Why is mathematics important in a grain silo? Grain looks like a solid pile, yet it can flow, rearrange and push against walls. Unlike a liquid, some of its weight is transferred to the wall through friction. Vertical stress can therefore approach a limiting value with depth instead of growing linearly without bound.
The Janssen model captures this surprising behaviour with a force balance and an exponential function. Bulk unit weight adds load; lateral stress pushes on the wall; wall friction carries part of the column’s weight; cross-sectional area and perimeter determine how strongly the wall participates. The result connects algebra, calculus and geometry to a practical granular system.
This is an educational static filling model, not a silo design procedure. Real silos experience nonuniform filling, discharge pressures, eccentric flow, consolidation, moisture, aeration, vibration and structural interaction. Current codes, measured material properties and qualified engineers are essential. Grain bins are also hazardous confined spaces; students should never enter or climb them.
- Separate grains from liquids
- Build the force balance
- Solve the exponential model
- Use geometry
- Work through an example
- Understand discharge and uncertainty
- Study limits
- Build student skills
- Read the FAQs
Why Grain Pressure Is Not Hydrostatic
A liquid reference model
For a stationary liquid of constant density rho, gauge pressure increases with depth z as p = rho g z. Each deeper horizontal layer supports the weight of liquid above, and wall shear is usually neglected in the hydrostatic balance.
If a liquid has unit weight gamma = rho g = 8 kilonewtons per cubic metre, pressure at 10 metres would be 80 kilopascals. This linear model is familiar and useful—but grain is not an ordinary liquid.
Granular contact networks
Grains touch one another at discrete contacts. Forces travel through irregular networks. The material can sustain shear and transfer horizontal pressure to walls. Friction at the wall supports part of the vertical load.
As depth increases, added grain weight is increasingly balanced by wall shear over the perimeter. The average vertical stress can approach an asymptote. This is often called the Janssen effect.
The grains near the bottom are not weightless. Rather, the wall carries a growing share of the overburden in the ideal model. The floor still supports a substantial load, and local stresses can be nonuniform.
Bulk density and unit weight
Bulk density is mass of grains divided by total occupied volume, including voids between grains. It differs from the solid material density of one kernel. Unit weight gamma = bulk density times gravitational acceleration.
If bulk density is 800 kilograms per cubic metre, gamma is about 800 times 9.81 = 7,848 newtons per cubic metre, or 7.85 kilonewtons per cubic metre.
Bulk density changes with grain type, moisture, filling method and consolidation. A value from one sample should not automatically represent a full silo.
Lateral stress ratio
The Janssen model relates average horizontal stress ph to average vertical stress pv through ph = K pv. K is a lateral stress ratio.
K is not necessarily one and is not a universal property. Its appropriate value depends on material state, friction, loading history and model assumptions. Soil mechanics uses active, passive and at-rest concepts, but a silo’s mobilised state during filling or discharge needs relevant data and standards.
Wall friction
Wall shear magnitude tau is modelled as mu ph, where mu is a wall-friction coefficient. Then tau = mu K pv.
Friction coefficient depends on wall material, roughness, grain condition, normal pressure and sliding state. The equation assumes the limiting friction is mobilised in a consistent direction. That assumption may not hold everywhere.
Arching
The transfer of load to walls is sometimes described as arching. It does not require one visible solid arch. It means internal stresses redirect load laterally so the boundary carries part of it.
Arching can also create flow problems near outlets, but outlet blockage and Janssen stress saturation are not identical phenomena. One static average model cannot describe every granular arch.
Did You Know? Doubling silo depth does not necessarily double bottom stress in the Janssen model. After several characteristic depths, the exponential curve becomes almost flat.
The Janssen Differential Equation
Slice the material
Consider a horizontal slice of thickness dz. Let cross-sectional area be A and wall perimeter U. Average vertical stress at depth z is pv(z). At z + dz it is pv + dpv.
The slice’s downward weight is gamma A dz. Wall friction acts upward with magnitude tau U dz = mu K pv U dz, using the average stress approximation.
Vertical force balance gives the increase in vertical force with depth:
A dpv = gamma A dz – mu K pv U dz.
Divide by A dz:
dpv/dz = gamma – mu K(U/A)pv.
Define hydraulic radius Rh = A/U. Then:
dpv/dz = gamma – (mu K/Rh)pv.
This first-order linear differential equation contains the mechanism. The first term adds weight. The second removes growth through wall support in proportion to current vertical stress.
