6. Author’s publications
6.6 Paper 2
Reprint of the paper
-
Post-Fire Structural Assessment of a Firefighting Training Facility: A Case Study (2021)
Contribution of the author of this thesis to the paper
-
M. Benýšek is the co-author of the paper (that was written under the supervision of R. Štefan).
-
M. Benýšek mainly carried out the numerical simulations of the fire stated in Section 3 of the paper (including results presentation).
-
The author’s contribution is 40 %.
Post-re structural assessment of a reghting training facility:
A case study
Petr Müllera, Martin Bený²eka, Radek tefana,∗
aDepartment of Concrete and Masonry Structures, Faculty of Civil Engineering, Czech Technical University in Prague, Thákurova 7, 166 29 Prague 6, Czech Republic
Abstract
The paper focuses on a case study of a post-re structural assessment of a training facility for reghters. The study is conducted in order to assess the risk of damage of the load-bearing structures of the facility due to the repeated exposure to high temperatures. Flames during the re trainings are made up by gas burners which are installed inside the building.
Burners produce ames, smoke and high temperatures (up to 1000◦C). The duration of the re trainings last 3 minutes in maximum. After the description of the analysed building, this study is divided into two main parts a temperature analysis and a post-re structural analysis. The Computational Fluid Dynamics (CFD) model of re, implemented in the Fire Dynamics Simulator (FDS), shows as very powerful tool to simulate specic re scenario.
This tool, coupled with traditional structural diagnosis enables conducting the post-re assessment and determining the level of structural damage.
Keywords: Post-re Structural Assessment, Structural Diagnosis, Fire Fighting Training Facility, Models of Fire, CFD Model, FDS Software
1. Introduction
The structures are normally assessed for re resistance during the building design before possible initialization of the re. However, analysis of structures after the exposure of re is also very important. So, it is necessary to determine the extent of structural damage. It depends on a specic thermal action, re size, duration of the re, ventilation conditions, etc. on one hand, and actual re resistance of subjected structure on the other hand.
Some specic buildings are specially designed for repeating controlled res. These build-ings usually serve for reghters as a training facility. It serves for the training of movement and orientation in a smoky space, exposure of reghters to the high temperatures, even-tually for the training of inhabitants evacuation. One of these objects is analysed in the present paper.
Fire can be idealized by models of re. The basic mathematical models of re are the nominal temperature-time curves, such as the standard temperature-time curve, external temperature-time curve, etc. These curves are only time-dependent and they are the most conservative models. The more sophisticated models of re are the natural models which can be simplied (local res, parametric temperature-time curves) or advanced (zone models or computational uid dynamics models CFD), see e.g. [111].
∗Corresponding author.
Email address: radek.stefan@fsv.cvut.cz (Radek tefan)
Preprint submitted to Journal of Fire Sciences July 28, 2021
Usually, when a structure has been exposed to re, a post-re assessment is needed in order to prove whether the structure is still safe and reliable. During the assessment, as much information about the re event as possible have to be gathered foremost duration of the re and maximum reached temperatures. Moreover, material tests proving the actual mechanical properties are needed almost in every case. Also, the possibility of irreversible changes of a static scheme of the structural system has to be checked. Then, based on aforementioned information, the calculation of residual load-bearing capacity can be carried out. Results of such a calculation are then used for the decision about the future usability of the building. Generally, three possible decisions can be made: (i) the building is safe and reliable enough without refurbishment, (ii) the structure has to be strengthened (or acting loadings reduced), or (iii) the building has to be demolished for the reason that the strengthening and refurbishment is not possible nor cost-eective.
In order to obtain higher level of certainty of the post-re assessment, the information gained experimentally on site or thanks to re brigade reports should be accompanied by theoretical thermal analysis modelling a certain re scenario. CFD models represent modern approach which can describe very specic re scenarios assuming real conditions in the analysed space. Results of such calculations can be then used for post-re structural diagnosis and assessment [1316].
The present paper focuses on a case study of a post-re structural assessment of a training facility for reghters. It was decided to conduct the post-re structural assessment in order to evaluate the extent of the possible negative eects the re trainings should have on the structure in order to ensure the object's operation in the future. The assessment consists of two tasks modelling of re and post-re structural analysis.
Modelling of re focuses on restoring real re scenario and evaluating the temperature evolution in the building. The Computational Fluid Dynamics (CFD) model of re, im-plemented in the Fire Dynamics Simulator (FDS) [12], is employed for this analysis. The simulations help to determine the suitable positions of the core cut-outs of concrete speci-mens which are used for the post-re structural analysis.
