Hand-considered hypotheses, 1701–1800
Food and agricultural processing, written one at a time. The subject is the physical plant: dryers, freezers, retorts, mixers, gas-flushed packs, silos, rheometers and conveyors. Each entry is given its own logical form — a misattribution, a hidden term, an identity, a bound, a threshold, a reversal, a count, an artefact that mimics an effect, a scale argument, an inversion of standard practice, a selection effect, a unit trap, a derived quantity, and several forms of my own — printed with the entry so that the structure, not only the topic, can be seen to differ. None has been tested. Where a number appears it is an estimate from the mechanism, not a measurement, and no entry makes a claim about what a food does to a body; the claims are about heat, water, gas, force and time.
Drying and sorption
Misattribution
H1701. Case hardening in convectively dried fruit is read as a skin of migrated solute, but the skin appears when evaporative cooling drops the surface below its glass transition, so the stiff layer is a glass and not a deposit. If that is so, drying the same slab in air humid enough to suppress evaporative cooling should leave the surface rubbery and the drying curve smooth, even though the solute flux reaching the surface is unchanged.
Hidden term
H1702. Every effective moisture diffusivity fitted from a drying curve carries the shrinkage term, because the fit is made in a coordinate frame that contracts while the sample does, and the true diffusivity differs by a factor of the order of the volume strain rate (estimate, not measured). Drying a pre-rigidised slab — freeze-dried and then resaturated inside a rigid mesh — should return a different fitted diffusivity for the same material, which is the one-experiment check.
Bound
H1703. Rehydration capacity has a ceiling no drying schedule can lift: it is the ratio of the original open pore volume to the pore volume that survives collapse, and collapse is irreversible once the matrix passes its glass transition. Micro-tomography of the same specimen before drying and after rehydration should show the ceiling directly, and any sample exceeding it has taken water into closed cells rather than into the pore network.
Threshold
H1704. The drying rate stops responding to air velocity at a value that belongs to the sample rather than to the dryer, because the knee sits where external and internal resistances are equal, at a Biot number near two (estimate). A velocity series run on slabs of two thicknesses should put the knee at the same Biot number rather than at the same velocity.
Reversal
H1705. A hygroscopic surface layer can slow drying in humid air and accelerate it in dry air, crossing over at the humidity where the layer's equilibrium uptake matches the vapour flux it can pass. The layer acts as a diffusion valve whose resistance grows with sorbed water, so its direction of effect depends on the driving force rather than on the coating. Coated and uncoated slabs across a humidity series should show the crossing point.
Count
H1706. An isothermal thin-layer drying curve supports at most three independent parameters — a rate constant, an equilibrium moisture and a shape — and models with five or six terms absorb the moisture-ratio transform instead of the physics. Fitting each candidate on one temperature and predicting another should show the extra parameters improving the fit while leaving the prediction untouched.
Artefact mimics effect
H1707. The long tail of a drying curve, read as a slowly relaxing bound-water fraction, can be the chamber's own headspace coming to equilibrium with the sample. Shrinking the headspace tenfold should shorten the tail tenfold while leaving the early curve unchanged, and if it does the tail was never in the sample.
Scale argument
H1708. Thin-layer data taken at a few millimetres cannot be extrapolated to a bed tens of centimetres deep, because the bed crosses from external to internal control somewhere in between and the crossing depth follows from the two resistances. A depth series in one dryer should show the drying time changing from linear to quadratic in depth at that point.
Inversion of standard practice
H1709. Tempering between drying passes is scheduled for the time moisture needs to equilibrate, yet the stress that cracks a kernel relaxes faster than the moisture profile flattens, so the usual practice holds the grain longer than the mechanics require and wastes dryer capacity. Cutting the pass at the relaxation time instead of the equilibration time should give the same breakage with visibly shorter tempering.
Selection effect
H1710. Published optima for drying temperature are often chosen by scoring appearance at a fixed elapsed time, which folds the rate effect into the colour kinetics and leaves an optimum belonging to the scoring protocol. Re-scoring the same runs at equal moisture content should move the optimum, and the size of the move measures how much of it was selection.