Units check
pv is force per area, kilonewtons per square metre or kilopascals. dpv/dz has units kPa per metre, equivalent to kilonewtons per cubic metre. Gamma has the same units.
Mu and K are dimensionless. Rh is length, so (mu K/Rh)pv has stress per length. The equation is dimensionally consistent.
Boundary condition
At the free surface z = 0, the simplest model takes pv(0) = 0 gauge stress. Surface surcharge or a roof load would change the boundary condition.
Depth z is measured downward. This sign convention makes weight positive in the stress-growth equation. If depth were measured upward, signs would change. State the coordinate before differentiating.
Physical interpretation
At shallow depth, pv is small, so wall-friction term is small. The derivative is close to gamma, like hydrostatic growth.
As pv grows, wall support grows. At the limiting stress, gamma equals (mu K/Rh)pv, so derivative becomes zero. Added depth no longer increases the average vertical stress in the ideal asymptotic model.
Analogy to other systems
The equation has the same mathematical form as charging toward a limit, first-order cooling toward ambient, or a tank with proportional outflow:
dy/dx = input – coefficient times y.
Different physics can share one differential equation. The mathematics transfers, while parameter meanings do not.
Stress Saturation with Depth
Solution
Let lambda = Rh/(mu K), a characteristic depth. Solving with pv(0) = 0 gives:
pv(z) = gamma lambda [1 – exp(-z/lambda)].
Horizontal stress is:
ph(z) = K pv(z) = gamma Rh/mu [1 – exp(-mu K z/Rh)].
The limiting vertical stress as z becomes large is pv,infinity = gamma lambda = gamma Rh/(mu K). Limiting horizontal stress is ph,infinity = gamma Rh/mu.
Characteristic depth for this silo
At z = lambda, stress reaches 1 – e to the power -1 = about 63.2 percent of its limiting value. At 2 lambda, it reaches 86.5 percent. At 3 lambda, 95.0 percent. At 4 lambda, 98.2 percent.
This gives a useful scale. Saying a silo is “deep” in the model means its fill depth is several characteristic depths, not merely many metres.
Shallow-depth limit
For small x, exp(-x) approximately equals 1 – x. With x = z/lambda:
pv approximately equals gamma lambda times z/lambda = gamma z.
Near the surface, the Janssen model approaches the linear overburden stress. Wall friction has had little depth over which to accumulate.
Effect of wall friction
Increasing mu decreases lambda and lowers limiting stresses. More friction means the wall carries weight over a shorter depth.
This should not be turned into “rougher walls are always safer.” Greater wall shear itself loads the structure, flow behaviour changes, and discharge effects may intensify. Design needs the full load case.
Effect of lateral ratio
Increasing K decreases vertical limiting stress because horizontal stress and wall friction mobilise more strongly. Interestingly, ph,infinity = gamma Rh/mu is independent of K in this simplified asymptote, because K cancels.
At finite depth K still affects the exponential approach. Model sensitivity should be evaluated rather than inferred from one limit.
Effect of geometry
Larger Rh increases characteristic depth and limiting stress. A wide silo has less wall perimeter per unit area, so the wall supports a smaller fraction of each slice’s weight.
This geometric insight is central: wall effects scale with perimeter, while weight scales with area.
Historical source
Janssen’s 1895 experiments and derivation have been translated and discussed in Experiments on Corn Pressure in Silo Cells. The historical study is foundational evidence, but modern design uses current standards and later research rather than the original equation alone.
Hydraulic Radius and Silo Shape
Definition
Hydraulic radius here is cross-sectional area divided by wall perimeter: Rh = A/U. It has units of length. It is not the same as physical radius except for particular shapes.
For a circular silo of diameter D, A = pi D squared/4 and U = pi D. Therefore Rh = D/4. For physical radius R = D/2, Rh = R/2.
For a square side b, A = b squared and U = 4b, so Rh = b/4. A circle and square with the same hydraulic radius have different detailed stress fields, even if the one-dimensional equation looks similar.
Circular example
For D = 6 metres, Rh = 1.5 metres. If mu = 0.4 and K = 0.5, characteristic depth lambda = 1.5/(0.4 times 0.5) = 7.5 metres.
For D = 12 metres with the same material assumptions, Rh doubles to 3 metres and lambda doubles to 15 metres. A 10-metre fill is relatively shallow in characteristic-depth terms for the wider silo.
Rectangular example
For dimensions 4 metres by 8 metres, A = 32 square metres and U = 24 metres, giving Rh = 1.333 metres. A square with equal area has side square root 32 = 5.657 metres and perimeter 22.627, giving Rh = 1.414 metres.
