Autonomous underwater vehicles (AUVs) are used in a wide variety of military and civilian applications. In this project, we used computational fluid dynamics (CFD) models to analyze the hydrodynamics of AUVs. In addition to fundamental analyses to estimate drag, CFD analyses were used to produce reduced-order maneuvering models (Coe & Neu, 2012; Coe & Neu, 2012; Coe & Neu, 2012). Virtual free-running model simulations, akin to those conducted in seakeeping basins and elsewhere, were developed in which a control algorithm was coupled with the CFD model and allowed to command the AUV’s propeller and control surfaces (Coe & Neu, 2013; Coe, 2013; Coe et al., 2013; Coe & Neu, 2011).
References
2013
Use of Overset Mesh to Allow Dynamic Deflection of Tight-Fitting Control Surfaces in CFD Simulations
Ryan G. Coe, and Wayne L. Neu
In Proceedings of the ASME 2013 32nd International Conference on Ocean, Offshore and Arctic Engineering, Jun 2013
The development of vehicle maneuvering simulations based within computational fluid dynamics (CFD) environments demands that vehicle control surfaces be dynamically deflected during such simulations. This paper details the process of developing and testing CFD simulation methods that allow for the deflection of a specific AUV’s control surfaces. This task is made particularly challenging by the geometry of the AUV, as its moving control surfaces fit very closely to stationary fixed strakes and the AUV’s hull (a fairly common trait among this class of vehicles). After ruling out embedded and deformable mesh approaches, an overset mesh method is applied. Steady-state simulations with this overset mesh show general agreement with static mesh simulations. The two approaches do, however, highlight the mesh sensitivity of CFD simulations in their ability to predict the onset of stall.
@inproceedings{Coe:2013ac,address={Nantes, France},author={Coe, Ryan G. and Neu, Wayne L.},bibtex_show=true,booktitle={Proceedings of the ASME 2013 32nd International Conference on Ocean, Offshore and Arctic Engineering},date-added={2020-10-07 15:14:11 -0600},date-modified={2025-06-25 12:01:07 -0600},doi={10.1115/OMAE2013-10931},eprint={https://asmedigitalcollection.asme.org/OMAE/proceedings-pdf/OMAE2013/55416/V007T08A051/4431565/v007t08a051-omae2013-10931.pdf},month=jun,note={V007T08A051},series={International Conference on Offshore Mechanics and Arctic Engineering},title={{Use of Overset Mesh to Allow Dynamic Deflection of Tight-Fitting Control Surfaces in CFD Simulations}},url={https://doi.org/10.1115/OMAE2013-10931},volume={7: CFD and VIV},year={2013},bdsk-url-1={https://doi.org/10.1115/OMAE2013-10931}}
Improved underwater vehicle control and maneuvering analysis with computational fluid dynamics simulations
The quasi-steady state-space models generally used to simulate the dynamics of underwater vehicles perform well in most steady flow scenarios, and are therefore acceptable for modeling today’s fleet of endurance-focused autonomous underwater vehicles (AUVs). However, with their usage of numerous assumptions and simplifications, these models are not well suited to certain unsteady flow situations and for use in the development of AUVs capable of performing more extreme maneuvers. In the interest of better serving efforts to design a new generation of more maneuverable AUVs, a tool for simulating vehicle maneuvering within computational fluid dynamics (CFD) based environments has been developed. Unsteady Reynolds-averaged Navier-Stokes (URANS) simulations are used in conjunction with a 6-degree-of-freedom (6-DoF) rigid-body kinematic model to provide a numerical test basin for vehicle maneuvering simulations. The accuracy of this approach is characterized through comparison with experimental measurements and quasi-steady state-space models. Three state-space models are considered: one model obtained from semi-empirical database regression (this is the method most commonly used in application) and two models populated with coefficients determined from the results of prescribed motion CFD simulations. CFD analyses focused on supporting the design of a general purpose AUV are also presented.