Unit trap
H1711. Dry-basis and wet-basis moisture contents diverge fastest where drying curves are steepest, so a rate quoted as a percentage gain on one basis can be a deficit on the other, with the difference exceeding typical scatter above roughly half a kilogram per kilogram dry basis (estimate). Re-plotting published curves on both bases shows how many reported improvements change sign.
Derived quantity
H1712. Drying efficiency, the energy carried off in vapour divided by the energy supplied, is neither a material nor a dryer constant, because its denominator contains wall losses and fan work that were never measured. Running the dryer empty at the same set point supplies the loss term, and subtracting it is the only route by which two machines' efficiencies become comparable.
Conservation argument
H1713. In a closed-loop heat-pump dryer the only water leaving the loop is the condensate, so condensed mass integrates the drying rate directly and can be logged without touching the sample. Any divergence between that integral and the gravimetric loss is a leak or an unaccounted sorption inside the loop, and the divergence itself is the diagnostic.
Non-invertibility
H1714. Sorption isotherms of starchy materials are not single-valued: desorption and adsorption at the same water activity differ, so a moisture meter calibrated on one branch misreads by an amount that grows with the number of humid-dry cycles the material has seen. Cycling a sample between two water activities and reading it on both branches quantifies the ambiguity without any need for a reference method.
Symmetry argument
H1715. A geometrically symmetric through-flow dryer should produce a symmetric moisture profile, so asymmetry in the finished bed is a map of air maldistribution obtained far more cheaply than a velocity survey. Blocking one inlet quadrant should move the asymmetry in a predictable direction, which turns the symmetry check into a calibrated instrument.
Turning point
H1716. The surface temperature of a drying solid rises sharply at the moment the surface stops being wet, and that turning point locates the transition with no mass measurement at all. The turn should coincide with the end of the constant-rate period in a companion gravimetric run; if it does not, the thermocouple is reading the boundary layer rather than the surface.
Fitting artefact
H1717. The monolayer capacity of the GAB equation is quoted as a material property, yet it moves when the water-activity range of the fit is changed, because the three parameters trade against one another outside the fitted window. Fitting one laboratory's isotherm over three different ranges should yield three monolayers, which settles whether the number belongs to the material or to the fit.
Saturation argument
H1718. Microwave drying cannot drive a product below a moisture floor, because the loss factor falls as water is removed and the material stops absorbing while water remains; the floor tracks the cavity field and the penetration depth rather than the drying air. A thickness series at fixed power should show the floor moving with penetration depth, not with air conditions.
Misattribution
H1719. Browning during low-temperature drying is usually assigned to Maillard chemistry, but much of the darkening at low water activity is stronger surface scattering from a pore structure that has coarsened. Index-matching the pores with a liquid of matching refractive index should remove the darkening without altering any chemistry.
Competitive exclusion
H1720. The constant-rate period is explained either by surface evaporation or by capillary supply, and the two cannot both be limiting: the period ends when supply falls below demand, which makes the critical moisture content a crossover of two rates rather than a fixed material constant. Raising the air velocity should therefore move the crossover moisture, and a genuine constant would not move.
Freezing and frozen storage
Threshold
H1721. Supercooling before ice nucleation in a food matrix is set by the largest pore of the most active foreign particle present, not by the matrix's water activity, so the nucleation temperature should be insensitive to solute concentration above a certain level. Filtering the mix through graded membranes and measuring nucleation temperature should track the filter cut-off rather than the solids content.
Misattribution
H1722. The length of a freezing plateau is read as latent heat, but part of it is the thermocouple's own lag inside a low-conductivity sample, which stretches the plateau without adding any heat. Repeating with a finer junction and extrapolating to zero diameter should shorten the plateau to its true length.
Identity
H1723. Recrystallisation during frozen storage and Ostwald ripening are the same coarsening process and must obey one rate law once mean radius is used, so two reported rates that disagree are describing different parts of the size distribution. Measuring the full distribution instead of a mean should collapse the disagreement without invoking any additional mechanism.
Bound
H1724. The ice crystal size reachable by fast freezing is bounded by the distance heat can diffuse during the time the sample spends between nucleation and glass transition, so that length, not the cooling rate on the dial, sets the floor. A thickness series at one surface temperature should show the same crystal size once thickness falls below the diffusion length.
