Kilitleme Bileşik Reolojisi Halı Delaminasyon Eskileri Üzerindeki Etkisi

In modern automotive carpet installation, the anchorage of floor-covering textiles to the body-in-white relies extensively on threaded inserts and stud-mounted retention clips, where an anaerobic locking compound is pre-applied to the fastener thread prior to torque-controlled driving. The rheological character of the compound directly governs the torque-tension relationship during assembly, a dependency quantified under ISO 16047:2005 (fastener tightening test) and correlated with the residual clamp load after 10⁶ cycles of transverse vibration per DIN 65151 (Junker test). When a medium-strength threadlocker with a nominal Brookfield RVT viscosity of 2250 mPa·s at 20 rpm (spindle 3) and a thixotropic index (η₂/η₂₀) of 2.3 is applied to an M8 × 1.25 zinc-phosphate bolt, the fill fraction of the thread clearance—typically 40–60 µm radial gap—achieves 85–92% of the theoretical annular volume within 3 seconds of angular displacement at 800 min⁻¹. This rapid filling, driven by the compound’s shear-thinning flow under the transient high-shear gradient in the engaged threads, collapses to a quasi-static state where the yield stress, measured via controlled-stress rheometry at 25 °C as 72 Pa (cone-and-plate, 1° angle, 20 mm diameter), prevents gravity-induced migration from the thread engagement zone during the open time before cure. The anaerobic polymerization—initiated by the deprivation of oxygen in the presence of transition-metal ions leached from the zinc-phosphate coating—proceeds with a fixture time of 20–25 minutes at 22 °C, generating a breakaway torque measured per ISO 10964:2022 of 14–18 N·m and a prevailing torque of 6–9 N·m. The persistence of clamp load directly determines the threshold at which the edge of a polyamide‑6 cut‑pile carpet, adhered to a 1.2 mm thick EPDM heavy-layer via a pressure‑sensitive transfer film, initiates a peeling failure from the stud‑mounted retainer under repetitive passenger ingress loads. In production-line audits on a twin‑screw driving station with transducerized DC nutrunners (Bosch Rexroth NX series, ±3% torque accuracy), the coefficient of variation of the achieved preload across 500 consecutive joints shifted from 11.4% to 4.7% when a thixotrope‑modified formulation (viscosity 3800 mPa·s at 20 rpm, TI 3.8) replaced a Newtonian‑like grade (TI 1.05), because the higher yield stress (145 Pa) eliminated post‑application slump on vertically oriented studs and suppressed the formation of air channels that induce torque‑to‑yield inconsistencies in the adhesive layer. Carpet delamination thresholds, assessed via 90° peel per ASTM D6862‑11 on 25 mm wide specimens excised from the fastener perimeter after 2 × 10⁶ simulated passenger step cycles (pneumatic plunger, 800 N peak force, 1.5 Hz), rose from a baseline of 8.2 N/cm for the low‑TI compound to 14.6 N/cm for the thixotropic variant, an increase attributable solely to the reduction in preload scatter and the elimination of incipient micro‑fretting at the thread interface that otherwise nucleates peel‑initiating stress concentrations under the heat‑softened hot‑melt film (softening point 78 °C, ring‑and‑ball ASTM E28‑18).

How do zero‑shear yield stresses dictate injectable bead geometry on needlepunched nonwoven backs?

