Chapter map

1. 1. Grid
2. 2. Kernel
3. 3. Velocity
4. 4. Interpolate
5. 5. Force
6. 6. Spread one
7. 7. Spread two
8. 8. Force total
9. 9. Work left
10. 10. Work right
11. 11. Move marker
12. 12. Spacing
13. 13. Time coupling
14. 14. Conclusion

Visit the How Mathematics Works hub for the wider series and its connections to education, reasoning, science, computing, design and careers.

---

1. Grid

This is where modelling choices begin to affect evidence. Use nodes x1=1,x2=2 with h=1.

A concrete situation

Boundary X=1.25.

What the mathematics does not license

Teaching analogue. That distinction matters because a correct formula can still answer the wrong operational question.

Try it, check it, transfer it

Draw positions. Then explain the result to someone who has not seen the formula. If the explanation needs a hidden assumption, add that assumption to the model record.

---

2. Kernel

The central mathematical idea is concrete. Use linear hat weights.

A concrete situation

Weights .75 and .25.

What the mathematics does not license

They sum to one. The habit of checking assumptions is one of the practical benefits of learning mathematics.

Try it, check it, transfer it

Record weights. Then explain the result to someone who has not seen the formula. If the explanation needs a hidden assumption, add that assumption to the model record.

---

3. Velocity

Here the geometry becomes operational. Set u1=2,u2=6.

A concrete situation

Grid values.

What the mathematics does not license

Keep units. The same reasoning helps students move from an exercise to a real dataset.

Try it, check it, transfer it

Record data. Then explain the result to someone who has not seen the formula. If the explanation needs a hidden assumption, add that assumption to the model record.

---

4. Interpolate

This step deserves a careful pause. Compute .75(2)+.25(6).

A concrete situation

Boundary U=3.

What the mathematics does not license

Use same kernel convention. That distinction matters because a correct formula can still answer the wrong operational question.

Try it, check it, transfer it

Record U. Then explain the result to someone who has not seen the formula. If the explanation needs a hidden assumption, add that assumption to the model record.

---

5. Force

The useful question is not only how to calculate, but what the calculation assumes. Set boundary F=4.

A concrete situation

Lagrangian force.

What the mathematics does not license

State direction. The habit of checking assumptions is one of the practical benefits of learning mathematics.

Try it, check it, transfer it

Record F. Then explain the result to someone who has not seen the formula. If the explanation needs a hidden assumption, add that assumption to the model record.

---

6. Spread one

A small example makes the mechanism visible. Compute .75F.

A concrete situation

f1=3.

What the mathematics does not license

Include h normalization when h differs. The same reasoning helps students move from an exercise to a real dataset.

Try it, check it, transfer it

Record f1. Then explain the result to someone who has not seen the formula. If the explanation needs a hidden assumption, add that assumption to the model record.

---

7. Spread two

This is where modelling choices begin to affect evidence. Compute .25F.

A concrete situation

f2=1.

What the mathematics does not license

Same kernel. That distinction matters because a correct formula can still answer the wrong operational question.

Try it, check it, transfer it

Record f2. Then explain the result to someone who has not seen the formula. If the explanation needs a hidden assumption, add that assumption to the model record.

---

8. Force total

The central mathematical idea is concrete. Add grid forces.

A concrete situation

Total=4.

What the mathematics does not license

Partition of unity. The habit of checking assumptions is one of the practical benefits of learning mathematics.

Try it, check it, transfer it

Record check. Then explain the result to someone who has not seen the formula. If the explanation needs a hidden assumption, add that assumption to the model record.

---

9. Work left

Here the geometry becomes operational. Compute F U.

A concrete situation

Value=12.

What the mathematics does not license

Instantaneous analogue. The same reasoning helps students move from an exercise to a real dataset.

Try it, check it, transfer it

Record work. Then explain the result to someone who has not seen the formula. If the explanation needs a hidden assumption, add that assumption to the model record.

---

10. Work right

This step deserves a careful pause. Compute f1u1+f2u2.

A concrete situation

Value=12.