Hidden term
H1725. Freezing point depression measured in a porous food carries a confinement term in addition to the colligative one, so one solute gives different depressions in different pore sizes. Graded pore glasses soaked in a single solution should show the depression varying with pore radius at fixed composition, which isolates the confinement contribution.
Reversal
H1726. A permeating cryoprotectant reduces ice damage at slow cooling and can increase it at fast cooling, because at fast rates it plasticises the matrix and lowers the glass transition before it has time to equilibrate osmotically. Crossing two cooling rates with two additive concentrations should expose the sign change in damage.
Scale argument
H1727. Freezing time scales with the square of thickness when conduction limits and linearly in thickness when surface resistance limits, so a centimetre-scale laboratory slab and a decimetre product sit on different branches and cannot be predicted from each other. A thickness series locates the crossover Biot number, and any plant prediction must be made on the branch the product actually occupies.
Count
H1728. Predicting a freezing curve requires three thermal properties plus the unfrozen-water curve, which is four independent quantities, and the three-property models appear to work only because the unfrozen-water curve has been folded into an effective heat capacity. Measuring unfrozen water independently should restore the fourth parameter and break the three-property fit.
Artefact mimics effect
H1729. A dense surface layer in micrographs of frozen desserts can be cutting smear from the microtome rather than a real ice-concentration zone at the surface. Sectioning the same block at two temperatures well below the ice point should thin or erase the layer at the colder knife, which identifies it as deformation rather than composition.
Unit trap
H1730. Freezing rates quoted in degrees per minute at the surface and at the thermal centre differ by roughly the Biot number, so a phrase such as rapid freezing means different things in different geometries. Reporting both traces converts the claim into an invariant, and only then do two studies become comparable.
Derived quantity
H1731. The thermal centre of a package is derived from geometry and boundary condition, not fixed at the geometric centre, and in an asymmetric pack it can sit off-axis by a distance comparable to the thickness difference. A thermocouple array across the pack should locate the true cold point and show it moving when the surrounding medium changes.
Inversion of standard practice
H1732. Raising air velocity in blast freezing is treated as unconditionally better, but for a packaged product the limiting resistances are the package surface film and the product's own conduction; past the point where those dominate, extra velocity adds dehydration without adding cooling. A velocity series on packaged and unpackaged samples should show the packaged ones flattening far earlier than the bare ones.
Threshold
H1733. Plate freezing beats air freezing only above a product thickness set by the ratio of contact resistance to internal conduction resistance, so thin products can freeze faster in air despite worse contact. Running one product on both systems across a thickness series should produce a crossover whose position is that ratio.
Selection effect
H1734. Ice crystal sizes reported from micrographs are selected by the section plane, and the bias grows with crystal anisotropy, so a sample that looks finer may simply have been cut across its long axis. Sectioning one block in three orthogonal planes should show the apparent size changing by a factor that measures the anisotropy directly.
Conservation argument
H1735. In a closed freezing chamber the heat removed equals the refrigerant's enthalpy rise, so integrating that rise over the run yields the ice fraction with no calorimetry at all. A gap against the measured latent plateau is chamber heat leak, and an empty-chamber run through the same profile estimates it.
Null mechanism
H1736. A time-temperature indicator that changes colour in frozen storage is sometimes credited to ice recrystallisation, but recrystallisation at constant temperature offers no mechanism to drive a colour change beyond ordinary diffusion, so an isothermal hold should reproduce any change produced by a fluctuating profile with the same integral. If it does not, the indicator is responding to something other than the ice.
Thermal processing
Misattribution
H1737. Come-up time in a steam retort is usually charged to the product's heat capacity, but a large part of it belongs to the condensation resistance of a steam-air mixture at the can surface. Running the same can in pure steam and in steam containing a few percent air should separate the two, since only the surface term moves.
Hidden term
H1738. A lethality integral computed from a single thermocouple includes the conduction shape factor between junction and food, so the probe reads colder than the cold point it is meant to represent. Two probe diameters at the same location should give two integrals, and extrapolation to zero diameter yields the product's own value.
Identity
H1739. Cook value and lethality are one integral referred to different reference temperatures, so any process change that raises one raises the other by a computable factor and there is no independent knob between them. Recomputing both from a single time-temperature record should show that fixed ratio and expose any optimisation claiming to move them apart.
