Chapter map

1. 1. Material speed
2. 2. Mesh speed
3. 3. Relative speed
4. 4. Time step
5. 5. Material move
6. 6. Mesh move
7. 7. Relative move
8. 8. Lagrangian limit
9. 9. Eulerian limit
10. 10. Cell volume
11. 11. Uniform field
12. 12. Rezone
13. 13. Remap
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. Material speed

A small example makes the mechanism visible. Set u=5.

A concrete situation

Material velocity in lab frame.

What the mathematics does not license

State units. Evidence becomes stronger when the limitation is tested instead of hidden.

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.

---

2. Mesh speed

This is where modelling choices begin to affect evidence. Set w=2.

A concrete situation

Grid moves in same direction.

What the mathematics does not license

State sign. Keeping this boundary clear makes the result easier to audit and transfer.

Try it, check it, transfer it

Record w. 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. Relative speed

The central mathematical idea is concrete. Compute u-w.

A concrete situation

Relative advection=3.

What the mathematics does not license

Subtract in one frame. A trustworthy workflow records the choice before seeing the final test result.

Try it, check it, transfer it

Record value. 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. Time step

Here the geometry becomes operational. Set dt=.1.

A concrete situation

Common update interval.

What the mathematics does not license

State time level. Evidence becomes stronger when the limitation is tested instead of hidden.

Try it, check it, transfer it

Record dt. 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. Material move

This step deserves a careful pause. Compute u dt.

A concrete situation

Displacement=.5.

What the mathematics does not license

Lab frame. Keeping this boundary clear makes the result easier to audit and transfer.

Try it, check it, transfer it

Record move. 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. Mesh move

The useful question is not only how to calculate, but what the calculation assumes. Compute w dt.

A concrete situation

Displacement=.2.

What the mathematics does not license

Lab frame. A trustworthy workflow records the choice before seeing the final test result.

Try it, check it, transfer it

Record move. 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. Relative move

A small example makes the mechanism visible. Compute (u-w)dt.

A concrete situation

Displacement=.3.

What the mathematics does not license

Also .5-.2. Evidence becomes stronger when the limitation is tested instead of hidden.

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.

---

8. Lagrangian limit

This is where modelling choices begin to affect evidence. Set w=u.

A concrete situation

Relative speed=0.

What the mathematics does not license

Mesh follows material. Keeping this boundary clear makes the result easier to audit and transfer.

Try it, check it, transfer it

Record limit. 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. Eulerian limit

The central mathematical idea is concrete. Set w=0.

A concrete situation

Relative speed=5.

What the mathematics does not license

Mesh is fixed. A trustworthy workflow records the choice before seeing the final test result.

Try it, check it, transfer it

Record limit. 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. Cell volume

Here the geometry becomes operational. Move both faces.

A concrete situation

Update volume from geometry.

What the mathematics does not license

Use consistent face velocities. Evidence becomes stronger when the limitation is tested instead of hidden.

Try it, check it, transfer it

Record volume. 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. Uniform field

This step deserves a careful pause. Advance a constant state.

A concrete situation

It should remain constant.

What the mathematics does not license

Test geometric conservation. Keeping this boundary clear makes the result easier to audit and transfer.

Try it, check it, transfer it

Record error. 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. Rezone

The useful question is not only how to calculate, but what the calculation assumes. Improve distorted nodes.

A concrete situation

Mesh quality rises.

What the mathematics does not license

State objective. A trustworthy workflow records the choice before seeing the final test result.

Try it, check it, transfer it

Record 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.

---

13. Remap

A small example makes the mechanism visible. Transfer conserved data.

A concrete situation

Global mass should match.

What the mathematics does not license

Check extrema. Evidence becomes stronger when the limitation is tested instead of hidden.

Try it, check it, transfer it

Record audit. 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

This is where modelling choices begin to affect evidence. State mesh motion,relative flux,GCL,remap and validation.

A concrete situation

Worked relative speed=3 and move=.3.

What the mathematics does not license

One-dimensional example. Keeping this boundary clear makes the result easier to audit and transfer.

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 arbitrary Lagrangian-Eulerian method describes a continuum on a mesh that may move with material, remain fixed in space or follow a chosen motion in between. Its central mathematics separates material velocity from mesh velocity.