What the mathematics does not license

Adjoint consistency. That distinction matters because a correct formula can still answer the wrong operational question.

Try it, check it, transfer it

Record check. Then explain the result to someone who has not seen the formula. If the explanation needs a hidden assumption, add that assumption to the model record.

---

11. Move marker

The useful question is not only how to calculate, but what the calculation assumes. Set X=1.75.

A concrete situation

Weights swap.

What the mathematics does not license

Recalculate. The habit of checking assumptions is one of the practical benefits of learning mathematics.

Try it, check it, transfer it

Record result. Then explain the result to someone who has not seen the formula. If the explanation needs a hidden assumption, add that assumption to the model record.

---

12. Spacing

A small example makes the mechanism visible. Change h.

A concrete situation

Kernel normalization changes.

What the mathematics does not license

Track dimensions. The same reasoning helps students move from an exercise to a real dataset.

Try it, check it, transfer it

Record formula. Then explain the result to someone who has not seen the formula. If the explanation needs a hidden assumption, add that assumption to the model record.

---

13. Time coupling

This is where modelling choices begin to affect evidence. Advance fluid and marker.

A concrete situation

Explicit stiffness may appear.

What the mathematics does not license

Check stability. That distinction matters because a correct formula can still answer the wrong operational question.

Try it, check it, transfer it

Record step. Then explain the result to someone who has not seen the formula. If the explanation needs a hidden assumption, add that assumption to the model record.

---

14. Conclusion

The central mathematical idea is concrete. State grid,markers,kernel,force law and validation.

A concrete situation

Worked U=3,total force=4,work=12.

What the mathematics does not license

Analogue is one-dimensional. The habit of checking assumptions is one of the practical benefits of learning mathematics.

Try it, check it, transfer it

Write conclusion. Then explain the result to someone who has not seen the formula. If the explanation needs a hidden assumption, add that assumption to the model record.

---

A seven-day practice lab

1. Day 1: define the decision, observation unit, X variables and Y or reference condition.
2. Day 2: reproduce the smallest worked example and check every dimension, unit and sign.
3. Day 3: build a transparent baseline and write what it assumes.
4. Day 4: fit the main method inside a training-only pipeline; record every parameter.
5. Day 5: challenge one assumption with a deliberately difficult dataset or operating regime.
6. Day 6: compare validation metrics with the practical workload created by errors.
7. Day 7: present one supported conclusion, one unresolved uncertainty and one safe next experiment.

Questions for a student, parent or mentor

• Which part of the model is a mathematical definition, and which part is a choice?
• What small calculation would catch a sign, scale or dimension error?
• What evidence comes from training, validation and the untouched test?
• What benign condition could look like a fault or prediction failure?
• How would the answer change for another response, asset, school project or time period?
• What should a human review before acting on the score?

A method-specific audit before deployment

The immersed boundary method couples a boundary described by moving material points to fluid variables stored on a fixed Eulerian grid. Regularised delta kernels spread boundary forces to the grid and interpolate grid velocity back to the boundary.

In a one-dimensional teaching analogue with grid spacing one, place a boundary point at X=1.25 between nodes x1=1 and x2=2. Use linear hat weights 0.75 and 0.25.

If grid velocities are u1=2 and u2=6, interpolation gives U=0.75 times 2 plus 0.25 times 6=3. The boundary point moves with a weighted local fluid velocity.

If the boundary exerts force F=4, spread it with the same weights: f1=3 and f2=1. The grid force sum is four, so total force is preserved in this analogue.

Using the same kernel in both directions creates a discrete work relationship: F times U equals f1 times u1 plus f2 times u2, and both sides equal twelve in the worked example.

Peskin’s 1972 method represented a moving heart-valve boundary by forces on a rectangular fluid mesh, linked through a semi-discrete delta function.

Did You Know? The fluid grid does not have to conform to the boundary. This avoids remeshing as a flexible structure moves, but spreads a sharp interface over a few grid cells.

The Eulerian fluid variables and Lagrangian boundary variables have different coordinates. Accurate coupling therefore depends on kernel support, moment conditions and grid resolution.

