Predictive Finite Element Shear Fracture Modeling under High Strain Rate Creasing Conditions
Predictive FEA of high strain rate creasing maps shear fracture mechanisms to optimize tool geometry and prevent score cracking on high-speed converting lines.

Strain
Paperboard conversion on modern high-speed folder-gluers subjects structural boxboard to severe localized mechanical deformation within milliseconds. Creasing operations executed at production line speeds between 300 and 800 meters per minute generate effective strain rates inside the scoring zone ranging from 102 s-1 to exceeding 104 s-1. Paperboard is an anisotropic composite.
Under quasi-static testing, multi-ply folding boxboard accommodates folding through controlled inter-ply delamination, allowing individual plies to slide over one another and form a internal hinge. Dynamic high strain rate deformation fundamentally alters this fracture mechanism. As deformation speed rises, polymer constituent behaviors within the wood fibre matrix shift from viscoelastic dissipation to brittle shear localization.
Shear stresses govern crease formation.
During the initial descent of a male creasing rule into a female matrix channel, the upper coated surface experiences intense out-of-plane compression while the lower back side undergoes complex combinations of bending tension and shear. At strain rates above 100 s-1, the hydrogen-bonded network of cellulose fibres and hemicellulose exhibits rate-dependent stiffening. The apparent tensile modulus along the machine direction increases, while the cross-direction shear strength degrades sharply under rapid cyclic displacement.
This rate sensitivity prevents traditional strain energy dissipation. High strain rates stiffen fibres. Instead of distributing deformation across a broad delamination zone, the board concentrates shear strain into narrow planes directly adjacent to the creasing rule edges, accelerating micro-crack initiation in the middle plies.
Fibre network stiffening under microsecond loading shifts energy absorption from diffuse ply shear into concentrated interfacial delamination.
Understanding the micro-mechanics of this deformation requires dissecting the multi-ply geometry of paperboard furnish. Top layers typically consist of bleached chemical pulp paired with mineral pigment coatings, middle layers contain mechanical pulp or recycled fibres with lower tensile elongation, and bottom layers feature bleached or unbleached chemical pulp to carry tensile loads during folding. Dynamic creasing forces the middle mechanical layer into intense out-of-plane shear.
When the rate of shear displacement outpaces the viscoelastic relaxation time of the native lignin-hemicellulose matrix, internal fracture occurs via rapid inter-ply debonding or intra-ply shear rupture. Coated layers crack under tension. FEA models that fail to incorporate rate-dependent shear yield criteria consistently underpredict the onset of score cracking on high-speed packaging lines.
| Ply Furnish Layer | Quasi-Static Modulus (ISO 1924-2) | Dynamic Modulus at 10^3 s^-1 | Quasi-Static Interlaminar Shear Strength | Dynamic Interlaminar Shear Strength at 10^3 s^-1 |
|---|---|---|---|---|
| Bleached Chemical Top Ply | 4200 | 5850 | 1.85 | 1.25 |
| Mechanical Middle Ply | 2100 | 3150 | 0.95 | 0.52 |
| Unbleached Chemical Back Ply | 3800 | 5200 | 1.60 | 1.10 |
| Recycled Fibre Middle Ply (CRB) | 1750 | 2800 | 0.72 | 0.38 |
The transition from ductile delamination to brittle shear fracture destabilizes the converting process window. When the middle ply fails to shear smoothly, bending moments transfer directly to the top coated surface, driving strain values past the ultimate elongation threshold of the mineral coating layer. The physical result manifests as visible score cracking along the exterior hinge, exposing underlying dark fibres or tearing white top sheets on finished pharmaceutical and food folding cartons.
Specifying stock based strictly on static tensile testing leads directly to unaccounted conversion scrap when running automated lines above 400 blanks per minute.

Formulation
Predictive finite element modeling of paperboard under high strain rate creasing demands robust constitutive formulations capable of capturing anisotropic elastoplasticity alongside rate-dependent damage evolution. Standard isotropic elastoplastic models fail to represent the orthotropic nature of paper, where machine direction, cross direction, and out-of-plane thickness direction exhibit stiffness ratios often exceeding 10:5:1. Constitutive laws must employ anisotropic yield surfaces, such as Hill ’48 or Hoffman criteria, adapted to incorporate strain-rate scaling functions for both yield stress and hardening moduli.
Calculations demand exact yield parameters.
Rate dependence is integrated through Cowper-Symonds or Johnson-Cook type formulations modified for orthotropic solids. The yield stress tensor scales dynamically as a function of the plastic strain rate invariant. Under dynamic shear, the yield surface expands asymmetric with respect to tension and compression.
Out-of-plane shear stress components dictate the initiation of internal damage long before total effective strain reaches equivalent uniaxial yield limits. Incorporating damage mechanics requires coupling the anisotropic yield function with a continuum damage parameter that degrades material stiffness tensors as plastic strain accumulates within specific directional planes.
Contractual compliance under ISO 1924-3 fails to protect converters against dynamic score fracture when strain rates exceed 100 s^-1.
Interlaminar failure simulation relies on Cohesive Zone Modeling (CZM) or virtual crack closure techniques embedded between discrete structural continuum plies. Cohesive element formulation requires traction-separation laws defined separately for Mode I normal opening and Mode II or Mode III transverse shear sliding. At elevated strain rates, the peak critical traction increases while the total fracture toughness decreases, compressing the bilinear softening area of the cohesive law.
The numerical model must govern this degradation using rate-dependent energy release rates. Delamination lowers bending stiffness.

