betaFeedback

WorksMathematics

Entropy-Compatible Barrier Schemes for Diffusive FENE Flows

Abstract

FENE-type conformation-tensor models impose a finite-extensibility constraint that is absent from Oldroyd–B flow: the conformation tensor must satisfy \CC\succ0 and \tr\CC<L^2. Positive definiteness alone is therefore insufficient, since a numerical state can remain positive while crossing the singular trace barrier. Even a trace-preserving logarithmic parametrization is not enough by itself: high-order reconstruction can remain inside the finite-extensibility domain while injecting artificial FENE entropy. We develop and analyze a barrier-preserving entropy-compatible discretization for FENE-P type flows with polymer center-of-mass molecular diffusion and for trace-singular FENE-family closures with the same entropy structure. The method combines a trace-barrier free energy, a finite-extensibility logarithmic parametrization, a least-damping entropy-compatible barrier-log reconstruction, molecular diffusion paired with the barrier entropy variable, compatible quadrature for polymeric work, and a scaled FENE stress variable for the small-Weissenberg limit. For admissible discrete states we prove finite-extensibility preservation at entropy quadrature points, existence and bisection computability of the maximal entropy-admissible reconstruction parameter, a fully discrete free-energy inequality with relaxation and molecular-diffusion barrier dissipation, a quantitative AP stress closure, and a fixed-discretization Newtonian limit. A conditional relative-entropy estimate is derived on compact subsets of the finite-extensibility domain. Numerical diagnostics verify barrier preservation, entropy-compatible reconstruction, energy decay, AP closure, coupled velocity–pressure–stress accuracy, and high-Weissenberg robustness near the trace constraint.

Introduction

The Oldroyd–B conformation tensor is constrained only by positive definiteness. In finitely extensible nonlinear elastic models the admissible set is smaller. If CC denotes the conformation tensor and L2L^2 is the squared maximum extension parameter, a FENE-type closure requires

C∈DL:={C∈S++d:tr⁡C<L2}.C \in\mathcal{D}_L := \left\{ C \in\mathbb{S}_{++}^{d} : \operatorname{tr} C < L^2 \right\}.
  • (v) an AP stress-closure estimate showing Sh=D(uh)+O(Wi)S_h = D(u_h) + O(\mathrm{Wi}) at fixed hh and Δt\Delta t;

  • (vi) a conditional relative-entropy estimate on compact subsets of the finite-extensibility domain;

  • (vii) numerical diagnostics that test the barrier, entropy correction, AP limit, coupled pressure–stress feedback, and high-Weissenberg behavior.

Related work on FENE closures, log-conformation variables, energy-stable viscoelastic schemes, and asymptotic-preserving relaxation discretizations is extensive; see, for example, [3, 5, 6, 7, 9, 2, 4, 8, 10]. The present paper is closest in spirit to structure-preserving conformation-tensor methods, but the finite-extensibility constraint introduces a trace barrier that is not present in Oldroyd–B. This extra constraint is the focus of the analysis.

IssueOldroyd–BFENE-P type model
Admissible setC≻0C \succ0C≻0C \succ0 and tr⁡C<L2\operatorname{tr} C < L^2
Entropy singularityλmin⁡(C)↓0\lambda_{\min}(C) \downarrow0λmin⁡(C)↓0\lambda_{\min}(C) \downarrow0 or tr⁡C↑L2\operatorname{tr} C \uparrow L^2
Stress lawC−IC-IfL(C)C−If_L(C)C-I with singular fLf_L
Solver parametrizationlog-conformation suffices for positivitylog-conformation must be combined with a trace barrier
Reconstructionlog reconstruction remains in S++d\mathbb{S}_{++}^{d}barrier-log reconstruction remains in DL\mathcal{D}_L and is filtered by FENE entropy
AP variable(C−I)/Wi(C-I)/\mathrm{Wi}(fL(C)C−I)/Wi(f_L(C)C-I)/\mathrm{Wi}

Table 1. Structural differences between Oldroyd–B and the FENE-P setting considered here.

The paper retains the entropy identity, finite-extensibility preservation, entropy-compatible barrier-log reconstruction, the discrete energy estimate, the AP closure, and the core benchmarks. This organization keeps the finite-extensibility mechanism visible while connecting each diagnostic to a structural claim.

Position relative to logarithmic and factor reconstructions. Classical log-conformation variables 1 and square-root or symmetric-factor variables 2 are designed primarily to preserve positive definiteness and to improve robustness in strongly stretched flows. For FENE-type models this is only the first admissibility condition. A positive tensor with tr⁡C≥L2\operatorname{tr} C \ge L^2 is outside the model domain, and a positive tensor with tr⁡C<L2\operatorname{tr} C < L^2 may still be unacceptable if a reconstruction step injects a mesh-scale burst of FENE entropy. Table 2 summarizes the distinction.

ReconstructionPositivityTrace barrierEntropy compatibility
Standard logbuilt innot built innot automatic for high-order states
Square root or factorbuilt innot built innot automatic for high-order states
Trace clippingenforceableenforceablegenerally not variationally consistent
Barrier logbuilt inbuilt inrequires entropy acceptance step
Barrier log with (5)built inbuilt inenforced by a least-damping FENE budget

Table 2. Comparison of reconstruction variables for FENE-type conformation tensors.

Richter, Iaccarino, and Shaqfeh 3 used a different and important finite-extensibility safeguard in DNS of FENE-P flow past a cylinder: the trace equation is advanced first and the positive root of a scalar algebraic relation is selected so that the FENE denominator remains positive. That idea is a trace-first admissibility update. The present construction has a different purpose. It treats the whole tensor as the constrained state variable, enforces C≻0C \succ0 and tr⁡C<L2\operatorname{tr} C < L^2 simultaneously by a barrier-log map, and then accepts high-order reconstructed states only through a FENE entropy budget. Thus the guarantee is not only denominator positivity in a component update, but compatibility with the fully discrete free-energy inequality.

Thus the proposed reconstruction has two layers. The barrier-log map enforces the geometry of the FENE state space, while the entropy acceptance rule enforces compatibility with the free-energy estimate. Both layers are needed for the fully discrete theorem below.

FENE-P model with polymer molecular diffusion

Let b=L2>db = L^2 > d and define

fb(C)=b−db−tr⁡C,Db={C∈S++d:tr⁡C<b}.f_b(C) = \frac{b-d}{b-\operatorname{tr} C}, \qquad\mathcal{D}_b = \{C \in\mathbb{S}_{++}^{d}: \operatorname{tr} C < b\}.

