Hand-considered hypotheses, 3901–4000

Packaging and barrier materials, written one at a time. Permeation and its measurement, the seal as a mechanical and a gas-tight joint, modified atmosphere as a balance of respiration and transmission, migration and scalping, the mechanics of protection in transit, and the physics that sets a shelf life. Each entry carries the logical form chosen for it, printed above the claim, so the variety of argument can be audited against the text rather than taken on trust. None has been tested; no citation is given because none is being invoked, and every quantity offered as an estimate is marked as one.

Permeation through the film

Misattribution
H3901. Apparent oxygen permeability rises as a solution-cast film is made thinner, which is usually charged to pinholes, but a skin of different free volume at the casting surface gives the same slope because its fixed thickness becomes a larger fraction of the whole. The two are separable by a thickness series, since a skin adds a fixed resistance that the series extrapolates to at zero thickness, while pinholes add a conductance proportional to their density and leave no such intercept.
Hidden term
H3902. Every permeability quoted for a hydrophilic polymer at a nominal humidity contains a plasticisation term, because the water sorbed in the amorphous phase lowers the glass transition and raises the diffusivity. The term is not a correction applied afterwards but part of the coefficient, and its size is set by the gap between the measurement humidity and the humidity at which the specimen was dried.
Crossover
H3903. In a biaxially drawn film, raising the draw ratio improves barrier by aligning chains, but past a draw ratio the microvoids nucleating at tie molecules add a Knudsen path, so permeability passes through a minimum rather than falling monotonically. The minimum is where void growth overtakes alignment gain, and a draw-ratio series at fixed temperature shows it as a reversal rather than a plateau.
Identity claim
H3904. The curvature of a glassy polymer's sorption isotherm and the slow tail of its permeation transient are one phenomenon seen from two sides, both being the relaxation of excess free volume toward its equilibrium occupancy. If that identity holds, a single long sorption run on a freshly quenched film should recover the permeation transient's time constant, which is a test no permeation cell can perform on its own.
Ratio invariant
H3905. Across a homologous series of penetrants in one polymer the permeability ratio is carried mainly by the solubility ratio, because the diffusivity ratio collapses onto a function of the penetrant's diameter alone. A selectivity claim that improves as the penetrant grows is therefore a solubility claim in disguise, and the deciding measurement is an isotherm rather than a transmission cell.
Threshold
H3906. A barrier coating cannot compensate for its own defect density: below a certain thickness the transport is set by the pinholes, whose areal density belongs to the substrate's roughness and not to the coating process. The threshold is where defect density times coating permeability equals the substrate's permeability, and thinning below it changes nothing measurable.
Inversion
H3907. Higher crystallinity normally raises barrier, but in a polymer already below its glass transition the crystallites can hold water at their interfaces, and the humidity-driven rise in permeability then becomes steeper as crystallinity increases. The inversion is invisible at moderate humidity and appears when the same specimens are compared at high humidity, so a single-condition comparison can rank two materials backwards.
Scaling argument
H3908. A package's gas ingress scales with its area while its capacity to absorb that gas scales with its volume, so the time to a critical internal concentration scales with the package's smallest dimension. A sachet and a drum made from the same film therefore differ in shelf life by geometry alone, which is why a film validated in one format is not validated in another.
Two-regime claim
H3909. Below the glass transition the activation energy of permeation is a diffusion term plus a sorption term of opposite sign, so the apparent Arrhenius slope can be smaller than the diffusion term taken alone. The two regimes meet at a break that appears in the permeability even when the diffusivity shows no break, and fitting a single slope across it produces an activation energy that belongs to neither.
Artefact
H3910. A large barrier gain credited to a platelet filler is often measured against a film of a different thickness, because the filled film is opaque and its thickness is taken from a different gauge than the clear control's. A matched-thickness comparison should halve the apparent effect, and the residual is the part that tortuosity can actually explain.
Non-monotonic window
H3911. A compatibiliser added to an immiscible blend improves barrier up to a loading and then worsens it, because past interfacial saturation the compatibiliser forms a continuous phase of its own and provides a path. The window's width is set by the interfacial area available, which a single extraction measurement can establish before any transmission test is run.
Symmetry argument
H3912. A film with dissimilar skin layers must show the same steady-state permeability from either face, so any measured asymmetry is a transient and must decay. Running the cell in both directions for long enough that the asymmetry vanishes is therefore not a refinement but the only way to report a number that belongs to the film.
Geometric limit
H3913. The permeability of a filled polymer cannot fall below the filler's own permeability once the filler's volume fraction passes percolation, because the filler becomes the continuous transport path. That limit is not a tortuosity factor and no improvement to the matrix can move it, which is why heavily filled barrier compounds plateau instead of continuing to improve.
Memory effect
H3914. A film exposed to humid air retains an enhanced permeability for hours after being returned to dry conditions, because desorption is slower than sorption and the amorphous phase stays dilated in the meantime. The decay constant of that memory is the film's own diffusion time, so the memory is a thickness measurement that a permeability laboratory performs without noticing.
Conservation audit
H3915. If a laminate's measured permeability is lower than every constituent layer's permeability divided by its thickness fraction, the accounting is impossible under series resistance, since interfaces can only add resistance where they are barriers. The audit is arithmetic and costs nothing, and it catches adhesion claims that would otherwise survive to a product trial.
Interaction term
H3916. A carbon dioxide and nitrogen mixture does not permeate a glassy polymer as the mole-fraction-weighted sum of its pure-gas values, because carbon dioxide plasticises the matrix and the mixture's effective diffusivity rises with its fugacity. The interaction term changes sign in a rubbery polymer, and that sign change is the cleanest evidence that plasticisation rather than competitive sorption is at work.

