Skip to content

3.7 Weak Formulation for BFSI

Virtual Work and Weak Form

The virtual work statement is used to enforce the three governing equations needed to solve for the nodal DOFs \(\mathbf{u}\), \(\mathbf{w}\) and \(e^{f}\), namely the mixture mass balance (2.15-6), the fluid momentum balance (2.15-7), and the solid momentum balance (2.15-8). We may rewrite the momentum balance equations to facilitate the enforcement of natural traction boundary conditions given in (2.15-16) and (2.15-17). Using \(\boldsymbol{\tau}^{f}=\varphi^{f}\boldsymbol{\tau}\), these become

\[ \begin{equation} \begin{aligned}\varphi^{s}\left(-\rho_{T}^{f}\mathbf{a}^{f}+\rho_{T}^{s}\mathbf{a}^{s}\right)= & -\frac{\varphi^{s^{2}}}{\varphi^{f}}\mathbf{\boldsymbol{\tau}}\cdot\grad\frac{\varphi^{f}}{\varphi^{s}}+\divg\left(-\varphi^{s}\boldsymbol{\tau}+\boldsymbol{\sigma}^{e}\right)\\ & +\varphi^{s}\left(-\rho_{T}^{f}\mathbf{b}^{f}+\rho_{T}^{s}\mathbf{b}^{s}\right)+\mathbf{k}^{-1}\cdot\mathbf{w}\,,\\ \rho_{T}^{f}\mathbf{a}^{f}= & -\grad p+\frac{1}{\phi^{f}}\mathbf{\boldsymbol{\tau}}\cdot\grad\varphi^{f}+\divg\boldsymbol{\tau}\\ & +\rho_{T}^{f}\mathbf{b}^{f}-\mathbf{k}^{-1}\cdot\mathbf{w}\,. \end{aligned} \label{eq:bfsi-gov-eqn-jump} \end{equation} \]

The virtual work statement for a Galerkin finite element formulation 1 is \(\delta W=0\), where

\[ \begin{equation} \begin{aligned}\delta W & =\int_{\Omega^{b}}\delta\mathbf{v}^{s}\cdot\left[\divg\left(-\varphi^{s}\boldsymbol{\tau}+\boldsymbol{\sigma}^{e}\right)-\frac{\varphi^{s^{2}}}{\varphi^{f}}\boldsymbol{\tau}\cdot\grad\frac{\varphi^{f}}{\varphi^{s}}+\mathbf{k}^{-1}\cdot\mathbf{w}\right]\,dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}^{s}\cdot\varphi^{s}\left(-\rho_{T}^{f}\left(\mathbf{b}^{f}-\mathbf{a}^{f}\right)+\rho_{T}^{s}\left(\mathbf{b}^{s}-\mathbf{a}^{s}\right)\right)\,dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\left[\divg\boldsymbol{\tau}-\grad p+\frac{1}{\varphi^{f}}\boldsymbol{\tau}\cdot\grad\phi^{f}-\mathbf{k}^{-1}\cdot\mathbf{w}\right]\,dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\rho_{T}^{f}\left(\mathbf{b}^{f}-\mathbf{a}^{f}\right)\,dv\\ & +\int_{\Omega^{b}}\delta J^{f}\left[\divg\mathbf{w}+\frac{\dot{J}^{s}}{J^{s}}-\frac{1}{J^{f}}\left(\varphi^{f}\dot{J}^{f}+\grad J^{f}\cdot\mathbf{w}\right)\right]\,dv\,. \end{aligned} \label{eq:bfsi-virtual-work-strong} \end{equation} \]

These integrals are evaluated in the current configuration of \(\Omega^{b}\). Here, \(\delta\mathbf{v}^{s}\) is the virtual solid velocity, \(\delta\mathbf{w}\) is the virtual relative fluid volumetric flux, and \(\delta J^{f}\) is the virtual fluid energy density. Integrating by parts and using the divergence theorem, the weak form of this statement may be written as \(\delta W=\delta W_{ext}-\delta W_{int}\) where the internal virtual work is

\[ \begin{equation} \begin{aligned}\delta W_{int} & =\int_{\Omega^{b}}\left[\left(-\varphi^{s}\boldsymbol{\tau}+\boldsymbol{\sigma}^{e}\right):\grad\delta\mathbf{v}^{s}-\delta\mathbf{v}^{s}\cdot\mathbf{k}^{-1}\cdot\mathbf{w}\right]\,dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}^{s}\cdot\varphi^{s}\left(\frac{\varphi^{s}}{\varphi^{f}}\boldsymbol{\tau}\cdot\grad\frac{\varphi^{f}}{\varphi^{s}}-\rho_{T}^{f}\mathbf{a}^{f}+\rho_{T}^{s}\mathbf{a}^{s}\right)\,dv\\ & +\int_{\Omega^{b}}\boldsymbol{\tau}:\grad\delta\mathbf{w}\,dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\left(\grad p+\mathbf{k}^{-1}\cdot\mathbf{w}\right)\,dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\left(\rho_{T}^{f}\mathbf{a}^{f}-\frac{1}{\varphi^{f}}\boldsymbol{\tau}\cdot\grad\varphi^{f}\right)\,dv\\ & +\int_{\Omega^{b}}\mathbf{w}\cdot\grad\delta J^{f}\,dv\,,\\ & +\int_{\Omega^{b}}\delta J^{f}\left[\frac{1}{J^{f}}\left(\varphi^{f}\dot{J}^{f}+\grad J^{f}\cdot\mathbf{w}\right)-\frac{\dot{J}^{s}}{J^{s}}\right]\,dv \end{aligned} \label{eq:bfsi-int-virtual-work} \end{equation} \]

and the external part is

\[ \begin{equation} \begin{aligned}\delta W_{ext} & =\int_{\partial\Omega^{b}}\delta\mathbf{v}^{s}\cdot\mathbf{t}^{\sigma}\,da\\ & +\int_{\Omega^{b}}\delta\mathbf{v}^{s}\cdot\varphi^{s}\left(-\rho_{T}^{f}\mathbf{b}^{f}+\rho_{T}^{s}\mathbf{b}^{s}\right)\,dv\\ & +\int_{\partial\Omega^{b}}\delta\mathbf{w}\cdot\mathbf{t}^{\tau}\,da+\int_{\Omega^{f}}\delta\mathbf{w}\cdot\rho_{T}^{f}\mathbf{b}^{f}\,dv\\ & +\int_{\partial\Omega^{b}}\delta J^{f}w_{n}\,da\,. \end{aligned} \label{eq:bfsi-ext-virtual-work} \end{equation} \]

where

\[ \begin{equation} \mathbf{t}^{\sigma}=-\varphi^{s}\mathbf{t}^{\tau}+\mathbf{t}^{e}\,.\label{eq:bfsi-sigma-traction} \end{equation} \]

Here, \(\mathbf{t}^{e}=\boldsymbol{\sigma}^{e}\cdot\mathbf{n}\) is the elastic traction, \(\mathbf{t}^{\tau}=\boldsymbol{\tau}\cdot\mathbf{n}\) is the true fluid viscous traction, and \(w_{n}=\mathbf{w}_{n}\cdot\mathbf{n}\) is the normal component of the relative fluid flux on the boundary \(\partial\Omega^{b}\), whose outward unit normal is \(\mathbf{n}\). The traction \(\mathbf{t}^{\sigma}\) emerges from the jump condition in (2.15-17). The integrands of the surface integrals represent the natural boundary conditions for this formulation. If boundary conditions are not set explicitly on \(\partial\Omega^{b}\), the natural boundary conditions are \(\mathbf{t}^{\sigma}=\mathbf{0}\), \(\mathbf{t}^{\tau}=\mathbf{0}\), and \(w_{n}=0\). These natural boundary conditions are consistent with the jump conditions presented above. Essential boundary conditions are prescribed on the solid displacement \(\mathbf{u}\), relative volumetric fluid flux \(\mathbf{w}\), and fluid dilatation \(e^{f}\), which are also consistent with the above jump conditions. In particular, an essential no-slip boundary condition may be prescribed on \(\Gamma^{bs}\) by setting \(\mathbf{w}=\mathbf{0}\). A symmetry plane may be prescribed with the essential boundary condition \(u_{n}\equiv\mathbf{u}\cdot\mathbf{n}=0\) and the natural boundary conditions \(\mathbf{t}^{\tau}=\mathbf{0}\) and \(w_{n}=0\).

In this formulation, the mixture traction is defined as \(\mathbf{t}=-p\mathbf{n}+\mathbf{t}^{e}+\varphi^{f}\mathbf{t}^{\tau}\), which may also be written as \(\mathbf{t}=-p\mathbf{n}+\mathbf{t}^{\sigma}+\mathbf{t}^{\tau}\). Because of the way we chose to split the internal and external virtual work in \eqref{eq:bfsi-int-virtual-work}-\eqref{eq:bfsi-ext-virtual-work}, \(\mathbf{t}\) is not a natural boundary condition in this formulation. In this expression for \(\mathbf{t}\), \(\mathbf{t}^{\sigma}\), and \(\mathbf{t}^{\tau}\) may be prescribed as natural boundary conditions, whereas \(p\) may be prescribed as an essential boundary condition on \(e^{f}\), using (2.15-9). However, there are two general scenarios where \(\mathbf{t}\) needs to be prescribed on a region of the boundary \(\partial\Omega^{b}\) with incomplete prior knowledge of \(p\), \(\mathbf{t}^{\sigma}\), or \(\mathbf{t}^{\tau}\): (1) When a BFSI boundary \(\Gamma^{b}\) represents a free surface (such as the fluid surface in channel flow), the mixture traction boundary condition requires that \(\mathbf{t}=\mathbf{0}\), in which case it is necessary to explicitly enforce \(-p\mathbf{n}+\mathbf{t}^{\sigma}+\mathbf{t}^{\tau}=\mathbf{0}\) as a constraint equation on that boundary, to impart the free surface \(\Gamma^{b}\) its natural shape. (2) At a biphasic-solid interface \(\Gamma^{bs}\), \(\mathbf{t}\) must balance the traction acting on the adjoining solid domain. Since \(\mathbf{u}\) is continuous across \(\Gamma^{bs}\) due to shared nodes, the solid natural boundary condition \(\mathbf{t}^{\sigma}\) of \(\mathbf{t}\) is already accounted for by the deformation, so that it is only necessary to prescribe the portion \(-p\mathbf{n}+\mathbf{t}^{\tau}\) of \(\mathbf{t}\) on the solid domain, thus \(-p\mathbf{n}+\mathbf{t}^{\tau}+\mathbf{t}^{s}=\mathbf{0}\) where \(\mathbf{t}^{s}\) is the (equal and opposite) traction acting on the solid domain. In both cases, the form of this traction boundary condition is the same, with \(\mathbf{t}^{e}\) and \(\mathbf{t}^{s}\) representing the tractions acting on \(\Gamma^{b}\) and \(\Gamma^{bs}\), respectively.

