Hand-considered hypotheses, 1001–1100

Written one at a time, not composed. The field is fluids: wall turbulence, free shear flow, convection, bubbles and drops, boiling, porous media, interfaces, suspensions. Each entry changes its logical form as well as its subject, and the form is printed with it. Every claim comes with something that could disprove it. None has been tested; numbers marked in parentheses as estimates are order-of-magnitude guesses, not measurements.

Wall turbulence and drag

Misattribution
H1001. Roughness is almost always indexed by a single equivalent sand-grain height, but at fixed sand-grain height the drag follows the skewness of the height distribution, because a few tall isolated elements set the virtual origin while a uniform distribution does not. The discriminating run needs three surfaces matched in k_s, differing only in skewness, in one oil channel at the same friction Reynolds number. Below a sand-grain height of about five viscous units (an estimate) the skewness should stop mattering, since the roughness no longer reaches the log region.
Hidden term
H1002. Temperature-dependent density puts a term into the shape factor of a heated boundary layer that is proportional to the wall heat flux and is normally absorbed into the pressure-gradient parameter. Two isothermal walls held at different temperatures, matched in momentum-thickness Reynolds number and in Clauser parameter, should then give shape factors differing by an amount that tracks the temperature ratio rather than the Reynolds number. The term is computable from the mean density profile, so the correction needs no new measurement.
Bound
H1003. Spanwise wall oscillation has a drag-reduction ceiling that is reached while the forcing amplitude is still rising, because beyond a certain Stokes-layer thickness the added shear dissipation grows faster than the streak disruption. Mapping the ridge of maximum reduction over a two-dimensional sweep of period against amplitude separates the two predictions: a line of constant Stokes thickness, or a line of constant wall velocity. Once that layer is thin compared with the channel, the ridge should also stop moving with Reynolds number.
Threshold
H1004. A roughness element alters drag only once it reaches far enough into the log region to displace the virtual origin, so the controlling ratio is element height to boundary-layer thickness and not height to viscous length. Building the same height-to-thickness ratio at two Reynolds numbers in different facilities tests it directly, since the roughness function must then agree even where the viscous-scaled height differs by a factor of three (an estimate).
Inversion of standard practice
H1005. Tripping a model is standard practice for guaranteeing a turbulent boundary layer, yet the trip's momentum deficit survives for tens of boundary-layer thicknesses and inflates low-Reynolds-number drag readings. Running one model with two trip heights and extrapolating to zero height recovers the untripped drag without a long fetch. A naturally transitional run on the same model supplies the endpoint of that extrapolation for free.
Selection effect
H1006. The maximum drag reduction asymptote is an envelope assembled from runs that happened to reach the highest reduction, not a property of the polymer, and it moves when the wall concentration is held fixed rather than left to deplete. With staged injection the local concentration is known at every station, and the envelope can then be watched as Reynolds number rises. A fixed-concentration loop is the only configuration in which two laboratories can compare the same envelope.
Count
H1007. Only three independent time scales survive in a turbulent channel, the viscous time, the outer time, and the time imposed by the pressure gradient, so a closure carrying more than three fitted constants is absorbing noise. Fitting on one Reynolds number and predicting a second without refitting exposes any surplus constants as free parameters.
Symmetry
H1008. The log law is fitted with one pair of constants for both walls of a channel, but unequal wall temperatures break the wall-normal symmetry of the mean profile and make the constants wall-specific. Asymmetric heating at matched friction Reynolds number is the clean test: the centerline should kink, and the kink must vanish when the two wall temperatures are equal.
Artefact mimics effect
H1009. Slip measured on a superhydrophobic wall in a recirculating loop can be dissolved gas emerging at the wall, which lowers the local viscosity and imitates slip while vanishing when the loop is degassed. The same texture run with degassed and gas-saturated water, with an inline dissolved-gas sensor, would decide the question.
Degeneracy
H1010. Riblets and streamwise travelling waves are thought to act by different routes, but at low amplitude both merely displace the virtual origin, so a surface carrying both should perform no better than the better one alone. Running each element on and off in one channel exposes sub-additivity if the routes are degenerate. The same test at high amplitude, where the wave forcing turns nonlinear, should show them ceasing to be degenerate.
Phase lag
H1011. Wall shear read by a Preston tube during a rapid pressure-gradient transient reports the wrong phase, because the tube integrates the outer velocity with a lag of order δ/U_e rather than responding on the inner time. Driving a known expansion, with wall shear read both ways at once, gives two zero crossings at different instants if the lag is real.
Unit trap
H1012. Drag reduction is quoted per part per million of polymer, but the controlling variable is concentration times molecular weight raised to a power near four-fifths (an estimate), so two polymers ranked per ppm can reverse order when ranked at equal overlap concentration. Two molecular weights compared at matched ppm and again at matched c M^0.8 demonstrate the reversal in a single sitting of rheometry.

