in Polycrystalline Microstructures”

THE DEPARTMENT OF CIVIL ENGINEERING

and

ADVISOR, SOMNATH GHOSH, PROFESSOR

ANNOUNCE THE THESIS DEFENSE OF

Doctoral Candidate

Ahmad Shahba

Tuesday, September 4, 2018

12:00pm (noon)

Malone 228

“Crystal Plasticity Finite Element Simulation of Deformation and Fracture in Polycrystalline Microstructures”

 

The mechanical response of metals and their alloys are governed by the deformation mechanisms in the underlying microstructure. High-fidelity modeling of deformation in metals requires development of proper constitutive laws at single crystal scale. Image-based crystal plasticity FE framework is regarded as one of the most powerful tools for deformation simulations, allowing the modelers to explicitly represent the elastic and plastic anisotropy of the material using physics-based laws in a computational domain which statistically represents the morphological and crystallographic properties of the microstructure.

In this work, a thermodynamically-consistent coupled crystal plasticity-crack phase field framework is derived to model fracture process in polycrystalline microstructures. The governing differential equations for the displacement and crack phase field are coupled via the Helmholtz free energy density (HFED). Using the volumetric-deviatoric decomposition of the elastic deformation gradient, a new HFED formulation is proposed which respects the unilateral damage conditions (tension-compression asymmetry of material response in the presence of cracks) and can be used for modeling fracture in anisotropic media under finite deformation conditions.

Numerical modeling of fracture is computationally daunting, partly due to the frequent convergence issues and occurrence of instabilities. Recognizing that the instabilities take place due to an excess energy, three viscous stabilization methods are proposed in this work to dissipate this excess energy and effectively overcome the instabilities. Unlike arc-length methods, the viscous stabilization is applicable for rate-dependent constitutive models and its implementation into any existing FE code is straightforward.

Crystal plasticity simulations of polycrystalline are generally carried out with linear tetrahedral elements due to their capability in conforming to complex geometries. These elements are known to suffer from volumetric locking in modeling (nearly-) incompressible materials, leading to numerical artifacts such as underestimation of displacements and overestimation of pressure levels. A modified F-bar-patch technique is developed in this work to alleviate volumetric locking in phase field modeling of ductile fracture.

In the course of plastic deformation, the local strain rate experienced by different material points in the microstructure could be orders of magnitude different from the applied macroscopic strain rate. It is of paramount significance to develop a unified crystal plasticity law which could be applied for a wide range of strain rates. Using the dislocation glide mechanisms in hcp metals, a unified flow rule is developed by combining the thermally-activated and drag-dominated processes. This unified law can be employed to model deformation over a wide range of strain rates and its explicit dependence of temperature makes it suitable for modeling high rate deformation of metals where adiabatic heating is significant.

Thesis Committee: Somnath Ghosh, James Guest, Jaafar El-Awady

Thesis Defense: Zhiye Li, “Micromechanical Studies of Glass Fiber Reinforced Epoxy Matrix Composites Undergoing Deformation and Damage at High Strain Rates”

THE DEPARTMENT OF CIVIL ENGINEERING

AND

ADVISOR SOMNATH GHOSH, PROFESSOR

ANNOUNCE THE THESIS DEFENSE OF

Doctoral Candidate

Zhiye Li

Friday, August 31, 2018

3:00pm

Latrobe 106

“Micromechanical Studies of Glass Fiber Reinforced Epoxy Matrix Composites Undergoing Deformation and Damage at High Strain Rates

This study develops an experimentally calibrated and validated 3D finite element model for simulating strain-rate dependent deformation and damage behavior in representative volume elements of S-glass fiber reinforced epoxy matrix composites. The fiber and matrix phases in the model are assumed to be elastic with their interfaces represented by potential-based and non-potential, rate-dependent cohesive zone models. Damage, leading to failure, in the fiber and matrix phases is modeled by a rate-dependent non-local scalar CDM model. The interface and damage models are calibrated using experimental results available in the literature, as well as from those conducted in this work. A limited number of tests are conducted with a cruciform specimen that is fabricated to characterize interfacial damage behavior. Validation studies are subsequently conducted by comparing results of FEM simulations with cruciform and from micro-droplet experiments. Sensitivity analysis are conducted to investigate the effect of mesh, material parameters and strain rate on the evolution of damage. Furthermore, their effects on partitions of the overall energy are also explored. Finally the paper examines the effect of microstructural morphology on the evolution of damage and its path.

Apart from exploring the damage mechanism, this study examines the effectiveness of periodic boundary conditions (PBCs), when applied to heterogeneous representative volume elements (RVEs) subjected to high strain-rate loading conditions. Even for a periodic multi-phase microstructure, the local stress and strain responses in the RVE under conditions of high strain-rate are not periodic. Stress waves propagate in the microstructure and interact with heterogeneities, resulting in reflection and transmission at the interfaces. To mitigate the limitations of PBCs, space and time dependent boundary conditions (STBCs), derived from analytical solutions to the 1D wave propagation problem, are proposed in this paper. This results in significant increase in the efficiency of the RVE analysis since it is not necessary to include larger RVEs. The paper introduces analytical solutions of the longitudinal and shear wave equations for elastic two-phase materials under time dependent boundary conditions. Subsequently a 3D composite RVE problem, is solved to investigate the efficacy of STBC. Results show that STBCs significantly improve the accuracy over PBCs for the same RVE. From this analysis, a strain-rate ≥ 10E5 per second is considered to be suitable transition point from periodic to space-time dependent boundary conditions for heterogeneous elastic composites.

