The Order of Layers# SLOP2895 · Week 8# Why does the same material set behave differently when reordered?
Weeks 2 to 7 changed what was in the specimen. This week nothing is added,
nothing is removed, and nothing is substituted. The only thing that moves is
position, and the assembly still behaves differently.
A Material Set Is a List# Bread, support, loose, moist, binding. Five materials, no arrangement.
Slide 1. Start with what a materials list actually specifies: membership.
Nothing here says what touches what.
An Assembly Is an Ordered List# The same five, now with a sequence. Everything this week follows from that
difference.
Reorder It: Nothing About the Materials Changes# Same components. Same masses. Same properties, every one of them.
Slide 3. Say this plainly, because the intuition fights it. Nobody swapped a
material for a different material.
What Changes Is the Set of Interfaces# Five layers admit four interfaces. Which four depends entirely on sequence.
Slide 4. This is the mechanism. The interface is where load is transferred and
where grip is held or lost — and the interfaces are what the ordering
decides.
Sixty Distinguishable Orders# Five layers, once the flip is discounted. Each presents a different pair of
materials at every junction.
Slide 5. Sixty is 5! / 2 = 60. The flip is discounted because turning the whole
stack over does not change which layer touches which.
Do not dwell on the arithmetic. The point is that “the same materials” names a
very large family of assemblies, not one.
Two Orders, One Material Set# Baseline order
Outer layer Support layer Loose layer Moist layer Binding layer Outer layer Revised order
Outer layer Support layer Binding layer — moved Loose layer Moist layer Outer layer The same five layers, exploded, assembled twice. Only the binding layer moves: in the revised order it sits above the loose layer rather than below the moist one. The material set is identical in both stacks, so every difference in how the two assemblies behave is a difference in which layer touches which. Only the binding layer moves.
Slide 6. The two stacks the whole week works from. In the revised order the
binding layer sits above the loose layer instead of below the moist one.
Give the room time here. Every later slide refers back to this pair.
Consequence One: Exposure# Only neighbours interact.
Slide 7. The first of two consequences worth testing. It sounds obvious and it
is routinely ignored in a materials list.
Only Touching Layers Interact# A layer’s effect on another depends on whether it reaches it.
Slide 8. A layer two positions away acts on its own neighbours, not across the
gap. Anything travelling further has to cross an interface to do it, and that
interface is a condition of its own.
A Moist Layer Against the Outer Layer# The boundary is loaded directly.
Slide 9. This is a Soak condition, in the Field Manual’s vocabulary: a moisture
front advances from a boundary into a porous layer.
The Same Layer, One Position Inward# A drier layer now stands between. A different boundary is loaded.
Slide 10. Identical material. Identical quantity. The front now has to cross an
interface before it reaches the outer layer, and it arrives later and weaker if
it arrives at all.
The Material Is Identical. The Assembly Is Not.# Slide 11. Land the first consequence before moving on. If a record says “the
bread went soft” without saying what the bread was touching, it has not
recorded the condition.
Consequence Two: The Load Path# Position decides what a layer carries.
Slide 12. The second consequence. This one is mechanical rather than about
transport, and it is the one that surprises people.
Distance From the Neutral Axis# In bending, a layer’s contribution to stiffness depends on how far it sits from
the axis that neither stretches nor compresses.
Slide 13. Standard beam behaviour, and the reason an I-beam puts its material
at top and bottom rather than in the middle.
This is a real structural relation, not a course invention — but the course
applies it to a layered assembly that only approximates a beam. Say so.
A Stiff Layer at the Centre# Close to the axis. It contributes little.
Slide 14. The same stiff component, doing almost nothing, because of where it
was placed.
The Same Layer at the Outer Face# Far from the axis. It contributes a great deal.
Position, Not Material, Decides Load-Bearing# Slide 16. The second consequence, landed. “Load-bearing filling” is a claim
about position as much as about substance.
What This Model Assumes# Discrete layers meeting at clean interfaces.
Slide 17. The honest limit. Real layers interpenetrate: crumb absorbs, films
migrate, edges blur. There is no clean plane in a real specimen.
The discrete model is accepted here because it is what makes order testable at
all. State it as an assumption, not as a fact about sandwiches.
Where the Model Breaks# A layer that soaks into its neighbour stops being two layers with an interface
between them.
Slide 18. And once that happens, “which interface failed” no longer has an
answer, because the interface is gone. This is a real limit on the week’s whole
method and students should be able to name it.
The Investigation# Fix one material set. Build it in two orders.
Slide 19. The practical half of the lecture starts here. The set is fixed
first, and stays fixed — that is what makes order the variable.
List Every Interface# For each order, name the two layers that form each junction.
Slide 20. Four interfaces per order, eight in total. A junction nobody listed
is a junction nobody predicted.
One Adjacency Note Per Interface# What do these two touching materials do to each other?
Slide 21. The note is the reasoning, written before the load is applied. It is
also the thing a marker reads to see whether the prediction was thought or
guessed.
Predict the First Failure# Name the interface, and say why, before loading anything.
Slide 22. Predict in writing. A prediction made after the result is not a
prediction, and the difference is not detectable later unless it was recorded
first.
Load Both Assemblies Identically# Same load. Same duration. Same support. Same reader.
Slide 23. Everything the Field Manual’s observation protocol asks for. If two
conditions changed, the comparison has nothing to attribute the difference to.
Compare Prediction With Result# The predicted interface and the observed one.
Slide 24. Both get recorded, whether or not they agree.
A Wrong Prediction With a Recorded Reason# Is the more useful record.
Slide 25. Say this deliberately and mean it. A confident wrong prediction with
its reasoning attached tells you which model failed. A right prediction with no
reasoning tells you nothing you can reuse.
This is also assessable: the mark is for the reasoning and the honesty, not for
being right.
The Failure Condition# An interface fails in one order and holds in the other — same materials, same
load, same interval.
Slide 26. Now order is the only remaining variable, and it is the one under
test. That is what makes the result attributable.
What Leaves the Session# An Assembly Rationale .
Slide 27. One artefact, four parts, on the next slide.
The Assembly Rationale# The fixed material set. Both orders drawn as labelled stacks. An adjacency note
at every interface. The predicted first failure, and the observed one.
Slide 28. Read these out. This is the specification the week is marked against,
and it is also the record Week 11 will need.
Where It Goes# The Failure Autopsy cannot attribute an interface failure without knowing the
order the specimen was built in.
Slide 29. Week 11 works from evidence produced by somebody else. If the order
was never recorded, a whole class of explanation is unavailable to the person
holding the fragments — and they will not know it is missing.
The same materials. A different structure.#