115

Chapter 7

Structural design

Introduction
Structural design is the methodical investigation of the stability, strength and rigidity of structures. The basic objective in structural analysis and design is to produce a structure capable of resisting all applied loads without failure during its intended life. The primary purpose of a structure is to transmit or support loads. If the structure is improperly designed or fabricated, or if the actual applied loads exceed the design specifications, the device will probably fail to perform its intended function, with possible serious consequences. A wellengineered structure greatly minimizes the possibility of costly failures.

Structural design process

A structural design project may be divided into three phases, i.e. planning, design and construction. Planning: This phase involves consideration of the various requirements and factors affecting the general layout and dimensions of the structure and results in the choice of one or perhaps several alternative types of structure, which offer the best general solution. The primary consideration is the function of the structure. Secondary considerations such as aesthetics, sociology, law, economics and the environment may also be taken into account. In addition there are structural and constructional requirements and limitations, which may affect the type of structure to be designed. Design: This phase involves a detailed consideration of the alternative solutions defined in the planning phase and results in the determination of the most suitable proportions, dimensions and details of the structural elements and connections for constructing each alternative structural arrangement being considered. Construction: This phase involves mobilization of personnel; procurement of materials and equipment, including their transportation to the site, and actual on-site erection. During this phase, some redesign may be required if unforeseen difficulties occur, such as unavailability of specified materials or foundation problems.

thrust, shear, bending moments and twisting moments), as well as stress intensities, strain, deflection and reactions produced by loads, changes in temperature, shrinkage, creep and other design conditions. Finally comes the proportioning and selection of materials for the members and connections to respond adequately to the effects produced by the design conditions. The criteria used to judge whether particular proportions will result in the desired behavior reflect accumulated knowledge based on field and model tests, and practical experience. Intuition and judgment are also important to this process. The traditional basis of design called elastic design is based on allowable stress intensities which are chosen in accordance with the concept that stress or strain corresponds to the yield point of the material and should not be exceeded at the most highly stressed points of the structure, the selection of failure due to fatigue, buckling or brittle fracture or by consideration of the permissible deflection of the structure. The allowable stress method has the important disadvantage in that it does not provide a uniform overload capacity for all parts and all types of structures. The newer approach of design is called the strength design in reinforced concrete literature and plastic design in steel-design literature. The anticipated service loading is first multiplied by a suitable load factor, the magnitude of which depends upon uncertainty of the loading, the possibility of it changing during the life of the structure and for a combination of loadings, the likelihood, frequency, and duration of the particular combination. In this approach for reinforced-concrete design, theoretical capacity of a structural element is reduced by a capacityreduction factor to provide for small adverse variations in material strengths, workmanship and dimensions. The structure is then proportioned so that depending on the governing conditions, the increased load cause fatigue or buckling or a brittle-facture or just produce yielding at one internal section or sections or cause elastic-plastic displacement of the structure or cause the entire structure to be on the point of collapse.

Philosophy of designing
The structural design of any structure first involves establishing the loading and other design conditions, which must be supported by the structure and therefore must be considered in its design. This is followed by the analysis and computation of internal gross forces, (i.e.

Design aids
The design of any structure requires many detailed computations. Some of these are of a routine nature. An example is the computation of allowable bending moments for standard sized, species and grades of dimension timber. The rapid development of the

116

Rural structures in the tropics: design and development

T1

T2

h1

60 A

if P=100N T1=T2=58

computer in the last decade has resulted in rapid adoption of Computer Structural Design Software that has now replaced the manual computation. This has greatly reduced the complexity of the analysis and design process as well as reducing the amount of time required to finish a project. Standard construction and assembly methods have evolved through experience and need for uniformity in the construction industry. These have resulted in standard details and standard components for building construction published in handbooks or guides.

Design codes
T2

T1

Many countries have their own structural design codes, codes of practice or technical documents which perform a similar function. It is necessary for a designer to become familiar with local requirements or recommendations in regard to correct practice. In this chapter some examples are given, occasionally in a simplified form, in order to demonstrate procedures. They should not be assumed to apply to all areas or situations.

Design of members in direct tension and compression

P

Tensile systems
Tensile systems allow maximum use of the material because every fibre of the cross-section can be extended to resist the applied loads up to any allowable stress. As with other structural systems, tensile systems require depth to transfer loads economically across a span. As the sag (h) decreases, the tensions in the cable (T1 and T2) increase. Further decreases in the sag would again increase the magnitudes of T1 and T2 until the ultimate condition, an infinite force, would be required to transfer a vertical load across a cable that is horizontal (obviously an impossibility). A distinguishing feature of tensile systems is that vertical loads produce both vertical and horizontal reactions. As cables cannot resist bending or shear, they transfer all loads in tension along their lengths. The connection of a cable to its supports acts as a pin joint (hinge), with the result that the reaction (R) must be exactly equal and opposite to the tension in the cable (T). The R can be resolved into the vertical and horizontal directions producing the forces V and H. The horizontal reaction (H) is known as the thrust. The values of the components of the reactions can be obtained by using the conditions of static equilibrium and resolving the cable tensions into vertical and horizontal components at the support points.

FORCE DIAGRAM FOR POINT A

T1

120 A

T2

h2

if P=100N then T1=T2=100N

Example 7.1
Two identical ropes support a load P of 5kN, as shown in the figure. Calculate the required diameter of the rope, if its ultimate strength is 30 MPa and a safety factor of 4.0 is applied. Also determine the horizontal support reaction at B.

FORCE DIAGRAM FOR POINT A

Chapter 7 Structural design

117

60 30 B

At support B, the reaction is composed of two components: Bv= T2 sin 30= 2.5 sin 30= 1.25kN
A

BH= T2 cos 30= 2.5 cos 30= 2.17kN

Short columns
P=5kN

T1

T2

A column which is short (i.e. the height is small compared with the cross-section area) is likely to fail because of crushing of the material. Note, however, that slender columns, which are tall compared with the cross-section area, are more likely to fail from buckling under a load much smaller than that needed to cause failure from crushing. Buckling is dealt with later.

Free body diagram

T= 4.3 kN

Short columns

5 kN

T2= 2.5 kN

The allowable stress in the rope is 30 = 7.5 N/ mm2 = 7.5 MPa 30 4 = 7.5 N/mm2 = 7.5 MPa 30 4 = 7.5 N/mm2 = 7.5 MPa 4
Stress = Force Force Area Stress = Force Area Stress = Area Therefore:

Slender columns

4.3 103 2 =4.3 573 mm Area required = 3 10 7.5 = 573 mm2 Area required = 3 4.3 107.5 = d2 = 573 mm2 Area required 7.5 A = r2 = 2 4 d A = r2 = 4

Example 7.2
A square concrete column, which is 0.5m high, is made of a nominal concrete mix of 1:2:4, with a permissible direct compression stress of 5.3 MPa (N / mm). What is the required cross-section area if the column is required to carry an axial load of 300kN?

Thus:
d= d= 4 573

A=
= 27 mm (min) 4 573 = 27 mm (min)

300 000 N F = = 56 604 mm2 5.3 N/mm2

i.e. the column should be minimum 238mm square.

118

Rural structures in the tropics: design and development

Design of simple beams Bending stresses

When a sponge is put across two supports and gently pressed downwards between the supports, the pores at the top will close, indicating compression, and the pores at the bottom will open wider, indicating tension. Similarly, a beam of any elastic material, such as wood or steel, will produce a change in shape when external loads are acting on it.
N C A C h

The moment caused by the external loads acting on the beam will be resisted by the moment of this internal couple. Therefore: M= MR= C (or T) h where: M = the external moment MR = the internal resisting moment C = resultant of all compressive forces on the cross- section of the beam T = resultant of all tensile forces on the cross-section of the beam h = lever arm of the reaction couple Now consider a small element with the area (R) at a distance (a) from the neutral axis (NA).
Compression
fc

Tension
a N A

fa

ymax

Figure 7.1 Bending effects on beams

The stresses will vary from maximum compression at the top to maximum tension at the bottom. Where the stress changes from compressive to tensile, there will be one layer that remains unstressed and this is called the neutral layer or the neutral axis (NA). This is why beams with an I-section are so effective. The main part of the material is concentrated in the flanges, away from the neutral axis. Hence, the maximum stresses occur where there is maximum material to resist them. If the material is assumed to be elastic, then the stress distribution can be represented by two triangular shapes with the line of action of the resultant force of each triangle of stress at its centroid. The couple produced by the compression and tension triangles of stress is the internal-reaction couple of the beam section.

ft

Note that it is common practice to use the symbol f for bending stress, rather than the more general symbol. Maximum compressive stress (fc) is assumed to occur in this case at the top of the beam. Therefore, by similar triangles, the stress in the chosen element is:

fa f f = c , fa = a c a ymax ymax
As force = stress area, then the force on the element= fa R= a (fc / ymax) R The resisting moment of the small element is: force distance (a) = a (fc / ymax) R a = Ra2 (fc / ymax)

Chapter 7 Structural design

119

The total resisting moment of all such small elements in the cross-section is: MR= Ra2 (fc / ymax) But Ra2 = I, the moment of inertia about the neutral axis, and therefore

C N A h T

Reinforced-concrete T-beams MR= I (fc / ymax) As the section modulus Zc= I / ymax, therefore MR = fc Zc = M; Similarly MR = ft Zt = M The maximum compressive stress (fc) will occur in the cross-section area of the beam where the bending moment (M) is greatest. A size and shape of crosssection, i.e. its section modulus (Z), must be selected so that the fc does not exceed an allowable value. Allowable working stress values can be found in building codes or engineering handbooks. As the following diagrams show, the concept of a resisting couple can be seen in many structural members and systems.
C N A T

In summary the following equation is used to test for safe bending: fw f= Mmax / Z where: fw = allowable bending stress f = actual bending stress Mmax = maximum bending moment Z = section modulus

Horizontal shear
The horizontal shear force (Q) at a given cross-section in a beam induces a shearing stress that acts tangentially to the horizontal cross-sectional plane. The average value of this shear stress is:

Q A

where A is the transverse cross-sectional area.

h

Rectangular beams
C

This average value is used when designing rivets, bolts and welded joints. The existence of such a horizontal stress can be illustrated by bending a paper pad. The papers will slide relative to each other, but in a beam this is prevented by the developed shear stress.
h

Girders and I beams (1/6 web area can be added to each flange area for moment resistance)

Sliding of layers

No sliding of layers

Rectangular reinforced-concrete beams (note that the steel bars are assumed to carry all the tensile forces).