Bound
H1740. Surface heat flux in a condensing retort is bounded by the condensate film rather than by the steam supply, so added baffles or more steam cannot push the flux past a ceiling set by film thickness and gravity. A flux gauge on the can wall should flatten while the steam supply is still being raised.
Percolation argument
H1741. Volumetric ohmic heating degrades into a current path through the more conductive phase once that phase percolates, so heating uniformity is governed by a threshold in volume fraction rather than by the conductivity ratio alone. Two-phase gels at graded volume fractions should show uniformity collapsing at the percolation fraction and at no other composition.
Reversal
H1742. Lengthening a holding tube raises the mean residence time but narrows the distribution, so the coldest element's thermal load can rise while the hottest element's falls; more holding is not uniformly more cook. Residence-time distributions measured at two tube lengths with the same tracer should show the two ends of the distribution moving in opposite directions.
Count
H1743. A tracer experiment resolves only as many residence-time moments as its sampling interval permits, and higher moments fitted beyond that are interpolation rather than measurement. Resampling the same record at half the rate should preserve the claimed moments if they are real and destroy them if they are not.
Artefact mimics effect
H1744. The apparent shear-thinning of a fluid passing through a heat exchanger is often the temperature dependence of viscosity along the tube rather than a non-Newtonian rheology. Running the same fluid through an isothermal tube at matched wall temperature should straighten the flow curve.
Scale argument
H1745. Fouling rate per unit area cannot be carried from a pilot exchanger to a plant exchanger by area alone, because wall shear stress differs and fouling follows it; the correct scale variable is shear, and its exponent is measurable from two pilot flow rates. Area scaling is the standard shortcut and it fails whenever the flow regime changes between the two sizes.
Unit trap
H1746. The z-value is a log-reduction change per degree, so a z computed from Celsius data and applied to Fahrenheit data is wrong by a factor of 1.8, and the error stays silent because both numbers look plausible. Recomputing a published z on the other scale is enough to show how often the conversion has been dropped.
Derived quantity
H1747. Thermal diffusivity extracted from a heating curve is a lumped quantity containing the surface heat transfer coefficient unless the Biot number is large, so below that it describes the apparatus rather than the material. Running one material at two surface coefficients should give two diffusivities, and only their convergence at high Biot number is the material's own.
Inversion of standard practice
H1748. High-temperature short-time processing is assumed to spare a viscous product, but when the limiting element is the wall layer its residence time grows with viscosity, so that layer can receive a larger load than it would in a slower process. A wall-mounted and a centreline thermocouple in the same run should show the wall layer's integral exceeding the centre's.
Conservation argument
H1749. Steam consumed minus condensate returned equals heat into the product plus losses, so the two sides must close and a persistent gap is air in the steam, which lowers the condensing temperature without appearing on any gauge. Closing the balance in a run with deliberately injected air calibrates the gap.
Selection effect
H1750. Reported optimal blanching times are often selected from a coarse grid of sampling times, and the optimum lands where the grid is densest, so refining the grid moves it. Re-analysing one published series on a finer time base should show the optimum migrating toward the inflection of the denaturation curve.
Threshold
H1751. The coldest point in a still retort sits on the axis only while conduction dominates; above a Rayleigh number set by the product's viscosity and the container's size, convection moves the cold point off-axis. A thermocouple array in containers of increasing diameter should show the cold point leaving the axis at a predictable size.
Non-invertibility
H1752. A single heat-penetration curve cannot yield both the surface heat transfer coefficient and the thermal diffusivity, because over the measurable time window the two enter the solution almost degenerately. Only changing the medium, and with it the surface coefficient, at fixed product breaks the degeneracy.
Mixing and emulsification
Misattribution
H1753. Mixing time measured by a conductivity probe is credited to the impeller's pumping capacity, but a substantial part of the recorded time is the probe's own response and its position in the flow. Two probes at mirrored positions, and one with a faster response, should return different mixing times for the same batch.
Hidden term
H1754. The power draw of a dough mixer contains an elastic term from stored energy in the dough, so equal power can correspond to different states of development. Running the mixer at two speeds to the same power should give different dough states, separable by a small oscillatory test on samples taken at each speed.
Identity
H1755. Under-mixing and over-mixing of a dough are the same departure from a work-input peak in opposite directions, so a single energy density cannot tell them apart and only the sign of the change in resistance can. A run carried past the peak with energy logged should show equal energy at two different states, one before and one after.
