Performance of Base-Isolated RC School Building under Blast Loading

Abstract

It is known that bomb-laden vehicles target many buildings as a result of terrorist activities. The effects of such attacks must be reduced and structures must be protected against blast effects. In high seismic hazard regions, buildings are designed to be earthquake resistant. One of the methods used to dampen earthquake effects on structures is base isolation. Base isolation is effective in distributing blast loads to the structure, similar to seismic loading. In this study, the effects of this distribution on the structure will be evaluated. The scope of the study encompasses the numerical estimation of explosive loads at different distances according to the explosive material carrying capacity of various vehicles and their effects on structures. Linear and nonlinear analysis methods were used to compare the dynamic behaviour of school buildings designed with and without base isolators. The Turkish Building Earthquake Code (2018), which includes the maximum seismic loads, was used in the design of the isolators. Numerical analysis was performed using SAP2000 software based on the finite element method. The blast loads were applied to the floors of the building using the direct integration method in the time domain. FEMA standards were used to determine the blast loads and the performance was evaluated in comparison with the numerical analysis results. As a result of the study, it was concluded that structures with base isolators are efficient in reducing the effects of an explosion at certain distances and these distances will affect the design of the shelter walls.

1. Introduction

Explosions have historically been a preferred method of terrorists for a variety of reasons. The main reason for this preference is that it is easy and fast in terms of process management and implementation. Bomb-laden vehicles are preferred as they can reach the target easily and the weight of the charges that they carry can be increased in proportion to the size of the vehicle. The damage to the structures because of the explosions varies according to the size of the explosive, the standoff distance, the characteristics of the structure and many other parameters [1]. Many researchers have conducted experimental and numerical studies on the estimation of blast loads. First of all, numerical studies were conducted to estimate the pressure of the blast load and various formulations were developed [2,3,4,5,6]. The usage of all kinds of tools and equipment with explosion risk, materials used in the defence industry, wars, terrorist activities, etc., all require explosion safety. For this reason, various regulations and reports related to explosion have been issued in many areas. Design of Structures to Resist Nuclear Weapons [2], Design of Structures to Resist the Effects of Accidental Explosions (Army Technical Manual) [7], Design of Blast Resistant Buildings in Petrochemical Facilities [8], and Unified Facilities Criteria [9] are various reports and regulations used nationally and internationally in the definition of blast loads. FEMA 426 (Risk Management Series Primer for Design of Commercial Buildings to Mitigate Terrorist Attacks, Providing Protection to People and Buildings) [2], FEMA 427 (Risk Management Series Primer for Design of Commercial Buildings to Mitigate Terrorist Attacks, Protecting Office, Retail, Multi-Family Residential, and Light-Industrial Facilities) [1], FEMA 452 (Risk Assessment—A How-To Guide to Mitigate Potential Terrorist Attacks Against Buildings) [10], FEMA 453 (Risk Management Series Design Guidance for Shelters and Safe Rooms) [11] and FEMA 430 (Risk Management Series Site and Urban Design for Security Guidance Against Potential Terrorist Attacks) [3] are published guides on the explosion of terrorist attacks. Karlos and Solomos [12] wrote a report containing blast load calculations. These regulations and reports are sources for estimating explosion effects on structures.

Table 1. Typical overpressure damage [1].

Incident Overpressure [kPa]	Damage
1.03–1.52	Typical window glass breakage
3.45–7.58	Minor damage to some buildings
7.58–12.41	Panels of sheet metal buckled
12.41–20	Failure of concrete block walls
34.48+	Collapse of wooden-framed buildings
27.58–48.26	Serious damage to steel-framed buildings
41.37–62.06	Severe damage to reinforced concrete structures
68.95–82.74	Probable total destruction of most buildings

shows the estimated damage table given in FEMA 426.

Minimum safe distances needed for preventing injury and damage for safe design and risk management of potential terrorist attacks are given in FEMA 430 (

Table 2. Injury or Damage Related to Stand-off [14].