Transmission measurement and its artefacts

Operator artefact
H3917. In a coulometric oxygen transmission test the sensor consumes oxygen, so at low carrier flow the partial pressure at the sensor is lower than at the film and the apparent flux rises as flow falls. A flow plateau is therefore a precondition for a valid number, and a high-barrier value taken below the plateau is a property of the plumbing.
Boundary layer argument
H3918. The dry side of a water-vapour cup carries a stagnant layer of fixed resistance, so as the specimen's resistance falls the measured rate approaches a constant regardless of thickness. That plateau belongs to the cell rather than to the film, and a thickness series is the only cheap way to see whether a reported rate sits on it.
Perimeter error
H3919. Masking a specimen leaves a sealant edge that wicks moisture laterally, so the exposed area exceeds the aperture and the error grows with the perimeter-to-area ratio. Two mask sizes on one film give the correction directly, and without it a small-aperture result is not comparable with a large-aperture one.
Unit trap
H3920. A transmission rate per unit area per day and a permeability per unit thickness differ by the thickness used, and that thickness is usually nominal rather than measured. A ten per cent thickness error therefore reappears as a ten per cent barrier claim, and the two kinds of number cannot be substituted for one another in a specification.
Invariance claim
H3921. A time-lag cell that does not hold its downstream pressure near zero measures a steady flux reduced by the ratio of downstream to upstream pressure, so a diffusivity taken from that slope without the correction is high by the same ratio while the lag itself is unchanged. The disagreement between a lag-derived and a slope-derived diffusivity is therefore a direct readout of the downstream pressure, and a cell that reports agreement is the one actually at zero.
Transient misread
H3922. A specimen conditioned at one humidity and measured at another carries a transient whose length is the film's diffusion time, so a reading taken early reports the transient and not the steady state. The transient's sign depends on whether the specimen was dried or wetted first, which is why two laboratories can disagree on a well-characterised film.
Selection effect
H3923. Any barrier comparison assembled across many laboratories over-represents films that survived mounting, and the survivors are systematically the thicker and the less brittle specimens. The bias runs in the direction of better barrier for the more fragile material, and a blinded mounting trial with equal rejection rates would expose it.
Calibration error
H3924. At an activation energy typical of polymer permeation a one-kelvin error in cell temperature moves the result by several per cent, which is the same size as many claimed formulation effects. A dummy specimen of known barrier run in the same cell at the same time converts that error from an assumption into a measurement.
Extrapolation error
H3925. A differential-pressure cell measures at an activity far above one atmosphere, where plasticising gases no longer follow Henry's law, so extrapolating to ambient fugacity overstates the permeability and therefore understates the barrier. The deviation is a function of fugacity rather than of pressure as such, and two fugacities are enough to detect it.
Degeneracy
H3926. An oxygen rate and a water-vapour rate measured on one film cannot separate pinholes from intrinsic permeability, because both mechanisms raise both rates in the same direction. A helium leak test breaks the degeneracy, since helium's diffusivity is high enough that a defect path dominates where a polymer path does not.
Hysteresis claim
H3927. The first measurement after a specimen change is biased by the previous specimen's sorbed gas desorbing into the carrier, and the bias is largest when the highest-barrier specimen follows the lowest. The order of measurement is therefore a variable, and randomising it costs less than modelling it.
Order of limits
H3928. Whether a specimen is dried before or after annealing decides its measured permeability, because drying first vitrifies free volume that an anneal would otherwise have collapsed, so the two sequences end at different states. The test is one anneal applied to two halves of a single film with the pretreatments swapped, and any difference is a history that the reported coefficient silently carries.

