Available online at www.sciencedirect.com
ScienceDirect Procedia Engineering 161 (2016) 137 – 142
World Multidisciplinary Civil Engineering-Architecture-Urban Planning Symposium 2016, WMCAUS 2016
Steel Connections Deterioration Coefficient Introduction for Post-Earthquake Fire Analysis Tudor Petrinaa,* a
UTCN, 15 C. Daicoviciu Str., 400020 Cluj-Napoca, Romania
Abstract This paper follows the results of an experimental testing programme on steel beam-to-column end-plate connections on scale 1:1, realized in 2014. After a seismic movement, connections are the important parts of a structure that may carry permanent damage. This effect is not taken into consideration by today’s advanced models for fire analysis like Vulcan, Safir and other. In the first part, the deduction of the deterioration coefficient is done. In the second part the examples of fire and post-earthquake fire analysis by applying the coefficient are done. The conclusions are in terms of difference in fire resistance of the structure for the above mentioned cases. ©©2016 Authors. Published by Elsevier Ltd. This 2016The The Authors. Published by Elsevier Ltd.is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of WMCAUS 2016. Peer-review under responsibility of the organizing committee of WMCAUS 2016 Keywords: post-earthquake fire; end-plate connections; deterioration coefficient; fire tests; fire resistance;
1. Introduction The experimental testing programme described in Petrina [1], had as objective the achieve of information on the real behaviour of steel beam-to-column end-plate connections, used for steel structures made by 3D frames, subjected to fire, after the deterioration of the nodes produced during an earthquake. The specimens were identically executed, the sensitivity of the nodes, when varying the fire action parameters and applying connection degradation by cyclic loads (according to ECCS document [2]) was studied by help of experimental tests. The post-earthquake tests were done during tests 4 and 6 of the above mentioned programme. The results of these tests were compared with test 1 and 2, when the
* Corresponding author. Tel.: +40 747929585. E-mail address:
[email protected]
1877-7058 © 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of WMCAUS 2016
doi:10.1016/j.proeng.2016.08.511
138
Tudor Petrina / Procedia Engineering 161 (2016) 137 – 142
specimen failed under mechanic action at 20°C; with test 3 when cyclic action was applied on the specimen, following a special procedure, and after that, subjected to mechanic action at a temperature of 20°C; with test 5 and 7 when mechanic action was applied on new specimens at 600°C and 400°C, respectively. According to these parallels, the magnitudes of applied forces and vertical displacements on the free end of the cantilever were analysed, understanding what safety measures should be applied when designing beam-to-column end-plate connections that have to resist earthquakes, accidental fire action and also under the combined action of fire after earthquake. The complexity of modelling the fire behaviour of nodes has concerned a great number of researchers in the last years, as for example Faggiano and Mazzolani [3], Bursi [4] or Puccinotti et al. [5]. Their experimental results referred mainly about moment-rotation relations. In Yassin [6], by considering the structure deterioration due to seismic movement, a 2D steel frame having one span was modelled and analysed under the post-earthquake fire action by using Safir and Ansys advanced models. A four minutes’ difference in the fire resistance of that frame was obtained as shown in figure 1:
Fig. 1. Difference in fire resistance of 2D frame as obtained by Yassin [6].