The molecularly diffusive FENE-P type system is written as

Re⁡(∂tu+u⋅∇u)+∇p−βΔu=(1−β)∇⋅S+f,(1)\operatorname{Re}(\partial_t u + u \cdot\nabla u) + \nabla p - \beta\Delta u = (1-\beta)\nabla\cdot S + f, \tag*{(1)}
∇⋅u=0,(2)\nabla\cdot u = 0, \tag*{(2)}
∂tC+u⋅∇C−(∇u)C−C(∇u)T=−1Wi⁡T(C)+εΔC,(3)\partial_t C + u \cdot\nabla C - (\nabla u)C - C(\nabla u)^{T} = -\frac{1}{\operatorname{Wi}}T(C) + \varepsilon\Delta C, \tag*{(3)}
T(C)=fb(C)C−I,S=T(C)Wi⁡.(4)T(C) = f_b(C)C - I, \qquad S = \frac{T(C)}{\operatorname{Wi}}. \tag*{(4)}

The normalization fb(I)=1f_b(I) = 1 makes C=IC = I the equilibrium conformation. The parameter ε=Pe⁡−1\varepsilon= \operatorname{Pe}^{-1} represents polymer center-of-mass molecular diffusion in the constant-polymer-density FENE-P closure. Periodic boundary conditions or the no-flux condition

∇C n=0\nabla C\,\mathbf n = 0

are assumed for this diffusive flux. If the polymer number density is not constant, the closure must be augmented by a concentration equation; that extension is outside the present analysis. The point here is that molecular diffusion must be paired with the FENE barrier entropy rather than treated as an arbitrary componentwise smoothing operator.

The elastic entropy is

Φb(C)=−log⁡det⁡C−(b−d)log⁡(b−tr⁡Cb−d).\Phi_b(C) = -\log\det C - (b-d)\log\left(\frac{b-\operatorname{tr} C}{b-d}\right).

It satisfies Φb(I)=0\Phi_b(I)=0 and

DΦb(C)[B]=(fb(C)I−C−1):B=Hb(C):B.D\Phi_b(C)[B] = \left(f_b(C)I-C^{-1}\right):B = H_b(C):B.

The Hessian is positive:

D2Φb(C)[B,B]=tr⁡(C−1BC−1B)+b−d(b−tr⁡C)2(tr⁡B)2.D^2\Phi_b(C)[B,B] = \operatorname{tr}\left(C^{-1}BC^{-1}B\right) + \frac{b-d}{(b-\operatorname{tr} C)^2}\left(\operatorname{tr} B\right)^2.

Thus Φb\Phi_b is convex on Db\mathcal{D}_b and blows up at both parts of the boundary: λmin⁡(C)↓0\lambda_{\min}(C)\downarrow0 and tr⁡C↑b\operatorname{tr} C\uparrow b.

Lemma 2.1 (trace-barrier buffer). Let C∈DbC\in\mathcal{D}_b and Φb(C)≤M\Phi_b(C)\le M. Then there is a positive function δb(M)\delta_b(M) such that

b−tr⁡C≥δb(M)>0.b-\operatorname{tr} C\ge\delta_b(M)>0.

One possible choice is

δb(M)=(b−d)exp⁡[−M+dlog⁡(b/d)b−d].\delta_b(M)=(b-d)\exp\left[-\frac{M+d\log(b/d)}{b-d}\right].

Proof. Since det⁡C≤(tr⁡C/d)d≤(b/d)d\det C\le(\operatorname{tr} C/d)^d\le(b/d)^d, one has −log⁡det⁡C≥−dlog⁡(b/d)-\log\det C\ge-d\log(b/d). From the entropy bound,

−(b−d)log⁡(b−tr⁡Cb−d)≤M+dlog⁡(b/d),-(b-d)\log\left(\frac{b-\operatorname{tr} C}{b-d}\right)\le M+d\log(b/d),

which gives the stated lower bound on b−tr⁡Cb-\operatorname{tr} C. □

Lemma 2.2 (relaxation dissipation). For every C∈DbC\in\mathcal{D}_b,

T(C):Hb(C)=∑i=1dλi(fb(C)−λi−1)2≥0.T(C):H_b(C)=\sum_{i=1}^{d}\lambda_i\left(f_b(C)-\lambda_i^{-1}\right)^2\ge0.

where λi\lambda_i are the eigenvalues of $C.

Proof. Because fb(C)f_b(C) is a scalar function of tr⁡C\operatorname{tr} C, it commutes with CC. Diagonalizing CC gives

(fbλi−1)(fb−λi−1)=λi(fb−λi−1)2.(5)\left(f_b\lambda_i-1\right)\left(f_b-\lambda_i^{-1}\right)=\lambda_i\left(f_b-\lambda_i^{-1}\right)^2. \tag*{(5)}

Summing over ii proves the claim. □

Lemma 2.3 (polymer molecular-diffusion entropy dissipation). Assume periodic boundary conditions or ∇Cn=0\nabla C n = 0 on ∂Ω\partial\Omega. For every smooth admissible conformation field,

∫ΩΔC:Hb(C) dx=−∫ΩIb(C) dx,\int_{\Omega} \Delta C : H_b(C)\,\mathrm{d}x = -\int_{\Omega} I_b(C)\,\mathrm{d}x,

where

Ib(C)=∑j=1d[tr⁡(C−1∂jCC−1∂jC)+b−d(b−tr⁡C)2(∂jtr⁡C)2]≥0.I_b(C) = \sum_{j=1}^{d} \left[ \operatorname{tr}\left(C^{-1}\partial_j C C^{-1}\partial_j C\right) + \frac{b-d}{(b-\operatorname{tr} C)^2}\left(\partial_j\operatorname{tr} C\right)^2 \right] \ge0.

Proof. Integrating by parts gives

∫ΩΔC:Hb(C) dx=−∑j=1d∫Ω∂jC:DHb(C)[∂jC] dx.\int_{\Omega} \Delta C : H_b(C)\,\mathrm{d}x = -\sum_{j=1}^{d}\int_{\Omega} \partial_j C : DH_b(C)[\partial_j C]\,\mathrm{d}x.

Since DHb=D2ΦbDH_b = D^2\Phi_b, the Hessian formula (2) gives the stated expression. □

Trace-singular FENE-family closures. Although the formulas below are written for the FENE-P coefficient fbf_b, the same algebra applies to a larger FENE-family class. Let

Tg(C)=g(tr⁡C)C−I,Hg(C)=g(tr⁡C)I−C−1,T_g(C) = g(\operatorname{tr} C)C - I,\qquad H_g(C) = g(\operatorname{tr} C)I - C^{-1},

where g∈C1([d,b))g \in C^1([d,b)), g(d)=1g(d)=1, g′(s)≥0g'(s)\ge0, and g(s)→∞g(s)\to\infty as s↑bs \uparrow b. Define

Gg(s)=∫dsg(r) dr,Φg(C)=−log⁡det⁡C+Gg(tr⁡C).G_g(s) = \int_d^s g(r)\,\mathrm{d}r,\qquad\Phi_g(C) = -\log\det C + G_g(\operatorname{tr} C).