Seals, closures and integrity

Lag
H3929. The interface temperature in a heat seal lags the jaw's set point by a time set by the jaw's thermal mass and the film's heat capacity, so a short dwell seals at a lower interface temperature than a long one on the same jaw. The lag turns a sharp fusion threshold into a knee in the strength-versus-dwell curve, and the knee moves with film thickness rather than with jaw temperature.
Count
H3930. A seal's leak rate is usually quoted as one number, but there are at least three parallel paths, the interface, the film's own permeability, and the cut edge's wicking, and they scale differently with pressure. A single-pressure test cannot apportion them, while a pressure series can, because only the interface path is viscous and only the edge path is humidity-sensitive.
Non-monotonic window
H3931. Hot-tack strength passes through a maximum with seal temperature, since below fusion the interface is weak and above it the melt's cohesive strength is low. The maximum's position depends on the jaw's opening speed, so a hot-tack ranking taken at one speed does not transfer to a machine that opens at another.
Geometric limit
H3932. A contaminated seal fails through a channel whose conductance scales with the cube of the gap height in the viscous regime, so a particle that doubles the gap raises the leak by nearly an order of magnitude. Leak-test sensitivity is therefore not a linear measure of seal quality, and a seal that passes a coarse tracer can still hold a channel a finer one would find.
Identity claim
H3933. Peel strength and burst strength rank seals differently by construction, because peel loads the weakest interface line while burst loads the whole perimeter in tension. A seal that is strong in peel and weak in burst has a corner defect, and the disagreement between the two tests is the diagnostic rather than an inconsistency.
Creep claim
H3934. A seal that passes a burst test can fail in creep at a lower stress held for a long time, because the seal layer is viscoelastic and its failure envelope is time-dependent. Burst strength therefore does not rank two sealants the way a warm storage trial does, and the ranking can invert within the range of warehouse temperatures.
Ratio criterion
H3935. Sealing through a contaminant succeeds when the particle is more compliant than the seal layer and thinner than the layer's thickness, because the melt flows around it and the jaw pressure closes the channel. The criterion is a ratio of particle thickness to seal-layer thickness, and it predicts a sharp transition rather than a gradual loss of integrity.
Operator artefact
H3936. A seal bar's thermocouple reads the jaw's metal rather than the interface, and the two differ by the contact resistance between jaw and film, which grows as the jaw's surface roughens with use. A jaw's calibration therefore drifts with its history, and the drift runs toward hotter indicated temperatures for a worn jaw.
Inversion of standard practice
H3937. Raising seal pressure is expected to improve a seal, but past the point where melt is squeezed out of the interface the seal layer thins and the bond weakens. The inversion is most visible with a low-viscosity sealant, and the optimum pressure is then a property of the sealant rather than of the machine.
Memory effect
H3938. A seal's strength keeps rising for hours after the jaws open, as the oriented seal layer relaxes and recrystallises, so a strength measured a minute after sealing understates the seal that ships. The rise is largest when the sealant's melting range sits close to the jaw temperature, and it is why green seals and aged seals rank materials differently.
Parallel-path audit
H3939. A fitment's leak rate is set by the spiral channel of the engaged thread rather than by the gasket's compression, because the channel is continuous and the gasket interrupts it only at its land. Leakage therefore falls with each engaged turn only until the channel's length exceeds the gasket's sealing width, after which it plateaus.
Sufficiency claim
H3940. A dye-penetrant test does not establish seal integrity when the dye's viscosity is high relative to the channel, because the test's sensitivity is governed by a capillary number rather than by the channel's existence. A seal that passes the dye and fails a gas test is not a contradiction, it is two tests measuring different channel sizes.
Symmetry argument
H3941. A pressurised pouch fails first at the seal edge nearer the load's resultant, so an off-centre seal gives a burst pressure that depends on which face is loaded rather than on the seal's average strength. Across a batch the preferred failure location is therefore a record of machine alignment, and turning the pouch over in the fixture should move the failure to the opposite edge.
Threshold
H3942. Ultrasonic sealing delivers energy at the anvil's knurl lands, so bond strength tracks the land area rather than the amplitude. There is a knurl spacing below which the bonds overlap into a continuous seam and above which they remain a spot pattern, and that threshold is a geometric ratio that no power setting can overcome.
Rate dependence
H3943. A seal's peel strength is a viscoelastic quantity, so two sealants can rank one way at a slow test rate and the other way at a fast one. The ranking that predicts field performance is the one at the rate the package actually sees in handling, and the crossover rate is measurable with a single instrument at three speeds.
Negative claim
H3944. A domed induction-seal lid is not evidence of a good bond, because the dome records the headspace's pressure rise during heating and persists after the seal has cooled and even after it has failed. The bond is established by torque and by pressure decay, and either can contradict the dome entirely.

