Hand-considered hypotheses, 3001–3100

Written one at a time. Each entry carries a different subject and, more importantly, a different logical form: some say a quantity has been misattributed, some assert a bound, some an inversion, some a hidden term, some an artefact that imitates a real effect. The form is printed above each claim so that the structural variety is visible rather than asserted. None has been tested. Numbers offered as magnitudes are order-of-magnitude estimates, marked as such, and no result is quoted from any source.

Hull form and resistance

Misattribution
H3001. The form factor from a low-speed resistance test is treated as a hull constant, yet the standard extrapolation fits a straight line whose intercept absorbs whatever wave term has not vanished at the lowest speed run. Fitting the same model over three different low-speed windows should give three different intercepts, a spread of a few per cent of the form factor (estimate), if the term is wave contamination rather than hull geometry.
Hidden term
H3002. Towing-tank water holds dissolved air that plates out as a microbubble carpet on the forward hull, subtracting a small term from skin friction (order one to three per cent, estimate) that no roughness allowance accounts for. A vacuum-degassing pass between two otherwise identical runs should move the resistance by more than the repeatability while leaving the wave pattern untouched.
Threshold
H3003. A bow bulb stops paying for itself above the Froude number at which its own crest falls behind the bow wave crest, and the crossover is governed by the bulb's length rather than the hull's displacement. A length series at fixed bulb volume, run across four Froude numbers, should show every bulb crossing the bare-bow curve at nearly the same bulb Froude number.
Reversal
H3004. In head seas the bulb that saves resistance in calm water can add to it, because pitching changes the bulb's submergence and therefore the phase at which its wave cancels the bow wave. Two bulbs with equal calm-water gain, tested at the same wave period, should reverse their ranking if the phase argument holds.
Scale argument
H3005. Hull roughness penalties are not Froude-scaled, so a model tested two orders of magnitude below the ship's Reynolds number cannot reproduce the roughness regime the ship lives in. The penalty has to be built from a roughness function evaluated at full-scale Reynolds number, which makes model-scale roughness comparisons a comparison of trip behaviour rather than hull condition. A geosim series spanning a factor of four in length, with one roughness applied to all, should show the penalty changing with scale in the direction the roughness function predicts.
Selection effect
H3006. Optimization studies report calm-water gains because that is the objective they were handed, so the surviving designs are selected on one criterion and cannot reveal what they paid elsewhere. Re-evaluating a batch of optimized hulls in one common wave scatter should expose a negative correlation between calm-water gain and added-resistance penalty.
Inversion of standard practice
H3007. Hulls are optimized at the contract design speed although most fuel is burned well below it, and optimizing at the speed actually steamed yields a fuller hull, a different bulb and a different running trim. Running two optimizations, one at each speed, with the same shape parameterization should show the two optima differing in prismatic coefficient by more than the tank's repeatability.
Unit trap
H3008. Saving claims for air lubrication are usually computed against bare-hull resistance, but the honest denominator is total resistance including the air's momentum flux and the compressor's shaft power. Recomputing one trial with the total denominator can flip the sign of the claimed gain, since the two denominators differ by roughly a factor of two (estimate), and the test is an energy balance over the same run.
Count
H3009. At fixed length, displacement and speed a hull's calm-water resistance is spanned by about four to five independent shape parameters, so an optimization reporting more is fitting tank noise. The check is a hold-out: re-optimize on half the runs and see which extra parameters survive.
Dimensional inconsistency
H3010. Added resistance is conventionally normalized by significant wave height squared, which builds in an assumption of second-order dependence; normalizing by steepness squared instead should collapse steepness series onto one curve if the response really is second-order. The two normalizations part company exactly where the data are said to show a nonlinear onset.
Sign error
H3011. In shallow water a ship's sinkage grows with speed up to the critical speed, but the trim change is assumed to keep its sign throughout; the claim is that the bow lifts while the ship still settles, because the pressure field's fore-aft asymmetry reverses before total sinkage peaks. A towed model at three depth-to-draft ratios should place the trim extremum at a lower speed than the sinkage extremum.
Interference
H3012. A catamaran's wave resistance is not twice a demihull's, because the two wave systems interfere constructively at one Froude number and destructively at another, with the spacing set by the transverse wavelength. A separation sweep at fixed Froude number should show the interference oscillating with a period computable from the dispersion relation.
Confound
H3013. A tank's blockage correction is derived for deep water, so in a shallow tank it confounds wall effects with depth effects and credits the walls for what the bottom did. Repeating one hull at two tank widths and the same depth-to-draft ratio should leave a residual that the standard correction cannot remove.
Bound
H3014. Wave resistance at fixed length and displacement has a floor set by the source distribution that minimizes the Michell integral, so a measured total resistance far below that floor indicates a force-calibration error rather than a breakthrough form. Running a calibration weight through the same dynamometer chain at the same loading rate is the check.
Averaging artefact
H3015. A resistance run averaged over the whole carriage pass includes an inertial term from the model's added mass during acceleration and deceleration, and the term grows with how quickly the carriage reaches speed. Two acceleration profiles at the same steady speed should give resistances differing by an amount proportional to the acceleration.

