contact | structural element calculators | paperback and pdf | « previous section | next section » | contents

Contents |

Introduction to structural design | Statics | Tributary areas | Equilibrium | Reactions | Internal forces | Indeterminate structure | Material properties |

The magnitude of internal forces and bending moments do not, by themselves, give any indication as to whether a particular structural element is safe or unsafe. Instead, the load or moment that an element can safely resist can only be determined when information about the element's cross section and material properties is considered: clearly, a large cross section is stronger than a small one. But "large" in what way? The cross-sectional properties relevant to the determination of structural safety and serviceability are different for tension elements, columns, and beams and are, therefore, discussed more fully in their appropriate context. What follows is a brief overview and summary of the major cross-section properties encountered in structural analysis and design, followed by a discussion of tension, compression, and bending.

Cross-sectional areas are easily determined: for rectangles, the area *A* = *B* × *H* (Figure 1.49*a*) and for circles, *A* = π*R*^{2} (Figure 1.49*c*). What may not be as immediately clear is that the -shaped cross section (Figure 1.49*b*) has an area, *A* = (*B* × *H*) – (*b* × *h*), and the circular ring (Figure 1.49*d*) has an area, *A* = π*R*^{2} – π*r*^{2}, where *R* is the outer and *r* is the inner radius.

Figure 1.49: Cross sections typically encountered as structural elements include *(a)* rectangles, *(b)* -shaped sections, *(c)* circles, and *(d)* circular rings

The moment of inertia, *I*_{x}, is defined as the sum of all elemental areas above or below the centroid (*x*-axis) of the cross section multiplied by the square of the distance from each of the individual elemental
centroids to the centroid of the cross section as a whole, or

(1.7)

where *y* is the distance from each elemental area (the elemental areas being *dA* = width × *dy*) to the centroid of the cross section, while *H*/2 and –*H*/2 represent the limits over which the integral is taken for the rectangle and -shaped section shown in Figure 1.49 (the same equation holds for the circular sections as well, except with the integral taken from *R*/2 to –*R*/2).
This property is useful in understanding the stiffness of a cross section when bent. It can be seen
that placing a good deal of the cross-sectional material away from the centroid — as in the -shaped section or, to a lesser extent, in the circular ring — increases the moment of inertia, and therefore the stiffness, since more "area" is multiplied by the square of a greater distance from the centroidal axis. Equation 1.7 can be solved as follows for rectangular and circular shapes:

(1.8)

Moments of inertia for the -shaped section and circular ring can be easily found by subtracting the smaller rectangle (or circle) from the larger one: for the -shaped section, *I*_{x} = *BH*^{3}/12 – *bh*^{3}/12; for the circular ring, *I*_{x} = π*R*^{4}/4 – π*r*^{4}/4.

For moments of inertia taken about the *y*-axis, the equations for rectangles and circles are easily
modified:

(1.9)

Moments of inertia for circular rings about the *y*-axis (Figure 1.49*d*) are determined as before: by subtracting the moment of inertia of the smaller from that of the larger circle. For moments of inertia of -shaped sections about the *y*-axis, however, it is not possible to simply subtract the smaller rectangles from the larger, as was done when computing the moment of inertia about the *x*-axis, since the centroids of the various parts being subtracted do not coincide.

Figure 1.50: Dimensions of an -shaped section oriented about its *y*-axis, for the calculation of moment of inertia

Instead, one must add the three moments of inertia of the two flanges and web, as shown in Figure 1.50, each taken about the *y*-axis:

(1.10)

The elastic section modulus, *S _{x}*, is a single parameter that measures a cross section's strength in bending. For symmetrical sections, such as those shown in Figures 1.49

(1.11)

For the circular shapes, *S _{x}* =

The plastic section modulus, *Z _{x}*, is used to determine the limit state of steel beams, defined as the point when the entire cross section has yielded. This property is unique to steel, since neither of the other materials we are considering (wood and reinforced concrete) has the necessary ductility to reach this state. Unlike the elastic section modulus,

Figure 1.51: Rectangular cross section showing shaded area and distance from centroid of shaded area to centroid of
the whole cross section, for calculation of plastic section modulus, *Z*_{x}

The radius of gyration of a cross section, *r* or ρ, is a distance — but one without any obvious physical meaning. It measures the cross section's resistance to buckling, when compressed, and is defined as follows:

(1.12)

where *I _{x}* is the moment of inertia about the

(1.13)