Injury and/or Damage	Stand-Off (m) (226 kg TNT)	Stand-Off (m) (2260 kg TNT)
Threshold of failure of concrete columns	9	18
Potentially lethal injuries	45	105
Injuries from wall fragments or to people in the open	45–75	105–150
Severe glass wounds (glass with applied film)	75	200
Severe glass wounds (unprotected glass)	150	300+
Minor cuts	240	300+

) [3]. In addition, safe following distances for explosive transport are determined by considering the maximum number of explosives that cars can carry based on their luggage capacity (

Table 3. Improvised Explosive Device (IED) Safe Stand-Off Distance [13].

Threat Description (TNT Equivalent)	1 Explosives Capacity [kg]	2 Mandatory Evacuation Distance [m]	Shelter-in-Place Zone [m]	3 Preferred Evacuation Distance [m]
Pipe Bomb	2.3	21 m	22–365 m	366 m
Suicide Bomber	9	34 m	35–518 m	519 m
Briefcase/Suitcase	23	46 m	47–563 m	564 m
Car	227	98 m	99–579 m	580 m
SUV/Van	454	122 m	123–731 m	732 m
Small Delivery Truck	1814	195 m	196–1158 m	1159 m
Container/Water Truck	4536	263 m	264–1554 m	1555 m
Semi-Trailer	27,216	475 m	476–2834 m	2835 m

¹ Based on the maximum amount of material that could reasonably fit into a container or vehicle. Variations possible. ² Governed by the ability of an unreinforced building to withstand severe damage or collapse. ³ Governed by the greater of fragment throw distance or glass breakage/falling glass hazard distance. These distances can be reduced for personnel wearing ballistic protection. Note that the pipe bomb, suicide bomb, and briefcase/suitcase bomb are assumed to have a fragmentation characteristic that requires greater stand-off distances than an equivalent amount of explosives in a vehicle.

) [13].

Developed using past explosion events, analytical relations related to the blast load are supported by experimental studies using scaling laws in the ongoing process [4,5,6,7,8,9,14,15,16,17]. Experimental studies were conducted based on members in column beams and storeys. In addition to these studies, numerical studies on the estimation of blast loads and their effects on the building system were also carried out [13,16,18,19,20,21,22,23,24,25,26,27].

Zhang and Philips (2015) [28], Zain Kangda et al. (2018) [29], and Tolani et al. (2022) [30] have indicated through their studies that seismic base isolators may be helpful against blast effects. Zhang and Philips (2015) [28] examined the effects of lead rubber bearings on structural responses to blast loads by creating a matrix model of multistorey structures in Matlab software. The blast load was applied to the structure by multiplying the force function by effective areas. Zain Kangda et al. (2018) [29] expanded on the work of Zhang and Philips (2015) [28] and employed various damping effects and blast models in their study. Tolani et al. (2022) [30] evaluated the effect of base isolators using a nonlinear analysis method for an isolated structure with different damping effects. When considering the behaviour of isolated structures under blast effects, further studies are needed. Codes for blast load design have been developed, which provide information on safe blast distances. However, no study in the literature compares these codes for the structural dynamic behaviour of seismic-isolated structures. The blast distances specified in the standards are based on general evaluations derived from observations. However, the resulting damage may vary due to structural differences. Therefore, numerical simulations are required for the dynamic behaviour of isolated structures subjected to blast loads. In this study, based on the assumption of an attack on an existing structure and using blast standards, the performance of seismically isolated and fixed-base structures was examined for the safe distances specified in the standards for a selected explosive charge weight. The performance analysis was performed for a three-storey reinforced concrete school building with fixed-base and base-isolated structural elements according to the Turkish Building Earthquake Code (2018), considering plastic rotations in the structural elements [31]. Thus, the realism of numerical blast predictions was comparatively evaluated by considering FEMA 427–426–430, etc., standards for blast load calculations.

2. Materials and Methods

Within the scope of the numerical study, design of blast load, calculation of safe standoff distance, examination of dynamic behaviour of isolated and non-isolated structures under the effects of blast loads, and comparison of building performances were made. At the end of the blast load analysis, the analyses were carried out using linear and nonlinear behaviour methods for the seismic isolated and fixed-base structure. The variation of base shear forces and storey displacements were studied. Moreover, performance analyses for all cases according to the damage limits given in the Turkish Building Earthquake Code 2018 [31] were made by calculating the plastic rotations that occur in the columns and beams section.

