Design of concrete structures-A.H.Nilson 13 thED Chapter 12

Design of concrete structures-A.H.Nilson 13 thED Chapter 12

Frames

ANALYSIS OF INDETERMINATE BEAMS AND FRAMES

s mụch of the concrete as is practical is placed in one single operation, Reinforcing steel is not terminated at the ends of a member but is extended through the joints into adjacent members At construction joints, special care is taken to bond the new con- crete to the old by carefully cleaning the latter, by extending the reinforcement through the joint, and by other means As a result, reinforced concrete structures usually rep- resent monolithic, or continuous, units A load applied at one location causes defor- mation and stress at all other locations Even in precast concrete construction, which resembles steel construction in that individual members are brought to the job site and joined in the field, connections are often designed to provide for the transfer of

‘moment as well as shear and axial load, producing at least partial continuity

‘The effect of continuity is most simply illustrated by a continuous beam, such as shown in Fig, 12.14, With simple spans, such as provided in many types of steel con- struction, only the loaded member CD would deform, and all other members of the structure would remain straight But with continuity from one member to the next through the support regions, as in a reinforced conerete structure, the distortion caused

by a load on one single span is seen to spread to all other spans, although the magni- tude of deformation decreases with increasing distance from the loaded member All members of the six-span structure are subject to curvature, and thus also to bending moment, as a result of loading span CD

Similarly, for the rigid-jointed frame of Fig 12.1b, the distortion caused by a load on the single member GH spreads to all beams and all columns, although, as before, the effect decreases with increasing distance from the load, All members are subject to bending moment, even though they may carry no transverse load,

If horizontal fore used by wind or seismic action, a frame, it deforms as illustrated by Fig 12.1c, Here, too, all members of the frame dis- tort, even though the forces act

| loading A member such as EH,

experience deformations and

points of loading, in contrast to the case of verti even without a directly applied transverse load, associated bending moment

In statically determinate structures, such as simple-span beams, the deflected shape and the moments and shears depend only on the type and magnitude of the loads

375

on the distortion of adjacent, rigidly connected members For a rigid joint such as joint

H in the frame shown in Fig 12.10 or c, all the rotations at the near ends of all mem bers framing into that joint must be the same For a correct design of continuous beams and frames, it is evidently necessary to determine moments, shears, and thrusts considering the effect of continuity at the joints

‘The determination of these internal forces in continuously reinforced concrete structures is usually based on elastic analysis of the structure at factored loads with methods that will be described in Sections 12.2 through 12.5 Such analysis requires knowledge of the cross-sectional dimensions of the members Member dimensions are initially estimated during preliminary design, which is described in Section 12.6 along with guidelines for establishing member proportions For checking the results of more exact analysis, the approximate methods of Section 12.7 are useful For many struc- tures, a full elastic analysis is not justified, and the ACI coefficient method of analysis described in Section 12.8 provides an adequate basis for design moments and shears Before failure, reinforced concrete sections are usually capable of considerable inelastic rotation at nearly constant moment, as was described in Section 6.9 This per-

In addition, the various combinations of factored loads specified in Table 1.2 must be used to determine the load cases that govern member design The subject of load placement will be addressed fi

to decrease rapidly with increasing distance from the load Since bending moments are

proportional to curvatures, the moments in more remote members are correspond-

ingly smaller than those in, or close to, the loaded span However, the loading shown

Frames DESIGN OF CONCRETE STRUCTURES | Chapter 12

in Fig 12.2a does not produce the maximum possible positive moment in CD In fact,

if additional live load were placed on span AB, this span would bend down, BC would bend up, and CD itself would bend down in the same manner, although to a lesser degree, as it is bent by its own load Hence, the positive moment in CD is increased if,

AB and, by the same reasoning, EF are loaded simultaneously By expanding the same reasoning to the other members of the frame, one can easily see that the checkerboard pattern of live load shown in Fig 12.2b produces the largest possible positive moments, not only in CD, but in all loaded spans Hence, two such checkerboard patterns are required to obtain the maximum positive moments in all spans,

