7.6 Steel Structure Design

Steel structure design covers how engineers size and detail steel members to safely carry loads. Since steel is one of the most common structural materials for buildings and bridges, understanding its properties and design principles is foundational to structural engineering practice.

This section covers the mechanical behavior of structural steel, the design of tension and compression members, beam design, and the philosophy behind limit state design.

Properties and Behavior of Structural Steel

Composition and Mechanical Properties

Structural steel is an alloy of iron and carbon, with small amounts of other elements like manganese, silicon, and nickel added to improve specific properties such as strength or weldability.

The stress-strain curve for steel has two key regions. In the elastic region, the material deforms proportionally to the applied stress and returns to its original shape when unloaded. Beyond the yield point, the material enters the plastic region, where permanent deformation occurs. Steel continues to carry increasing load until it reaches its ultimate tensile strength, after which it necks and eventually fractures.

This ductile behavior is a major advantage for structural safety. A steel member will visibly deform well before it actually fails, giving warning that something is wrong.

Key properties to know:

• Young's modulus (E): approximately 200 GPa, giving steel high stiffness
• High strength-to-weight ratio: makes steel efficient for large structures like skyscrapers and long-span bridges
• Steel grades have different yield strengths and compositions for different applications:
  • A36 steel: yield strength of 250 MPa, commonly used in building frames
  • A572 Grade 50 steel: yield strength of 345 MPa, often used in bridges and heavy structures

Environmental Considerations and Material Behavior

Steel is susceptible to corrosion when exposed to moisture and oxygen. Protective measures like galvanization (zinc coating) or paint systems are standard practice.

Temperature significantly affects steel's performance:

• High temperatures reduce both strength and stiffness. Steel loses strength rapidly above 500°C, which is why fireproofing is critical. Common methods include intumescent coatings (which swell to insulate the steel when heated) and concrete encasement.
• Low temperatures can cause certain steel grades to become brittle, meaning they fracture suddenly without the usual warning deformation.

Thermal expansion must also be accounted for, especially in long-span structures. Steel's coefficient of thermal expansion is approximately $12 \times 10^{-6}$ per °C. Engineers use expansion joints or sliding bearings to accommodate this movement and prevent stress buildup.

Design of Steel Tension and Compression Members

AISC Specifications and Design Approaches

The American Institute of Steel Construction (AISC) publishes the standard specifications used for steel design in the United States. Two design philosophies are provided:

Load and Resistance Factor Design (LRFD) is the primary modern approach. It applies separate factors to loads (to account for uncertainty in loading) and to resistance (to account for uncertainty in material strength and behavior):

$\phi R_n \geq \sum \gamma_i Q_i$

where $\phi$ = resistance factor, $R_n$ = nominal strength, $\gamma_i$ = load factors, and $Q_i$ = load effects.

Allowable Strength Design (ASD) is an alternative method that divides the nominal strength by a single safety factor:

$R_n / \Omega \geq \sum Q_i$

where $\Omega$ = safety factor. Both methods are calibrated to produce similar designs, but LRFD is more widely used in current practice.

Tension Member Design

Tension members (think bracing rods, truss chords in tension, or hangers) are designed by checking two possible failure modes. The member's nominal tensile strength is the lesser of:

1. Yielding of the gross section: The entire cross-section yields. $P_n = F_y A_g$

2. Fracture of the net section: The member fractures through the reduced area at bolt holes or other openings. $P_n = F_u A_e$

where $F_y$ = yield strength, $A_g$ = gross cross-sectional area, $F_u$ = ultimate tensile strength, and $A_e$ = effective net area.

The effective net area $A_e$ accounts for shear lag, which occurs when not all elements of a cross-section are connected (for example, when only one leg of an angle is bolted). In that case, stress doesn't distribute evenly across the full section, so the effective area is reduced.

A third check, block shear rupture, must also be performed at connections. This failure mode involves a combination of shear tearing along one plane and tension fracture along another.

Compression Member Design

Compression members (columns, struts, truss chords in compression) are governed by buckling rather than simple material yielding. A slender column can buckle at a load well below what the material itself could handle in pure compression.

The key parameter is the slenderness ratio:

$KL/r$

• $K$ = effective length factor (accounts for end conditions: pinned, fixed, etc.)
• $L$ = unbraced length of the member
• $r$ = radius of gyration of the cross-section

A higher slenderness ratio means the member is more prone to buckling. The AISC specification provides two equations for nominal compressive strength, depending on whether the member is stocky or slender:

• For $KL/r \leq 4.71\sqrt{E/F_y}$ (inelastic buckling controls): $P_n = [0.658^{(F_y/F_e)}] F_y A_g$

• For $KL/r > 4.71\sqrt{E/F_y}$ (elastic buckling controls): $P_n = 0.877 F_e A_g$

where $F_e$ = Euler elastic buckling stress.