Constitutive Rate-Sensitivity Formulations
Mathematical modeling of the dynamic yield stress tensor utilizes an extended orthotropic rate-dependent parameterization. The instantaneous yield stress along a principal axis scales according to the local plastic strain rate:
sigma_y(epsilon_dot) = sigma_0
In this relationship, sigma_0 represents the static yield strength, epsilon_dot is the equivalent plastic strain rate, while C and P represent material rate-sensitivity constants calibrated via high-speed tensile and shear experiments. Shear components require distinct rate parameters due to the pronounced velocity sensitivity of inter-fibre bond slippage. Neglecting the directional dependence of rate sensitivity parameters introduces severe numerical errors in localized shear strain predictions.

Failure Envelopes in Multi-Axial Stress Fields
Multi-axial stress field calculations in the creasing zone require coupled strain-energy density criteria. The combined stress state includes high out-of-plane compression under the creasing rule nose coupled with intense transverse shear along the channel shoulders. Fracture initiation occurs when the quadratic interaction of shear stress and transverse compression satisfies the modified Hoffman criterion:
C_11 (sigma_11 – sigma_22)^2 + C_22 (sigma_22 – sigma_33)^2 + C_33 (sigma_33 – sigma_11)^2 + C_44 tau_12^2 + C_55 tau_23^2 + C_66 tau_31^2 = 1
Material constants C_ij change as functions of accumulated damage. As micro-cracking progresses inside the middle mechanical furnish, the shear stiffness terms C_55 and C_66 undergo exponential degradation. Mill technical datasheets frequently cite static bursting strength to claim high-speed creasing suitability, dismissing high strain rate score cracking as an operational tooling alignment issue on the customer converting floor.

Discretization
Numerical implementation of high strain rate creasing in FEA codes like Abaqus Explicit or LS-DYNA demands rigorous dynamic discretization strategies. Implicit solvers struggle with the severe geometric non-linearities, rapid contact state changes, and localized material instabilities characteristic of high-speed scoring. Explicit time integration solves these challenges by bypassing global stiffness matrix inversion, calculating nodal accelerations directly from internal and external force vectors.
Explicit integration maintains stability. However, explicit schemes are conditionally stable, governed by the Courant-Friedrichs-Lewy (CFL) condition, which restricts the integration time step based on the smallest element dimension and the dilatational wave speed of the paperboard matrix.
Element selection governs numerical accuracy and stability. Fully integrated solid continuum elements suffer from shear locking under severe bending, artificially stiffening the score profile. Reduced integration three-dimensional solid elements mitigate locking but require active hourglass control parameters to prevent non-physical zero-energy deformation modes.
Solid-shell elements or specialized multi-layered continuum shell formulations provide superior performance by calculating out-of-plane strain gradients across multiple integration points through the thickness while avoiding the element size restrictions of pure 3D solid meshes. Element size distorts damage evolution.
| Element Type and Formulation | Through-Thickness Element Layers | Characteristic Mesh Size | Stable Time Step | Computation Time Relative Factor | Shear Band Resolution Accuracy |
|---|---|---|---|---|---|
| C3D8R (3D Solid Reduced Integration) | 3 | 0.150 | 4.2e-8 | 1.0x | Poor (Hourglass Modes Present) |
| C3D8R with Enhanced Hourglass Control | 6 | 0.050 | 1.4e-8 | 7.8x | Moderate (Slight Shear Over-stiffening) |
| SC8R (Continuum Shell) | 6 | 0.050 | 1.6e-8 | 5.2x | Good (Accurate Bending Hinge) |
| C3D8I (Incompatible Modes Solid) | 12 | 0.025 | 6.8e-9 | 28.4x | Excellent (Captures Localized Micro-cracks) |
| COH3D8 (Cohesive Elements Interleaved) | 6 Solid + 5 Cohesive | 0.035 | 9.1e-9 | 18.6x | Superior (Explicit Interfacial Delamination) |
Model spatial resolution requires adaptive or highly refined fixed meshes in the contact region beneath the creasing rule. Element dimensions through the thickness must not exceed 0.03 millimeters to capture steep shear strain gradients across individual ply interfaces. Mass scaling must be applied with extreme caution; while scaling mass increases the stable time step, artificial inertial forces distort the true high-rate dynamic response, masking brittle shear fracture mechanisms.
- Spurious hourglassing modes emerge when using reduced integration elements with insufficient artificial damping, distorting local strain calculations along channel shoulders.
- Element deletion instability occurs when constitutive models instantly drop element stress to zero upon reaching critical damage, launching high-frequency numerical shockwaves across the mesh.
- Contact penalty stiffness spikes generate sudden drops in the global time step when rigid creasing rule geometries contact soft top-coat element surfaces.
- Inertial distortion from aggressive mass scaling alters the physical ratio of kinetic to internal energy, invalidating dynamic fracture predictions.
When purchasing paperboard grades destined for high-speed automated packaging lines, technical purchasing agreements must mandate compliance with ISO 187 conditioning protocols prior to lab validation, establishing that property deviations outside specified dynamic shear tolerances void supplier performance guarantees.

