Forecasting Fire Smoke Safety in Buildings and Hybrid Building Simulations Using Data Assimilation

Student: Cheng-Chun Lin, Ph.D.

Sponsor: Natural Science and Engineering Research Council of Canada (NSERC) Discovery Grants

Publications:

• Cheng-Chun Lin and L. Wang. 2015. Real-time forecasting of building fire growth and smoke transport. Submitted to Fire Safety Journal.
  
• Cheng-Chun Lin and L. Wang. 2014. Forecasting smoke transport in a compartment fire using Ensemble Kalman Filter. Journal of Fire Sciences. DOI: 10.1177/0734904114548837 (IF: 1.26)
• Cheng-Chun Lin and L. Wang. 2013. Forecasting simulations of indoor environment using data assimilation via an Ensemble Kalman filter. Building and Environment, Vol. 64: 169–176. (IF: 2.43)

Motivation:

• Data assimilation is a technique to combine numerical simulations with measured data from sensors to forecast future events, or reconstruct past events.
• In building simulations, measured data are often used to validate/calibrate a building model. If the comparison is good, the model is believed to be ok; If it is not good, reasons have to be provided.
• Data assimilation can make more use of measured data by combining them with numerical model, allowing one to "make sense" of the observations.

Research Highlights:

1. Indoor Contaminant Transport Prediction

   

2. Forecasting Building Smoke Transport During Fires

Real-time Forecasting of Building Fire and Smoke Spread for a 3-zone Case

(the case is 1:5 scale of the NIST fire test)

At 180 seconds, heat release rate increased from 1.5 kW to 2.8 kW. The system was able to respond dynamically: the forecasted parameters (lines) updated dynamically according to the measured data (dots).