The square has slightly smaller perimeter for the same area and therefore slightly larger Rh. But corners and wall flexibility make actual stress distributions more complex than the average model.
Perimeter-to-area ratio
The coefficient U/A = 1/Rh directly controls wall-support term. Small cells have large perimeter relative to area, so wall effects are strong. Large cells behave more hydrostatically over greater depths.
This is a recurring scaling law. Surface or boundary effects often scale differently from volume effects. As size grows, the balance changes.
Equivalent diameter
Engineers sometimes express a noncircular section through an equivalent hydraulic diameter Dh = 4A/U = 4Rh. This does not make a rectangular silo physically circular; it only preserves area-to-perimeter ratio for a specific calculation.
Hopper geometry
The vertical-wall derivation assumes constant cross-section. A hopper narrows with depth and redirects forces along sloping walls. Its geometry requires separate balances and current design provisions.
Do not extend Rh from the cylinder unchanged into a cone. Area, perimeter, wall angle and flow state vary.
Worked Example: A Six-Metre-Diameter Silo
Consider an ideal circular vertical silo with diameter D = 6.0 m, fill depth H = 10.0 m, bulk unit weight gamma = 8.0 kN/m3, wall friction coefficient mu = 0.40 and lateral stress ratio K = 0.50. These are illustrative, not design values.
Geometry
Area A = pi D squared/4 = 9pi = 28.274 square metres. Perimeter U = pi D = 18.850 metres. Hydraulic radius Rh = A/U = 1.500 metres.
Characteristic depth
Lambda = Rh/(mu K) = 1.5/(0.4 times 0.5) = 7.5 metres.
Fill depth ratio H/lambda = 10/7.5 = 1.333. Exponential factor exp(-1.333) is about 0.2636. Saturation fraction is 1 – 0.2636 = 0.7364.
Vertical stress at the base level
Limiting vertical stress pv,infinity = gamma lambda = 8 times 7.5 = 60 kPa.
At 10 metres, pv = 60 times 0.7364 = 44.18 kPa.
A no-wall-friction overburden estimate gamma H would be 8 times 10 = 80 kPa. The Janssen average is lower because wall shear supports part of the weight.
Horizontal wall stress
ph = K pv = 0.5 times 44.18 = 22.09 kPa at 10 metres. Limiting horizontal stress is gamma Rh/mu = 8 times 1.5/0.4 = 30 kPa.
Horizontal stress rises with depth toward 30 kPa in this model. It is not uniform over height.
Average floor force
If base stress were uniform at the average pv, vertical force on the flat base would be pv A = 44.18 kN/m2 times 28.274 m2 = about 1,249 kN.
Total grain weight is gamma A H = 8 times 28.274 times 10 = about 2,262 kN. Difference, about 1,013 kN, is supported through integrated wall shear in the ideal equilibrium.
This is a force-balance check. A real base stress is not necessarily uniform, and hopper or structural details change the distribution.
Integrated wall shear check
Wall shear at depth z is tau = mu K pv(z). Its vertical resultant over height is integral tau U dz. Instead of integrating numerically, conservation says total wall support = weight – base force = about 1,013 kN.
Students can integrate the exponential and verify the same result. Agreement checks both calculus and units.
Stress at characteristic depths
At z = 7.5 m, pv = 60(1 – e^-1) = 37.93 kPa. At z = 15 m, pv = 51.88 kPa. At z = 22.5 m, pv = 57.01 kPa.
Adding 7.5 metres raises stress by 37.93 kPa in the first interval, 13.95 in the second and 5.13 in the third. The increments shrink exponentially.
Sensitivity to wall friction
If mu falls from 0.40 to 0.30, lambda = 1.5/(0.30 times 0.50) = 10 m. Limiting pv becomes 80 kPa. At z = 10 m, pv = 80(1 – e^-1) = 50.57 kPa.
Lower friction increases base stress in the model because less weight is carried by the wall. But the wall shear distribution and flow behaviour also change.
Sensitivity to bulk density
If gamma increases 5 percent while mu, K and geometry stay fixed, stresses increase 5 percent because gamma is a linear multiplier. In reality denser packing may also change K or friction, so independent variation is a local sensitivity, not a complete prediction.