2. Coefficient deduction A parallel between the response of the deteriorated connection / new connection under the fire action at 400°C shows us a difference in terms of forces of approximately 55%. The curves are similar for the two tested cases, having one jump corresponding to the fail of the first row of bolts. This was at an applied force of 110kN for the prior deteriorated connection and at 218kN for the second case. The second step in the shape of the curve corresponds to the fail of the second row of bolts at a force of 67kN for the deteriorated case and 150kN for the new connection at 400°C. Here, the same degradation of 55% was noticed. Making the same parallel at 600°C, we obtained a difference in the two curves of about 12%, but at 600°C, the bolts kept only 22% of their capacity. So, the behaviour at 600°C is mainly influenced by bolts degradation due to high temperature. Moreover, from numerical analysis made in Petrina [1], by help of advanced model for fire analysis Vulcan, we got that for this exact connection fixed in the test stand, for the designed (mechanic) load, the critical temperature would be around 535°C, when the complete failure of connection under fire occurs. Taking into considerations all these, when using steel beam-to-column end-plate bolted connections for columns made from H profile and beams made from I profile, the author suggests the introduction of a coefficient for performing fire analysis of structures, in a range from 0.55 to 0.45. The coefficient refers to structure’s nodes EI stiffness reduction, made in order to calibrate numerical nonlinear models able to make fire or high-temperature analysis, making possible also a post-earthquake fire analysis. This analysis should be done for buildings situated in zones with high probability of earthquakes. 3. Coefficient introduction for the studied sub-structure A steel beam, column, connection sub-structure was designed following all prescriptions mentioned in Petrina [7].
139
Tudor Petrina / Procedia Engineering 161 (2016) 137 – 142
The column is made up by H type profile having the thickness of flanges of 15mm and the thickness of web of 10mm and a total length of 3m. The cantilever beam is of I type profile with the thickness of flanges of 15mm and the thickness of web of 8mm and the total length of 2 m like in figure 2.
Fig. 2. Studied sub-structure.
Fig. 3. Fire analysis on the studied sub-structure.
In the following, a calibration of the advanced model Vulcan is done in order to analyze the sub-structure under postearthquake fire and see the difference in terms of maximum deflection and fire resistance of the ensemble. The substructure was modelled being considered fixed in the test stand, for better correlation with test results.
Displacement [mm]
20 0 -20 0
10
20
30
40 Time [min]
-40 -60 -80 -100 -120
Node #1 Node #6 Node #7 Node #13
-140 -160 -180 Fig. 4. Time-vertical displacement diagram corresponding the fire analysis.
The fire was considered on the entire sub-structure as shown in figure 3, following a bilinear real time-temperature curve. In the image, the studied structure is coloured red, with blue is the test stand. At a load corresponding to the designed values, the fire resistance in the case of fire analysis of not deteriorated connection is 24.5 minutes and the
140
Tudor Petrina / Procedia Engineering 161 (2016) 137 – 142
critical temperature is 542°C. The sign of failure is given by the deflection rate. After that, another analysis was performed, by introducing the degradation coefficient. The EI stiffness of the connection was reduced to 50% in accordance with chapter 2 of this study and the same magnitude of force was applied on the free end of the cantilever. The vertical displacement diagram is presented in figure 4 for the not deteriorated connection (fire analysis) and in figure 5 for the deteriorated connection (post-earthquake fire analysis). In the diagrams, node 1 is the free end of the cantilever. 20 0 Displacement [mm]
-20
5
0
10
15
20
25
30 Time [min]
-40 -60
Node #1 Node #6 Node #7 Node #13
-80 -100 -120 -140 -160 Fig. 5. Time-vertical displacement diagram corresponding the post-earthquake fire analysis.
The fire resistances and the maximum deflection of the cantilever beam are shown in table 1. One may notice that in the case of the fire analysis on the studied sub-structure, the fire resistance is four minutes greater than in the case of the post-earthquake fire. A parallel between the displacement curves in the two cases underlines the lower capacity of the deteriorated connection under fire as in figure 6. Table 1. Fire and post-earthquake fire analysis on the sub-structure by deterioration coefficient introduction. FIRE RESISTANCE [minutes] 32 28
FIRE POST-EARTHQUAKE FIRE
MAXIM. DISPLACEMENT [mm] 167 149
0
Displacement [mm]
Ͳ5
Ͳ20
0
5
10
15
20
25
30
35 Time [minutes]
Ͳ40 Ͳ60 Ͳ80 Ͳ100 Ͳ120 Ͳ140 Ͳ160
-new connection -deteriorated connection
Ͳ180
Fig. 6: Time-vertical displacement diagram corresponding to the fire / post-earthquake fire analysis.
