A delay differential model is developed for an infectious disease in which latent infection, early infectiousness, reporting, unreported symptomatic infection, and removal are represented separately. The incidence term is nonnegative, unreported cases add to rather than subtract from infectious pressure, and a latent accounting state preserves population balance across the fixed delay. Continuous nonnegative histories are specified on the complete delay interval. Positivity, boundedness, the population identity, and the invasion reproduction number are established. A functional-correction procedure is formulated as a coupled mean closure followed by Picard iteration; this avoids treating compartment averages as independent quantities and provides a computable local error criterion. Reproducible synthetic experiments use a population of \(10^{6}\), a four-day latency, and a baseline reproduction number of \(2.20\). Without a transmission change, incidence peaks on day \(100.50\) at \(1881.58\) infections per \(100{,}000\) persons per day, with \(84.35\%\) cumulative infection by day 180. A \(60\%\) transmission reduction beginning on day 35 lowers the day-42 effective reproduction number to \(0.876\), the peak to \(25.93\) per \(100{,}000\) per day, and cumulative infection to \(1.68\%\). Holding the reproduction number fixed while increasing the delay from zero to eight days shifts the peak from day \(51.95\) to day \(140.08\) and reduces its height from \(3334.63\) to \(1332.50\) per \(100{,}000\) per day. On a one-day correction interval, the maximum scaled error decreases from \(1.52\times10^{-1}\) at first order to \(7.92\times10^{-3}\) at second order and \(1.99\times10^{-6}\) at fifth order. These experiments are numerical benchmarks rather than forecasts; disease-specific prediction requires calibrated data and identifiability analysis.
Set-valued operations are useful when one input state admits several chemically distinct outcomes, but an ordinary powerset does not distinguish alternative outcomes from co-products and does not preserve stoichiometric multiplicity. This paper develops a type-consistent framework in which atomic species generate a free commutative multiset monoid, elementary hyperoperations return nonempty sets of product multisets, and \((m,n)\)-superhyperoperations act between explicitly defined iterated nonempty powersets. Domain-specific feasibility predicates impose balance, energetic, kinetic, and selection constraints without changing the common algebraic type. Two general results are established: elementary balance is preserved along every finite admissible path, and the \((0,1)\) construction reduces exactly to the underlying hyperoperation. The framework is specialized to biochemical, geochemical, photochemical, electrochemical, radiochemical, neurochemical, and medicinal-chemical systems. Machine-checkable stoichiometric residuals are reported for representative channels. The quantitative checks include the Daniell-cell value \(E^\circ_{\mathrm{cell}}=1.103\,\mathrm{V}\), the photon energy \(hc/(365\,\mathrm{nm})=3.3968\,\mathrm{eV}\), and evaluated \(^{212}\mathrm{Bi}\) branching fractions of \(64.06\%\) and \(35.94\%\), which sum to \(100.00\%\). These checks validate the internal consistency of the formal examples; they are not presented as new laboratory measurements.
Acute viral conjunctivitis can generate rapidly growing outbreaks in schools and other close-contact settings, but intervention comparisons are meaningful only when the underlying transmission model is dimensionally consistent and its reproduction numbers are reproducible. We formulate a closed-population susceptible–exposed–infectious–recovered (SEIR) model with two distinct controls: a dimensionless reduction in effective contact, \(u_1\), and treatment-supported case management, \(u_2\), which increases removal from infectious contact at a maximum rate \(\tau\). The basic reproduction number is \(R_0=\beta/\gamma\), the controlled invasion number is \(R_c=\beta(1-u_1)/(\gamma+\tau u_2)\), and the state-dependent effective reproduction number is \(R_e(t)=R_cS(t)/N\). These expressions satisfy the required identity \(R_c(0,0)=R_0\) and yield an explicit control threshold. The baseline scenario is calibrated to the published estimate \(R_0=2.64\) for acute haemorrhagic conjunctivitis, with a 1.5-day latent period and a 5-day infectious period. Contact reduction alone crosses the invasion threshold at \(u_1=0.621\), whereas the assumed case-management range does not do so alone. A combined scenario \((u_1,u_2)=(0.50,0.80)\) gives \(R_c=0.660\) and limits new infections to \(0.289\%\) of initially susceptible individuals in a 10,000-person illustrative cohort, compared with \(90.938\%\) without control. A 10,000-draw uncertainty analysis gives a \(93.9\%\) probability that the combined scenario maintains \(R_c<1\). Independent numerical solvers agree to within \(4.4\times10^{-10}\) persons, and population is conserved to within \(7.7\times10^{-11}\) persons. The results support combined intervention thresholds as transparent scenario analyses; they do not constitute a clinical ranking of treatments or a population-specific forecast without local calibration.
This work develops a reduced-order model for the transverse deformation of a skeletal-muscle fascicle during an imposed contraction cycle. The muscle is represented as a tensioned fiber coupled to a Kelvin–Voigt foundation, so that elastic storage and rate-dependent dissipation are included explicitly. A through-thickness displacement ansatz is introduced without the coordinate singularity present in earlier formulations, and compatible fixed-end and quiescent initial conditions are imposed. Spatial variation of fiber tension is treated by a bounded perturbation expansion. Projection onto the fixed-end eigenfunctions yields a hierarchy of damped modal equations whose solutions are written with the correct mode-dependent Green function and time convolution. The second-order approximation is evaluated for a reproducible dimensionless test problem and compared with an independent conservative finite-difference solution advanced by the average-acceleration Newmark method. For a tension-heterogeneity amplitude of \(0.20\), the global relative root-mean-square difference is \(0.441\%\), the maximum time-resolved relative \(L^2\) difference after the initial loading interval is \(3.16\%\), and the maximum midpoint displacement difference is \(8.04\times10^{-4}\). Grid refinement from 201 to 401 spatial points changes the final finite-difference solution by \(6.49\times10^{-6}\) in relative \(L^2\) norm. The formulation is intended as a transparent benchmark and rapid reduced-order approximation; it does not replace three-dimensional active-muscle models or constitute clinical validation.