@phdthesis{Coe:2013ab,address={Blacksburg, VA},author={Coe, Ryan Geoffrey},bibtex_show=true,date-added={2020-10-07 12:50:49 -0600},date-modified={2025-06-25 12:01:07 -0600},school={Virginia Tech},title={Improved underwater vehicle control and maneuvering analysis with computational fluid dynamics simulations},url={http://hdl.handle.net/10919/23777},year={2013},bdsk-url-1={http://hdl.handle.net/10919/23777}}
CFD-Based Maneuvering Simulations for Autonomous Underwater Vehicles
Ryan G. Coe, Brian R. McCarter, and Wayne L. Neu
In Virginia Space Grant Consortium Student Research Conference, Jun 2013
@inproceedings{Coe:2013aa,address={Norfolk, VA},author={Coe, Ryan G. and McCarter, Brian R. and Neu, Wayne L.},bibtex_show=true,booktitle={Virginia Space Grant Consortium Student Research Conference},date-added={2020-10-07 15:14:40 -0600},date-modified={2025-06-25 12:01:07 -0600},title={{CFD}-Based Maneuvering Simulations for Autonomous Underwater Vehicles},url={http://www.vsgc.odu.edu/awardees/20122013/abstracts/Papers - Grad/Coe, Ryan - Paper.pdf},year={2013},bdsk-url-1={http://www.vsgc.odu.edu/awardees/20122013/abstracts/Papers%20-%20Grad/Coe,%20Ryan%20-%20Paper.pdf}}
2012
Virtual planar motion mechanism tests in a CFD environment
Ryan Coe, and Wayne Neu
In Virginia Space Grant Consortium Student Research Conference, Jun 2012
A study is currently underway to better understand and influence the maneuvering characteristics of autonomous underwater vehicles (AUVs). A two-pronged approach, using traditional quasi-steady state-space modeling as well as maneuvering experiments performed in unsteady Reynolds-averaged Navier-Stokes simulations (URANS), has been adopted to provide the greatest possible insight into vehicle modeling. State-space models must be populated with parameters describing the vehicle of interest. This paper focuses on the use of a virtual planar motion mechanism (PMM) method to find hydrodynamic maneuvering characteristics within a computational fluid dynamics (CFD) environment.
@inproceedings{Coe:2012aa,author={Coe, Ryan and Neu, Wayne},bibtex_show=true,booktitle={Virginia Space Grant Consortium Student Research Conference},date-added={2020-10-07 12:59:38 -0600},date-modified={2025-06-25 12:01:07 -0600},title={Virtual planar motion mechanism tests in a {CFD} environment},url={https://www.academia.edu/18038713/VIRTUAL_PLANAR_MOTION_MECHANISM_TESTS_IN_A_CFD_ENVIRONMENT},year={2012},bdsk-url-1={https://www.academia.edu/18038713/VIRTUAL_PLANAR_MOTION_MECHANISM_TESTS_IN_A_CFD_ENVIRONMENT}}
Planar Motion Mechanism (PMM) testing provides a means of determining the performance characteristics (often referred to as control derivatives or maneuvering coefficients) that populate vehicle quasi-steady state-space models. While the general nature of these tests is fairly well established, the details of execution and post processing methods vary between experimenters. This study employs numerical simulations to examine the use of a number of common force decomposition models across a range of amplitudes of oscillation. The results show that the force decomposition models examined perform similarly well at low amplitude, however accuracy appears to decline at higher amplitudes of motion with the increasing prevalence of unmodeled viscous flow phenomena.
