1. Design Fires A Structural Engineering Perspective Dr M Gillie

2. Content Existing fire models Existing fire models –Standard Fire –Swedish curves –Parametric fires CFD models CFD models Future directions Future directions –Rein’s fires Structural behaviour in travelling fires Structural behaviour in travelling fires

3. Standard Fire Test Defined using a furnace Test (ASTM-E- 119, ISO-834, BS-476 Part 8) Defined using a furnace Test (ASTM-E- 119, ISO-834, BS-476 Part 8) Critical Temperature

4. Pros and Cons Has many limitations Has many limitations –Not based on real fire data –Test repeatability difficult –No cooling phase –Uniform heating –Uses gas temperature “not fair” But But –Widely used –Can be useful for crudely comparing products

5. Compartment Fires

6. Energy balance for a compartment – Swedish Method QcQc QWQW QWQW QLQL QRQR

8. Assumptions in Swedish method No heat built-up in pre-flashover phase of fire No heat built-up in pre-flashover phase of fire Temperature uniform in the compartment Temperature uniform in the compartment Uniform heat transfer coefficient in compartment boundaries Uniform heat transfer coefficient in compartment boundaries All combustion takes place in the compartment All combustion takes place in the compartment

9. Pros and Cons Limitations Limitations –Crude –Rather severe –Implicit expressions (Eurocode parametric curves solve this) –***Uniform fire…*** –***…hence maximum size of compartment*** But But –“Not bad” –Can be used in performance-based design

10. Zone Models Extension of parametric models Extension of parametric models Assume uniform temperatures in each zone Assume uniform temperatures in each zone Normally computer based Normally computer based Several commercial codes available eg. Ozone, CFast Several commercial codes available eg. Ozone, CFast Similar drawbacks and benefits to parametric curves Similar drawbacks and benefits to parametric curves

11. Fires Travel Large compartment test at BRE showed travelling fire behaviour Large compartment test at BRE showed travelling fire behaviour Travel times of the order of 45 minutes within compartment 5m deep Travel times of the order of 45 minutes within compartment 5m deep Crib Ventilation 5.5m

12. …other Evidence

13. Defining Temperature Loading Need simple model that can capture effects of Need simple model that can capture effects of –Travelling fire –Ventilation conditions –Cooling Preliminary work undertaken on simplified travelling fire loading by Rein et. al. Preliminary work undertaken on simplified travelling fire loading by Rein et. al. NOTE: CFD calculations too complex for design work and… NOTE: CFD calculations too complex for design work and… …can not predict burning behaviour anyway …can not predict burning behaviour anyway

14. Rein’s Fire Model Rein proposes a near field and far field model of temperature loading Rein proposes a near field and far field model of temperature loading Fire field moves around compartment Fire field moves around compartment

15. Proposed Fire Model 200-600C 1200C Compartment Structure Cool Hot Cool Need size of fire, gas temperatures and nature of travel!

16. Proposed Gas Temperatures 1200C Distance from seat of fire T Size of fire Size of fire Good ventilation Limited ventilation

17. Travelling Behaviour Typical office fire load =570MJ/m 2 Typical office fire load =570MJ/m 2 Rate of burning = 500 kW/m 2 Rate of burning = 500 kW/m 2 Therefore 19 minutes in every location Therefore 19 minutes in every location Total fire time= 19A c /A f Total fire time= 19A c /A f Local fire T=0 Local fire T>0 Local fire Later on Fire compartment

18. Travelling Behaviour What “path” does the fire take? What “path” does the fire take? We can’t know We can’t know Depends on details of compartment and chance Depends on details of compartment and chance

19. Influence on Structural Behaviour y x Travelling fire Composite frame “Cardiington like”

20. Influence of Temperatures

22. Conclusions Current assumptions probably conservative Current assumptions probably conservative Travelling fires in large compartments may lead to significantly lower design forces Travelling fires in large compartments may lead to significantly lower design forces –Current assumptions very severe Possibilities for significant savings in fire protection Possibilities for significant savings in fire protection Work to date of conceptual nature Work to date of conceptual nature