For both of these cases, the resulting virtual work on the free surface \(\Gamma^{b}\) or the interface \(\Gamma^{bs}\) takes the form

\[ \begin{equation} \delta F=-\int_{\Gamma}\delta\mathbf{v}^{s}\cdot\left(-p\mathbf{n}+\mathbf{t}^{\tau}\right)\,da\,,\label{eq:bfsi-traction-virtual-work} \end{equation} \]

where the elemental area \(da\) on \(\Gamma\) may be evaluated from the covariant basis vectors \(\mathbf{g}_{\alpha}\) (\(\alpha=1,2\)),

\[ \begin{equation} da=\left|\mathbf{g}_{1}\times\mathbf{g}_{2}\right|\,d\eta^{1}d\eta^{2}\,,\label{eq:bfsi-diff-area} \end{equation} \]

where

\[ \begin{equation} \mathbf{g}_{\alpha}=\frac{\partial\mathbf{x}\left(\eta^{1},\eta^{2}\right)}{\partial\eta^{\alpha}}\,,\label{eq:bfsi-covar-vectors} \end{equation} \]

and \(\mathbf{x}\left(\eta^{1},\eta^{2}\right)\) is the parametric representation of \(\Gamma\), defined on the solid constituent. The outward normal \(\mathbf{n}\) to \(\Omega^{b}\) on \(\Gamma\) is evaluated from

\[ \begin{equation} \mathbf{n}=\frac{\mathbf{g}_{1}\times\mathbf{g}_{2}}{\left|\mathbf{g}_{1}\times\mathbf{g}_{2}\right|}\,.\label{eq:bfsi-norm-vector} \end{equation} \]

As a result, the virtual work can be rewritten as

\[ \begin{equation} \delta F=-\int_{\Gamma}\delta\mathbf{v}^{s}\cdot\left(-p\mathbf{I}+\boldsymbol{\tau}\right)\cdot\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\,d\eta^{1}d\eta^{2}\,.\label{eq:bfsi-traction-final} \end{equation} \]

In FEBio this boundary condition is called BFSI traction, which the user must explicitly prescribe on free surfaces \(\Gamma^{b}\) and deformable interfaces \(\Gamma^{bs}\). The code automatically determines which of these two types of interfaces is being considered.

BFSI Linearization

Using the virtual work integral \(\delta W\) such that

\[ \begin{equation} \delta W+D\delta W\left[\Delta\mathbf{u}\right]+D\delta W\left[\Delta\mathbf{w}\right]+D\delta W\left[\Delta J^{f}\right]\approx0\,,\label{eq:Virtual-Work-Expanded-BFSI} \end{equation} \]

it may be expanded as

\[ \begin{equation} \begin{aligned} & D\delta W_{int}\left[\Delta\mathbf{u}\right]+D\delta W_{int}\left[\Delta\mathbf{w}\right]+D\delta W_{int}\left[\Delta J^{f}\right]\\ & -D\delta W_{ext}\left[\Delta\mathbf{u}\right]-D\delta W_{ext}\left[\Delta\mathbf{w}\right]-D\delta W_{ext}\left[\Delta J^{f}\right]\\ & \approx\delta W_{ext}-\delta W_{int}\,. \end{aligned} \label{eq:linearized-work-split-bfsi} \end{equation} \]

The linearizations of integrals are performed in the material frame of the solid domain of \(\Omega^{b}\), allowing us to linearize \(\delta W_{int}\) along increments \(\Delta\mathbf{u}\), \(\Delta\mathbf{w}\), or \(\Delta J^{f}\) by simply bringing the directional derivative operator inside the integrals of Eqs. \eqref{eq:bfsi-int-virtual-work}-\eqref{eq:bfsi-ext-virtual-work}. For notational convenience, we let \(J\equiv J^{s}\) and \(\mathbf{F}\equiv\mathbf{F}^{s}\). Thus, the conversion of the internal virtual work to the material frame of the solid produces

\[ \begin{equation} \begin{aligned}\delta W_{int}= & \int_{\Omega^{b}}\left(-\varphi_{r}^{s}\mathbf{T}_{d}+\mathbf{F}\cdot\boldsymbol{\sigma}^{e}\cdot\mathbf{F}^{T}\right):\Grad\delta\mathbf{v}\cdot\mathbf{F}^{-1}dV\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot\varphi_{r}^{s}\left(\frac{\varphi^{s}}{\varphi^{f}}\boldsymbol{\tau}\cdot\mathbf{F}^{-T}\cdot\Grad\left(\frac{\varphi^{f}}{\varphi^{s}}\right)-\rho_{T}^{f}\mathbf{a}^{f}+\rho_{T}^{s}\dot{\mathbf{v}}^{s}\right)dV\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot-J^{2}\mathbf{F}^{-T}\cdot\mathbf{K}^{-1}\cdot\mathbf{F}^{-1}\cdot\mathbf{w}dV\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot J\left(\rho_{T}^{f}\mathbf{a}^{f}+\mathbf{F}^{-T}\cdot\Grad p\right)dV\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot J\left(-\frac{1}{\varphi^{f}}\boldsymbol{\tau}\cdot\mathbf{F}^{-T}\cdot\Grad\varphi^{f}+J\mathbf{F}^{-T}\cdot\mathbf{K}^{-1}\cdot\mathbf{F}^{-1}\cdot\mathbf{w}\right)dV\\ & +\int_{\Omega^{b}}J\boldsymbol{\tau}:\Grad\delta\mathbf{w}\cdot\mathbf{F}^{-1}dV\\ & +\int_{\Omega^{b}}J\Grad\delta J^{f}\cdot\mathbf{F}^{-1}\cdot\mathbf{w}+\delta J^{f}\left(J\frac{\varphi^{f}}{J^{f}}\frac{D^{f}J^{f}}{Dt}-\dot{J}^{s}\right)dV\thinspace, \end{aligned} \label{eq:int-virtual-work-material-BFSI} \end{equation} \]

where \(\mathbf{S}=J\,\mathbf{F}^{-1}\cdot\boldsymbol{\sigma}^{e}\cdot\mathbf{F}^{-T}\) is the second Piola-Kirchhoff stress for the solid constituent of the mixture, \(\mathbf{W}=J\mathbf{F}^{-1}\cdot\mathbf{w}\) is the Piola transformation of \(\mathbf{w}\), and \(dV=J^{-1}dv\) is an elemental volume of \(\Omega^{f}\) in its material frame 1. In addition, \(\mathbf{K}^{-1}=J^{-1}\mathbf{F}^{T}\cdot\mathbf{k}^{-1}\cdot\mathbf{F}\) is the inverse of the permeability tensor in the material frame. Note that in the material frame, the fluid acceleration \(\mathbf{a}^{f}\) is

\[ \begin{equation} \mathbf{a}^{f}=\frac{1}{\varphi^{f}}\left(\dot{\mathbf{w}}+\varphi^{f}\dot{\mathbf{v}}^{s}-\frac{\varphi^{s}}{\varphi^{f}}\frac{\dot{J}}{J}\mathbf{w}+\frac{1}{J}\left(\frac{1}{\varphi^{f}}\Grad\mathbf{w}+\Grad\mathbf{v}^{s}-\frac{1}{\varphi^{f^{2}}}\mathbf{w}\otimes\Grad\varphi^{f}\right)\cdot\mathbf{W}\right)\,.\label{eq:material-fluid-accel-BFSI} \end{equation} \]

The linearization of \(\delta W_{int}\) is then performed along an increment \(\Delta\mathbf{u}\), and the integral is reverted back to the spatial frame, yielding for the displacement equations