Transition, structures, and closure

Misattribution
H1013. Streak spacing is conventionally credited to a linear instability of the mean profile, yet the spacing may simply be inherited from whatever spanwise wavelength disturbed the leading edge. Spanwise-periodic roughness at two wavelengths, placed at two streamwise stations, tells local selection from downstream inheritance. The two possibilities differ in whether the spacing can still be changed after the roughness ends.
Threshold
H1014. Pipe transition is not a property of the pipe at a critical Reynolds number but the point where the laminar basin of attraction becomes smaller than the disturbance the rig can supply. Injecting puffs of controlled amplitude across a range of Reynolds numbers maps the sustaining threshold, which ought to fall smoothly with no sharp critical number. The threshold amplitude, rather than a Reynolds number, would then be the quantity a theory must predict.
Identity
H1015. Hairpin vortices and streak instabilities are one structure seen through two conditional averages, which is why the two literatures never contradict each other directly. Applying both averaging schemes to a single high-resolution dataset and checking whether each recovers the other's defining signature would settle the identity.
Hidden term
H1016. The turbulent kinetic energy budget of a channel carries a work term from the mean pressure gradient acting on the turbulent mass flux, negligible at constant density but not when density varies with temperature. Computing a channel with a strong wall temperature gradient exposes the term directly, since it moves energy between mean and fluctuating fields at a rate comparable with production.
Reversal
H1017. Near a heated wall the turbulent Prandtl number changes sign when the molecular Prandtl number is low enough, below roughly one-fifth (an estimate), so heat travels up the mean temperature gradient in liquid metals. Simultaneous velocity and temperature measurement in a liquid-metal channel would show the sign change at a distance from the wall that moves with Prandtl number.
Scale argument
H1018. The pre-multiplied spectrum in the log region peaks at a wavelength proportional to distance from the wall, which would make the apparent outer peak a projection of a self-similar hierarchy rather than a separate structure. Plotting spectra from a dozen wall distances against wavelength over height is the whole test. A collapse would not prove the hierarchy, but a failure would end the idea.
Artefact mimics effect
H1019. Amplitude modulation of small scales by large ones is partly an artefact of the filter that defines the large scales, since the filter leaks small-scale content into the large-scale signal and biases the coefficient upward. Synthetic signals of prescribed coupling, passed through the same filter, calibrate the bias. If modulation survives that correction it is physical; if not, it belonged to the filter.
Inversion of standard practice
H1020. Adding particles to a gas is expected to raise the effective viscosity and therefore the drag, but below a volume fraction near one part in a hundred thousand (an estimate) particles gathering in low-speed streaks may damp streak instability and lower skin friction. Sweeping loading in a dilute aerosol channel at fixed charge state is enough to see whether a drag minimum precedes the expected rise.
Bound
H1021. The flux Richardson number of a stably stratified shear flow cannot exceed a ceiling fixed by the ratio of buoyancy time to shear time, and the approach to that ceiling is what saturates mixing efficiency. Stratified grid turbulence with independent control of shear and buoyancy should show the saturation at a Richardson number set by the shear rather than by the initial perturbation.
Conservation audit
H1022. Turbulent kinetic energy budgets measured with hot wires fail to close because the wire rectifies the fluctuating velocity, and the rectification is largest where the intensity is largest. Integrating the budget twice, once with each instrument, localizes the missing term and shows whether it accounts for the residual.
Degeneracy
H1023. Strong acceleration and strong convex curvature both relaminarize a boundary layer by thinning the stress-bearing layer, so a flow with both should be no more stable than the more stabilizing of the two. A curved accelerating channel with independent control of acceleration and curvature would test that sub-additivity.
Unobservable
H1024. Dissipation is not measurable by any single probe, because the smallest scales are smaller than the wire and what is reported is a filtered dissipation whose filter depends on wire diameter. Three matched wires of different diameter, in one flow, return three estimates differing by a factor set by the filter rather than by the flow. That would make quoted dissipation values instrument-specific at the level of tens of percent (an estimate).