Figure 7.2 Shearing effects on beams

120

Rural structures in the tropics: design and development

However, the shear stresses are not equal across the cross-section. At the top and bottom edge of the beam they must be zero, because no horizontal shear stresses can develop. If the shear stresses at a certain distance from the neutral axis are considered, their value can be determined according to the following formula:

Q A y I b

where: t = shear stress Q = shear force A = area for the part of the section being sheared off y = perpendicular distance from the centroid of PA to the neutral axis I = moment of inertia for the entire cross-section b = width of the section at the place where shear stress is being calculated.
y

3Q 3Q Q = = 1.5 2bd 2A A 3Q 3Q Q max = = = 1.5 2bd 2A A 3Q 3Q Q For rectangular sections max = = = 1.5 2A A 3Q 2bd Q = 2Q= 1.5 max 3Q 3 Q A 2 a max = = =1 .5 2bd 2A 3Q A Q For square sections max = 2 = 1.5 A 2a Q 3Q max = 2 = 1.5 A 16Q 2 a 4Q For circular sections 3Q max = Q 2 = 3 A 3 D max = 2 = 1.5 A 2a 16Q 4Q = max = 3D2 3A For I-shaped sections of steel 16 beams, Q 4a Qconvenient = = max Qall shearing approximation is to assume that 3D2 3A resistance max d t is afforded by the web plus the part of the flanges that 16Q 4Q = max = Q 2of the web. forms a continuation A 3D 3 max d t Q Thus: max d t Q For I-sections max d t

max =

where: d = depth of beam t = thickness of web If timber and steel beams with spans normally used in buildings are made large enough to resist the tensile and compressive stresses caused by bending, they are usually strong enough to resist the horizontal shear stresses also. However, the size or strength of short, heavily loaded timber beams may be limited by these stresses.

Deflection of beams
y

Centroid for area A

b y

Excessive deflections are unacceptable in building construction, as they can cause cracking of plaster in ceilings and can result in jamming of doors and windows. Most building codes limit the amount of allowable deflection as a proportion of the members length, i.e. 1/180, 1/240 or 1/360 of the length. For standard cases of loading, the deflection formulae can be expressed as:

max = Kc
Q

WL3 EI

It can be shown that the maximum shear stress tmax in a beam will occur at the neutral axis. Thus, the following relations for the maximum shear stress in beams of different shapes can be deduced, assuming the maximum shear force (Q) to be the end reaction at a beam support (column).

Maximum horizontal shear force in beams

where: max= maximum deflection (mm) Mmax f = Kf = constant depending on the type of loading and the cw Z end support conditions W = total load (N) L = effective span (mm) E = modulus of elasticity (N/mm) I = moment of inertia (mm4) It can be seen that deflection is greatly influenced by the span L, and that the best resistance is provided by beams which have the most depth (d), resulting in a large moment of inertia.

Chapter 7 Structural design

w =

4 Qmax 16 Qmax = 3A 3d2

121

Note that the effective span is greater than the clear span. It is convenient to use the centre to centre distance of the supports as an approximation of the effective span. Some standard cases of loading and resulting deflection for beams can be found later in this section.

For I-shaped cross-sections of steel beams

w =

Qmax A

Design criteria
The design of beams is dependent upon the following factors: 1. Magnitude and type of loading 2. Duration of loading 3. Clear span 4. Material of the beam 5. Shape of the beam cross-section

where: tw = allowable shear stress t = actual shear stress Qmax = maximum shear force A = cross-section area Like allowable bending stress, allowable shear stress varies for different materials and can be obtained from a building code. Maximum shear force is obtained from the shear-force diagram. 3. Deflection In addition, limitations are sometimes placed on maximum deflection of the beam (max):

WL3 are designed using the following formulae: Beams = K max c EI

1. Bending stress

fw f =

Mmax Z

where: fw = allowable bending stress f = actual bending stress Mmax = maximum bending moment Z = section modulus This relationship derives from simple beam theory and

max = Kc

WL3 EI

Example 7.3
Consider a floor where beams are spaced at 1200mm and have a span of 4000mm. The beams are seasoned cypress with the following properties: fw= 8.0N/mm, tw = 0.7 MPa (N/mm), E= 8.400 MPa (N/mm), density 500 kg/m Loading on floor and including floor is 2.5 kPa. Allowable deflection is L/240
12 m

Mmax f = max M fmax y max INA = max ymax INA Mmax fmax = fmax andMmax INA = y max ymax INA INA INA= Z ymax =Z ymax
The maximum bending stress will be found in the INA = Z section I of the beam where the maximum bending NA y max= Z momentyoccurs. The maximum moment can be obtained max from the bending-moment diagram. 3 Qmax 3Qmax w = 3 Q = 3Q 2. stress 2 A max w = Shear = 2bdmax 2bd A For 2 rectangular cross-sections:

3 Qmax 3Qmax =3 Qmax =3Qmax ww = 2 A = 2bd 2bd 2A 4 Qmax 16 Qmax w = 4 Q = 16 Q 3 circular A max = cross-sections: 3d2 max w = For 3A 3d2 4 Qmax 16 Qmax = 4 Qmax =16 Q 2 d ww = 3 A = 3max 3A 3d2 Qmax w = Q w = A max A
12 m 1 2 m

1 2 m

1 2 m

12 m

1 2 m

4m

122

Rural structures in the tropics: design and development

Choose a 100 mm by 225 mm timber. The timber required is a little less than that assumed. No Assume a 100 mm by 250 mm cross-section for the recalculations are 6 required 0.78 106unless it is estimated that a 6Z 216 mm if a smaller size d = size=timber would be = beams. smaller adequate b 100 6initially. 0.78 106 6Z had been assumed d = = = 216 mm b 100 (ii) Beam mass= 0.1 0.25 500 9.81= 122.6N/m 6 66 0.78 10 6.Z 0 78 for 10 6 Z vi) 6 = 0.12kN/m Check loading: = shear mm = 216 mm d = =d = = 216 b 100 b 100 Total w = 3+0.12 = 3.12kN/m

(i) Beam loading: w= 1.2m 2.5kN/m2= 3kN/m

3Qmax 3 6.24 103 = 0.42 MPa = 2A 100 225 (iii) Calculate reactions and draw shear-force and 3Qmax 2 3 6.24 103 = 0.42 MPa = = bending-moment diagrams 2A 2 100 225 3 6.24 103 3Q 3 3Qmax As 3 the 6.24 10 safe load for the timber is 0.7 N/mm (MPa) the max = = = 0.42 MPa = 0.42 MPa = = section is adequate resistance to horizontal shear. 2A 225 2 100 in 225 2A 2 100 =
W= 312 kN/m

4m 624 kN

vii) Check deflection 5to ensure WL3 that it is less than 1/240 of max = the span (from Table 7.1) 384 EI 3 5 WL max = 384 EI 3 5 WL3 5 WL max = max = 384 EI 384 EI

SFD

- 624

bd3 100 2253 I= = = 95 106 mm4 where: 12 12 3 3 225 bd 100 E=I8400 MPa = = (N/mm) = 95 106 mm4 12 12 3 3 3 100 225 225 bd3 100 bd 4 106 mm4 = = 95 106 mm = 95 I= = I= 12 12 12 12

W= 3.12kN/m 4m= 12.48kN= 12.48 103N 3 .48 103 43 109 12 5 L= 4 10 mm max = = 13 mm 384 8400 95 106 5 12.48 103 43 109 max = = 13 mm 384 8400 95 106 wl2 M max= = 624 kN BMD 9 3 43 109 8 12 .3 48 10 5 3 10 4 10 5 12.48 max = max = = 13 mm = 13 mm 6 384 384 8400 10 95 106 95 8400 The allowable deflection, 400/240= 16.7 >13. The beam is therefore satisfactory. iii) Calculate maximum bending moment (Mmax) using the equation for a simple beam, uniformly loaded (see Bending moments caused by askew loads Table 7.1) If the beam is loaded so that the resulting bending moment is not about one of the main axes, the moment has to be resolved into components acting about the main w L2 32 .12 42 6 axes. The stresses are then calculated separately relative 2 = 6.24 kNm= 6.24 10 / Nmm M 4 w L = 3.12 = =8 6.24 kNm= 6.24 106 / Nmm M max = max = 8 to each axis and the total stress is found by adding the 8 8 stresses caused by the components of the moment. iv) Find the required section modulus (Z)

Zreq

M max 6.624 106 Zmax =10 = 0.78 106 mm3 M req = 6.24 = = fw =80.78 106 mm3 fw 8

Example 7.4

v) Find a suitable beam depth, assuming 100 mm breadths: From Table 6.3, the section modulus for a rectangular shape is Z= 1/6 bd2

Design a timber purlin that will span rafters 2.4m on centre. The angle of the roof slope is 30 and the purlin will support a vertical dead load of 250 N/m and a wind load of 200N/m acting normal to the roof. The allowable bending stress (fw) for the timber used is 8 MPa. The timber density is 600 kg/m. 1. Assume a purlin cross-section size of 50 mm 125mm. Find an estimated self-load. W = 0.05 0.125 600 9.81 = 37N/m The total dead load becomes 250 + 37= 287N/m

6 0.78 106 6Z = = 216 mm d = b 100

130 103 f= f=

52 103

Chapter 7 Structural design

323 103 103 103 + = 2.5 + 2 = 4.5 N/mm2 = 4.5 MPa 130 103 52 103
3 3

123

323 10 103 10 + = 2.5 + 2 = 4.5 N/mm2 = 4.5 MPa 130 103 52 103 bd2 50 1002 Zx = = = 83 103 mm3 6 6 2. Find the components of the loads relative to the This size is safe, but a smaller size may be satisfactory. 2 1002 50 bd main axes. Try 50 mm 100mm. Zx = = = 83 103 mm3 6 6 Wx= 200N/m + 287 N/m cos 30= 448.5N/m 2 2 100 502 3 bd 100 bd2 Z 50 = = 83 10= 42 3103 mm3 Zx = =y = mm 6 6 6 6 Wy= 287N/m sin 30= 143.5N/m
W1=250 N/m

W2=200 N/m

Zy =

bd2 100 502 = = 42 103 mm3 6 6

2 2 50 bd 323 103100 103 103= 42 103 mm3 Z f y= = 6 =3 + 6 = 3.9 + 2.5 = 6.4 N/mm2 = 6.4 MPa 83 10 42 103 3 323 10 103 103 closer to the allowable This To save 6.4 MPa + is much = 3.9 + 2.5 = 6.4 N/mm2 = stress. 3 3 10 83 10money, 42 50 mm 75 mm should also be tried. In this x 3 3 therefore 50 mm 100mm is chosen. case f>f and w10 323 10 103 f= + = 3.9 + 2.5 = 6.4 N/mm2 = 6.4 MPa 83 103 42 103 Universal steel beams y Steel beams of various cross-sectional shapes are commercially available. Even though the properties of their cross-sections can be calculated with the formulae given in the section Design of members in direct 2 tension and compression, it is easier to obtain them WL wL = Mmax = from handbook tables. These tables will also take into 8 8 consideration the effect of rounded edges, welds, etc. Sections of steel beams are indicated with a 3. Calculate the bending moments about each axis for WL load. wL2 combination of letters and a number, where the letters a uniformly distributed The purlin is assumed = Mmax = 8 represent the shape of the section and the number to be a simple beam. 8 represents the dimension, usually the height, of the wx L2 448.5 2.42 section in millimetres, e.g. IPE 100. In the case of HE Mmax x =WL wL =2 = 323 103 Nmm sections, the number is followed by a letter indicating 8 8 = Mmax = 8 8 the thickness of the web and flanges, e.g. HE 180B. Mmax x Mmax z An example of an alternative method of notation is = +2 fw 2.4 wx L2 f 448.5 3 Z Z 305 102 UB 25, i.e. a 305 mm by 102 mm universal Mmax x = = =y323 10 Nmm x 8 8 beam weighing 25 kg/m. The following example demonstrates another 2 wy M L 143.5 z2.42 Mmax 3 max x method of taking into account the self-weight of the =f =+ fw= 103 10 Nmm 2 y = w L2 Mmax 448.5 2.4 8Z 83 Nmm Zy = x = = x 323 10 structural member being designed. 8 8

f=

Mmax x

Mmax y

2 2 2 4. The actual stress in the timber must be no greater 2 wy L 143.5 2.4 125 50 bd 3 Mmax = = 103 1033 mm Nmm Z = = = 130 10 than the allowable stress. y x8 86 6 M M f = max x + max z fw Z Zy Mmax x Mmax x z 2 2 f 2 2.4 + wy L2 f = 143.5 125 50 bd w 3 Z = = =y 103 10 Nmm Z = Zx = = 130 103 mm3 x 8 6 68