Boundary forces may come from fibres, springs, bending energies or constitutive laws. A numerically stable coupling cannot validate an incorrect material model.

Explicit coupling can become stiff when structures are strong or light. Implicit and semi-implicit schemes trade additional solves for larger stable time steps.

Leakage can occur if boundary markers are too far apart relative to the fluid grid. Marker spacing, kernel width and incompressibility solve must be studied together.

Students can verify partition of unity, first moments, total spread force and the discrete work identity on tiny grids. These checks reveal indexing and normalization errors.

Applications include heart valves, swimming organisms, flexible fibres, particulate flow and fluid-structure interaction. Contact and high Reynolds number flow require additional care.

A diffuse numerical interface is not the same as a physically thick boundary. Report the kernel footprint and test how results change with refinement.

For parents and students, the transferable lesson is that two descriptions can cooperate through a carefully balanced translation that preserves totals in both directions.

A responsible report states the fluid equations, grid, boundary markers, force law, delta kernel, interpolation, spreading, coupling scheme, time step, leakage tests, convergence and validation.

---

Frequently asked questions

What is the immersed boundary method?

A fluid–structure method coupling moving boundary points to a fixed fluid grid through regularised delta kernels.

Why avoid a body-fitted mesh?

The structure can move and deform without rebuilding the fluid grid around it.

What does force spreading do?

It transfers Lagrangian boundary force to nearby Eulerian grid nodes.

What does velocity interpolation do?

It estimates boundary velocity from nearby grid velocities.

What can cause leakage?

Marker spacing, kernel support, grid resolution and the fluid solver can leave gaps in effective coupling.

What must be reported?

Fluid equations, grid, markers, force law, kernel, coupling, time scheme, leakage, convergence and validation.

---

Primary source and fact-check note

Charles S. Peskin, “Flow Patterns Around Heart Valves: A Numerical Method,” Journal of Computational Physics 10, 1972. Definitions and core claims in this article were checked against this primary publication. Calculations shown here were independently recomputed. Examples are educational illustrations; they are not reported industrial performance results.

Useful next reading

• Finite difference method
• Finite volume method
• Adaptive mesh refinement
• Numerical analysis hub
• Study and learning methods

Keep the options open. Mathematics education supports clearer questions, transferable problem-solving skills and informed choices across science, engineering, computing, AI, statistics, design and many other careers. It does not guarantee intelligence, income, admission or employment; those outcomes depend on far more than one subject or model.

Method audit: representation, transfer and evidence

Kernel audit

Check partition of unity, moments and support.

Transfer audit

Verify interpolation, spreading and discrete work consistency.

Resolution audit

Compare marker spacing with fluid-grid spacing.

Coupling audit

Test explicit or implicit stability, leakage and convergence.

Mechanism workshop: calculate, challenge and transfer

Weight table

Calculate hat-kernel weights at several marker positions.

Constant test

Interpolate a constant velocity exactly.

Force test

Spread force and sum grid values.

Work test

Compare Lagrangian and Eulerian power.

Marker sweep

Move X across a cell and check smoothness.

Kernel swap

Compare support and moment conditions.

Leak test

Increase marker spacing and measure crossing flow.

Stiffness test

Strengthen springs and find stable time steps.

Student explanation

Explain two grids translating force and velocity.

Parent explanation

Relate balanced translation to teamwork.

Evidence notebook: reconstruct every step

Evidence notebook 1: Grid

Step 1 isolates grid. Use nodes x1=1,x2=2 with h=1. The checked result is: Boundary X=1.25. Keep dimensions, kernel or lattice conventions, node and direction order, normals, sign, time level and precision beside every number. The safeguard is: Teaching analogue. Draw positions. Change one input, recompute the local rule and global invariant, and explain whether the difference comes from quadrature, collision, streaming, interface transport, force transfer, spatial resolution, time integration, rounding or a changed model. Ask a second reader to reproduce the arithmetic and finish with the narrow claim the evidence supports.