The FENE-P entropy corresponds to g=fbg=f_b and Gg(s)=−(b−d)log⁡((b−s)/(b−d))G_g(s)=-(b-d)\log((b-s)/(b-d)).

Proposition 2.4 (common entropy mechanism for FENE-family closures). For every C∈DbC\in D_b and every symmetric BB,

DΦg(C)[B]=Hg(C):B,D\Phi_g(C)[B] = H_g(C):B,
D2Φg(C)[B,B]=tr⁡(C−1BC−1B)+g′(tr⁡C)(tr⁡B)2.D^2\Phi_g(C)[B,B] = \operatorname{tr}(C^{-1}BC^{-1}B) + g'(\operatorname{tr} C)(\operatorname{tr} B)^2.

Moreover,

Tg(C):Hg(C)=∑i=1dλi(g(tr⁡C)−λi−1)2≥0,T_g(C):H_g(C) = \sum_{i=1}^{d}\lambda_i\left(g(\operatorname{tr} C)-\lambda_i^{-1}\right)^2 \ge0,

and, for ∇⋅u=0,\nabla\cdot u=0,

−((∇u)C+C(∇u)T):Hg(C)=−2Tg(C):∇u.-\left((\nabla u)C+C(\nabla u)^{\mathsf{T}}\right):H_g(C) = -2T_g(C):\nabla u.

Proof. The derivative and Hessian follow from the chain rule and the derivative of −log⁡det⁡C-\log\det C. Since g(tr⁡C)g(\operatorname{tr} C) is scalar, TgT_g and HgH_g are diagonal in the eigenbasis of CC, giving the relaxation identity. The stretching identity follows from C:∇u=C:D(u)C:\nabla u=C:D(u) and I:∇u=∇⋅u=0I:\nabla u=\nabla\cdot u=0. □\square

This proposition is the structural reason for calling the method FENE-type rather than only FENE-P. Once a closure has the trace-singular entropy variable HgH_g, the barrier-log state space, entropy-compatible reconstruction, and discrete coupling argument below carry over with fbf_b replaced by g(tr⁡C)g(\operatorname{tr}C). The numerical section uses FENE-P because it is the standard Peterlin closure and makes the finite-extensibility denominator explicit.

Continuous energy law

The FENE trace barrier is compatible with the usual polymeric-work cancellation. Testing (1) by uu gives the polymeric work term

−(1−β)∫ΩS:∇u dx=−1−βWi∫ΩT(C):∇u dx.-(1-\beta)\int_{\Omega} S:\nabla u\,\mathrm{d}x=-\frac{1-\beta}{\mathrm{Wi}}\int_{\Omega}T(C):\nabla u\,\mathrm{d}x.

Testing (3) by Hb(C)H_b(C) yields the stretching contribution

−((∇u)C+C(∇u)T):Hb(C).-\left((\nabla u)C+C(\nabla u)^{\mathsf{T}}\right):H_b(C).

Since Hb(C)=fbI−C−1H_b(C)=f_bI-C^{-1} and ∇⋅u=0\nabla\cdot u=0,

−((∇u)C+C(∇u)T):(fbI−C−1)=−2fbC:∇u+2I:∇u=−2T(C):∇u.-\left((\nabla u)C+C(\nabla u)^{\mathsf{T}}\right):(f_bI-C^{-1}) = -2f_bC:\nabla u+2I:\nabla u=-2T(C):\nabla u.

Multiplication by (1−β)/(2Wi)(1-\beta)/(2\mathrm{Wi}) therefore cancels the polymeric work exactly.

Theorem 3.1 (continuous barrier energy identity). Assume a smooth solution with C(t,x)∈DbC(t,x)\in D_b and periodic or entropy-compatible boundary conditions. Then

ddt[Re2∥u∥L22+1−β2Wi∫ΩΦb(C) dx]+β∥∇u∥L22+1−β2Wi2∫ΩT(C):Hb(C) dx+(1−β)ε2Wi∫ΩIb(C) dx=(f,u).(6)\frac{\mathrm{d}}{\mathrm{d}t}\left[\frac{\mathrm{Re}}{2}\lVert u\rVert_{L^2}^{2}+\frac{1-\beta}{2\mathrm{Wi}}\int_{\Omega}\Phi_b(C)\,\mathrm{d}x\right]+\beta\lVert\nabla u\rVert_{L^2}^{2} +\frac{1-\beta}{2\mathrm{Wi}^{2}}\int_{\Omega}T(C):H_b(C)\,\mathrm{d}x +\frac{(1-\beta)\varepsilon}{2\mathrm{Wi}}\int_{\Omega}I_b(C)\,\mathrm{d}x=(f,u). \tag*{(6)}

where IbI_b is the molecular-diffusion entropy density in Lemma 2.3.

Proof. The kinetic-energy identity follows from the incompressibility cancellation of the convection term. The entropy derivative follows from DΦb=HbD\Phi_b=H_b. The stretching contribution cancels the polymeric work as shown above. Relaxation gives the nonnegative term in Lemma 2.2, and polymer molecular diffusion gives the nonnegative barrier dissipation in Lemma 2.3. Combining the terms gives (3.1). □\square

The trace-barrier term in IbI_b is the new contribution relative to Oldroyd–B. It penalizes gradients of tr⁡C\operatorname{tr}C more strongly near the finite-extensibility boundary. This term is also the reason that a discrete diffusion operator should be paired with the same barrier entropy rather than treated as a componentwise Laplacian detached from the entropy calculation.

Barrier logarithmic parametrization

A standard log-conformation parametrization C=Exp⁡ΨC = \operatorname{Exp}\Psi enforces C≻0C \succ0 but does not enforce tr⁡C<b\operatorname{tr} C < b. We use instead the map

Bb(Ψ)=bExp⁡Ψ1+tr⁡(Exp⁡Ψ),Ψ=ΨT.\mathcal{B}_b(\Psi) = \frac{b\operatorname{Exp}\Psi}{1 + \operatorname{tr}(\operatorname{Exp}\Psi)}, \qquad\Psi= \Psi^{T}.

Then Bb(Ψ)∈Db\mathcal{B}_b(\Psi) \in\mathcal{D}_b for every symmetric Ψ\Psi, because

tr⁡Bb(Ψ)=btr⁡(Exp⁡Ψ)1+tr⁡(Exp⁡Ψ)<b.\operatorname{tr}\mathcal{B}_b(\Psi) = \frac{b\operatorname{tr}(\operatorname{Exp}\Psi)}{1 + \operatorname{tr}(\operatorname{Exp}\Psi)} < b.

The map is onto Db\mathcal{D}_b. If C∈DbC \in\mathcal{D}_b, set

Exp⁡Ψ=Cb−tr⁡C.\operatorname{Exp}\Psi= \frac{C}{b - \operatorname{tr} C}.