Modified atmosphere and headspace

Model selection
H3945. The residual oxygen after a gas flush is set by the mixing inside the headspace during the flush, not by the flush volume, so a well-mixed dilution model predicts a lower residual than a plug-flow model at the same volume. Flushing at two nozzle positions with the same volume separates the two, and the difference is the mixing term the volume alone cannot supply.
Crossover
H3946. A larger headspace dilutes the initial oxygen but also stores proportionally more of it, so for a given film the time to a critical oxygen level has a minimum at an intermediate headspace. The optimum sits where stored oxygen and ingress are equal, and it moves with the package's surface-to-volume ratio rather than with the film alone.
Stability condition
H3947. An equilibrium oxygen level in a respiring package is stable only if the film's transmission rises faster with concentration than the produce's consumption does, otherwise the equilibrium is unstable and the package swings between anaerobic and aerobic states. The condition is a comparison of slopes at the design point, and it explains why two films with equal permeability there can behave quite differently.
Hidden term
H3948. The carbon dioxide in a package's headspace is not the carbon dioxide in the package, because a large fraction dissolves in the product's water, so headspace measurement underestimates the total by a partition term. That term is set by the product's volume and solubility, and it is why a measured headspace can fall while the total is conserved.
Diffusion front
H3949. An oxygen scavenger's rate is limited by how fast its reaction front advances into its own material, and that front moves as the square root of time, so the scavenger's capacity is only partly used at the end of shelf life. The unused fraction follows from the scavenger's thickness and diffusivity rather than from its chemistry, which is why thicker scavenger parts are inefficient.
Activation lag
H3950. A moisture-activated scavenger in a dry package has an induction period before it consumes oxygen at its rated rate, set by the time water needs to diffuse into the scavenger's matrix. A package sealed dry therefore behaves as if it carried no scavenger for the early part of its life, and the lag is visible as a plateau in the headspace curve.
Scaling argument
H3951. A pinhole's contribution grows relative to the film's permeation as a package shrinks, because the defect's flow is independent of area while the film's is proportional to it. Small packages are therefore leak-dominated even with an excellent film, and the crossover size is set by the defect's conductance rather than by the film's permeability.
Parallel network
H3952. Water condensing on a package's inner surface creates a local region of high water activity and altered gas transmission, so the film behaves as a parallel network of wet and dry areas rather than as a single permeability. The effect concentrates where the film runs coldest, which is usually a seal or a corner, and it is why chilled packs leak first at corners.
Sampling artefact
H3953. A headspace sample drawn through a septum carries the needle's dead volume and the septum's own permeability, so a small sample is biased toward the air that leaked in during insertion. The bias scales inversely with sample volume, which is enough to explain disagreements between laboratories sampling the same pack.
Relaxation time
H3954. A package's headspace reaches its equilibrium composition over hours rather than minutes, because dissolved gas must diffuse out of the product and sorbed gas must equilibrate with the film. A reading taken early reports the filling gas, and the sign of its error depends on whether the product was saturated before sealing.
Coefficient mismatch
H3955. Respiration and permeation have different temperature coefficients, so a package balanced at one temperature becomes anaerobic above it and over-oxygenated below it. The mismatch rather than the absolute permeability is the design variable, and it predicts a safe temperature window narrower than the tolerance of either property.
Perimeter effect
H3956. Gas entering through a seal's edge is proportional to the perimeter while the film's transmission is proportional to the area, so a long narrow package is edge-dominated. The same film in two shapes therefore yields two shelf lives, and the difference is geometric rather than material.
Oversizing
H3957. Pairing a high-barrier film with a scavenger can waste the scavenger, because its rate exceeds the film's ingress and its capacity is spent on the headspace's initial oxygen rather than on leakage. The combination has an optimum in which the scavenger is sized to the ingress, and oversizing buys nothing once the initial oxygen is gone.
Sign reversal
H3958. Raising a package's initial carbon dioxide increases what dissolves in the product and can lower what remains in the headspace, so the headspace reading moves opposite to the total. A mass balance that includes the product reconciles them, and the balance is itself the experiment.