In addition to maximum span moments, it is often necessary to investigate min- imum span moments Dead load, acting ans, usually produces only positive span moments, However, live load, placed as in Fig 12.2a, and even more so

in Fig 12.2b, is seen to bend the unloaded spans upward, i.e., to produce negative moments in the span If these negative live load moments are larger than the generally positive dead load moments, a given girder, depending on load position, may be sub- ject at one time to positive span moments and at another to negative span moments It

‘must be designed to withstand both types of moments; i it must be furnished with tensile steel at both top and bottom Thus, the loading of Fig 12.2b, in addition to giv- ing maximum span moments in the loaded spans, gives minimum span moments in the unloaded spans

Maximum negative moments at the supports of the girders are obtained, on the other hand, if loads are placed on the two spans adjacent to the particular support and

in a corresponding pattern on the more remote girders A separate loading scheme of this type is then required for each support for which maximum negative moments are

to be computed

In each column, the largest moments occur at the top or bottom While the load- ing shown in Fig 12.2c results in large moments at the ends of columns CC’ and DD’, the reader can easily be convinced that these moments are further augmented if addi- tional loads are placed as shown in Fig 12.2d

It is seen from this brief discussion that, to calculate the maximum possible moments at all critical points of a frame, live load must be placed in a great variety of different schemes In most practical cases, however, consideration of the relative mag- nitude of effects will permit limitation of analysis to a small number of significant cases

‘The ACI Code requires that structures be designed for a number of load combinations,

as discussed in Section 1.7 Thus, for example, factored load combinations might include (1) dead plus live load, (2) dead plus fluid plus temperature plus live plus soil plus snow load, and (3) three possible combinations that include dead, live, and wind load, with some of the combinations including snow, rain, soil, and roof live load While each of the combinations may be considered as an individual loading condition, experience has shown that the most efficient technique involves separate analyses for each of the basic loads without load factors, that is, a full analysis for unfactored dead load only, separate analyses for the various live load distributions described in Section 12.2a, and separate analyses for each of the other loads (wind, snow, etc.) Once the separate analyses are completed, itis a simple matter to combine the results using the appropriate load factor for each type of load This procedure is most advantageous because, for example, live load may require a load factor of 1.6 for one combination,

Frames

ANALYSIS OF INDETERMINATE BEAMS AND FRAMES 379

a value of 1.0 for another, and a value of 0.5 for yet another Once the forces have been alculated for each combination, the combination of loads that governs for each mem- ber can usually be identified by inspection

SIMPLIFICATIONS IN FRAME ANALYSIS

Considering the complexity of many practical building frames and the need to account for the possibility of alternative loadings, there is evidently a need to simplify This can be done by means of certain approximations that allow the determination of moments with reasonable accuracy while substantially reducing the amount of com- putation

Numerous trial computations have shown that, for building frames with a rea- sonably regular outline, not involving unusual asymmetry of loading or shape, the influence of sidesway caused by vertical loads can be neglected In that case, moments

s are determined with sufficient accuracy by dividing the entire

's of one continuous beam, plus the top and bottom columns framing into that particular beam, Placing the live loads on the beam in the most unfavorable manner permits sufficiently accurate determination

of all beam moments, as well as the moments at the top ends of the bottom columns and the bottom ends of the top columns For this partial structure, the far ends of the columns are considered fixed, except for such first-floor or basement columns where soil and foundation conditions dictate the assumption of hinged ends Such an approach

is explicitly permitted by ACI Code 8.9, which specifies the following for floor and roof members:

2 ‘The arrangement of live load may be limited to combinations of (a) factored dead load

‘on all spans with full factored live load on two adjacent spans, and (b) factored dead Joad on all spans with full factored live load on alternate spans

When investigating the maximum negative moment at any joint, negligible error will result if the joints second removed in each direction are considered to be com- pletely fixed Similarly, in determining maximum or minimum span moments, the joints at the far ends of the adjacent spans may be considered fixed Thus, individual portions of a frame of many members may be investigated separately