Then

Bb(Ψ)=bC/(b−tr⁡C)1+tr⁡C/(b−tr⁡C)=C.\mathcal{B}_b(\Psi) = \frac{bC/(b - \operatorname{tr} C)}{1 + \operatorname{tr} C/(b - \operatorname{tr} C)} = C.

Thus (4) is a natural finite-extensibility analogue of the log-conformation map.

Proposition 4.1 (Constraint enforcement by parametrization). If the nonlinear algebraic solve is performed in a symmetric variable Ψh\Psi_h and the accepted conformation tensor is defined by Ch=Bb(Ψh)C_h = \mathcal{B}_b(\Psi_h) at entropy quadrature points, then

Ch(xq)≻0,tr⁡Ch(xq)<bC_h(x_q) \succ0, \qquad\operatorname{tr} C_h(x_q) < b

at every quadrature point xqx_q, independent of the Newton iterate size before acceptance.

Proof. The matrix exponential is positive definite, and the scalar denominator in (4) is strictly positive. The trace calculation above gives the strict upper bound. □

This parametrization is not required for the energy proof; one may also use a line-search Newton method that rejects candidates outside Db\mathcal{D}_b. Its advantage is practical: the finite-extensibility constraint is built into the unknown. The proof below is written in the accepted conformation tensor, so it applies to either realization.

Entropy-compatible barrier-log reconstruction

The barrier-log map solves the geometric part of admissibility, but it does not by itself guarantee compatibility with the discrete FENE free energy. A high-order reconstruction can remain inside DbD_b and still add artificial elastic entropy. We therefore add an entropy-compatible reconstruction layer.

Let Ψ~q\widetilde{\Psi}_q be a physical predictor at quadrature point xqx_q, and let Ψ~q\widetilde{\Psi}_q be a raw high-order barrier-log reconstruction. Define

Ψq(θ)=Ψ~q+θ(Ψ~q−Ψ~q),Cq(θ)=Bb(Ψq(θ)),0≤θ≤1.\Psi_q(\theta) = \widetilde{\Psi}_q + \theta(\widetilde{\Psi}_q - \widetilde{\Psi}_q), \qquad C_q(\theta) = B_b(\Psi_q(\theta)), \qquad0 \le\theta\le1.

For a nonnegative budget τh\tau_h, the accepted parameter is

θ⋆=max⁡{θ∈[0,1]:∑qwqΦb(Cq(θ))≤∑qwqΦb(Cq(0))+τh}.\theta_\star= \max\left\{ \theta\in[0,1] : \sum_q w_q\Phi_b(C_q(\theta)) \le\sum_q w_q\Phi_b(C_q(0)) + \tau_h \right\}.

The accepted tensor is Cq⋆=Cq(θ⋆)C_q^\star= C_q(\theta_\star). If the raw reconstruction is already entropy-compatible, then θ⋆=1\theta_\star= 1. Otherwise the method damps only the entropy-incompatible part along the barrier-log segment.

Lemma 5.1 (convex FENE entropy profile on barrier-log paths). Let C(θ)=Bb(Ψ0+θE)C(\theta) = B_b(\Psi_0 + \theta E). Then J(θ)=Φb(C(θ))J(\theta) = \Phi_b(C(\theta)) is convex on [0,1][0,1]. More explicitly,

Φb(Bb(Ψ))=blog⁡(1+tr⁡(eΨ))−tr⁡Ψ+γb,\Phi_b(B_b(\Psi)) = b\log(1+\operatorname{tr}(e^\Psi))-\operatorname{tr}\Psi+\gamma_b,

where γb\gamma_b is independent of Ψ\Psi.

Proof. Set Z(Ψ)=1+tr⁡(eΨ)Z(\Psi) = 1+\operatorname{tr}(e^\Psi). Since Bb(Ψ)=beΨ/ZB_b(\Psi) = b e^\Psi/Z, one has

det⁡Bb(Ψ)=bdetr⁡ΨZ−d,b−tr⁡Bb(Ψ)=b/Z.\det B_b(\Psi) = b^d e^{\operatorname{tr}\Psi}Z^{-d}, \qquad b-\operatorname{tr}B_b(\Psi) = b/Z.

Substitution into (2) gives (5.1). The map Ψ↦log⁡(1+tr⁡eΨ)\Psi\mapsto\log(1+\operatorname{tr}e^\Psi) is the spectral log-sum-exp function with an additional zero mode and is convex on symmetric matrices. The term −tr⁡Ψ-\operatorname{tr}\Psi is affine. Hence JJ is convex along every affine path. □

Proposition 5.2 (least-damping entropy-compatible reconstruction). The set in (5) is a closed interval containing 0. Hence θ⋆\theta_\star exists, is the largest admissible parameter, and can be computed by bisection. The accepted tensor satisfies

Cq⋆∈Db,∑qwqΦb(Cq⋆)≤∑qwqΦb(Cq(0))+τh.C_q^\star\in D_b, \qquad\sum_q w_q\Phi_b(C_q^\star) \le\sum_q w_q\Phi_b(C_q(0))+\tau_h.

Proof. Admissibility in DbD_b follows from Proposition 4.1 for every θ∈[0,1]\theta\in[0,1]. By Lemma 5.1, the entropy profile is continuous and convex; therefore its sublevel set is a closed interval. Since θ=0\theta=0 satisfies the inequality, the interval is nonempty and has a largest element. Bisection applies because admissibility is monotone along this interval. □

Proposition 5.3 (asymptotic inactivity). Assume the raw barrier-log reconstruction defect Eq=Ψ~qe−Ψ~qbE_q=\widetilde{\Psi}^{e}_q-\widetilde{\Psi}^{b}_q is O(hk+1)O(h^{k+1}) and satisfies

∑qwqD{Φb∘Bb}(Ψ~qb)[Eq]=O(h2k+2).\sum_q w_q D\{\Phi_b\circ\mathcal{B}_b\}(\widetilde{\Psi}^{b}_q)[E_q]=O(h^{2k+2}).

If τh=cτh2k+2\tau_h=c_\tau h^{2k+2} with cτc_\tau large enough for the leading consistency constant, then θ⋆=1\theta_\star=1 for sufficiently small hh. In the marginal active case, 1−θ⋆=O(hk+1)1-\theta_\star=O(h^{k+1}) under a nondegenerate endpoint derivative.

Proof. Taylor expansion of the convex entropy profile gives

∑qwq{Φb(Cq(1))−Φb(Cq(0))}=O(h2k+2).\sum_q w_q\{\Phi_b(C_q(1))-\Phi_b(C_q(0))\}=O(h^{2k+2}).

under the stated first-variation condition and compactness of the path. The mesh-scaled budget therefore accepts the raw endpoint for sufficiently small hh. If the endpoint is marginally active, convexity and the nondegenerate derivative give the stated bound on 1−θ⋆1-\theta_\star. □

When C⋆C^\star is used consistently in the stress force, stretching term, relaxation term, and entropy quadrature, the fully discrete energy estimate below is unchanged except for the explicit budget contribution (1−β)τh/(2Wi)(1-\beta)\tau_h/(2\mathrm{Wi}). Thus the reconstruction is not merely trace-preserving; it is compatible with the FENE free-energy balance.