From Equation 1.9, the moment of inertia about the *y*-axis used to compute the minimum radius of
gyration for a rectangular cross section is *I _{y}* =

Elements subjected to tension provide us with the simplest mathematical model relating internal force and stress:

(1.14)

This equation is simple and straightforward because it corresponds to the simplest pattern of strain that can develop within the cross section of a structural element. As shown in Figure 1.52, this strain is assumed to be uniformly distributed across the entire cross section; for this reason, the stress can be defined as force per unit area. Classical "strength of materials" texts use the symbol, σ, for axial stress, so that we get:

(1.15)

where *P* is the internal force at a cross section with area, *A*. By axial stress, we mean stress "acting" parallel to the longitudinal axis of the structural element, or stress causing the element to strain in the direction of its longitudinal axis. Tension is an axial stress causing elongation; compression is an axial stress causing shortening or contraction.

Figure 1.52: Illustrations of *(a)* tension element; and *(b)* free-body diagram cut at any cross-section with area, A

Where bolt holes reduce the cross-sectional area of a tension element, the remaining area at the
cross section, *A _{n}*, is called the net area. Failure or "rupture" of an element stressed in tension occurs
at a failure surface defined by the location and quantity of such bolt holes. Where the holes are arrayed
in an orthogonal grid, as shown in Figure 1.53

Figure 1.53: Net area of a cross section, shown in black, with *(a)* one possible failure surface when bolt holes are arrayed in orthogonal grid; and *(b)* multiple possible failure surfaces when bolt holes are staggered

When we discuss particular structural materials, stresses are often represented by the letter *F*
rather than σ, and capitalized when referring to allowable, yield or ultimate stresses in timber and
steel. For example, *F _{y}* refers to the yield stress of steel;

(1.16)

The elongation of an element in tension can be computed based on the definition of modulus of
elasticity given in Equation 1.6; since *E* = σ/ε, and substituting *P**/**A* for σ and (elongation)/(original length) for ε, we get:

(1.17)

Solving for elongation, and letting *L* = original length, the equation becomes:

(1.18)

** Problem definition.** Compute the elongation, or change in length, for a steel bar with a cross-sectional area of 4 in

** Solution overview.** Find elongation = (

*Problem solution*

From Equation 1.18, elongation = (*PL*)/(*AE*) = (10 kips × 36 in.)/(4 in^{2} × 29,000 ksi) = 0.0031 in.

Columns are vertical elements subjected to compressive stress; nothing, however, prevents us from applying the same design and analysis methods to any compressive element, whether vertical, horizontal or inclined. Only axially-loaded compression elements (with no bending moments present) will be considered here.

Compression is similar to tension, since both types of structural action result in a uniform distribution of axial stress over a cross section taken through the element. But allowable stress in compression is often limited by the phenomenon of buckling, in which the element deforms out of its axial alignment at a stress that may be significantly lower than the stress causing compressive crushing.

To understand why an axially-loaded column will buckle rather than simply compress, consider the case of an eccentrically-loaded column, as shown in Figure 1.54. Unlike a beam whose internal bending moments are not influenced by load-induced deflections (Figure 1.55), the eccentrically-loaded column will deflect more than might be expected
if only the initial moment, *M*_{1}, is considered, since the "initial deflections" increase bending moments throughout the column, in turn causing further deflection, as shown in Figure 1.54*b*. What the mathematician Leonard Euler (1707–1783) figured out was that these deflections increase rapidly in the vicinity of a particular ("critical") load, at which point the column is assumed to fail, and that the value of this load is independent of the initial eccentricity. In other words, even with the smallest imaginable deviation from axiality, a column is assumed to buckle at some critical load. Since no perfectly axial columns (or loads) can exist, all columns behaving elastically are assumed to buckle at the critical buckling stress derived by Euler:

(1.19)

where

*E* = the modulus of elasticity

*K* = a coefficient that depends on the column's end constraints (see Appendix Table A-1.2)

*L* = the unbraced length of the column

*r* = the radius of gyration with respect to the unbraced length (sometimes given the symbol, ρ), equal to ,

where *I* is the least moment of inertia and *A* is area of the cross section

Figure 1.54: Increase of bending moment in a column due to load-induced deflection

Figure 1.55: No increase of bending moment in a beam due to load-induced deflection

For the typical case in which the unbraced length is the same for both axes of the column, *r* (or *I*)
is taken as the smaller of the two possible values, i.e., *r _{min}* (or

The strength of wood and steel columns is limited in two ways: either they will crush at their maximum compressive stress, or buckle at some critical stress that is different from, and independent of, their strength in compression. Euler's equation for critical buckling stress works well for slender columns, but gives increasingly inaccurate results as the slenderness of columns decreases and the effects of crushing begin to interact with the idealized conditions from which Euler's equation was derived. Figure 1.56 shows schematically the relationship between Euler critical buckling stress, crushing strength, and test results for columns with different slenderness ratios. It can be seen that only for slender columns can the Euler curve be used as a basis for design.