Figure 12.3 demonstrates the application of the ACI Code requirements for live load on a three-span subframe The loading in Fig 12.3a results in maximum positive moments in the exterior spans, the minimum positive moment in the center span, and the maximum negative moments at the interior faces of the exterior columns The loading shown in Fig 12.3) results in the maximum positive moment in the center span and minimum positive moments in the exterior spans The loading in Fig 1 results in maximum negative moment at both faces of the interior columns Since the structure is symmetrical, values of moment and shear obtained for the loading shown

in Fig, 12.3c apply to the right side of the structure as well as the left Due to the sim- plicity of this structure, joints away from the spans of interest are not treated as fixed

‘Moments and shears used for design are determined by combining the moment and shear diagrams for the individual load cases to obtain the maximum values along each span length, The resulting envelope moment and shear diagrams are shown in Figs 12.3d and ¢, respectively The moment and shear envelopes (note the range of

Design of Concrete Sietures,Thireowh Indeterminate Beams and Eamet —

tion

380 DESIGN OF CONCRE: Chapter 12

FIGURE 12.3

Subframe loading as required

by ACI Code 8.9: Loading

for (a) maximum positive

‘moments in the exterior

spans, the minimum positive

‘moment in the center span,

and the maximum negative

and minimum positive

moments in the exterior

spans; and (c) maximum

requirements for shear reinforcement

ard to columns, ACI Code 8.8 indic:

In rẻ

1 Columns shall be designed to resist the axial forces from factored loads on all floors

or roof and the maximum moment from factored loads on a single adjacent span of

Frames

ANALYSIS OF INDETERMINATE BEAMS AND FRAMES 381

the floor or roof under consideration The loading condition giving the maximum ratio

‘of moment to axial load shall also be considered

2 In frames or continuous construction, consideration shall be given to the effect of unbalanced floor or roof loads on both exterior and interior columns and of eccentric loading due to other causes

3 In computing moments in columns due to gravity loading, the far ends of columns built integrally with the structure may be considered fixed

4, The resistance to moments at any floor or roof level shall be provided by distributing the moment between columns immediately above and below the given floor in pro- portion to the relative column stiffness and conditions of restraint

Although it is not addressed in the ACI Code, axial loads on columns are usu- ally determined based on the column tributary areas, which are defined based on the midspan of flexural members framing into each column, The axial load from the trib- utary area is used in design, with the exception of first interior columns, which are typ- ically designed for an extra 10 percent axial load to account for the higher shear expected in the flexural members framing into the exterior face of first interior columns The use of this procedure to determine axial loads due to gravity is conser- vative (note that the total vertical load exceeds the factored loads on the structure) and

is adequately close to the values that would be obtained from a more detailed frame analysis

MetHops For Etastic ANALYSIS

Many methods have been developed over the years for the elastic analysis of continu- ous beams and frames The so-called classical methods (Ref 12.1), such as applica- tion of the theorem of three moments, the method of least work (Castigliano’s second theorem), and the general method of consistent deformation, will prove useful only in the analysis of continuous beams having few spans or of very simple frames, For the more complicated cases generally encountered in practice, such methods prove exceedingly tedious, and alternative approaches are preferred,

For many years moment distribution (Ref 12.1) provided the basic analytical tool for the analysis of indeterminate concrete beams and frames, originally with the aid of the slide rule and later with handheld programmable calculators For relatively small problems, moment distribution may still provide the most rapid results, and it is often used in current practice, However, with the widespread availability of comput- ers, manual methods have been replaced largely by matrix analysis, which provides rapid solutions with a high degree of accuracy (Refs, 12.2 and 12.3)

Approximate methods of analysis, based either on careful sketches of the shape

of the deformed structure under load or on moment coefficients, still provide a means for rapid estimation of internal forces and moments (Ref, 12.4) Such estimates are useful in preliminary design and in checking more exact solutions for gross errors that

‘might result from input errors In structures of minor importance, approximations may even provide the basis for final design

In view of the number of excellent texts now available that treat methods of analy- sis (Refs 12.1 t0 12.4 to name just a few), the present discussion will be confined to an evaluation of the usefulness of several of the more important of these, with particular ref- erence to the analysis of reinforced concrete structures Certain idealizations and approx- imations that facilitate the solution in practical cases will be described in more detail