Migration, sorption and scalping

Diffusion limit
H3959. An additive's migration into food is limited by its diffusion through the polymer rather than by its solubility in the food, so the migration curve rises with the square root of time and then plateaus as the surface layer depletes. The plateau's height is set by total additive and partition coefficient, and measuring two thicknesses separates them.
Contact-pressure transfer
H3960. Ink components transferred to a package's inner face during winding are governed by winding tension and residual solvent rather than by ink formulation alone, because the transfer is driven by contact pressure. The same ink on two winders then gives two migration results, and the difference is the pressure history rather than the chemistry.
Solubility matching
H3961. Aroma scalping is largest for compounds whose solubility parameter matches the sealant's, so a polymer chosen as a good moisture barrier can be a poor aroma barrier for the compound that matters. Polymer rankings by scalping therefore do not transfer between aroma compounds, and a single-compound test is not predictive of a mixture.
Vapour-phase path
H3962. Migration from recycled fibre into a dry food can proceed through the gas phase rather than by contact, so its rate is set by the migrant's vapour pressure and the package's air exchange instead of by contact area. The path is identifiable because it changes with headspace volume, while contact migration does not.
Partition anomaly
H3963. The partition coefficient of a migrant between a polar polymer and a fatty food is not the ratio of solubilities, because the food swells the polymer's surface and the swollen layer's solubility differs from the bulk's. The anomaly is largest for migrants comparable in size to the swelling penetrant, and it disappears when the food is pre-saturated with the migrant itself.
Activation mismatch
H3964. Migration's activation energy exceeds permeation's because migration is diffusion in a matrix that the migrant itself plasticises, so an accelerated test at elevated temperature overestimates the room-temperature result by an amount that grows with migrant size. The overestimate is not a constant factor, and two temperatures are the minimum needed to detect it.
Ratio criterion
H3965. A functional barrier works only if its thickness exceeds the migrant's diffusion length over the shelf life, so a layer with low permeability but insufficient thickness fails anyway. The criterion is thickness against the square root of diffusivity times time, and it predicts a sharp failure rather than a gradual leakage as the layer thins.
Spatial confound
H3966. A seal's heat history raises its crystallinity and lowers the diffusivity of migrants there, so the seal migrates less than the wall although it is the same material. The asymmetry is measurable between two samples cut from one package, and it confounds any migration test that does not record where its specimen came from.
Sorption trap
H3967. A laminating adhesive can trap the solvent an ink would otherwise release, so a low-odour ink can produce a higher final headspace concentration than a conventional one when the adhesive's sorption capacity is large. Ink rankings by odour therefore depend on the lamination, and the trap is identifiable by a desorption run on the laminate alone.
Worst case
H3968. The contaminant concentration in recycled content is not its mean but the concentration in the fraction that was printed or otherwise contaminated, so a worst-case estimate built on the mean underestimates by the ratio of the two fractions. Sorting and analysing the fractions separately yields that ratio, and it is the number a design should carry.
Plasticisation coupling
H3969. Water sorbed into a polymer lowers its glass transition and raises every migrant's diffusion coefficient, so a migration test at high humidity can exceed a dry one by more than the humidity's effect on the food. The coupling is the mechanism, and it is why one package migrates differently in a tropical and a temperate market.
Contact efficiency
H3970. Migration into a powder is limited by the contact area between particles and film rather than by the migrant's solubility, so a coarser powder migrates less although its composition is unchanged. The effect is geometric and vanishes when the powder is compacted against the film, which is the test that separates it from a solubility effect.
Reversal
H3971. A thicker package can migrate more total mass while releasing less per unit area, because it holds a larger reservoir that replenishes the depleted surface layer. The two measures therefore rank packages differently, and only one of them is the quantity a specification usually limits.
Count
H3972. A migration result is a sum over at least three resistances, the polymer's diffusion, the interface's partition, and the food's own diffusion, and one test cannot separate them. Two film thicknesses and two stirring rates give four equations, after which the resistances are identifiable without any new instrument.

