Steven F. Wojtkiewicz

Probabilistic Numerical Simulation of Pavement Performance using MEPDG

ABSTRACT It is widely acknowledged that the accurate simulation of complex engineering systems, such as nuclear power reactors, modern weapon systems, and aircraft, requires probabilistic analysis due to inherent uncertainties in their models parameters. The demand for and the complexity of probabilistic analysis prompted Sandia National Laboratories to develop a versatile software toolkit, DAKOTA, adaptable to various engineering applications. Pavements are another example of a complex engineering system requiring probabilistic modeling due to the uncertain nature of most of the pavement performance models parameters, including traffic, climate, material properties, and pavement structure. The deterministic pavement performance models vary from simplistic empirical relationships to complex mechanistic-empirical computational algorithms. Due to DAKOTAs independence of choice of analysis tool, it is a natural candidate to perform probabilistic aspects of pavement performance prediction. The paper presents a software framework for probabilistic modeling of pavement performance, which combines deterministic performance prediction models from the MEPDG and probabilistic analysis tools from DAKOTA. The power of this approach is demonstrated by analyzing the effect of variability in the asphalt concrete AC mix design on the variability of the pavement performance prediction.

Bridge life extension using semiactive vibration control

This paper focuses on the use of a control system to extend the life of a highway bridge. The safe life of a bridge can be more than tripled if the peak strain levels it experiences are reduced by just 33%. As of 2012, over 5000 bridges in the country have been deemed to be structurally deficient. Hence, the use of a vibration control system to extend the lives of bridges can be of tremendous societal impact. This paper utilizes a dynamic model of the Cedar Avenue tied arch steel bridge in Minnesota to investigate avenues for peak strain reduction. Simulations show that the use of passive structural modification devices such as stiffeners and dampers is inadequate to reduce the key resonant peaks in the frequency response of the bridge. Both active and semiactive vibration control strategies are then pursued. Active vibration control can effectively reduce all resonant peaks of interest, but is practically difficult to implement on a bridge due to power, size, and cost considerations. Semiactive control with a variable orifice damper in which the damping coefficient is changed in realtime using bridge vibration feedback can be practically implemented. Simulation results show that, when employed with multiple devices, the proposed semiactive control system can reduce the response at all critical resonant frequencies. Further analysis reveals that the location and number of actuators on the bridge is critical for controlling these specific resonant frequencies.

Real-time hybrid simulation using the convolution integral method

This paper proposes a real-time hybrid simulation method that will allow complex systems to be tested within the hybrid test framework by employing the convolution integral (CI) method. The proposed CI method is potentially transformative for real-time hybrid simulation. The CI method can allow real-time hybrid simulation to be conducted regardless of the size and complexity of the numerical model and for numerical stability to be ensured in the presence of high frequency responses in the simulation. This paper presents the general theory behind the proposed CI method and provides experimental verification of the proposed method by comparing the CI method to the current integration time-stepping (ITS) method. Real-time hybrid simulation is conducted in the Advanced Hazard Mitigation Laboratory at the University of Connecticut. A seismically excited two-story shear frame building with a magneto-rheological (MR) fluid damper is selected as the test structure to experimentally validate the proposed method. The building structure is numerically modeled and simulated, while the MR damper is physically tested. Real-time hybrid simulation using the proposed CI method is shown to provide accurate results.

Optimal design of flexible pavements using a framework of DAKOTA and MEPDG

Pavement design optimisation is an active area of research. Due to a large number of parameters, such as thickness of layers, material properties, climatic conditions, affecting pavement performance, it is usually not feasible to determine an optimal design using a trial and error approach. In order to make the design calculation computationally tractable, the process can be posed as an optimisation problem. Previous investigations in this vein have suffered from the limitations of a specific pavement analysis tool, specific design goals and specific optimisation algorithms. This paper presents a general computational framework, combining Mechanistic-Empirical Pavement Design Guide and Design Analysis Kit for Optimisation and Terascale Applications, to overcome these shortcomings. The frameworks promise is demonstrated through its application to a minimum cost pavement design problem using both direct and surrogate-based (SBO) optimisation approaches. The SBO formulation is shown to achieve significant savings in required computational time with a minimal loss of accuracy in the determined optimal design.

Use of GPU computing for uncertainty quantification in computational mechanics: A case study

Graphics processing units GPUs are rapidly emerging as a more economical and highly competitive alternative to CPU-based parallel computing. As the degree of software control of GPUs has increased, many researchers have explored their use in non-gaming applications. Recent studies have shown that GPUs consistently outperform their best corresponding CPU-based parallel computing alternatives in single-instruction multiple-data SIMD strategies. This study explores the use of GPUs for uncertainty quantification in computational mechanics. Five types of analysis procedures that are frequently utilized for uncertainty quantification of mechanical and dynamical systems have been considered and their GPU implementations have been developed. The numerical examples presented in this study show that considerable gains in computational efficiency can be obtained for these procedures. It is expected that the GPU implementations presented in this study will serve as initial bases for further developments in the use of GPUs in the field of uncertainty quantification and will i aid the understanding of the performance constraints on the relevant GPU kernels and ii provide some guidance regarding the computational and the data structures to be utilized in these novel GPU implementations.