The isolators have been designed with reference to the limits of displacement per unit. Lead core rubber elastomers were used for the base isolation. A suitable size was selected for the lead core diameter and elastomer diameter of the isolator to match the column dimensions. The effective yield strength of the isolation unit is approximately equal to the characteristic strength at TBEC 2018. The target damping ratio for the lead core rubber isolators is in the range of 15% to 35%, depending on the mechanical properties of the isolator. The yield strength of the lead material was assumed to be 11 MPa and the characteristic strength of the isolator was obtained. The cyclic behaviour of the lead core rubber elastomer was determined from the data obtained.

For the design of the explosion load, safe explosion distances were determined according to FEMA standards [1,2,11,32], and the explosion load function proposed by Mills [33] was used for these distances. According to Mills’ explosion load function, explosion effects are transmitted according to scaled distances so that the explosion load will have different effects on each floor. Therefore, the finite element method was used in the analysis, and SAP2000 software was used to apply the loads in the time domain using the direct integration method.

3. Theory

Explosive properties, standoff distances and the environment of the explosion are important in blast load design. The most commonly used correlation, which exemplifies the time-dependent behaviour of the explosion, is the equation developed by Friedlander (Equation (1)) [18].

$$\begin{array}{l}{P}_s(t)=P_o+P_{so}(1-t/t_o)\exp (-bt/t_o)\end{array}$$

(1)

Here, P_o is the ambient air pressure, P_so is the peak incident overpressure, t_o is the total duration of blast wave and b is the parameter controlling the rate of wave amplitude decay coefficient.

A number of full-scale tests are desired while experimentally studying (or for other purposes) the effect of explosions on structures. However, these full-scale tests may be risky and costly. Hopkinson (1915) and Cranz (1926) developed the Hopkinson–Cranz Scaling Law (Cube–Root Scaling Law) to expand the applicability of experiments conducted at different scales [18,23]. There is a linear relation between explosive weight and standoff distance according to Hopkinson’s law (Equation (2)).

$$Z=R/{\sqrt[3]{W}}_e$$

(2)

Here, R is the standoff distance [m] from the blast source to the blasting point, and W_e is the equivalent explosive weights [kg]. The variety of explosives has led to a universal acceptance for the explosive parameters required in calculations. TNT, whose detonation property is similar to many explosives, was chosen in the generalization. The equivalent weight of TNT that relates the weight of the explosive to the equivalent weight of TNT will be calculated according to Equation (3) using the ratio of heat generated during blasting [23].

$$W_e=W_{\exp }{H}_{\exp }^d/H_{TNT}^d$$

(3)

Here, W_e is the equivalent weight [kg] of TNT, W_exp is the weight [kg] of the actual explosive, H^d_exp is the actual explosive heat [MJ/kg], and H^d_TNT is the heat of explosion of TNT [MJ/kg]. Mills, who expressed the scaled distance in kPa in peak incident overpressure calculations, developed the formula in Equation (4) [33].

$$P_{SO}=1772/Z^3-114/Z^2+108/Z$$

(4)

The explosion occurs as an instant impact load. Blast load profile (Figure 1) is defined to be the time-dependent function of the explosion. The time-varying blast load will be impacted on an existing base-isolated structure at building storey levels in the content of numerical work using SAP2000 software [34], in which analyses are based on the finite element method.

The design of seismic base isolators, according to TBEC 2018, is calculated based on the maximum displacement demand, target damping, and target period values. Base isolators reduce relative storey drifts by extending the structure’s period. As a result, the isolator stiffness is determined according to a longer target period value than a fixed-base structure. The isolators are sized by controlling unit deformation limits. The algorithm used in this study for isolator design is presented in the following Figure 2.

Explosion and Seismic İsolation

Blast loads differ from earthquake effects as they are not proportional to the mass of the building and cannot be equally distributed to the structural members. Therefore, modeling of linear or nonlinear behaviour, rigid diaphragm acceptance, plastic hinge, and other such definitions used in earthquake resistant structure design may not be realistic under random loads such as explosions. One of the methods used in damping earthquake loads on structures is base isolation. Although it is mandatory to use base isolation in hospitals in Turkey today, it can also be used for reinforcement purposes. Seismic dampers used in the base isolation limit interstorey drift and prevent structural members from being strained by extending the period of the building [20]. The dampers used in base isolation ensure that the structures have a different behaviour compared to fixed-base systems by ensuring the horizontal movement of the structures. One important aspect that this situation reveals is to ensure that the blast loads, which do not show uniform distribution, are distributed more uniformly throughout the structure [28]. Since the effects of the explosion are not as long-lasting as the effects of the earthquake, the loading is completed before the building starts to move. In this case, the effects of explosion and earthquakes differ in terms of the damage they cause [35] (Figure 3).