Propulsion, wake and boundary layer

Misattribution
H3016. The wake fraction from a pitot rake behind a model is not the ship's wake fraction, because the model's boundary layer is proportionally thicker and its wake deficit deeper. Two models at different scales and the same Froude number should give wake fractions differing by the ratio of momentum thicknesses, not by a constant.
Hidden term
H3017. Thrust deduction measured behind a hull carries the rudder's blockage, so a ship with a larger rudder looks as though it has a worse hull-propeller interaction when it only has a larger obstacle. The same hull with two rudder sizes at matched propeller loading should give deductions differing by the rudder's drag.
Identity claim
H3018. Open-water and behind-hull efficiency differ by wake fraction, thrust deduction and relative rotative efficiency, and those same three numbers set the blade's unsteady load margin; a propulsor claiming an efficiency gain without moving any of them must be changing the hull's flow. A wake survey before and after fitting is the discriminating measurement.
Reversal
H3019. Contra-rotating propellers gain at high loading but can lose at low loading, because the aft rotor works in a swirl field the forward rotor has already partly straightened, leaving less swirl to recover. Testing the pair at two thrust coefficients with matched total thrust should show the efficiency difference changing sign.
Inversion of standard practice
H3020. The instinct that a larger propeller is always more efficient fails in shallow water, where tip clearance sets the maximum diameter and the optimum sits at the clearance limit rather than past it. A diameter series at fixed clearance ratio should show efficiency rising to the limit and then falling as the tips approach the bottom.
Selection effect
H3021. Energy-saving devices are selected on a single model-test loading, so the survivors tend to be those with the flattest gain curve rather than the highest peak. Testing the whole set at three loadings should reorder the ranking and reveal the selection.
Scale argument
H3022. Momentum thickness at the propeller plane scales roughly with Reynolds number to the minus one fifth, so a model's wake is thicker and its wake fraction higher than the ship's. The correction is a function of scale alone, which makes a two-scale test the only direct measurement of it.
Count
H3023. Hull-propeller interaction is carried by three numbers, namely wake fraction, thrust deduction and relative rotative efficiency, so a device claiming to shift four independent interaction terms is shifting something outside that set, most likely hull resistance. Fitting the device with the propeller removed and measuring hull resistance isolates the intruder.
Unit trap
H3024. A five per cent power saving at fixed speed is about a one and a half per cent speed gain at fixed power, because power varies with the cube of speed (estimate). Quoting the two interchangeably overstates the benefit, and both reductions should be reported from the same trial.
Degeneracy
H3025. Different wake fields and different blade designs can yield the same open-water thrust and torque while producing different unsteady blade loads, so matching the open-water curve cannot separate them. Blade-root strain gauges in a deliberately non-uniform inflow are the discriminating measurement.
Budget argument
H3026. An air-lubrication budget closes only when compressor shaft power and the air's buoyancy work are both charged against the friction saving, and several reported net gains are positive on the friction line and negative at the shaft. An energy balance over the whole plant, rather than a resistance comparison, is the test.
Failure of collapse
H3027. Thrust and torque coefficients should collapse across propeller diameters at fixed advance ratio once blade Reynolds number passes a threshold; below it the collapse fails, and the failure is a scale artefact rather than a design difference. Repeating the sweep across diameters at fixed advance ratio should show the coefficients converging above the threshold.
Rectification
H3028. A propeller in waves converts an oscillating inflow into a steady thrust rise because thrust depends on the square of the inflow, which is rectification rather than a wave thrust. A fixed propeller in an oscillating flow at constant shaft speed should show a mean thrust rise proportional to the velocity variance.
Lag
H3029. A ship accelerating hard carries its old wake for a convective time, so thrust measured during acceleration records the previous steady condition. Repeating the acceleration at two rates should shift the apparent thrust transient by exactly the difference in convective time.

