2.3 Earthquake Resistant Steel Building Design

The experiences gleaned from past strong earthquakes prove that the initial conceptual design of a building is extremely important to the behaviour of the building during an earthquake. It was shown repeatedly that no static analysis can assure a good dissipation of energy and favourable distribution of damage in irregular buildings, e.g., structures with large asymmetry or distinctively soft storeys. The responsibility for a good initial conceptual design lies with the architect, as well as with the structural engineer providing numerical proof of the structure's safety.

Picture 2.6 - Earthquake Distorted Steel Frame

2.3.1 General Design Considerations in Earthquake Areas

Some general principles for the design of structures to be erected in earthquake areas are given here. It should be pointed out that earthquake resistant structures can be designed without consideration to these principles. Compliance with these principles will, however, substantially reduce the possibility of dynamic effect occurrences, which cannot be predicted by linear analysis. The high cost of an earthquake resistant structure is reduced by use of these lower values in comparison with a non-earthquake resistant structure. It also seems that the combination of good design and simple analysis results in safer structures than the combination of bad design and refined analysis.

First,, the seismic activity of the construction site must be determined. According to Eurocode 8 and other regulations, the elastic response spectrum to be used for design purposes depends on several parameters such as the seismic zone, the type of seismic action, the local soil conditions and the viscous damping ratio of the structure.

The seismic zone can be characterized by means of the severity of the seismic action. This characterization is accomplished by normalizing the response spectra to a certain level of peak ground acceleration. Usually, the response spectrum for vertical motion is defined as a percentage of the response spectrum for the two orthogonal horizontal directions. In Eurocode 8 the suggested percentage is 70%.

The maximum acceleration to be used in each region in Europe is defined according to microzonation studies (The identification of separate individual areas having different potentials for hazardous earthquake effects) for each zone, depending on the local seismic hazard parameters.

In addition, the dynamic behaviour of a simple structure is easy to understand and to compute. The risk is low in forgetting any special aspect of performance such as an interaction of parts with different rigidity. Overall simplicity leads to simple detailing.

Furthermore, any discontinuity in the design brings a stress concentration and, potentially, a local failure mechanism. Energy dissipation in the structure should be as high as possible. There should, therefore, be many dissipative zones in the structure. As a result, the objective should be a global failure mechanism. The non-homogeneous behaviour of a structure with major discontinuity leads to tedious calculations and difficult design of the connection areas.

Continuity and uniform distribution of strength in the horizontal direction of a building generally means symmetry, if possible almost axisymmetry. Plan layout of vertical resisting elements also should recognise the need for a high global torsional stiffness.

Sometimes tall buildings can be constructed as an architectural choice, addressing environmental concerns and the consumer needs. But one must be careful in considering the coping width in the slenderness affect of building height due to the overturning affect of an earthquake. High slenderness may, however, be useful in some cases. In general, the more slender a structure, the worse the overturning effect of an earthquake.

Additionally, earthquake action generates special torsional effects in structures, mainly because the results of inertia forces generated by the earthquake is applied at the mass center "M" of each floor of the structure. The latter generally does not coincide with the torsion center "S" of the earthquake resistant structure. The resultant force times the distance to that center gives a torsional moment "Mt". In multi-storey frames, the torsional moment from one particular floor is increased by the resulting moment of the floors above. In most structures, the approach to evaluate this torsional moment is partly rational (the distance between S and M) and partly statistical, because the load distribution in a structure is not well known at the design stage and changes through the life of the structure. Codes indicate how to evaluate this second term. There may also be a second cause for torsional action. The earthquake itself results mainly in the vertical propagation of a shear wave so that two points of the structure may be moving differently at one time. This torsion origin is normally important for structures which are very large in plan, e.g., bridges. To resist the torsional action, the structure must be given adequate torsional rigidity. The best solution is obtained by putting the earthquake resistant part of the structure close to the perimeter of the structure as a whole and all around it, complying with the symmetry principle. It must be pointed out that the classical "one vertical core" structure of earthquake-free areas is not effective because it lacks torsional rigidity. It should simply be avoided in asymmetrical layouts.

Moreover, building diaphragms transfer horizontal inertia forces, resulting from the motion applied to the floor masses and their loading. Diaphragms must be structures of low deformability and capable of efficiently distributing the horizontal action between the various vertical resistant structures. Diaphragms may be provided in many ways, for example, concrete slabs, composite slabs, trusses or frames. Diaphragms must be properly linked to the vertical rigidity elements. The links must be able to transmit the horizontal inertia force.

Important loads should not be placed where they generate inertia forces under earthquake loading. For example, a library should, by preference, be at ground level. An X-Ray installation should be close to the center of rotation. Masses should be reduced whenever possible. For instance, using light floor systems rather than traditional slabs can bring drastic reductions in inertia forces and result in substantial economy in the framework. Similar choices should be made for partition walls, infills, claddings, etc.