In the end, this study proves that effective material properties are also influenced by stress wave propagation and strain rate. The relation of wavefront motion and macroscopic stiffness of RVE, and strain rate effect of the composites material properties are investigated. Various simulations at different strain rates are conducted to show that the homogenized models proposed in this study have considerable accuracy.

Thesis Defense: Jiaxin Zhang, “Uncertainty Quantification from Small Data: A Multimodel Approach”

THE DEPARTMENT OF CIVIL ENGINEERING

AND

ADVISOR MICHAEL SHIELDS, ASSISTANT PROFESSOR

ANNOUNCE THE THESIS DEFENSE OF

Doctoral Candidate

Jiaxin Zhang

Tuesday, August 7, 2018

9:00am

Hodson 311

“Uncertainty Quantification from Small Data: A Multimodel Approach”

As a central area of computational science and engineering (CSE), uncertainty quantification (UQ) is playing an increasingly important role in computationally evaluating the performance of complex mathematical, physical and engineering systems. UQ includes the quantification, integration, and propagation of uncertainties that result from stochastic variations in the natural world as well as uncertainties created by lack of statistical data or knowledge and uncertainty in the form of mathematical models. A common situation in engineering practice is to have a limited cost or time budget for data collection and thus to end up with sparse datasets. This leads to epistemic uncertainty (lack of knowledge) along with aleatory uncertainty (inherent randomness), and a mix of these two sources of uncertainties (requiring imprecise probabilities) is a particularly challenging problem.

A novel methodology is proposed for quantifying and propagating uncertainties created by lack of data. The methodology utilizes the concepts of multimodel inference from both information-theoretic and Bayesian perspectives to identify a set of candidate probability models and associated model probabilities that are representative of the given small dataset. Both model-form uncertainty and model parameter uncertainty are identified and estimated within the proposed methodology. Unlike the conventional method that reduces the full probabilistic description to a single probability model, the proposed methodology fully retains and propagates the total uncertainties quantified from all candidate models and their model parameters. This is achieved by identifying an optimal importance sampling density that best represents the full set of models, propagating this density and reweighting the samples drawn from the each of candidate probability model using Monte Carlo sampling. As a result, a complete probabilistic description of both aleatory and epistemic uncertainty is achieved with several orders of magnitude reduction in Monte Carlo-based computational cost.

Along with the proposed new UQ methodology, an investigation is provided to study the effect of prior probabilities on quantification and propagation of imprecise probabilities resulting from small datasets. It is illustrated that prior probabilities have a significant influence on Bayesian multimodel UQ for small datasets and inappropriate priors may introduce biased probabilities as well as inaccurate estimators even for large datasets. When a multi-dimensional UQ problem is involved, a further study generalizes this novel UQ methodology to overcome the limitations of the independence assumption by modeling the dependence structure using copula theory. The generalized approach achieves estimates for imprecise probabilities with copula dependence modeling for a composite material problem. Finally, as applications of the proposed method, an imprecise global sensitivity analysis is performed to illustrate the efficiency and effectiveness of the developed novel multimodel UQ methodology given small datasets.

Post-Doctoral Fellowship in Optimization of Energy and Food Systems

Post-Doctoral Fellowship in Optimization of Energy and Food Systems

Applications are invited for a full-time postdoctoral position at the Department of Civil Engineering and the Center for Systems Science and Engineering at Johns Hopkins University under the supervision of Sauleh Siddiqui at the Mathematical Optimization for Decisions Lab (MODL). The postdoc will spend 50% of their effort on existing projects involving equilibrium problems, bilevel optimization, and machine learning with applications to energy and food systems. The other 50% of the postdoc’s effort will be spent on a project of the postdoc’s choosing, for which a one-page application will be required. The details of the one-page application are below, but the proposal must be highly creative, original, and pertain to at least one of the topics mentioned in this paragraph.

The postdoc will work closely with an international multidisciplinary team of faculty and Ph.D. students from a number of institutions. In addition, the PI will make every effort to mentor the postdoc for transition into a faculty position. This includes guidance on grant-writing, teaching opportunities, and translation of research. Women and Underrepresented Minorities are highly encouraged to apply. This is a year-long postdoc which can potentially be extended up to two years upon satisfactory performance and availability of funding.

Requirements:

– PhD degree in applied or computational mathematics, computer science, engineering, economics, or a closely related field.

– Research experience in one or more of the following fields: optimization, equilibrium, machine learning.

– Experience with GAMS, MATLAB, Python, or other programming languages.

