{"id":"30c5b7eb-8b0b-44f6-8f81-10a50db6c111","arxiv_id":"2607.21164","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A discrete two-node Cosserat rod model with compliance-form stresses is derived from a mixed Petrov-Galerkin rod FEM via midpoint/trapezoidal quadrature, and demonstrated on three benchmarks.","lead":"Researchers derive a fast 'chain of blocks' rod model from a rigorous finite-element rod theory, showing the simplified model is a special case of the expensive one. This gives soft-robotics and animation simulators a way to keep the speed of discrete rod models while inheriting the accuracy and stability of a mixed finite-element formulation.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (10) has an element-length scaling error: K_c = L_el·diag(...) is the inverse of the K_c that §4's derivation requires; the printed constitutive law makes stiffness scale as L_el^2.","rationale":"The reader identified the reliance on the unpublished [15] and the unproven index-1 DAE as the weakest assumptions. These are valid secondary concerns, but the most load-bearing issue is an internal mathematical error: Eq. (10) is inconsistent with the derivation in §4 and with physical dimensions. Since the abstract's central claim is that the discrete model arises rigorously from the mixed FEM, and the discrete model's governing equations as printed are not the ones derived, the claim is not supported as written. The numerical examples appear to use the corrected scaling, so the paper is likely correctable, but the manuscript in its current form contains a false equation in a core component. This warrants a rejection (or at minimum a major revision) rather than conditional acceptance based on external verification.","tokens_in":13671,"tokens_out":17503,"duration_ms":150661,"concrete_test":"Single-element axial test: set L_el=2, EA=1, apply a prescribed stretch γ−γ0=0.01 to the free node. From Eq. (9) with K_c as printed, the internal force is n = EA·L_el^2·0.01 = 0.04. Re-derive the same test from the midpoint-integrated mixed virtual work (28), which gives n = EA·(γ−γ0) = 0.01. If the discrete equilibrium with an external force F=0.01 yields γ−γ0=0.01, then Eq. (10) must be corrected to K_c = (1/L_el)·diag(...). Reproduce this check with the published code to confirm the discrepancy.","verdict_should_be":"REJECT","load_bearing_attack":"The discrete constitutive law in Eqs. (9)–(11) is dimensionally inconsistent with the mixed-FEM derivation in §4. Eq. (10) defines K_c,i = L_el_i·diag(EA, GA, GA, GJ, EI, EI) and Eq. (11) defines l_i = L_el_i·(γ_i−γ0_i, κ_i−κ0_i). Substituting into Eq. (9), K_c^{-1}λ − l = 0 gives λ = K_c·l = L_el^2·diag(...)·(γ−γ0). This predicts an internal force n = EA·L_el^2·(γ−γ0), whereas the physically correct and §4-derived relation is n = EA·(γ−γ0). The error is traceable: §4 starts from the mixed FEM virtual work (28), whose compliance term is δn^T(K_γ^{-1}n − (γ−γ0)); after midpoint integration this becomes L_el·δn^T(K_γ^{-1}n − (γ−γ0)). For the discrete term δλ^T(K_c^{-1}λ − l) in (33) to match, one must have K_c^{-1} = L_el·K_γ^{-1}, i.e., K_c = (1/L_el)·diag(...), not L_el·diag(...). Thus the printed Eq. (10) has a reciprocal error. A reader implementing the paper as written would obtain a stiffness that vanishes or grows with mesh size, breaking the claimed equivalence to the FEM and the numerical convergence. This is an internal inconsistency independent of the unpublished reference [15]; it directly invalidates the central claim as stated.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a discrete Cosserat rod model representing the rod as a chain of rigid nodes coupled by compliant elements with independent stress degrees of freedom. The central claim is that this discrete model arises rigorously from a mixed Petrov–Galerkin Cosserat rod FEM at linear kinematic interpolation order when internal virtual work is integrated by the midpoint rule and external/inertial contributions by the trapezoidal rule. Section 4 derives the discrete equilibrium and dynamic equations from the mixed FEM virtual work, and three numerical examples (two-segment helix, tendon-driven manipulator, Wilberforce pendulum) are used to demonstrate accuracy, mesh convergence, and physical behavior.","tokens_in":14104,"tokens_out":21918,"duration_ms":179283,"significance":"If the derivation is correct, the paper provides a valuable bridge between mixed FEM and discrete rod models: it yields a banded, two-node coupling structure that is convenient for GPU-based implementations, preserves compliance-based constitutive laws, and may inherit locking-free