Figure 1.56: Schematic relationship between critical stress and column slenderness

Like all structural elements, beams are both stressed and subject to deformations when loaded. Both of these considerations must be accounted for in the design of beams.

* Deflection.* While the elongation or contraction of axially-loaded members along their longitudinal axes is usually of little consequence, beams may experience excessive deflection perpendicular
to their longitudinal axes, making them unserviceable. Limits on deflection are based on several
considerations, including minimizing vibrations, thereby improving occupant comfort; preventing
cracking of ceiling materials, partitions, or cladding supported by the beams; and promoting positive
drainage (for roof beams) in order to avoid ponding of water at midspan. These limits are generally
expressed as a fraction of the span,

(1.20)

where *w* = distributed load (lb/in. or kips/in.); *L* = span (in.); *E* = modulus of elasticity (psi or ksi); and *I* = moment of inertia (in^{4}). When using Equation 1.20 with *L* in feet, w in lb/ft or kips/ft, *E* in psi or ksi (compatible with load, *w*), and *I* in in^{4}, as is most commonly done, multiply the expression by 12^{3}
to make the units consistent.

* Bending stress.* Beams are stressed when they bend because the action of bending
causes an elongation on one side, resulting in tension, and a shortening on the other side, resulting in compression. By exaggerating the curvature of the beam as it bends, this elongation and shortening can be visualized. Exactly
where the tension and compression are depends on how the beam is loaded and how it is supported.

For *simply-supported* beams with downward-acting loads (i.e., with gravity loads), the beam is stretched on the bottom (tension) and shortened on the top (compression) as shown in Figure 1.57.

Figure 1.57: Behavior of a simply supported beam

For *cantilevered* beams fixed at one end, with downward-acting loads, the beam is stretched on the top and shortened on the bottom (Figure 1.58).

Figure 1.58: Behavior of a cantilevered beam

For *continuous* beams spanning over several supports, the changing curvature causes the position of tension and compression zones to reverse a number of times over the length of the beam, as illustrated in Figure 1.59.

Figure 1.59: Behavior of a continuous beam

The relative position of tension and compression within the beam's cross section is directly related
to the sign of the bending moment at that cross section. As can be seen from Figure 1.60*a*, a
counterclockwise moment on the right side of a free-body diagram is equivalent to a distribution
of bending stress with compression on the top and tension on the bottom of the beam: so-called
"positive" bending (and "positive" bending moment). Figure 1.60*b* shows a free-body diagram cut
through a cantilever beam with "negative" bending — i.e., tension on the top and compression on
the bottom corresponding to a clockwise moment as shown. The reversing curvature of a continuous
beam, such as that shown in Figure 1.59, corresponds precisely to a reversal in the sign of the
bending moment.

Figure 1.60: Comparison of "positive" and "negative" bending in *(a)* a simply-supported beam, and *(b)* a cantilevered beam

As shown in Figure 1.61, points of inflection (points where the curvature changes) always occur at points of zero moment.

Figure 1.61: Continuous beam showing correspondence of points of inflection (change from positive to negative curvature)
and points of zero moment

Bending stresses within these beams can be computed if we assume that the stretching and shortening that take place at any cross section are linear; i.e., a straight line connecting a stretched point with a shortened point on any cross-sectional cut will accurately describe the shape of the beam throughout the entire cross section (Figure 1.62).

Figure 1.62: Shortening and stretching (compression and tension) at a typical beam cross-section

Three observations can be made, once this assumption is accepted: (1) maximum elongation
and shortening occur at the top and bottom of the beam (the "extreme fibers"); (2) a surface exists
somewhere between the extreme fibers that is neither elongated nor shortened — this "plane"
is called the "neutral axis" or "neutral surface"; and (3) strain can be defined as the elongation or
shortening of any portion of the beam, divided by its original (unloaded) length. Since the original
length is a constant, a strain diagram has the same shape as an "elongation-shortening diagram." For
materials with linear stress-strain relationships (where stress equals strain times a constant *modulus
of elasticity*), a stress diagram will also have the same shape as the strain or "elongation-shortening
diagram." Figure 1.63 compares these diagram shapes for materials with linear stress-strain relationships.

