On this page, you can find links to our internal MSDL report publications, as well as abstracts and bibliographic information for publications in peerreviewed journal articles and conferences. Older publications by Dr. Collette before he founded the MSDL are also included. Where allowed by copyright, full text downloaded or preprints of articles are also provided. Papers where source code and data is provided for provenance will have a link below the abstract to access the code or data. Zotero metadata is also provided for the publications.
Department of Naval Architecture
and Marine Engineering
210 Naval Architecture and Marine Engineering
2600 Draper Dr., Ann Arbor MI 48109

Dylan Temple; Chair: Matthew D. Collette. "A MultiObjective Collaborative Optimization Framework to Understand Tradeoffs Between Naval Lifetime Costs Considering Production, Operation, and Maintenance". P.h.D University of MichiganAnn Arbor. 2015
The lifetime cost of naval vessels is an increasingly important factor to ship owners and, subsequently, to ship designers. A vessel's lifetime cost is composed of various cost categories such as production, operation, and maintenance. The impact of each of these categories is important and in many instances they may be competing with each other. Design decisions regarding the hull form and structure will dictate these costs, however, in what way decisions will impact them is difficult to understand. This is especially true for naval vessels as their service life is uncertain, and changes to the operational life of a vessel can have significant unforeseen costs with respect to maintenance and operation. In order to reduce the overall lifetime cost the tradeoffs between these different categories must be understood. This thesis explores a linked resistance, production, and maintenance costing model and develops a novel enhanced multidisciplinary optimizer capable of solving the resulting problem. Most work in cost optimization has focused on reducing a single category of cost and considering other disciplines operational constraints at best. This type of sequential or singlediscipline optimization does not reveal the tradespace to the designer and may result in nonoptimal designs being developed when considering the full lifecycle cost of the vessel. Unfortunately understanding these tradeoffs is difficult and traditional multiobjective optimization algorithms are unable to resolve the Paretofronts effectively. Presented here is a framework to aid designers in finding these trade spaces using a multidisciplinary optimization environment. In order to realistically represent the problem being solved a maintenance costing algorithm is developed that tracks physical damage throughout a ship's lifetime. Given that the design life of a vessel may be prolonged a probabilistic service life is implemented to account for this uncertainty. A hydrodynamic search method is also developed that facilitates efficiently searching large design spaces using a minimal number of design variables. These models allow for the development of tradespaces that reflect the nuances of the naval design problem. In order to utilize these models to understand the tradeoffs in lifetime cost an enhanced multidisciplinary optimization framework is developed. This algorithm uses novel techniques to facilitate solving this difficult design problem. The algorithm (eMOCO) is adopted from a multiobjective collaborative optimization framework with two enhancements. The first is the use of a decision support process, goal programming, at the subsystem level in order to allow the discipline optimizers to reduce objective functions local to that discipline. This means that the disciplinelevel solutions that returned to the systemlevel optimizers are minimized with respect to their local variables. Secondly, a new singleobjective genetic algorithm is developed specifically as a disciplinelevel optimizer in distributed MDO architectures. This novel GA, called the locallyelitist genetic algorithm (LEGA,) allows the discipline problem to be solved in a single execution of the disciplinelevel optimizer. These enhancements, tailored specifically to the naval design problem, facilitate solving for these difficult and unique tradespaces. This model is used to develop trade spaces between production, maintenance, and resistance in order to understand the interaction between the different categories of cost. The results show that the tradespaces are difficult to fully resolve and the use of a multidisciplinary environment is necessary. They also show that by developing the tradespaces unique understanding into the interaction between cost categories can be found that allow an engineer to design ships that have minimal lifetime cost and are robust to changes in operation or service life.
Zhu, J., M. Collette. "A dynamic discretization method for reliability inference in Dynamic Bayesian Networks." Reliability Engineering and System Safety (June 2015 Volume 138):24252
The material and modeling parameters that drive structural reliability analysis for marine structures are subject to a significant uncertainty. This is especially true when timedependent degradation mechanisms such as structural fatigue cracking are considered. Through inspection and monitoring, information such as crack location and size can be obtained to improve these parameters and the corresponding reliability estimates. Dynamic Bayesian Networks (DBNs) are a powerful and flexible tool to model dynamic system behavior and update reliability and uncertainty analysis with life cycle data for problems such as fatigue cracking. However, a central challenge in using DBNs is the need to discretize certain types of continuous random variables to perform network inference while still accurately tracking lowprobability failure events. Most existing discretization methods focus on getting the overall shape of the distribution correct, with less emphasis on the tail region. Therefore, a novel scheme is presented specifically to estimate the likelihood of lowprobability failure events. The scheme is an iterative algorithm which dynamically partitions the discretization intervals at each iteration. Through applications to two stochastic crackgrowth example problems, the algorithm is shown to be robust and accurate. Comparisons are presented between the proposed approach and existing methods for the discretization problem.
Temple, D.W., M. Collette. "Minimizing lifetime structural costs: Optimizing for production and mantenance under service life uncertainty." Marine Structures (January 2015 Volume 40):6072
Current naval shipowners are being forced to extend the service lives of their aging vessels from budgetary and political constraints. This is causing them to incur significant costs due to maintaining the structure of these older ships to keep the ships in operation. These increasing costs make it desirable to design new naval structures with their minimization in mind, as well as ensuring that such vessels are robust to changes in expected service life with respect to their total lifetime cost. However, such structures will necessarily have higher production costs, therefore, an optimization framework is presented to estimate both production and maintenance costs for a naval vessel's internal structure and develop tradespaces between these two competing objectives in order to find designs that represent a balance of both.
Hodapp, D., M. Collette and A.W. Troesch. 2015. "Stochastic nonlinear fatigue crack growth predictions for simple specimens subject to representative ship strucutral loading sequences." International Journal of Fatigue 70 (January):3850
Recent work by the authors investigated an extension of the finite element analysis of plasticityinduced crack closure to nonstationary, ship structural loading sequences by taking advantage of their inherent timedependent nature in which the larger loading cycles tend to be clustered together. In doing so, firstorder load interactions are presumed to arise from the random occurrence and severity of physical storms encountered by ships and offshore structures throughout their service lives. This material hysteresis is captured through a timedependent crack “opening” level (Kop) which is based on the evolution of a rateindependent, incremental plasticity model simulating combined nonlinear kinematic and isotropic hardening. The result is a mechanistic rather than phenomenological numerical model requiring only experimentally measured fatigue crack growth rates under constant amplitude, cyclic loading (e.g., ASTM E64713) and a full material constitutive model defined through experimental push–pull tests for the same material. This approach permits a consideration of material behaviors which are physically relevant to structural steels, yet necessarily omitted in the similar application of a stripyield model.
The present paper generalizes the model originally proposed by the authors to now consider arbitrary storm model loading sequences taken from highfidelity, timedomain seakeeping codes. To predict the fatigue fracture induced by variable amplitude stress records with upwards of 5×106 timedependent cycles, a consistent modeling reduction is applied based on the Ordered Overall Range (OOR) or racetrack counting method. The resultant crack growth behavior is demonstrated to converge remarkably well for sufficiently small refined mesh sizes. Using this model, and by considering different arrangements of the same stress record, the importance of nonlinearities (i.e., those associated with ship response as well as material hysteresis) are emphasized.
