The model record

What goes in

Purpose. The engineering question this model exists to answer, in one sentence. Everything else follows from it, and a model without one will eventually be asked the wrong question.

Responsibility. Who built it, who checked it, who reviewed it, and when.

Software. Name, version, solver. Versions matter: results can change between them, and 'we used the current version' is not reproducible two years later.

Model type and analysis type. One-dimensional, two, three; linear, second-order, nonlinear, modal. And why — the justification for linear analysis is a validation statement.

Units and origin. Stated once, explicitly. Half the errors in Module 12 begin here.

Materials, sections, supports, releases, constraints. What was assumed, with the reasoning for anything not obvious. 'Bases pinned' needs a reason; 'bases modelled with rotational springs of 8 000 kN·m/rad from the geotechnical report' needs a reference.

Loads and combinations. Including which combinations were not run and why.

Warnings. What the solver reported, and what was decided about each.

Validation and verification carried out. What check, what it compared against, what the result was. 'Checked' is not a record.

Sensitivity studies. Which variables, what range, what the answer did.

Limitations. What this model cannot be used for.

Exclusions are as important as inclusions

This is the part that gets left out, and it is the part a reviewer most needs.

A model record that says what was included lets a reviewer check those things. A record that also says what was deliberately excluded, and why lets them disagree — which is the only way a review adds anything.

'Non-structural partitions excluded: they contribute stiffness at serviceability but cannot be relied on for the life of the building, and their layout is not yet fixed.'

A reviewer can now engage with that. They might agree, or point out that the vibration assessment depends on exactly that stiffness. Without the sentence they cannot tell whether partitions were considered and dismissed, or never thought about.

The same applies to load cases not run, behaviours assumed negligible, and members left out of the model. Say so, and say why.

Writing a limitation that survives

A limitation will be read by someone who has not read the rest of the record, six months later, looking for permission to do something. Write it accordingly:

Weak: 'Model is for global analysis only.'

Strong: 'This model uses one-dimensional elements throughout and contains no representation of connection geometry. It must not be used for connection design, for local stress at any joint, or for any assessment of the column base other than the overall force and moment transmitted to it.'

The second names what the model cannot do and what somebody might mistakenly try to use it for. That is the sentence that stops the misuse.

What this is not

The workspace in this course produces a Queensferry model review record. It is a teaching artefact and a reasonable structure for organising the information. It is not a category 2 or category 3 checking form, it is not a substitute for whatever your organisation's quality procedures require, and it makes no claim to satisfy any external requirement.