Results table
| Quantity | Value | Interpretation |
|---|---|---|
| Hydraulic radius | 1.50 m | Area/perimeter for 6 m circle |
| Characteristic depth | 7.50 m | 63.2% of limit at this depth |
| Limiting vertical stress | 60.0 kPa | Ideal asymptote |
| Vertical stress at 10 m | 44.18 kPa | 73.64% of limit |
| Horizontal stress at 10 m | 22.09 kPa | K times vertical stress |
| Ideal base force | About 1,249 kN | Uniform-average illustration |
Filling, Discharge and Measurement
Filling history
Rain filling, central filling and off-centre filling create different packing and stress paths. Falling grain can create impact and density variations. The static model assumes a representative uniform state.
Stress may depend on whether friction is fully mobilised and in which direction. Janssen parameters are therefore effective model inputs, not simple constants detached from procedure.
Discharge pressures
When an outlet opens, flow zones form and stresses redistribute. Dynamic wall pressures can exceed or differ from static filling predictions. Eccentric discharge can create asymmetric loads.
A silo safe under one static calculation is not thereby safe during discharge. Modern codes specify load cases and factors; those provisions must be followed.
Mass flow and funnel flow
In mass flow, much of the stored material moves when discharge occurs. In funnel flow, a channel moves while surrounding material may remain stagnant. Geometry, wall friction and material flow properties influence the regime.
Flow regime affects residence time, segregation, wall loads and reliability. Predicting it requires more than the vertical Janssen equation.
Pressure sensors
Wall pressure cells measure local interaction and can be sensitive to installation stiffness, size and calibration. A sensor may disturb the stress field it measures. Grain contacts create fluctuations.
Average multiple sensors and report variability. One reading should not be treated as the cross-sectional mean without evidence.
Weighing and load cells
Base or leg load cells can estimate stored mass, but wall friction and structural load paths matter. Temperature, creep and support reactions can change readings.
Mass inventory may combine load cells, level measurements and calibration curves. Each has uncertainty.
Parameter estimation
Given measured pv versus depth, one can fit pv = Pinf(1 – exp(-z/lambda)). Then infer combinations of gamma, Rh, mu and K. The data may identify Pinf and lambda better than mu and K separately.
This is an identifiability problem. Because lambda contains the product mu K, stress data alone may not distinguish the two without independent measurements.
Uncertainty propagation
Limiting vertical stress pv,infinity = gamma Rh/(mu K). For small independent relative uncertainties, squared relative uncertainty is approximately the sum of squared relative uncertainties in gamma, Rh, mu and K.
If gamma and Rh each have 3 percent uncertainty, mu 10 percent and K 12 percent, combined is square root(0.03 squared + 0.03 squared + 0.10 squared + 0.12 squared) = about 16.2 percent.
The friction and lateral-ratio assumptions dominate. Reporting a limiting stress to four decimal places would be misleading.
What the Janssen Model Leaves Out
Nonuniform stresses
The model uses cross-sectional averages. Real force chains create local variations. Wall stiffness and shape influence distribution.
Material anisotropy
Deposited grains can develop direction-dependent fabric. One constant K may not describe all depths or load paths.
Consolidation and time
Stored grain can settle, creep or change moisture. Bulk density and friction evolve. Temperature gradients can cause movement and condensation.
Flexible walls
Wall deformation changes geometry and stress interaction. A rigid-wall assumption may not hold for thin corrugated bins.
Discharge dynamics
Flowing material produces transient and asymmetric loads, especially near transitions and eccentric outlets. Static pressure saturation is insufficient.
Hopper transition
The cylinder-to-hopper junction can concentrate stress. Sloping walls introduce normal and tangential components not in the constant-section slice.
Aeration and pore pressure
Airflow through grain creates pressure gradients and can alter effective stresses. Rapid filling or discharge may involve air effects.
Structural response
Pressure is only a load. Wall hoop force, buckling, stiffeners, foundation, fatigue and connections require structural analysis and load combinations.
Model-scope table
| Janssen assumption | Mathematical benefit | Real extension |
|---|---|---|
| Constant cross-section | One hydraulic radius | Hoppers and variable geometry |
| Constant gamma, mu and K | Closed-form exponential | Depth and history dependence |
| Cross-sectional average | One stress value per depth | Local force chains and asymmetry |
| Static equilibrium | Simple force balance | Filling impacts and discharge dynamics |
| Fully mobilised wall friction | Predictable wall support | Partial or reversed mobilisation |
| Rigid walls | Fixed geometry | Wall-structure interaction |
Misconceptions Worth Correcting
“Grain pressure is just rho g h”
Near the surface it can resemble linear overburden, but wall friction redirects load. The Janssen model predicts saturation with depth.