Evidence notebook 2: Kernel

Step 2 isolates kernel. Use linear hat weights. The checked result is: Weights .75 and .25. Keep dimensions, kernel or lattice conventions, node and direction order, normals, sign, time level and precision beside every number. The safeguard is: They sum to one. Record weights. Change one input, recompute the local rule and global invariant, and explain whether the difference comes from quadrature, collision, streaming, interface transport, force transfer, spatial resolution, time integration, rounding or a changed model. Ask a second reader to reproduce the arithmetic and finish with the narrow claim the evidence supports.

Evidence notebook 3: Velocity

Step 3 isolates velocity. Set u1=2,u2=6. The checked result is: Grid values. Keep dimensions, kernel or lattice conventions, node and direction order, normals, sign, time level and precision beside every number. The safeguard is: Keep units. Record data. Change one input, recompute the local rule and global invariant, and explain whether the difference comes from quadrature, collision, streaming, interface transport, force transfer, spatial resolution, time integration, rounding or a changed model. Ask a second reader to reproduce the arithmetic and finish with the narrow claim the evidence supports.

Evidence notebook 4: Interpolate

Step 4 isolates interpolate. Compute .75(2)+.25(6). The checked result is: Boundary U=3. Keep dimensions, kernel or lattice conventions, node and direction order, normals, sign, time level and precision beside every number. The safeguard is: Use same kernel convention. Record U. Change one input, recompute the local rule and global invariant, and explain whether the difference comes from quadrature, collision, streaming, interface transport, force transfer, spatial resolution, time integration, rounding or a changed model. Ask a second reader to reproduce the arithmetic and finish with the narrow claim the evidence supports.

Evidence notebook 5: Force

Step 5 isolates force. Set boundary F=4. The checked result is: Lagrangian force. Keep dimensions, kernel or lattice conventions, node and direction order, normals, sign, time level and precision beside every number. The safeguard is: State direction. Record F. Change one input, recompute the local rule and global invariant, and explain whether the difference comes from quadrature, collision, streaming, interface transport, force transfer, spatial resolution, time integration, rounding or a changed model. Ask a second reader to reproduce the arithmetic and finish with the narrow claim the evidence supports.

Evidence notebook 6: Spread one

Step 6 isolates spread one. Compute .75F. The checked result is: f1=3. Keep dimensions, kernel or lattice conventions, node and direction order, normals, sign, time level and precision beside every number. The safeguard is: Include h normalization when h differs. Record f1. Change one input, recompute the local rule and global invariant, and explain whether the difference comes from quadrature, collision, streaming, interface transport, force transfer, spatial resolution, time integration, rounding or a changed model. Ask a second reader to reproduce the arithmetic and finish with the narrow claim the evidence supports.

Evidence notebook 7: Spread two

Step 7 isolates spread two. Compute .25F. The checked result is: f2=1. Keep dimensions, kernel or lattice conventions, node and direction order, normals, sign, time level and precision beside every number. The safeguard is: Same kernel. Record f2. Change one input, recompute the local rule and global invariant, and explain whether the difference comes from quadrature, collision, streaming, interface transport, force transfer, spatial resolution, time integration, rounding or a changed model. Ask a second reader to reproduce the arithmetic and finish with the narrow claim the evidence supports.

Evidence notebook 8: Force total

Step 8 isolates force total. Add grid forces. The checked result is: Total=4. Keep dimensions, kernel or lattice conventions, node and direction order, normals, sign, time level and precision beside every number. The safeguard is: Partition of unity. Record check. Change one input, recompute the local rule and global invariant, and explain whether the difference comes from quadrature, collision, streaming, interface transport, force transfer, spatial resolution, time integration, rounding or a changed model. Ask a second reader to reproduce the arithmetic and finish with the narrow claim the evidence supports.

Evidence notebook 9: Work left

Step 9 isolates work left. Compute F U. The checked result is: Value=12. Keep dimensions, kernel or lattice conventions, node and direction order, normals, sign, time level and precision beside every number. The safeguard is: Instantaneous analogue. Record work. Change one input, recompute the local rule and global invariant, and explain whether the difference comes from quadrature, collision, streaming, interface transport, force transfer, spatial resolution, time integration, rounding or a changed model. Ask a second reader to reproduce the arithmetic and finish with the narrow claim the evidence supports.

