About SimDesign

SimDesign is an open-source Python framework for the simulated design of buildings following historical and modern seismic design procedures. It is developed under the Built Environment Data (BED) initiative, and this website is its web service: it runs the framework for you, so that building portfolios, their designs and their nonlinear models can be generated without installing anything. It currently covers reinforced concrete moment-resisting frame buildings, with design classes for Europe.

Why simulated design

Analytical seismic vulnerability models often rely on a few archetype structural models to represent an entire building class. Such models tend to oversimplify the seismic behaviour of individual buildings, and fail to capture the building-to-building variability and the many sources of uncertainty within a class, particularly at a regional scale.

SimDesign takes a probabilistic approach instead. For a building class, it generates a portfolio of building realisations and designs each of them with the seismic design codes and practices of its design class, accounting for the variability in geometry, materials and construction quality.

How it works

1. 1

   Building Class Information Model (BCIM)

   A building class is defined by its design class, its number of storeys and its design lateral load coefficient. For a class, the BCIM samples a portfolio of building realisations: storey heights, bay widths and floor layout, slab, beam and column types, material grades, construction quality and masonry infills.

2. 2

   Building Design Information Model (BDIM)

   Each realisation goes through an iterative simulated design procedure that follows the design rules of its class, from member sizing to reinforcement detailing. Its construction quality then modifies the design details, such as stirrup spacing, concrete cover and material strengths, to their expected in situ values. The results form the BDIM.

3. 3

   Building Nonlinear Structural Model (BNSM)

   Finally, each design becomes a nonlinear structural model in OpenSees (OpenSeesPy or Tcl), with scripts for gravity, modal and nonlinear static (pushover) analyses, ready for further analysis.

The models support the development of probabilistic seismic demand models, fragility functions and vulnerability models for large-scale seismic risk assessment, such as the European Seismic Risk Model (ESRM20).

How to cite

The Built Environment Data Framework for Simulated Design and Vulnerability Modelling in Earthquake Engineering

Volkan Ozsarac, Nuno Pereira, Hossameldeen Mohamed, Xavier Romão, Gerard J. O'Reilly

Earthquake Engineering & Structural Dynamics, 54(11), 2651–2670, 2025

The seismic vulnerability modelling of different categories of buildings analytically requires methodologies capable of capturing the wide range of building standards, construction practices, architectural layouts, earthquake design scenarios and available knowledge. Current vulnerability models employ varying assessment approaches, building taxonomies, representations of seismic loading and, in some cases, rely on a limited number of archetype structural models to represent an entire building class. Consequently, these structural models are likely to oversimplify the seismic behaviour of an individual building, fail to adequately capture the reality of building-to-building variability and inadequately account for multiple sources of uncertainty, particularly when applied to regional contexts. Addressing these issues requires a probabilistic approach where seismic vulnerability is assessed using models of building portfolios that can reflect features related to engineering design practice, as well as construction variability and quality. To achieve this objective, this article introduces a collaborative framework for the simulated design of buildings, along with the open-source software package developed for its integration into the Built Environment Data initiative, a planned service within the European Plate Observing System. The simulated design framework accommodates the design of buildings using both historical and modern seismic design procedures and regulations, while capturing building-to-building variability. Following the design process, the framework generates OpenSees computational models for nonlinear analysis, facilitating the development of probabilistic seismic demand models, fragility functions and vulnerability models. The framework's capabilities are demonstrated through examples that highlight notable distinctions among the building classes under consideration and emphasise the importance of the attributes involved in the design process. The collaborative nature of the framework presented here enables the earthquake engineering community to contribute to a growing database of seismic design practices, encompassing a wide range of design codes.

Read More

View publicationCopy citation

How to contribute

The framework is collaborative: the earthquake engineering community can contribute the design practices of other countries and periods. SimDesign is built to be extended. Each material and lateral load-resisting system, such as reinforced concrete moment frames, is a subpackage of simdesign, and within it each design class and each modelling approach is a module of its own, built on a shared base library. The design practice of another country or period, or another way of modelling the buildings, is added as a new branch of the tree below, without changing the existing ones. Likewise, SimDesign can be extended with other materials (precast concrete, steel, masonry) and lateral load-resisting systems. Materials and lateral load-resisting systems below are named after the GEM Building Taxonomy v4.0. The contribution guide describes how to submit a contribution.

Open source

SimDesign is free software, licensed under the GNU Affero General Public License v3.0. Its source code is maintained in a GitHub repository, and the simdesign package can be installed from PyPI to run the framework directly in Python, with the full set of options.