{"id":"b3925ac4-cb51-4b2c-8c87-69a694701eab","arxiv_id":"1909.08695","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A micromorphic finite-element formulation for nonlinear flexoelectricity avoids C1-continuous shape functions and is demonstrated on 3D benchmark problems.","lead":"The paper builds a finite-element framework for flexoelectricity, the polarization of a dielectric by strain gradients, using an extra 'micromorphic' field so standard C0 elements can be used. If it works, it makes 3D nonlinear flexoelectric simulations much easier, though the fully coupled model still awaits experimental validation.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Penalty-limit of the C0 micromorphic scheme is not demonstrated: Sec. 4.3.2 reports a persistent |F−F| mismatch at X=0, where |G| and the flexoelectric response peak, so the claimed flexoelectric solution may be an artifact of the finite-penalty regularization.","rationale":"The reader identified the constitutive assumption in Eq. (13) and the finite penalty p as the weakest load-bearing premise. My concern refines this into an internal numerical-consistency problem: even if Eq. (13) were accepted as a constitutive model, the C0 penalty method has not been shown to converge to the flexoelectric gradient limit, and the one diagnostic the paper reports points in the opposite direction in the boundary region where flexoelectric effects are largest. This is not an experimental-validity dispute; it is a question about whether the numerical object computed is the one claimed. A convergence study with a manufactured or reference solution would settle it. The lack of shipped code makes independent verification harder, but it does not change the overall conditional verdict: the paper remains a useful methods contribution, provided the missing p/h consistency property is established.","tokens_in":43391,"tokens_out":12614,"duration_ms":138529,"concrete_test":"Use a manufactured solution of the p-to-infinity gradient theory (or, alternatively, the same cantilever problem solved with a C1-continuous or finite-difference reference) and run the C0 micromorphic scheme on uniform meshes h, h/2, h/4, h/8 with p=5000 mu, 5e4 mu, and 5e5 mu, including choices with p proportional to h^{-1}. Track max|F_h−F_h| at X=0, the L2 error in G over the beam, and the point C-D potential difference. If the boundary value of |F−F| does not decrease as p and h vary and the potential difference does not converge to the reference value, then the C0 penalty construction does not reproduce the claimed flexoelectric solution.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step is the claim that Eq. (11), with finite penalty p, makes G=GradF a stand-in for Grad Grad phi in a C0 setting. For this to hold numerically, the H1-conforming micro-deformation field F_h must be able to track the macro deformation gradient F_h=Grad phi_h as p grows. But F_h is only piecewise polynomial and generally discontinuous across element faces, whereas F_h is C0-continuous. The paper's own Fig. 10(c) shows the consequence: |F−F| is large at the clamped end X=0, and the text states that this mismatch persists 'irrespective as to the choice of the penalty term p.' Since |G| is also largest at X=0, the flexoelectric energy (13) and the resulting electric potential are generated in the very region where the micro-macro tie is least reliable. No p- or h-convergence study is reported, and Section 5 concedes that the optimal functional setting is open. Thus the numerical flexoelectric potentials and size effects in Section 4 are not established as approximations to the claimed flexoelectric gradient theory: with finite p they belong to a different regularized theory with an arbitrary parameter, and with p approaching infinity the discrete mismatch prevents pointwise recovery of the limit. The central claim is therefore conditional on a consistency property that is neither proved nor tested.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a micromorphic formulation of nonlinear flexoelectricity. An independent micro-deformation field F and its material gradient G are introduced as additional degrees of freedom, and a scale-bridging penalty term couples F to the macroscopic deformation gradient F. In this way the higher-order gradient content of flexoelectricity is said to be accommodated within standard C0-continuous finite elements. The governing equations are derived from a stationary energy (Dirichlet) principle, the monolithic finite element formulation and tangent blocks are given, and three-dimensional numerical examples (a strip with a hole and a cantilever beam) illustrate size-dependent stiffening and flexoelectric-induced electric potentials. The manuscript also presents a classification of reduced models (M-, E-, EM-, FG-, FM-Elasticity) and