Blast penetration and propagation inside a building L. OUERDANE Process & Technologies Division, Technip France

Size: px

Start display at page:

Download "Blast penetration and propagation inside a building L. OUERDANE Process & Technologies Division, Technip France"

Clifton Curtis
6 years ago
Views:

1 Blast penetration and propagation inside a building L. OUERDANE Process & Technologies Division, Technip France

2 Table of contents 1. Blast wave generation 2. Blast penetration into a building 3. Blast propagation in channels of a building 4. Conclusion 5. Wish list 2

3 1. Blast wave generation 3

4 Introduction Objective: generation of a given blast wave in order to assess interaction with structures far from combustion area (far field), i.e. combustion is not the issue 4

5 Introduction Objective: generation of a given blast wave in order to assess interaction with structures far from combustion area (far field), i.e. combustion is not the issue Blast loads for structural assessment (oil & gas industry) 5

6 Introduction Blast propagation for Defense & Space Assessment of safety distances for complex geometries Structural response of safety barrier (e.g. blast wall, barricades, ) Design of high explosives storage buildings 6

7 Blast wave generation Several methods available in FLACS require blast wave propagation : Vessel burst / pressure balloon, short overpressure duration UVCE without congestion increasing artificially the laminar flame velocity via specifications in a user-defined specie Gas explosion in a bangbox (congestion and confinement) Desired blast profile obtained and used as an input Peak pressure P + Pulse duration t + Sharp rise time t + Picture of a shock wave INERIS 7

8 Blast wave generation Comparison ME / FLACS UVCE without congestion FLACS

9 Blast wave generation Comparison ME / FLACS Gas explosion in a bangbox Evolution of blast wave profile function of distance No stiffening of blast wave profile while propagating 9

Blast wave generation Conclusion For large pulse duration value (200-500 ms) Large domain required for expansion wave far field

recommendation : cover pressure front/pressure rise region by at least 5 grid cells. Problem : Exceed no.

10 Blast wave generation Conclusion For large pulse duration value ( ms) Large domain required for expansion wave far field propagation (> 1 km) Going away from explosion source: Peak pressure OK but no stiffening of pressure wave profile (nonphysical) Gexcon recommendation : cover pressure front/pressure rise region by at least 5 grid cells. Problem : Exceed no. of control volumes Use of dumps along the path of the wave 26 dumps required in this case, i.e. 26 different grids h=500 m h=90 m New option in FLACS of plane blast wave generation?? 10

11 Blast wave generation Conclusion Validation needs Validation of blast capacity is less exhaustive than for the reactive phase Defense & Space in France heavily relies on analytical/empirical approaches like those presented in UFC , : Need to check FLACS performance in similar configuration in order to be able to extrapolate to different configurations Need to check whether these simple approach are actually valid Validations to be performed internally (on-going on spare time): Enclosure effects on blast propagation Blast propagation in corridors Blast in multi cellular building Blast penetration in confined area 11

12 2. Blast penetration into a building 12

13 Blast penetration into a building Use of the generated blast wave as input via rdfile Import of the building geometry Creation of new (smaller) domain + refined grid FLACS simulation Variable P Exterior blast wave entering a building 13

14 Blast penetration into a building FLACS vs analytical methods Comparison of overpressure at the entrance of the building with analytical methods based on shock waves (Defense / Nuclear industry) Key parameters Outside blast wave peak and duration Angle of impact of the incident blast wave Geometric configuration of openings Fill-up volume Output : Reduced average overpressure inside the building Influence of angle of impact α of the incident pressure wave on inside peak pressure Experimental chart from ISL 14

volume ofchamber t f = fill up time depending on blast duration 15 Ratio of interior

15 Blast penetration into a building Experimental chart method (nuclear explosion) ISL P Ef = incident exterior pressure P f = interior pressure F = area of opening V = volume ofchamber t f = fill up time depending on blast duration 15 Ratio of interior and exterior peak pressure of a blast wave entering a volume V Experimental chart from ISL

16 Blast penetration into a building UFC method U.S. Department of Defense Shock loads entering windows or doors P Pi A 0 න d(p P i ) = C L 0 V 0 t න dt 0 Where : C L = leakage pressure coefficient function of the pressure difference P P i P = applied exterior pressure P i = interior pressure A 0 = area of openings V 0 = Structure volume t = duration of exterior pressure 16

Blast penetration into a building FLACS vs analytical methods Results : Ratio of peak pressure Pint / Pext FLACS simulation Method used Experimental charts ISL

17 Blast penetration into a building FLACS vs analytical methods Results : Ratio of peak pressure Pint / Pext FLACS simulation Method used Experimental charts ISL UFC FLACS simulation PMAX inside the building after blast wave propagation Similar reduction of the overpressure inside the building 17

18 3. Blast propagation in channels of a building 18

19 Blast propagation in channels of a building FLACS simulation Variable P 19

20 Blast propagation in channels of a building Good assessment of the overpressure propagation in channels? Grid guidelines for inside openings (doors, windows )? Verification with analytical method of blast waves in tunnels gives similar magnitudes 20

21 4. Conclusion 21

22 Conclusion Difficulties generating the initial blast wave: huge domain, long calculation durations, several tests needed, no sharp rise time Blast wave penetration: Good modelisation with FLACS compared to experimental results Grid guidelines available? Entrance should be resolved with a sufficient number of grid cells Same condition than vent openings with at least 6-8 grid cells seems to work 22 Blast propagation in channels: results OK but grid guidelines needed Regular grid along path of the wave in channel Limit number of grid cells in the smallest dimension of the channel required? 1-2 grid cells sufficient through small openings? Same condition than vent openings?

23 5. Wish list 23

24 Wish list More validation of blast capacity for different configurations Creation of a module in FLACS or boundary conditions in order to directly generate a given blast wave Goal: enhance the realization of blast propagation studies Automation of multiple dumps in order to maintain the sharp rise time of a blast wave Parallelization of blast solver Automatic resolution of far-field issue 24

25 Thank you

The rapid evolution of

The rapid evolution of Shock Testing Miniaturized Products by George Henderson, GHI Systems Smaller product designs mandate changes in test systems and analysis methods. Don t be shocked by the changes. Figure 1. Linear Shock