4. Finite Element Modelling

In this study, an existing school building was used in the analyses and a seismic base isolation design was made for this building. The existing structure was modeled as base-isolated using lead rubber bearings and the finite element method. In the building structural system, the columns were 350 × 600 mm in size and the beams were 300 × 600 mm in size, and the building storey plan is given in Figure 4. TBEC 2018 [31] was taken as a reference in the design of the dampers used in the base isolation. Detailed structure features are given in

Table 4. Features of structures and isolators.

	Feature of Building
Grade of Concrete	C25/30
Grade of Steel	S420
Storey Height (Units: m)	3.80
Size of beams (Units: mm) Reinforcement Arrangements for beam sections	300 × 600 3ϕ14 + 2ϕ12
Size of columns (Units: mm) Reinforcement Arrangements for column sections	350 × 600–600 × 350 16ϕ16
Slab Thickness (Units: mm)	140
	Feature of Isolators
Number of isolators	28
Design period of DBE, TD (Units: s)	2.5
Effective Stiffness, ke (Units: kN/m)	405.69
Post-Elastic Stiffness/Lead Core Stiffness, k2/k1	0.077
Yielding Force Fy = FQ (Units: kN)	115.60
Total Design Displacement; DD (Units: mm)	93
Maximum Displacement DM (Units: mm)	778
Damping	0.21

.

In order to determine the rotation rates at the column and beam ends, plastic hinges were defined at the ends of the columns and beams while modeling the structures in nonlinear calculation methods. Additionally, moment–curvature diagrams were used in defining plastic hinges. In the definition of column joints, plastic hinges were defined for each column by entering the vertical reaction force values (normal force) of each storey (Figure 5).

The existing structure was concrete class C25 and the Mander confined concrete model [36] was used (Figure 6). Longitudinal reinforcements were S420 steel grade and transverse reinforcement was S220 steel grade.

Scaling laws were used to estimate the blast loads to be impacted on the structure. Peak incident overpressure was obtained by obtaining blast load functions by keeping the blast load constant at different distances. Thus, an estimation was made about possible building damage using FEMA 426 [2] data (Table 1). In addition, an assumption that a hatchback was detonated was made to be used in calculations according to the explosive carrying capacity of the vehicles. Blast loads were obtained using Hopkinson scaling laws. Explosion effects were applied to each storey of the building at different distances, and load functions expressing the effects of the vehicles exploding at different distances on the storey levels were calculated and are given in Figure 7.

When the maximum explosion pressures are evaluated according to the distances, it is estimated that structures exposed to blast pressure in the range of 68.95–82.74 kPa will collapse according to Table 1. In the range of 12.41–20 kPa, it is expected that there may be collapses in the walls. In this case, a structure exposed to 227 kg TNT explosion from 5 m and 10 m distance is expected to be at the collapse level. Although it is estimated that the structure will be damaged at a distance of R_G = 15 m and R_G = 20 m, it is expected that the structural system will not be damaged at a distance of R_G = 20 m. According to Table 2, it is seen that the columns reach the damage threshold when an explosion with the effect of 226 kg TNT occurs at 9 m. For this reason, in order to compare the explosion effects in structures with and without base isolation, performance analyses were conducted based on the assumption that a hatchback-type vehicle with a carrying capacity of 227 kg TNT for R_G = 10 m (Performance analysis 1) and R_G = 15 m (Performance analysis 2) was detonated (Figure 8a). In order to transfer the blast load to the structure, the unit area loads were calculated at each floor level along the surface of the building, and dynamic analyses were carried out in this way (Figure 8b).

5. Numerical Analysis and Results

In the first stage of the study, modal analysis was performed in order to examine the effect of seismic isolation on the free vibration behaviour of the structure and, as a result of the analysis, free vibration mode shapes and frequencies were obtained for each system (Figure 9).