Evidence notebook 10: Work right

Step 10 isolates work right. Compute f1u1+f2u2. The checked result is: Value=12. Keep dimensions, kernel or lattice conventions, node and direction order, normals, sign, time level and precision beside every number. The safeguard is: Adjoint consistency. Record check. Change one input, recompute the local rule and global invariant, and explain whether the difference comes from quadrature, collision, streaming, interface transport, force transfer, spatial resolution, time integration, rounding or a changed model. Ask a second reader to reproduce the arithmetic and finish with the narrow claim the evidence supports.

Evidence notebook 11: Move marker

Step 11 isolates move marker. Set X=1.75. The checked result is: Weights swap. Keep dimensions, kernel or lattice conventions, node and direction order, normals, sign, time level and precision beside every number. The safeguard is: Recalculate. Record result. Change one input, recompute the local rule and global invariant, and explain whether the difference comes from quadrature, collision, streaming, interface transport, force transfer, spatial resolution, time integration, rounding or a changed model. Ask a second reader to reproduce the arithmetic and finish with the narrow claim the evidence supports.

Evidence notebook 12: Spacing

Step 12 isolates spacing. Change h. The checked result is: Kernel normalization changes. Keep dimensions, kernel or lattice conventions, node and direction order, normals, sign, time level and precision beside every number. The safeguard is: Track dimensions. Record formula. Change one input, recompute the local rule and global invariant, and explain whether the difference comes from quadrature, collision, streaming, interface transport, force transfer, spatial resolution, time integration, rounding or a changed model. Ask a second reader to reproduce the arithmetic and finish with the narrow claim the evidence supports.

Evidence notebook 13: Time coupling

Step 13 isolates time coupling. Advance fluid and marker. The checked result is: Explicit stiffness may appear. Keep dimensions, kernel or lattice conventions, node and direction order, normals, sign, time level and precision beside every number. The safeguard is: Check stability. Record step. Change one input, recompute the local rule and global invariant, and explain whether the difference comes from quadrature, collision, streaming, interface transport, force transfer, spatial resolution, time integration, rounding or a changed model. Ask a second reader to reproduce the arithmetic and finish with the narrow claim the evidence supports.

Evidence notebook 14: Conclusion

Step 14 isolates conclusion. State grid,markers,kernel,force law and validation. The checked result is: Worked U=3,total force=4,work=12. Keep dimensions, kernel or lattice conventions, node and direction order, normals, sign, time level and precision beside every number. The safeguard is: Analogue is one-dimensional. Write conclusion. Change one input, recompute the local rule and global invariant, and explain whether the difference comes from quadrature, collision, streaming, interface transport, force transfer, spatial resolution, time integration, rounding or a changed model. Ask a second reader to reproduce the arithmetic and finish with the narrow claim the evidence supports.

Failure cases and responsible recovery

Kernel weights do not sum

Total force and constants are lost. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Different transfer kernels

Discrete work balance can break. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Markers too sparse

Fluid leaks through the boundary. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Sharp-interface claim

Regularisation spreads forces. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Explicit stiffness ignored

Time steps may become tiny. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Material law assumed

Coupling cannot validate structure physics. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Grid refinement only

Marker spacing must refine too. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Application validation skipped

Numerical balance is not biological truth. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Verified case study: fourteen steps from model to decision

1. Define the model

Write the PDE or kinetic rule, variables, units, domain and boundary or initial data.

2. Identify the representation

Name boundary densities, lattice populations, a level-set field or Lagrangian markers.

3. State the transfer rule

Write the kernel, collision-streaming map, Hamilton-Jacobi speed or delta coupling.

4. Calculate a tiny case

Use the smallest nontrivial example with labelled indices.

5. Check a conserved quantity

Verify source weight, mass, volume, force or work as eligible.

6. Audit signs and geometry

Check normals, directions, level-set sign, marker coordinates and Jacobians.

7. Test a singular or sharp feature

Approach a kernel source, interface, wall or concentrated force.