Frames DESIGN OF CONCRETE STRUCTURES | Chapter 12

Moment Distribution

In 1932, Hardy Cross developed the method of moment distribution to solve problems

in frame analysis that involve many unknown joint displacements and rotations For the next three decades, moment distribution provided the standard means in engineer- ing offices for the analysis of indeterminate frames Even now, it serves as the basic analytical tool if computer facilities are not available

In the moment distribution method (Ref 12.1), the fixed-end moments for each member are modified in a series of cycles, each converging on the precise final result,

to account for rotation and translation of the joints The resulting series can be termi- nated whenever one reaches the degree of accuracy required After member end moments are obtained, all member stress resultants can be obtained from the laws of statics,

Ithas been found by comparative analyses that, except in unusual cases, building- frame moments found by modifying fixed-end moments by only two cycles of moment distribution will be sufficiently accurate for design purposes (Ref 12.5)

Matrix Analysis

Use of matrix theory makes it possible to reduce the detailed numerical operations required in the analysis of an indeterminate structure to systematic processes of matrix manipulation that can be performed automatically and rapidly by computer Such methods permit the rapid solution of problems involving large numbers of unknowns Asa consequence, less reliance is placed on special techniques limited to certain types

of problems, and powerful methods of general applicability have emerged, such as the direct stiffness method (Refs 12.2 and 12.3) By such means, an “exact” determina- tion of internal forces throughout an entire building frame can be obtained quickly and

at small expense Three-dimensional frame analysis is possible where required A large number of alternative loadings can be considered, including dynamic loads Some engineering firms prefer to write and maintain their own programs for structural analysis particularly suited to their needs However, most make use of read- ily available programs that can be used for a broad range of problems Input—includ- ing loads, material properties, structural geometry, and member dimensions—is pro- vided by the user, often in an interactive mode Output includes joint displacements and rotations, plus moment, shear, and thrust at critical sections throughout the struc- ture A number of programs are available, e.g., PCA-FRAME (Portland Cement Association, Skokie, Hlinois) and others from numerous private firms Most of these programs perform analysis of two or three-dimensional framed structures subject to static or dynamic loads, shear walls, and other elements in a small fraction of the time formerly required, providing results to a high degree of accuracy Generally, ordinary desktop computers suffice

IDEALIZATION OF THE STRUCTURE

It is seldom possible for the engineer to analyze an actual complex redundant struc- ture, Almost without exception, certain idealizations must be made in devising an ana- lytical model, so that the analysis will be practically possible, Thus, three-dimensional members are represented by straight lines, generally coincident with the actual cen-

It is evident that the usual assumption in frame analysis that the members are prismatic, with constant moment of inertia between centerlines, is not strictly correct

A beam intersecting a column may be prismatic up to the column face, but from that point to the column centerline it has a greatly increased depth, with a moment of iner- tia that could be considered infinite compared with that of the remainder of the span

A similar variation in width and moment of inertia is obtained for the columns Thus,

to be strictly correct, the actual variation in member depth should be considered in the analysis Qualitatively, this would increase beam support moments somewhat and decrease span moments In addition, it is apparent that the critical section for design for negative bending would be at the face of the support, and not at the centerline, since for all practical purposes an unlimited effective depth is obtained in the beam across the width of the support

It will be observed that, in the case of the columns, the moment gradient is not very steep, so that the difference between centerline moment and the moment at the top

or bottom face of the beam is small and can in most cases be disregarded However, the slope of the moment diagram for the beam is usually quite steep in the region of the support, and there will be a substantial difference between the support centerline moment and face moment If the former were used in proportioning the member, an unnecessarily large section would result It is desirable, then, to reduce support moments found by elastic analysis to account for the finite width of the supports

In Fig 12.4, the change in moment between the support centerline and the sup- port face will be equal to the area under the shear diagram between those two point: For knife-edge supports, this shear area is seen to be very nearly equal to V2 Actually, however, the reaction is distributed in some unknown way across the width

of the support This will have the effect of modifying the shear diagram as shown by the dashed line; it has been proposed that the reduced area be taken as equal to Val 3