Figure 1.63: Elongation, strain, and stress diagrams for a linear, elastic material

For materials with nonlinear stress-strain relationships, a stress diagram can be pieced together
by plotting points from a stress-strain curve for the material. Thus, a steel beam stressed beyond its
elastic region would have stress and strain distributions as shown in Figure 1.64. The elongation and
shortening, shown in Figure 1.64*a*, and therefore the strain, shown in Figure 1.64*b*, is assumed to
remain linear even when the stress, shown in Figure 1.64*d* through Figure 1.64*f*, becomes nonlinear.
In Figure 1.64*c*, the stress at the extreme fibers of the cross section just reaches the limit of elastic
behavior (with stress, σ* _{y}*) which corresponds to the so-called elastic moment,

Figure 1.64: Elongation, strain, and stress diagrams for an elastic-plastic material such as steel showing *(a)* elongation and shortening of the actual material; *(b)* strain diagrams; *(c)* stress diagram at the point where the outer fiber has just yielded; *(c)* stress diagram corresponding to strain just beyond the elastic limit; *(e)* stress diagram corresponding to continued strain beyond the elastic limit; *(f)* stress diagram corresponding to the plastic moment (where the entire cross-section has yielded); and *(g)* stress-strain diagram

The shape of the stress diagram is a key element in determining the magnitudes of stresses within the beam: when combined with the cross-sectional shape, the requirements of equilibrium can be used to find the magnitudes of the stresses. Typical stress diagrams are shown in Figure 1.65 corresponding to the allowable moment for wood and the limit states for steel and reinforced concrete.

*Allowable stress design.* As an example of how the stress-moment relationship is computed using the
allowable stress design method, consider a free-body diagram cut from a rectangular cross section
of width, *b*, and height, *h* (assuming a linear stress-strain relationship resulting in a linear stress diagram), as shown in Figure 1.65*a*. From the requirements of horizontal equilibrium, the total compressive
force, *C*, must equal the total tension force, *T*. For this to occur, the neutral axis must be at
the center of the beam, and the maximum compressive stress must equal the maximum tension stress. Any other linear distribution of stresses would be inconsistent with these requirements of equilibrium. The couple of equal and opposite forces represented by *C* and *T*, multiplied by the moment arm between them, must equal the bending moment, *M*, caused by the loads acting on the beam. The basic bending stress equation derives from this simple fact: *M* equals *C* (or *T*) times the moment arm, τ; that is:

(1.21)

Accounting for beam width, *b*, *C* = (1/2)(*F _{b}*)(

(1.22)

Defining "*bh*^{2}/6" as the *section modulus*, *S*, and "*bh*^{3}/12" as the moment of inertia, *I*, for a rectangular cross section, and solving for the maximum allowable stress, *F _{b}*, we get the basic bending stress equations for allowable stress design:

(1.23)

where *F _{b}* is the allowable bending stress for the material (psi or ksi),

*Stress-moment relationships.* Equation 1.23 shows the relationship between bending stress, bending
moment, and section modulus for a material stressed within its linear-elastic range. It is the basis for
wood beam design. Steel and reinforced concrete, however, are no longer designed on the basis of
assumed linear-elastic behavior. Even so, the basic relationship between moment, stress, and some
sort of section modulus property remains essentially the same for all three materials, as can be seen
by comparing the stress and resultant force diagrams shown in Figure 1.65.

Figure 1.65: Bending stresses acting on rectangular cross sections corresponding to the *(a)* allowable moment for
wood; and the limit states (maximum moment at failure) for *(b)* steel; and *(c)* reinforced concrete

While specific derivations will be covered in the chapters that follow, the requirements of horizontal equilibrium (*C* = *T*)
and rotational equilibrium (*M* = *C* × τ = *T* × τ) lead to design equations with essentially the same form for all three materials: Equation 1.24 (solving for the required section modulus, *S*, in Equation 1.23) applies to allowable stress design in wood; Equation 1.25 to allowable strength design in steel; and
Equation 1.26 to strength design in reinforced concrete.