Example 7.5
Design a steel beam, to be used as a lintel over a door opening, which is required to span 4.0 m between centres of simple supports. The beam will be carrying a 220 mm thick and 2.2 m high brick wall, weighing 20kN/m. Allowable bending stress is 165MPa. Uniformly distributed load caused by brickwork is 0.22 2.2 4.0 20= 38.7kN. Assumed self-weight for the beam is 1.5kN. (Note: the triangular load distribution for bricks above the lintel would result in a slightly lower load value). 40.2 4.0 distributed load W = 38.7 Total + 1.5 = WL uniformly Mmax = = = 20.1 kNm = 20.1 106 Nmm 40.2 kN 8 8

5. Try the assumed purlin size of 50 125mm.

bd2 125 502 = 125 =2 = 52 103 mm3 bd2 Zy 50 6 103 mm3 6 Zx = = = 130 1252 6 bd2 506 Zx = = = 130 103 mm3 6 6 2 125 502 bd Zy = = = 52 103 mm3 6 6
f= 323 103 103 103 + = 2.5 + 2 = 4.5 N/mm2 = 4.5 MPa 130 103 52 103

Mmax =

WL 40.2 4.0 = = 20.1 kNm = 20.1 106 Nmm 8 8

bd2 125 502 Zy = = = 52 103 mm3 23 23 6 6 323 103 10 10 125 50 bd 2.5 23 4.5 3N/mm2 = 4.5 MPa f =Zy = +10 =mm =+ = 52 103 52 6 103 130 6 bd2 50 1002
3 3

Zreq =

20.1 106 = 0.122 106 mm3 = 122 cm2 165 20.1 106 = 0.122 106 mm3 = 122 cm2 165

Zreq =

124

Rural structures in the tropics: design and development

Suitable sections as found in a handbook would be:

Section Zx-x Mass

INP 160 IPE 180 HE 140A HE 120A

117 cm 146cm 155 cm 144 cm

17.9 kg/m 18.8 kg/m 24.7 kg/m 26.7 kg/m

Although the total value of the load has increased, the maximum shear force remains the same but the maximum bending is reduced when the beam is cantilevered over the supports.

Choose INP 160 because it is closest to the required section modulus and has the lowest weight. Then recalculate the required Z using the INP 160 weight: 4.0 17.9 9.81= 702 N, which is less than the assumed self-weight of 1.5 kN. A recheck on the required Z reveals a value of 119cm, which is close enough.

Continuous beam

Continuous beams
A single continuous beam extending over a number of supports will safely carry a greater load than a series of simple beams extending from support to support. Consider the shear force and bending moment diagrams for the following two beam loadings:
8m 5 kN/m

Simple beam

20 kN +20

20 kN

Although continuous beams are statically indeterminate WL and the calculations are complex, BM = approximate values 6 can be found with simplified equations. Conservative equations for two situations are as follows: Load concentrated between supports: BM = Load uniformly distributed: BM =

WL 12

WL 6

M max = 40 kNm

It is best to treat the two end sections as simple beams. WL BM = Standard cases of beam loading 12 A number of beam loading cases occur frequently and it is useful to have standard expressions available for them. Several of these cases will be found in Table 7.1.
2m

2m

8m 5 kN/m

Composite beams
In small-scale buildings the spans are relatively small and, with normal loading, solid rectangular or square sections are generally the most economical beams. However, where members larger than the available sizes and/or length of solid timber are required, one of the following combinations may be chosen: 1. Arranging several pieces of timber or steel into a structural frame or truss. 2. Universal steel beams. 3. Built-up timber sections with the beam members nailed, glued or bolted together into a solid section, or with the beam members spaced apart and only connected at intervals. 4. Strengthening the solid timber section by the addition of steel plates to form a flitch-beam. 5. Plywood web beams with one or several webs. 6. Reinforced-concrete beams.

30 kN/m +20 +10

30 kN/m

-10 -20 M max = 30 kNm

-10

-10

Chapter 7 Structural design

125

Table 7.1

Beam equations
Loading diagram
W a A L a+b=L b B

Shear force at x: Qx

Bending moment at x: Mx

Deflection at x: x

QA =

Wb L Wa L W 2 W 2

Wab L When a = b Mc = Mc = WL 4 WL 8 L 2

Wa2 b2 3EIL

QB = QA =

Total W =wL x A L B

M max = at x =

max

5WL3 384 EI L 2

QB = -

at x =

Total W = wL 2 W A L x B

QA = QB = -

2W wL = 3 3 W wL =3 6

M max = 0.064mL2 at x = 0.577 L

max

= 0.00652

wL4 EI

at x = 0.519

Total W = wL 2 W A L B

QA = QB = -

W wL = 2 4 W wL =2 4

M max = at x =

wL2 12 L 2

max

wL4 120 EI L 2

at x =

A L Total W = wL

QA = QB = W

M A = - WL

WL3 3EI

QA = W
B

A L Total W = wL 2 W A L W A a L a+b=L Total W = wL A L Total W = wL 2 W A x W b

QA = 0

MA = -

WL wL2 =2 2

WL3 wL4 = 3EI 3EI

QA = W
B

MA = -

QA = 0

WL wL2 =3 6

wL4 30 EI

QA =
B

Wb L Wa L W 2

MA = MB = -

Wab 2 L2 Wa b L2 WL 12
2

QB = QA =
B

Wa3b3 3EIL

W QB = 2 QA = 2W 3 W 3

M A = MB = -

WL3 384 EI wL4 764 EI

MA = MB = -

WL wL2 =10 20 WL wL2 =15 30

max

B L W W

QB = -

at x = 0.475L

QA =
L R1 W W L R1 R2 R2 W

W 2

M max =

WL * 6

max

wL3 192 EI

QA =

W 2

M max =

WL * 12

max

wL3 384 EI

126

Rural structures in the tropics: design and development

Built-up timber beams

When designing large members, there are advantages in building up solid sections from smaller pieces because these are less expensive and easier to obtain. Smaller pieces also season properly without checking. The composite beams may be built up in ways that minimize warping and permit rigid connections between columns and beams. At the same time the importance of timber defects is decreased, because the load is distributed to several pieces, not all with defects in the same area.

Columns
Although the column is essentially a compression member, the manner in which it tends to fail and the amount of load that causes failure depend on: 1. The material of which the column is made. 2. The shape of cross-section of the column. 3. The end conditions of the column. The first point is obvious: a steel column can carry a greater load than a timber column of similar size. Columns with a large cross-section area compared with the height are likely to fail by crushing. These short columns have been discussed earlier.

Built-up solid beam

Buckling of slender columns

If a long, thin, flexible rod is loaded axially in compression, it will deflect a noticeable amount. This phenomenon is called buckling and occurs when the stresses in the rod are still well below those required to cause a compression/shearing-type failure. Buckling is dangerous because it is sudden and, once started, is progressive.

Built-up solid column

Ra

fte

m me

be

rs

sp

d ace

ap

art

Tie member

Figure 7.3 Built-up beams and trusses

Built-up solid beams are normally formed by using vertical pieces nailed or bolted together: Nailing is satisfactory for beams up to about 250 mm in depth, although these may require the use of bolts at the ends if the shear stresses are high. Simply multiplying the strength of one beam by the number of beams is satisfactory, provided that the staggered joints occur over supports. Built-up sections with the members spaced apart are used mainly where the forces are tensile, such as in the bottom chords of a truss. Where used in beams designed to resist bending, buckling of the individual members may have to be considered if those members have a large depth-to-width ratio. However, this can be avoided by appropriate spacing of stiffeners that connect the spaced members at intervals. Where the loading is heavy, the beam will require considerable depth, resulting in a large section modulus to keep the stresses within the allowable limit. If sufficient depth cannot be obtained in one member, it may be achieved by combining several members, such as gluing the members together to form a laminate.

Although the buckling of a column can be compared with the bending of a beam, there is an important difference in that the designer can choose the axis about which a beam bends, but normally the column will take the line of least resistance and buckle in the direction where the column has the least lateral unsupported dimension. As the loads on columns are never perfectly axial and the columns are not perfectly straight, there will always be small bending moments induced in the column when it is compressed. There may be parts of the cross-section area where the sum of the compressive stresses caused by the load on the column could reach values larger than the allowable or even the ultimate strength of the material.

Chapter 7 Structural design

127

Therefore the allowable compressive strength cw is reduced by a factor k, which depends on the slenderness ratio and the material used. Pbw = k cw A where: Pbw = allowable load with respect to buckling k = reduction factor, which depends on the slenderness ratio cw = allowable compressive stress A = cross-section area of the column When the load on a column is not axial but eccentric, a bending stress is induced in the column as well as a direct compressive stress. This bending stress will need to be considered when designing the column with respect to buckling.

Rotation

2. Fixed in position but not in direction (pinned).

Slenderness ratio
As stated earlier, the relationship between the length of the column, its lateral dimensions and the end fixity conditions will strongly affect the columns resistance to buckling. An expression called the slenderness ratio has been developed to describe this relationship:
P P

KL l = r r

where: = slenderness ratio K = effective length factor whose value depends on how the ends of the column are fixed L = length of the column r = radius of gyration (r = I / A) l = effective length of the column (K L) There are four types of end condition for a column or strut:
P Rotation

3. Fixed in direction but not in position.

side movement

4. Fixed in position and in direction. The consideration of the two end conditions together results in the following theoretical values for the effective length factor (Kp is the factor usually used in practice).