– Demonstrated ability to work independently as well as collaboratively with excellent written and oral communication skills.

– Interdisciplinary research experience is preferred

– A record of research accomplishment as reflected in publications in peer-reviewed journals and conferences and presentations at scientific meetings.

Anticipated start date is September 1, 2018. Review of applications will begin July 10th, 2018 and continue until the position is filled. Complete applications should include the following (in a single pdf file) to Siddiqui@jhu.edu:

(1) A cover letter

(2) A full curriculum vitae

(3) Up to two research publications and/or preprints

(4) The names and contact information for three references

(5) (Optional) A one-page original research proposal with the following headings: Motivation, Research Questions, Research Approach, Methods, Data Sources, Timeline. Proposals will be judged on creativity and originality, so think big!

Thesis Defense: Hwanpyo Kim “Simulation of non-Gaussian/non-stationary stochastic processes: beyond second-order orthogonality”

THE DEPARTMENT OF CIVIL ENGINEERING

AND

ADVISOR MICHAEL SHIELDS, ASST. PROFESSOR

ANNOUNCE THE THESIS DEFENSE OF

Doctoral Candidate

Hwanpyo Kim

Tuesday, April 24, 2018

12:00pm

Malone 107

“Simulation of non-Gaussian/non-stationary stochastic processes: beyond second-order orthogonality”

The theory of stochastic processes and their generations are indispensable to characterize wind fluctuations, ocean waves, and earthquake excitations among other quantities in engineering. To computationally analyze and simulate these stochastic systems, practical realization of samples of stochastic processes is essential. The object of this thesis is to introduce new state-of-the-art methodologies for the generation of stochastic processes with non-Gaussianity/non-stationarity possessing higher-order properties than the second-order orthogonality.

A new type of Iterative Translation Approximation Method (ITAM) using the Karhunen-Loève expansion was developed for simulating non-Gaussian and non-stationary processes utilizing translation process theory. The proposed methodology enhances the accuracy of simulated processes in matching a prescribed autocorrelation, maintains the computational efficiency, and resolves limitations caused by utilizing evolutionary power spectra for non-stationary processes.

A new generalized stochastic expansion, the bispectral representation method (BSRM), expanded from the traditional spectral representation method is introduced to simulate skewed nonlinear stochastic processes. With new orthogonal increments to satisfy the conditions of the Cramér spectral representation up to third order orthogonality, the BSRM generates samples that match both the power spectrum and bispectrum of the process by modeling complex nonlinear wave interactions.

A model of phase angle distributions to characterize phase coupling in higher-order stochastic processes is presented. Relationships between the trigonometric moments of circular distributions of phase differences and higher-order cumulant spectra are derived. The prescribed properties are shown to accurately model quadratic and cubic phase couplings in simple stochastic processes and can easily be extended to general n-wave couplings.

Lastly, as applications of the prescribed methods, wind pressure and turbulent wind velocity time histories are generated with SRM, ITAM, and BSRM and applied to two different nonlinear dynamic structural systems. For structures having material and geometrical nonlinearities, performance of an elastic perfectly-plastic structure and the buffeting response of a long-span bridge with coupled aerodynamic forces are examined. The structures are investigated to observe the effect of higher-order properties of the excitations on the response when compared to conventional second-order Gaussian and non-Gaussian excitations.

Thesis Defense: Wei Jang, “Machine Learning and Optimization for Healthcare and Energy Systems”

The Department of Civil Engineering
and
Advisor Sauleh Siddiqui, Assistant Professor
Announce the Thesis Defense of
Doctoral Candidate

 

Wei Jiang

Thursday, March 22, 2018

2:00-4:00pm

Latrobe 106

 

“Machine Learning and Optimization for Healthcare and Energy Systems

 

Abstract:

Healthcare and energy systems provide critical service to our society. Recent advancement in information technology has enabled these systems to keep retrieving and storing data. In this dissertation, we used machine learning, optimization techniques, and data from healthcare and energy systems to build predictive models and discover new knowledge to guide decision-making and improve the efficiency and sustainability of these systems. We also used optimization techniques to improve the efficiency of hyperparameter tuning for machine learning algorithms. Specifically, we built a dynamic daily prediction model for predicting heart failure patients’ 30-day readmission risk. We built a prediction model to predict xerostomia (dry mouth) for head and neck cancer patients treated with radiotherapy and identified the influence pattern of radiation dose across head and neck on xerostomia. Using an economic equilibrium model combined with optimization techniques for calibration, we built the first global trade model for wood chip and analyzed how local renewable energy policy in the United States could affect the global wood chip trade and lead to deforestation in other world regions. Finally, we created a new method for tuning the hyper-parameter for support vector machines by solving the problem as a bilevel optimization problem using stochastic gradient descent combined with dual coordinate descent method. We showed that the new method is more efficient than ad hoc empirical approaches. In summary, we demonstrated how machine learning and optimization techniques can improve the efficiency of healthcare and energy systems, and how optimization techniques can advance machine learning algorithms.

 

 

Back to top