\[ \begin{equation} \begin{aligned}D\delta W_{int}\left[\Delta\mathbf{u}\right]= & \int_{\Omega^{b}}\alpha_{f}\left(\grad\Delta\mathbf{u}\cdot\boldsymbol{\sigma}^{e}:\grad\delta\mathbf{v}+\grad\delta\mathbf{v}:\boldsymbol{\mathcal{C}}:\grad\Delta\mathbf{u}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}\varphi^{s}\boldsymbol{\tau}:\grad\delta\mathbf{v}\cdot\left(\grad\Delta\mathbf{u}+\frac{\varphi^{s}}{\varphi^{f}}\divg\Delta\mathbf{u}\mathbf{I}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}\varphi^{s}\grad\delta\mathbf{v}:\boldsymbol{\mathcal{C}}^{\tau}:\left(-\frac{\varphi^{s}}{\varphi^{f^{2}}}\left(\divg\Delta\mathbf{u}\right)\mathbf{D}^{w}+\mathbf{M}\cdot\grad\Delta\mathbf{u}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}\varphi^{s}\grad\delta\mathbf{v}:\frac{1}{\varphi^{f^{2}}}\boldsymbol{\mathcal{C}}^{\tau}:2\frac{\varphi^{s}}{\varphi^{f}}\left(\divg\Delta\mathbf{u}\right)\mathbf{w}\otimes\grad\varphi^{f}dv\\ & +\int_{\Omega^{b}}\alpha_{f}\varphi^{s}\grad\delta\mathbf{v}:\frac{1}{\varphi^{f^{2}}}\boldsymbol{\mathcal{C}}^{\tau}:\mathbf{w}\otimes\left(-\grad^{T}\Delta\mathbf{u}\cdot\grad\varphi^{f}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}\varphi^{s}\grad\delta\mathbf{v}:\frac{1}{\varphi^{f^{2}}}\boldsymbol{\mathcal{C}}^{\tau}:\mathbf{w}\otimes\left(-\left(\divg\Delta\mathbf{u}\right)\grad\varphi^{f}+\varphi^{s}\grad\left(\divg\Delta\mathbf{u}\right)\right)dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot-\alpha_{f}\frac{\varphi^{s^{2}}}{\varphi^{f}}\boldsymbol{\tau}\cdot\left(2\frac{\varphi^{s}}{\varphi^{f}}\divg\Delta\mathbf{u}+\grad^{T}\Delta\mathbf{u}\right)\cdot\grad\frac{\varphi^{f}}{\varphi^{s}}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot\alpha_{f}\frac{\varphi^{s^{2}}}{\varphi^{f}}\boldsymbol{\tau}\cdot\frac{1}{\varphi^{s}}\grad\left(\divg\Delta\mathbf{u}\right)dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot\alpha_{f}\frac{\varphi^{s^{2}}}{\varphi^{f}}\boldsymbol{\mathcal{C}}^{\tau}:\left(-\frac{\varphi^{s}}{\varphi^{f^{2}}}\left(\divg\Delta\mathbf{u}\right)\mathbf{D}^{w}+\mathbf{M}\cdot\grad\Delta\mathbf{u}\right)\cdot\grad\frac{\varphi^{f}}{\varphi^{s}}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot\alpha_{f}\frac{\varphi^{s^{2}}}{\varphi^{f^{3}}}\boldsymbol{\mathcal{C}}^{\tau}:2\frac{\varphi^{s}}{\varphi^{f}}\left(\divg\Delta\mathbf{u}\right)\mathbf{w}\otimes\grad\varphi^{f}\cdot\grad\frac{\varphi^{f}}{\varphi^{s}}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot-\alpha_{f}\frac{\varphi^{s^{2}}}{\varphi^{f^{3}}}\boldsymbol{\mathcal{C}}^{\tau}:\mathbf{w}\otimes\left(-\grad^{T}\Delta\mathbf{u}\cdot\grad\varphi^{f}\right)\cdot\grad\frac{\varphi^{f}}{\varphi^{s}}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot-\alpha_{f}\frac{\varphi^{s^{2}}}{\varphi^{f^{3}}}\boldsymbol{\mathcal{C}}^{\tau}:\mathbf{w}\otimes\left(-\divg\Delta\mathbf{u}\grad\varphi^{f}+\varphi^{s}\grad\left(\divg\Delta\mathbf{u}\right)\right)\cdot\grad\frac{\varphi^{f}}{\varphi^{s}}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot-\alpha_{f}\varphi^{s}\frac{\rho_{T}^{f}}{\varphi^{f}}\left(\varphi^{s}\left(\divg\Delta\mathbf{u}\right)\dot{\mathbf{v}}^{s}+\varphi^{f}\frac{\alpha_{m}}{\alpha_{f}\beta\Delta t^{2}}\Delta\mathbf{u}\right)dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot\alpha_{f}\varphi^{s}\frac{\rho_{T}^{f}}{\varphi^{f}}\left(\frac{\varphi^{s}}{\varphi^{f}}\left[\left(\left(-\frac{1}{\varphi^{f}}\right)\frac{\dot{J}}{J}+\frac{\gamma}{\beta\Delta t}\right)\divg\Delta\mathbf{u}-\grad^{T}\Delta\mathbf{u}:\mathbf{L}^{s}\right]\mathbf{w}\right)dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot-\alpha_{f}\varphi^{s}\frac{\rho_{T}^{f}}{\varphi^{f}}\left(-\frac{\varphi^{s}}{\varphi^{f^{2}}}\divg\Delta\mathbf{u}\grad\mathbf{w}+\frac{\gamma}{\beta\Delta t}\grad\Delta\mathbf{u}\right)\cdot\mathbf{w}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot-\alpha_{f}\varphi^{s}\frac{\rho_{T}^{f}}{\varphi^{f^{3}}}\left[\left(\left(\varphi^{s}+1\right)\frac{1}{\varphi^{f}}\grad\varphi^{f}\divg\Delta\mathbf{u}-\varphi^{s}\grad\left(\divg\Delta\mathbf{u}\right)\right)\cdot\mathbf{w}\right]\mathbf{w}dv \end{aligned} \label{eq:Wint-lin-u-u-BFSI} \end{equation} \]
\[ \begin{aligned} & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot\alpha_{f}\varphi^{s}\frac{\rho_{T}^{f}}{\varphi^{f}}\left(\mathbf{L}^{f}\cdot\grad\Delta\mathbf{u}\cdot\mathbf{w}+\varphi^{s}\left(\divg\Delta\mathbf{u}\right)\mathbf{a}^{f}\right)dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot\alpha_{f}\rho^{s}\frac{\alpha_{m}}{\alpha_{f}\beta\Delta t^{2}}\Delta\mathbf{u}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot\alpha_{f}\left(-2\left(\divg\Delta\mathbf{u}\right)\mathbf{k}^{-1}+\grad^{T}\Delta\mathbf{u}\cdot\mathbf{k}^{-1}\right)\cdot\mathbf{w}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot\alpha_{f}\left(\mathbf{k}^{-1}\cdot\grad\Delta\mathbf{u}+\left(\mathbf{k}^{-1}\underbar{\otimes}\mathbf{k}^{-1}\right):\mathsf{k}:\grad\Delta\mathbf{u}\right)\cdot\mathbf{w}dv\thinspace, \end{aligned} \]

for the fluid flux equations

\[ \begin{equation} \begin{aligned}D\delta W_{int}\left[\Delta\mathbf{u}\right]= & \int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\frac{\rho_{T}^{f}}{\varphi^{f}}\left(-\frac{\varphi^{s}}{\varphi^{f}}\left[\left(-\frac{1}{\varphi^{f}}\frac{\dot{J}}{J}+\frac{\gamma}{\beta\Delta t}\right)\divg\Delta\mathbf{u}-\grad^{T}\Delta\mathbf{u}:\mathbf{L}^{s}\right]\mathbf{w}\right)dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\frac{\rho_{T}^{f}}{\varphi^{f}}\left(\varphi^{s}\left(\divg\Delta\mathbf{u}\right)\dot{\mathbf{v}}^{s}+\varphi^{f}\frac{\alpha_{m}}{\alpha_{f}\beta\Delta t^{2}}\Delta\mathbf{u}\right)dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\frac{\rho_{T}^{f}}{\varphi^{f}}\left(-\frac{\varphi^{s}}{\varphi^{f^{2}}}\divg\Delta\mathbf{u}\grad\mathbf{w}+\frac{\gamma}{\beta\Delta t}\grad\Delta\mathbf{u}\right)\cdot\mathbf{w}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\frac{\rho_{T}^{f}}{\varphi^{f}}\frac{1}{\varphi^{f^{2}}}\left[\left(\left(\varphi^{s}+1\right)\frac{1}{\varphi^{f}}\grad\varphi^{f}\left(\divg\Delta\mathbf{u}\right)-\varphi^{s}\grad\left(\divg\Delta\mathbf{u}\right)\right)\cdot\mathbf{w}\right]\mathbf{w}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot-\alpha_{f}\frac{\rho_{T}^{f}}{\varphi^{f}}\mathbf{L}^{f}\cdot\grad\Delta\mathbf{u}\cdot\mathbf{w}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\left(1-\frac{\varphi^{s}}{\varphi^{f}}\right)\left(\divg\Delta\mathbf{u}\right)\rho_{T}^{f}\mathbf{a}^{f}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\left(\left(\divg\Delta\mathbf{u}\right)\mathbf{I}-\grad^{T}\Delta\mathbf{u}\right)\cdot\grad pdv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\left(2\left(\divg\Delta\mathbf{u}\right)\mathbf{k}^{-1}-\grad^{T}\Delta\mathbf{u}\cdot\mathbf{k}^{-1}\right)\cdot\mathbf{w}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot-\alpha_{f}\left(\mathbf{k}^{-1}\cdot\grad\Delta\mathbf{u}+\left(\mathbf{k}^{-1}\underbar{\otimes}\mathbf{k}^{-1}\right):\mathsf{k}:\grad\Delta\mathbf{u}\right)\cdot\mathbf{w}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\frac{1}{\varphi^{f}}\boldsymbol{\tau}\cdot\left(2\frac{\varphi^{s}}{\varphi^{f}}\left(\divg\Delta\mathbf{u}\right)\mathbf{I}+\grad^{T}\Delta\mathbf{u}\right)\cdot\grad\varphi^{f}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot-\alpha_{f}\frac{\varphi^{s}}{\varphi^{f}}\boldsymbol{\tau}\cdot\grad\left(\divg\Delta\mathbf{u}\right)dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot-\alpha_{f}\frac{1}{\varphi^{f}}\boldsymbol{\mathcal{C}}^{\tau}:\left(-\frac{\varphi^{s}}{\varphi^{f^{2}}}\left(\divg\Delta\mathbf{u}\right)\mathbf{D}^{w}+\mathbf{M}\cdot\grad\Delta\mathbf{u}\right)\cdot\grad\varphi^{f}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot-\alpha_{f}\frac{1}{\varphi^{f^{3}}}\boldsymbol{\mathcal{C}}^{\tau}:2\frac{\varphi^{s}}{\varphi^{f}}\left(\divg\Delta\mathbf{u}\right)\mathbf{w}\otimes\grad\varphi^{f}\cdot\grad\varphi^{f}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\frac{1}{\varphi^{f^{3}}}\boldsymbol{\mathcal{C}}^{\tau}:\mathbf{w}\otimes\left(-\grad^{T}\Delta\mathbf{u}\cdot\grad\varphi^{f}\right)\cdot\grad\varphi^{f}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\frac{1}{\varphi^{f^{3}}}\boldsymbol{\mathcal{C}}^{\tau}:\mathbf{w}\otimes\left(-\left(\divg\Delta\mathbf{u}\right)\grad\varphi^{f}+\varphi^{s}\grad\left(\divg\Delta\mathbf{u}\right)\right)\cdot\grad\varphi^{f}dv\\ & +\int_{\Omega^{b}}\alpha_{f}\boldsymbol{\tau}:\grad\delta\mathbf{w}\cdot\left(\left(1-\frac{\varphi^{s}}{\varphi^{f}}\right)\left(\divg\Delta\mathbf{u}\right)\mathbf{I}-\grad\Delta\mathbf{u}\right)dv \end{aligned} \label{eq:Wint-lin-w-u-BFSI} \end{equation} \]
\[ \begin{aligned} & +\int_{\Omega^{b}}\alpha_{f}\grad\delta\mathbf{w}:\boldsymbol{\mathcal{C}}^{\tau}:\left(-\frac{\varphi^{s}}{\varphi^{f^{2}}}\left(\divg\Delta\mathbf{u}\right)\mathbf{D}^{w}+\mathbf{M}\cdot\grad\Delta\mathbf{u}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}\grad\delta\mathbf{w}:\boldsymbol{\mathcal{C}}^{\tau}:2\frac{\varphi^{s}}{\varphi^{f^{3}}}\left(\divg\Delta\mathbf{u}\right)\mathbf{w}\otimes\grad\varphi^{f}dv\\ & +\int_{\Omega^{b}}-\alpha_{f}\grad\delta\mathbf{w}:\boldsymbol{\mathcal{C}}^{\tau}:\frac{1}{\varphi^{f^{2}}}\mathbf{w}\otimes\left(-\grad^{T}\Delta\mathbf{u}\cdot\grad\varphi^{f}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}\grad\delta\mathbf{w}:\boldsymbol{\mathcal{C}}^{\tau}:\frac{1}{\varphi^{f^{2}}}\mathbf{w}\otimes\left(-\left(\divg\Delta\mathbf{u}\right)\grad\varphi^{f}+\varphi^{s}\grad\left(\divg\Delta\mathbf{u}\right)\right)dv\thinspace, \end{aligned} \]

