CA08 exercise · Method-selection and coupling exercise

Hybrid Method Selection and Coupling

Choose suitable hybrid method combinations for multi-region acoustic problems and compare weak iterative coupling with a monolithic coupled reference.

45–75 minLinked to Hybrid MethodsDashboard progress enabled

Aim

Choose method partitions and inspect coupling strength

Hybrid methods combine solvers because one method is rarely optimal everywhere. This exercise asks you to classify subregions and then run a toy weak-coupling iteration.

1

Choose a hybrid modelling strategy

A vibrating machine housing radiates into an unbounded exterior fluid. A natural method split is structural FEM for the housing and BEM/Rayleigh radiation for the exterior acoustic field.

What is one-way FEM–BEM radiation coupling?

2

Compute the monolithic reference of a toy coupled system

For the toy system in the Python code, the monolithic solution is approximately x = 0.3261 and y = 0.2536.

Python exercise

Run weak versus monolithic coupling

The code compares a monolithic coupled solution with a sequential fixed-point iteration.

Expected observation

The weak iteration approaches the monolithic solution when feedback is moderate. Strong coupling can slow convergence or destabilise the iteration.

Teaching note

This is a numerical analogy for partitioned hybrid methods. Real hybrid acoustic coupling also needs physical interface consistency and validation.

Written submission

Hybrid validation checklist

Interface checks

List three checks you would perform before trusting a hybrid FEM–BEM or FEM–SEA result.

0 words
Saved automatically in this browser draft.

Submit solution

Classroom submission workflow

When a student is logged in, the submission is stored in Supabase and becomes visible to the supervisor grading page across devices. Local browser storage remains as fallback.

Supabase not configured

Checking login...

Draft status

0%

0 checks · 0 answers · 0 runs

Student identity

Enter temporary identity until login is active.