Jets, wakes, and mixing

Misattribution
H1025. Jet spreading is attributed to large eddies, but in a finite enclosure the rate may be set by the streamwise pressure variation at the jet edge. Discharging the same nozzle into a large plenum and into a narrow duct, at matched exit conditions, separates the two if the rates differ.
Threshold
H1026. A jet becomes self-similar only after a distance that scales with the lip boundary-layer thickness rather than with the nozzle diameter, so equal-diameter nozzles with different lips reach similarity at different stations. Lip suction to thin the boundary layer would move the virtual origin while leaving the diameter unchanged.
Identity
H1027. The mixing transition in a compressible shear layer and the enhancement seen behind a shock are one process, both being the amplification of oblique modes by a mean density gradient. Producing the density gradient by heating instead of by a shock, at matched convective Mach number, is the discriminating configuration.
Hidden term
H1028. Entrainment in a turbulent jet contains a contribution from pressure work at the jet edge that is normally folded into a constant entrainment coefficient, which then becomes a function of the co-flow ratio. Measurements at several co-flow ratios, with the edge pressure field mapped separately, would split the two contributions.
Reversal
H1029. A bluff-body wake can narrow rather than widen when freestream turbulence is raised, because the extra disturbance triggers earlier roll-up and reattaches the shear layers. Varying grid intensity and scale ahead of a cylinder locates a minimum wake width, expected near a few percent (an estimate) if the reversal is real.
Count
H1030. A turbulent wake carries just two conserved fluxes, momentum and buoyancy, so its profile has two independent shape parameters regardless of how many are fitted. Two-parameter fits at several downstream stations leave structureless residuals only if the count is right.
Scale argument
H1031. The far-field decay exponent of grid turbulence is set by the ratio of integral scale to tunnel width, not by the mesh Reynolds number, because the largest eddies feel the walls and stop growing. One grid in two tunnels of different cross-section, at matched mesh Reynolds number, settles it by giving two decay exponents. If the exponent instead follows mesh Reynolds number, the wall effect is negligible and the standard decay law survives.
Inversion of standard practice
H1032. Screens are installed upstream to homogenize a test section, yet a screen also imposes its mesh as a preferred length that can lock the downstream turbulence to it. Swapping the screen for one of different mesh, everything else fixed, moves the integral scale while leaving the mean profile alone.
Phase lag
H1033. A forced shear layer responds not at the frequency of maximum growth rate but at the frequency whose convective lag places roll-up where the forcing is strongest. Sweeping forcing frequency and tracking the phase between forcing and roll-up location would give a phase speed equal to the eddy convection velocity, not the wave speed of linear theory.
Unit trap
H1034. Mixing efficiency is a ratio of potential to kinetic energy and is dimensionless only when both are integrated over the same control volume. Recomputing reported efficiencies with a common volume, or measuring both fluxes on one grid, should remove the occasional values that appear to exceed the theoretical ceiling.

Convection and stratification

Misattribution
H1035. Rough walls raise the Nusselt number of a convection cell, and the gain is credited to extra surface area when it may instead come from the roughness fixing plume spacing and thereby thinning the thermal boundary layer. Two roughness patterns of equal area but different pitch would separate area from spacing.
Threshold
H1036. At low Rayleigh number, roughness can lower heat transport by trapping fluid and thickening the thermal boundary layer before plumes are triggered, so the sign of the roughness effect changes at a crossover. Interchangeable floors in one cell, swept in Rayleigh number, expose the crossover, expected where roughness height is comparable with boundary-layer thickness (an estimate).
Identity
H1037. The ultimate regime and the classical boundary-layer regime may be one state observed at different Prandtl numbers, with the apparent transition marking which term controls the boundary layer. One apparatus spanning a decade in Prandtl number at fixed Rayleigh number is the test: the transition must move with Prandtl number.
Hidden term
H1038. Rotating convection carries a heat-flux term from Ekman pumping that is usually attributed to the sidewall, and the term scales with the square root of the Ekman number. Replacing a no-slip sidewall with a stress-free one at matched rotation rate removes the term and shifts the Nusselt number by a predictable amount.
Reversal
H1039. Tilting a convection cell first lowers and then raises the heat transport, because a small tilt aligns plumes with gravity while a large one drives a circulation that scours the boundary layers. A tilt sweep at fixed Rayleigh number in one cell traces the non-monotonic curve and locates its minimum.
Bound
H1040. The large-scale circulation of a Rayleigh-Bénard cell has a maximum mean speed set by the balance of buoyancy against wall friction, and the bound is reached once the aspect ratio exceeds about two (an estimate). Three aspect ratios at matched Rayleigh and Prandtl numbers give the same wind speed if the bound is real.
Selection effect
H1041. Scatter in published Nusselt-Rayleigh exponents comes from experiments selecting the Rayleigh range where their apparatus is most stable, and that range correlates with aspect ratio. Replotting the compiled data against aspect ratio rather than Rayleigh number should reveal the correlation.
Degeneracy
H1042. Penetrative and double-diffusive convection give the same heat-flux scaling while the stabilizing gradient is weak, because both are governed by the ratio of buoyancy frequencies. Adding a tracer whose diffusivity differs from that of the destabilizing component breaks the degeneracy and separates the two.