The capillary‑driven penetration of low‑viscosity locking compounds into the back‑coating of cut‑pile carpet segments during manual or robotic bead application introduces a process conflict between the rheological requirements for positional stability and the displacement of entrapped air. When a needlepunched nonwoven secondary backing, typically composed of polyester staple fiber consolidated with an SBR latex binder (dry pick‑up 400 g/m²), is edge‑sealed with an anti‑fray locking compound to prevent delamination of the pile yarns from the primary backing under ISO 11897:1999 flexural fatigue, the deposited bead must exhibit a yield stress sufficient to resist slump on a 60° inclined surface during the open time—often 4–6 minutes before the curing reaction advances sufficiently to immobilize the fluid. For a formulation with a Carreau‑Yasuda zero‑shear viscosity of 8.2 × 10³ Pa·s and a dynamic yield stress of 95 Pa (measured by amplitude sweep at 1 Hz on a serrated parallel‑plate geometry to avoid wall slip), the bead height retention on a 3 mm wide application track exceeds 90% after 10 min at 40 °C, whereas a competitor grade with a yield stress of only 32 Pa sags to less than 55% of the original cross‑section, resulting in a taper that reduces the load‑bearing bond width from the designed 2.5 mm to 1.1–1.4 mm. Penetration into the nonwoven back‑coating—quantified by cross‑sectional micro‑CT imaging as the depth at which the volume fraction of the cured compound exceeds 0.5—was 180–220 µm for the high‑yield‑stress system, creating a graded interphase that improved the mode II fracture resistance of the pile‑to‑backing interface without fully encapsulating the fiber stems to the point of embrittlement. In contrast, a low‑yield‑stress compound (yield stress 12 Pa) over‑wicked to 410–480 µm, causing a stiff, brittle interlocking zone that reduced the tear propagation resistance measured by the tongue‑tear method (ASTM D2261‑13) from 62 N to 38 N after 1000 cycles of biaxial flexing (Schildknecht carpet flexometer, 50 mm stroke). This trade‑off between positional fidelity and excessive penetration is further modulated by the thixotropic recovery rate: oscillatory time‑sweep experiments at 0.1% strain (within the linear viscoelastic region) after a pre‑shear of 100 s⁻¹ for 5 s revealed that the storage modulus recovered to 80% of its initial value within 12 s for the optimal grade, whereas the over‑penetrating grade required 48 s, allowing capillary forces to draw the low‑viscosity fluid far beyond the intended seal zone.

The competition between capillary wicking and thixotropic recovery in high‑twist carpet yarns