Finally, the shape of the design response spectrum indicates that earthquake forces are lower for structures characterized by a predominant high period (T) of vibration. This characteristic can sometimes be used at the start of a design, especially when more refined data are available for a particular site. For instance, in a site with thick alluvium layers, which is characterized by a response spectrum with relatively high amplitudes in the high-period range and low amplitudes in the low-period range, a very rigid structure would fit better than a flexible one. The opposite choice would apply to rock areas.

2.3.2 Structural Design Systems for Steel Buildings

Nowadays, steel rolled products, such as H or tube sections, are available in a wide variety of types and dimensions, larger than in the past. They may be used to produce a wide range of structural elements and connections.

 

Picture 2.7 - Typical Steel Sections

 

Steel elements have the advantage of easily constructed large elements. They may be considered the most appropriate building material to ensure large structure seismic resistance. The performance and ductility of the structural elements and connections may be affected by many factors.

Structural steel elements often have high slenderness and slender cross-sections due to high strength. Consequently, various types of buckling may occur, such as flexural buckling of the whole element, lateral-torsional buckling and local buckling of plate elements.

In general, structures through inelastic lateral load dissipation capacity can withstand considerable deformations without failure by dissipating large amounts of seismic energy. There are essentially two structural systems used to resist horizontal seismic actions:

1. Braced structural frames (BSF) or truss bracing.
2. Moment resisting frames (MRF) or simple frames.

Additionally, diaphragms and shear walls are considered with these structure systems.

Diaphragms are horizontal resistance elements, generally floors and roofs, that transfer the lateral forces between the vertical resistance elements (shear walls or frames). Basically, a diaphragm acts as a horizontal I-beam. That is, the diaphragm itself acts as the web of the beam and its edges act as flanges.

Shear walls are vertical walls that are designed to receive lateral forces from diaphragms and transmit them to the ground. The forces in these walls are predominantly shear forces in which the fibers within the wall try to slide past one another.

When you build a house of cards, you design a shear wall structure, and you soon learn that sufficient card "walls" must be placed at right angles to one another or the house will collapse.

If you were to connect your walls together with tape, it is easy to see that the strength of this house of cards would be immediately become greatly increased. This illustrates a very important point: In general, the earthquake resistance of any building is highly dependent upon the connections joining the building's larger structural members, such as walls, beams, columns and floor-slabs.

Shear walls, in particular, must be strong in themselves and also strongly connected to each other and to the horizontal diaphragms. In a simple building with shear walls at each end, ground motion enters the building and creates inertial forces that move the floor diaphragms. This movement is resisted by the shear walls and the forces are transmitted back down to the foundation

2.3.2.1 Braced Structural Frames

Braced frames develop their resistance to lateral forces by the bracing action of diagonal members. The braces induce forces in the associated beams and columns so that all work together like a truss, with all members subjected to stresses that are primarily axial.

A Concentrically Braced Frame (CBF) has minor eccentricities in the joints of the frame that are accounted for in the design.

 

Picture 2.8 - 

 

An Eccentrically Braced Frame (EBF) has elements that are strictly controlled to combine a stiffening effect due to the diagonal braces with yielding in the link beams. Eccentrically braced frames are present only in conforming buildings.

Steel Concentrically Braced Frames (CBF) are strong, stiff and ductile, and are therefore ideal for seismic framing systems. The quality of the seismic response of CBFs is determined by the performance of the brace. In order to achieve the best performance from a CBF, the brace must fail before any other component of the frame does. Under current design codes, these connections can become oversized to meet requirements, decreasing the yield length of the brace, causing the CBFs to fail earlier than expected.

EBFs address the desire for a laterally stiff framing system with significant energy dissipation capability to accommodate large seismic forces. A typical EBF consists of a beam, one or two braces and columns. Its configuration is similar to traditional braced frames with the exception that at least one end of each brace must be eccentrically connected to the frame.

 

Picture 2.9 - Typical Eccentrically Braced Frames

 

The eccentric connection introduces bending and shear forces in the beam adjacent to the brace. The short segment of the frame where these forces are concentrated is the link. EBF lateral stiffness is primarily a function of the ratio of the link length to the beam length. As the link becomes shorter, the frame becomes stiffer, approaching the stiffness of a concentric braced frame. As the link becomes longer, the frame becomes more flexible, approaching the stiffness of a moment frame.

2.3.2.2 Moment Resisting Frames

When seismic resistance is provided by moment resistant frames, lateral forces are resisted primarily by the joints between columns and beams. These joints become highly stressed and the details of their construction are very important. Moment frames use, as a last-resort resistance strategy, the energy absorption obtained by permanent deformation of the structure prior to ultimate failure. For this reason, moment resistant frames generally are steel structures with bolts or welded joints in which the natural ductility of the material advantageous. However, properly reinforced concrete frames that contain a large amount of steel reinforcing are also effective as ductile frames. They will distort and retain resistance capacity prior to failure and will not fail in a brittle manner.

Architecturally, moment resistant frames offer a certain advantage over shear walls or braced frames because they tend to provide structures that are much more unobstructed internally than shear wall structures, which facilitates the design of accompanying architectural elements such as exterior walls, partitions, and ceilings and the placement of building contents such as furniture and loose equipment. Nevertheless, moment resistant frames require special construction and detailing and, therefore, are more expensive than shear walls or braced frames.