Bubbles, drops, and disperse flow

Misattribution
H1043. Drag reduction in a bubbly channel is credited to bubble buoyancy, but it may come from bubbles collecting in the streak region and breaking the streaks. Running downward, so buoyancy opposes the mean flow, keeps the reduction only if the streak mechanism is the cause.
Threshold
H1044. Turbulent breakup of a bubble requires not only a critical Weber number but also a residence time in the high-strain region long enough for deformation to grow. Known strain rate and residence time in a converging-diverging channel place breakup at a Weber number that depends on residence time. Breakup maps built on Weber number alone would then be facility-specific.
Identity
H1045. Pseudo-turbulence in a bubbly flow and particle-induced turbulence in a dilute suspension are one fluctuation field generated by the same wake-shedding instability and scaled by disperse-phase size. Matching a bubbly flow against a solid-particle flow in size, loading, and Reynolds number aligns their low-frequency spectra if the fields are the same.
Hidden term
H1046. The sound speed of a bubbly liquid includes a contribution from each bubble's radial inertia that is normally dropped, and the contribution becomes significant above a few percent void fraction. Filling a resonator with bubbly water at controlled void fraction turns the claim into a frequency shift that tracks bubble size.
Reversal
H1047. The lift on a bubble in a shear flow can reverse sign at high shear, because the vorticity-induced lift is overcome by deformation-induced lift once the bubble aligns with the strain. Controlled shear and bubble size in a Couette cell locate the reversal at a critical Weber number.
Scale argument
H1048. Coalescence in a turbulently stirred emulsion is governed by the ratio of film drainage time to the eddy turnaround time at the drop scale, so rates collapse on that ratio rather than on the Weber number. Stirring a cell at variable viscosity ratio tests the collapse over a range of drop sizes.
Artefact mimics effect
H1049. Turbulence suppression reported in vertical bubbly pipe flow can be a blind-zone artefact of the phase-detection probe, which discards high-velocity core samples in proportion to local void fraction. Comparing probe statistics with an optical measurement at the same station would show the suppression shrinking as the blind zone is corrected. If the correction removes the suppression, the effect belongs to the instrument.
Inversion of standard practice
H1050. Raising the shear rate is the standard way to break a surfactant-laden drop, but beyond a point the Marangoni stress grows faster than the viscous stress and the drop behaves as if it were more viscous. Stepping the shear ramp while measuring interfacial tension in situ shows breakup only after the surfactant is stripped from the interface.
Count
H1051. A monodisperse bubbly flow at low void fraction needs three independent closure quantities, void fraction, slip velocity and interfacial area, so a model adding a fourth without a new measurement is unidentifiable. A parameter-identifiability analysis on one dataset shows the fourth trading off against the other three.
Unobservable
H1052. Instantaneous interfacial area may not be recoverable from line-averaged void-fraction chords at any chord count, because different area distributions produce identical chord-length statistics. Reconstructing one synthetic flow from two chord counts returns the same statistics from different area fields.