Fully discrete scheme

Let Vh×QhV_h\times Q_h be an inf-sup stable velocity–pressure pair and MhM_h a symmetric tensor space. Let (⋅,⋅)Q(\cdot,\cdot)_Q denote a positive quadrature rule used consistently in the entropy terms and the polymeric work. At time level n+1n+1, find

(uhn+1,phn+1,Chn+1)∈Vh×Qh×Mh(u_h^{n+1},p_h^{n+1},C_h^{n+1})\in V_h\times Q_h\times M_h

such that Chn+1(xq)∈DbC_h^{n+1}(x_q)\in D_b and

Thn+1=fb(Chn+1)Chn+1−I,Shn+1=Thn+1Wi,T_h^{n+1}=f_b(C_h^{n+1})C_h^{n+1}-I,\qquad S_h^{n+1}=\frac{T_h^{n+1}}{\mathrm{Wi}},

at quadrature points. The momentum step is

Re⁡(uhn+1−uhnΔt,vh)+Re⁡ ch(uhn+1;uhn+1,vh)+β(∇uhn+1,∇vh)−(phn+1,∇⋅vh)−(1−β)(Shn+1,∇vh)Q=(fn+1,vh),\operatorname{Re}\left(\frac{u_h^{n+1}-u_h^n}{\Delta t},v_h\right)+\operatorname{Re}\,c_h(u_h^{n+1};u_h^{n+1},v_h)+\beta(\nabla u_h^{n+1},\nabla v_h) -(p_h^{n+1},\nabla\cdot v_h)-(1-\beta)(S_h^{n+1},\nabla v_h)_Q=(f^{n+1},v_h),

with

(qh,∇⋅uhn+1)=0.(q_h,\nabla\cdot u_h^{n+1})=0.

The conformation update is

(Chn+1−ChnΔt,Bh)Q+ah(uhn+1;Chn+1,Bh)−((∇uhn+1)Chn+1+Chn+1(∇uhn+1)T,Bh)Q+1Wi(Thn+1,Bh)Q+εdh(Chn+1,Bh)=0.\left(\frac{C_h^{n+1}-C_h^n}{\Delta t},B_h\right)_Q+a_h(u_h^{n+1};C_h^{n+1},B_h)-\left((\nabla u_h^{n+1})C_h^{n+1}+C_h^{n+1}(\nabla u_h^{n+1})^{T},B_h\right)_Q+\frac{1}{\mathrm{Wi}}(T_h^{n+1},B_h)_Q+\varepsilon d_h(C_h^{n+1},B_h)=0.

The forms are chosen so that the velocity convection is skew-symmetric, conformation transport has zero entropy contribution for discretely incompressible velocity, and the stretching term is evaluated at the same quadrature points as the stress force. The form dhd_h is the discrete polymer molecular-diffusion operator and is paired with the barrier entropy variable:

dh(Ch,Hb(Ch))=Ib,h(Ch)≥0d_h(C_h,H_b(C_h))=\mathcal{I}_{b,h}(C_h)\ge0

for conforming periodic or no-flux discretizations, or at least dh(Ch,Hb(Ch))≥Ib,h(Ch)d_h(C_h,H_b(C_h))\ge\mathcal{I}_{b,h}(C_h) for stabilized variants. This is the discrete counterpart of Lemma 2.3. If a high-order reconstruction is used before the conformation state enters the coupled step, Chn+1C_h^{n+1} denotes the entropy-compatible accepted tensor from Section 5.

Define the discrete energy

Ehn=Re2∥uhn∥L22+1−β2Wi∑qwqΦb(Chn(xq)).\mathcal{E}_h^n=\frac{\mathrm{Re}}{2}\lVert u_h^n\rVert_{L^2}^2+\frac{1-\beta}{2\mathrm{Wi}}\sum_q w_q\Phi_b(C_h^n(x_q)).

Theorem 6.1 (discrete barrier energy inequality). Assume the accepted conformation tensor satisfies Chn+1(xq)∈DbC_h^{n+1}(x_q)\in D_b at all entropy quadrature points and the discrete transport and stretching forms have the compatibility properties stated above. Then

Ehn+1−Ehn+Re2∥uhn+1−uhn∥L22+Δt β∥∇uhn+1∥L22+Δt 1−β2Wi2∑qwqThn+1(xq):Hb(Chn+1(xq))+Δt (1−β)ε2WiIb,h(Chn+1)≤Δt (fn+1,uhn+1)+1−β2Wiτh.\mathcal{E}_h^{n+1}-\mathcal{E}_h^n+\frac{\mathrm{Re}}{2}\lVert u_h^{n+1}-u_h^n\rVert_{L^2}^2+\Delta t\,\beta\lVert\nabla u_h^{n+1}\rVert_{L^2}^2 +\Delta t\,\frac{1-\beta}{2\mathrm{Wi}^2}\sum_q w_q T_h^{n+1}(x_q):H_b(C_h^{n+1}(x_q)) +\Delta t\,\frac{(1-\beta)\varepsilon}{2\mathrm{Wi}}\mathcal{I}_{b,h}(C_h^{n+1}) \le\Delta t\,(f^{n+1},u_h^{n+1})+\frac{1-\beta}{2\mathrm{Wi}}\tau_h.

In particular, if the right-hand side is controlled by Young’s inequality, the discrete entropy remains bounded, and Lemma 2.1 gives a positive trace buffer at quadrature points.

Proof. Set vh=uhn+1v_h=u_h^{n+1} in (6). The skew-symmetric convection and pressure terms vanish, and the standard identity

(a−b,a)=12(∥a∥2−∥b∥2+∥a−b∥2)(a-b,a)=\frac{1}{2}\left(\lVert a\rVert^2-\lVert b\rVert^2+\lVert a-b\rVert^2\right)

gives the kinetic increment. In (6) use the entropy test

Bh=Hb(Chn+1)=fb(Chn+1)I−(Chn+1)−1B_h = H_b(C_h^{n+1}) = f_b(C_h^{n+1})I - (C_h^{n+1})^{-1}

at quadrature points. Convexity of Φb\Phi_b gives

Hb(Chn+1):(Chn+1−Chn)≥Φb(Chn+1)−Φb(Chn).H_b(C_h^{n+1}) : (C_h^{n+1} - C_h^n) \ge\Phi_b(C_h^{n+1}) - \Phi_b(C_h^n).