Response Modification Approach for Safe Extension of Bridge Life

A large portion of highway bridges in the United States are reaching or have reached their intended design lives. To avoid replacing a large number of bridges simultaneously, methodologies to safely extend their lives are important to help avoid high replacement costs and to schedule bridge replacement over a longer time window. This paper proposes an approach to extend the fatigue life of vulnerable steel bridges through a response modification apparatus, consisting of a mechanical amplifier and a response modification device, which provides supplemental stiffness and damping to the bridge. Because of the relatively small deflections encountered under typical service loads, the use of a mechanical amplifier allows for a smaller apparatus and enables a more efficient device to provide adequate response modification forces to the bridge. Herein, the use of a scissor jack as the mechanical amplifier is proposed for use in bridge applications, and its utility in concert with a passive stiffness device is demonstrated by application to a simple beam structure. Reductions in moment ranges of 37% and safe life extensions of 300% are achieved on a simple beam model with the proposed response modification apparatus.

Computationally Efficient Design of Semiactive Structural Control in the Presence of Measurement Noise

Designing control strategies for smart structures, such as those with semiactive devices, is complicated by the nonlinear nature of the feedback control, secondary clipping control, and other additional requirements such as device saturation. The authors have previously developed an approach for semiactive control system design, based on a nonlinear Volterra integral equation (NVIE) that provides a low-order computationally efficient simulation of such systems, for state feedback semiactive clipped-optimal control. This paper expands the applicability of the approach by demonstrating that it can also be adapted to accommodate more realistic cases when, instead of full-state feedback, only a limited set of noisy response measurements is available to the controller. This extension requires incorporating a Kalman filter estimator, which is linear, into the nominal model of the uncontrolled system. The efficacy of the approach is demonstrated by a numerical study of a 100-DOF frame model, excited by a filtered Gaussian random excitation, with noisy acceleration sensor measurements to determine the semiactive control commands. The results show that the proposed method can achieve more than two orders of magnitude improvement in computational efficiency while retaining a comparable level of accuracy.

Lifetime Extension of a Realistic Model of an In-Service Bridge through a Response Modification Approach

Many highway bridges in the United States are reaching their intended design lives and are in need of attention. To avoid replacing these bridges simultaneously, methodologies to safely extend their lives are important to help avoid high replacement costs and to allow for bridge replacement to occur over a longer time window. This paper expands an approach to extend the fatigue life of vulnerable steel bridges through a response modification (RM) apparatus, consisting of a mechanical amplifier and a RM device, which provide efficient supplemental stiffness and damping to the bridge. Because of the relatively small deflections encountered under typical service loads, the use of a mechanical amplifier allows for a more efficient and less intrusive apparatus to provide the required RM forces imparted to the bridge for increased bridge life. This paper presents a parameter study exploring the flexibility of the elements in the RM apparatus, the length of the apparatus, and the characteristics of the RM device. The analyses also add superstructure damping and other model improvements to a comprehensive numerical model of a realistic in-service bridge. The paper carries out dynamic analyses in the frequency domain to ensure the robustness of the RM apparatuses for various loading frequencies. The most important characteristic of the RM apparatus is found to be the cross-sectional area of the members. Additionally, the parameter study showed that a smaller apparatus is most cost effective. Finally, the dynamic analyses showed that the use of a semiactive device may be beneficial to the effectiveness of the RM approach.

Computationally efficient design of optimal output feedback strategies for controllable passive damping devices

Designing control strategies for smart structures, such as those with semiactive devices, is complicated by the nonlinear nature of the feedback control, secondary clipping control and other additional requirements such as device saturation. The usual design approach resorts to large-scale simulation parameter studies that are computationally expensive. The authors have previously developed an approach for state-feedback semiactive clipped-optimal control design, based on a nonlinear Volterra integral equation that provides for the computationally efficient simulation of such systems. This paper expands the applicability of the approach by demonstrating that it can also be adapted to accommodate more realistic cases when, instead of full state feedback, only a limited set of noisy response measurements is available to the controller. This extension requires incorporating a Kalman filter (KF) estimator, which is linear, into the nominal model of the uncontrolled system. The efficacy of the approach is demonstrated by a numerical study of a 100-degree-of-freedom frame model, excited by a filtered Gaussian random excitation, with noisy acceleration sensor measurements to determine the semiactive control commands. The results show that the proposed method can improve computational efficiency by more than two orders of magnitude relative to a conventional solver, while retaining a comparable level of accuracy. Further, the proposed approach is shown to be similarly efficient for an extensive Monte Carlo simulation to evaluate the effects of sensor noise levels and KF tuning on the accuracy of the response.

Efficient Frequency Response of Locally Uncertain Linear Structural Systems

The frequency response analysis of large, linear structural models subjected to deterministic, dynamic loads is oftentimes a computationally intensive task. The situation is only exacerbated when stiffness and damping uncertainties are present in the system’s description. However, this parametric uncertainty is often contained within a small number of localized features. These features involve only a small portion of the total degrees-of-freedom of the structural model. In this paper, this locality of uncertainty is exploited, and an exact method is presented for the frequency analysis of locally uncertain systems subjected to deterministic inputs based on the well-known Sherman-Morrison-Woodbury formula. The proposed method yields exact responses for the perturbed systems and is not affected by magnitude nor the type (e.g., normal, lognormal, etc.) of the uncertainties with the sole restriction that the system remains stable with probability one. A numerical example is presented to illustrate the signifi...