Boiling and phase change

Misattribution
H1053. Critical heat flux in a channel is attributed to a wall-blanketing instability, but in a short channel it may be limited by the pressure drop available to expel vapour. Matched subcooling at two system pressures separates the wall instability from the expulsion limit.
Threshold
H1054. Flow boiling in a microchannel becomes unstable below a threshold mass flux not because of the pump but because the channel's vapour compressibility exceeds the loop's, with the threshold set by the ratio of channel volume to loop volume. Changing the accumulator volume moves the threshold by the predicted ratio.
Identity
H1055. Dryout in a heated tube and critical heat flux in a pool are the same event whenever the tube diameter exceeds the capillary length, both being a failure to resupply liquid to the wall. Decreasing the tube diameter makes the two criteria merge at that length.
Hidden term
H1056. Condensation on a vertical surface carries a term from wave-induced thinning of the film that is usually folded into the Nusselt correlation, and the term varies with film Reynolds number raised to a power near one-third (an estimate). Local film thickness and local heat flux measured on the same surface make the correlation residual track thickness fluctuation.
Reversal
H1057. Subcooled boiling first raises and then lowers the pressure drop, because bubbles initially increase mixture viscosity and later lubricate the wall by reducing the effective shear. Local pressure taps along a heated channel catch the maximum at a void fraction of a few percent (an estimate).
Bound
H1058. The Leidenfrost temperature has an upper bound set by the Rayleigh-Taylor wavelength of the vapour film, and the bound is approached once the surface is rough enough that chemistry no longer matters. Sweeping smooth and rough surfaces of one material through the temperature range shows their Leidenfrost points converging.
Artefact mimics effect
H1059. Reported nucleate-boiling site densities are limited by the resolution of the microscope used to count them, so the density is really the density of sites larger than the resolution and the apparent power-law exponent is set by the counting. Counting one surface at three magnifications changes the exponent, which is the whole point.
Phase lag
H1060. The wall temperature response to a step in heat flux during flow boiling lags by the wall's diffusion time rather than the fluid's, so a measured boiling curve is a convolution of wall inertia with the fluid response. Walls of two thicknesses give curves that collapse only after deconvolution with the wall impulse response.

Porous media

Misattribution
H1061. The Forchheimer coefficient is attributed to inertia, but part of it is the same viscous drag that produces Darcy's law evaluated at the true tortuous velocity, so fitting both to one dataset double-counts. Packs of increasing tortuosity at matched permeability decide whether the coefficient follows tortuosity or Reynolds number.
Threshold
H1062. A yield-stress fluid begins to move in a porous medium only above a pressure gradient scaling with the yield stress over the pore throat, and a core threshold exceeds the local one because flow samples the throat distribution. Cores of the same mean throat size, but different spreads, give different thresholds.
Identity
H1063. Early tracer breakthrough and long tailing in a column are two ends of one residence-time distribution shaped by dead-end pores, so a model fitting only one describes half the flow. A column with its dead-end fraction measured by imaging would let the same distribution predict both ends.
Hidden term
H1064. The effective stress law for a soft poroelastic solid carries a term from the solid's own compressibility that is absorbed into the Biot coefficient and then varies with strain. Drained and undrained tests across a stress range, on one sample, reveal the coefficient drifting.
Reversal
H1065. Relative permeability can rise as saturation falls in a narrow band near the residual, because disconnected ganglia reconnect through films and open a low-resistance path. A core flood resolved finely in saturation reveals the non-monotonic segment.
Scale argument
H1066. Dispersion in a layered porous medium is controlled by the ratio of transverse dispersivity to layer thickness rather than by Péclet number, so columns of different diameter at matched Péclet number disperse differently. Layered packs of controlled thickness collapse on that ratio, or they do not.
Inversion of standard practice
H1067. Sand packs are vibrated to maximize density, but a denser pack has a narrower throat distribution and can be less permeable, because the loss of the few large throats outweighs the gain in porosity. Two packs at matched porosity, one vibrated and one not, differ in permeability if the throat distribution is what matters.
Degeneracy
H1068. Capillary pressure and relative permeability are both moments of one throat-size distribution, so fitting them separately leaves a redundant degree of freedom in any two-phase model. Deriving both from a microtomographic throat distribution, then predicting the other, exposes the redundancy.
Unit trap
H1069. Permeability is reported in darcy, a unit of area, so a tight core is not a different kind of rock but the same rock with smaller pores, and comparisons that treat permeability as a material class conflate geometry with scale. Cores sintered from different grain sizes to the same permeability differ in Forchheimer coefficient if geometry is being read as scale. The pair is easy to make, and the test would show whether the darcy is being read as a material label.
Derived quantity
H1070. The Kozeny constant is a product of tortuosity and specific surface, so a measurement reporting it as fixed has absorbed a change in one factor into the other. Tortuosity by resistivity and surface area by gas adsorption, on the same packs, show the product varying while neither factor is constant.
Bound
H1071. The residual non-wetting saturation left after a flood is bounded by the pore-throat aspect ratio rather than by the volume injected, once the capillary number is low. Floods at matched capillary number but different rates return the same residual only if that bound holds.
Symmetry
H1072. Injection and production are not symmetric in a porous medium even for reversible flow, because the advancing front is unstable while the retreating one is stable. A core with ports at both ends, flooded both ways with the same fluids, gives two different breakthrough curves.
Artefact mimics effect
H1073. Stress-dependent permeability of a core can be the confining sleeve changing the flow area rather than the rock changing its pores. Jacketing one core in two sleeve materials of different stiffness changes its stress sensitivity at the same effective stress. Sleeve compliance is rarely reported, which is why the effect could persist unnoticed.
Bound
H1074. The rate at which a porous medium can be produced without mobilizing fines is bounded by the interstitial velocity at the wellbore, not by the Darcy velocity, so a safe field rate is lower by a factor of porosity than a rule that applies the same critical velocity to Darcy flow. A radial cell with graded permeability puts the bound at a rate that scales with porosity.