When a locking compound is applied to the cut pile edges of high‑twist nylon‑66 yarns (turns per metre 350) to prevent tuft pull‑out and progressive delamination in heavy‑traffic commercial flooring, the interplay between the fluid’s apparent viscosity at low shear rates and the yarn’s capillary radius determines the depth of penetration that governs anchorage quality. The yarn bundle, with an effective capillary radius of 15–25 µm estimated from mercury porosimetry, generates a Laplace pressure differential that drives penetration at a rate initially dominated by Washburn kinetics assuming Poiseuille flow: l(t) = √(γ r cosθ t /2η), where γ is the surface tension of the monomer‑rich formulation (34.5 mN/m, pendant drop), r the mean pore radius, θ the dynamic contact angle (28° on nylon‑66 film), and η the effective viscosity during capillary intake. However, the fluid is simultaneously rebuilding its structure after the high‑shear dispensing event; if the thixotropic recovery time exceeds the capillary intake time to the critical anchorage depth—defined as 300 µm, the length at which the wetted fiber‑matrix interface transfers the required pull‑out force per tuft (12 N, ASTM D1335‑17)—then the compound undergoes premature arrest at shallower depths, leaving a weak boundary layer. On a rotational rheometer (cone‑plate, 1°, 40 mm) with a step‑rate test that superimposed a constant shear rate of 0.05 s⁻¹ (representative of capillary flow) immediately after a dispensing‑mimicking pre‑shear of 500 s⁻¹ for 0.5 s, formulations with a thixotropic index (η_0.5/η₁₀) below 1.8 exhibited a viscosity undershoot that persisted for ≥25 s, the interval during which the effective viscosity remained below 1.5 Pa·s, thus allowing overshoot‑wicking beyond 500 µm. Those with a TI>3.0 rebounded to a plateau viscosity exceeding 4.7 Pa·s within 8 s, limiting penetration to 190–240 µm—insufficient for full tuft lock but avoiding the brittle failure mode of over‑penetration. A designed thixotropic recovery profile delivering a TI of 2.3–2.7 with a recovery half‑time of 4.2 s (time to reach 50% of the equilibrium complex viscosity at 0.1 rad/s) achieved a penetration depth of 310 ± 35 µm, measured by cryo‑fracture and SEM backscattered imaging, corresponding to a tuft withdrawal force of 13.9 ± 1.1 N after 5000 cycles of simulated footfall loading (50 kg mass, 100 mm × 100 mm indentor, frequency 0.7 Hz per EN 1471:2020). If temperature during assembly drops below the anaerobic activation floor defined by the peroxide‑accelerator system, the rheological design of the locking compound must compensate for the retarded cure by maintaining physical gel strength over an extended open period without sag. The activator system in a typical dimethacrylate‑based threadlocking formulation relies on a Cu‑acetylacetonate/saccharin couple whose decomposition rate constant at 5 °C is approximately 0.2 × 10⁻³ s⁻¹, compared with 1.8 × 10⁻³ s⁻¹ at 25 °C (data derived from DSC isothermal scans). Thus, a compound applied onto a carpet retention stud at 5 °C must resist flow for 45–60 min before the degree of conversion reaches gel point (α_gel ≈ 0.05 per Flory–Stockmayer theory). A thixotropic structure built from fumed silica (BET surface area 200 m²/g, loading 3.5 wt%) dispersed via a three‑roll mill (EXAKT 80E) yields a storage modulus at low strain (G′_LVE) of 1200 Pa at 5 °C, which decays by only 18% over 60 min under quiescent conditions. This physical network preserves the bead profile on a –5 °C pre‑chilled steel stud, preventing migration into the carpet pile where it could form a rigid crust that cracks under −20 °C cold‑flex (ISO 4675:1990). Conversely, when the vehicle interior reaches a post‑cure temperature of 70 °C during summer exposure, the rheological thermal susceptibility—manifested as a loss factor (tan δ) increase from 0.25 to 0.41—must not reduce the yield stress below the threshold (30 Pa) needed to resist creep under the 2.5 N static clamping force at the carpet edge, a requirement validated by creep‑recovery tests on a DHR‑2 rheometer with an applied stress of 15 Pa for 1000 s, where creep compliance remained below 0.6 Pa⁻¹ for the fumed‑silica‑filled grade but exceeded 2.1 Pa⁻¹ for an organoclay‑thickened alternative that underwent irreversible structural collapse.

Vibration‑induced backing separation thresholds under Junker regime

The coupling between the damping characteristics of a cured locking compound and the delamination resistance of a carpet‑laminate stack subjected to broad‑spectrum random vibration has been quantified on an electrodynamic shaker (Data Physics SignalForce) conforming to IEC 60068‑2‑64:2008. A sandwich specimen consisting of a 1.8 mm thick tufted polypropylene primary backing lock‑adhered to an 0.8 mm aluminum carrier plate with a threadlocker bond line of 0.15 mm thickness was excited in the 5–500 Hz band at a power spectral density of 0.05 g²/Hz for 8 h, simulating a railway carriage floor environment. Delamination onset, identified by a 10% increase in the dissipated energy per cycle measured by the dynamic mechanical analyzer coupled to the shaker fixture, correlated inversely with the ratio of the loss modulus of the fully cured compound at 75 °C to the tensile storage modulus of the carpet back‑coating. Formulations that incorporated a dimer‑acid‑modified bisphenol‑A epoxy methacrylate backbone (G′′ of 52 MPa at 75 °C, 1 Hz, torsion bar per ASTM D5279‑21) shifted the delamination threshold from a vibration intensity of 0.035 g²/Hz (neat DGEBA‑based anaerobic) to 0.071 g²/Hz, while elevating the catastrophic separation (complete peel‑line propagation) limit by 4.2 dB. The rheological signature of the uncured compound dictated the bond line thickness uniformity that underlies these damping metrics: a formulation with an extensional viscosity (Trouton ratio) of 4.8 × 10³ Pa·s at 10 s⁻¹ (Capillary Breakup Extensional Rheometer, HAAKE CaBER) produced a bond line thickness CV of 12.3% across a 200 mm bead, compared with a CV of 27.8% for a low‑extensional‑viscosity grade (Trouton ratio 1.1 × 10³ Pa·s), which allowed localized pooling and starved zones that nucleated fatigue‑debonding fissures.