If the conformation tensor is obtained by the entropy-compatible reconstruction of Section 5, the accepted endpoint contributes at most the budget τh\tau_h to the quadrature entropy; otherwise τh=0\tau_h = 0. The stretching term cancels the polymeric work in the momentum equation by the same pointwise identity used in the continuous proof. Relaxation is nonnegative by Lemma 2.2, and diffusion contributes the discrete Hessian dissipation Ib,hI_{b,h}. Summing the momentum and entropy relations gives (6.1). □

AP stress closure and Newtonian limit

The finite-extensibility stress is nonlinear in CC, but the AP variable is simple:

S=fb(C)C−IWi.S = \frac{f_b(C)C-I}{\mathrm{Wi}}.

The conformation equation can be rewritten as a relaxation equation for SS. At fixed hh and Δt\Delta t, insert Thn+1=WiShn+1T_h^{n+1}=\mathrm{Wi}S_h^{n+1} into (6). The relaxation term becomes (Shn+1,Bh)Q(S_h^{n+1},B_h)_Q, while the leading stretching term gives the rate-of-strain tensor. All time-difference, transport, diffusion, and nonlinear stretching remainders are multiplied by Wi after the stress substitution.

Proposition 7.1 (AP closure). Assume the discrete states remain uniformly bounded in the norms entering the conformation residual and stay in a compact subset of DbD_b. Then, for fixed hh and Δt\Delta t,

∥Shn+1−D(uhn+1)∥Mh′≤CWi,\left\|S_h^{n+1}-D(u_h^{n+1})\right\|_{M_h'} \le C\mathrm{Wi},

where CC is independent of Wi in the tested small-Weissenberg range.

Proof. After substituting Thn+1=WiShn+1T_h^{n+1}=\mathrm{Wi}S_h^{n+1} in (6), the leading balance is

(Shn+1,Bh)Q=(D(uhn+1),Bh)Q+Wi Rhn+1(Bh),(S_h^{n+1},B_h)_Q=(D(u_h^{n+1}),B_h)_Q+\mathrm{Wi}\,\mathcal{R}_h^{n+1}(B_h),

with the normalization of DD matching the stretching convention. The residual Rhn+1\mathcal{R}_h^{n+1} contains the time difference, transport, diffusion, and higher-order stretching terms. The assumed uniform bound gives

∣Rhn+1(Bh)∣≤C∥Bh∥Mh,\left|\mathcal{R}_h^{n+1}(B_h)\right| \le C\lVert B_h\rVert_{M_h},

which proves (7.1). □

Theorem 7.2 (fixed-discretization Newtonian limit). *Let Wij→0\mathrm{Wi}_j\to0 and assume the corresponding admissible discrete solutions are uniformly bounded and remain in a compact subset of DbD_b. Then any convergent subsequence has a limit satisfying the discrete incompressible Navier–Stokes scheme with total viscosity equal to the solvent viscosity plus the polymeric contribution from the closure Sh=D(uh)S_h = \mathbf{D}(u_h).

Proof. The finite-dimensional compactness gives a convergent subsequence. The AP closure identifies the stress limit as D(uh)\mathbf{D}(u_h). Passing to the limit in the momentum equation gives the discrete Newtonian system with the polymeric stress absorbed into the viscous operator. □\square

Relative-entropy error estimate

For convergence rates one needs a compact spectral and trace set

Kλ,δ={C∈Db:λI⪯C, b−tr⁡C≥δ},λ,δ>0.K_{\lambda,\delta} = \{C \in\mathcal{D}_b : \lambda I \preceq C,\ b - \operatorname{tr} C \ge\delta\}, \qquad\lambda,\delta> 0.

On this set the entropy Hessian is bounded above and below, so the relative entropy

Φb(C∣C^)=Φb(C)−Φb(C^)−Hb(C^):(C−C^)\Phi_b(C|\widehat{C}) = \Phi_b(C) - \Phi_b(\widehat{C}) - H_b(\widehat{C}) : (C - \widehat{C})

is equivalent to ∥C−C^∥F2\lVert C - \widehat{C}\rVert_{F}^{2}. The constants deteriorate as λ↓0\lambda\downarrow0 or δ↓0\delta\downarrow0, which is unavoidable because the entropy becomes singular at both boundaries.

Theorem 8.1 (conditional error estimate). Let (u,p,C)(u,p,C) be a smooth solution on [0,T][0,T] with C(t,x)∈Kλ,δC(t,x) \in K_{\lambda,\delta}. Assume stable projections, a quadrature-consistent discretization of the coupling terms, and nonlinear solver residuals of order hk+1+Δth^{k+1} + \Delta t. If the discrete solution remains in a slightly larger compact subset of Db\mathcal{D}_b, then

max⁡0≤n≤N[∥u(tn)−uhn∥L22+∑qwqΦb(C(tn,xq)|Chn(xq))]≤CT(h2k+2+Δt2).\max_{0 \le n \le N} \left[ \lVert u(t_n) - u_h^n\rVert_{L^2}^{2} + \sum_q w_q \Phi_b\left(C(t_n,x_q)\middle|C_h^n(x_q)\right) \right] \le C_T\left(h^{2k+2} + \Delta t^2\right).

Proof. The proof follows the relative-entropy stability argument. Subtract the projected exact equations from the discrete equations, test the velocity error by itself, and test the conformation error in the entropy variable. The compatible quadrature cancels the leading polymeric-work error. The Hessian bounds on Kλ,δK_{\lambda,\delta} control the nonlinear FENE coefficients and convert relative entropy into squared tensor error. Consistency and solver residuals contribute C(h2k+2+Δt2)C(h^{2k+2} + \Delta t^2), while lower-order terms are bounded by the current error. A discrete Gronwall inequality gives the estimate. □\square

The theorem is intentionally conditional: it gives a classical rate in a resolved regime away from the finite-extensibility boundary. The energy theorem and trace-barrier buffer are the mechanisms that help keep the computation in such a regime.

Numerical verification

The numerical diagnostics are designed to test the mechanisms used in the analysis: trace-barrier preservation, entropy-compatible reconstruction, energy decay, AP closure, coupled velocity–pressure–stress feedback, and high-Weissenberg behavior near the finite-extensibility boundary. The reported benchmarks are kept compact and tied directly to the structural claims.

TestMechanismExpected outcome
Barrier map testBb(Ψ)∈Db\mathcal{B}_b(\Psi) \in\mathcal{D}_bstrict tr⁡C/L2<1\operatorname{tr} C/L^2 < 1
Entropy-compatible reconstructionFENE entropy budget along barrier-log pathraw entropy excess removed with largest θ\theta
Energy decaybarrier entropy and relaxationmonotone free energy without forcing
AP closurescaled stress S=T/WiS = T/\mathrm{Wi}error proportional to Wi\mathrm{Wi}
Coupled manufactured solvepressure, velocity, FENE stress feedbacksecond-order trend
High-Wi stretch sweepweak relaxation near trace barrierpositive trace buffer

Table 3. Main numerical diagnostics and the structural claim each test addresses.