‘The fact that the reaction is distributed will modify the moment diagram as well as the shear diagram, causing a slight rounding of the negative moment peak, as shown in the figure, and the reduction of Val 3 is properly applied to the moment diagram after the peak has been rounded This will give nearly the same face moment as would be obtained by deducting the amount Val 2 from the peak moment

Another effect is present, however: the modification of the moment diagram due

to the increased moment of inertia of the beam at the column, This effect is similar to

Reduction of negative and

positive moments in a frame,

With this said, there are two other approaches that are often used by structural designers The first is to analyze the structure based on the simple line diagram and to reduce the moment from the column centerline to the face of the support by Val 2

without adjusting for the higher effective stiffness within the thickness width of the col-

umn, The moment diagram, although somewhat less realistic than represented by the

of reinforcing steel from the flexural members framing into the column (usually from two different directions) and from the column itself The somewhat higher percentage

of reinforcement required at midspan usually causes little difficulty in concrete place- ment The second approach involves representing the portion of the “beam” within the width of the column as a rigid link that connects the column centerline with the clear span of the flexural member The portion of the column within the depth of the beam

in also be represented using a rigid link Such a model will produce moment diagrams similar to the lower curve in Fig 12.4, without additional analysis The second approach is both realistic and easy to implement in matrix analysis programs

It should be noted that there are certain conditions of support for which no reduction in negative moment is justified For example, when a continuous beam carried by a girder of approximately the same depth, the negative moment in the beam

at the centerline of the girder should be used to design the negative reinforcing steel

Selection of reasonable values for moments of inertia of beams and columns for use

in the frame analysis is far from a simple matter The design of beams and columns is based on cracked section theory, i.e., on the supposition that tension concrete is inef- fective It might seem, therefore, that moments of inertia to be used should be deter- mined in the same manner, i.e., based on the cracked transformed section, in this way accounting for the effects of cracking and presence of reinforcement Things are not this simple, unfortunately

Consider first the influence of cracking For typical members, the moment of inertia of a cracked beam section is about one-half that of the uncracked gross concrete section, However, the extent of cracking depends on the magnitude of the moments rel- ative to the cracking moment In beams, no flexural cracks would be found near the inflection points Columns, typically, are mostly uncracked, except for those having relatively large eccentricity of loading A fundamental question, 100, is the load level to consider for the analysis Elements that are subject to cracking will have more exten- sive cracks near ultimate load than at service load Compression members will be unaf- fected in this respect Thus, the relative stiffness depends on load level

ssults from the fact that the effective cross section of beams varies along a span In the positive bending region, a beam usually has a T sec- tion For typical T beams, with flange width about 4 to 6 times web width and flange

ss from 0.2 to 04 times the total depth, the gross moment of inertia will be about 2 times that of the rectangular web with width b,„ and depth h, However, in the negative bending region near the supports, the bottom of the section is in compression

‘The T flange is cracked, and the effective cross section is therefore rectangular

‘The amount and arrangement of reinforcement are also influential In beams, if bottom bars are continued through the supports, as is often done, this steel acts compression reinforcement and stiffens the section, In columns, reinforcement ratios are generally much higher than in beams, adding to the stiffnes

Given these complications, it is clear that some simplifications are necessary It

is helpful to note that, in most cases, it is only the ratio of member stiffnesses that influences the final result, not the absolute value of the stiffnesses The stiffness ratios

Frames DESIGN OF CONCRETE STRUCTURES | Chapter 12

may be but little affected by different assumptions in computing moment of inertia if there consistency for all members

In practice, it is generally sufficiently accurate to base stiffness calculations for sis on the gross concrete cross section of the columns In continuous T , cracking will reduce the moment of inertia to about one-half that of the uncracked section Thus, the effect of the flanges and the effect of cracking may nearly incel in the positive bending region In the negative moment regions there are no flanges; however, if bottom bars continue through the supports to serve as compres- sion steel, the added stiffness tends to compensate for lack of compression flange

‘Thus, for beams, generally a constant moment of inertia can be used, based on the rec tangular cross-sectional area byt