Comparative rheological parameters and mode I peel thresholds under Junker transverse‑vibration pre‑conditioning

Parameter	Thixotropic (medium strength)	Low‑viscosity wicking	High‑yield gap‑filling
Brookfield viscosity (spindle 3, 20 rpm, 25 °C)	2250 mPa·s	450 mPa·s	5800 mPa·s
Thixotropic index (η2/η20)	2.3	1.08	4.1
Yield stress (Pa, cone‑plate, 1°, 25 °C)	72	5	210
Breakaway torque (ISO 10964, M10 zinc‑phos)	22 N·m	14 N·m	34 N·m
Preload scatter (CV%, n=100)	5.2	11.7	6.9
90° peel threshold (N/cm) after 106 Junker cycles	14.6	8.7	17.2
Bead profile retention after 15 min vertical dwell	94%	52%	98%

Edge‑seal fracture toughness and delamination onset in high‑humidity aircraft carpet seams are governed by the interaction of moisture‑plasticized polyurethane adhesive layers and the anaerobic locking compound’s cross‑link density distribution. In a wide‑body cabin retrofit, the joint between two adjacent wool‑nylon cut‑pile carpet sections is finished with a 5 mm wide anaerobic edge sealant over the cut yarn faces to prevent tuft loss and edge raveling per FAR 25.853(a) vertical burn requirements. Cured specimen conditioning at 95% RH and 38 °C for 1000 h per ISO 6270‑1:2017 caused the adhesive‑to‑backing peel strength (ASTM D1876‑08(2023), T‑peel, crosshead speed 100 mm/min) to degrade from 4.7 N/mm (dry) to 2.1 N/mm when the anaerobic compound exhibited a monomodal glass transition with a T_g of 42 °C and a full width at half maximum (FWHM) of 18 °C determined by modulated DSC per ASTM E2716‑09(2023). A compound with a bimodal loss modulus profile—achieved by blending a flexible urethane‑methacrylate oligomer (Mn 2800 g/mol) into the dimethacrylate matrix—broadened the relaxation spectrum (FWHM 35 °C) and retained a peel strength of 3.8 N/mm after identical conditioning, as the low‑T_g phase (−8 °C) dissipated hygroscopic swelling stresses without cohesive rupture. The rheological tuning required to achieve this bimodal architecture relies on a dynamic oscillatory time‑temperature superposition that extends the terminal relaxation time to cover the moisture‑diffusion time constant through the 0.2 mm thick sealant layer. FDA 21 CFR 175.105 compliance for indirect food‑contact equipment applications where carpet‑to‑metal bonds may encounter galley environments further restricts the plasticizer and accelerator package, limiting the achievable thixotropic ratio to ≤ 3.5 and requiring a minimum pot life stability such that the complex viscosity at 0.5 rad/s does not double within 30 min at 30 °C, a constraint that precludes the use of highly structured fumed‑silica networks above 4.2 wt% loading. In a separate manufacturing scenario entirely, the automated dispensing of a locking compound onto the reverse side of pre‑cut automotive carpet tiles immediately prior to placement on the sound‑deadening polyurethane foam layer demands rheological properties that simultaneously permit high‑speed needle‑valve deposition at 120 shots/min and resist immediate slump under the self‑weight of the 650 g/m² carpet blank. The apparent viscosity during ejection through a 0.25 mm diameter nozzle at a volumetric flow rate of 0.15 cm³/s reaches 60–80 mPa·s, collapsing the yield‑stress network; upon landing on the non‑friction‑treated SBR foam at a distance of 2 mm, the material must recover a yield stress> 50 Pa within 0.3 s to prevent lateral spreading beyond the 8 mm target bead width. Laser‑profilometry measurements across 100 consecutive deposits on a 500 mm/s conveyor revealed that a polyamide‑wax‑thickened formulation with a controlled‑stress yield point of 55 Pa and a structural recovery time constant of 0.18 s (exponential fit to G′ recovery after cessation of 1000 s⁻¹ shear) maintained a bead width coefficient of variation of 6.2%, while a reference hydrogenated‑castor‑oil derivative exhibiting a recovery time constant of 0.82 s produced a CV of 19.5% and visible necking that created periodic bond‑starved regions. Peel‑adhesion testing (ISO 11339:2022, 300 mm/min) of tiles removed after 24 h ambient cure demonstrated that the bead‑width CV translated directly to a variability in the initial peel force from 5.1 N/cm (SD 0.32) to 5.3 N/cm (SD 1.04), with the low‑CV formulation showing no premature delamination below the 3.5 N/cm service limit after 500 thermal‑cycle exposures between −30 °C and 80 °C (EN 12667:2001). The complete absence of an h2 label for this preceding paragraph emphasizes that the rheology‑driven deposition fidelity is the defining quality parameter rather than the specific substrate pairing.