discusses the penalty limit p→∞ as a route to gradient elasticity.","tokens_in":43856,"tokens_out":4927,"duration_ms":52195,"significance":"If the numerical consistency and validation gaps were closed, this would be a useful contribution: it offers a practical C0 finite element route to fully nonlinear flexoelectricity in three dimensions, with explicit constitutive forms, a complete tangent structure, and an open-source implementation based on deal.II and automatic differentiation. The variational derivation and the model classification are valuable and clearly presented. The central claim, however, is only partially supported by the evidence in the manuscript: no convergence study is reported, the finite-penalty mismatch between F and F is admitted at the boundary point where |G| and the flexoelectric response peak, and the full FM-Elasticity model is not validated against an independent reference solution or experiment. The paper is therefore a promising starting point rather than a fully established numerical framework for flexoelectricity.","major_comments":[{"comment":"The manuscript's own results show a persistent mismatch between the micro-deformation gradient field F and the macroscopic deformation gradient F at X=0, where |G| and the flexoelectric response are largest, and the text states that this mismatch persists 'irrespective as to the choice of the penalty term p'. Because the flexoelectric energy in Eq. (13) is linear in G, the computed electric potential in this region is generated by a field whose tie to the deformation gradient is explicitly unreliable. Since no p- or h-convergence study is presented, the numerical solutions in Section 4 are not established as approximations to the limiting gradient theory. I recommend a quantitative investigation: h-refinement at fixed p, p-refinement, and a comparison with an independent C1 reference solution on a simple geometry, together with an assessment of how the mismatch affects the quantities of interest such as the potential difference between points C and D.","section":"Sec. 4.3.2, Fig. 10(c)"},{"comment":"The FM-Elasticity formulation is validated only against benchmark problems for E-Elasticity and M-Elasticity; no validation of the fully coupled flexoelectric response is provided against experiment, an analytical solution, or an independent numerical method. Section 5 explicitly defers experimental validation to future work. Since the central claim is that the model captures flexoelectric polarization and size-dependent effects, at least one quantitative benchmark is needed, for example a comparison with the linearized analytical solution for a bent beam or with the C1 finite element results of Yvonnet and Liu for a simplified geometry. Without such a test, the constitutive choices in Eq. (13) and the resulting potentials remain unverified model predictions.","section":"Sec. 5 (Discussion) and Sec. 4.2"},{"comment":"The claim that the micromorphic scheme recovers the flexoelectric gradient theory as p→∞ is not established. The scale-bridging term (Eq. (11)) penalizes the L2 difference between the piecewise polynomial micro-deformation F_h and the continuous deformation gradient F_h = Grad phi_h; because F_h is generally discontinuous across element faces while F_h is continuous, the penalty limit cannot deliver pointwise equality. The persistent gap shown in Fig. 10(c) is consistent with this observation. The authors should either prove a Gamma-convergence or projection-type result for the discrete limit, or reformulate the claim to present FM-Elasticity with finite p as a regularized model in its own right, with p as a constitutive parameter whose influence is systematically studied.","section":"Sec. 2.4.3, Eq. (13) and Remark 6"},{"comment":"No mesh-refinement or polynomial-order study is reported for the coupled three-field formulation, despite the non-standard choice of tri-quadratic approximations for phi and F with a tri-linear approximation for the electric potential. For a mixed-type nonlinear problem of this structure, stability and convergence of the monolithic Newton iteration cannot be taken for granted. A demonstration of h-convergence (for example, tracking the tip deflection and the potential difference on the cantilever problem) and a brief study of the sensitivity to the penalty parameter p are essential to support the numerical claims of the paper.","section":"Secs. 3 and 4"}],"minor_comments":[{"comment":"The contraction notation in Eq. (13), involving a vector f^T E, a third-order tensor G, and a second-order tensor I, is not covered by the scalar-product definitions in the Notation section, which define only contractions of same-order tensors. The precise index contraction should be written out explicitly.","section":"Notation, Eq. (13)"},{"comment":"Reference [46] (Vu, Steinmann and Possart, International Journal for Numerical Methods in Engineering, 70(6):685–704) appears to be a duplicate of reference [40]; the bibliography should be deduplicated and the citation in Sec. 4.2 should point to the intended entry.","section":"References"},{"comment":"In panels (c) and (d) of Fig. 10, the axes labelled 'X' and '|G|' are ambiguous because the same x-axis scale is used for different lines; please label the axes explicitly and clarify which quantity is plotted along the horizontal line A–B versus the vertical line C–D.","section":"Fig. 10"},{"comment":"The sentence 'Several choices for p were investigated and all produced similar behaviour' is too vague to be useful; if these investigations are not reported in detail, the statement should be removed or replaced by a figure or table.","section":"Sec. 4.3.2"}],"recommendation":"major_revision","confidential_remarks":"The paper fits the scope of an applied mechanics journal and the micromorphic formulation is a plausible contribution. My main concern is the gap between the advertised 'C0 framework for flexoelectricity' and the numerical evidence: the admitted penalty mismatch at the point of maximal flexoelectric response and the absence of a convergence or validation study are load-bearing for the central claim. The authors should be encouraged to add a benchmark, a convergence analysis, and a systematic p-sensitivity study; with those additions, the contribution could become publishable. I do not see a basis for rejection on novelty or correctness-of-derivation grounds."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Zack, here's my take. The core idea is to introduce a micromorphic deformation gradient field F and penalize its deviation from the macroscopic F, so the flexoelectric term can be written in terms of Grad F rather than second gradients of the displacement. That lets them use standard C0 elements in 3D, which is genuinely new relative to Yvonnet & Liu's C1 Argyris triangles, Abdollahi's meshless 2D, and Mao's linearized mixed formulation. The variational derivation is standard but clean, and they provide all the tangent blocks. The E-elasticity benchmark reproduces Vu et al., which is a good sanity check. The numerical examples show the expected size-dependent stiffening and flexoelectric potentials that scale linearly with the coefficient, as the energy in Eq. (13) implies.\n\nThe soft spot is the one they flag themselves in Sec. 4.3.2 and then don't fully address. Figure 10(c) shows that |F−F| does not shrink at the clamped end, and the text admits this persists regardless of p. Since |G| is also largest there, the flexoelectric potential is generated precisely where the micromorphic/macroscopic tie is least reliable. With finite p the scheme solves a regularized problem with an extra parameter, and the p→∞ limit is not demonstrated. There is no p- or h-convergence study. So the claim that the micromorphic approach accommodates the second-gradient flexoelectric theory is not yet supported by numerical evidence that the regularization is actually converged. The authors say the optimal functional setting is open, which is honest, but it means the central claim is conditional.\n\nThe other weakness is calibration: the coupled FM-elasticity model is not compared to experiment or to an independent reference solution, and the flexoelectric energy is a proposal. The paper itself calls experimental validation critical future work. For a methods paper that's acceptable, but it limits what one can infer from the computed potentials.\n\nStill, the paper is worth engaging with. The formulation is coherent, the derivations are complete, and the approach will likely be reused. Before building on it, I would want a convergence test in p on a smooth problem and a clearer discussion of whether the clamped-end mismatch is a modeling artifact or a discretization limitation. It deserves a serious referee; with revisions addressing the penalty limit and a benchmark, it would be solid.","headline":"A genuinely useful C0 finite-element framework for finite-strain flexoelectricity, but the numerical evidence that it converges to the targeted gradient theory is thinner than the headline claim suggests.","tokens_in":44221,"tokens_out":2477,"would_cite":true,"duration_ms":29356,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A penalty-coupled micromorphic field lets ordinary C0 finite elements capture flexoelectricity and size effects.","keywords":["flexoelectricity","micromorphic continua","finite element method","strain gradient","electromechanical coupling","size-dependent response","nonlinear electro-elasticity"],"falsifier":"Measure the electric potential across a bent cantilever of a known flexoelectric material (for example a relaxor ceramic) at