Performance limits were calculated according to TBEC 2018 (Turkish Building Earthquake Code-2018) (

Table 5. Performance damage limits for TBEC 2018 [31].

Performance Limits
LS-1 (Immediate occupancy)	θp = 0
LS-2 (Life safety)	θP = 0.75 θP(LS-3)
LS-3 (Collapse Prevention)	θp = (2/3(θu − y)Lp (1 − 0.5 Lp/Ls) + 4.5ϕudb)
LS: Limit state

). Plastic rotations in construction elements were evaluated according to TBEC 2018 damage limits.

As a result of the explosion analyses, the effect of explosives at different distances on the behaviour of the isolation units was also examined and the results were obtained comparatively. Isolator capacity is calculated as (D_M) 0.778 m. When a vehicle with the effect of 227 kg TNT explodes at a distance of 5 m and 10 m, it is seen that the isolation unit capacities targeted according to TBEC 2018 are exceeded and deformed, but if the same vehicle explodes at a distance of 15 m and 20 m, the isolation unit remains in the safe zone (Figure 10).

In order to compare the building performances at different standoff distances, the shear forces occurring at the base of the building and the storey displacements were compared for the limit R_G = 10 m where the islator capacity was exceeded, and the distance R_G = 15 m where the isolation unit was in the safe zone. The reason for not making any comparison for R_G = 5 m is the very high excess of the capacity. The high explosion pressure at 5 m, according to the data given in Table 2, also supports the analysis results. The time dependent variation of the base shear forces occurring in the structure as a result of the explosion at R_G = 10 m and R_G = 15 m is given comparatively in Figure 11 for base-isolated and non-isolated structures, considering the linear and nonlinear behaviour. The change of distance significantly differentiates the explosion effects occurring on the structure. It is seen that the base shear forces decrease in the building where seismic base isolation was applied. When linear analysis and nonlinear analysis methods are compared, nonlinear behaviour and permanent deformations occurred in the structure in large explosive effects (close standoff distance or increase in explosive amount) and smaller base shear forces occurred due to ductility according to linear analysis.

Since it was estimated that two structures with and without seismic isolation would collapse for R_G = 10 m, a detailed comparison of plastic hinge formations in the base-isolated and non-isolated structures was made for R_G = 15 m, and the results are given in Figure 11. When Figure 12 is examined, although it was seen that there was a dense plastic hinge formation and some columns that became mechanisms in the seismic non-isolated state, the structural frame members in the isolated state were largely protected from explosion effects. In addition, there was no collapse at the ends of the structural frame members in seismic isolation.

Base isolation also affected the rotation rates in the structural members under the effect of explosion. Therefore, in order to evaluate the performance of base-isolated and non-isolated structures, the performance evaluations of structures exposed to explosive effects at distances R_G = 10 m and R_G = 15 m were compared with the plastic rotation limit values of the damage levels in TBEC 2018 of the plastic rotation rates occurring in the column and beam sections as a result of the explosion. The plastic rotations of the columns and beams of seismic isolated and fixed-base structures, which would be caused by vehicle explosion effects at different distances, were calculated and compared according to the TBEC 2018 damage limit values (Figure 13 and Figure 14).

When the rotation rates of the beams were compared, it was observed that although less rotation occurred in seismic isolated structures for the first storey compared to the fixed-base structures, more rotation occurred in the seismic isolated structure for R_G = 10 m compared to the fixed-base structure for the second storey. However, there was no significant difference between two different building systems for R_G = 15 m. When the third storey beams were evaluated, it was seen that there was no difference between seismic isolated and fixed-base structures. The reason for this is that the explosion causes local damage and, as the number of storeys increases, the effects of the explosion decrease as the distance increases. When comparing performance analysis 1 and 2 for beams, all of the beams were in the limited damage zone for R_G = 15 m, while part of the beams reached a controlled damage level for the first and second storeys for R_G = 10 m.