@inproceedings{Coe:2012ac,address={Hampton Roads, VA, USA},author={Coe, Ryan G. and Neu, Wayne L.},bibtex_show=true,booktitle={OCEANS2012},date-added={2020-10-07 12:52:07 -0600},date-modified={2025-06-25 12:01:07 -0600},doi={10.1109/OCEANS.2012.6405027},issn={0197-7385},keywords={computational fluid dynamics;flow;marine vehicles;motion control;numerical analysis;state-space methods;amplitude effects;virtual PMM tests;planar motion mechanism testing;control derivatives;maneuvering coefficients;quasi-steady state-space models;numerical simulations;force decomposition models;oscillation amplitudes;motion amplitudes;unmodeled viscous flow phenomena;marine vehicle control design;computational fluid dynamics;Force;Oscillators;Vehicles;Numerical models;Hydrodynamics;Testing;Damping},month=oct,pages={1-5},title={Amplitude effects on virtual {PMM} tests},url={https://doi.org/10.1109/OCEANS.2012.6405027},year={2012},bdsk-url-1={https://doi.org/10.1109/OCEANS.2012.6405027}}
Asymmetrical wake and propulsor effects on control surface effectiveness on AUVs
This study considers the influences of wake asymmetries and propulsor effects on the forces and moments created by control surfaces. Traditional quasi-steady state-space models developed for autonomous underwater vehicles (AUVs) tend to neglect these effects. Reynolds-averaged Navier-Stokes (RANS) simulations were used to assess the impact of asymmetrical inflow due to forward appendages as well as changes in the flow field created by an operating propeller on control surface effectiveness. For the AUV tested, substantial asymmetries in the flow field near the upper and lower rudders create significant differences in their respective performances. This discrepancy between the rudders has the potential to create considerable and unsuspected maneuvering reactions. The presence of the propeller was also seen to noticeably influence the performance of the control surfaces.
@inproceedings{Coe:2012ab,address={Hampton Roads, VA, USA},author={Coe, Ryan G. and Neu, Wayne L.},bibtex_show=true,booktitle={OCEANS2012},date-added={2020-10-07 12:53:07 -0600},date-modified={2025-06-25 12:01:07 -0600},doi={10.1109/OCEANS.2012.6404941},issn={0197-7385},keywords={autonomous underwater vehicles;flow;hydrodynamics;motion control;Navier-Stokes equations;propellers;vehicle dynamics;wakes;asymmetrical wake effects;propulsor effects;control surface effectiveness;AUV;autonomous underwater vehicles;quasi-steady state-space models;Reynolds-averaged Navier-Stokes simulations;RANS;asymmetrical inflow impact assessment;upper rudders;lower rudders;unsuspected maneuvering reactions;quasi-steady state-space hydrodynamic models;Propellers;Sea surface;Vehicles;Computational modeling;Actuators;Underwater vehicles;Surface treatment},month=oct,pages={1-4},title={Asymmetrical wake and propulsor effects on control surface effectiveness on {AUVs}},url={https://doi.org/10.1109/OCEANS.2012.6404941},year={2012},bdsk-url-1={https://doi.org/10.1109/OCEANS.2012.6404941}}
2011
Vehicle Control in a CFD Environment
Ryan G. Coe, and Wayne L. Neu
In Grand Challenges on Modeling and Simulation Conference, Oct 2011
Work in progress toward a tool for synchronous control algorithm design and vehicle hydrodynamic analysis via CFD is discussed. A commercial CFD code, with minor augmentations, will serve as a high-fidelity hydrodynamic model to be run in parallel with and loosely coupled to a candidate control algorithm allowing the vehicle to be computationally flown under the influence of that control. A number of components of this tool have been developed and are described. These include a momentum based propeller model which simulates both added axial and swirl velocities and a control surface force parameterization which will be used as a first approximation while a mesh morphing scheme for deflecting control surfaces is implemented. The accuracy of the CFD approach is demonstrated by comparison with experimental data for drag on a DARPA SUBOFF model and for skin friction and flow separation on a prolate spheroid at an angle of attack. Unsteady nonlinear effects are demonstrated with a simulation of a spheroid propagating through water in the presence of an oscillating sway force. The direction of future work is also presented.
@inproceedings{Coe:2011aa,address={The Hague, Netherlands},author={Coe, Ryan G. and Neu, Wayne L.},bibtex_show=true,booktitle={Grand Challenges on Modeling and Simulation Conference},date-added={2020-10-07 15:17:57 -0600},date-modified={2025-06-25 12:01:07 -0600},title={Vehicle Control in a {CFD} Environment},url={http://dl.acm.org/citation.cfm?id=2348229.2348281},year={2011},bdsk-url-1={http://dl.acm.org/citation.cfm?id=2348229.2348281}}