1. Total freedom of rotation and side movement like the top of a flagpole. This is the weakest end condition.

128

Rural structures in the tropics: design and development

K=10

Columns and struts with both ends fixed in position and effectively restrained in direction would theoretically have an effective length of half the actual length. However, in practice this type of end condition is almost never perfect and therefore somewhat higher values for K are used and can be found in building codes. In fact, in order to avoid unpleasant surprises, the ends are often considered to be pinned (Kp = 1.0) even when, in reality, the ends are restrained or partially restrained in direction. The effective length can differ with respect to the different cross-sectional axes:
y

Both ends pinned

x ly

K=20

ly

lx

One end fixed in direction and position, the other free

K=05 Kp=065

1. A timber strut that is restrained at the centre has only half the effective length when buckling about the y-y axis as when buckling about the x-x axis. Such a strut can therefore have a thickness of less than its width.

Both ends fixed in direction and position

d

l B

Kp 085

07

One end pinned, the other fixed in direction and position

2. In the structural framework, the braces will reduce the effective length to l when the column A-B is buckling sideways but, as there is no bracing restricting buckling forwards and backwards, the effective length for buckling in these directions is 3l. Similarly, the bracing struts have effective lengths of 1/2 d and d respectively.

Chapter 7 Structural design

129

load of 15 kN. Allowable compressive stress (cw) for the timber is 5.2MPa.

l = 2L

15 kN

Pin
l=L

3. The leg of a frame, which is pinned to the foundation, has the effective length l = 2 L but, if the top is effectively secured for sideways movement, the effective length is reduced to l = L.

l = 2L

4. In a system of post and lintel where the bottom of the post is effectively held in position and secured in direction by being cast in concrete, the effective length l= 2 L.

L = 3 000
b

Axially loaded timber columns

Timber columns are designed with the following formulae: 1. In this case, the end conditions for buckling about the x-x axis are not the same as about the y-y axis. Therefore both directions must be designed for buckling (Where the end conditions are the same, it is sufficient to check for buckling in the direction that has the least radius of gyration). Find the effective length for buckling about both axes. Buckling about the x-x axis, both ends pinned: lx= 1.0 3 000= 3000mm Buckling about the y-y axis, both ends fixed: ly= 0.65 3 000= 1950mm

KL and Pbw = k cw A r

Note that in some building codes a value of slenderness ratio in the case of sawn timber is taken as the ratio between the effective length and the least lateral width of the column l/b.

Example 7.6
Design a timber column that is 3 metres long, supported as shown in the figure and loaded with a compressive

Table 7.2

Reduction factor (k) for stresses with respect to the slenderness ratio for wood columns
Slenderness ratio l/b l/r k 2.9 10 1.0 5.8 20 1.00 8.7 30 0.91 11.5 40 0.81 14.4 50 0.72 17.3 60 0.63 20.2 70 0.53 23.0 80 0.44 26.0 90 0.35 28.8 100 0.28 34.6 120 0.20 40.6 140 0.14 46.2 160 0.11 52.0 180 0.40

b= least dimension of cross-section; r= radius of gyration

x =
130

lx 3 000 = = 83 rx 36.1

Rural structures in the tropics: design and development

2. Choose a trial cross-section, which should have its largest lateral dimension resisting the buckling d 125 the largest effective length. Try rx = about = the axis = 36.with 1 mm 250 3 mm 2 3125mm. The section properties are:

l 3 000 xl = x 1 = = 83 rx 950 36.= 1 90 gives ky = 0.16 y = y = ry 21.7

d bd 125 d= = 125 50 = 6250mm rx =A = 36 .125= 136 mm rx= = = .1 mm 2 32 3 2 32 3

Pwx= 0.41 5.2 9375 = 19988 N, say 20 kN Pwy= 0.35 5.2 9375 = 17063 N, say 17 kN

The allowable load with respect to buckling on the b d 50125 ly 000 1 950 F 15 ry = 4 mm rx = = = = 14 =.36 .1 mm = cross-section = MPa 90 with 75 mm 125mm is therefore c column = y== r =.7 1.6 2 3 2 3 21 2 3 2 3 A 9Although y375 17 kN. this is bigger than the actual load, d 125 rx = = = 36.1 mm further iterations to find the precise section to carry the 2 b 3 b2 50 3 50 ry =ry = = = = 14 .414 mm = .4 mm 15kN are not necessary. 2 d32 3 2 32 3 125 The compressive stress in the chosen cross-section rx = = = 36.1 mm 2 3 b 2 3 50 will be: 3 000 = 14.4 mm ry Find = lx the = 3. allowable x = 2 r 3= 36 2 3 = 83 load with regard to buckling on F 15 000 b the x 50 .1 for buckling in both directions. column c = = = 1.6 MPa ry = = = 14.4 mm A 9 375 2 3 2 3 l lx 3 000 3 000 gives k = 0.41 (from Table 7.2) x b =xx = = 50= = 83 x = 83 rx= rx36.1 36 .1.4 mm ry = = 14 This is much less than the allowable compressive 2 3 2 3 stress, which makes no allowance for slenderness. 3 000 ly = lx1 = 950 = 83 gives k = 0.16 (from Table 7.2) y = x = = 135 y r 14.4 36.1 Axially loaded steel columns lxry 3x000 x = = = 83 The allowable loads for steel columns with respect to rx l 1 36 .1 l 950 1 950 yk y P buckling can be calculated in the same manner as for = A y w= = =c = 135 y l= = 135 3 000 r 14 . 4 r 14 . 4 x y timber. However, the relation between the slenderness P = 0.41 5.2 6250mm= 13325 N wx = =y = 83 x 36 .5.2 1 6250mm= 5200 N x ratio and the reduction factor (k) is slightly different, Pwy=r0.16 ly 1 950 125 = = = 135 as seen in Table 7.3. rx =y = 36.1 mm r 14.4 2 3 y allowable load with respect to buckling is 4. As the ly 1 950 y = smaller = = 135 the actual load, a bigger cross-section Example 7.7 ry125 14.than 4 125 needs to be chosen. Try 75 mm 125 mm and repeat Calculate the safe load on a hollow square steel stanchion rx = rx = 36 .136 mm = .1 mm ly =321 950 2 3 steps 2 and 3. whose external dimensions are 120mm 120mm. The y = = = 135 ry 75 14.4 walls of the column are 6 mm thick and the allowable ry = 125 = 21.7 mm rx = = 36 . 1 mm Section properties: compressive stress cw = 150 MPa. The column is 2 3 2 3 4 metres high and both ends are held effectively in 125 rA= = 75 = 36 . 1 mm x 125 = 9375mm position, but one end is also restrained in direction. 75 2r 75 3 ry = .721 mm = = 21 = .7 mm y The effective length of the column l = 0.85L= 0.85 2 32 3 3 125 BD 1204 1084 I bd3 = 3400mm. 4000 rx = = 36.1 mm rx = ry = = = = 46.6 mm 2 3 75 A 12 ( BD bd ) 12 (1202 1082) ry = = 21.7 mm 2 3 75 BD3 bd3 1204 1084 I ry = = 21.7 mm rx = ry = = = = 46.6 mm A 12 ( BD bd ) 12 (1202 1082) 2 3

75 ry = = 21.7 mm Find the 2 3 allowable buckling load for the new crosssection:
x = lx 3 000 = = 83 rx 36.1

l 3 400 = = 73 gives k = 0.72 by interpolation r 46.6

gives kx = 0.41

l 3 400 2 = = = 73 Pw= k cw A= 0.72 150 (120 - 1082)= 295kN. r 46.6 b= D 12 0.87 D 4 b=

D 12 0.87 D 4 Reduction factor (k) for stresses with respect to the slenderness ratio for steel columns
Table 7.3

= y
k k

ly

ry

1 950 =10 = 20 90 21.7

0.95 160 0.23 150

30 0.92 170 0.21

40 0.90 180 0.19

50 0.86 190 0.17

60 0.81 200 0.15

70 0.74 210 0.14

80 0.67 220 0.13

90 0.59 230 0.12

100 0.51 240 0.11

110 0.45 250 0.10

120 0.39 300 0.07

130 0.34 350 0.05

140 0.30

0.97

0.26

c =

F 15 000 = = 1.6 MPa A 9 375

l 4000 = = 13.3 b 300

Chapter 7 Structural design

l 4000 = = 13.3 b 300

131

l 4000 = = 13.3 b 300

ry =
3 4 bd3 loaded 1084 columns BD 120 I Obviously, by the law of superposition, the added Axially concrete f = = = 46.6 mm 2 2 + 1 i.e. stresses of the two load effects must be below the building codes permit the use of plain concrete A 12 (Most BD bd ) ) 12 (120 108 Pcw fw f allowable stress. only in short columns, that is to say, columns where the + 1 i.e. ratio of the effective length to least lateral dimension Pcw fw does not exceed 15, i.e. l/r 15. If the slenderness ratio is f Therefore between 10 and 15, the allowable compressive strength + 1 i.e. 3 400 f P l cw w must The tables of figures relating to l/b in = be = reduced. = 73 .6 slenderness ratio are only approximate, axial compressive stress bending stress place rof a46 true + 1 as radii of gyration depend on both b and d values in allowable compressive stress allowable bending stress axial compressive stress bending stress the cross-section and must be used with caution. In the + 1 allowable compressive stress allowable bending stress case of a circular column:

b=

D 12 0.87 D , where 4

D= the diameter of the column.

axial compressive stress bending stress + 1 allowable compressive stress allowable bending stress c f + 1 which can be transferred to: fw k cw c f + 1 fw k cw

Table 7.4

M f P1 c + cw + cw 1 K A k fw Z f

Permissible compressive stress (Pcc) in concrete for columns (MPa or N/mm2)

Concrete mix Slenderness ratio, l/b

cw

Example 7.9
15

Prescribed C10 C15 C20 Nominal 1:3:5 1:2:4 1:1.5:3

10

11

12

13

14

3.2 3.9 4.8

3.1 3.8 4.6

3.0 3.7 4.5

2.9 3.6 4.3

2.8 3.5 4.2

2.7 3.4 4.1

Determine within 25 mm the required diameter of a timber post loaded as shown in the figure. The bottom of the post is fixed in both position and direction by being cast in a concrete foundation. Allowable stresses timber used are cw= 9 MPa and fw= 10 MPa. for D2 the 2002 A= = = 31 400 mm2 4 4
P = 30 kN e = 500 F = 5 kN

3.1 3.8 4.7

3.0 3.7 4.5

2.9 3.6 4.4

2.8 3.5 4.2

2.7 3.4 4.1

2.6 3.3 4.0

Higher stress values may be permitted, depending on the level of work supervision.