ACI Code 8.6.1 states that any set of reasonable assumptions may be used for computing relative stiffnesses, provided that the assumptions adopted are consistent throughout the analysis ACI Commentary R8.6.1 notes that relative values of stiffness are important and that two common assumptions are to use gross E/ values for all members or to use half the gross EI of the beam stem for beams and the gross El for the columns Additional guidance is given in ACI Code 10.11.1, which specifies the section properties to be used for frames subject to sidesway Thirty-five percent of the gross moment of inertia is used for beams and 70 percent for columns This differs from the guidance provided in ACI Commentary 8.6.1 but, except for a factor of 0.70, matches the guidance provided in the earlier discussion

If floor beams are cast monolithically with reinforced concrete walls (frequently the case when first-floor beams are carried on foundation walls), the moment of iner- tia of the wall about an axis parallel to its face may be so large that the beam end could

be considered completely fixed for all practical purposes If the wall is relatively thin

or the beam particularly massive, the moment of inertia of each should be calculated, that of the wall being equal to br°12, where 1 is the wall thickness and Ð the wall width tributary to one beam,

If the outer ends of concrete beams rest on masonry walls, as is sometimes the case, an assumption of zero rotational restraint (i.e., hinged support) is probably clos-

For columns supported on relatively small footings, which in turn rest on com- pressible soil, a hinged end is generally assumed, since such soils offer but little resis- tance to rotation of the footing If, on the other hand, the footings rest on solid rock,

or if a cluster of piles is used with their upper portion encased by a concrete cap, the effect is to provide almost complete fixity for the supported column, and this should

be assumed in the analysis Columns supported by a continuous foundation mat should likewise be assumed fixed at their lower ends

be used rather than relative /-L values

A common situation in beam-and-girder floors and concrete joist floors is illus- trated in Fig 12.5 The sketch shows a beam-and-girder floor system in which longi tudinal beams are placed at the third points of each bay, supported by transverse gird- crs, in addition to the longitudinal beams supported directly by the columns If the transverse girders are quite stiff, it is apparent that the flexural stiffness of all beams

in the width w should be balanced against the stiffness of one set of columns in the longitudinal bent If, on the other hand, the girders have little torsional stiffness, there would be ample justification for making two separate longitudinal analyses, one for the beams supported directly by the columns, in which the rotational resistance of the columns would be considered, and a second for the beams framing into the girders, in which case hinged supports would be assumed In most it would be sufficiently accurate to consider the girders stiff torsionally and to add directly the stiffness of all beams tributary to a single column, This has the added advantage that all longitudinal beams will have the same cross-sectional dimensions and the same reinforcing steel, which will greatly facilitate construction, Plastic redistribution of loads upon over- loading would generally ensure nearly equal restraint moments on all beams before collapse as assumed in design, Torsional moments should not be neglected in design- ing the girders

Frames DESIGN OF CONCRETE STRUCTURES | Chapter 12

on a structure is often dominated by the weight of the slab Obviously, a preliminary estimate of member sizes must be one of the first steps in the analysis Subsequently, with the results of the analysis at hand, members are proportioned, and the resulting dimensions compared with those previously assumed If necessary, the assumed sec- tion properties are modified, and the analysis is repeated Since the procedure may become quite laborious, it is obviously advantageous to make the best possible origi- nal estimate of member sizes, in the hope of avoiding repetition of the analysis

In this connection, it is worth repeating that in the ordinary frame analysis, one

is concerned with relative stiffnesses only, not the absolute stiffnesses, If, in the orig- inal estimate of member sizes, the stiffnesses of all beams and columns are overesti- mated or underestimated by about the same amount, correction of these estimated sizes after the first analysis will have little or no effect Consequently, no revision of the analysis would be required If, on the other hand, a nonuniform error in estimation

is made, and relative stiffnesses differ from assumed values by more than about 30 percent, a new analysis should be made