Calibration
Accurate FEA predictions depend entirely on high-fidelity material characterization across equivalent strain rate regimes. Standard tensile testing per ISO 1924-2 operates at strain rates near 10-3 s-1, yielding strength figures that overestimate high-speed deformability. Extracting true dynamic shear and tensile properties requires specialized test rigs, including servo-hydraulic high-speed tensile machines, split Hopkinson pressure bars modified for low-impedance soft solids, and custom double-shear dynamic test fixtures.
High strain rates alter bond compliance.
Characterizing cohesive zone parameters for Mode II shear delamination involves dynamic End-Notched Flexure (ENF) testing. Specimen geometry consists of laminated paperboard strips with pre-cracked interfaces loaded at displacement velocities matching converting press line speeds. High-speed digital camera arrays record crack tip propagation at capture rates exceeding 500,000 frames per second.
Data processing uses optical flow algorithms to measure displacement fields directly, enabling calculation of dynamic fracture toughness values without relying on compliance calibration approximations designed for rigid metals.
Dynamic fracture toughness of middle-ply mechanical pulp drops by up to 40 percent when strain rates escalate from static to 10^3 s^-1.
Tool geometry sets local strain. Calibrating rate-sensitivity parameters requires multi-axial optimization routines that match FEA force-displacement curves against experimental creasing force profiles. Direct measurement of local surface strain fields during physical scoring validation relies on ultra-high-speed stereoscopic Digital Image Correlation (DIC).

Can Digital Image Correlation Resolve Intralaminar Shear Bands?
Resolving localized shear bands within an individual 50-micrometer paperboard ply requires optical magnification systems paired with high-frame-rate sensors. Standard speckle patterns applied via spray paint add unwanted stiffness to thin paperboard plies, invalidating strain measurements. Advanced optical setups utilize high-resolution piezo-actuated airbrushes to deposit sub-micron carbon particle patterns directly onto the polished cross-section edge of the board.
Stereoscopic camera arrangements operating with long-working-distance microscope optics achieve spatial resolutions of 2.1 micrometers per pixel. This resolution allows direct mapping of out-of-plane shear strain distributions across individual ply boundaries during microsecond creasing strikes, exposing micro-delamination events before macro-cracks appear on the board surface.
- Mount the cut paperboard sample in a rigid high-speed micro-creasing test bench aligned perpendicular to dual synchronized high-speed cameras.
- Apply a high-contrast sub-micron speckle pattern to the specimen cross-section using an atomized carbon suspension.
- Set illumination using high-intensity pulsed LED arrays to eliminate motion blur at camera exposure times below one microsecond.
- Execute the creasing strike at a controlled velocity of 5.0 meters per second while triggering camera acquisition at 1,000,000 frames per second.
- Export cross-correlation strain fields to extract local Mode II displacement jumps across internal ply interfaces.
- Import displacement field data into optimization software to fit non-linear cohesive zone traction-separation laws.
Calibrating material models against low-speed lab bench tests guarantees failure when production lines accelerate to operational velocity.