“Pressure stops increasing because lower grain has no weight”
No. Added weight is balanced increasingly by wall shear. The load path changes.
“Hydraulic radius equals silo radius”
For a circular cross-section, Rh = D/4 = physical radius/2. It is area divided by perimeter.
“More wall friction always reduces every structural load”
It lowers base stress in the ideal model but increases the share transferred to walls. Static and dynamic load cases still need evaluation.
“A fitted curve proves mu and K separately”
The characteristic depth contains their product. Additional measurements or assumptions are needed to identify them individually.
“The static equation designs the outlet”
No. Outlet flow, arching, ratholing, discharge rate and dynamic loads need granular-flow and structural analysis.
Normalising and Testing the Model
Build a dimensionless view
After students understand the dimensional equation, they can make the pattern clearer by defining \(\zeta=z/\lambda\). The normalised pressure is
\[ \frac{p_v}{\gamma\lambda}=1-e^{-\zeta}. \]
This compact form says that silos with very different dimensions can share the same curve when depth is measured in units of \(\lambda\). At \(z=\lambda\), pressure has reached about 63.2% of its asymptote; at \(2\lambda\), about 86.5%; and at \(3\lambda\), about 95.0%. These landmarks give students a fast reasonableness check without a calculator.
The dimensionless form also separates shape from scale. The curve \(1-e^{-\zeta}\) is universal inside the Janssen assumptions, while \(\gamma\lambda\) supplies the pressure scale. This is a transferable modelling habit: identify a characteristic length or time, then ask whether apparently different systems collapse onto one normalised relationship.
Compare sensitivity before refining inputs
The limiting pressure can be rewritten as
\[ p_{v,\infty}=\frac{\gamma R_h}{\mu K}. \]
This exposes proportional relationships. A 10% increase in bulk unit weight or hydraulic radius gives a 10% increase in ideal limiting pressure if other inputs stay fixed. A 10% increase in \(\mu\) or \(K\) does not give a 10% decrease exactly; the new multiplier is \(1/1.10\), about 0.909, so the reduction is about 9.1%.
Students can use this to rank uncertainties. If wall friction is poorly known while diameter is measured accurately, refining diameter to another decimal place may contribute less than testing a realistic friction range. Sensitivity analysis therefore supports better measurement choices, not just more calculations.
Check limiting cases
A strong model should behave sensibly at its mathematical edges. At the surface, \(z=0\), the exponential is one and the predicted vertical stress is zero. For very small depth, the approximation \(e^{-z/\lambda}\approx1-z/\lambda\) gives \(p_v\approx\gamma z\). Near the top, therefore, the Janssen equation initially resembles the hydrostatic line. Wall transfer becomes increasingly visible as depth grows.
If wall friction tends towards zero, \(\lambda=R_h/(\mu K)\) grows very large. The apparent saturation moves far away and the shallow-depth hydrostatic form dominates. If the silo becomes wider while other properties stay fixed, \(R_h\) and \(\lambda\) increase, so a deeper column is needed before the curve bends strongly. These limits agree with the physical story and help reveal algebraic mistakes.
The equation should not be pushed carelessly to impossible parameters. Negative friction, negative unit weight or \(K=0\) do not describe the intended passive granular system. A singular expression is sometimes mathematics warning us that an assumption has left its valid domain.
Turn a curve into a design question
Suppose two candidate bins store the same material. One has twice the hydraulic radius but identical \(\mu\) and \(K\). Its characteristic depth and limiting vertical pressure are both twice as large. Yet its wall perimeter and floor area also change, so total force does not follow from pressure alone. Students should calculate both stress and the area on which it acts.
That distinction is crucial in engineering: stress is force per area, while structural members respond to integrated loads, bending moments and local concentrations. The Janssen curve is therefore an input to a wider structural calculation, not a complete design certificate.
How Students Can Learn This Mathematics
Compare liquid and grain models
Plot p = gamma z and the Janssen exponential with the same gamma. Explain why the curves begin together and separate.
Derive the slice balance
Draw a differential slice. Label top and bottom stress forces, weight and wall shear. Divide carefully by area and thickness.
Solve the differential equation
Use integrating factors or separation after rewriting. Apply pv(0) = 0. Differentiate the solution to verify it satisfies the equation.
Explore dimensionless depth
Plot pv/pv,infinity against z/lambda. Every parameter set collapses onto 1 – e^-x in the ideal model. This is a powerful nondimensionalisation.