Seakeeping and wave loads

Misattribution
H3030. Roll damping from a small-amplitude decay test is assigned to skin friction, but the dominant term there is eddy-making at the bilge keels' roots. Fitting root fairings, which change the vortex but not the wetted area, should cut the small-amplitude damping while leaving the large-amplitude value almost unchanged.
Hidden term
H3031. The linear roll damping coefficient from a decay contains the frequency dependence of added inertia, so a still-water decay and a forced oscillation at the same amplitude do not measure the same coefficient. Forcing at two frequencies while holding amplitude should separate the two terms.
Threshold
H3032. Parametric roll needs the encounter frequency near twice the roll natural frequency, but the wave-height threshold at which it starts is set by the amplitude dependence of damping, not by the GM variation alone. A wave-height sweep at fixed period should show the threshold moving when a bilge keel changes the damping.
Reversal
H3033. Increasing bow flare adds reserve buoyancy but can raise vertical acceleration in short waves, so a shape change that improves one seakeeping metric can worsen the impact load. Two bows of equal displacement, tested in the same short-wave train, should show the flared bow heaving less and slamming harder.
Scale argument
H3034. Slamming pressure scales with relative velocity squared, but a model's structure filters the peak, so model-scale pressures underestimate full-scale peaks unless structural dynamics are scaled too. A rigid and a flexible model of the same bow, in identical waves, should give different peak pressures.
Identity claim
H3035. Green water on the deck and bow slamming are one water-particle event seen at two stations, so deck water height should track bow vertical velocity with a phase lag set by the deck's distance from the bow. Simultaneous pressure and height records at three stations should show one event rather than two.
Count
H3036. A linear seakeeping response at a given heading is fixed by six degrees of freedom and two wave parameters, so an empirical model with more than about eight fitted coefficients is absorbing nonlinearity or noise. A hold-out test on a second heading should expose the extra coefficients.
Artefact mimicking effect
H3037. Reported springing on a model in a wave basin can be the wave-maker's own second-order free waves, which travel at a different speed and appear at twice the primary frequency. An empty-basin run at the same settings should show the same spectral line with no model present.
Symmetry argument
H3038. A symmetric hull in head seas must have zero mean yaw moment, so a measured nonzero mean indicates model misalignment or an oblique wave field rather than a physical asymmetry. A repeat with the model rotated by one hundred and eighty degrees should reverse the sign of the mean if alignment is the cause.
Confound
H3039. Added resistance measured in an irregular sea is dominated by the few largest waves, so two basins generating the same spectrum with different group structure give different answers. The test is to generate one spectrum with two phase sets and compare the means.
Degeneracy
H3040. At high encounter frequency the heave and pitch added masses are not separately identifiable from a single response measurement, because the two enter the equations in one combination. Two frequencies and two headings are needed to break the degeneracy.
Averaging artefact
H3041. Significant roll amplitude is a statistic of a non-stationary record, so its value depends on the window length whenever the record contains a transient. Reducing the same record with three window lengths should give three significant amplitudes, and their spread is the artefact.