Second part of the assessment consists of post-re structural analysis. At rst, prelim-inary and detailed inspection of the building is conducted. Based on gained ndings both concrete and steel specimens are extracted from structure and tested in laboratory in order to obtain actual material properties after the exposure to high temperature.
The results of the material tests are compared with theoretical material deterioration models with respect to the temperatures obtained using FDS simulations.
After that, residual structural performance is evaluated and appropriate refurbishment is designed.
The analysed building is described in Section 2. Input data for re simulations and the models of re created in FDS software are given in Section 3. Post-re structural analysis consisting of visual assessment, structural diagnosis, evaluation of results, thermal analysis of selected elements and assessment of residual load-bearing capacity is stated in Section 4.
In Section 5, summarizing conclusions are given.
2. Description of the building
The analysed building, see Fig. 1, is a part of reghters headquarters and is being used as a training simulator for reghter apprentices, who practice to ght real-scale res. For such purposes, several gas burners producing ames, smoke and high temperatures (up to
2
1000 ◦C) are installed inside the building. Although very high temperatures are reached during the trainings (according to the thermal power of each burner), the trainings last only for several minutes, usually no more than 3 minutes.
Figure 1: The analysed building.
The analysed building is a two-oor structure with one underground oor. Overall di-mensions of the building are approximately 13 m ×8 m and the height is 7.5 m. The structural system consists of reinforced-concrete (RC) walls and slab combined with steel oor beams. Roof structure consists of steel rafters and purlins and trapezoidal steel sheets.
RC walls are250 mmand 200 mmthick. RC slab above underground oor is200 mmthick.
Steel oor beams are I-shaped proles (IPE180). In order to mitigate the structural damage caused by repeating high temperatures exposure, all rooms where the burners are installed are equipped with protective sots and tiling made of steel sheets. The gap between the structure and protective layer is ventilated by fans during the training and also certain time after its ending. Thus, the structures are not exposed to the ames, high temperatures and extinguishing directly. Regardless of the positive eects of the protective measures, the structural assessment of the whole building was demanded by inner policy of the Czech Fire Rescue Service to specify the extent of structural damage caused by repeating re trainings.
For the reason that the building was originally not designed for this purpose, but it has been serving in this way for more than 10 years, the following questions are about to be answered:
To what extent do re trainings damage the structure?
Is the structure reliable and safe enough to continue serving in the same manner in long-term views?
Is it necessary to repair the structure or to strengthen it?
Based on the task, structural assessment was conducted in order to provide answers to the aforementioned questions. Within the assessment, the following steps were carried out:
Detailed visual assessment of a structure.
3
Structural diagnosis consisting of extracting samples of both concrete and steel, on-site non-destructive testing (NDT) of concrete and laboratory destructive testing of concrete and steel specimens, and evaluation of test results. Since the possibilities of diagnosis testing were limited, the material deterioration was also calculated using the theoretical material deterioration models proposed in [1719], with respect to the tem-peratures obtained using FDS simulations (see Section 3). Results of both approaches were compared.
Calculation of residual load-bearing capacity.
Design of necessary refurbishment.
The oor plans of the analysed building, with the locations of the gas burners, are shown in Figs. 2 and 3.
GAS BURNER
13000
2000 2500 1550 900 1040 1000 1000 1000 2010
25001000100010002500 8000
2015 1000 1000 1000 985 1000 1000 1000 1000 1000 2000
448010002515
GAS BURNER
GARAGE ENTRANCE MAIN ENTRANCE
GAS BURNER
1,5 MW
1,5 MW 3,0 MW
1.03 back room
1.01 front room 1.02 garage
SIMULATION 1 SIMULATION 2
SIMULATION 3
Figure 2: The rst oor of the analysed building.
4
100 2000 1000 1000 1000 2990 1000 1000 1000 2010 100 13000
30025001000100010002500300 8000 GAS BURNER 1,5 MW
2.07 back room
2.06 living room 2.02 dining room
2.05 bedroom
2.01 coridor 2.04 kitchen
2.03 room
SIMULATION 4
Figure 3: The second oor of the analysed building.
3. Modelling of re
For the modelling of re in the analysed building, it was needed to create the CFD simulations which were done in the FDS software [12]. The gas burners were used, positions and values of the heat release rate are depicted in Figs. 2 and 3. Based on the results, the appropriate positions (in areas with the highest temperatures) for the concrete core cut-outs were determined. These concrete core cut-outs were subsequently used for the structural analysis, see Section 4. LPG is used as the fuel of the burners. The trainings last only for several minutes while most often they last up to 3 minutes. As no burner was installed in the underground oor, this oor is not analysed in this study.