Z=

D3 2003 = = 785 400 mm2 32 32

L = 3 000

Example 7.8
A concrete column with an effective length of 4 metres has a cross-section of 300 mm by 400 mm. Calculate the allowable axial load if a nominal 1:2:4 concrete mix is to be used. Slenderness ratio

r=

D 200 = = 50 mm 4 4

l 4000 = = 13.3 b 300 =

Hence Table 7.4 gives Pcc= 3.47N/mm by interpolation. Pw= Pcc A= 3.47 300 400= 416.4kN.
+ 1 loaded i.e. Eccentrically timber and steel columns

l 6 300 = = 126 r 50

f Pcw fw Where a column is eccentrically loaded, two load effects need to be considered: The axial compressive stress caused by the load and the bending stresses caused by the eccentricity of the load. xial compressive stress bending stress + 1 wable compressive stress allowable bending stress

The load of 5 kN on the cantilever causes a bending moment of M= F e= 5kN 0.5m= 2.5kNm in the post below the cantilever. The effective length of the post= L K= 3000 2.1= 6300mm. Try a post with a diameter of 200mm. The cross-sectional properties are:

K A

fw

cw

D2 2002 = = 31 400 mm2 4 4 132 D2 2002 A= = = 31 400 mm2 4 D3 2003 4 Z= = = 785 400 mm2 32 32 D2 2002 A= = = 31 400 mm2 4 4 D3 2003 Z= = = 785 400 mm2 32 32 D3 2003 D 400 200 2 Z= = 785 r== = mm = 50 mm 32 32 4 4 A= D3 2003 Z = D 200 = = 785 400 mm2 r = 32 = =32 50 mm 4 4 D 200 r= = = 50 mm 4 4 l 6 300 The slenderness ratio = = 126 P1 = M r + cw 50 cw D 200 M K P fw Z 1A cw r= = = 50+mm cw 4 4 K A fw Z Interpolation in Table 7.3 gives k P1 l cw = 0.18 M 6 300 126 cw = += = K A r fw 50 Z l 6 P1 M += cw 300 = = 126 cw K A fr Z 50 w

Rural structures in the tropics: design and development

30 000 9 2.5 106 l + 6 300 = 8.17 N/mm2 9 N/mm2 = = 10 = 126 000 9 785 2 .5 106 0.1830 31 400 400 r + 50 = 8.17 N/mm2 9 N/mm2 0.18 31 400 10 785 400 6 30 000 9 2.5 10 + = 8.17 N/mm2 9 N/mm2 8 31 400 10 785 400 the post has a diameter of 200 mm, it will be 6 30 000 9 If 2.5 10 2 2 + able to carry = 8.17 N/mm 9 N/mm the loads, but the task was to determine 0.18 31 400 10 785 400 the diameter within 25 mm. Therefore a diameter of 6 tried. 300 175mm should also be = = 144 6. 300 43 75 = = 144 43.75 6 300 = = 144 k = 0.13 43.75 6 300 = = 144 43.75 30 000 9 2.5 106 + = 23 N/mm2 > 9 N/mm2 000 9 167 2.5 106 0.1330 24050 10 480 + = 23 N/mm2 > 9 N/mm2 0.13 24050 10 167 480 6 30 000 9 2.5 10 2 2 + This diameter = 23 N/mm 9 N/mm is too > small, so a diameter of 200mm 13 24050 10 167 480 should be chosen. It will be appreciated that the choice 30 000 9 2.5 106 2 + of effective = 23 based >end N/mm 9 N/mm length on fixity2has a great effect 0.13 24050 10 167 480 on the solution.

Many agricultural buildings have walls built of blocks or bricks. The same design approach as that shown for plain concrete with axial loading can be used. The maximum allowable compressive stresses must be ascertained, but the reduction ratios can be used as before. l 2 800 = = 23.3 b 120 Example 7.10

Determine the maximum allowable load per metre of a 120mm thick wall, with an effective height of 2.8 metres and made from concrete grade C 15: (a) when the load is 2 800 l 2 800 lload = 23 central; (b) when the is eccentric =800 == 23 .3 .3by 20mm. l 2 120 =bb 120 = 23.3 b 120 l 2 800 = 23.3 3.3 Slenderness ratio, = (2.8 2.0 ) = 2.27 N/mm2 = 2.27 M = 2.8 b P120 cw 5 Interpolation gives:

3.3 2 (2 )= Pcw= =2 2 . 8 3.3 (2 .8 0= 2.27 N/mm = 2.27 MPa 2 P . 8 .8 2.2 0.) 2.27 N/mm =2 .27 MPa 3 . 3 (5 Pcw = cw 2.8 25 .8 2.0 ) = 2.27 N/mm2 = 2.27 MPa 5 20 3.3 = 0.167 (2.8 2.0 ) = 2.27 N/mm2 e== Pcw = 2.8 2.27 MPa b 120 5 Pw = A Pcw = 1.0 0.12 Allowable load 2.27 106 = 272.4 kN/m wall 20 e e =20 0.167 Ratio of eccentricity = 120 == 0.167 e b 20 =b 120 = 0.167 b 120

Plain and centrally reinforced concrete walls

Basically walls are designed in the same manner as e 20 columns, but there are a few differences. A wall is A double interpolation = gives: = 0.167 1.06 106 b Pw =120 = 127.2 kN/m A Pcw = 1.0 0 .12 distinguished from a column by having a length that is 1 000 more than five times the thickness. Pcw= 1 .06N/mm = 1 .06MPa Plain concrete walls should have a minimum thickness of 100mm. Allowable load 6 Where the load on the wall is eccentric, the wall 10 1.06 6 10 1.06 Pw = = 127 .2 kN/m wall =A A P Pcw= =11 .00. 0 .12 6 P = 127 .2 kN/m wall . 0 12 10 1 . 06 must have centrally placed reinforcement of at least w 1 000 Pw = A = 127.2 kN/m wall Pcw =cw 1.0 0 .12 1 000 0.2 percent of the cross-section area if the eccentricity 1 000 ratio e/b exceeds 0.20. This reinforcement may not be 1.06 106 e wall Pw Central = 127 .2 kN/m = A Pcw = 1.0 0 .12 included in the load-carrying capacity of the wall. reinforcement is not required because < 20 1 000 b

e e < 20 < 20 e b < b20 b e < 20

Chapter 7 Structural design

133

Table 7.5

Allowable compressive stress, Pcw for concrete used in walls (N/mm)

Concrete grade or mix Slenderness ratio l/b 0.00 Ratio of eccentricity of the load e/b Plain concrete walls 0.10 0.20 Centrally reinforced concrete walls 0.30 0.40 0.50

C20

25 20 15 10

2.4 3.3 4.1 4.8 2.0 2.8 3.4 3.9 2.3 2.7 3.2 2.3 3.2 4.1 4.7 3.0 3.7 4.3 2.7 3.3 3.8

1.7 2.3 3.0 3.7 1.3 1.9 2.4 3.0 1.6 2.0 2.5 1.6 2.2 2.9 3.6 2.1 2.7 3.4 1.8 2.3 2.9

0.9 1.4 2.0 2.7 0.7 1.1 1.7 2.2 1.0 1.4 1.8 0.8 1.3 1.9 2.6 1.3 1.9 2.5 1.0 1.6 2.1

0.8 0.8 0.8 0.7 0.7 0.7 0.5 0.5 0.5 0.8 0.8 0.8 0.7 0.7 0.7 0.6 0.6 0.6

0.4 0.4 0.4 0.35 0.35 0.35 0.3 0.3 0.3 0.4 0.4 0.4 0.35 0.35 0.35 0.3 0.3 0.3

0.3 0.3 0.3 0.25 0.25 0.25 0.2 0.2 0.2 0.3 0.3 0.3 0.25 0.25 0.25 0.2 0.2 0.2

C15

25 20 15 10

C10

20 15 10

1:1:3 1:1:2

25 20 15 10

1:2:3

20 15 10

1:2:4

20 15 10

Higher values of stress may be permitted, depending on the level of work supervision.

Trusses
It can be seen from the stress distribution of a loaded beam that the greatest stress occurs at the top and bottom extremities of the beam.

For these situations where bending is high but shear is low, for example in roof design, material can be saved by raising a framework design. A truss is a pinpointed framework.

fc C N T A

C h T N A

ft

This led to the improvement on a rectangular section by introducing the I-section in which the large flanges were situated at a distance from the neutral axis. In effect, the flanges carried the bending in the form of tension stress in one flange and compression stress in the other, while the shear was carried by the web.

A truss concentrates the maximum amount of materials as far away as possible from the neutral axis. With the resulting greater moment arm (h), much larger moments can be resisted. Resistance of a truss at a section is provided by: M= C h= T h

134

Rural structures in the tropics: design and development

where: C = T in parallel chords and: C = compression in the top chord of the truss. T = tension in bottom chord of a simply supported truss. h= vertical height of truss section. If either C or T or h can be increased, then the truss will be capable of resisting heavier loads. The value of h can be increased by making a deeper truss. Allowable C- or T-stresses can be increased by choosing a larger cross-section for the chords of the truss, or by changing to a stronger material. A framework or truss can be considered as a beam with the major part of the web removed. This is possible where bending stresses are more significant than shear stresses. The simple beam has a constant section along its length, yet the bending and shear stresses vary. The truss, comprising a number of simple members, can be fabricated to take into account this change in stress along its length. The pitched-roof truss is the best example of this, although the original shape was probably designed to shed rainwater. Roof trusses consist of sloping rafters that meet at the ridge, a main tie connecting the feet of the rafters and internal bracing members. They are used to support a roof covering in conjunction with purlins, which are laid longitudinally across the rafters, with the roof cover attached to the purlin. The arrangement of the internal bracing depends on the span. Rafters are normally divided into equal lengths and, ideally, the purlins are supported at the joints so that the rafters are only subjected to axial forces. This is not always practicable because purlin spacing is dependent on the type of roof covering. When the purlins are not supported at the panel joints, the rafter members must be designed for bending as well as axial force. See Figure 7.2. The internal bracing members of a truss should be triangulated and, as far as possible, arranged so that long members are in tension and compression members are short to avoid buckling problems. The outlines in Figure 7.3 give typical forms for various spans. The thick lines indicate struts. The lattice girder, also called a truss, is a plane frame of open web construction, usually with parallel chords

or booms at top and bottom. There are two main types, the N- (or Pratt) girder and the Warren girder. They are very useful in long-span construction, in which their small depth-to-span ratio, generally about 1/10 to 1/14, gives them a distinct advantage over roof trusses. Steel and timber trusses are usually designed assuming pin-jointed members. In practice, timber trusses are assembled with bolts, nails or special connectors, and steel trusses are bolted, riveted or welded. Although these rigid joints impose secondary stresses, it is seldom necessary to consider them in the design procedure. The following steps should be considered when designing a truss: 1. Select general layout of truss members and truss spacing. 2. Estimate external loads to be applied including self-weight of truss, purlins and roof covering, together with wind loads. 3. Determine critical (worst combinations) loading. It is usual to consider dead loads alone, and then dead and imposed loads combined. 4. Analyse the framework to find forces in all members. 5. Select the material and section to produce in each member a stress value that does not exceed the permissible value. Particular care must be taken with compression members (struts), or members normally in tension but subject to stress reversal caused by wind uplift. Unless there are particular constructional requirements, roof trusses should, as far as possible, be spaced to achieve minimum weight and economy of materials used in the total roof structure. As the distance between trusses is increased, the weight of the purlins tends to increase more rapidly than that of the trusses. For spans up to around 20 m, the spacing of steel trusses is likely to be about 4metres and, in the case of timber, 2metres. The pitch, or slope, of a roof depends on locality, imposed loading and type of covering. Heavy rainfall may require steep slopes for rapid drainage; a slope of 22 is common for corrugated steel and asbestos roofing sheets. Manufacturers of roofing material usually make recommendations regarding suitable slopes and fixings.