and for the fluid dilatation equations

\[ \begin{equation} \begin{aligned}D\delta W_{int}\left[\Delta\mathbf{u}\right]= & \int_{\Omega^{b}}\alpha_{f}\grad\delta J^{f}\cdot\left(\divg\Delta\mathbf{u}\mathbf{I}-\grad\Delta\mathbf{u}\right)\cdot\mathbf{w}dv\\ & +\int_{\Omega^{b}}\delta J^{f}\alpha_{f}\frac{1}{J^{f}}\left(\dot{J}^{f}\divg\Delta\mathbf{u}+\grad J^{f}\cdot\left(\left(\divg\Delta\mathbf{u}\right)\mathbf{I}-\grad\Delta\mathbf{u}\right)\cdot\mathbf{w}\right)dv\\ & +\int_{\Omega^{b}}-\delta J^{f}\alpha_{f}\left(\divg\mathbf{v}^{s}+\frac{\gamma}{\beta\Delta t}\right)\divg\Delta\mathbf{u}+\grad^{T}\Delta\mathbf{u}:\mathbf{L}^{s}dv\thinspace, \end{aligned} \label{eq:Wint-lin-J-u-BFSI} \end{equation} \]

where \(\mathbf{M}=\frac{\gamma}{\beta\Delta t}\mathbf{I}-\mathbf{L}^{s}\), and \(\boldsymbol{\mathcal{C}}\) is the fourth-order spatial elasticity tensor associated with \(\boldsymbol{\sigma}^{e}\) 1. As done in the CFD formulation (Section Temporal Discretization and Linearization) 2, we have introduced the fourth-order tensor \(\boldsymbol{\mathcal{C}}^{\tau}\) representing the tangent of the viscous stress with respect to the rate of deformation,

\[ \begin{equation} \boldsymbol{\mathcal{C}}^{\tau}=\frac{\partial\boldsymbol{\tau}}{\partial\mathbf{D}^{f}}\,,\label{eq:visc-stress-tan-D-1} \end{equation} \]

where \(\mathbf{D}^{f}\) is the fluid rate of deformation tensor (the symmetric part of \(\mathbf{L}^{f}\)). The tensors \(\boldsymbol{\mathcal{C}}_{d}\) and \(\boldsymbol{\mathcal{C}}\) depend on the choice of constitutive relations for the fluid and solid constituents of \(\Omega^{f}\), respectively. In addition, \(\mathsf{k}\) is the spatial fourth order permeability tensor with respect to right Cauchy-Green tensor \(\mathbf{C}\). The linearized equations include the generalized-\(\alpha\) parameters because the virtual work is evaluated at the intermediate time step, while the increment itself (\(\Delta\mathbf{u}\) in this case) is at the current time step 3.

Following the same procedure, the linearizations of \(\delta W_{int}\) along an increment \(\Delta\mathbf{w}\) is given by

\[ \begin{equation} \begin{aligned}D\delta W_{int}\left[\Delta\mathbf{w}\right]= & \int_{\Omega^{b}}-\alpha_{f}\frac{\varphi^{s}}{\varphi^{f}}\grad\delta\mathbf{v}:\boldsymbol{\mathcal{C}}^{\tau}:\left(\grad\Delta\mathbf{w}-\frac{1}{\varphi^{f}}\Delta\mathbf{w}\otimes\grad\varphi^{f}\right)dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot-\alpha_{f}\varphi^{s}\frac{\rho_{T}^{f}}{\varphi^{f}}\left(\left(\frac{\alpha_{m}}{\gamma\alpha_{f}\Delta t}-\frac{\varphi^{s}}{\varphi^{f}}\frac{\dot{J}}{J}-\frac{1}{\varphi^{f^{2}}}\left(\grad\varphi^{f}\cdot\mathbf{w}\right)\right)\mathbf{I}+\mathbf{L}^{f}\right)\cdot\Delta\mathbf{w}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot-\alpha_{f}\varphi^{s}\frac{\rho_{T}^{f}}{\varphi^{f^{2}}}\grad\Delta\mathbf{w}\cdot\mathbf{w}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot\alpha_{f}\frac{\left(\varphi^{s}\right)^{2}}{\left(\varphi^{f}\right)^{2}}\boldsymbol{\mathcal{C}}^{\tau}:\left(\grad\Delta\mathbf{w}-\frac{1}{\varphi^{f}}\Delta\mathbf{w}\otimes\grad\varphi^{f}\right)\cdot\grad\left(\frac{\varphi^{f}}{\varphi^{s}}\right)dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot-\alpha_{f}\mathbf{k}^{-1}\cdot\Delta\mathbf{w}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\frac{\rho_{T}^{f}}{\varphi^{f}}\left(\left(\frac{\alpha_{m}}{\gamma\alpha_{f}\Delta t}-\frac{\varphi^{s}}{\varphi^{f}}\frac{\dot{J}}{J}-\frac{1}{\varphi^{f^{2}}}\left(\grad\varphi^{f}\cdot\mathbf{w}\right)\right)\mathbf{I}+\mathbf{L}^{f}\right)\cdot\Delta\mathbf{w}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\frac{\rho_{T}^{f}}{\varphi^{f^{2}}}\grad\Delta\mathbf{w}\cdot\mathbf{w}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\mathbf{k}^{-1}\cdot\Delta\mathbf{w}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\frac{1}{\varphi^{f^{2}}}\boldsymbol{\mathcal{C}}^{\tau}:\left(\frac{1}{\varphi^{f}}\Delta\mathbf{w}\otimes\grad\varphi^{f}-\grad\Delta\mathbf{w}\right)\cdot\grad\varphi^{f}dv\\ & +\int_{\Omega^{b}}\alpha_{f}\frac{1}{\varphi^{f}}\grad\delta\mathbf{w}:\boldsymbol{\mathcal{C}}^{\tau}:\left(\grad\Delta\mathbf{w}-\frac{1}{\varphi^{f}}\Delta\mathbf{w}\otimes\grad\varphi^{f}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}\grad\delta J^{f}\cdot\Delta\mathbf{w}+\delta J^{f}\alpha_{f}\frac{1}{J^{f}}\grad J^{f}\cdot\Delta\mathbf{w}dv\thinspace, \end{aligned} \label{eq:Wint-lin-w-BFSI} \end{equation} \]

whereas the linearizations of \(\delta W_{int}\) along an increment \(\Delta J^{f}\) is

\[ \begin{equation} \begin{aligned}D\delta W_{int}\left[\Delta J^{f}\right]= & \int_{\Omega^{b}}-\alpha_{f}\varphi^{s}\boldsymbol{\tau}_{J}^{\prime}:\grad\delta\mathbf{v}\Delta J^{f}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot\alpha_{f}\varphi^{s}\left(\frac{\varphi^{s}}{\varphi^{f}}\boldsymbol{\tau}_{J}^{\prime}\cdot\grad\frac{\varphi^{f}}{\varphi^{s}}+\frac{\rho_{T}^{f}}{J^{f}}\mathbf{a}^{f}\right)\Delta J^{f}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\left(p^{\prime\prime}\grad J^{f}\Delta J^{f}+p^{\prime}\grad\Delta J^{f}\right)dv\\ & +\int_{\Omega^{b}}\delta\mathbf{w}\cdot-\alpha_{f}\left(\frac{\rho_{T}^{f}}{J^{f}}\mathbf{a}^{f}\Delta J^{f}+\frac{1}{\varphi^{f}}\boldsymbol{\tau}_{J}^{\prime}\Delta J^{f}\cdot\grad\varphi^{f}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}\boldsymbol{\tau}_{J}^{\prime}:\grad\delta\mathbf{w}\Delta J^{f}dv\\ & +\int_{\Omega^{b}}\delta J^{f}\alpha_{f}\frac{1}{J^{f}}\left(\varphi^{f}\frac{\alpha_{m}}{\gamma\alpha_{f}\Delta t}-\frac{1}{J^{f}}\left(\varphi^{f}\dot{J}^{f}+\grad J^{f}\cdot\mathbf{w}\right)\right)\Delta J^{f}dv\\ & +\int_{\Omega^{b}}\delta J^{f}\alpha_{f}\frac{1}{J^{f}}\grad\Delta J^{f}\cdot\mathbf{w}dv\thinspace. \end{aligned} \label{eq:Wint-lin-J-BFSI} \end{equation} \]

