Week 8: The Order of Layers
The Week 8 lecture — assembly order as a structural variable, and how to build one material set twice so that the ordering is the only thing under test
19 April 2027
Thirty slides on one claim: reordering an assembly changes none of its materials and all of its interfaces.
The lecture builds the claim in two halves. The first is the mechanism — five layers admit four interfaces, sixty distinguishable orders are available to them, and only neighbouring layers interact, so moving one layer changes which boundary is loaded and which layer sits far enough from the neutral axis to carry anything. The second half is the method: fix a material set, build it in two orders, write an adjacency note at every interface, predict the first failure in writing, then load both identically and compare.
It also states where the model stops. The course treats an assembly as discrete layers meeting at clean interfaces. Real layers interpenetrate, and once one layer has soaked into its neighbour there is no interface left to attribute a failure to. The discrete model is accepted here because it is what makes order testable at all, not because it is what a specimen is.
The week page carries the same argument as prose, with the exploded pair at full size. It is the readable equivalent of the deck and needs no presenter.