Bound
H1756. Powder mixing cannot homogenise below the scale of the largest cohesive cluster in the feed, so the achievable coefficient of variation is bounded by the feed's cluster size distribution rather than by the machine. Sieving the feed to remove clusters should lower the achievable variation with no change to the mixer.
Threshold
H1757. Segregation of a binary powder appears only above a size ratio, because below it the percolation velocity of the fine fraction is too low to separate against the mixing action, and the threshold is a ratio rather than a size difference (estimate, of order one third). Graded size ratios in one mixer should keep homogeneity until the ratio crosses that value.
Reversal
H1758. Raising impeller speed in wet granulation can reduce homogeneity above a Froude number, because the regime shifts from distributive mixing to intensive densification and granules grow faster than they are dispersed. A speed series with granule size measured at fixed liquid content should show the reversal.
Artefact mimics effect
H1759. Emulsion stability scored by turbidity depends on the dilution used for the reading, since dilution leaves time for coalescence before the measurement is complete. Measuring one emulsion at two dilution ratios and extrapolating to zero dilution separates the sample's stability from the measurement's.
Scale argument
H1760. Mixing time in a stirred tank does not scale with volume; it scales with the number of impeller revolutions needed to homogenise, so geometric similarity at constant tip speed gives a longer time in the larger vessel. Two geometrically similar tanks traced with a dye should need equal revolution counts and unequal times.
Unit trap
H1761. Energy density in homogenisation is quoted per volume and per mass, and the two differ by the product's density; near 1.05 grams per cubic centimetre the gap is about five percent (estimate), the same size as many reported effects. Recomputing a published comparison on the other basis tests whether the effect survives the conversion.
Count
H1762. A laser diffractometer resolves about three independent parameters for a unimodal droplet distribution, so a five-parameter bimodal fit to the same scattering is underdetermined. Fitting a known bimodal standard with both models should show the five-parameter fit recovering the standard only by accident.
Derived quantity
H1763. The mixing index taken from a coefficient of variation depends on the mass of each sample, so two laboratories using different thief volumes report different indices for one batch. Expressing the index per sample mass, or measuring at two thief sizes, makes the numbers comparable.
Inversion of standard practice
H1764. For cohesive powders the standard instruction is to add liquid slowly, but above a cohesion threshold fast addition yields smaller granules because nucleation outruns layering. Two addition rates at the same total liquid should invert the usual size ranking.
Packaging atmosphere
Misattribution
H1765. Rising headspace oxygen in a package is usually charged to film permeability, but a seal leak produces the same rise and scales with seal length rather than film area. A pressure-decay test plus packages of two seal lengths at equal area separates the two contributions.
Hidden term
H1766. Measured oxygen transmission through a film includes the boundary layers on both sides, so a high-permeability film reads below its true value and the bias depends on the carrier gas flow. Raising the flow at fixed film should move the reading until the boundary layer is swept away, and the plateau is the film's own value.
Identity
H1767. Vacuum and modified-atmosphere packaging differ only in their initial condition, and the transient composition converges to the same equilibrium fixed by respiration and film permeation. Loading two packages with different initial headspaces and the same product should show the compositions meeting at one point.
Bound
H1768. Achievable carbon dioxide in a package is bounded by the film's selectivity and the product's respiratory quotient, so no film can push the level past a computable ceiling at a given oxygen level. Films of graded selectivity should approach that ceiling without crossing it.
Threshold
H1769. Panel collapse of a sealed package happens below a pressure deficit set by the laminate's bending stiffness and the headspace volume, which makes the threshold a structural quantity rather than a gas quantity. Packages of two film gauges and two headspace volumes should collapse at deficits predicted by plate theory.
Reversal
H1770. Increasing film thickness reduces oxygen ingress but also lengthens the package's thermal lag, so during cooling the product stays warmer and respires longer, and past some thickness the gas composition can worsen. A thickness series run through one cooling profile should show the composition turning.
Artefact mimics effect
H1771. An oxygen scavenger that appears to saturate is often being read by a sensor that consumes oxygen itself, so the plateau belongs to the instrument. A blank package carrying the same sensor and no product should show the same plateau, which identifies its owner.
