Abstract
Accurate and efficient response sensitivities for fluid–structure interaction (FSI) simulations are important for assessing the uncertain response of coastal and off-shore structures to hydrodynamic loading. To compute gradients efficiently via the direct differentiation method (DDM) for the fully incompressible fluid formulation, approximations of the sensitivity equations are necessary, leading to inaccuracies of the computed gradients when the geometry of the fluid mesh changes rapidly between successive time steps or the fluid viscosity is nonzero. To maintain accuracy of the sensitivity computations, a quasi-incompressible fluid is assumed for the response analysis of FSI using the particle finite element method and DDM is applied to this formulation, resulting in linearized equations for the response sensitivity that are consistent with those used to compute the response. Both the response and the response sensitivity can be solved using the same unified fractional step method. FSI simulations show that although the response using the quasi-incompressible and incompressible fluid formulations is similar, only the quasi-incompressible approach gives accurate response sensitivity for viscous, turbulent flows regardless of time step size.
Similar content being viewed by others
References
Arnold D, Brezzi F, Fortin M (1984) A stable finite element for the stokes equations. Math Stat 21(4):337–344
Bebamzadeh A, Haukaas T, Vaziri R, Poursartip A, Fernlund G (2010) Application of response sensitivity in composite processing. J Compos Mater 44(15):1821–1840
Becker P, Idelsohn S, Oñate E (2015) A unified monolithic approach for multi-fluid flows and fluidstructure interaction using the particle finite element method with fixed mesh. Comput Mech 55(6):1091–1104
Chock G, Carden L, Robertson I, Olsen M, Yu G (2013) Tohoku tsunami-induced building failure analysis with implications for U.S. tsunami and seismic design codes. Earthq Spectra 29(S1):S99–126
Chock GYK, Robertson I, Riggs HR (2011) Tsunami structural design provisions for a new update of building codes and performance-based engineering. In: ASCE solutions of coastal disasters (COPRI). Anchorage, pp 423–435
Chopra AK (2007) Dynamics of structures: theory and applications to earthquake engineering, 3rd edn. Prentice-Hall, Englewood Cliffs
Conte JP, Barbato M, Spacone E (2004) Finite element response sensitivity analysis using force-based frame models. Int J Numer Methods Eng 59(13):1781–1820
Conte JP, Vijalapura PK, Meghalla M (2003) Consistent finite-element response sensitivity analysis. J Eng Mech 129(12):1380–1393
Hübner B, Walhorn E, Dinkler D (2004) A monolithic approach to fluid-structure interaction using space-time finite elements. Comput Methods Appl Mech Eng 193(23–26):2087–2104
Fujimura K, Kiureghian AD (2007) Tail-equivalent linearization method for nonlinear random vibration. Prob Eng Mech 22(1):63–76
Gu Q, Barbato M, Conte J, Gill P, McKenna F (2012) Opensees-snopt framework for finite-element-based optimization of structural and geotechnical systems. J Struct Eng 138(6):822–834
Guo Q, Jeffers A (2014) Direct differentiation method for response sensitivity analysis of structures in fire. Eng Struct 77:172–180
Haukaas T, Scott MH (2006) Shape sensitivities in the reliability analysis of nonlinear frame structures. Comput Struct 84(15–16):964–977
Idelsohn S, Marti J, Limache A, Oñate E (2008) Unified lagrangian formulation for elastic solids and incompressible fluids: application to fluid-structure interaction problems via the pfem. Comput Methods Appl Mech Eng 197:1762–1776
Kiran R, Li L, Khandelwal K (2016) Complex perturbation method for sensitivity analysis of nonlinear trusses. J Struct Eng. To appear
Kleiber M, Antunez H, Hien T, Kowalczyk P (1997) Parameter sensitivity in nonlinear mechanics. Wiley, New York
Mase GT, Smelser RE, Mase GE (2009) Continuum mechanics for engineers, 3rd edn. CRC Press, Boca Raton
McAllister T (2014) The performance of essential facilities in Superstorm Sandy. In: ASCE structures congress. Boston, pp 2269–2281
McKenna F, Fenves GL, Scott, MH (2000) Open system for earthquake engineering simulation. University of California, Berkeley. http://opensees.berkeley.edu
McKenna F, Scott MH, Fenves GL (2010) Nonlinear finite-element analysis software architecture using object composition. J Comput Civil Eng 24(1):95–107
Oñate E, Franci A, Carbonell JM (2014) Lagrangian formulation for finite element analysis of quasi-incompressible fluids with reduced mass losses. Int J Numer Methods Fluids 74(10):699–731
Oñate E, Idelsohn S, Pin FD, Aubry R (2004) The particle finite element method. An overview. Int J Comput Methods 1(2):267–307