(1.24)

(1.25)

(1.26)

In each case, the section modulus term (*S*, *Z*, or *bd*^{2}) must be greater or equal to the bending moment divided by a bending stress term. The stress terms in Equations 1.24, 1.25, and 1.26 vary:
for wood, an adjusted allowable stress, *F _{b}'*, is used directly; for steel, the yield stress,

The triangular stress distribution in allowable stress design for wood corresponds to the elastic
section modulus, *S* = *bd*^{2}/6, defined in Equation 1.22 for rectangular cross sections. For steel, the
plastic section modulus, *Z*, is used, and is equal to *bd*^{2}/4 for a rectangular section — this is easily derived from the equilibrium of stresses shown in Figure 1.65*b*, although it should be noted that
rectangular solid shapes are virtually nonexistent in steel beams. The term, *bd*^{2}, used in reinforced
concrete Equation 1.26, has no official status as a "section modulus," yet it consists of the same basic
variables and has the same units as wood's *S* and steel's *Z*.

*Bending design methods.* Equations 1.24, 1.25, and 1.26 are "design" equations, since they provide
guidance for the size and shape of bending elements that are capable of resisting a given bending
moment. In practice, after bending moments are determined (for example, by the construction of
load, shear, and moment diagrams; from moment value tables; or with the use of structural analysis
software), the required section modulus term is calculated, and a cross section is then selected. In
the case of wood and steel, tables of standard cross sections and their corresponding section moduli
facilitate the direct selection of appropriate shapes. The design of a reinforced concrete beam is less
direct, since the ratio of steel to concrete may vary, producing a range of acceptable *bd*^{2} terms, each
of which may sponsor a range of choices for cross-sectional dimensions *b* and *d*.

* Shear stress.* Internal forces perpendicular to the longitudinal axis of beams may also exist along with bending moments at any cross section, consistent with the requirements of equilibrium (see, for example, the force

(1.27)

where τ = shear stress at a distance, *y*, from the extreme fiber (psi or ksi); *V* is the total shear force at the cross section (lb or kips); *Q* is the "static moment" of the partial cross-sectional area (from
the extreme fiber to the distance, *y*) about the neutral axis of the cross section (in^{3}); *I* is the moment of inertia of the cross section (in^{3}); and *b* is the width of the cross section at a distance, *y*, from the extreme fiber (in.).

*Rectangular sections.* For rectangular cross sections, the maximum shear stress, which occurs at the
neutral axis, becomes:

(1.28)

where *h* is the height of the rectangular cross section; all other variables are as defined above for
Equation 1.27. Alternatively, one can solve for the required cross-sectional area, Areq = *bh* (in^{2}) as the
basis for designing or analyzing a rectangular beam for shear, corresponding to an allowable shear
stress, τ_{allow} (psi or ksi) for maximum shear force, *V* (lb or kips). In this case, one gets:

(1.29)

This is the basis for checking shear in timber beams, which are almost always rectangular (Figure 1.66). Reinforced concrete beams behave in a more complex manner, and special procedures for dealing with shear, or diagonal tension, have been developed.

Figure 1.66: Distribution of shear stress on a rectangular cross-section

In the vicinity of supports, loads are transferred by compression directly to those supports (Figure 1.67), and the maximum shear force is therefore somewhat less than the computed maximum value. In the design of wood and reinforced concrete beams, the shear force within a distance, *d*, of the face of the supports can be considered equal to the value of the shear force at that distance, *d*. For wood beams, d is the total beam height; for reinforced concrete, it represents the effective depth, measured to the centerline of the tension reinforcement.

Figure 1.67: Reduction of shear force, *V*_{max}, in the vicinity of the beam's reaction (support)

*Wide-flange sections.* For steel wide-flange sections, the maximum shear stress, also at the neutral
axis, can be found by computing the static moment, *Q*, of the partial area (above the neutral axis)
about the neutral axis and solving Equation 1.27, as shown in Figure 1.68.

Figure 1.68: Distribution of shear stress on a flanged cross-section, and calculation of maximum shear stress, τ_{max}

For steel wide-flange shapes, simplified procedures have been developed, based on the average stress on the cross section, neglecting the overhanging flange areas; that is:

(1.30)

where τ_{max} = the maximum shear stress within the cross section, *V* = the total shear force at the cross section, *d* = the cross-sectional depth, and *t _{w}* = the web thickness (see Figure 1.69).

Figure 1.69: Comparison of actual and assumed maximum shear stress, τ_{max}, for a steel wide-flange beam

© 2020 Jonathan Ochshorn; all rights reserved. This section first posted November 15, 2020; last updated November 15, 2020.