8. Refine

Improve panels, lattice, interface grid and marker spacing consistently.

9. Separate errors

Distinguish modelling, discretisation, quadrature, iteration and rounding.

10. Compare a baseline

Use an exact, manufactured, direct or independently implemented reference.

11. Count cost

Record unknowns, nonzeros or dense entries, cell updates, memory and elapsed time.

12. Validate

Compare application outputs with independent measurements or accepted benchmarks.

13. Communicate limits

State what the result supports and what remains untested.

14. Preserve evidence

Save equations, data, code, meshes, parameters, logs and corrected arithmetic.

Learner lab: ten checks for transferable understanding

Lab question 1

Create a reproducible mini-investigation of Eulerian grids, Lagrangian markers, delta kernels, force, velocity, work and leakage. For task 1, state one prediction, calculate a transparent case, record units and conventions, compare an independent baseline, identify one failure mode, and explain the result once for a student and once for a parent. End with one practical next step and one claim the evidence does not support.

Lab question 2

Create a reproducible mini-investigation of Eulerian grids, Lagrangian markers, delta kernels, force, velocity, work and leakage. For task 2, state one prediction, calculate a transparent case, record units and conventions, compare an independent baseline, identify one failure mode, and explain the result once for a student and once for a parent. End with one practical next step and one claim the evidence does not support.

Lab question 3

Create a reproducible mini-investigation of Eulerian grids, Lagrangian markers, delta kernels, force, velocity, work and leakage. For task 3, state one prediction, calculate a transparent case, record units and conventions, compare an independent baseline, identify one failure mode, and explain the result once for a student and once for a parent. End with one practical next step and one claim the evidence does not support.

Lab question 4

Create a reproducible mini-investigation of Eulerian grids, Lagrangian markers, delta kernels, force, velocity, work and leakage. For task 4, state one prediction, calculate a transparent case, record units and conventions, compare an independent baseline, identify one failure mode, and explain the result once for a student and once for a parent. End with one practical next step and one claim the evidence does not support.

Lab question 5

Create a reproducible mini-investigation of Eulerian grids, Lagrangian markers, delta kernels, force, velocity, work and leakage. For task 5, state one prediction, calculate a transparent case, record units and conventions, compare an independent baseline, identify one failure mode, and explain the result once for a student and once for a parent. End with one practical next step and one claim the evidence does not support.

Lab question 6

Create a reproducible mini-investigation of Eulerian grids, Lagrangian markers, delta kernels, force, velocity, work and leakage. For task 6, state one prediction, calculate a transparent case, record units and conventions, compare an independent baseline, identify one failure mode, and explain the result once for a student and once for a parent. End with one practical next step and one claim the evidence does not support.

Lab question 7

Create a reproducible mini-investigation of Eulerian grids, Lagrangian markers, delta kernels, force, velocity, work and leakage. For task 7, state one prediction, calculate a transparent case, record units and conventions, compare an independent baseline, identify one failure mode, and explain the result once for a student and once for a parent. End with one practical next step and one claim the evidence does not support.

Lab question 8

Create a reproducible mini-investigation of Eulerian grids, Lagrangian markers, delta kernels, force, velocity, work and leakage. For task 8, state one prediction, calculate a transparent case, record units and conventions, compare an independent baseline, identify one failure mode, and explain the result once for a student and once for a parent. End with one practical next step and one claim the evidence does not support.

Lab question 9

Create a reproducible mini-investigation of Eulerian grids, Lagrangian markers, delta kernels, force, velocity, work and leakage. For task 9, state one prediction, calculate a transparent case, record units and conventions, compare an independent baseline, identify one failure mode, and explain the result once for a student and once for a parent. End with one practical next step and one claim the evidence does not support.

Lab question 10

Create a reproducible mini-investigation of Eulerian grids, Lagrangian markers, delta kernels, force, velocity, work and leakage. For task 10, state one prediction, calculate a transparent case, record units and conventions, compare an independent baseline, identify one failure mode, and explain the result once for a student and once for a parent. End with one practical next step and one claim the evidence does not support.