3.1. Input data for re simulations
The conditions in the analysed building during the trainings and the appropriate input data employed for the FDS simulations are summarised below:
only one burner is active at the same time,
doors and gates are fully opened (both indoors and outdoors),
some windows are without lling and for the simulation, these windows are assumed as permanently opened,
some windows (highlighted by dashed lines with x markers in the plans) are covered by metal sheet, these windows are omitted in the simulations, see Figs. 2 and 3,
5
space between the sot and the load-bearing structure in the rst and the second oor (except room No. 1.03) is intensively force-ventilated, ventilators are in progress during the training and couple minutes after its ending; the ventilation was neglected in FDS simulations,
protective tiling of walls were also neglected in FDS simulations,
space between the sot and the load-bearing structure in the rst oor in room No.
1.03 is not intensively force-ventilated,
burners are equipped by the shielding plates which inuence the ow of the hot gasses.
3.2. Fire simulations in FDS
The Fire Dynamics Simulator (FDS) [12] software was used for the simulations of re in the analysed building to predict the temperature distribution in the space and to determine the most suitable positions for core cut-outs of the concrete for following analysis of the structure. As a pre-processor, the Pyrosim software [20] was used. For visualization of results, the Smokeview software [21] was applied. The overall model of the building is shown in Fig. 4.
Figure 4: Overall visualization of the analysed building created in the Pyrosim software [20] (side wall is invisible).
During the simulations, the material properties are constant and they are considered according to EN 1992-1-2 [17]. The number of cells for the computational mesh is 143000, the cell size is 0.2 m×0.2 m×0.2 m, division method is uniform. Other parameters are as follows:
re simulation time 180 seconds according to the real training duration,
ambient temperature20 ◦C,
ambient pressure 1013.25 hPa,
relative humidity40.0 %,
6
simulation type Very Large-Eddy Simulation (VLES) [12],
temperatures were measured via thermocouples under the ceiling (mesh0.5 m×0.5 m), see Fig. 5.
Figure 5: Visualization of the thermocouples in Pyrosim [20] (room 1.02; front wall is invisible).
The heat release rates of burners in simulations were set to1.5 MWor3.0 MW, according to the real power of the burner, see Figs. 2 and 3. The area of the burners was set as 0.2 m×0.2 m. Each burner starts in t = 0.1 s with maximum released energy. With the respect of the burners location, four simulations in FDS were created:
No. 1: burner in the middle of the room 1.02, HRR = 3.0 MW, see Fig. 6,
No. 2: burner in the corner of the room 1.02, HRR = 1.5 MW, see Fig. 7,
No. 3: burner in the corner of the room 1.03, HRR = 1.5 MW, see Fig. 9,
No. 4: burner in the corner of the room 2.07, HRR = 1.5 MW, see Fig. 10.
3.3. Results
The results from the simulations No.1 and No.2 are shown in Figs. 6 and 7. The front wall is set as invisible. Temperatures are shown graphically.
Figure 6: Simulation No. 1, t = 180 s, left gas temperatures, right wall temperatures (front wall is invisible); visualised by the Pyrosim software [20].
7
Figure 7: Simulation No. 2, t = 180 s, left gas temperatures, right wall temperatures (front wall is invisible); visualised by the Pyrosim software [20].
The graphs of temperatures from thermocouples and the heat release rate from simulation No. 1 are shown in Fig. 8 for the detailed description of the results. For the comparison, ve thermocouples above the burner with the highest temperatures were chosen. The gas burner was set in this case to3.0 MW, see Fig. 8. In simulation No. 1, the gas temperatures around the ceiling reach the maximum value approximately 920 ◦C, see see Fig. 8.
In simulation No. 2, the gas temperatures reach only approx. 620◦Cwhich was assumed because the burner had a lower value of the HRR.
Figure 8: Simulation No. 1, left heat release rate measured in FDS [12], right ve thermocouples above the burner with the highest gas temperatures; visualised by the Pyrosim software [20] and the FMC software [22].
The results of simulation No. 3, burner 1.5 MW, are shown in Fig. 9. It can be seen that the hot gases go through the spiral staircase into the room 2.07, where uncovered steel structures are exposed to high temperatures. The shielding plates which direct the ow of the hot gasses are placed around the burner. Hence, the hot gasses ow back to the side wall, see Fig. 9. Space above the ceiling of the room no. 1.03 is not force-ventilated, thus the high temperatures aect the ceiling structures.
8
Figure 9: Simulation No. 3,t= 180 s, left vectors of gas temperatures with smoke, right wall temperatures (back wall is invisible); visualised by the Pyrosim software [20].
In simulation No. 3, the gas temperatures around the ceiling reach the maximum value approximately 570 ◦C, see Fig. 9.