Oñate E, Idelsohn SR, Celigueta MA, Rossi R (2007) Advances in the particle finite element method for the analysis of fluid-multibody interaction and bed erosion in free surface flows. Comput Methods Appl Mech Eng 197(19–20):1777–1800
Ryzhakov P, Rossi R, Idelsohn S, Oñate E (2010) A monolithic Lagrangian approach for fluid-structure interaction problems. Comput Mech 46:883–899
Scott MH, Filippou FC (2007) Exact response gradients for large displacement nonlinear beam-column elements. J Struct Eng 133(2):155–165
Scott MH, Franchin P, Fenves GL, Filippou FC (2004) Response sensitivity for nonlinear beam-column elements. J Struct Eng 130(9):1281–1288
Simo JC, Hughes TJR (1998) Computational inelasticity. Springer, New York
Zhang Y, Der Kiureghian A (1993) Dynamic response sensitivity of inelastic structures. Comput Methods Appl Mech Eng 108:23–36
Zhu M, Scott MH (2014) Improved fractional step method for simulating fluid-structure interaction using the pfem. Int J Numer Methods Eng 99(12):925–944
Zhu M, Scott MH (2014) Modeling fluid-structure interaction by the particle finite element method in OpenSees. Comput Struct 132:12–21
Zhu M, Scott MH (2016) Direct differentiation of the particle finite element method for fluid-structure interaction. J Struct Eng 142(3):04015159
Zhu M, Scott MH (2016) Unified fractional step method for lagrangian analysis of quasi-incompressible fluid and nonlinear structure interaction using the pfem. Int J Numer Methods Eng. doi:10.1002/nme.5321
Zienkiewicz O, Taylor R (2005) The finite element method for solid and structural mechanics, vol 2, 6th edn. Elsevier, Oxford
Zienkiewicz O, Taylor R, Zhu J (2005) The finite element method: its basis and fundamentals, vol 1, 6th edn. Elsevier, Butterworth-Heinemann, Oxford
Acknowledgments
This material is based on the work supported by the National Science Foundation under Grant No. 0847055. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix 1: MINI element
The MINI fluid element adds a bubble node for velocity at its center in order to satisfy the LBB condition [1].
Using the nodes in Fig. 9, the shape functions for velocity and pressure in a 2D MINI element are defined as
where \(L_1\), \(L_2\), and \(L_3\) are area coordinates [34]
The spatial derivatives of the velocity and pressure shape functions used in the viscous matrix and gradient operator defined in Eqs. (14) and (17) are evaluated as
and
where A is twice the element area
and the intermediate variables are
The current coordinates \(x_i\) and \(y_i\) of each corner node are determined from the nodal displacements relative to their initial coordinates, \(x_i^0\) and \(y_i^0\), at the start of the simulation
where \(u_i\) and \(w_i\) are the horizontal and vertical nodal displacements. Using the shape functions of the MINI element, the mass matrix in Eq. (13) can be expressed in closed-form as
while the fluid viscous matrix in Eq. (14) is
The gradient operator in Eq. (17) is expressed as
and the external force vector in Eq. (19) is
Finally, the pressure mass matrix in Eq. (20) is evaluated and lumped
The expressions in Eqs. (56)–(63) will be used to show how the sensitivity equations are obtained from differentiation of the response equations.
Appendix 2: Evaluation of geometric derivatives
The geometric derivatives in Eqs. (33) and (34) are evaluated exactly using the MINI element formulation. This appendix demonstrates evaluation of the derivatives in matrices \({\mathbf H}\) and \({\mathbf T}\) but can be extended to any other fluid elements. First, the derivatives of the element area A defined in Eq. (53) are evaluated
Then, the derivatives of the variables defined in Eq. (54) are evaluated, for example,
The derivatives of \(c_2\), \(c_3\), \(d_1\), \(d_2\), and \(d_3\) can be obtained similarly. With the derivatives in Eqs. (64 65)–(67), the matrices \({\mathbf H}\) and \({\mathbf T}\) in Eqs. (35) and (36) can be evaluated on a column by column basis. For the 2D MINI element, both \({\mathbf H}\) and \({\mathbf T}\) have six columns, each of which corresponds to the evaluation of the partial derivatives in Eqs. (35) and (36) with respect to \(u_1\), \(u_2\), \(u_3\), \(w_1\), \(w_2\), and \(w_3\) for columns one through six, respectively. For instance, the first column of the matrices \({\mathbf H}\) and \({\mathbf T}\) corresponds to the displacement \(u_1\)
where
The derivatives \(\frac{\partial {\mathbf M}}{\partial {A}}\), \(\frac{\partial {\mathbf K}}{\partial {A}}\), \(\frac{\partial {\mathbf K}}{\partial {c_a}}\), etc. are evaluated based on Eqs. (56)–(63).
Rights and permissions
About this article
Cite this article
Zhu, M., Scott, M.H. Direct differentiation of the quasi-incompressible fluid formulation of fluid–structure interaction using the PFEM. Comp. Part. Mech. 4, 307–319 (2017). https://doi.org/10.1007/s40571-016-0123-6
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s40571-016-0123-6