Headroom
Translating numerical predictions into converting success establishes the physical headroom of a specific board grade on high-speed packaging machinery. Creasing headroom defines the allowable range of creasing rule penetration depths and matrix channel widths that produce an acceptable fold hinge without inducing surface crack failure or internal ply crush. High strain rate FEA modeling reveals that as machine line speed increases, the usable converting window shrinks.
A crease geometry that performs cleanly at 150 meters per minute often causes total score line fracture at 500 meters per minute under identical depth settings due to rate-induced embrittlement of the furnish.
Optimizing creasing headroom through modeling reduces expensive trial-and-error tooling changes on the factory floor. Simulation allows packaging engineers to evaluate alternative rule profiles, such as rounded rule radii or asymmetric channel geometry, prior to cutting physical steel counter-plates. Furthermore, modeling quantifies the exact structural trade-offs between score crack risk and subsequent carton folding resistance.
A shallow crease prevents coating fracture but increases the moment required to fold the blank, causing carton bulged panels during high-speed filling operations.
| Operational Parameter | Conventional Recycled Board (CRB) – Un-optimized Crease | Premium Solid Bleached Board (SBB) – Standard Setup | FEA-Optimized Recycled Board (CRB) – Tailored Tooling |
|---|---|---|---|
| Basis Weight / Grammage | 350 | 350 | 350 |
| Line Speed | 450 | 600 | 600 |
| Peak Strain Rate inside Crease | 1200 | 1600 | 1600 |
| FEA Predicted Shear Damage Index | 0.88 (High Risk) | 0.32 (Low Risk) | 0.45 (Acceptable) |
| Observed Score Crack Defect Rate | 4.25% | 0.15% | 0.35% |
| Delivered Tonnes Processed | 50.0 | 50.0 | 50.0 |
| Substrate Landed Cost per Tonne | 1,150 | 1,850 | 1,150 |
| Total Material Landed Outlay | 57,500 | 92,500 | 57,500 |
| Scrap Loss Value | 2,443 | 138 | 201 |
| Tooling Adjustment Investment | 0 | 0 | 2,500 |
| Net Effective Cost per 1000 Blank Yield | 42.02 | 64.88 | 42.18 |
Consider a practical converting operation processing 50 tonnes of 350 gsm paperboard into pharmaceutical folding cartons. Running standard Coated Recycled Board (CRB) at high production speeds without optimizing rule geometries produces a score crack defect rate of 4.25 percent due to dynamic shear failure in the weak recycled middle ply. Substituting virgin Solid Bleached Board (SBB) eliminates defects due to superior fibre length and tear resistance, but inflates material expenditure by over 35,000 USD for the job.
Utilizing FEA to redesign the creasing matrix profile specifically for the dynamic shear limits of CRB allows line speeds of 600 blanks per minute while keeping scrap losses under 0.4 percent. The material savings dwarf the computational modeling investment after a single production run.
Optimizing matrix channel width via dynamic shear modeling reclaims up to 35 percent of lost converting headroom on recycled board grades.
Decision criteria for stock selection must balance sheet physics against finished pack yield requirements. When moving to high-speed converting lines, procurement decisions guided solely by static material spec sheets invite severe operational losses.
- Verify dynamic shear compliance by requiring mills to supply strain-rate sensitivity parameters alongside standard burst and tensile certificates.
- Recalibrate matrix channel geometries for lower-grade recycled stocks to offset rate-induced delamination resistance drops.
- Audit warehouse relative humidity to ensure delivered pallets maintain target moisture content before entering high-speed converting lines.
- Implement FEA-driven score design prior to ordering custom cutting and creasing tooling for major packaging production runs.
Does the current constitutive framework accurately capture the coupling between microsecond shear damage accumulation and long-term springback angle relaxation in folded cartons?

Boundaries
While predictive FEA modeling under high strain rate creasing conditions delivers massive operational advantages, current computational frameworks possess defined physical limits. Moisture gradients across the paperboard thickness represent a persistent challenge in numerical accuracy. Paperboard absorbs and desorbs atmospheric moisture rapidly.
A sheet entering the creasing station with a dry top coat and a high-moisture core exhibits radical variations in local strain rate sensitivity that standard homogeneous-ply continuum models fail to capture. Moisture softens the starch matrix. Ambient temperature changes on the press floor similarly alter the viscoelastic relaxation spectrum of synthetic binder resins present in mineral coatings.
Non-linear viscoelastic-viscoplastic multi-ply formulations require vast computational resources. Simulating a complete package blank containing dozens of intersecting score lines at full high-resolution element density remains impractical for routine packaging design workflows. Engineers must apply sub-modeling techniques, isolating individual high-risk score intersections for refined dynamic explicit mesh analysis while treating the surrounding sheet as simplified structural shell elements.
Pallet conditioning prevents score cracking. Boundary condition approximations at the edges of sub-models can introduce fictitious stress reflections if non-reflecting wave boundaries are improperly defined in the explicit solver.
Future developments in computational paper science target multi-scale simulation techniques, directly linking discrete fibre network models to continuum macro-scale FEA solvers. Discrete Element Modeling (DEM) of individual wood fibres, hemicellulose matrix bonds, and mineral filler particles provides unmatched physical insight into microsecond shear band development. Coupling DEM micro-models with explicit macro-scale FEA allows direct simulation of fibre pull-out, inter-fibre bond fracture, and coating layer delamination without relying on phenomenological damage parameters.
As high-speed computing power expands, integrating multi-scale physical models directly into automated packaging procurement workflows will establish new standards for substrate specification, eliminating the gap between mill-certified material swatches and high-speed converting reality.