The burner with HRR 1.5 MW for simulation No. 4 is placed in room 2.07, see Fig. 3.
The results are shown below, see Fig. 10. In this case, gas temperatures reach approximately 770 ◦C and the shielding plates direct the ow of the hot gasses. The HRR for simulations No. 3 and No. 4 are shown in Fig. 11. The burners have in both cases 1.5 MW.
9
Figure 10: Simulation No. 4, t= 180 s, top left and bottom vectors of gas temperatures with smoke, top right wall temperatures (back wall is invisible); visualised by the Pyrosim software [20].
Figure 11: Heat release rate measured in FDS [12], left simulation No. 3, right simulation No. 4;
visualised by the FMC software [22].
10
4. Post-re structural analysis 4.1. Visual assessment
The post-re structural assessment has begun with preliminary inspection of the whole building. The function of the burners was demonstrated by a short test, see Figs. 12 and 13.
Figure 12: Example of re test in garage in the rst oor.
Figure 13: Example of re test near spiral staircase in the rst oor.
Within the preliminary inspection the construction system together with protection sub-structures was studied. Potential critical spots were identied according to positions of gas burners and visible deterioration. Also, information needed for conducting the FDS sim-ulations were obtained from the building's technicians, including specication of burners thermal power, their shape, position and type of the used gas. As a result of the preliminary inspection, extent of forthcoming structural diagnosis was determined.
As a following step of post-re assessment, detailed inspection of the building was carried out. Areas and structures inuenced by the re trainings were identied and drawn into structural drawings of all oors, see Figs. 14 and 15. Such areas were marked as ZONE 1 ZONE 3 with assumed maximal reached temperatures and duration. Individual structural elements were classied into damage classes according to their presumed damage level.
11
The classication was done in accordance with the methodology proposed in [23]. This classication of each aected element was also drawn into the mentioned drawings.
As a second part of the detailed inspection output the forthcoming structural diagnosis was planned. Spots for extracting cores out of RC walls and slab were dened as well as spots for extraction of steel specimens out of the oor beams. Then spots for conducting the rebound hammer tests were determined. All of mentioned information were drawn into the structural drawings. Table of used symbols can be found in Fig. 16.
GAS BURNER
2000 2500 1550 900 1040 1000 1000 1000 2010
25001000100010002500 8000
2015 1000 1000 1000 985 1000 1000 1000 1000 1000 2000
448010002515
Figure 14: Shape of the rst oor with drawn results of visual inspection.
Based on the performed visual assessment it can be stated that RC structures did not exhibit signicant damage. The structural element surfaces were directly visible as the protective tilling and sots were removed at the time of inspection. No extensive surface damage was found except of several places near the windows where hot gasses leave the building during the trainings. No spots with spalled cover layer were found. The reinforce-ment was thus nowhere exposed directly to air. Also, no buckled or ruptured rebars were found. The surface of concrete was either coloured to black by soot or had natural concrete colour. RC elements did not exhibit extensive deections.
Steel oor elements in ceiling structures above the rst and second oor are in very good condition since no evidence of corrosion, extensive deections, buckling or distortion was found. This is probably thanks to the eective system of ventilated sots. Based on this analysis, it can be concluded that the elements were aected by re trainings only to negligible extent. However, there is one exception the sot in the room with spiral staircase in the rst oor (ZONE 2) was not ventilated. Therefore, hot air could accumulate in the space above sot, even though its majority ew to the second oor through staircase opening due to the chimney eect. Nevertheless, steel beams in this location are much more corroded,
12
TRAPEZOIDAL STEEL SHEETS + STEEL RAFTERS AND PURLINSvariable+6,550 TRAPEZOIDAL STEEL SHEETS
+ STEEL RAFTERS AND PURLINSvariable+6,550
STEEL RAFTERS IPE120 STEEL PURLINS
IPE140
100 2000 1000 1000 1000 2990 1000 1000 1000 2010 100
13000
30025001000100010002500300 8000
12,5°12,5°
200 1325 1250 1250 1300 1275 1275 2350195019502350 1300 1250 1250 1325 200
MAX. TEMP. [°C]
Figure 15: Shape of the second oor with drawn results of visual inspection.
SPOT FOR EXTRACTING VERTICAL CONCRETE
SPOT FOR EXTRACTING HORIZONTAL CONCRETE CORE SPECIMEN FROM FLOOR SLAB, CORE DIAMETER D=100 mm
SPOT FOR EXTRACTING HORIZONTAL CONCRETE CORE SPECIMEN FROM FLOOR SLAB, CORE DIAMETER D=100 mm