behavior from the underlying mixed formulation. The paper also ships reproducible code and reports clean second-order convergence against an analytical helix solution. These strengths make the result potentially important for soft robotics, computer graphics, and multibody dynamics. However, the central claim currently rests on two specific technical points that need correction or clarification: a scaling error in the element constitutive law and an incomplete treatment of the virtual-rotation interpolation.","major_comments":[{"comment":"Equation (10) defines K_{c,i} = L_i^el diag(EA, GA, GA, GJ, EI, EI), while Eq. (11) defines l_i = L_i^el (γ_i−γ0_i, κ_i−κ0_i). Substituting into Eq. (9), K_{c,i}^{-1}λ_i − l_i = 0 gives λ_i = K_{c,i} l_i = (L_i^el)^2 diag(...)(γ_i−γ0_i). This is dimensionally inconsistent (n_i would have units N·m²) and makes the effective stiffness vanish as the mesh is refined. The §4 derivation, after the midpoint rule applied to Eq. (28), yields the compliance term L_i^el δn^T(K_γ^{-1}n − (γ−γ0)), which requires K_{c,i}^{-1} = L_i^el K_γ^{-1}, i.e., K_{c,i} = (1/L_i^el) diag(...). As printed, Eq. (10) is a reciprocal L_el error that breaks the claimed equivalence to the mixed FEM and contradicts the observed convergence. The authors must correct Eq. (10) and verify that the accompanying code uses the corrected expression.","section":"§3, Eq. (10) with §4, Eq. (33)"},{"comment":"The derivation linearly interpolates the virtual rotations as δϕ_lin(s_el) = (δϕ_{i−1}+δϕ_i)/2, but δϕ_i is defined as the virtual rotation in the nodal frame B_i, while the stress and strain quantities in Eq. (28) are in the element frame. For finite inter-element rotations, these are different frames, and a coordinate-wise average of body-frame virtual rotations is not frame-invariant. The resulting force Jacobian W_c in Eq. (20) contains no rotation matrices between the nodal and element frames for the moment blocks (bottom rows). The authors should either specify a consistent frame for the virtual-rotation interpolation (e.g., inertial or element frame) and show the corresponding transformations in Eq. (33), or restrict the rigor claim to small inter-element rotation increments. Without this clarification, the derivation from the mixed FEM is incomplete.","section":"§4, Eq. (30) and Eq. (20)"}],"minor_comments":[{"comment":"The text states that the second target helix has height h/2, but the parameterization in Eq. (42) gives a vertical rise of c·β(L) = h for the second segment. Either the text or the parameterization should be corrected; the numerical results themselves are still well-defined.","section":"§5, first example"},{"comment":"The statement that Eqs. (12)–(14) constitute an index-1 DAE is asserted without proof. A brief justification (e.g., invertibility of K_c and differentiability of l_c(q)) would be helpful, since the dynamic solver selection depends on this property.","section":"§3, after Eq. (21)"},{"comment":"Please state how the error ∥r(s)−r*(s)∥ is evaluated at non-nodal points (e.g., interpolation of the discrete centerline) and define the norm used, so the convergence results are fully reproducible.","section":"§5, Fig. 2"},{"comment":"The derivation in §4 relies on the unpublished preprint [15] for the mixed FEM internal virtual work expression. The paper should explicitly note this dependence, since the rigor of the present derivation is conditional on the correctness of that formulation.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The Eq. (10) scaling error is most likely a typo that can be fixed in revision, but the authors should double-check their numerical implementation against the corrected law. The virtual-rotation interpolation issue in §4 is more subtle and, if not resolved, undermines the claimed rigorous derivation. The paper's heavy reliance on the same group's unpublished preprint [15] is acceptable for a communication, but it should be clearly flagged. If the authors can address both major points, the paper would make a solid contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Frankly, the paper's central claim does not survive its own equations. In Eq. (10), K_c,i is defined as L_i diag(EA, GA, GA, GJ, EI, EI). Substituting into Eq. (9) gives an internal force n = L_i^2 EA (gamma - gamma_0), so stiffness scales with element length squared. Section 4's quadrature reduction of the mixed FEM requires exactly the reciprocal: K_c = (1/L_i) diag(...) and l_i = L_i (gamma - gamma_0), so that K_c^{-1} lambda - l matches L_i [K_gamma^{-1} n - (gamma - gamma_0)]. As printed, the constitutive law is dimensionally inconsistent