Here, \(p^{\prime}\) and \(p^{\prime\prime}\) respectively represent the first and second derivatives of \(p\left(J^{f}\right)\). We have also defined \(\boldsymbol{\tau}_{J}^{\prime}\) as the tangent of the viscous stress \(\boldsymbol{\tau}\) with respect to \(J^{f}\),

\[ \begin{equation} \boldsymbol{\tau}_{J}^{\prime}=\frac{\partial\boldsymbol{\tau}}{\partial J^{f}}\,.\label{eq:visc-stress-tan-J-1} \end{equation} \]

For the external work, when \(\mathbf{t}^{\sigma}\), \(\mathbf{t}^{\tau}\), all \(\mathbf{b}\) and \(w_{n}\) are prescribed, the linearizations simplify to

\[ \begin{equation} D\delta W_{ext}\left[\Delta\mathbf{u}\right]=\int_{\Omega^{b}}\delta\mathbf{w}\cdot\alpha_{f}\divg\Delta\mathbf{u}\rho_{T}^{f}\mathbf{b}^{f}dv\,,\label{eq:Wext-lin-u-BFSI} \end{equation} \]
\[ \begin{equation} D\delta W_{ext}\left[\Delta\mathbf{w}\right]=0\,,\label{eq:Wext-lin-w-BFSI} \end{equation} \]
\[ \begin{equation} \begin{aligned}D\delta W_{ext}\left[\Delta J^{f}\right]= & \int_{\Omega^{b}}\delta\mathbf{w}\cdot-\alpha_{f}\frac{\rho_{T}^{f}}{J^{f}}\mathbf{b}^{f}\Delta J^{f}dv\\ & +\int_{\Omega^{b}}\delta\mathbf{v}\cdot\alpha_{f}\varphi^{s}\frac{\rho_{T}^{f}}{J^{f}}\mathbf{b}^{f}\Delta J^{f}dv\,. \end{aligned} \label{eq:Wext-lin-J-BFSI} \end{equation} \]

As discussed in Section Mass and Momentum Balance, we define the fluid dilatation \(e^{f}=J^{f}-1\) as an alternative essential variable, since initial and boundary conditions \(e^{f}=0\) are more convenient to handle in a numerical scheme than \(J^{f}=1\). It follows that \(\grad J^{f}=\grad e^{f}\) and \(\dot{J}^{f}=\dot{e}^{f}\). Therefore the changes to the above equations are minimal, simply requiring the substitution \(J^{f}=1+e^{f}\) and \(\Delta J^{f}=\Delta e^{f}\).

BFSI Spatial Discretization

The degrees of freedom \(\mathbf{u}\left(\mathbf{x},t\right)\), \(\mathbf{w}\left(\mathbf{x},t\right)\), \(J^{f}\left(\mathbf{x},t\right)\) are spatially interpolated over the domain \(\Omega^{f}\) using the same interpolation functions \(N_{a}\left(\mathbf{x}\right)\), with \(a=1\) to \(n\), where \(n\) is the number of nodes in an element,

\[ \begin{equation} \begin{aligned}\mathbf{u}\left(\mathbf{x},t\right) & =\sum_{a=1}^{n}N_{a}\left(\mathbf{x}\right)\mathbf{u}_{a}\,,\\ \mathbf{w}\left(\mathbf{x},t\right) & =\sum_{a=1}^{n}N_{a}\left(\mathbf{x}\right)\mathbf{w}_{a}\,,\\ J^{f}\left(\mathbf{x},t\right) & =\sum_{a=1}^{n}N_{a}\left(\mathbf{x}\right)J_{a}^{f}\,. \end{aligned} \label{eq:spatial-interpol-BFSI} \end{equation} \]

Here, \(\mathbf{u}_{a}\), \(\mathbf{w}_{a}\), and \(J_{a}^{f}\) are the nodal values of the degrees of freedom that evolve over time. These relations may be used to evaluate \(\mathbf{L}^{s}\), \(\mathbf{L}^{w}\), \(\divg\mathbf{w}\), \(\dot{\mathbf{w}}\), \(\grad J^{f}\), \(\dot{J^{f}}\), etc. Similar interpolations are used for virtual increments \(\delta\mathbf{v}\), \(\delta\mathbf{w}\) and \(\delta J^{f}\), as well as real increments \(\Delta\mathbf{u}\), \(\Delta\mathbf{w}\) and \(\Delta J^{f}\). In practice, interpolations \(N_{a}\) are performed in the parametric space of each finite element, which is a material frame.

Now, the discretized form of \(\delta W_{int}\), using eq. \eqref{eq:bfsi-int-virtual-work}, may be written as

\[ \begin{equation} \begin{aligned}\delta W_{int} & =\sum_{a}\delta\mathbf{v}_{a}\cdot\mathbf{f}_{a}^{u}\\ & +\sum_{a}\delta\mathbf{w}_{a}\cdot\mathbf{f}_{a}^{w}\\ & +\sum_{a}\delta J_{a}^{f}f_{a}^{J}\,, \end{aligned} \label{eq:Wint-discret-BFSI} \end{equation} \]

where

\[ \begin{equation} \begin{aligned}\mathbf{f}_{a}^{u}= & \int_{\Omega^{b}}\left(-\varphi^{s}\boldsymbol{\tau}+\boldsymbol{\sigma}^{e}\right)\cdot\grad N_{a}dv\\ & +\int_{\Omega^{b}}N_{a}\left(\varphi^{s}\left(\frac{\varphi^{s}}{\varphi^{f}}\boldsymbol{\tau}\cdot\grad\frac{\varphi^{f}}{\varphi^{s}}-\rho_{T}^{f}\mathbf{a}^{f}+\rho_{T}^{s}\dot{\mathbf{v}}^{s}\right)-\mathbf{k}^{-1}\cdot\mathbf{w}\right)dv\,,\\ \mathbf{f}_{a}^{w}= & \int_{\Omega^{b}}\boldsymbol{\tau}\cdot\grad N_{a}+N_{a}\left(\rho_{T}^{f}\mathbf{a}^{f}+\grad p-\frac{1}{\varphi^{f}}\boldsymbol{\tau}\cdot\grad\varphi^{f}+\mathbf{k}^{-1}\cdot\mathbf{w}\right)dv\,,\\ f_{a}^{J}= & \int_{\Omega^{b}}\grad N_{a}\cdot\mathbf{w}+N_{a}\left(\frac{\varphi^{f}}{J^{f}}\frac{D^{f}J^{f}}{Dt}-\frac{\dot{J}^{s}}{J^{s}}\right)dv\,. \end{aligned} \label{eq:Wint-discret-parts-BFSI} \end{equation} \]

The discretized form of \(D\delta W_{int}\left[\Delta\mathbf{u}\right]\) becomes

\[ \begin{equation} \begin{aligned}D\delta W_{int}\left[\Delta\mathbf{u}\right] & =\sum_{a}\delta\mathbf{v}_{a}\cdot\sum_{b}\mathbf{K}_{ab}^{uu}\cdot\Delta\mathbf{u}_{b}\\ & +\sum_{a}\delta\mathbf{w}_{a}\cdot\sum_{b}\mathbf{K}_{ab}^{wu}\cdot\Delta\mathbf{u}_{b}\\ & +\sum_{a}\delta J_{a}^{f}\sum_{b}\mathbf{k}_{ab}^{Ju}\cdot\Delta\mathbf{u}_{b}\,, \end{aligned} \label{eq:Wint-lin-u-discret-FSI} \end{equation} \]

