Building models · simulation · evidence
Computational Architecture & Building Performance
Connect building questions to explicit models, inspectable assumptions and reproducible calculations.
Start with a question
Choose a useful entry point
Interactive example
Building program and area reconciliation
Reconcile gross, assignable and residual support areas before using a program total in other decisions.
ExploreMaha-authored method
A reproducible design-evidence package
A design-evidence package should let another person reconstruct a bounded calculation and see what it cannot establish.
ExploreMaha-authored method
Model checking is not code compliance
A model checker can establish that a declared condition holds in its input. Whether that condition captures an applicable legal requirement is a separate question.
Explore
Design foundations
4 articles
Define the question, the geometry and the information a model must preserve.
Source-linked guide
What can computational architecture establish?
A computational model makes a design question explicit enough to calculate, compare and challenge. It does not turn an attractive concept into an approved building.
ExploreMaha-authored method
Parametric geometry: constraints before shapes
A parametric study should expose the variables that change a design and the constraints that remain fixed.
ExploreMaha-authored method
Separate requirements, assumptions and decisions
A target records what an owner wants; an assumption lets a study proceed; a decision records what an authorized person accepted.
ExploreSource-linked guide
BIM and IFC: exchange with a declared purpose
A shared building model is useful only when the receiving workflow gets the information it requires.
Explore
Performance studies
4 articles
Inspect energy, light, airflow and material comparisons without confusing simulations with measurements.
Source-linked guide
What belongs behind a building-energy result?
An energy result needs a model, an operating scenario and an explicit comparison basis—not just a percentage saving.
ExploreSource-linked guide
Daylight and glare: inspect the question behind the image
A persuasive lighting image and a numerical daylight study are different deliverables.
ExploreSource-linked guide
Natural ventilation and mixed-mode boundaries
A ventilation concept needs operating conditions and fallback behavior before it can support a performance claim.
ExploreMaha-authored method
Material comparisons need matching boundaries
Before comparing two material claims, check that they describe the same function, quantity and study boundary.
Explore
Checks and evidence
4 articles
Try a calculation, challenge assumptions and preserve the boundary of each result.
Interactive example
Building program and area reconciliation
Reconcile gross, assignable and residual support areas before using a program total in other decisions.
ExploreMaha-authored method
Model checking is not code compliance
A model checker can establish that a declared condition holds in its input. Whether that condition captures an applicable legal requirement is a separate question.
ExploreMaha-authored method
Sensitivity without a false best design
Sensitivity analysis shows how a result responds to assumptions; it does not decide which objective matters most.
ExploreMaha-authored method
A reproducible design-evidence package
A design-evidence package should let another person reconstruct a bounded calculation and see what it cannot establish.
Explore
Connected knowledge
Continue across domains