In one dimension, let material move at u=5 while the mesh moves at w=2 in consistent distance-per-time units. Relative to the mesh, advection occurs at u-w=3.

Over a time step of 0.1, the material moves 0.5, the mesh moves 0.2 and the material moves 0.3 relative to the grid. The three displacements agree because 0.5-0.2=0.3.

A Lagrangian choice sets w=u, so the relative advective speed is zero and mesh points follow material. An Eulerian choice sets w=0, so the full material speed crosses the fixed grid.

The arbitrary choice lets a mesh follow an important boundary without accepting every deformation of the material. A mesh-motion equation, smoothing rule or optimisation decides where interior nodes go.

Conservation laws must be transformed consistently to the moving control volume or coordinate description. Relative fluxes use u-w, while changing cell volumes contribute geometric terms.

Did You Know? Even a perfectly uniform physical field can develop numerical error on a moving mesh if the discretisation does not preserve the geometric conservation law. Geometry is part of the numerical balance.

Purely Lagrangian meshes can tangle or become highly skewed under shear. Purely Eulerian meshes avoid that motion but may blur material interfaces through advection. ALE chooses a compromise rather than eliminating both difficulties.

When mesh quality deteriorates, rezoning moves nodes to better positions and remapping transfers solution data. The remap should be conservative for conserved quantities and should not create new extrema without justification.

Mesh velocity must be used in boundary conditions and numerical fluxes. Updating coordinates without updating the equations produces an inconsistent method even if the picture looks plausible.

Time integration couples physical motion and geometric motion. The solver must define when positions, face velocities, volumes and fluxes are evaluated.

Students can compare Lagrangian, Eulerian and intermediate mesh choices on the same one-dimensional advection problem, then check mass and a uniform-field solution after each update.

Applications include fluid-structure interaction, free surfaces, metal forming, impacts, moving pistons and compressible flow. Large topological change may still require remeshing or another interface method.

Mesh refinement is not enough if elements become badly shaped. Convergence studies should report quality measures, remap frequency and conservation error as well as nominal element size.

For parents and students, the transferable lesson is that a coordinate system can move strategically, but every measurement must account for how the ruler itself moved.

A responsible ALE report states material equations, mesh velocity, transformed fluxes, mesh-motion rule, geometric conservation, rezoning, remap, boundary conditions, time integration, mesh quality, convergence and validation.

---

Frequently asked questions

What does ALE mean?

The mesh may move with material, stay fixed, or follow another prescribed motion.

Why does u minus w appear?

Material crosses a mesh moving at velocity w with relative velocity u-w.

What is the geometric conservation law?

A consistency condition linking mesh motion and changing cell geometry so uniform states remain uniform.

Why rezone and remap?

Rezoning improves mesh quality; remapping transfers the solution to the improved mesh.

Does a moving mesh remove advection error?

It can reduce some relative motion, but mesh distortion, remap error and physical discretisation remain.

What must be reported?

Material equations, mesh velocity, relative flux, geometry update, rezoning, remap, boundaries, convergence and validation.

---

Primary source and fact-check note

C. W. Hirt, A. A. Amsden and J. L. Cook, “An Arbitrary Lagrangian–Eulerian Computing Method for All Flow Speeds,” Journal of Computational Physics 14, 1974. 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

• Immersed boundary method
• Finite volume method
• Adaptive mesh refinement
• Numerical analysis hub
• How Mathematics Works

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

Frame audit

Separate material, mesh and relative velocities.

Geometry audit

Update coordinates, face speeds and cell measures consistently.

Conservation audit

Test a uniform state and conserved totals.

Remap audit

Measure transfer error, new extrema and mesh-quality change.

Mechanism workshop: calculate, challenge and transfer

Velocity table

Compare u,w and u-w.

Limit test

Set w=u and w=0.

Moving cell

Update two face positions and volume.

Uniform test

Preserve a constant field.

Flux test

Use relative face velocity.

Mesh quality

Measure skewness before and after rezoning.

Remap mass

Compare totals across meshes.

Time-level audit

Label geometry and flux evaluation times.

Student explanation

Explain a ruler that moves with the problem.

Parent explanation

Relate coordinate choice to viewpoint.