Compare shapes
Calculate Rh for circles, squares and rectangles of equal area. Discuss perimeter effects and why identical Rh does not guarantee identical local stress.
Make a safe model
Use a small transparent container with dry rice under teacher supervision and external force sensing only. Never enter, climb or reach into real bins. A tabletop model demonstrates trends, not structural loads.
Fit synthetic data
Generate noisy stress-depth values and estimate Pinf and lambda. Plot residuals. Observe parameter correlation.
Run sensitivity analysis
Vary mu and K by plausible percentages. Identify which uncertainty dominates. Avoid presenting one nominal curve as certainty.
Connect related mathematics
Wastewater networks use gravity flow but treat a fluid, as shown in Why Mathematics? | Wastewater Networks, Pipe Slopes and Gravity Flow. Comparing the two clarifies why granular shear changes the governing equation.
Keep pathways open
Students may explore civil, structural, agricultural, mechanical or process engineering, food science, granular physics and safety. Mathematics supports these pathways but does not guarantee admission or professional responsibility. Verify current official programme requirements.
For Singapore cohorts, check current MOE and SEAB terminology for Posting Groups, Full Subject-Based Banding, G1/G2/G3 subjects and the Singapore-Cambridge Secondary Education Certificate. Where applicable, use SEC Additional Mathematics Examination for G2/G3 rather than applying older labels automatically.
Guidance for Parents and Teachers
Begin with the surprising observation: grain pressure need not remain hydrostatic. Let students propose mechanisms before showing the wall-friction balance.
Insist on a free-body diagram. The exponential formula is meaningful only when weight, wall shear and vertical stress forces are visible.
Use units to distinguish bulk unit weight from density. Kilograms per cubic metre is mass density; kilonewtons per cubic metre is unit weight.
Ask for limits. At z = 0 stress should be zero in the simple boundary condition. At shallow depth it should approach gamma z. At great depth it should approach a finite asymptote.
Highlight safety. Grain can engulf, suffocate and trap people; machinery and dust add hazards. Classroom mathematics must never encourage entry into a silo or bin.
Separate model insight from design authority. A correct Janssen calculation is not a building approval. Modern codes and qualified engineers assess all load cases.
Value honest uncertainty. Friction and stress ratios are difficult to know exactly. Sensitivity bands teach more than decorative decimals.
Frequently Asked Questions
What is the Janssen effect?
It is the tendency for granular vertical stress to approach a limiting value because wall friction supports part of the material’s weight.
What is hydraulic radius in a silo?
Cross-sectional area divided by wall perimeter. For a circular silo it equals diameter divided by four.
Why is grain different from water?
Grains sustain shear and transfer horizontal stress through contact networks. Wall friction redirects load, while a simple hydrostatic liquid model neglects such wall support.
What does K represent?
It relates average horizontal stress to average vertical stress through ph = Kpv. Its appropriate value depends on material state and loading history.
What does lambda mean?
Lambda = Rh/(mu K) is the characteristic depth. At that depth vertical stress reaches about 63.2 percent of its limiting value.
Does pressure become exactly constant after three lambda?
No. The exponential approaches the limit continuously. At three lambda it is about 95 percent of the limit.
Why does a wider silo have greater bottom stress?
It has more area relative to perimeter. Wall support is less effective per unit stored weight, so characteristic depth and limiting vertical stress increase.
Can the model predict discharge pressure?
Not reliably by itself. Discharge changes flow and stress, often creating dynamic and asymmetric loads.
Can tabletop rice confirm full-scale safety?
No. Scale, friction, particle size, wall stiffness and dynamics differ. It is only a learning demonstration.
Is silo work dangerous?
Yes. Engulfment, confined spaces, machinery and dust can be fatal. Only trained authorised personnel following current safety procedures should work around or inside storage structures.
Useful Next Reading
- Experiments on Corn Pressure in Silo Cells: translation and comment on Janssen’s paper
- University of Edinburgh thesis discussing silo pressure models
- University of Alabama at Birmingham: Silo Analysis and Design
- Why Mathematics? | Wastewater Networks, Pipe Slopes and Gravity Flow
- Why Mathematics? | Concrete Mix Design, Water–Cement Ratio and Strength
The mathematics of a grain silo begins with one thin slice. Weight pushes down, the wall pushes back through friction, and an exponential curve emerges. That small force balance explains why boundary geometry can reshape an entire load path. Just as importantly, the limits of the model teach students when a beautiful equation must hand responsibility to measurement, standards and professional judgement.