and the claimed equivalence to the mixed FEM is false. A reader implementing Eqs. (9)-(11) would see the rod soften without bound as the mesh is refined, contradicting the convergence reported in the helix example. This is not a subtle sign error; it breaks the paper's main selling point. Now the credit. The idea of presenting a two-node discrete Cosserat rod as the midpoint/trapezoidal quadrature reduction of a p=1 mixed Petrov-Galerkin FEM is genuinely worth having. The banded force Jacobian in (20) is clean, and the derivation in Section 4 would be elegant once the reciprocal is fixed. The numerical examples are well chosen: variable cross-section helix, tendon-driven tapered rod, and the Wilberforce pendulum. The helix and tendon results show the expected trends. The weaker parts beyond the scaling error: the 'absence of locking' is inherited from ref. [15], an unpublished same-group preprint, and no baseline comparison is given. In the Wilberforce example, the bob's moment of inertia is 'appropriately tuned' to get the pi/2 phase shift, but the value is not reported, so the reader cannot reproduce the tuning. These are addressable, but as they stand they leave the strongest dynamic claim unsupported. So: the paper deserves a serious referee because the core idea is valuable and the error looks fixable, but it should not be accepted as is. I would send it back for major revision with the scaling correction, a demonstration of locking-free behavior against a non-mixed baseline, and a reported inertia for the pendulum. My verdict would be reject-and-resubmit rather than accept.","headline":"The bridging idea is genuinely useful, but Eq. (10) has a reciprocal element-length error that breaks the paper's central equivalence claim.","tokens_in":642,"tokens_out":1061,"would_cite":false,"duration_ms":55725,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["65M60","74K10","74S05"],"pacs":[],"model":"deepseek-v4-flash","headline":"A chain of rigid bodies with compliant links is proven to be exactly a degree-one mixed Petrov–Galerkin Cosserat rod finite element, provided internal work is integrated at element midpoints and external and inertial work at nodes by the tr","keywords":["Cosserat rod","discrete rod model","mixed finite element","Petrov–Galerkin","Hellinger–Reissner","locking-free","multibody dynamics","differential-algebraic equations"],"falsifier":"Discretize a clamped, nearly inextensible rod with the chain model and with the companion mixed FEM using exact quadrature at the same mesh; if the two converge to different equilibrium shapes, or if the chain model locks (e.g., shear oscillations) while the exact-quadrature FEM does not, the claimed equivalence fails. A cheaper check: replace the midpoint rule in the internal virtual work with exact integration and observe whether the discrete strain definitions still coincide with equation (7).","tokens_in":13583,"feed_emoji":"🪢","tokens_out":4144,"duration_ms":42079,"temperature":0.7,"pith_summary":"This paper claims that a discrete Cosserat rod model — a chain of rigid nodes coupled by compliant forces and moments — is not a heuristic construction but exactly a degree-one mixed Petrov–Galerkin Cosserat rod finite element, provided the internal virtual work is integrated by the midpoint rule and the external and inertial virtual work by the trapezoidal rule. The chain model's strain measures come from the relative kinematics of adjacent node pairs, and the constitutive law is written in compliance form with independent stress unknowns. The significance is twofold: the model inherits the locking-free robustness of the underlying mixed FEM, and it exposes the two-node banded coupling of computer-graphics discrete rods rather than the dense coupling of strain-parameterized reduced-order models. The paper verifies this claim with static helix and tendon-driven examples plus a dynamic Wilberforce pendulum.","feed_headline":"Two-node rod chains equal a mixed Cosserat FEM","feed_subtitle":"Midpoint internal and trapezoidal external quadrature turn a p=1 Petrov–Galerkin beam into a compliant rigid-body chain.","key_machinery":"The engine is the mixed Petrov–Galerkin variational principle in Hellinger–Reissner form: internal force and moment resultants are independent stress degrees of freedom, conjugate to compliance-weighted strain measures. Linear interpolation of the kinematics, combined with the midpoint rule for internal quadrature, yields the discrete strain definitions and the two-node coupling blocks; the trapezoidal rule yields the diagonal lumped inertia. The whole argument rests on the identity between