Moorings and station-keeping

Misattribution
H3042. Damping in taut polyester moorings is usually assigned to line drag, but most of the dissipation is the rope's internal hysteresis. A dry bench test at the same tension amplitude and frequency should reproduce most of the measured damping with no water present.
Hidden term
H3043. Dynamic tension measured at a fairlead includes sheave friction and the line's own inertia, and the inertia term is in phase with acceleration rather than velocity. A frequency sweep at constant amplitude should separate the two by their phase.
Bound
H3044. Station-keeping capacity is bounded by the most heavily loaded line's mean tension plus its dynamic amplification, not by the sum of breaking strengths. Adding lines beyond the count at which load sharing saturates should not raise capacity, which a line-count series reveals.
Threshold
H3045. Polyester rope stiffness rises sharply above a mean-load threshold associated with bedding-in, so below it the rope is a different spring. A mean-load sweep at fixed amplitude should show the dynamic stiffness jumping at the threshold and the jump vanishing after sufficient cycling.
Reversal
H3046. A larger line diameter adds capacity but also adds drag, and in a current the drag grows with diameter while capacity grows with area, so beyond some diameter the heavier line carries more load than it adds. Two diameters in the same current should show the crossover in dynamic tension.
Count
H3047. A symmetric spread mooring with any number of lines resists only three independent horizontal quantities, namely surge force, sway force and yaw moment, so a design table treating per-line stiffnesses as independent is over-parameterized. Applying a pure yaw moment should reproduce the line tensions from those three resultants alone.
Selection effect
H3048. Mooring failure rates compiled from inspection records over-represent lines that were inspected, and inspection follows failures, so the reported rate is a selection artefact. A database recording inspection dates as well as failures should show the rate falling as coverage rises.
Unit trap
H3049. The mooring safety factor is defined on minimum breaking load while the design load is a long-duration extreme derived from a short simulation, so the ratio is not a probability and cannot be converted into one without the load distribution. A trial that reports both the factor and the fitted exceedance probability from the same load record would show whether the two are consistent, and by how much the factor misstates the risk.
Derived quantity
H3050. Quasi-static and dynamic mooring stiffness differ by the line's added mass and drag, so using the static value to compute the vessel's natural period gives a period that is too long. A free-decay test in still water and a forced-oscillation test should bracket the true period.
Inversion of standard practice
H3051. Catenary moorings are designed to keep the line off the seabed at extreme offset, but allowing earlier lift-off can reduce peak tension because the touchdown point then moves and the catenary lengthens. Two designs differing only in lift-off offset should rank differently at the extreme than at the mean.
Memory effect
H3052. A mooring line's tension-offset curve is hysteretic because of rope bedding-in and chain link rotation, so one static curve cannot predict the response to a transient. Cycling to the same offset at two rates should trace two curves whose enclosed area is the dissipated energy.
Scale argument
H3053. A model mooring line cannot satisfy Froude and Cauchy scaling at once, so every model test chooses one and biases the dynamic tension. Testing the same system at two model scales with the same choice should show the bias shrinking predictably.
Saturation
H3054. Mooring damping saturates when line drag is dominated by the line's own velocity, so adding lines adds capacity but not damping. A line-count series at fixed vessel amplitude should show damping per line falling as the count rises.
Rectification
H3055. Slow-drift excitation is not a force at the wave frequency but a rectification of the wave envelope, so its spectrum is the square of the envelope spectrum and depends on group structure. Two seas with the same spectrum but different grouping should give different slow-drift responses.
Interference
H3056. Thruster-hull interaction reduces effective thrust more at low advance speed, so a dynamic-positioning capacity curve is not a scaled bollard-pull curve. Thrust measured at three speeds and two headings should show the interaction changing with speed rather than with thrust alone.