Mechanical protection

Resonance
H3973. The acceleration a product sees in a drop is set by the product's own internal resonance when the cushion's transmitted pulse carries energy at that frequency, so a product that is rigid on the bench can be fragile in the package. The cushion's pulse width is the design variable, and it is fixed by thickness and stiffness rather than by drop height.
Equivalence claim
H3974. Drop damage is set by the velocity change at impact rather than by height, so drops from different heights onto different surfaces are equivalent when their velocity changes match. The equivalence is exact for a rigid product and approximate for a cushioned one, and it is why the impact surface specification matters as much as the height.
Creep rate
H3975. A stacked corrugated box's creep is governed by the liner's moisture content more than by load duration, because fibre creep rises steeply with sorbed water. A modest humidity rise therefore shortens a stack's safe life more than a comparable temperature rise, and the two are separable in a controlled chamber.
Geometry dependence
H3976. A film's puncture resistance depends on probe geometry as well as on the film, so two films can rank one way with a sharp probe and the other way with a blunt one. Only a probe shaped like the product's corner yields a ranking that transfers to the field, and a standard using a blunter one measures a different property.
Rate dependence
H3977. Tear resistance is rate-dependent because the tear tip's process zone is viscoelastic, so a pendulum test at one speed ranks materials differently from a slow tear. The crossover speed is where the process zone's relaxation time meets the test's time scale, and three speeds locate it.
Load path
H3978. A corner drop loads the package through its stiffest path, so the corner sees a higher acceleration than a face at the same height and cushion. A test that drops only on faces therefore underestimates the demand, and the corner is the design case even when every face passes.
Threshold
H3979. A cushion stops protecting once its thickness is less than the stopping distance for the drop, and beyond that point acceleration is governed by the product's own stiffness. The threshold is a thickness rather than a stress, and it explains why fitting a stiffer cushion can make a package worse.
Stress concentration
H3980. A sealed pouch's burst pressure is set by the stress concentration at the seal's edge, so a stronger film can yield a weaker pouch if the seal's edge geometry is sharper. The concentration factor follows from the seal's thickness ratio, which is why seal design and film selection cannot be optimised apart.
Ratio criterion
H3981. A rigid corner inside a flexible package punctures when the corner radius is small relative to the film's strain at break and thickness, so increasing the radius can beat increasing the thickness. The two are not interchangeable, and the ratio of radius to thickness predicts which of them will help.
Load case
H3982. A vacuum package's film is loaded by the pressure difference and supported by the product, so its stress peaks where the product's surface is convex and unsupported. Leaks therefore begin at a convexity rather than at a seal, which makes package geometry rather than film strength the controlling variable.
Thermal mismatch
H3983. A rigid package's lid deflects in retort because lid and body expand differently, and the deflection concentrates stress at the seal's edge. The mismatch belongs to the material pair, so a lid that passes a pressure test at one temperature can fail at another with the same pressure.
Rate-dependent modulus
H3984. A cushioning foam's modulus is rate-dependent, so a cushion curve from a slow compression test does not predict a drop, and a foam tuned for vibration can be too stiff for impact. The regimes are separated by the foam's relaxation time, which means one material can be right for one and wrong for the other.
Spectral claim
H3985. Random vibration damages a package through cycles accumulated near the product's resonance, so the damage is set by the power spectral density at that frequency rather than by overall RMS. Two spectra with equal RMS can differ in damage by a large factor, and the resonance's width decides by how much.
Humidity path
H3986. A corrugated box loses compression strength as moisture displaces the hydrogen bonds in the fibre, and the loss is steeper than the moisture gain because the load-bearing bonds sit at the flute tips. The controlling variable is therefore humidity rather than temperature, and a box that passes in a dry warehouse can fail in a humid one at the same load.