 

Picture 2.10 - Typical Moment Resisting Frames

 

2.3.3 Connections

There are many types and varieties of connections, and each has different rotational characteristics that affect the frame behaviour. Butt welding, fillet welding, bolting, and riveting may be employed for aseismic connections, either individually or in combination. As fully bolted or riveted connections tend to be large and expensive, fully welded connections or a combination of welding and bolting are the most frequently used. Bolts have the advantage of providing more damping to frames than welds.

Connections should be designed to make fabrication and erection of the framework as simple and rapid as possible.

Conclusive design criteria for beam-to-column joints are not yet available for seismic conditions. Until the recent past, relatively few cyclic load tests had been performed on joints commonly used in Europe. At present many experimental investigations are in progress in different European laboratories. They deal with cyclic behaviour of rigid and semi-rigid joints, both for bare steel and composite constructions.

The connection types are grouped into two main categories

1. Welded connections
2. Bolted connections

2.3.3.1 Welded Connections

Structural welding is a process by which the parts that are to be connected are heated and fused, with supplementary molten metal at the joint. A relatively small depth of material will become molten, and upon cooling, the structural steel and weld metal will act as one continuous part where they are joined. The additional metal is deposited from a special electrode, which is part of the electric circuit that includes the connected part.

For considering the seismic behaviour of welded connections, two major criteria have been taken into account:

If stiffeners are added to the parts of the connection which are most responsible for its flexibility, the amount of energy absorption decreases but the load level is increased.

If elements are added to a joint which do not substantially modify the deformation mechanism but increase the local strength of the structural elements, then there will be an increase of energy absorption and load level provided that the collapse is ductile.

Picture 2.11 - Typical Welded Moment Resisting Connections

For this type of connection the plastic rotation of the beam is mainly developed by the extension of plastic deformation near the connection. Generally, in order to control the extension of the plastic region in the element in the vicinity of the connection, the beam-to-column connection must have an ultimate bending moment which is greater than the full bending resistance of the attached element. For this reason Eurocode 8 requires that the resistance of the connection be greater than the resistance of the adjacent connected element:

R_d ³ 1,20 R_fy

Where R_d is the resistance of the connection according to Eurocode 3 and R_fy is the yielding resistance of the connected part. Connections made by means of butt-welds or full penetration groove welds are deemed to satisfy this overstrength criterion

2.3.3.2 Bolted Connections

There are different types of bolted connections. They can be categorized based on the type of loading.

• Tension member connection and splice. It subjects the bolts to forces that tend to shear the shank.
• Beam end simple connection. It subjects the bolts to forces that tend to shear the shank.
• Hanger connection. The hanger connection puts the bolts in tension

The bolts are subjected to shear or tension loading.

• In most bolted connection, the bolts are subjected to shear.
• Bolts can fail in shear or in tension.
• You can calculate the shear strength or the tensile strength of a bolt

Simple connection: If the line of action of the force acting on the connection passes through the center of gravity of the connection, then each bolt can be assumed to resist an equal share of the load.

 

Picture 2.12 - Typical Bolted Connections

 

The strength of the simple connection will be equal to the sum of the strengths of the individual bolts in the connection.

2.3.4 Beams

One of the important factors that influence the overall performance of a steel framed structure is the bending capacity of the beams and girders. In steel, failure is usually by instability and in order to perform satisfactorily in a major earthquake, the moment rotation curve must be developing a long plateau. Ultimate failure usually is not caused by lateral buckling but is triggered by local buckling.

 

Picture 2.13 - Typical Bolted Connections

 

Experimental investigations carried out on cantilever beams under repeated and reversing loading has shown that the development of local buckling in the flanges does not signal an immediate loss of the moment resistance. The beams are able to sustain loads substantially higher than those that cause initial flange buckling. This behaviour is attributed to the considerable post-buckling strength of the plate elements. However, after the occurrence of the maximum load in the subsequent load cycles, the moment resistance deteriorates. This deterioration is higher with increasing width-thickness ratio (b/t) of the flanges as a consequence of the early occurrence of local instability in the flange elements.

The severe distortions of the flanges tend to induce torsional displacement of the section. They are associated with a lower load that would develop in pure flexural buckling.

2.3.5 Columns Under Bending and Compression

Columns are the last basic units of construction in steel frame structure that must be considered. Over the years, much work has been done on column research, but much of this has been done with axial loadings with little or no eccentricity to determine basic column formulas. It is only recently that extensive work has been performed on beam columns where large bending moments are applied at the same time as the axial loads. Much of early work has been done for the purpose of developing criteria for plastic design in steel but does not extend into the ranges of load combinations customary in providing for earthquake resistance in major steel construction. In high rise buildings the general range of usable column loadings is in a fairly narrow range, and many of the tests are too far out of this range to provide much usable information.

Local buckling and torsion effects reduce the strength of the column. In sections where local buckling occurs, the shape and value of the moment rotation curve remain the same as in compact sections up to the point where the flange buckles locally

• << Prev
• Next >>