Risers, pipelines and vortex-induced motion

Misattribution
H3057. VIV lock-in in a sheared current is described as shedding matching a natural frequency, but the response organizes into cells whose local shedding locks to the local velocity. A riser with two natural frequencies close together, towed in a linear shear, should show two cells rather than one.
Hidden term
H3058. A rotating rig's apparent current has a vertical component wherever the model sits off the rig's centreline, so measured VIV amplitude includes an axial flow the analysis does not model. Offsetting the model by a known amount should change the response in a way predictable from the geometry.
Identity claim
H3059. VIV of a riser and VIM of a platform are one instability viewed in a current and in waves, because the wave's orbital velocity acts as an equivalent current at the structure's natural frequency. A platform in regular waves at the VIM frequency and in a current of matching rms velocity should give the same response shape.
Bound
H3060. A strake's suppression of VIV is bounded by the drag it adds, and below a coverage fraction it adds drag without suppressing. A coverage series at fixed current should show a threshold below which net fatigue life is worse than the bare pipe's.
Threshold
H3061. A riser's touchdown zone has a curvature threshold above which the pipe lifts off the soil and the touchdown point becomes a region. Reducing top tension in steps should show the zone length jumping at the threshold rather than growing smoothly.
Reversal
H3062. Raising a riser's top tension lowers its static stress but can raise its VIV fatigue by moving the natural frequency into a more energetic current band. A tension sweep in a measured current profile should show fatigue reaching a minimum before rising.
Scale argument
H3063. VIV model tests at low Reynolds number cannot be scaled by a single factor because wake mode transitions depend on Reynolds number, so amplitude ratios between model and prototype are not constant. Two model scales at the same reduced velocity should disagree by more than the uncertainty if the dependence is real.
Count
H3064. In a sheared current the VIV response has at most a few excited modes because the shedding correlation length limits participation, so a fit with many modes is aliasing the response. Reconstructing the response from the first three modes should reproduce the measured curvature.
Degeneracy
H3065. Two current profiles can give the same VIV amplitude and dominant frequency while producing different fatigue at the touchdown, so amplitude alone cannot discriminate design options. Comparing curvature spectra at the touchdown is the discriminating measurement.
Edge effect
H3066. On-bottom stability is assessed with a uniform current, but a span's ends see a boundary layer thinned by the adjacent soil, so the effective span length is set by the soil profile rather than the flow. A span with two end geometries should show different lateral resistance at the same nominal length.
Order of limits
H3067. Pipeline lateral resistance is governed by drained soil strength at slow loading and undrained strength at fast loading, and a storm's loading rate decides which applies, so the two limits can differ by a factor of order two (estimate). Cyclic loading at two rates should show the transition.
No-go
H3068. A free-spanning pipeline cannot be stabilized by adding weight beyond the point where the added weight lowers the span's natural frequency into the current's energetic band, so no single weight works for all current speeds. A weight series at three speeds should show the failure region widening rather than closing.