Interfaces and thin films

Misattribution
H1075. The slowing of a spreading drop on a smooth surface is attributed to contact-line friction, but some of it may be evaporative cooling raising viscosity near the contact line. Spreading the same drop in vapour near saturation, rather than in dry air, separates the two.
Threshold
H1076. A thin film ruptures only below a thickness set by the balance of disjoining pressure against capillary pressure, and the critical thickness depends on how the film was made because the disjoining pressure is hysteretic. Films drained to a thickness and films spin-coated to the same thickness rupture at different times, which no thickness-only theory allows. If rupture time depends on preparation at fixed thickness, disjoining pressure is not a function of thickness alone.
Identity
H1077. The coffee-ring deposit and the Marangoni deposit are the same edge-directed capillary flow driven by two different gradients, so reversing the surfactant gradient should convert one into the other without changing the flow topology. Particle tracking in a drop with a controlled surfactant source catches the flow reversing while the deposit pattern stays recognizably similar.
Hidden term
H1078. The drag of a superhydrophobic channel contains a term from the pressure drop across the menisci of the gas layer that is usually omitted and that grows as the channel height falls. Two channel heights with identical texture differ in drag by exactly the meniscus term.
Reversal
H1079. Drag on a textured surface can fall and then rise as post pitch increases, because small pitch traps gas while large pitch lets the interface sag into contact. Posts of increasing pitch in one channel put the drag minimum at an intermediate pitch.
Scale argument
H1080. Whether a texture is slippery to a flowing liquid depends on the ratio of gas-layer thickness to texture pitch rather than on the coating's chemistry, so surfaces of equal chemistry but different pitch have different slip. A pitch series at one chemistry collapses on that ratio. No new coating is needed, only new masks.
Inversion of standard practice
H1081. Surfactant is added to lower surface tension and speed wetting, but at a moving front the same surfactant builds a Marangoni stress that opposes the front, so dosing at the reservoir can slow imbibition. Dosing the meniscus locally, against dosing the reservoir at equal total surfactant mass, gives two different front speeds.
Count
H1082. Foam stability at fixed liquid fraction has two free parameters at most, the slope of the disjoining pressure curve and the interfacial dilatational modulus, so rankings using more are not identifiable. Two surfactants at matched equilibrium tension but different modulus differ in foam stability only through the modulus.
Conservation audit
H1083. The gas volume of a bubble grown slowly on a needle is smaller than the volume metered by the syringe, because gas dissolves into the liquid during growth, and the deficit scales with growth time. Weighing the dissolved gas before and after closes the balance; growing at two rates shows the deficit scaling with time.
Degeneracy
H1084. The critical capillary number for drop breakup and the critical viscosity ratio are degenerate in the Taylor description, so experiments varying only one cannot locate the boundary. A two-dimensional sweep of both variables in one shear cell turns the boundary into a curve rather than a point.
Phase lag
H1085. The contact line of an advancing meniscus overshoots after a pressure step, with a lag set by the viscous time over the wetted length. A fast pressure step in a capillary tube, filmed at high speed, reveals an overshoot whose period scales with the square of the wetted length.
Unobservable
H1086. The contact angle of a moving meniscus in a transparent cell is not measurable from outside, because the optical path integrates meniscus curvature over the cell depth and returns a depth average. Two cell depths and one fluid give two apparent angles if this is right. The correction is a geometrical factor computable from cell depth and meniscus shape.

