New computing architectures allow sophisticated, powerful fire modeling simulations that are crucial for the future of fire safety. However, the computing resources that are needed to use these tools has historically been prohibitive.
Fire Dynamic Simulator Used for Forensic Analysis
Two pieces of software, Fire Dynamic Simulator (FDS) and Smokeview (SMV), have been developed for fire modeling simulations over the last 15 years. They can be used to understand real-life fires after they occur or predict the consequences before they happen. Shortly after the software was developed by the National Institute of Standards and Technology (NIST), it was used to analyze how a blaze in Washington D.C. led to the deaths of two firefighters.
The fire team entered the basement and first floor of the building after deeming the structure safe. However, within two minutes, rapidly expanding fire in the basement forced firefighters to flee. Two personnel were lost, and a third was severely burned.
At the time, no one understood why the fire had grown so quickly or why the fire developed the way it did. The forensics team turned to NIST and FDS for answers. By inputting the specifications of the building, material thermal properties, and the origin of the fire, fire protection engineers and scientists were able to deduce the cause of the fire’s rapid expansion: firefighters had broken two windows in order to enter the basement, which allowed oxygen into the basement. In turn, the new air fueled the fire’s expansion and caused unexpected changes in another location of the building. Without FDS and SMV, officials may never have fully understood the dynamics of why the fire behaved the way it did.
Cloud Overcomes an FDS Weakness
The Achilles Heel of FDS is also its greatest strength – resolution.
The program breaks down environments of interest into small cells and performs complex calculations on every individual cell as it marches through time. This high resolution approach enables users of FDS to study fires in ways that were possible in the recent past.
On the other hand, the program requires a staggering amount of processing speed and memory. Historically, organizations needed to access small server farms in order to use it for large commercial or industrial structures. But, as processing technology becomes more affordable and accessible, fire analysis can be more practical and more relevant than ever before. For example, cloud computing allows us to rent all the space we need on a case-by-case basis.
By Jeff Harrington, CEO and Founder of Harrington Group, Inc.