Offshore structures and second-order hydrodynamics

Misattribution
H3069. A semi-submersible's air gap is set by the extreme crest, but the minimum gap is set by the platform's heave phase, which can raise the deck relative to the crest. Two platforms with the same extreme crest but different heave damping should show different minimum gaps.
Hidden term
H3070. Wave-in-deck loads measured on a model include the deck's trapped air, which is not Froude-scaled because air compressibility depends on absolute pressure. Venting the deck should change the load more at model scale than at full scale, testable with two vent areas.
Threshold
H3071. Ringing on a gravity-based structure appears above a wave steepness threshold set by second-order sum-frequency forcing, not by wave height alone. A steepness series at constant height should show the onset at a fixed steepness.
Reversal
H3072. A heave plate lowers the natural frequency by adding mass and can move the structure into a worse wave band, so a plate that reduces the resonant response in one sea state can increase it in another. A plate-size series tested in two spectra should show the ranking reversing.
Identity claim
H3073. Slow-drift force and wave-drift damping are one second-order interaction viewed in frequency and time domains, so a discrepancy between a frequency-domain prediction and a time-domain simulation indicates a missing damping term rather than a new force. Adding the damping should make the two converge.
Count
H3074. The second-order force on a fixed structure is fixed by the first-order potential on the free surface and on the body, so a third-order empirical correction is usually absorbing a discretization error. Refining the free-surface mesh should shrink the correction.
Scale argument
H3075. The Keulegan-Carpenter number governs the drag-inertia transition but does not fix Reynolds number, so tests at the same KC and different Re give different coefficients. Running one KC at three Re should show the coefficients converging above a threshold.
Confound
H3076. VIM measured in waves is confounded by slow-drift excitation in the same frequency band, so a platform that appears to VIM may simply be drifting. Repeating the test in current alone and in waves alone should show which signal the motion follows.
Selection effect
H3077. Platforms reported as having low VIM are those with high mooring damping, not necessarily those with better hull shapes, so the database is confounded by mooring design. Grouping by mooring stiffness should remove the apparent shape effect.
Inversion of standard practice
H3078. Designing for the hundred-year wave can be less conservative than designing for the ten-year wave plus a fatigue check, because the smaller event's steepness drives the impact loads that the larger event's height does not. Both conditions must be run before either can be called governing.
Unit trap
H3079. A model's displacement is a volume at model scale while a prototype's is a mass in tonnes, so comparing them directly mixes a volume with a mass. The check is to weigh the model's ballast water and compare it with the prototype's displacement divided by the cube of the scale ratio, both expressed as masses.
Feedback
H3080. Mooring stiffness changes the platform's natural period, which changes the second-order excitation, which changes the tension that set the stiffness, so a decoupled analysis misses the loop. Coupled and decoupled runs of one system should diverge most near the platform's period.
Limiting case
H3081. At zero current a strake's drag penalty vanishes but its added mass still shifts the natural frequency, so a drag-only comparison at model scale misses the frequency shift. A still-water decay test with and without strakes should show the period change.
Symmetry argument
H3082. A four-column semi with symmetric columns must have zero mean yaw in head seas, so a measured yaw indicates a current, a mooring asymmetry or a model misalignment. A still-water yaw test should separate the mooring contribution from the hydrodynamic one.