Suspensions and complex fluids

Misattribution
H1087. Shear thickening in a dense suspension is attributed to particle contacts, though dilatant migration of particles away from the shearing surface may account for the remainder by raising concentration in the bulk. Concentration measured across a Couette gap changes at the onset of thickening.
Threshold
H1088. A flowing suspension is inhomogeneous below a critical shear rate because shear-induced migration pushes particles toward the center, and the onset tracks the ratio of particle radius to pipe radius. Two pipe diameters and one suspension put the onsets at rates that collapse on that ratio.
Identity
H1089. Yield stresses measured by a vane and by a cone-and-plate on the same paste differ because both record a fracture event whose stress concentration depends on geometry, not because the material has two yield stresses. Vanes of two diameters agree only if the event is a material property.
Hidden term
H1090. Apparent wall slip of a polymer melt in a die includes a compressibility contribution from near the entrance that is normally assigned to true slip. A long die and a short die at the same wall shear differ in apparent slip if compressibility is the source.
Reversal
H1091. The pressure drop of a thixotropic fluid in a pipe depends on the direction of the ramp, because structure rebuilds during low-rate steps and is broken during high-rate ones. A staircase ramp up and down, with step duration controlled, widens the hysteresis loop as the steps lengthen.
Scale argument
H1092. Whether a granular layer flows or jams follows the ratio of imposed shear rate to the particles' own fluctuation rate rather than packing fraction alone, so one packing can do both depending on wall roughness. Rough and smooth Couette walls at one packing give transitions at different shear rates.
Inversion of standard practice
H1093. Fluidized beds are run above minimum fluidization, but pulsing gas just below that velocity can mix better at lower pressure drop, because each pulse fluidizes locally without sustaining a global bubble phase. Pulsed gas against steady gas at matched time-averaged velocity shows better mixing for the pulsed case.
Count
H1094. A dilute suspension has at most two independent viscometric functions, viscosity and first normal stress difference, so a third reported at the same shear rate is a combination of the two. One suspension measured in cone-and-plate and in Couette exposes the third quantity as a derived combination.
Conservation audit
H1095. Solids flux in a horizontal slurry pipe is not the product of mean concentration and mean velocity, because the two profiles are correlated and the covariance can be a substantial fraction of the total. Both profiles measured at once expose the missing covariance term.
Degeneracy
H1096. Two suspensions matched in viscosity at one shear rate can differ at another, so a single-point viscosity is not a material property and rankings built on it are arbitrary. Sweeping shear on suspensions matched at one rate makes the curves cross.

Scaling and measurement

Unit trap
H1097. Reynolds number is quoted without the length that defines it, so two experiments at the same value can differ by orders of magnitude in scale separation, which is the quantity that actually controls the flow. Replotting compiled data against scale separation instead of Reynolds number collapses cases that currently disagree.
Derived quantity
H1098. The drag coefficient folds in a normalizing area, so a body that is better at one normalization can look worse at another, and only comparisons at matched area and matched Reynolds number survive. Two bodies with different area conventions reverse rank under the two conventions.
Bound
H1099. No Reynolds-stress closure can be both local and exact, because the stress at a point depends on the pressure field, which is nonlocal, so every local model carries an error bounded below by the pressure-strain integral scale. A local model against direct simulation at two Reynolds numbers shows the error failing to fall with resolution. That floor would appear as a resolution-independent bias in every a-posteriori test.
Threshold
H1100. A turbulence model calibrated at one Reynolds number fails above a threshold in scale separation, and below it the failure is invisible because the fitted constant absorbs it. One model across a decade of Reynolds number, with error plotted against scale separation, locates the threshold. Below that point calibration and physics are indistinguishable, and this is why the failure is usually reported as scatter.

Status and plan

100 of the planned set, written singly. The forms are tracked so that repetition would be visible rather than hidden: eighteen of them appear here, and no two entries open the same way. None has been tested.

Several of these are cheap to check with equipment most fluids laboratories already own, and a few would need a new rig. The list is offered as ideas, not as results.

If one is worth an experiment, it is H1086: that the contact angle of a moving meniscus read through a transparent cell is a depth average, so two cells of different depth give different apparent angles for the same fluid. It needs two cells, one camera, and an afternoon.

Incoprea Research · hand-considered hypotheses 1001–1100 · 3 November 2026 · ~/Desktop/Nov3/