where

\[ \begin{equation} \begin{aligned}\mathbf{K}_{ab}^{uu}= & \int_{\Omega^{b}}\alpha_{f}\left(\left(\boldsymbol{\sigma}^{e}:\grad N_{b}\otimes\grad N_{a}\right)\mathbf{I}+\grad N_{a}\cdot\boldsymbol{\mathcal{C}}\cdot\grad N_{b}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}\varphi^{s}\boldsymbol{\tau}\cdot\left(\frac{\varphi^{s}}{\varphi^{f}}\grad N_{a}\otimes\grad N_{b}+\grad N_{b}\otimes\grad N_{a}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}\varphi^{s}\grad N_{a}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\grad N_{b}\cdot\left(-\frac{\varphi^{s}}{\varphi^{f^{2}}}\mathbf{D}^{w}+\mathbf{M}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}\frac{\varphi^{s}}{\varphi^{f^{2}}}\grad N_{a}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\left(2\frac{\varphi^{s}}{\varphi^{f}}\grad\varphi^{f}\otimes\grad N_{b}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}\frac{\varphi^{s}}{\varphi^{f^{2}}}\grad N_{a}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\left(\grad N_{b}\otimes\grad\varphi^{f}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}\frac{\varphi^{s}}{\varphi^{f^{2}}}\grad N_{a}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\left(-\grad\varphi^{f}\otimes\grad N_{b}+\varphi^{s}\grad\grad N_{b}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}N_{a}\frac{\varphi^{s^{2}}}{\varphi^{f}}\boldsymbol{\tau}\cdot\left(2\frac{\varphi^{s}}{\varphi^{f}}\grad\frac{\varphi^{f}}{\varphi^{s}}\otimes\grad N_{b}+\grad N_{b}\otimes\grad\frac{\varphi^{f}}{\varphi^{s}}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\frac{\varphi^{s^{2}}}{\varphi^{f}}\boldsymbol{\tau}\cdot\frac{1}{\varphi^{s}}\grad\grad N_{b}dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\frac{\varphi^{s^{2}}}{\varphi^{f}}\grad\frac{\varphi^{f}}{\varphi^{s}}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\grad N_{b}\cdot\left(-\frac{\varphi^{s}}{\varphi^{f^{2}}}\mathbf{D}^{w}+\mathbf{M}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\frac{\varphi^{s^{2}}}{\varphi^{f^{3}}}\grad\frac{\varphi^{f}}{\varphi^{s}}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\left(2\frac{\varphi^{s}}{\varphi^{f}}\grad\varphi^{f}\otimes\grad N_{b}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\frac{\varphi^{s^{2}}}{\varphi^{f^{3}}}\grad\frac{\varphi^{f}}{\varphi^{s}}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\left(\grad N_{b}\otimes\grad\varphi^{f}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}N_{a}\frac{\varphi^{s^{2}}}{\varphi^{f^{3}}}\grad\frac{\varphi^{f}}{\varphi^{s}}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\left(-\grad\varphi^{f}\otimes\grad N_{b}+\varphi^{s}\grad\grad N_{b}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}\frac{\varphi^{s}}{\varphi^{f}}\rho_{T}^{f}N_{a}\left(\varphi^{s}\dot{\mathbf{v}}^{s}\otimes\grad N_{b}+\varphi^{f}\frac{\alpha_{m}}{\alpha_{f}\beta\Delta t^{2}}N_{b}\mathbf{I}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}\rho_{T}^{f}N_{a}\frac{\varphi^{s^{2}}}{\varphi^{f^{2}}}\left[\left(\left(-\frac{1}{\varphi^{f}}\right)\frac{\dot{J}}{J}+\frac{\gamma}{\beta\Delta t}\right)\mathbf{w}\otimes\grad N_{b}-\mathbf{w}\otimes\left(\mathbf{L}^{s^{T}}\cdot\grad N_{b}\right)\right]dv\\ & +\int_{\Omega^{b}}-\alpha_{f}N_{a}\varphi^{s}\frac{\rho_{T}^{f}}{\varphi^{f}}\left(-\frac{\varphi^{s}}{\varphi^{f^{2}}}\mathbf{L}^{w}\cdot\mathbf{w}\otimes\grad N_{b}+\frac{\gamma}{\beta\Delta t}\left(\grad N_{b}\cdot\mathbf{w}\right)\mathbf{I}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}N_{a}\rho_{T}^{f}\frac{\varphi^{s}}{\varphi^{f^{4}}}\left(\varphi^{s}+1\right)\left(\grad\varphi^{f}\cdot\mathbf{w}\right)\mathbf{w}\otimes\grad N_{b}dv \end{aligned} \label{eq:Wint-lin-uu-discret-BFSI} \end{equation} \]
\[ \begin{aligned} & +\int_{\Omega^{b}}\alpha_{f}N_{a}\rho_{T}^{f}\frac{\varphi^{s^{2}}}{\varphi^{f^{3}}}\mathbf{w}\otimes\grad\grad^{T}N_{b}\cdot\mathbf{w}dv\\ & +\int_{\Omega^{b}}\alpha_{f}\frac{\varphi^{s}}{\varphi^{f}}\rho_{T}^{f}N_{a}\left(\mathbf{L}^{f}\left(\grad N_{b}\cdot\mathbf{w}\right)+\varphi^{s}\mathbf{a}^{f}\otimes\grad N_{b}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\frac{\alpha_{m}}{\alpha_{f}\beta\Delta t^{2}}\rho^{s}N_{b}\mathbf{I}dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\left(-2\left(\mathbf{k}^{-1}\cdot\mathbf{w}\right)\otimes\grad N_{b}+\grad N_{b}\otimes\left(\mathbf{k}^{-1}\cdot\mathbf{w}\right)\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\left(\left(\grad N_{b}\cdot\mathbf{w}\right)\mathbf{k}^{-1}+\left(\mathbf{k}^{-1}\underbar{\otimes}\mathbf{k}^{-1}\right):\mathsf{k}:\left(\grad N_{b}\cdot\mathbf{w}\right)\mathbf{I}\right)dv\,, \end{aligned} \]
\[ \begin{equation} \begin{aligned}\mathbf{K}_{ab}^{wu}= & \int_{\Omega^{b}}\alpha_{f}\frac{\rho_{T}^{f}}{\varphi^{f}}N_{a}\left(\varphi^{s}\dot{\mathbf{v}}^{s}\otimes\grad N_{b}+\varphi^{f}\frac{\alpha_{m}}{\alpha_{f}\beta\Delta t^{2}}N_{b}\mathbf{I}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}\frac{\rho_{T}^{f}}{\varphi^{f}}N_{a}\frac{\varphi^{s}}{\varphi^{f}}\left(-\frac{1}{\varphi^{f}}\frac{\dot{J}}{J}+\frac{\gamma}{\beta\Delta t}\right)\mathbf{w}\otimes\grad N_{b}dv\\ & +\int_{\Omega^{b}}\alpha_{f}\frac{\rho_{T}^{f}}{\varphi^{f}}N_{a}\frac{\varphi^{s}}{\varphi^{f}}\mathbf{w}\otimes\left(\mathbf{L}^{s^{T}}\cdot\grad N_{b}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\frac{\rho_{T}^{f}}{\varphi^{f}}\left(-\frac{\varphi^{s}}{\varphi^{f^{2}}}\left(\mathbf{L}^{w}\cdot\mathbf{w}\right)\otimes\grad N_{b}+\frac{\gamma}{\beta\Delta t}\left(\grad N_{b}\cdot\mathbf{w}\right)\mathbf{I}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\frac{\rho_{T}^{f}}{\varphi^{f^{4}}}\left(\varphi^{s}+1\right)\left(\grad\varphi^{f}\cdot\mathbf{w}\right)\mathbf{w}\otimes\grad N_{b}dv\\ & +\int_{\Omega^{b}}-\alpha_{f}N_{a}\frac{\rho_{T}^{f}}{\varphi^{f^{3}}}\varphi^{s}\mathbf{w}\otimes\left(\grad\grad^{T}N_{b}\cdot\mathbf{w}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}\frac{\rho_{T}^{f}}{\varphi^{f}}N_{a}\mathbf{L}^{f}\left(\grad N_{b}\cdot\mathbf{w}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\left(1-\frac{\varphi^{s}}{\varphi^{f}}\right)\rho_{T}^{f}\mathbf{a}^{f}\otimes\grad N_{b}dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\left(\grad p\otimes\grad N_{b}-\grad N_{b}\otimes\grad p\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\left(2\left(\mathbf{k}^{-1}\cdot\mathbf{w}\right)\otimes\grad N_{b}-\grad N_{b}\otimes\left(\mathbf{k}^{-1}\cdot\mathbf{w}\right)\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}N_{a}\left(\left(\grad N_{b}\cdot\mathbf{w}\right)\mathbf{k}^{-1}+\left(\mathbf{k}^{-1}\underbar{\otimes}\mathbf{k}^{-1}\right):\mathsf{k}:\left(\grad N_{b}\cdot\mathbf{w}\right)\mathbf{I}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\frac{1}{\varphi^{f}}\boldsymbol{\tau}\cdot\left(2\frac{\varphi^{s}}{\varphi^{f}}\grad\varphi^{f}\otimes\grad N_{b}+\grad N_{b}\otimes\grad\varphi^{f}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}N_{a}\frac{1}{\varphi^{f}}\boldsymbol{\tau}\cdot\varphi^{s}\grad\grad N_{b}dv\\ & +\int_{\Omega^{b}}-\alpha_{f}N_{a}\frac{1}{\varphi^{f}}\grad\varphi^{f}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\grad N_{b}\cdot\left(-\frac{\varphi^{s}}{\varphi^{f^{2}}}\mathbf{D}^{w}+\mathbf{M}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}N_{a}\frac{1}{\varphi^{f^{3}}}\grad\varphi^{f}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\left(2\frac{\varphi^{s}}{\varphi^{f}}\grad\varphi^{f}\otimes\grad N_{b}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}N_{a}\frac{1}{\varphi^{f^{3}}}\grad\varphi^{f}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\left(\grad N_{b}\otimes\grad\varphi^{f}\right)dv \end{aligned} \label{eq:Wint-lin-wu-discret-BFSI} \end{equation} \]
\[ \begin{aligned} & +\int_{\Omega^{b}}\alpha_{f}N_{a}\frac{1}{\varphi^{f^{3}}}\grad\varphi^{f}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\left(-\grad\varphi^{f}\otimes\grad N_{b}+\varphi^{s}\grad\grad N_{b}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}\boldsymbol{\tau}\cdot\left(\left(1-\frac{\varphi^{s}}{\varphi^{f}}\right)\grad N_{a}\otimes\grad N_{b}-\grad N_{b}\otimes\grad N_{a}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}\grad N_{a}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\grad N_{b}\cdot\left(-\frac{\varphi^{s}}{\varphi^{f^{2}}}\mathbf{D}^{w}+\mathbf{M}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}\frac{1}{\varphi^{f^{2}}}\grad N_{a}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\left(2\frac{\varphi^{s}}{\varphi^{f}}\grad\varphi^{f}\otimes\grad N_{b}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}\frac{1}{\varphi^{f^{2}}}\grad N_{a}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\left(\grad N_{b}\otimes\grad\varphi^{f}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}\frac{1}{\varphi^{f^{2}}}\grad N_{a}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\left(-\grad\varphi^{f}\otimes\grad N_{b}+\varphi^{s}\grad\grad N_{b}\right)dv\,, \end{aligned} \]
\[ \begin{equation} \begin{aligned}\mathbf{k}_{ab}^{Ju}= & \int_{\Omega^{b}}\alpha_{f}\left(\grad N_{b}\otimes\mathbf{w}-\left(\grad N_{b}\cdot\mathbf{w}\right)\mathbf{I}\right)\cdot\grad N_{a}dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\frac{1}{J^{f}}\left(\dot{J}^{f}\grad N_{b}+\left(\grad N_{b}\otimes\mathbf{w}-\grad N_{b}\cdot\mathbf{w}\mathbf{I}\right)\cdot\grad J^{f}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}N_{a}\left(\frac{\dot{J}}{J}+\frac{\gamma}{\beta\Delta t}\right)\grad N_{b}+\mathbf{L}^{s^{T}}\cdot\grad N_{b}dv\,, \end{aligned} \label{eq:Wint-lin-Ju-discret-BFSI} \end{equation} \]