Evidence notebook: reconstruct every step

Evidence notebook 1: Material speed

Step 1 isolates material speed. Set u=5. The checked result is: Material velocity in lab frame. Keep units, coordinates, particle or mesh labels, kernel or shape conventions, sign, time level and precision beside every number. The safeguard is: State units. Record u. Change one input, recompute the local mechanism and the relevant invariant, and explain whether the difference comes from interpolation, deposition, neighbour support, free energy, mesh motion, remapping, 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: Mesh speed

Step 2 isolates mesh speed. Set w=2. The checked result is: Grid moves in same direction. Keep units, coordinates, particle or mesh labels, kernel or shape conventions, sign, time level and precision beside every number. The safeguard is: State sign. Record w. Change one input, recompute the local mechanism and the relevant invariant, and explain whether the difference comes from interpolation, deposition, neighbour support, free energy, mesh motion, remapping, 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: Relative speed

Step 3 isolates relative speed. Compute u-w. The checked result is: Relative advection=3. Keep units, coordinates, particle or mesh labels, kernel or shape conventions, sign, time level and precision beside every number. The safeguard is: Subtract in one frame. Record value. Change one input, recompute the local mechanism and the relevant invariant, and explain whether the difference comes from interpolation, deposition, neighbour support, free energy, mesh motion, remapping, 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: Time step

Step 4 isolates time step. Set dt=.1. The checked result is: Common update interval. Keep units, coordinates, particle or mesh labels, kernel or shape conventions, sign, time level and precision beside every number. The safeguard is: State time level. Record dt. Change one input, recompute the local mechanism and the relevant invariant, and explain whether the difference comes from interpolation, deposition, neighbour support, free energy, mesh motion, remapping, 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: Material move

Step 5 isolates material move. Compute u dt. The checked result is: Displacement=.5. Keep units, coordinates, particle or mesh labels, kernel or shape conventions, sign, time level and precision beside every number. The safeguard is: Lab frame. Record move. Change one input, recompute the local mechanism and the relevant invariant, and explain whether the difference comes from interpolation, deposition, neighbour support, free energy, mesh motion, remapping, 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: Mesh move

Step 6 isolates mesh move. Compute w dt. The checked result is: Displacement=.2. Keep units, coordinates, particle or mesh labels, kernel or shape conventions, sign, time level and precision beside every number. The safeguard is: Lab frame. Record move. Change one input, recompute the local mechanism and the relevant invariant, and explain whether the difference comes from interpolation, deposition, neighbour support, free energy, mesh motion, remapping, 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: Relative move

Step 7 isolates relative move. Compute (u-w)dt. The checked result is: Displacement=.3. Keep units, coordinates, particle or mesh labels, kernel or shape conventions, sign, time level and precision beside every number. The safeguard is: Also .5-.2. Record check. Change one input, recompute the local mechanism and the relevant invariant, and explain whether the difference comes from interpolation, deposition, neighbour support, free energy, mesh motion, remapping, 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: Lagrangian limit

Step 8 isolates lagrangian limit. Set w=u. The checked result is: Relative speed=0. Keep units, coordinates, particle or mesh labels, kernel or shape conventions, sign, time level and precision beside every number. The safeguard is: Mesh follows material. Record limit. Change one input, recompute the local mechanism and the relevant invariant, and explain whether the difference comes from interpolation, deposition, neighbour support, free energy, mesh motion, remapping, 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: Eulerian limit

Step 9 isolates eulerian limit. Set w=0. The checked result is: Relative speed=5. Keep units, coordinates, particle or mesh labels, kernel or shape conventions, sign, time level and precision beside every number. The safeguard is: Mesh is fixed. Record limit. Change one input, recompute the local mechanism and the relevant invariant, and explain whether the difference comes from interpolation, deposition, neighbour support, free energy, mesh motion, remapping, 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: Cell volume

Step 10 isolates cell volume. Move both faces. The checked result is: Update volume from geometry. Keep units, coordinates, particle or mesh labels, kernel or shape conventions, sign, time level and precision beside every number. The safeguard is: Use consistent face velocities. Record volume. Change one input, recompute the local mechanism and the relevant invariant, and explain whether the difference comes from interpolation, deposition, neighbour support, free energy, mesh motion, remapping, 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: Uniform field