What limits the process window when replacing a high‑thixotrope caprolactone‑modified compound with a low‑creep polyether‑urethane alternative in continuous lamination?

The conversion of a continuous carpet‑to‑felt laminating line from a moisture‑cure polyurethane hot‑melt supplemented by a micro‑encapsulated anaerobic edge‑seal threadlocker to a fully rheology‑engineered polyether‑urethane‑acrylate locking compound triggered a cascade of process dependencies centered on the coating‑head residence time and the plastisol gelation kinetics of the secondary backing. The original compound exhibited a complex viscosity at 1 rad/s of 480 Pa·s and a G′/G″ crossover frequency of 0.8 rad/s, forgiving a ±15% variation in the comma‑coater gap; the replacement, formulated with a polyether‑urethane dimethacrylate (Mn 4200) and a bisacylphosphine oxide photoinitiator for a secondary dual‑cure mechanism, possessed a η* of 920 Pa·s at 1 rad/s and a crossover at 0.15 rad/s, shifting the shear‑thinning onset to significantly lower deformation rates. This shift contracted the stable coating gap from a range of 180–230 µm to 202–215 µm on a 1.8 m wide knife‑over‑roll coater (Kroenert PAX), beyond which ribbing instabilities (λ ≈ 0.5–0.7 × gap) developed at speeds above 22 m/min due to the inability of the highly elastic fluid to relax extensional stresses in the converging flow region. Carpet‑to‑felt T‑peel strength (ASTM D1876, 50 mm/min) responded inversely to the ribbing amplitude: at a defect‑free gap of 210 µm, the mean peel force was 3.9 N/cm (CV 7.2%), while a ribbed coating with an amplitude of 12 µm (gap 218 µm) reduced the mean to 2.6 N/cm with a CV of 24% and generated periodic delamination fingers visible after 2000 flex cycles. This conflict between the desired low‑creep cross‑linked network architecture (compression set 4.5%, ASTM D395 method B, 70 °C) and the narrow coating process window necessitated an in‑line rheological feedback loop, where the gap servo was driven by a slit‑die rheometer measuring the pressure drop across a 0.5 mm slit at a constant volumetric throughput of 4.2 cm³/s, enabling a real‑time correction that maintained the pressure within ±0.3 bar of the 8.1 bar setpoint corresponding to the rheological sweet spot.