Ridge Internal bracing Purlin Rafter

Pu
Eaves

rlin

ccu

rs

be

tw

n ee

joi

nts

Main tie

Bending

Figure 7.4 Truss components

Chapter 7 Structural design

135

PALLADIAN TRUSS

PRATT TRUSS

W or BELGIAN TRUSS

UP TO 8 m span

FAN TRUSS UP TO 12 m span

PRATT TRUSS

N - GIRDER LONG SPAN CONSTRUCTION

WARREN GIRDER

Figure 7.5 Types of trusses

Member Dead Imposed Dead + imposed Load D Load I Load D+I

Wind Load W

Design Load

80m

A simplified approach can be taken if the intention is to use a common section throughout. Once the layout has been chosen, the member that will carry the maximum load can be established. An understanding of the problems of instability of compression members will lead the designer to concentrate on the top chord or rafter members. A force diagram or method of sections can then be used to determine the load on these members and the necessary size.

20m

LAYOUT CHOSEN (nodes at purlin points)

Example 7.11
A farm building comprising block walls carries steel roof trusses over a span of 8 metres. Roofing sheets determine the purlin spacings. Design the roof trusses.

Assume a force analysis shows maximum rafter forces of approximately 50kN in compression (D + I) and 30kN in tension (D + W), outer main tie member 50kN tension (D + I) and 30kN compression (D + W). A reversal of forces caused by the uplift action of wind will cause the outer main tie member to have 50kN of tension and 30kN of compression.

075m

To enable the designer to determine the maximum design load for each member, the member forces can be evaluated either by calculation or graphical means, and the results tabulated as shown:

138m

138m

08

136

Rural structures in the tropics: design and development

Consulting a structural engineering handbook reveals that a steel angle with a section of 65 mm 50 mm 6 mm and an effective length of 1.8 m can safely carry 29kN in compression. Rafter: Using two angles back-to-back will be satisfactory because the distance between restraints is only 1.38 m. (Note that angles must be battened together along the length of the rafter). Main Tie: The 65mm 50mm 6mm section can carry the required tensile force. Although its length is a little greater than 1.8m, the compressive load brought about by wind uplift is safe as the design codes allow a greater slenderness ratio for intermittent loads such as wind. Finished Design: Note the use of a sole plate to safely distribute the load to the blockwork wall to ensure that the bearing stress of the blocks is not exceeded. See Figure 7.4.

Frames
Apart from the roof truss, there are a number of other structural frames commonly used in farm building construction. They include portal frames, pole barns and post-and-beam frames.

A single-bay portal frame consists of a horizontal beam or pitched rafters joined rigidly to vertical stanchions on either side to form a continuous plane frame. For design purposes, portal frames can be classified into three types: fixed base, pinned base (two pins), pinned base and ridge (three pins). The rigid joints and fixed bases have to withstand bending moments and all bases are subjected to horizontal, as well as vertical, reactions. Hence foundation design requires special attention. The externally applied loads cause bending moments, shear forces and axial forces in the frame. Portal frames are statically indeterminate structures and the complexity of the analysis precludes coverage here. However, the results of such calculations for a number of standard cases of loading are tabulated in handbooks. Using these and the principle of superposition, the designer can determine the structural section required for the frame. Determining the maximum values of the bending moment, shear force and axial force acting anywhere in the frame allows the selection of an adequate section for use throughout the frame. Care must be exercised to ensure that all joints and connections are adequate.

350 80 250 160

2 x 12 mm holes in 6 mm thick soleplate

150

100

Rafter 2 -65 x 60 x 6

800 1 375
1 375

10 Pitch 50 x 50 x 6 L Longitudinal ties 2 050 3 950

Notes: All welds to be 4mm llet All bolts to be M16 Gusset plates to be 8mm thick Internal bracing shown 65 x 50 x 6 to use common section (size can be reduced if others available) All sections in grade 43 steel Purlin supports: 70 x 70 x 6 with 2 x 6 holes

Figure 7.6 Finished design of the roof truss

750

350

Chapter 7 Structural design

137

Portal frames may be made of steel, reinforced concrete or timber. With wider spans the structural components become massive if timber or reinforced concrete is used. Hence, steel frames are most common for spans over 20 m. At the eaves, where maximum bending moments occur, the section used will need a greater depth than at other points in the frame.

Example 7.12
Design the roof of a building using block walls, timber posts and rafters as shown in the figures below.
Design section

Knee braces

Timber posts

RC PORTAL FRAME

It is assumed that the knee braces reduce the effective span of the rafters between the central wall and the timber posts. The moments and forces involved are as shown in the diagram below. Self-weights and service load have been estimated. Continuity over post and brace has been disregarded. This provides a simple but safe member.
277 kN 277 kN 138 m knee brace S.F.D. (kN) 277 kN 074 m

TIMBER PORTAL FRAME

post 503

138 m

Figure 7.7 Portal or rigid frame

226 -051 3119

-328 2427

Pole barns are usually built with a relatively simple foundation, deeper than usual, and backfilled with rammed earth. Pole barns are braced between columns and rafters in each direction. The braces serve to reduce the effective length of compression members and the effective span of rafters and other beam members. This leads to a structure that is simple to analyse and design, and can be a low-cost form of construction. A shed-type building is a simple construction consisting of beams (horizontal or sloping), supported at their ends on walls or posts. There may be one or more intermediate supports depending on the width of the building. Purlins running longitudinally support the roof cover. As the principal members are simple or continuous beams (very often timber of rectangular section), the stress-analysis aspect of the design is straightforward. When the beam is supported by timber posts, the design of the posts is not difficult because the load is assumed to be axial. Like the poles in the pole barn, the foundation can consist of a simple pad of concrete beneath the post, or the base of the post can be set into concrete.

B.M.D. (kNm)

Self-weights and service load have been estimated. Continuity over post and brace have been disregarded. This provides a simple safe member. Maximum shear force = 5kN Maximum bending moment= 3120kN/ mm. Try two rafters at 38 200 (back to back)
Maximum shear stress = 3Q 3 5 000 = 2bd 2 76 200

= 0.49 N/mm2 = 0.49 MPa

Maximum bending stress =

M = I Z 3 120 103 6 = = 6 .2 N/mm2 = 6 .2 MPa 76 2002

My

M M bending stress =138y = Z I My M m bending stress = 3 120 = 103 6 = I = 6 .2 N/mm2 = 6 .2 MPa Z 76 2002 3 3 120 10 6 = = 6 .2 N/mm2 = 6 .2 MPa 2 76 of 200 Tables allowable stresses indicate that most hardwoods, but not all softwoods, are adequate.
The load transferred to the outer wall by rafters is a little over 3 kN. Assuming that the strength of the blocks is at least 2.8 MPa (N/mm), the area required is: 3 000 = 1 072 mm2 2.8 3 000 = 1 072 mm2 2.8 As the rafter underside is 76 mm, the minimum interface across the wall is:

Rural structures in the tropics: design and development

joining area than is possible with lapped members. This is often an important factor in nailed and glued joints. Arrangement of members on a single centre line is usually possible with gussets. When full-length timber is not available for a member, a butt joint with cover plates can be used to join two pieces together. This should be avoided, if possible, for the top members (rafters) of a truss and positioned near mid-span for the bottom member (main tie).

1 072 = 14 mm 76 1 072 = 14 mm 76 Hence there is no problem of load transfer to the wall.

Assume posts are 100 mm 100 mm and 2.5 m long, then l / b= 25 and Table 7.2 gives K= 0.38 With c= 5.2 MPa (N/mm) allowable for design, 0.38 5.2N/mm 1002 20kN. Therefore the load is within the safety margin.
Plywood gussets

Cover plates

Connections Timber structure

The methods used to join members include lapped and butt connectors. Bolt and connector joints, nailed joints and glued joints, and sometimes a combination of two types, are examples of lapped connections. Butt connections require the use of plates or gussets. In all cases the joints should be designed by calculating the shear forces that will occur in the members. If two members overlap, the joint is called a singlelap joint. If one is lapped by two other members, i.e. sandwiched between them, it is called a double-lap joint. With a single lap, the joint is under eccentric loading. While for small-span trusses carrying light loads this is not significant, when the joints carry large loads eccentricity should be avoided by the use of double-lap joints. Double members are also used to obtain a satisfactory arrangement of members in the truss as a whole. Sandwich construction enables the necessary sectional area of a member to be obtained by the use of relatively thin timbers, with any double members in compression being blocked apart and fixed in position to provide the necessary stiffness.
Figure 7.8 Butt joints

Bolt and connector joints

Simple bolted joints should only be used for lightly loaded joints because the bearing area at the hole (hole diameter member thickness) and the relatively low bearing stress allowed for the timber compared with that of the steel bolt, may cause the timber hole to elongate and fail. Timber connectors are metal rings or toothed plates used to increase the efficiency of bolted joints. They are embedded half into each of the adjacent members and transmit loads from one to the other. The type most commonly used for light structures is the toothed-plate connector, a mild-steel plate cut and stamped to form triangular teeth projecting on each side that embed in the surfaces of the members on tightening the bolt that passes through the joint. The double-sided toothed connector transmits the load and the bolt is assumed to take no load.

Butt joints
The use of gussets permits members to butt against each other in the same plane, avoids eccentric loading on the joints and, where necessary, provides a greater

Glued joints
Glues made from synthetic resins produce the most efficient form of joint, as strong as or even stronger than the timber joint, and many are immune to attack by dampness and decay. With this type of joint,

Chapter 7 Structural design

139

all contact surfaces must be planed smooth and the necessary pressure applied during setting of the glue. Bolts or nails that act as clamps are often used and left in place. The members may be glued directly to each other using lapped joints, or single-thickness construction may be used by the adoption of gussets. As with nailed joints, lapped members may not provide sufficient gluing area and gussets must then be used to provide the extra area. Glued joints are more often used when trusses are prefabricated because control over temperature, joint fit and clamping pressure is essential. For home use, glue is often used together with nail joints.

Plywood gusset

Figure 7.10 Truss gussets

Figure 7.9 Double-sided toothed plate connector

Nailed joints
Although joining by nails is the least efficient of the three methods mentioned, it is an inexpensive and simple method, and can be improved upon by using glue in combination with the nails. When trusses are prefabricated in factories, nailing plates are often used to connect the member. These fasteners come in two types: 1. A thin-gauge plate called a pierced-plate fastener, which has holes punched regularly over its surface to receive nails. The pierced plate can also be used for on-site fabrication. 2. A heavier plate with teeth punched from the plate and bent up 90 degrees, called a toothed-plate fastener, or connector. This type, in which the teeth are an integral part of the plate, must be driven in by a hydraulic press or roller.