Step 11 isolates uniform field. Advance a constant state. The checked result is: It should remain constant. Keep units, coordinates, particle or mesh labels, kernel or shape conventions, sign, time level and precision beside every number. The safeguard is: Test geometric conservation. Record error. Change one input, recompute the local mechanism and the relevant invariant, and explain whether the difference comes from interpolation, deposition, neighbour support, free energy, mesh motion, remapping, 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: Rezone

Step 12 isolates rezone. Improve distorted nodes. The checked result is: Mesh quality rises. Keep units, coordinates, particle or mesh labels, kernel or shape conventions, sign, time level and precision beside every number. The safeguard is: State objective. Record positions. Change one input, recompute the local mechanism and the relevant invariant, and explain whether the difference comes from interpolation, deposition, neighbour support, free energy, mesh motion, remapping, 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: Remap

Step 13 isolates remap. Transfer conserved data. The checked result is: Global mass should match. Keep units, coordinates, particle or mesh labels, kernel or shape conventions, sign, time level and precision beside every number. The safeguard is: Check extrema. Record audit. Change one input, recompute the local mechanism and the relevant invariant, and explain whether the difference comes from interpolation, deposition, neighbour support, free energy, mesh motion, remapping, 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 mesh motion,relative flux,GCL,remap and validation. The checked result is: Worked relative speed=3 and move=.3. Keep units, coordinates, particle or mesh labels, kernel or shape conventions, sign, time level and precision beside every number. The safeguard is: One-dimensional example. Write conclusion. Change one input, recompute the local mechanism and the relevant invariant, and explain whether the difference comes from interpolation, deposition, neighbour support, free energy, mesh motion, remapping, 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

Mesh speed omitted

Relative flux is wrong. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Coordinates moved only

Equations no longer match geometry. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Uniform state changes

Geometric conservation is violated. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Lagrangian mesh always best

Shear can tangle cells. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Rezone without remap audit

Conserved quantities may change. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Remap called exact

Interpolation and limiting add error. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Element size alone reported

Skewness and quality matter. Build a small counterexample, restore the missing condition and recalculate before making the claim.

Moving picture as proof

Convergence and validation are still required. 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 governing field, particle, continuum or conservation equations, variables, units, domain and boundary data.

2. Choose the representation

Name particles, grid values, order parameters or moving mesh coordinates.

3. State the transfer rule

Write the shape function, kernel, chemical potential or relative flux.

4. Calculate a tiny case

Use the smallest nontrivial example with labelled indices and coordinates.

5. Check an invariant

Verify charge, constant reproduction, mass, free energy, uniform flow or conserved total as eligible.

6. Audit signs and units

Check charge signs, kernel dimensions, chemical-potential convention and mesh-velocity direction.

7. Challenge a difficult region

Test a boundary, sparse neighbourhood, diffuse interface or distorted element.

8. Refine

Improve particles, grid, interface resolution or mesh quality consistently.

9. Separate errors

Distinguish modelling, sampling, interpolation, discretisation, remap, iteration and rounding.

10. Compare a baseline

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

11. Count cost

Record particles, neighbours, grid degrees of freedom, 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, parameters, meshes, seeds, logs and corrected arithmetic.

Learner lab: ten checks for transferable understanding

Lab question 1

Create a reproducible mini-investigation of material velocity, mesh velocity, relative flux, moving geometry, rezoning and remapping. 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 material velocity, mesh velocity, relative flux, moving geometry, rezoning and remapping. 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 material velocity, mesh velocity, relative flux, moving geometry, rezoning and remapping. 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 material velocity, mesh velocity, relative flux, moving geometry, rezoning and remapping. 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 material velocity, mesh velocity, relative flux, moving geometry, rezoning and remapping. 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 material velocity, mesh velocity, relative flux, moving geometry, rezoning and remapping. 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 material velocity, mesh velocity, relative flux, moving geometry, rezoning and remapping. 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 material velocity, mesh velocity, relative flux, moving geometry, rezoning and remapping. 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 material velocity, mesh velocity, relative flux, moving geometry, rezoning and remapping. 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 material velocity, mesh velocity, relative flux, moving geometry, rezoning and remapping. 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.