Scale argument
H1772. Package gas exchange cannot be predicted from film transmission alone when the surface-to-volume ratio changes, so a pouch and a tray of the same film reach different compositions. Scaling by that ratio rather than by the film is required, and the apparent shelf-life exponent is then not one.
Unit trap
H1773. Transmission rates quoted per atmosphere and per bar differ by about one percent, but thickness normalisation and the choice of driving force can multiply that into a factor of two, so a claimed improvement can be a units slip. Recomputing a published rate in SI units from its stated conditions exposes the slip.
Derived quantity
H1774. Shelf life derived from a package is not a physical constant: it contains an acceptability limit that is a specification, so two shelf lives measured against different limits are not comparable. Recomputing one dataset against the other's limit should either reconcile the numbers or expose the difference as definitional.
Selection effect
H1775. Reported modified-atmosphere compositions are measured after equilibration, and packages that leak fail to equilibrate and are quietly dropped, so the reported mean is biased toward tight packages. Recording the leak rates of the excluded packages would size the bias.
Conservation argument
H1776. In a sealed package the moles of each gas change only by permeation and by consumption, so a rise in nitrogen with no nitrogen source in the product is a leak, and the ratio of the nitrogen change to the oxygen change is its signature. Two packages with different leak sizes should fall on a line of that slope.
Storage
Misattribution
H1777. Caking of a stored powder at the bottom of a sack is usually charged to moisture, but that layer also carries the weight of everything above it, and creep at that stress can bond particles with no moisture change at all. A load-free column stored beside a loaded one at the same humidity should cake only if moisture is responsible.
Hidden term
H1778. Crumb firmness used as a staling measure contains a moisture-redistribution term in addition to starch retrogradation, so a firm loaf is not necessarily a retrograded one. Hermetic and permeable storage at one temperature should diverge in firmness while retrogradation measured calorimetrically stays equal.
Bound
H1779. Condensation on a silo wall requires the wall temperature to fall below the dew point of the interstitial air, and that is bounded by the ratio of wall thermal resistance to grain thermal resistance, so no ambient swing produces condensation when the ratio is small enough. Wall-temperature logging in two silos of different construction should show the bound.
Reversal
H1780. Aeration is meant to cool stored grain, but below a certain airflow it can warm it, because humid air entering the bulk condenses and releases latent heat faster than the airflow removes sensible heat. An airflow series with inlet and outlet psychrometry should show the sign of the net heat change reversing.
Identity
H1781. Moisture migration in a grain bulk and thermal diffusion are one transport viewed in two frames, linked by a Soret coefficient, so the two reported rates must agree once normalised by the temperature gradient. Re-analysis of a migration experiment with the temperature field measured should collapse the two rates onto one curve.
Count
H1782. A storage trial with four chambers cannot resolve the independent effects of temperature, humidity, time and initial moisture, because the design confounds them, and any claim to have separated four effects from four chambers is aliasing. Printing the aliasing structure of a fractional design shows which conclusions the data can actually support.
Scale argument
H1783. The thermal time constant of a grain bulk scales with the square of its dimension, so a two-hundred-gram laboratory sample equilibrates in days while a two-hundred-tonne silo takes months (estimate). A trial that does not run for the larger time constant cannot observe the phenomenon it claims to test.
Unit trap
H1784. Relative humidity describes air and water activity describes the product; the two are numerically equal only at equilibrium, and using one where the other belongs reverses the predicted direction of moisture transfer. A storage experiment measuring both should show the reversal exactly where they diverge.
Food rheology
Misattribution
H1785. Apparent yield stress of a food gel from a vane measurement is credited to the network, but the vane's end effects add a term scaling with its aspect ratio. Vanes of two aspect ratios in the same gel should give two yield stresses that converge once the end correction is applied.
Hidden term
H1786. Measured viscosity of a paste contains wall slip, whose contribution varies inversely with gap, so a gap series separates it and the fitted shear-thinning exponent shifts with gap until slip is removed. Three gaps at one shear rate are enough to extract the slip velocity.
Identity
H1787. Thixotropy and shear-thinning can look identical in a loop test but are not the same: a step down in shear rate gives a time-dependent decay for thixotropy and an instantaneous drop for shear-thinning. The step response, not the loop area, is the discriminator.