Shelf-life physics

Linearisation limit
H3987. The rule that shelf life halves every ten kelvin is a local linearisation of an Arrhenius rate and holds only where the activation energy is constant. A failure that is a physical change with a small activation energy has a much weaker temperature dependence, and the rule's error grows toward the extremes of the range it is used across.
Diffusion control
H3988. The time for a product to reach a critical moisture content is set by the package's transmission and by the product's internal diffusion, and either can control, so a slowly diffusing product looks better packaged than it is. Measuring the product's moisture profile at one time, rather than its average, separates the two.
Buffer capacity
H3989. A desiccant extends shelf life by its capacity divided by the transmission rate, so once the capacity exceeds what the shelf life needs, the extension is linear in desiccant mass and independent of chemistry. That linearity is the test, and a desiccant that fails to show it is rate-limited rather than capacity-limited.
Confounded measures
H3990. An oxidation's induction period is set by antioxidant depletion, and the depletion rate is set by oxygen ingress, so the induction period measures the package's barrier as much as the product's stability. Any accelerated test that changes temperature and oxygen together confounds the two, and only a barrier series at fixed temperature separates them.
Reversal
H3991. In an accelerated shelf-life test, moisture gain and oxygen ingress accelerate by different factors, so two packages can rank one way in the test and the other way in storage. The reversal follows from the two activation energies, and a test at two conditions is the minimum that detects it.
Areal invariant
H3992. A package's ultraviolet blocker protects by absorbing light in a layer whose thickness is set by its concentration, so protection depends on the blocker's areal density rather than on the concentration as such. A thin, concentrated layer is therefore defeated by light entering through a pinhole or through the package's own fluorescence, and the two formulations that look equal on a specification are not.
Threshold variance
H3993. A shelf life fixed by a sensory threshold inherits the threshold's variance, and the variance of the crossing time is amplified when the quality curve is flat near that threshold. Two products with equal mean life can therefore have very different failure distributions, and the flatter one fails with less warning.
Path dependence
H3994. The effect of a temperature excursion depends on when it occurs whenever the quality kinetics are nonlinear, because the same thermal dose early and late leaves different remaining quality. Bookkeeping must therefore integrate the rate over time rather than sum equivalent days, and the difference is largest near the end of life.
Percolation
H3995. A powder cakes when liquid bridges form a percolating network between particles, so the critical moisture is a percolation threshold rather than a solubility limit. The threshold follows the particle size distribution, which is why a finer powder cakes at lower moisture with identical chemistry.
Overpressure
H3996. A retort package's internal pressure is set by its headspace gas and the product's vapour pressure, so the overpressure needed to keep the seal unloaded depends on fill temperature rather than on the retort's set point. A package validated at one fill temperature can fail at another under an identical retort profile.
Cold brittleness
H3997. At frozen temperatures a film's permeability is low but its toughness is also low, so a package that is leak-free at ambient can crack in handling at −18 °C and present as a barrier failure. The transition is the film's glass transition, which makes it a mechanical design limit rather than a permeation one.
Regime change
H3998. Oxygen stored in a headspace scales with volume while ingress scales with area, so a large headspace protects by dilution early and is dominated by ingress late. The crossover time follows the volume-to-area ratio, and it is the number that should be compared when package sizes are changed.
Missing term
H3999. A strongly hygroscopic product buffers the headspace's humidity, so the package's transmission does not translate directly into a moisture gain rate and the product can appear better packaged than it is. The buffer capacity is the plateau of the sorption isotherm, and it is the term most shelf-life estimates leave out.
Closure argument
H4000. A package's shelf life is the minimum over its failure paths, and the paths are not independent because a mechanical failure changes the barrier while a barrier failure changes the mechanical load. Improving the strongest path therefore cannot extend life, and identifying the minimum path is the only intervention that can.

Status and plan

One hundred entries, numbered in sequence. Each was composed on its own, with the argument shape fixed before the subject, so that no two share a form; the label above each claim is there to be checked against the sentence beneath it rather than accepted as a taxonomy.

The physics stays inside the pack: the polymer as a solution-diffusion medium, the cell that measures it and the artefacts that flatter it, the seal as a channel and a viscoelastic joint, the gas budget of a modified-atmosphere pack, what leaves the material and what the product takes up, the cushion and the box, and the rate processes that decide when a shelf life ends. No claim here concerns a brand, a regulation, or a person.

If one of these is worth an afternoon, it is H3965: a functional barrier is usually specified by its permeability, and the claim is that thickness against the migrant's diffusion length decides it, so a layer that looks excellent on a datasheet fails because it is too thin for the time it must hold. It needs two layer thicknesses, one migrant, and a series of storage times, and it would either explain a class of failures that are currently blamed on the resin or retire the explanation.

Incoprea Research · hand-considered hypotheses 3901–4000 · 3 November 2026 · ~/Desktop/Nov3/