141
Tudor Petrina / Procedia Engineering 161 (2016) 137 – 142
4. Example – Steel structure post-earthquake fire analysis An example of applying the deterioration coefficient of the connection was studied for the case of a building made by 3D steel frames and reinforced concrete plates having the 2 openings of 10m on one direction and the 3 openings of 7.5m on the other direction, with 3 stories. The columns, beams and their connections are of the type studied in the previous chapter. The fire and post-earthquake fire analysis was performed by the help of the advanced model Vulcan, which takes into consideration the material and geometry non-linearity at different temperatures attained by the layers of the materials. The software offers the possibility of simulation of all parts of the structure like plate, reinforcement, steel frames. A case of an ignition in a corner compartment as in figure 7 is presented in the following. Fire behaviour of the structure in the 2 cases was observed and compared. The collapse of the studied structure is shown in figure 7.
Fig. 7. Collapse of structure due to fire.
The fire resistance of the structure differs for the two cases, as shown in table 2, due to connection deterioration by 50% for the second case and the difference between the maximum displacements corresponding to collapse is equal to 21 mm. Table 2. Fire and post-earthquake fire analysis of the real building by deterioration coefficient introduction. FIRE RESISTANCE [minutes]
MAXIM. DISPLACEMENT [mm]
FIRE
77
147
POST-EARTHQUAKE FIRE
71
168
A parallel between the vertical displacements of the nodes from an affected beam, for the two cases, is shown in the diagram from figure 8. For the analysed structure, taking into consideration the deterioration of the connections due to earthquake, the fire resistance decreases by 8% and the maximum displacements increase by 12%. 5. Conclusions In this work, a deterioration coefficient due to earthquake action was experimentally deduced and applied then on a numerical fire analysis performed on the experimentally tested sub-structure and also on a real structure. The goal was to notice the behaviour of a structure under fire and also under post-earthquake fire. The calibration of the advanced model by deterioration coefficient introduction showed a different behaviour of the structure in the case of postearthquake fire. The results obtained by applying this procedure (experimentally deduce the coefficient + calibration of advanced model) were similar to results of other authors which applied numerical methods to find out the deterioration of the structure.
142
Tudor Petrina / Procedia Engineering 161 (2016) 137 – 142
0 Vertical displacement [mm]
Ͳ20
Ͳ20
0
20
40
60
80
100
Time [min]
Ͳ40 Ͳ60 Ͳ80 Ͳ100 Ͳ120 Ͳ140
Node#585FIRE Node#727FIRE Node#790FIRE Node#585POSTͲEFIRE Node#727POSTͲEFIRE
Ͳ160 Node#790POSTͲEFIRE Ͳ180 Fig. 8. Time-vertical displacement diagram corresponding to fire/post-earthquake fire analysis of the real building.
References [1] T. Petrina, Numerical and experimental analysis regarding fire resistance of structures – PhD Thesis, UT Press, Romania, 2014 (in Romanian) [2] Technical Committee 1 (ECCS) Recommended Testing Procedure for Assessing the Behaviour of Structural Steel Elements under Cyclic Loads, no. 45, 1986 [3] B. Faggiano, F.M. Mazzolani. Methodology for robustness assessment of structures subjected to fire following earthquake through a performance base approach. COST TU0904, 2010 [4] O.S. Bursi. Prefabricated composite beam-to-concrete filled tube or partially reinforced concrete encased column connections for severe seismic and fire loadings – Final Report RFSCR-03034, 2008 [5] R. Puccinotti, O.S. Bursi, J.M. Franssen, T. Lennon, Seismic-induced fire resistance of composite welded beam-to-column joints with concrete filled tubes, Fire Safety Journal. 46 (2011) 335-347 [6] M. H. Yassin. Post earthquake fire performance of building structures. Ph.D. Thesis, Concordia University, Quebec, Canada, 2010 (in English) [7] T. Petrina, Computation of structural fire resistance of steel sections, Acta Technica Napocensis (ISSN 1221-5848). 54 (2011), 308-317