The experienced designer can estimate member sizes with surprising accuracy, Those with little or no experience must rely on trial calculations or arbitrary rules, modified to suit particular situations In building frames, the depth of one-way slabs (discussed at greater length in Chapter 13) is often controlled by either deflection requirements or the negative moments at the faces of the supporting beams Minimum depth criteria are reflected in Table 13.1, and negative moments at the face of the sup- port can be estimated using coefficients deseribed in Section 12.8, A practical mini- mum thickness of 4 in is often used, except for joist construction meeting the require- ments of ACI Code 8.11 (see Section 18.2d)

Beam sizes are usually governed by the negative moments and the shears at the supports, where their effective section is rectangular Moments can be approximated

by the fixed-end moments for the particular span, or by using the ACT moment coef- ficients (see Section 12.8) In most cases, shears will not differ greatly from simple beam shears Alternatively, many designers prefer to estimate the depth of beams at about 3 in per foot of span, with the width equal to about one-half the depth

For most construction, wide, relatively shallow beams and girders are preferred

to obtain minimum floor depths, and using the same depth for all flexural members allows the use of simple, low-cost forming systems Such designs can significantly reduce forming costs, while incurring only small additional costs for concrete and reinforcing steel It is often wise to check the reinforcement ratio - based on the assumed moments to help maintain overall economy - ~ 0.012 in preliminary design will give - ~ 0.01 in a final design, if a more exact analysis is used, Obviously, mem- ber dimensions are subject to modification, depending on the type and magnitude of the loads, methods of design, and material strength,

Column sizes are governed primarily by axial loads, which can be estimated quickly, although the presence of moments in the columns is cause for some increase

of the area as determined by axial loads For interior columns, in which unbalanced

‘moments will not be large, a 10 percent increase may be sufficient, while for exterior columns, particularly for upper stories, an increase of 50 percent in area may be appro-

in the former

For minimum forming costs, it is highly desirable to use the same column

ns throughout the height of a building This can be accomplished by using rength concrete on the lower stories (for high-rise buildings, this should be strength concrete available) and reducing concrete strength in upper stories, propriate For columns in laterally braced frames, the preliminary design of the lower-story columns may be based on zero eccentricity using 0.80: P, = P, A total reinforcement ratio - , ~ 0,02 should be used for the column with the highest axial load With a value of - , ~" 0.01 for the column with the lowest axial load on higher stories, the column size is maintained, reducing /7 when - „ drops below 1 percent Although ACI Code 10.9.1 limits » , to a range of 1 to 8 percent, the effective minimum value of

«is 0,005 based on ACI Code 10.8.4, which allows the minimum reinforcement to be calculated based on a reduced effective area A,, not less than one-half the total area (this, provision cannot be used in regions of high seismic risk), For columns in lateral load— resisting frames, a subframe may be used to estimate the factored bending moments due to lateral load on the lower-story columns The subframe illustrated in Fig 12.6 consists of the lower two stories in the structure, with the appropriate level of fixity at the base The upper flexural members in the subframe are treated as rigid Factored lat- eral loads are applied to the structure Judicious consideration of factors such as those just discussed, along with simple models, as appropriate, will enable a designer to pro-

‘duce a reasonably accurate preliminary design, which in most cases will permit a sat isfactory analysis to be made on the first trial

an indeterminate structure, it is necessary to estimate the proportions of its members

relative stiffness, upon which the analysis depends These dimen- sions can be obtained on the basis of approximate analysis Also, even with the avail- ability of computers, most engineers find it desirable to make a rough check of

ults, using approximate means, to detect gross errors Further, for structures of minor importance, it is often satisfactory to design on the basis of results obtained by rough calculation, For these reasons, many engineers at some stage in the design process estimate the values of moments, shears, and thrusts at critical locations, using approximate sketches of the structure deflected by its loads

Provided that points of inflection (locations in members at which the bending

moment is zero and there is a reversal of curvature of the elastic curve) can be located

accurately, the stress resultants for a framed structure can usually be found on the bas

of static equilibrium alone Each portion of the structure must be in equilibrium under

the application of its external loads and the internal stress resultants For the fixed-end beam in Fig 12.74, for example, the points of inflection under

uniformly distributed load are known to be located 0.211/ from the ends of the span