several beam thicknesses and compare with the model's predicted linear scaling of polarization with the flexoelectric coefficient and concentration of the micro-gradient at the clamped end. If the measured potential does not grow linearly with $\\upsilon$ or does not show the predicted size-dependent distribution, the assumed form of Eq. (13) is ruled out. A numerical check is also possible: compute the same cantilever with increasing penalty $p$ and check that the potential converges; the paper reports that near the clamped boundary $\\bar F$ cannot be tied to $F$ no matter how large $p$ is, so this is where the approximation should be tested.","tokens_in":43193,"feed_emoji":"⚡","tokens_out":8205,"duration_ms":79399,"temperature":0.7,"pith_summary":"Flexoelectricity couples the gradient of deformation to electric polarization, but the governing equations contain second gradients of the motion, which normally force finite element approximations to be globally $C^1$-smooth. This paper claims that a micromorphic reformulation removes that obstacle: an independent micro-deformation field $\\bar F$, penalised to stay close to the deformation gradient $F$, supplies the needed gradient information while all fields remain $C^0$. The result is a fully nonlinear, three-dimensional finite element formulation of flexoelectricity that also reproduces size-dependent stiffening. The payoff is that complex flexoelectric geometries become accessible with standard finite element technology rather than specialised $C^1$ elements.","feed_headline":"Ordinary finite elements can now capture flexoelectricity","feed_subtitle":"A penalty-coupled micromorphic field supplies the missing gradient, opening 3D simulation of strain-gradient polarization.","key_machinery":"The load-bearing device is the micromorphic micro-deformation $\\bar F$ and its gradient $\\bar G = \\mathrm{Grad}\\,\\bar F$, promoted to independent degrees of freedom. The scale-bridging energy $\\tfrac12 p[\\bar F - F]:[\\bar F - F]$ with penalty parameter $p$ lets $\\bar G$ approximate the true second gradient $G = \\mathrm{Grad}\\,\\mathrm{Grad}\\,\\boldsymbol{\\varphi}$; as $p\\to\\infty$ the formulation reproduces gradient elasticity, while finite $p$ keeps the problem $C^0$. The proposed flexoelectric energy $\\psi_{\\mathrm{flexo}} = \\upsilon [f^T E]\\cdot \\bar G : I$ is linear in $\\bar G$, which makes the flexoelectric contribution to the dielectric displacement scale linearly with the flexoelectric coefficient. A parameter-classification diagram shows how the same code reduces to ordinary elasticity, electro-elasticity, micromorphic elasticity and gradient elasticity by sending selected parameters to zero or infinity.","core_discovery":"The central claim is that flexoelectricity can be formulated as a micromorphic continuum problem and solved with conventional $C^0$-continuous finite elements. The authors introduce the micro-deformation $\\bar F$ as an independent field whose gradient $\\bar G = \\mathrm{Grad}\\,\\bar F$ plays the role of the second gradient of the macroscopic motion; a penalty-like scale-bridging energy $\\tfrac12 p[\\bar F - F]:[\\bar F - F]$ ties $\\bar F$ to $F$, so in the limit $p\\to\\infty$ the micromorphic gradient approaches $G = \\mathrm{Grad}\\,\\mathrm{Grad}\\,\\boldsymbol{\\varphi}$. The flexoelectric energy is taken to be linear in $\\bar G$, coupling it to the pulled-back electric field, and the Dirichlet principle supplies the coupled balance equations for the macroscopic motion, the micro-deformation and the electric potential. Numerical examples in three dimensions demonstrate the size-dependent stiffening of a strip with a hole and the flexoelectric potential generated in a bent cantilever.","pith_inferences":["Because the penalty parameter $p$ is finite in practice, the computed $\\bar G$ is a regularised proxy for $G = \\mathrm{Grad}\\,\\mathrm{Grad}\\,\\boldsymbol{\\varphi}$; a systematic study of how the predicted polarization depends on $p$, especially near clamped boundaries, would quantify how much of the response is model artifact.","The linearity of Eq. (13) in $\\bar G$ is a strong, testable prediction: if experiments on a single material show a nonlinear relation between strain gradient and polarization, the proposed energy must be replaced by a higher-order term, though the micromorphic machinery itself would survive.","The same penalty-micromorphic idea could be applied directly to other fourth-order coupled problems such as strain-gradient piezoelectricity or flexomagnetic coupling, since the mathematical obstruction—missing $C^1$ continuity—is identical."],"forward_implications":["A single $C^0$ finite element code now covers a spectrum of theories—nonlinear elasticity, electro-elasticity, micromorphic elasticity, gradient elasticity and