Barrier preservation, entropy correction, and energy decay.

The first diagnostic compares a plain log-conformation map with the barrier-log map (4). The background dimension is d=2d = 2 and L2=12L^2 = 12. The raw log map preserves positive definiteness but can produce tr⁡C>L2\operatorname{tr} C > L^2 for large logarithmic stretches. The barrier-log map keeps the same unconstrained variable while mapping it into Db\mathcal{D}_b.

amplitudeλmin⁡(exp⁡Ψ)\lambda_{\min}(\exp\Psi)tr⁡(exp⁡Ψ)/L2\operatorname{tr}(\exp\Psi)/L^2tr⁡(Bb(Ψ))/L2\operatorname{tr}(\mathcal{B}_b(\Psi))/L^2barrier status
0.56.065E-011.013E-015.487E-01admissible
1.52.231E-012.053E-017.114E-01admissible
2.58.208E-025.205E-018.620E-01admissible
3.53.020E-021.381E+009.430E-01admissible
4.51.111E-023.725E+009.781E-01admissible

Table 4. Parametrization diagnostic for L2=12L^2 = 12. The barrier-log map enforces the finite-extensibility trace constraint for all tested logarithmic amplitudes.

The next diagnostic isolates the reconstruction layer and compares it with trace-only denominator repair. We start from a near-barrier tensor with tr⁡C/b=0.88\operatorname{tr} C/b = 0.88 and apply the same symmetric high-order perturbation in different variables. The standard log state can cross the trace barrier immediately. A trace repair rescales the state back below bb, which mimics the role of trace-first safeguards in denominator control, but it leaves a very large FENE entropy jump because the repaired state is placed extremely close to the singular barrier. The barrier-log endpoint remains admissible by construction, yet it may still add substantial FENE entropy. The entropy-compatible barrier-log reconstruction selects the largest admissible θ\theta satisfying eq:5, so the accepted state controls both the trace barrier and the entropy increment.

aalog tr⁡C/b\operatorname{tr} C/brepair ΔΦ\Delta\Phibarrier-log tr⁡C/b\operatorname{tr} C/bbarrier-log ΔΦ\Delta\Phientropy-log ΔΦ\Delta\Phiθ⋆\theta_\star
0.251.060210.4380.8983.1170.1000.035
0.501.302210.6570.9166.6580.1000.017
0.751.625210.9080.93110.5630.1000.012
1.002.055211.1850.94514.7670.1008.64E-3
1.252.622211.4800.95619.2120.1006.91E-3
1.503.371211.7900.96623.8470.1005.76E-3

Table 5. Controlled near-barrier reconstruction comparison for b=L2=20b = L^2 = 20. The trace repair keeps the denominator positive but does not control the FENE entropy increment; the entropy-compatible barrier-log reconstruction enforces a prescribed entropy budget.

Wi\mathrm{Wi}methodmax tr⁡C/b\operatorname{tr} C/bmin b−tr⁡Cb - \operatorname{tr} Cmax ΔΦ\Delta\Phifailed step
100standard log0.9959.94E-279.5626
100sqrt factor1.0008.98E-4164.0968
100trace repair1.0002.00E-5237.57–
100barrier log0.9911.87E-173.05–
100entropy log0.8193.62E+016.00–
400standard log0.9519.90E-138.229
400sqrt factor0.9968.35E-282.5024
400trace repair1.0002.00E-5249.04–
400barrier log1.0002.63E-5244.12–
400entropy log0.8193.62E+016.00–

Table 6. Iterated high-Weissenberg stretch/reconstruction stress test. A failed step means the method crossed the FENE barrier during the iteration. The trace-repair baseline remains trace-admissible only by repeatedly placing the tensor within 2×10−52 \times10^{-5} of the barrier, while the entropy-compatible reconstruction preserves a macroscopic buffer.

Next we evolve a homogeneous relaxation-diffusion diagnostic with no forcing. Figure 1 shows the normalized free energy and the maximum trace ratio. The free energy decreases, while the trace remains below the finite-extensibility threshold.

Barrier relaxation diagnostic showing normalized free energy and maximum trace ratio

Figure 1. Barrier relaxation diagnostic. The FENE free energy decays while the maximum trace ratio remains strictly below one.

9.2. AP closure. The AP test fixes Δt=0.05\Delta t = 0.05 and varies Wi. The error is the dual norm of Sh−D(uh)S_h - D(u_h) at the final time. The observed slope is essentially first order in Wi, as predicted by the AP closure estimate eq:7.1.

Wiclosure errorratemax tr⁡C/L2\operatorname{tr} C/L^2
8.0E−028.0\mathrm{E}{-02}6.41E−036.41\mathrm{E}{-03}–0.1830.183
4.0E−024.0\mathrm{E}{-02}3.18E−033.18\mathrm{E}{-03}1.011.010.1810.181
2.0E−022.0\mathrm{E}{-02}1.58E−031.58\mathrm{E}{-03}1.011.010.1800.180
1.0E−021.0\mathrm{E}{-02}7.89E−047.89\mathrm{E}{-04}1.001.000.1800.180
5.0E−035.0\mathrm{E}{-03}3.94E−043.94\mathrm{E}{-04}1.001.000.1800.180

Table 7. AP closure diagnostic with fixed Δt=0.05\Delta t = 0.05.

9.3. Fully coupled manufactured solution. The manufactured solution uses a periodic velocity, mean-zero pressure, and conformation tensor defined through the barrier map

Cex(t,x)=Bb(Ψex(t,x)).C_{\mathrm{ex}}(t,x) = \mathcal{B}_b\left(\Psi_{\mathrm{ex}}(t,x)\right).

This construction guarantees Cex∈DbC_{\mathrm{ex}} \in D_b and allows the manufactured forcing to exercise the nonlinear FENE stress, pressure projection, tensor transport, stretching, relaxation, and diffusion terms simultaneously. Table 8 reports the observed errors at final time T=0.05T = 0.05.

NNΔt\Delta t∥u−uh∥\lVert u-u_h\rVertrate∥C−Ch∥\lVert C-C_h\rVertratemin⁡(b−tr⁡Ch)\min(b-\operatorname{tr} C_h)
161.25E-031.84E-03–2.91E-03–5.82E+00
245.56E-048.11E-042.021.30E-031.995.81E+00
323.13E-044.58E-041.997.31E-042.005.80E+00
481.39E-042.04E-041.993.25E-042.005.80E+00

Table 8. Fully coupled manufactured FENE solve. Errors are discrete L2L^2 norms at T=0.05T=0.05.

FENE AP closure convergence plot, showing measured and slope-one curves

Figure 2. FENE AP closure. The scaled FENE stress converges linearly to the Newtonian closure at fixed time step.