Beyond overall member buckling, engineers must also check for local buckling of individual elements (flanges and webs), which can occur before the full member buckles.

Analysis and Design of Steel Beams

Flexural Design and Buckling Considerations

The goal of flexural design is to ensure a beam can resist the bending moments caused by applied loads. For a compact steel section (one where local buckling won't occur before the full cross-section yields), the maximum moment capacity is the plastic moment:

$M_p = F_y Z_x$

where $Z_x$ = plastic section modulus. This represents the condition where the entire cross-section has yielded in either tension or compression.

However, beams that are not adequately braced along their compression flange can fail by lateral-torsional buckling (LTB). This is a combined sideways and twisting instability that reduces the beam's moment capacity below $M_p$. The unbraced length of the compression flange is the critical variable, and AISC provides equations to calculate the reduced capacity when LTB governs.

Sections are classified as compact, non-compact, or slender based on the width-to-thickness ratios of their flanges and webs. Each classification has different design equations reflecting the likelihood of local buckling.

Shear Design and Composite Action

Shear in a steel beam is carried primarily by the web. The nominal shear strength is:

$V_n = 0.6 F_y A_w C_v$

where $A_w$ = area of the web (depth × web thickness) and $C_v$ = web shear coefficient (equals 1.0 for most rolled W-shapes, but can be less for thin-webbed plate girders).

Additional checks at support points and concentrated load locations include web yielding and web crippling, which are localized failure modes.

In regions where both high shear and high moment occur simultaneously, moment-shear interaction may need to be checked.

Composite action between steel beams and concrete floor slabs is common in building construction. By welding shear studs to the top flange of the beam, the concrete slab and steel beam act together as a single unit. This significantly increases both stiffness and moment capacity, often allowing the use of lighter steel sections. Designers can specify either full or partial composite action depending on the number of shear connectors provided.

Serviceability and Connection Design

Beyond strength, beams must meet serviceability requirements so the structure performs well under everyday use:

• Deflection limits: Typical limits are L/360 for live load deflection and L/240 for total load deflection, where L is the beam span.
• Vibration control: Floor beams must have adequate natural frequency and limited acceleration to avoid perceptible bouncing, particularly in open-plan offices or pedestrian bridges.

Connection design is critical because connections transfer forces between members. The two main categories are:

• Moment connections transfer both shear and bending moment. These are typically welded, bolted with extended end plates, or a combination. They create rigid frame behavior.
• Shear connections (also called simple connections) transfer only vertical shear. Examples include single-angle, double-angle, and shear tab connections. These allow the beam end to rotate freely.

Connection design must account for the forces involved, the geometry of the connected members, and practical fabrication and erection considerations.

Limit State Design vs Load and Resistance Factor Design

Principles of Limit State Design

Limit state design is the overarching philosophy: a structure is designed so that it does not reach any condition (a "limit state") where it becomes unfit for its intended purpose. There are two categories:

Ultimate limit states relate to safety and maximum load-carrying capacity:

• Strength failure (yielding, buckling, fracture)
• Stability failure (overturning, sliding)
• Fatigue failure under repeated cyclic loading

Serviceability limit states relate to the structure's performance under normal use:

• Excessive deflections
• Objectionable vibrations
• Cracking in concrete elements (relevant for composite construction)
• Durability issues such as corrosion

Load and Resistance Factor Design (LRFD) Methodology

LRFD is a specific implementation of limit state design. It applies separate factors to loads and resistances to account for the different sources of uncertainty in each.

Load combinations represent realistic scenarios of simultaneous loading. For example:

$1.2D + 1.6L + 0.5S$

where $D$ = dead load, $L$ = live load, and $S$ = snow load. The higher factor on live load (1.6) reflects the greater uncertainty in predicting live loads compared to dead loads (1.2).

Resistance factors ($\phi$) are specific to each failure mode:

• Tension yielding: $\phi = 0.90$
• Compression members: $\phi = 0.90$
• Flexure in beams: $\phi = 0.90$

The general design check remains:

$\phi R_n \geq \sum \gamma_i Q_i$

These factors were calibrated using probability-based methods to achieve a consistent reliability index ($\beta$), typically between 3.0 and 4.0. This index represents the statistical margin of safety, accounting for uncertainties in loads, material properties, and analysis methods. A $\beta$ of 3.0 corresponds roughly to a probability of failure of about 1 in 1,000 over the structure's lifetime.