whereas that of \(D\delta W_{int}\left[\Delta\mathbf{w}\right]\) becomes

\[ \begin{equation} \begin{aligned}D\delta W_{int}\left[\Delta\mathbf{w}\right] & =\sum_{a}\delta\mathbf{v}_{a}\cdot\sum_{b}\mathbf{K}_{ab}^{uw}\cdot\Delta\mathbf{w}_{b}\\ & +\sum_{a}\delta\mathbf{w}_{a}\cdot\sum_{b}\mathbf{K}_{ab}^{ww}\cdot\Delta\mathbf{w}_{b}\\ & +\sum_{a}\delta J_{a}^{f}\sum_{b}\mathbf{k}_{ab}^{Jw}\cdot\Delta\mathbf{w}_{b}\,, \end{aligned} \label{eq:Wint-lin-w-discret-BFSI} \end{equation} \]

where

\[ \begin{equation} \begin{aligned}\mathbf{K}_{ab}^{uw}= & \int_{\Omega^{b}}\alpha_{f}\frac{\varphi^{s}}{\varphi^{f}}\grad N_{a}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\left(\frac{1}{\varphi^{f}}N_{b}\grad\varphi^{f}-\grad N_{b}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}\varphi^{s}\frac{\rho_{T}^{f}}{\varphi^{f}}N_{a}\left(\left(\frac{\alpha_{m}}{\gamma\alpha_{f}\Delta t}-\frac{\varphi^{s}}{\varphi^{f}}\frac{\dot{J}}{J}-\frac{1}{\varphi^{f^{2}}}\left(\grad\varphi^{f}\cdot\mathbf{w}\right)\right)\mathbf{I}+\mathbf{L}^{f}\right)N_{b}dv\\ & +\int_{\Omega^{b}}-\alpha_{f}\varphi^{s}\frac{\rho_{T}^{f}}{\varphi^{f^{2}}}N_{a}\left(\grad N_{b}\cdot\mathbf{w}\right)\mathbf{I}dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\frac{\left(\varphi^{s}\right)^{2}}{\left(\varphi^{f}\right)^{2}}\grad\frac{\varphi^{f}}{\varphi^{s}}\cdot\boldsymbol{\mathcal{C}}_{d}\cdot\left(-\frac{1}{\varphi^{f}}N_{b}\grad\varphi^{f}+\grad N_{b}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}N_{a}N_{b}\mathbf{k}^{-1}dv\,,\\ \mathbf{K}_{ab}^{ww}= & \int_{\Omega^{b}}\alpha_{f}N_{a}\frac{\rho_{T}^{f}}{\varphi^{f}}\left(\left(\frac{\alpha_{m}}{\gamma\alpha_{f}\Delta t}-\frac{\varphi^{s}}{\varphi^{f}}\frac{\dot{J}}{J}-\frac{1}{\varphi^{f^{2}}}\left(\grad\varphi^{f}\cdot\mathbf{w}\right)\right)\mathbf{I}+\mathbf{L}^{f}\right)N_{b}dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\frac{\rho_{T}^{f}}{\varphi^{f^{2}}}\left(\grad N_{b}\cdot\mathbf{w}\right)\mathbf{I}dv\\ & +\int_{\Omega^{b}}\alpha_{f}N_{a}\left(\frac{1}{\varphi^{f^{2}}}\grad\varphi^{f}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\left(\frac{1}{\varphi^{f}}N_{b}\grad\varphi^{f}-\grad N_{b}\right)+N_{b}\mathbf{k}^{-1}\right)dv\\ & +\int_{\Omega^{b}}\alpha_{f}\frac{1}{\varphi^{f}}\grad N_{a}\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\left(-\frac{1}{\varphi^{f}}N_{b}\grad\varphi^{f}+\grad N_{b}\right)dv\,,\\ \mathbf{k}_{ab}^{Jw}= & \int_{\Omega^{b}}\alpha_{f}N_{b}\left(\grad N_{a}+N_{a}\frac{1}{J^{f}}\grad J^{f}\right)dv\,, \end{aligned} \label{eq:Wint-lin-w-discret-parts-BFSI} \end{equation} \]

and finally, for \(D\delta W_{int}\left[\Delta J^{f}\right]\) the equations become

\[ \begin{equation} \begin{aligned}D\delta W_{int}\left[\Delta J^{f}\right]= & \sum_{a}\delta\mathbf{v}_{a}\cdot\sum_{b}\mathbf{k}_{ab}^{uJ}\Delta J_{b}^{f}\\ & +\sum_{a}\delta\mathbf{w}_{a}\cdot\sum_{b}\mathbf{k}_{ab}^{wJ}\Delta J_{b}^{f}\\ & +\sum_{a}\delta J_{a}^{f}\sum_{b}k_{ab}^{JJ}\Delta J_{b}^{f}\,, \end{aligned} \label{eq:Wint-lin-J-discret-BFSI} \end{equation} \]

where

\[ \begin{equation} \begin{aligned}\mathbf{k}_{ab}^{uJ}= & \int_{\Omega^{b}}\alpha_{f}\varphi^{s}N_{a}N_{b}\left(\frac{\varphi^{s}}{\varphi^{f}}\boldsymbol{\tau}_{J}^{\prime}\cdot\grad\frac{\varphi^{f}}{\varphi^{s}}+\frac{\rho_{T}^{f}}{J^{f}}\mathbf{a}^{f}\right)dv\\ & +\int_{\Omega^{b}}-\alpha_{f}N_{b}\varphi^{s}\boldsymbol{\tau}_{J}^{\prime}\cdot\grad N_{a}dv\,,\\ \mathbf{k}_{ab}^{wJ}= & \int_{\Omega^{b}}\alpha_{f}N_{a}\left(p^{\prime}\grad N_{b}+N_{b}\left(p^{\prime\prime}\grad J^{f}-\frac{\rho_{T}^{f}}{J^{f}}\mathbf{a}^{f}-\frac{1}{\varphi^{f}}\boldsymbol{\tau}_{J}^{\prime}\cdot\grad\varphi^{f}\right)\right)dv\Delta J_{b}^{f}\\ & +\int_{\Omega^{b}}\alpha_{f}N_{b}\boldsymbol{\tau}_{J}^{\prime}\cdot\grad N_{a}dv\,,\\ k_{ab}^{JJ}= & \int_{\Omega^{b}}\alpha_{f}N_{a}\frac{1}{J^{f}}\left(\left(\varphi^{f}\frac{\alpha_{m}}{\gamma\alpha_{f}\Delta t}-\frac{1}{J^{f}}\left(\varphi^{f}\dot{J}^{f}+\grad J^{f}\cdot\mathbf{w}\right)\right)N_{b}+\grad N_{b}\cdot\mathbf{w}\right)dv\,, \end{aligned} \label{eq:Wint-lin-J-discret-parts-BFSI} \end{equation} \]

for external virtual work in eq. \eqref{eq:bfsi-ext-virtual-work}, the discretized equations are

\[ \begin{equation} \begin{aligned}\delta W_{ext} & =\sum_{a}\delta\mathbf{v}_{a}\cdot\mathbf{f}_{a}^{u}\\ & +\sum_{a}\delta\mathbf{w}_{a}\cdot\mathbf{f}_{a}^{w}\\ & +\sum_{a}\delta J_{a}^{f}f_{a}^{J}\,, \end{aligned} \label{eq:Wext-discret-BFSI} \end{equation} \]

where

\[ \begin{equation} \begin{aligned}\mathbf{f}_{a,ext}^{u} & =\int_{\partial\Omega^{b}}N_{a}\mathbf{t}^{\sigma}da+\int_{\Omega^{b}}N_{a}\varphi^{s}\left(-\rho_{T}^{f}\mathbf{b}^{f}+\rho_{T}^{s}\mathbf{b}^{s}\right)dv\,,\\ \mathbf{f}_{a,ext}^{w} & =\int_{\partial\Omega^{b}}N_{a}\mathbf{t}^{\tau}da+\int_{\Omega^{b}}\rho_{T}^{f}N_{a}\mathbf{b}^{f}dv\,,\\ f_{a,ext}^{J} & =\int_{\partial\Omega^{b}}N_{a}w_{n}da\,, \end{aligned} \label{eq:Wext-discret-parts-BFSI} \end{equation} \]

and the discretized forms of the linearized external virtual work are

\[ \begin{equation} \begin{aligned}D\delta W_{ext}\left[\Delta\mathbf{u}\right] & =\sum_{a}\delta\mathbf{w}_{a}\cdot\sum_{b}\mathbf{K}_{ab}^{wu}\cdot\Delta\mathbf{u}_{b}\,,\\ D\delta W_{ext}\left[\Delta J^{f}\right] & =\sum_{a}\delta\mathbf{v}_{a}\cdot\sum_{b}\mathbf{k}_{ab}^{uJ}\Delta J_{b}^{f}+\sum_{a}\delta\mathbf{w}_{a}\cdot\sum_{b}\mathbf{k}_{ab}^{wJ}\Delta J_{b}^{f}\,, \end{aligned} \label{eq:Wext-lin-discret-BFSI} \end{equation} \]

where

\[ \begin{equation} \begin{aligned}\mathbf{K}_{ab,ext}^{wu}= & \int_{\Omega^{b}}\alpha_{f}N_{a}\rho_{T}^{f}\mathbf{b}^{f}\otimes\grad N_{b}dv\,,\\ \mathbf{k}_{ab,ext}^{uJ}= & -\int_{\Omega^{b}}\alpha_{f}\varphi^{s}\frac{\rho_{T}^{f}}{J^{f}}N_{a}N_{b}\mathbf{b}^{f}dv\,,\\ \mathbf{k}_{ab,ext}^{wJ}= & -\int_{\Omega^{b}}\alpha_{f}\frac{\rho_{T}^{f}}{J^{f}}N_{a}N_{b}\mathbf{b}^{f}dv\,. \end{aligned} \label{eq:Wext-lin-discret-parts-BFSI} \end{equation} \]