When the rotation ratios of the columns were compared, it was observed that although less rotation occurred in seismic isolated structures compared to fixed-base structures for the first storey, more rotation occurred in the seismic isolated structure for R_G = 10 m compared to the fixed-base structure for the second storey. However, when the rotation rates occurring in the first storey columns were examined, it was observed that high rotation rates were formed that would lead the fixed-base structure to the collapse zone. For R_G = 15 m, high rotations occurred in the rigid bonded system in the first storey columns, while the seismic isolated building columns remained at limited damage level.

As a result of the performance analyses, both structures were in the state of collapse under the effect of 227 kg TNT in performance analysis 1 (R_G = 10 m). On the other hand, in performance analysis 2 (R_G = 15 m), the seismic isolated structure was at the advanced damage level, while the non-seismic isolated structure was in the state of collapse under the effect of 227 kg TNT. In Figure 15, ‘IO’ indicates immediate occupancy, ‘LS’ indicates life safety, ‘CP’ indicates collapse prevention, and ‘C’ indicates collapse. The percentage of exceeding performance limits has been expressed as an exceedance of these limits.

Consequently, performance analyses of seismic isolated and non-isolated structures were conducted to examine the nonlinear behaviour of seismic isolated and non-isolated structures under explosion effects, and to examine the behaviour of structural members. While the seismic isolated structure provided the performance level of preventing collapse, the structure without seismic isolation did not. In addition, considering the plastic hinge densities, it was observed that the plastic hinge density shifted from the first storey to the upper storey in the seismic isolated condition. When the performance-based evaluation was made according to nonlinear analysis methods, it was concluded that seismic isolated and fixed-base structures did not provide life safety for both performance analysis cases, although the isolators used in the base isolation had a positive effect on the structure performance in damping the blast loads.

6. Conclusions

In this study, nonlinear analyses of an existing school building under the effect of blast loads applied from different distances were performed and the building performances were examined according to TBEC 2018. In the analyses, lead rubber bearings were used as an isolator to examine the behaviour of the structure against blast loads when it was fixed base or base isolated. The results were obtained comparatively by calculating the base shear force, storey displacements, and the amount of rotation in beams and columns. In light of the obtained data, the following conclusions were drawn.

• It was observed that the lead rubber bearings used in the base isolation were effective in damping the explosion effects, damping the impact loads by reducing the interstorey drift and reducing the base shear forces on the structure.
• When the base isolation was evaluated according to nonlinear methods, it was seen that the base isolation reduced the reaction of the building against explosion effects, changed the mode shape of the building, and affected the plastic rotation rates in the building.
• Considering the nonlinear behaviour, plastic rotations concentrated on the base storeys in the non-isolated structure helped to reduce the explosion effect by dispersing to the upper storey of the seismic isolated structure.
• The results of the analysis coincide with the data in the tables created by FEMA (Federal Emergency Management Agency) for building design against terrorist attacks. (FEMA 427-426-430, etc.).
• It has been observed that the strength of structures to resist the effects of explosions is affected in different ways at different distances. The fact is that isolators do not help to resist the effects of explosions at close range, while isolators do help to resist the effects of explosions at long distances.
• In this study, it was observed that base-isolated structures, not designed for blast loads but only for seismic loads, performed better than fixed-base structures. This suggests that, in areas where isolators are not used and where the seismic risk is low, the use of isolators can be effective in reducing the effects of explosions.
• The determination of safe explosion distances is of great importance in evaluating the effects of explosions. For instance, shelter walls designed for base-isolated structures require less distance compared to walls designed for fixed-base structures. In densely populated areas where it may not be possible to widen the safety area, the use of isolated structures is seen as advantageous.

The protection of people and buildings against blast loading is very important. Thus, this study has been proposed using existing guidelines for blast load prediction and blast resistance performances of structural components, building base isolation effectiveness under the blast loading. Since the structures under the effects of the explosion are damaged locally on the basis of element, the collapses in the structural system elements can cause failure in the structure. Base isolation causes some general structural damage by distributing the loads at certain rates under the effects of explosion. In this case, more realistic approaches can be made for the damage estimation of the members by using traditional methods compared to structures without seismic isolation. Since explosions cause different effects than earthquake loads, they also affect the behaviour in structural members. The evaluation of the explosion effects according to the seismic regulations can be done by considering the ductile behaviour of the elements. However, under the effects of explosion, it may be possible for the structural members to become brittle.