flexoelectric micromorphic elasticity—by taking limits of the same constitutive parameters.","For the bent cantilever, the model predicts that the flexoelectric potential scales linearly with the flexoelectric coefficient and concentrates where the deformation gradient is concentrated, giving a computable signature of where polarization should appear in experiments.","Because micromorphic elasticity is built in, the formulation simultaneously captures size-dependent stiffening, so predictions for small structures include the mechanical size effect that accompanies flexoelectricity.","The converse flexoelectric effect can be added by introducing a micromorphic electric field and a similar scale-bridging energy, extending the same machinery to actuation problems."],"supporting_citations":[{"why":"Supplies the micromorphic hyperelasticity theory and the quadratic energy forms for the elastic, micromorphic and scale-bridging contributions.","marker":"[45]"},{"why":"The preceding nonlinear flexoelectricity model at finite strain that requires C1-continuous elements; the present method avoids that restriction.","marker":"[37]"},{"why":"Provides the coupled nonlinear electro-elasticity formulation whose benchmarks validate the E-Elasticity limit of the new code.","marker":"[39]"},{"why":"Supplies the strip-with-hole electro-elasticity results that the E-Elasticity computations reproduce.","marker":"[46]"},{"why":"Establishes the micromorphic route to gradient-type continua and motivates the parameter-limit reduction scheme.","marker":"[21]"}],"fun_headline_variants":["Micromorphic fields bring flexoelectricity to standard FEM","Penalty-bridged gradients unlock flexoelectric 3D simulations","Modeling flexoelectricity with a micromorphic twist","Flexoelectric effect tamed by micromorphic finite elements","A micromorphic path to simulate flexoelectricity in solids"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the constitutive choice in Eq. (13) that flexoelectric energy is linear in the micromorphic gradient $\\bar G$, together with the finite penalty coupling $p$ standing in for the true second gradient; the paper itself notes near concentrations the tie $\\bar F \\approx F$ cannot be sharpened regardless of $p$, and that experimental validation of the full FM-Elasticity model is future work.","fun_headline_variants_meta":{"raw":{"variants":["Micromorphic fields bring flexoelectricity to standard FEM","Penalty-bridged gradients unlock flexoelectric 3D simulations","Modeling flexoelectricity with a micromorphic twist","Flexoelectric effect tamed by micromorphic finite elements","A micromorphic path to simulate flexoelectricity in solids"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000141,"raw_usage":{"total_tokens":1111,"prompt_tokens":842,"completion_tokens":269,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":458,"completion_tokens_details":{"reasoning_tokens":189}},"tokens_in":458,"tokens_out":269,"duration_ms":3221,"temperature":1.0,"reasoning_tokens":189,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:22:34.276093+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the electric potential across a bent cantilever of a known flexoelectric material (for example a relaxor ceramic) at several beam thicknesses and compare with the model's predicted linear scaling of polarization with the flexoelectric coefficient and concentration of the micro-gradient at the clamped end. If the measured potential does not grow linearly with $\\upsilon$ or does not show the predicted size-dependent distribution, the assumed form of Eq. (13) is ruled out. A numerical check is also possible: compute the same cantilever with increasing penalty $p$ and check that the potential converges; the paper reports that near the clamped boundary $\\bar F$ cannot be tied to $F$ no matter how large $p$ is, so this is where the approximation should be tested.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the micromorphic hyperelasticity theory and the quadratic energy forms for the elastic, micromorphic and scale-bridging contributions."},{"cited_title":"Yvonnet and L","cited_arxiv_id":null,"evidence_quote":"The preceding nonlinear flexoelectricity model at finite strain that requires C1-continuous elements; the present method avoids that restriction."},{"cited_title":"Pelteret, D","cited_arxiv_id":null,"evidence_quote":"Provides the coupled nonlinear electro-elasticity formulation whose benchmarks validate the E-Elasticity limit of the new code."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the strip-with-hole electro-elasticity results that the E-Elasticity computations reproduce."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the micromorphic route to gradient-type continua and motivates the parameter-limit reduction scheme."}],"review_version":1}