Bound
H1788. The elastic modulus of a wet foam cannot exceed a ceiling set by interfacial tension over bubble radius, however stiff the continuous phase, because the bubbles deform first. Foams of graded bubble size at fixed air fraction should show the modulus tracking radius rather than matrix modulus above a threshold.
Threshold
H1789. Strain hardening in a dough appears above a strain rate set by its longest relaxation time, and below that rate the response is Newtonian, so what looks like hardening at low rates is often gravity thinning of the filament. Two filament diameters at the same rate should give different apparent hardening if gravity is responsible.
Reversal
H1790. A small amount of high-molecular-weight polymer can lower extensional viscosity at low rates while raising it at high rates, because the coil is unexpanded at low rates and strain-hardens only above the coil-stretch transition. A rate series at two concentrations should show the crossing.
Artefact mimics effect
H1791. The gel point located from a frequency sweep is often the instrument's compliance limit rather than the sample's transition, so a stiffer transducer moves the apparent gel point. Running one sample on two instruments of known compliance separates the two.
Derived quantity
H1792. The consistency index from a power-law fit is not a material property, since it trades against the exponent and therefore depends on the shear-rate window. Two laboratories using different windows should disagree by a factor computable from their exponents, and the disagreement should vanish on a common window.
Grain and seed handling
Misattribution
H1793. Grain breakage in handling is usually charged to impact velocity, but much of it comes from repeated loading below the single-impact fracture stress, which accumulates damage with no visible crack. A drop tower programmed for many sub-threshold impacts should produce breakage that no single impact of the same total energy can.
Hidden term
H1794. The angle of repose of a seed lot contains wall friction from the base it is poured onto, so a rough plate and a smooth one give two angles for the same seed. A wall-less method such as a rotating drum should give an angle below both and independent of base material.
Bound
H1795. The filling rate of a silo without segregation is bounded by the percolation velocity of the fine fraction through the flowing coarse matrix rather than by the conveyor's capacity, and above that rate the fines are trapped where they land. Filling at two rates with tracers in the fine fraction should show the transition.
Threshold
H1796. Kernel cracking during drying is governed by the moisture gradient rather than by the moisture content, so a slowly dried kernel at low moisture can stay sound while a quickly dried one at the same average moisture cracks. Drying at two rates to one final moisture, with the gradient measured by sectioning, should separate the two.
Reversal
H1797. In pneumatic conveying, particle breakage falls as air velocity rises, reaches a minimum, and then climbs again, because dilute-phase impact energy grows faster than particle concentration falls. A velocity series at fixed solids throughput should show that minimum, and its position should move with the bend radius rather than with the velocity alone.
Count
H1798. A single impact-acoustic sensor resolves only as many vibrational modes as its bandwidth allows, so hardness classifications beyond a few classes are not identifiable from that signal alone. Broadband excitation with a known sensor response should show how many modes are actually recoverable.
Selection effect
H1799. Seed-lot vigour estimates are computed from the seeds that germinated, excluding the fraction that did not, so the reported mean rises with the lot's heterogeneity rather than with its quality. Including the excluded fraction, or reporting the whole distribution, should remove the artefact.
Scale argument
H1800. A single-kernel drop test at high velocity cannot predict breakage in a bucket elevator, because the elevator applies many low-velocity impacts whose damage accumulates subcritically, and the scaling variable is the number of cycles rather than the velocity. A rig reproducing the elevator's impact count at its velocity should predict the plant's breakage.
Status
100 of 1,000. Written individually, with the logical form tracked so that repetition of structure is visible rather than hidden. The topics were chosen to cover the eight unit operations named at the top, and the forms were rotated deliberately: no form appears twice in succession, and the more unusual forms — non-invertibility, percolation, competitive exclusion, null mechanism — are used where the physics actually takes that shape.
None has been tested; the value of a list like this is the ideas, and the value of a test is separate. Estimates are marked in the text and are order-of-magnitude reasoning from the mechanism, not data.
If one of these is worth an experiment, it is H1787 — that thixotropy and shear-thinning are separable by a step down in shear rate and by nothing else. It needs a rheometer, a paste and an afternoon.
Incoprea Research · hand-considered hypotheses 1701–1800 · 3 November 2026 · ~/Desktop/Nov3/