Fouling, coatings and surface mechanics

Misattribution
H3083. The drag penalty of a fouled hull is assigned to roughness height, but at low speed much of it comes from the boundary layer's virtual origin shifting over patchy fouling. Uniform and patchy fouling with the same mean height, on the same flat plate, should give different penalties.
Hidden term
H3084. A fouling-release coating's drag includes a term from the coating's own compliance, which lets the boundary layer slip slightly, and the term scales with the coating's shear modulus. Two coatings of equal thickness and roughness but different modulus should differ in drag.
Threshold
H3085. Biofouling settlement switches from settlement-limited to growth-limited above a ratio of wall shear to recruit adhesion strength, and the threshold is species-specific. A flume with a shear gradient should show the transition as a sharp front rather than a gradual decline.
Reversal
H3086. Slow steaming saves fuel per mile but can increase fouling, because the speed falls below the shear that inhibits settlement while staying above the shear that inhibits growth. Two ships on one route at different speeds should show fouling mass rising as speed falls.
Identity claim
H3087. The roughness function of a fouled surface matches that of a sand-grain surface only above a coverage threshold; below it patchiness dominates and the equivalent sand-grain height is not a surface property. A coverage series should show the equivalence failing at low coverage.
Count
H3088. At most three statistical descriptors, mean height, coverage and patch size, can carry a fouled hull's drag penalty, so a model using more is overfitting a small sample. A hold-out hull should expose the extra descriptors.
Scale argument
H3089. A fouled flat plate tested at model scale is often below the fully rough regime, so its penalty cannot be transferred to a ship by a constant factor. The transfer requires the roughness function evaluated at the ship's roughness Reynolds number, so a two-scale plate test is the discriminating experiment.
Selection effect
H3090. Hull-cleaning trials select heavily fouled hulls because that is where a saving is measurable, so the reported saving does not apply to an average hull. A trial stratified by initial fouling level should show the saving falling with initial cleanliness.
Unit trap
H3091. Speed loss from fouling is quoted at fixed power, but at fixed revolutions the loss is smaller and at fixed speed the power penalty is larger, and the three are linked by the propeller curve. Reducing one trial's data all three ways is the test, and the three numbers should be mutually consistent only if the propeller curve is known.
Contamination
H3092. Marine growth on a riser's strakes changes both the effective diameter and the strake's suppression, so a measured VIV reduction credited to strakes may be the growth's own damping. Clean and grown strakes at the same reduced velocity should separate the two.
Feedback
H3093. An antifouling coating leaches biocide into a boundary layer that is depleted at low speed, so its effective lifetime depends on the ship's speed history rather than on elapsed time. Two coupons, one continuously stirred and one still, should release at different rates and fail at different times.
Lag
H3094. A fouling community's response to a coating lags the coating's leaching by the settlement window of the dominant recruit, so a coating that appears to fail in month eighteen may have failed in month six. Monthly settlement panels on the same coating should show the failure earlier than a hull survey does.
Threshold
H3095. There is a fouling thickness below which cleaning removes the community and above which it leaves a basal layer that regrows faster, because the removal mechanism changes from peeling to cutting. A thickness series cleaned by one method should show regrowth rate jumping at the threshold.

Scaling, testing and inference

Artefact mimicking effect
H3096. A towing carriage's structural vibration aliases into the resistance record when sampling is keyed to the position encoder, producing a spurious low-Froude hump. Recording the carriage's own accelerometer alongside the load cell should show the hump vanishing when the vibration is filtered at source.
Derived quantity
H3097. The correlation allowance is defined as the difference between trial and prediction, so it cannot test the prediction that produced it; an improvement in the allowance is an improvement in the fit, not in the physics. A blind prediction lodged before the trial is the only test.
Selection effect
H3098. Model-test databases over-represent hulls that were tested because someone doubted them, so the distribution of measured form factors is not the distribution of hulls afloat. Weighting by hulls built rather than hulls tested should shift the apparent population mean.
Inversion of standard practice
H3099. Model tests are used to predict full-scale behaviour, yet the largest uncertainty usually sits in the full-scale environment, so a short instrumented trial can constrain the model more than another model iteration. Running prediction and trial in parallel and comparing uncertainty bands, not means, is the test.
Bound
H3100. No tank comparison can resolve a difference smaller than the facility's run-to-run repeatability, and that repeatability is set less by the dynamometer than by the water's temperature history over the day. Two identical runs separated by a full diurnal cycle should disagree more than two back-to-back runs, which makes the bound measurable rather than assumed.

Status and plan

100 of a longer series. The subjects here are ship hydrodynamics, propulsion, seakeeping, moorings, risers, offshore structures and fouling; chemical degradation is deliberately absent because it is treated in a separate file.

The forms were tracked while writing so that repetition would be visible: no two entries share an opening, and the labels differ across roughly thirty logical types. None has been tested; a hypothesis list is worth the ideas in it, and nothing more, until someone builds the experiment.

If one of these deserves a tank slot, it is H3042 — that a polyester mooring's damping is mostly internal hysteresis and should survive a dry bench test. It needs a load frame and a rope, not a basin.

Incoprea Research · hand-considered hypotheses 3001–3100 · 3 November 2026 · ~/Desktop/Nov3/