FIGURE 12.7

Analysis of fixed-end beam

by locating inflection points

Since the moment at these points is zero, imaginary hinges can be placed there with-

out modifying the member behavior The individual segments between hinges can be analyzed by statics, as shown in Fig 12.70 Starting with the center segment, shears equal to 0.2891 must act at the hinges These, together with the transverse load, pro-

duce a midspan moment of 0.0417w/2, Proceeding next to the outer segments, a down-

ward load is applied at the hinge representing the shear from the center segment This

together with the applied load, produces support moments of 0,0833w/2 Note that, for

this example, since the correct position of the inflection points was known at the start, the resulting moment diagram of Fig 12.7c agrees exactly with the true moment di

am for a fixed-end beam shown in Fig 12.7d, In more practical cases inflection

points must be estimated, and the results obtained will only approxi

‘The use of approximate analysis in determining stress resultants in frames is

illustrated by Fig 12.8 Figure 12.8a shows the geometry and loading of a two-member rigid frame In Fig 12.8h an exaggerated sketch of the probable deflected shape is

nate the true values

Frames DESIGN OF CONCRETE STRUCTURES | Chapter 12

given, together with the estimated location of points of inflection On this basis, the central portion of the girder is analyzed by statics, as shown in Fig 12.84, to obtain girder shears at the inflection points of 7 kips, acting with an axial load P (still not determined) Similarly, the requirements of statics applied to the outer segments of the girder in Fig 12.8¢ and e give vertical shears of I1 and 13 kips at B and C, respec- tively, and end moments of 18 and 30 fi-kips at the same locations Proceeding then

to the upper segment of the column, shown in Fig 12.86 with known axial load of

11 kips and top moment of 18 fi-kips acting, a horizontal shear of 4.5 kips at the inflection point is required for equilibrium, Finally, static analysis of the lower part of the column indicates a requirement of 9 fi-kips moment at A, as shown in Fig 12.8¢

‘The value of P equal to 4.5 kips is obtained by summing horizontal forces at joint B

‘The moment diagram resulting from approximate analysis is shown in Fig 12.8h For comparison, an exact analysis of the frame indicates member end moments

of 8 f-kips at A, 16 f-kips at B, and 28 ft-kips at C The results of the approximate analysis would be satisfactory for design in many cases; if a more exact analysis is to

be made, a valuable check is available on the magnitude of results

A specialization of the approximate method described, known as the portal method, is commonly used to estimate the effects of sidesway due to lateral forces act ing on multistory building frames, For such frames, it is usual to assume that hoi zontal loads are applied at the joints only If this is true, moments in all members vary linearly and, except in hinged members, have opposite signs close to the midpoint of each member

For a simple rectangular portal frame having three members, the shear forces are the same in both legs and are each equal to half the external horizontal load If one of the legs is more rigid than the other, it would require a larger horizontal force to place it horizontally the same amount as the more flexible leg Consequently, the portion

of the total shear resisted by the stiffer column is larger than that of the more flexible column,

In multistory building frames, moments and forces in the girders and columns of each individual story are distributed in substantially the same manner as just discussed for single-story frames The portal method of computing approximate moments, shears, and axial forces from horizontal loads is, therefore, based on the following three simple propositions:

1, The total horizontal shear in all columns of a given story is equal and opposite

to the sum of all horizontal loads acting above that story

2 The horizontal shear is the same in both exterior columns; the horizontal shear

in each interior column is twice that in an exterior column

3 The inflection points of all members, columns and girders, are located midway between joints

Although the last of these propositions is commonly applied to all columns, including those of the bottom floor, the authors prefer to deal with the latter separately, depending on conditions of foundation If the actual conditions are such as practically

to prevent rotation (foundation on rock, massive pile foundations, etc.), the inflection points of the bottom columns are above midpoint and may be assumed to be at a dis- tance 2h-3 from the bottom, If little resistance is offered to rotation, e.g., for relatively small footings on compressible soil, the inflection point is located closer to the bottom and may be assumed to be at a distance /-3 from the bottom, or even lower (With ideal hinges, the inflection point is at the hinge, i., at the very bottom.) Since shears and corresponding moments are largest in the bottom story, a judicious evaluation of