Pierced plate fastener

Toothed plate fastener

Figure 7.11 Nailing plates for truss construction

140

Rural structures in the tropics: design and development

Table 7.6

Minimum nailing distances

ro d1 d1 ro
Loaded edge of member 2

FO RC E

r o

r o

eb

d11

d11

eb

Nailing area

rb

Nailing area x ro d1 d11 rb e0 eb

0
ro Loaded end member 1

5d 5d 5d 5d 5d 5d 5d

5d 5d 5d 5d 5d 5d 5d

10d 10d 10d 10d 10d 10d 10d

5d 5.5d 6d 6.5d 7d 7.5d 8d

8d 8d 8d 8d 8d 8d

15d 15d 15d 15d 15d 15d 15d

10 20 30 40 50 60
E

r o r o

Nailing area rb

ro

Loaded edge Loaded end member 1

d : r0 : d1 : d11 : rb : e0 : eb :

Diameter of the nail (mm). Distance from the extreme row of nails to the unloaded edge of member. Distance between two nails in the nailing area, measured perpendicular to the axis of the member. Distance between two nails measured parallel to the axis of the member. Distance from the extreme row of nails to the loaded edge of the member. Distance from the nearest row of nails to the unloaded end of member. Distance from the nearest row of nails to the loaded end of the member.

FO

RC

Stability
In order to permit the development of the full load at each nail and to avoid splitting the wood, minimum spacing between nails and distances from the edges and ends of the member are necessary. Nailing patterns for use on timber structures are usually available locally. They depend on the quality and type of nails and timber used, and are based on the safe lateral nail load. The Housing Research and Development Unit of the University of Nairobi investigated timber nailed joints made with spacings in accordance with the continental standard for timber joints, which proved to be satisfactory. The main principles are given in tables 7.6 and 7.7. Stability problems in a building are chiefly the result of horizontal loads, such as those resulting from wind pressure, storage of granular products against walls, soil pressure against foundations and sometimes earthquakes. Overturning of foundation walls and foundation piers and pads is counteracted by the width of the footing and the weight of the structure. Only in special cases will it be necessary to give extra support in the form of buttresses. Overturning of external walls is counteracted by the support of perpendicular walls and partitions. Note, however, that not all types of wall, for example framed walls, are adequately rigid along their length without diagonal bracing. If supporting walls are widely spaced and/or the horizontal loads are large, extra support can be supplied by the construction of piers, columns or buttresses. See Chapter 8. Diagonal bracing is used to make framed walls and structures stiff. Long braces should preferably transfer the load with a tensile stress to avoid buckling. Braces

Connections in steel structures

Connections may be bolted, riveted or welded. The principal design considerations are shear, tension and compression, and the calculations are relatively straightforward for the types of design covered.

Chapter 7 Structural design

141

Table 7.7

Basic lateral loads per nail

Continental nail diameter (mm) Kenya nail diameter (mm) Basic lateral nail load (N) 1.8 90 2.0 2.1 2.4 2.65 2.8 3.1 3.35 3.4 3.75 3.8 4.2 4.5 4.6 5.1 5.6 5.6 6.0 6.1

120 140 190 220 250 310 350 370 430 450 540 600 630 750 880 960 1 000

(If the case of pre-bored nail holes 0.8 times nail diameter, the lateral nail load can be increased by 25%)

Connections in single shear at bottom boom of truss

Bolts in shear and tension at ridge of portal frame

Figure 7.12 Connections for steel frames

are usually supplied in pairs, i.e. on both diagonals, so that one will always be in tension independently of the wind direction. If the framed wall is covered with a sheet material, such as plywood, chipboard or metal sheets, the lateral forces on the frame can be counteracted by shear in the sheets. This design requires the sheets to be securely fixed to the frame, both horizontally and vertically. The sheets must be strong enough to resist buckling or failure through shear.

the frames may need extra support from longitudinal bracing. Tension rods are frequently used.

Wind force

Afternate direction of wind force

Braces

Sheet material thoroughly xed to the frame

Figure 7.13 Bracing for portal frame

Masonry and concrete walls that are stiff and capable of resisting lateral wind loading are called shear walls. Portal or rigid frame buildings are normally stable laterally, when the wind pressure acts on the long sides. However, when the wind loads occur at the gable ends,

Post-and-beam or shed-frame buildings will, in most cases, require wind bracing, both along and across the building because there are no rigid connections at the top of the wall to transfer loads across and along the building. The same applies to buildings employing roof trusses. End bracing should be installed.

142

Rural structures in the tropics: design and development

Walls with long spans between the supporting crosswalls, partitions or buttresses tend to bend inwards under the wind load, or outwards if bulk grain or other produce is stored against the wall. At the bottom of the wall this tendency is counteracted by the rigidity of the foundation (designed not to slide) and the support of a floor structure. The top of the wall is given stability by the support of the ceiling or roof structure, or a specially designed wall beam that is securely anchored to the wall. The designer must consider the ability of the building to withstand horizontal loading from any and all directions, without unacceptable deformation.

3. Bearing on ground: The normal pressure between the base of the wall and the soil beneath can cause a bearing failure of the soil, if the ultimate bearing capacity is exceeded. Usually the allowable bearing pressure will be one-third of the ultimate value. Note that the pressure distribution across the base is not constant.

Retaining walls Wall failure

Walls are commonly used to retain soil on sloping sites, water in a pond or bulk products within a storage area. There are several limiting conditions which, if exceeded, can lead to the failure of a retaining wall. Each must be addressed in designing a wall. 1. Overturning: This occurs when the turning moment resulting from lateral forces exceeds that exerted by the self-weight of the wall. The factor of safety against overturning should be at least two.

Bearing pressure

4. Rotational Slip: The wall and a large amount of the retained material rotate about point O if the shear resistance developed along a circular arc is exceeded. The analysis is too complex to include here.

Rotation

Overturning

2. Sliding: The wall will slide if the lateral thrust exceeds the frictional resistance developed between the base of the wall and the soil. The factor of safety against sliding should be about two.

5. Wall material failure: The structure itself must be capable of withstanding the internal stresses set up, that is to say, the stresses must not exceed allowable values. Factors of safety used here depend on the material and the level of the designers knowledge of the loads actually applied. Naturally, both shear and bending must be considered, but the most critical condition is likely to be tension failure of the front facet.

Sliding

Joint failure in blockwork

Chapter 7 Structural design

143

Gravity walls and dams are dependent on the effect of gravity, largely from the self-weight of the wall itself, for stability. Other types of wall rely on a rigid base, combined with a wall designed against bending, to provide an adequate structure.

7H/

90

Tension bending failure

Example 7.13
Design of a gravity wall retaining water
06 m

Pressure exerted by retained material

3m

Liquid pressure
The pressure in a liquid is directly proportional to both the depth and the specific weight of the liquid (w) which is the weight per unit volume, w = g (N/m), where: = density of liquid (kg/m) g= gravitational acceleration (9.81m/s2)
E D

18 m

Consider a mass concrete dam with the crosssection shown, which retains water to 3m depth. Assume: Ground safe bearing capacity: 300 kPa. Coefficient of sliding friction at base: 0.7. Specific weight of concrete: 23kN/m. wH2 9.8 103 32 P= = = 44.1 kN 2 2 1. Find water force P: All calculations per metre length of wall: wH2 9.8 103 32 P= = = 44.1 kN 2 2 9.8 2 wH 103 32 P= = = 44.1 kN 2 2 (acting onemetre up face) W = A speci c weight (0.6 + 1.8) 2. Find= mass length wall: kN 3 of onemetre 23 = 82.8of 2 W = A speci c weight (0 ) .6 + 1 . 8weight speci c W = A = 3 23 = 82.8 kN (0.62+ 1.8) = 3 23 = 82.8 kN 2 action of w: Taking moments of area 3. Find line of about vertical face:

Free surface Liquid H P H/ 3

The pressure at a given depth acts equally in all directions, and the resultant force on a dam or wall face is normal to the face. The pressure from the liquid can be treated as a distributed load with linear variation in a triangular load form, with a centroid two-thirds of the way down the wet face. wH2 P= 2 p = gH = wH (N/m2) and:

P=

2 wH2 acting at a depth of H 3 2

A X + A2 X2 X= 1 1 It should be noted that a wall retaining a material that is A1 + A2 2 saturated H (waterlogged) must resist this liquid pressure (3 0.6 0.3) + ( 0.6 3 1.0) 3 = = 0.65 m in addition to the lateral pressure from the retained A1X1 + A2 X2 1.8 + 1.8 material. X= AA X ++A A X 22 2 X = 1 11 (3 0 . 6 0 . 3 A1 + A2 ) + ( 0.6 3 1.0) = = 0.65 m 1. ++ (3 0.6 1 0..8 3) (8 0.6 3 1.0) = = 0.65 m 1.8 + 1.8

144

b = Pb =

Pc =

W 82.8 = = 46 kPa A 1 1.8

MI Ymax

Rural structures in the tropics: design and development

Hence the self-weight of the wall acts 0.25m to left of the base centre W 82line. .8 Pc = = = 46 kPa A 1 1.8 4. Find the vertical compressive stress on the base:
3 W bd 82 .8 1 1.83 Pc =I == = = 46 kPa = 0 .486 m4 A 12 1 1.8 12

10. Check overturning

Pc =

W 82 .8 = MI = b = Pb = 46 kPa A 1 1.8 Ymax 5. Find the moment about the centre line of the base
MI M = (1 44.1) -b (0.25 =P = 82.8); (clockwise) - (anticlockwise) b M = 23.4 kNm Ymax 1.8 YP = MI = 0.9 m max 6. Find the bending stresses/pressures b = = b Y 2
max

P = 441 kN 1m W = 828 065 115

bd3 1 MI 1.83 where: I = b = P =b = =0 .486 m4 Ymax 12 12 bd3 1 1.83 = = 0 .486 My m4 W + 12 = P 12= I A bd3 1 1.83
12 = 12 = 0 .486 m4

Overturning moment about D= 44.1 1= 44.1 kNm Stabilising moment about D= 82.8 1.15= 95.22kNm Factor of safety against overturning= 94.22 / 44.1= 2.16

I =
I =

The wall is safe against overturning. 11. Check sliding

1.8 = 0.9 m 2 bd3 1 1.83 I = = = 0 .486 m4 12 12 1.8 23 4 0 .486 = = 0.9 m b =Y Pmax = = 12.6 kPa b 1.8 2 0 .9 Y = = 0.9 m
Ymax=
max

P W F=W

7. Find theW actual stresses/pressures M y =P= + I A My 1.8W Ymax = =P = = + 0.9 m I 2A My My W W P = P= =P = + + E= 58.6 kPa (compression) E= 46 A + 12.6 A I = I D= PD= 46 - 12.6= 33.4 kPa (compression)

23.Compression 4 0 .486 (Note: only indicates the resultant P, and b = P b = = 12 .6 kPa .9 W would0 intersect the base line within its middle third). My W =P= + 23A .4 0 .486 8. Compare maximum pressure with allowable bearing I b = P b = 12.6 kPa 23 .4 0 .= 486 0 . 9 capacity : b = P b = = 12.6 kPa 0 .9 Pmax= 58.6 kPa
This is less than the allowable safe bearing capacity of the soil. Hence the wall-soil interface is safe against 23.4 0 .486 b = Pb = bearing failure. = 12.6 kPa 0 .9 9. Compare actual stresses in the wall with allowable values: Maximum stress = 58.6 kPa (compression) and no tensile stress at any point across wall. Hence the wall material is safe.