midpoint-evaluated linear interpolation and the relative kinematics of adjacent node pairs.","core_discovery":"The central discovery is an exact equivalence between two apparently different objects: a rigid-body-chain Cosserat rod and a mixed finite element at linear kinematic interpolation order. Starting from the internal virtual work of a companion mixed Petrov–Galerkin formulation and evaluating it at element midpoints reproduces the discrete strains and the banded force-direction Jacobian; applying the trapezoidal rule to external and inertial virtual work produces the lumped nodal masses, inertias, and loads. The discrete rod is therefore the p=1 mixed FEM with specific low-order quadrature, and its equations of motion form a differential-algebraic system whose algebraic part is just the compli","pith_inferences":["Inference: If the equivalence holds generally, error analyses and locking studies for the underlying mixed FEM transfer directly to graphics-style discrete rod solvers, potentially giving rigorous accuracy guarantees to algorithms developed for visual simulation.","Inference: The specific quadrature choices are likely what make the discrete strains objective and the coupling local; testing alternative quadrature rules in the same variational setting would generate a family of discrete rods with different accuracy and locking trade-offs.","Inference: A natural testable extension is to use the banded structure with local relaxation for frictional contact between multiple rods, a direction the paper names as future work.","Inference: The equivalence suggests that the compliance-form of the discrete rod is not an approximation but the mixed method's natural structure, so hybrid rigid–soft robot models built from such chains can rely on standard saddle-point solvers."],"forward_implications":["The rigid-body-chain model is a rigorously derived FEM discretization, so its convergence theory and locking-free behavior can be inherited from the mixed method rather than argued heuristically.","Because coupling is only between adjacent nodes, the mass matrix is constant and diagonal and the force Jacobian is banded, making local and parallel (GPU-style) relaxation solvers a natural next step.","The compliance-split DAE form keeps the nonlinearity in the algebraic constraint, which the paper argues is more robust and efficient than condensing it into the kinetic equations.","The formulation handles variable cross sections and stiffness discontinuities without special treatment, as demonstrated by the two-helix and tendon-driven examples.","The Wilberforce pendulum example shows the model reproduces coupled longitudinal–torsional energy exchange with a near π/2 phase shift."],"fun_headline_variants":["Rod chains are mixed FEM with midpoint quadrature","Cosserat rod chain equals a mixed finite element","Discrete rod is a mixed FEM in disguise","Midpoint rule turns beam FEM into nodal chains","Two-node rod model matches mixed Cosserat FEM"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The derivation inherits every equation from a companion unpublished mixed FEM, so if that formulation's stress interpolation or virtual-work sign conventions are wrong the discrete model inherits the error; separately, the dynamic DAE is assumed index-1 for all configurations without proof.","fun_headline_variants_meta":{"raw":{"variants":["Rod chains are mixed FEM with midpoint quadrature","Cosserat rod chain equals a mixed finite element","Discrete rod is a mixed FEM in disguise","Midpoint rule turns beam FEM into nodal chains","Two-node rod model matches mixed Cosserat FEM"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000279,"raw_usage":{"total_tokens":1487,"prompt_tokens":728,"completion_tokens":759,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":472,"completion_tokens_details":{"reasoning_tokens":686}},"tokens_in":472,"tokens_out":759,"duration_ms":6761,"temperature":1.0,"reasoning_tokens":686,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T08:17:32.307785+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Discretize a clamped, nearly inextensible rod with the chain model and with the companion mixed FEM using exact quadrature at the same mesh; if the two converge to different equilibrium shapes, or if the chain model locks (e.g., shear oscillations) while the exact-quadrature FEM does not, the claimed equivalence fails. A cheaper check: replace the midpoint rule in the internal virtual work with exact integration and observe whether the discrete strain definitions still coincide with equation (7).","supporting_citations":[],"review_version":1}