A Semi-Automated Approach to CAD input into Field Based Fire Modelling Tools
This research project examines the role of Computer Aided Design (CAD) data in the design cycle of modelling tools used, in particular, for Computational Fluid Dynamics (CFD) based Fire Modelling tools.
The evaluation of the fire safety aspects of buildings is one of the most important parts of the design phase for architects. Over the last decade or two, major advances have been made in the development of tools for assisting architects in the evaluation of building design. These range from tools to evaluate (1) the look-and-feel of the building and (2) Safety-aspects associated with the design.
Automatic or Semi-Automatic importing of buildings within CFD based modelling tools, is pursued in fire modelling applications for two primary reasons. Firstly, with such a system it may be possible to import large and complex building structures with minimal effort from existing CAD based data, for example the commonly used Drawing eXchange File (DXF) format. DXF is possibly the most widespread CAD exchange format in use by CAD packages on small computer systems. It was developed primarily by the AutoCAD developers and has thus received its popularity mainly from the high number of AutoCAD users. Most CAD systems can export and most also import DXF, at least for 2D data.
The approach used in this project for importing geometrical data via CAD files results in a substantially reduced “Build Phase” of the fire model development, which in most practical applications can represent a substantial part of the entire analysis. Furthermore, this would also make fire field modelling accessible to a wider audience. Secondly, it is also hoped that such an environment would aid the designers to build scenarios based on either whole or parts of the imported building structures. Once the overall system is in place, it is envisaged that the overall design cycle would take much less time to build the model(s) than is currently possible, for large and complex simulations. The development and use of these techniques is being explored through the use of the SMARTFIRE fire field model.
This project concentrates on research into the use of underlying feature recognition techniques to assist in importing of buildings from external CAD for the solution of fire scenarios simulated using SMARTFIRE. Its main aim is to improve the turn around time for engineers who normally would spend considerable amounts of time re-building the geometry from scratch in the SMARTFIRE problem setup user interface. The Semi-automatic approach of importing CAD drawings would reduce the initial overheads involved in geometry build phase. The technique automatically recognises objects such as doors, windows, walls, etc, for both layered (easily) and non-layered (with difficulty) CAD files. Typically, a DXF CAD file consists of the following information: Points, Lines, Polylines, Circle, Arc, Text, etc. However, none of these are defined as objects that can be easily used for defining room objects within buildings for non-layered files. The room objects need to be built from the information provided. This is very important, since almost all field-modelling tools require domains that will need to be meshed. Once the room objects have been constructed, then a building object can be constructed using a collection of rooms. The stumbling block in this methodology is how does one construct “room” objects from a collection of lines, polylines, etc.
The approach adopted for the research is termed as the Vertex Graph Traversal Algorithm for Room Identification. This is exemplified using the example below. Consider a room defined using lines and polylines.
Step 1: Importing a DXF file. (Fig. 1a)
Step 2: Convert Doors and Windows symbols to lines and merge vertices – This
ensures a closed contour of walls has been generated. (Fig. 1b)
Step 3: Project a ray from the centre until a wall has been detected. The
wall contour is then detected all around the closed room contour to ascertain
the lines that make the room object. (Fig. 1c)
Step 4: Collect all connecting lines to define a room object. (Fig. 1d)
(a) (b)
(c)
(d)
Figure 1: (a) Import CAD Drawing (b) Convert door and window
(c) Ray tracing for walls (d) Detected room contour
Once the above phase has been applied to reasonably clean DXF files, users of
the developed prototype framework can use it to generate models to be imported
into SMARTFIRE. This is illustrated in Figure 2, where a subset building
structure of 13 rooms is extracted from the original building that consisted 50
rooms. Figure 2b, depicts the intermediate feature recognised interpretation of
the original DXF file. This converts various objects to their database
representation so that they are easily handled. Figure 2c depicts the
three-dimensional model extracted, from part of the DXF file, as a scenario that
is of interest to the modeller. The model is then imported into SMARTFIRE
where the mesh is automatically generated. This is depicted in Figure 3.
Figure 2: (a) Part of original DXF drawing (b) Feature recognised output
(c) Building model generated from “rooms”
Results from the research have shown that the initial CAD drawing is critical to the level of automation possible. An interface and the underlying feature recognition techniques to address the above task have been implemented in a prototype system. This is currently being used to assess various potential methodologies that would provide a usable and practical tool for engineers with both “old” and “newly” created drawings. The final phase of the task which will allow the altering of the imported building objects via the SMARTFIRE problem setup component has also been implemented and tested on a range of cases. Work is currently underway to finalise the methodology that will be embedded within a future release version of the SMARTFIRE software.
Figure 3: Typical mesh generated via SMARTFIRE
For information relating to the SMARTFIRE fire field model visit our
SMARTFIRE web pages. You will
find a range of publications relating to SMARTFIRE and our other research on the
FSEG Publications pages both CMS press and external publications.