High-Weissenberg robustness near the trace barrier

The final benchmark stresses the finite-extensibility boundary. We prescribe an extensional velocity field and initialize CC so that tr⁡C/L2\operatorname{tr} C/L^2 is already large. The test is run at increasing Wi with L2=20L^2=20, ε=2×10−3\varepsilon=2\times10^{-3}, and T=0.12T=0.12. Larger Wi weakens relaxation and increases stretch, but the barrier update keeps the conformation tensor inside DbD_b.

Wifinal energymax tr⁡C/L2\operatorname{tr} C/L^2min L2−tr⁡CL^2-\operatorname{tr} Crejected barrier steps
102.18E-020.8313.38E+000
502.94E-020.9101.80E+000
1003.51E-020.9588.40E-010
2004.10E-020.9794.20E-011

Table 9. High-Weissenberg barrier robustness at T=0.12T=0.12.

The last row shows one line-search rejection when Wi =200=200. The accepted state remains strictly within the barrier. This diagnostic is useful because a positivity-only method would not detect the finite-extensibility failure until the FENE coefficient becomes singular or changes sign numerically.

High-Weissenberg finite-extensibility diagnostic

Figure 3. High-Weissenberg finite-extensibility diagnostic. Larger Wi drives the trace closer to the barrier, but the accepted conformation remains strictly admissible.

Discussion

The FENE trace constraint changes what a structure-preserving method must prove. Positivity is still necessary, but it is no longer the complete admissibility condition. The stress coefficient fb(C)f_b(C) depends on the distance to the trace boundary, and the polymer molecular-diffusion term is entropy dissipative only when it is paired with the FENE barrier variable. A stable method must therefore preserve the accepted tensor in the set DbD_b, control reconstruction-level FENE entropy increments, and use that same tensor consistently in the stress force, stretching term, relaxation term, molecular diffusion, and entropy quadrature.

The barrier-log map (4) is one practical way to enforce the finite-extensibility geometry. It is a direct analogue of the log-conformation map, but its range is the FENE admissible set rather than the entire positive definite cone. The entropy-compatible reconstruction (5) adds the missing balance condition: among admissible barrier-log states, it chooses the least-damping state whose FENE entropy increment is within the prescribed budget. The energy proof itself is independent of the nonlinear parametrization; it requires only that the accepted tensor be admissible, entropy-compatible, and used consistently in the discrete coupling identities.

The AP part of the analysis also differs from Oldroyd–B in detail. The conformation perturbation C−IC-I is not the best stress variable because the FENE factor changes with tr⁡C\operatorname{tr} C. The regular stress variable is instead T(C)/WiT(C)/\mathrm{Wi}. This choice avoids singular scaling in the momentum equation and gives the correct Newtonian limit at fixed discretization parameters.

Conclusion

We have proposed and analyzed a barrier-preserving, entropy-compatible, asymptotic-preserving discretization for molecularly diffusive FENE-P type viscoelastic flow. The key finite-extensibility feature is the admissible set Db={C≻0,tr⁡C<b}\mathcal{D}_b = \{C \succ0,\operatorname{tr} C < b\}, which is enforced by a trace-barrier entropy and can be built into the nonlinear solve through a barrier logarithmic parametrization. High-order barrier-log reconstructions are filtered by a least-damping FENE entropy constraint, so admissibility and entropy compatibility are enforced simultaneously. Compatible quadrature makes the cancellation between FENE polymeric work and conformation stretching exact at the fully discrete level, while polymer molecular diffusion contributes the barrier Hessian dissipation. The resulting scheme satisfies a free-energy inequality with relaxation and molecular-diffusion barrier dissipation, preserves a positive trace buffer under an entropy bound, and has a regular small-Weissenberg limit in the scaled FENE stress variable. The numerical diagnostics verify the main mechanisms in both moderate and near-barrier regimes.

Acknowledgments

The author acknowledges financial support from the National Natural Science Foundation of China (NSFC, Grant No. 12501602), the Education Department of Hunan Province (Grant No. 24C0055), the Science and Technology Department of Hunan Province (Grant No. 2025JJ60052), and the Scientific Research Start-up Fund of Xiangtan University (Grant No. KZ0810769).

School of Mathematics and Computational Science, Xiangtan University, Xiangtan, Hunan, China

Email address: [email protected]

References

  1. [1]N. Balci, B. Thomases, M. Renardy, and C. R. Doering, Symmetric factorization of the conformation tensor in viscoelastic fluid models, J. Non-Newtonian Fluid Mech., 166 (2011), pp. 546–553.
  2. [2]J. W. Barrett and E. Süli, Existence of global weak solutions to finitely extensible nonlinear bead-spring chain models for dilute polymers, Math. Models Methods Appl. Sci., 21 (2011), pp. 1211–1289.
  3. [3]R. B. Bird, R. C. Armstrong, and O. Hassager, Dynamics of Polymeric Liquids, Volume 1: Fluid Mechanics, 2nd ed., Wiley, New York, 1987.
  4. [4]S. Boyaval, T. Lelievre, and C. Mangoubi, Free-energy-dissipative schemes for the Oldroyd-B model, ESAIM Math. Model. Numer. Anal., 43 (2009), pp. 523–561.
  5. [5]R. Fattal and R. Kupferman, Constitutive laws for the matrix-logarithm of the conformation tensor, J. Non-Newtonian Fluid Mech., 123 (2004), pp. 281–285.DOI
  6. [6]R. Fattal and R. Kupferman, Time-dependent simulation of viscoelastic flows at high Weissenberg number using the log-conformation representation, J. Non-Newtonian Fluid Mech., 126 (2005), pp. 23–37.
  7. [7]M. A. Hulsen, R. Fattal, and R. Kupferman, Flow of viscoelastic fluids past a cylinder at high Weissenberg number: stabilized simulations using matrix logarithms, J. Non-Newtonian Fluid Mech., 127 (2005), pp. 27–39.
  8. [8]S. Jin, Efficient asymptotic-preserving schemes for some multiscale kinetic equations, SIAM J. Sci. Comput., 21 (1999), pp. 441–454.
  9. [9]R. Keunings, On the high Weissenberg number problem, J. Non-Newtonian Fluid Mech., 20 (1986), pp. 209–226.
  10. [10]H. Liu and P. Yu, A finite difference method for the FENE dumbbell model of polymeric fluids, SIAM J. Numer. Anal., 49 (2011), pp. 2167–2193.
  11. [11]D. Richter, G. Iaccarino, and E. S. G. Shaqfeh, Simulations of three-dimensional viscoelastic flows past a circular cylinder at moderate Reynolds numbers, J. Fluid Mech., 651 (2010), pp. 415–442.

Paper details

vlnc:W_ys2pm4re
Version
v1 · latest
Published
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…
Contents