Matrix-Based Parametric Modelling of the Drafting Zone
Authors/Creators
Description
The drafting zone of a cotton ring spinning machine is governed by a complex set of interacting parameters — roller surface velocities (Vᵢ), fibre counts at nip cross-sections (nᵢ), normal forces (Fᵢ), roller radii (rᵢ, bᵢ), and Shore A hardness values (sᵢ) — whose interdependencies are presently managed through costly trial-and-error experimentation. This paper proposes a unified matrix-based modelling framework that formally represents all six parameter classes across a standard 3-over-3 double-apron drafting system. The framework comprises eight interconnected matrices: a 6×3 system state matrix, a velocity–fibre continuity matrix, a Hertzian contact compliance matrix, a hardness–geometry coupling matrix, a stochastic fibre migration matrix, a friction coefficient matrix, and a 6×6 master influence matrix. Published industry values for 100% cotton carded ring-spun yarn (30 Ne) are used to populate each matrix, sourced from peer-reviewed studies including ScienceDirect, Bagwan et al. (2016), Islam et al. (2020), Siddiqui & Yu (2015), and Textile & Leather Review (2024). Analysis of the master coupling matrix reveals that break draft velocity is the single most influential parameter (relative contribution 35.58% to CVm%), that nip force is not independently controllable but is geometrically determined via the Hertz product E_eff × E_eff, and that the observed stiffness gradient (83°–75°–65° Shore A from back to front) encodes a deliberate contact-mechanics taper. The framework is shown to be directly translatable into a state-space model suitable for digital twin construction, inverse optimisation, and physics-constrained machine learning. Significant research gaps identified include stochastic off-diagonal force coupling, micronaire-dependent friction coefficients, and roller-flexure nonlinearities in the hardness–geometry matrix.
Files
Matrix-Based_Parametric_Modelling_Enhanced_v2.pdf
Files
(600.1 kB)
| Name | Size | Download all |
|---|---|---|
|
md5:3fe604450add17c74649fc9ba3c6d89c
|
600.1 kB | Preview Download |