Frictional resistance= mW mW= 0.7 82.8= 58kN Horizontal thrust= P= 44.1kN As the required factor against sliding is 2, there is a deficiency of (2 44.1) - 58= 30.2kN. Additional anchorage against sliding should be provided.

Example 7.14
Design a circular water tank with the following dimensions/properties: Diameter 5m, depth of water 3m Water weighs 9.8 103N/m Pressure (P) at a depth of 3m

Chapter 7 Structural design

145

For a dry material, the angle of repose is usually equal to the angle of shearing resistance of the material. This angle of shearing resistance is called the angle of internal friction (). The angle of friction is the essential property of a granular material on which Rankines theory is based. This theory enables the lateral pressure to be expressed as a proportion of the vertical pressure, which was shown depend on specific weight and depth only. P(before) H 29to .4 3 3 = = 44.1 kN/m 2 2 In this case, at a depth h the active lateral pressure is given by: P3H 29.4 3 = = 44.1 kN/m 2 P2 =kwh where: k= a constant dependent on the materials involved.

1 sin Although k =there is some friction between the retained 1+ sinwall material and the face, usually this is disregarded, giving a relatively simple relationship for k: k=
P3 = wH = 9.8 103 3 = 29.4 kPa This acts vertically over the entire base; therefore the base should be designed for a uniformly distributed load (UDL) of 29.4 kPa.

1 sin 1 + sin

where: = the angle of friction

pa =

1 sin wH (N/m2 ) 1 + sin

Pressure P3 also acts laterally on the side wall at its where: bottom edge. This pressure decreases linearly to zero at 1 psin total force per metre of wall-face (N) a= pa = wH (N/m2 ) the water surface. 1 + sin1 sin wH2 (N/m length of wall) Pa = 2 5 1 + sin 2 = 577.3 kN Total force on base= P3 AB = 29.4 4

(acting at the centre of the base) Total force on the side per metre of perimeter wall:

Pa= total force per metre of wall face (N) This gives the approximate horizontal resultant force on a vertical wall face when it is retaining material that is level with the top of the wall. If the surface of the retained material is sloping up from the wall at an angle equal to its angle of repose, a modification is required.

P3H 29.4 3 = = 44.1 kN/m run 2 2

(acting onemetre above base)

Pressure exerted by granular materials

Granular materials such as sandy soils, gravelly soils and grain possess the property of internal friction (friction between adjacent grains), but are assumed not to possess the property of cohesion. If a quantity of such material in 1 sin a dry condition is tipped on to k = it will form a conical heap: the shape a flat surface, 1 + sin maintained by this internal friction between grains. The angle of the sloping side is known as the angle of repose.

Example 7.15

1 sin wH2 P = retaining soil Wall 1 + sin 2

SOIL

P=

1 sin 35 18.6 22 kN/m 2 1 + sin 35

Timber beams

pa =

1 sin wH (N/m2 ) 1 + sin

Angle of reponse

Steel posts set in concrete

Pa =

1 sin wH2 146 1 + sin 2

Rural structures in the tropics: design and development

1 sin wH2 P =the wall shown Consider retaining loose sandy soil to 1 + sin 2 a depth of 2 metres. Tables provide angle of friction equal to 35 and specific weight equal to 18.6 kN/m. Assuming a smooth vertical surface and horizontal soil surface, Rankines theory gives: P= 1 sin wH2 1 + sin 2 1 sin 35 18.6 22 kN/m length of wall 2 1 + sin 35

2m

S=4m

P=

P= 10.1 kN/m length of wall.

S 4 If steel posts are placed at 2.5m centres, each post can tan 45 + = tan (45 + 13.5) = 2 1.63 = 3.26 m be approximated to a vertical cantilever beam 2.5 m 2 2 2 1 sin 35 18.6 22 a total distributed load of 10.1 2.5 kN/m P = long, carrying 2 linear variation from zero at the top to 1 += sin 35 kN of PRESSURE 25.25 DIAGRAM a maximum at the base. The steel post and foundation S 4 tan 45 + = tan (45 + 13.5) = 2 1.63 = 3.26 m concrete must be capable of resisting the applied 2 load, 2 2 principally in bending but also in shear. Critical height is:

The timber crossbeams can be analysed as beams simply supported over a span of 2.5 m, S each tancarrying 45 + = 2 equal 2 to the a uniformly distributed load. This load is product of the face area of the beam and the pressure in the soil at a depth indicated by the centroid of the area of the beam face.

4 1 sin 4 S tan (45 = + + .5 tan 13 .m 5) 2 1.63 = 3.26 m 45 == 3 = wH (45 ) + 13 =2 tan 2 1.P 63 .26 2 2 1 + sin 2

pa =

1 sin wh 1 + sin

Design in the same way as the shallow bin because the 1 sin 2 metres. depth ofP grain = is only wH 1 + sin Maximum pressure at the base of the wall:

if beam face is 0.3 m high, h = 2.0 0.15= 1.85 m P = 0.27 18.6 1.85 = 9.29 kN/m2
S tan for a square binon of the sidebeam length S. 45 + Total uniformly distributed load 2 2 = 9.29 0.3 2.5 = 6.97 kN

1 sin 27 7.7 2 = 5.78 kN/m2 1 sin 1+P sin 27 1 sin = wH P= wH 1 + sin 1 + sin = = 1 sin 27 7.7 2 = 5.78 kN/m2 1 + sin 27

5.78 22 or resultant force P = = 11.56 kN/m 1 sin 27 2 1 sin 27 = 2 7.7 2 = 5.78 kN/m2 = 7.7 2 2 = 5.78 kN/m 1 + sinthe 27base of the wall). /3m above 1 + sin (acting 27
2

sin The1maximum bending moment at the centre of the pa = wh span can be determined and the beam section checked. 1 + sin

5.78 2 Note that the design wall is complex if it consists .56the kN/m P= = 11of of a plate 2of uniform thickness, but if the wall is thought of as comprising a number of vertical members cantilevered from the floor, an approach similar to that Example 7.16 2 wall be used. Grain storage bin 5.78 2can 5for .78 the 22 soil-retaining PkN/m = = 11.56 kN/m = 11.56 (The theory given does not apply to deep bins). A P = 2 2 shallow bin can be defined as one with a sidewall height Designing for EARTHQUAKES of less than In areas where earthquakes occur frequently, buildings must be designed to resist the stresses caused by S tremors. While the intensity of tremors can be much tan 45 + for a square bin of side length S. greater in loosely compacted soil than in firm soil or 2 2 solid bedrock, one- and two-storey buildings are at greater risk on firm ground or bedrock because of the Consider a square bin of side length 4 metres retaining shorter resonance periods. shelled maize/corn to a depth of 2 metres. Assume = 27; specific weight is 7.7kN/m.

Chapter 7 Structural design

147

Casualties are most likely to be caused by the collapse of walls causing the roof to fall, and the failure of projecting elements such as parapets, water tanks, non-monolithic chimneys and loose roof coverings. Outbreaks of fire caused by fractures in chimneys or breaks in mains supply lines present an additional hazard. While small buildings with timber frame walls, or a wooden ring beam supported by the posts of a mudand-pole wall, can resist quite violent earthquakes, the following measures will increase the resistance of a large building to collapse from earth tremors: Use a round or rectangular shape for the building. Other shapes such as L T or U should be divided into separate units. To be effective, this separation must be carried down through to the foundation. Avoid large spans, greatly elongated walls, vaultand-dome construction and wall openings in excess of one-third of the total wall area. Construct a continuously reinforced footing that rests on uniform soil at a uniform depth even on sloping ground. Fix the roof securely, either to a continuously reinforced ring beam on top of the walls, or to independent supports, which will not fail even if the walls collapse. Avoid projecting elements, brittle materials and heavy materials on weak supports. Avoid combustible materials near chimneys and power lines. Ductile structures have many joints that can move slightly without failing, e.g. bolted trusses. Such structures have a greater capacity to absorb the energy of earthquake waves. A symmetrical, uniformly distributed ductile framework with the walls securely fixed to the frame is suitable for large buildings. Masonry walls are sensitive to earthquake loads and tend to crack through the joints. It is therefore important to use a good mortar and occasionally reinforcing will be required.

allowable flange and web thickness is 25 mm. If the allowable maximum tensile strength and compressive stress are 125 MPa and 70 MPa respectively. Determine the size of the flange. 8. A short hollow cylindrical column with an internal diameter of 200 mm and external diameter of 250 mm carries a compressive load of 600 kN. Find the maximum permissible eccentricity of the load if (a) the tensile stress in the column must not exceed 15 MPa; (b) the compressive stress must not exceed 76 MPa. 9. Design a section of a trapezoidal masonry retaining wall 10 metres high, to retain earth weighing 16 000 N/m3. The angle of repose for the earth is 25 when the earth surface is horizontal and level with the top of the wall. The weight of the masonry wall is 25 000 N/m3. 10. A reinforced concrete beam is 200 mm wide, has an effective depth of 450 mm and four 20 mm diameter reinforcing bars. If the section has to resist a bending moment of 50 106 N mm, calculate the stresses in steel and concrete. The modular ratio of steel to concrete is equal to 18.

Further reading
Al Nageim, H., Durka, F., Morgan, W. & Williams, D.T. 2010. Structural mechanics: loads, analysis, materials and design of structural elements. 7th edition. London, Pearson Education. Nelson, G.L., Manbeck, H.B. & Meador, N.F. 1988. Light agricultural and industrial structures: analysis and design. AVI Book Co. Prasad, I.B. 2000. A text book of strength of materials. 20th edition. 2B, Nath Market, Nai Sarak, Delhi, Khanna Publishers. Roy, S.K. & Chakrabarty, S. 2009. Fundamentals of structural analysis with computer analysis and applications. Ram Nagar, New Delhi, S. Chand and Co. Ltd. Salvadori, M. & Heller, R. 1986. Structure in architecture: the building of buildings. 3rd edition. Englewood Cliffs, New Jersey, Prentice-Hall. Whitlow, R. 1973. Materials and structures. New York, Longman.

Review questions
1. Define structural design. 2. Briefly describe the structural design process. 3. Why is it important to take into account deflection of structural elements during design phase? 4. Outline factors that influence design of beams. 5. Which measures improve the resistance of buildings to earthquake? 6. Calculate section moduli for a T section, flange 150 mm by 25 mm, web thickness 25 mm and overall depth, 150 mm. 7. A 10 m long T section beam is simply supported, with the flange uppermost, from the right-hand end and at a point 2.5 m from the left-hand end. The beam is to carry a uniformly distributed load of 8 kN/m over the entire length. The