Combining these results, from the linearized virtual work equation of (3.5-5) we can represent the system of equations in a compact matrix form as

\[ \begin{equation} \left[\begin{array}{ccc} \mathbf{K}_{ab}^{uu} & \mathbf{K}_{ab}^{uw} & \mathbf{k}_{ab}^{uJ}-\mathbf{k}_{ab,ext}^{uJ}\\ \mathbf{K}_{ab}^{wu}-\mathbf{K}_{ab,ext}^{wu} & \mathbf{K}_{ab}^{ww} & \mathbf{k}_{ab}^{wJ}-\mathbf{k}_{ab,ext}^{wJ}\\ \mathbf{k}_{ab}^{Ju} & \mathbf{k}_{ab}^{Jw} & k_{ab}^{JJ} \end{array}\right]\left[\begin{array}{c} \Delta\mathbf{u}\\ \Delta\mathbf{w}\\ \Delta J^{f} \end{array}\right]=\left[\begin{array}{c} \mathbf{f}_{a,ext}^{u}-\mathbf{f}_{a}^{u}\\ \mathbf{f}_{a,ext}^{w}-\mathbf{f}_{a}^{w}\\ f_{a,ext}^{J}-f_{a}^{J} \end{array}\right]\,.\label{eq:Virtual-work-matrix-BFSI} \end{equation} \]

BFSI Traction Interface

The virtual work \(\delta F\) on a biphasic-FSI interface was given in eq. \eqref{eq:bfsi-traction-final}. The linearizations of \(\delta F\) are given by

\[ \begin{equation} \begin{aligned}D\delta F\left[\Delta\mathbf{u}\right]= & \int_{\partial\Omega^{b}}\delta\mathbf{v}\cdot\alpha_{f}\frac{\varphi^{s}}{\varphi^{f}}\left(\divg\Delta\mathbf{u}\right)\mathbf{T}_{d}\cdot\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\thinspace d\eta^{1}d\eta^{2}\\ & +\int_{\partial\Omega^{b}}-\delta\mathbf{v}\cdot\boldsymbol{\mathcal{C}}^{\tau}:\alpha_{f}\left(-\frac{\varphi^{s}}{\varphi^{f^{2}}}\left(\divg\Delta\mathbf{u}\right)\mathbf{D}^{w}+\mathbf{M}\cdot\grad\Delta\mathbf{u}\right)\cdot\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\thinspace d\eta^{1}d\eta^{2}\\ & +\int_{\partial\Omega^{b}}-\delta\mathbf{v}\cdot\boldsymbol{\mathcal{C}}^{\tau}:\alpha_{f}\left(2\frac{\varphi^{s}}{\varphi^{f^{3}}}\left(\divg\Delta\mathbf{u}\right)\mathbf{w}\otimes\grad\varphi^{f}\right)\cdot\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\thinspace d\eta^{1}d\eta^{2}\\ & +\int_{\partial\Omega^{b}}-\delta\mathbf{v}\cdot\boldsymbol{\mathcal{C}}^{\tau}:\alpha_{f}\left(-\frac{1}{\varphi^{f^{2}}}\mathbf{w}\otimes\left(-\grad^{T}\Delta\mathbf{u}\cdot\grad\varphi^{f}\right)\right)\cdot\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\thinspace d\eta^{1}d\eta^{2}\\ & +\int_{\partial\Omega^{b}}\delta\mathbf{v}\cdot\boldsymbol{\mathcal{C}}^{\tau}:\alpha_{f}\left(\frac{\varphi^{s}}{\varphi^{f^{2}}}\mathbf{w}\otimes\left(-\frac{1}{J}\left(\divg\Delta\mathbf{u}\right)\grad J+\grad\left(\divg\Delta\mathbf{u}\right)\right)\right)\cdot\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\thinspace d\eta^{1}d\eta^{2}\\ & +\int_{\partial\Omega^{b}}-\delta\mathbf{v}\cdot\alpha_{f}\boldsymbol{\tau}\cdot\left(\frac{\partial\Delta\mathbf{u}}{\partial\eta^{1}}\times\mathbf{g}_{2}+\mathbf{g}_{1}\times\frac{\partial\Delta\mathbf{u}}{\partial\eta^{2}}\right)\thinspace d\eta^{1}d\eta^{2}\,,\\ D\delta F\left[\Delta\mathbf{w}\right] & =\int_{\partial\Omega^{b}}-\delta\mathbf{v}\cdot\alpha_{f}\boldsymbol{\mathcal{C}}^{\tau}:\left(\frac{1}{\varphi^{f}}\grad\Delta\mathbf{w}-\frac{1}{\varphi^{f^{2}}}\Delta\mathbf{w}\otimes\grad\varphi^{f}\right)\cdot\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\thinspace d\eta^{1}d\eta^{2}\,,\\ D\delta F\left[\Delta J^{f}\right] & =\int_{\partial\Omega^{b}}-\delta\mathbf{v}\cdot\alpha_{f}\left(-p^{\prime}\mathbf{I}+\boldsymbol{\tau}_{J}^{\prime}\right)\cdot\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\Delta J^{f}\thinspace d\eta^{1}d\eta^{2}\,. \end{aligned} \label{eq:BFSI-int-lin} \end{equation} \]

The discretized forms of this virtual work is

\[ \begin{equation} \delta F=\sum_{a}\delta\mathbf{v}_{a}\cdot\mathbf{f}_{a}\,,\label{eq:BSI-int-discret} \end{equation} \]

where

\[ \begin{equation} \mathbf{f}_{a}=\int_{\partial\Omega^{b}}-N_{a}\left(-p\mathbf{I}+\boldsymbol{\tau}\right)\cdot\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\thinspace d\eta^{1}d\eta^{2}\,.\label{eq:BFSI-int-discret-parts} \end{equation} \]

The discretized forms of the linearizations are

\[ \begin{equation} \begin{aligned}D\delta F\left[\Delta\mathbf{u}\right] & =\sum_{a}\delta\mathbf{v}_{a}\cdot\sum_{b}\mathbf{K}_{ab}^{uu}\cdot\Delta\mathbf{u}_{b}\,,\\ D\delta F\left[\Delta\mathbf{w}\right] & =\sum_{a}\delta\mathbf{v}_{a}\cdot\sum_{b}\mathbf{K}_{ab}^{uw}\cdot\Delta\mathbf{w}_{b}\,,\\ D\delta F\left[\Delta J^{f}\right] & =\sum_{a}\delta\mathbf{v}_{a}\cdot\sum_{b}\mathbf{k}_{ab}^{uJ}\Delta J_{b}^{f}\,, \end{aligned} \label{eq:BFSI-int-lin-discret} \end{equation} \]

where

\[ \begin{equation} \begin{aligned}\mathbf{K}_{ab}^{uu}= & \int_{\partial\Omega^{b}}\alpha_{f}N_{a}\frac{\varphi^{s}}{\varphi^{f}}\left(\boldsymbol{\tau}\cdot\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\right)\otimes\grad N_{b}\thinspace d\eta^{1}d\eta^{2}\\ & +\int_{\partial\Omega^{b}}-\alpha_{f}N_{a}\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\cdot\boldsymbol{\mathcal{C}}_{d}\cdot\grad N_{b}\cdot\left(-\frac{\varphi^{s}}{\varphi^{f^{2}}}\mathbf{D}^{w}+\mathbf{M}\right)\thinspace d\eta^{1}d\eta^{2}\\ & +\int_{\partial\Omega^{b}}-\alpha_{f}N_{a}\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\left(2\frac{\varphi^{s}}{\varphi^{f^{3}}}\grad\varphi^{f}\otimes\grad N_{b}\right)\thinspace d\eta^{1}d\eta^{2}\\ & +\int_{\partial\Omega^{b}}-\alpha_{f}N_{a}\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\left(\frac{1}{\varphi^{f^{2}}}\grad N_{b}\otimes\grad\varphi^{f}\right)\thinspace d\eta^{1}d\eta^{2}\\ & +\int_{\partial\Omega^{b}}\alpha_{f}N_{a}\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\cdot\boldsymbol{\mathcal{C}}^{\tau}\cdot\mathbf{w}\cdot\frac{\varphi^{s}}{\varphi^{f^{2}}}\left(-\frac{1}{J}\grad J\otimes\grad N_{b}+\grad\grad N_{b}\right)\thinspace d\eta^{1}d\eta^{2}\\ & +\int_{\partial\Omega^{b}}-\alpha_{f}N_{a}\left(-p\mathbf{I}+\boldsymbol{\tau}\right)\cdot\boldsymbol{\mathcal{A}}\left\{ -\mathbf{g}_{2}\frac{\partial N_{b}}{\partial\eta^{1}}+\mathbf{g}_{1}\frac{\partial N_{b}}{\partial\eta^{2}}\right\} \thinspace d\eta^{1}d\eta^{2}\,,\\ \mathbf{K}_{ab}^{uw}= & \int_{\partial\Omega^{b}}-\alpha_{f}N_{a}\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\cdot\boldsymbol{\mathcal{C}}_{d}\cdot\frac{1}{\varphi^{f}}\left(\grad N_{b}-\frac{1}{\varphi^{f}}N_{b}\grad\varphi^{f}\right)\thinspace d\eta^{1}d\eta^{2}\,,\\ \mathbf{k}_{ab}^{uJ}= & \int_{\partial\Omega^{b}}-\alpha_{f}N_{a}N_{b}\left(-p^{\prime}\mathbf{I}+\boldsymbol{\tau}_{J}^{\prime}\right)\cdot\left(\mathbf{g}_{1}\times\mathbf{g}_{2}\right)\thinspace d\eta^{1}d\eta^{2}\,. \end{aligned} \label{eq:BFSI-int-lin-discret-parts} \end{equation} \]

Note that \(\boldsymbol{\mathcal{A}}\left\{ \cdot\right\}\) represents the skew-symmetric tensor form of the bracketed vector.


  1. Bonet, Javier; Wood, Richard D.. "Nonlinear continuum mechanics for finite element analysis." Cambridge University Press (1997). 

  2. Ateshian, Gerard A; Shim, Jay J; Maas, Steve A; Weiss, Jeffrey A. "Finite Element Framework for Computational Fluid Dynamics in FEBio." J Biomech Eng, vol. 140 (2018). 

  3. Jansen, Kenneth E; Whiting, Christian H; Hulbert, Gregory M. "A generalized-\(\alpha\) method for integrating the filtered {Navier}--{Stokes} equations with a stabilized finite element method." Comput. Methods Appl. Mech. Engrg., vol. 190, pp. 305--319 (2000).