1015

TESTS ON COMPOSITE PILE CAPPING BEAMS

MAQSOOD AHMED RAJA Research Student

Department of Civil Engineering The City Univer51ty

Northampton Square - London EClV OHB

PHILIP R. S. SPEARE Lecturer

Department of Civil En~ineering The City Universlty

Northampton Square - London ECl V OHB

ABSTRACT

Seven full scale composite beams were tested to destruction to study the effects of uniformly distributed top load on their bending, shear and deflection characteristics. The beams were basically of similar shape with different geometric properties. Experimental results are compared with analysis based on the British and United States' Codes. Flexural and shear design recommendations are compared and conclusions are drawn as to their applicability. The effects of varying important parameters like span and shear reinforcement are studied. Other parameters such as tension reinforcement, relative concrete and brick areas were also varied together to study their combined effect. The degree to which each of the parameters influences strength is discussed and suggestions are made for modifications in the design practices for the beams.

NOTATION

As .... ...... ....... .. ...... ..... .. .. .. ............ ...... area of tension steel

Av,Asv .............. ..... ....... ...... ..... ... ....... area of shear reinforcement

b ... .. ...... ...... .... ........ .. ............ ... .. ... .. .. overall width of the beam

bc .. ... ... .. ........ .. .... ..... .. .................. .... . width of concrete portion in the web of the beam

1016

bv .. ...... ...... ....... ..... .................. .. ........ width of the section in shear

d ......... .. .. .. ............ .. .. ... .. ........ .......... . depth of the beam

db .... .. ........ ... .......... ...... .................. .. depth of brickwork portion of the beam

de ...................................... ................ depth of concrete base

f'e ............... ..... ......... ... ..... .. ....... .. ....... characteristic strength of concrete

fy .............. ...... .. .. .. ....... .... .......... .... .... yield strength of steel

fyv .. ............................................... characteristic shear strength of shear reinforcement

h .... ...... ...... .. ............ .. .. ...... .... .. ... overall height of the beam

L ......... ..... ... .. .... ..... .... ..... ... ... ... ... span

Mu ........ .. .... .. ...... .. ........ .. .............. ultimate moment capacity of the section

sv .. ... .. ........... .. .......................... .. . spacing of shear reinforcement

Ve .... ... .. ........................................ concrete shear stress

Vs ......... ... .... .. ................. .. ............ shear stress in shear reinforcement

Ve .. .. .... .. ..... .. .... .. .... .. .. ...... .. .......... shear capacity of concrete

Vs .... ...... ......... .. .......... .. ............ .. .. shear capacity of shear reinforcement

Vu .... .. .... .. ........ .. ........ .. .. .. .. .......... ultimate shear capacity of the section

W ........ .................................. .. .. .. Totalload on span L

Zt ...... ... .... ........ .. .. .. .... .. ................ section modulus for tensile flexural stress

pw ...... .. .... .. .................. ...... .......... percentage of tensile steel (As/bd)

INTRODUCTION

Composite pile capping beams have been used for low rise buildings over soils of low bearing capacity in the recent past [1] . Tests were conducted at The City University to ascertain their load bearing characteristics.The beams consisted of a concrete base with its central portion projecting vertically upwards in the form of an inverted T-beam. The projection was sandwiched between two walls each being one brick thick. The tensile reinforcement was placed in the lower portion of the concrete base. The shear reinforcement consisted of a standard 8503 wire fabric placed in the concrete infill and spanned the entire length of the beam. Figure 1 shows a typical composite beam.

TEST PROGRAM

Seven specimens were tested to destruction. Six specimens were simply supported. Specimen 7 was continuous over an intermediate support at midspan. A number of equally spaced 30 ton jacks were used to load the beams. The jacks rested on short lengths of I-beams which were placed in a channel section, of low bending stiffness, covering the entire length of the beam. The jacks reacted against an I-beam which was held down with the help of cross members and high yield steel bars reacting against the strong floor. Table 1, based on the notation of Figure 1, shows the cross-sectional dimensions and other properties of the specimens. The value of the section modulus in tension, Zt, has been worked out for the areas transformed to those of brickwork. In each case, the ratio of elastic moduli of the

•

1017

constituent materiais was calculated using the 8ritish codes [6,7) and ali the areas were then transformed to equivalent brickwork areas.

S.No Span

mm

1. 5000

2. 2300

3. 2300

4. 2500

5. 2500

6. 3750

7. 2x2500

h

\t-- b-* b

H

d

. · · 1J. 1--____ -' 5f mm

Figure 1. Cross-Section of a Typical Composite Pile Capping Beam.

TABLEl Geometric and other properties of specimens

b be db de h Shear Tensile

mm mm mm mm mm Mesh Steel

280 75 510 250 760 8503 3Y16

300 95 450 300 750 8503 2Y16+ Y8

300 95 450 300 750 2Y16+Y8

280 75 510 250 760 8503 3Y16

280 75 510 250 760 3Y16

280 75 540 250 790 8503 3Y16

280 75 540 250 790 8503 3Y16

2Y16Top

ZI

mm3

3.27x107

3.24x107

3.21X107

3.27x107

3.25x107

3.41X107

3.67x107

1018

The Electrical strain gauges were connected to shear and tension reinforcement at suitable locations. Deflection gauges were placed at midspan and quarter spans. Demec gauges were fitted on the brickwork across the whole length of the beam. They spaned the vertical mortar joints between bricks and were useful in giving an early warning of any cracks that were not yet visible. Figure 2 shows the general arrangement of the specimen, the loading rig and the instrumentation.

The load was increased in gradual steps from zero to that causing failure. At each stage, the strains were recorded with the help of a microcomputer. Deflections were noted from the dial gauges. Similarly, Demec gauge readings were taken. The cracks were plotted on the wall at each stage. Their trajectories were also noted and this information fed into a main frame computer. A plot could thus be obtained showing progressive cracking at each stage of the loading. The program also had the facility of recording the strains at each stage and printing them at the locations of relevant strain gauges.

Figure 2. Generalloading and instrumentation arrangement.

EFFECTS OF PARAMETRIC VARIATIONS

Some specimens were constructed for the sole purpose of studying the effects of variation of only important parameters like shear fabric and span to depth ratio. Others were designed to evaluate the combined effect of variation of a number of less important parameters. A detailed discussion of the effect of ali these factors on different aspects of structural action follows.

WEB STEEL FABRIC

Specimens 2 and 3 were similar except for the absence of shear mesh in the concrete infill of specimen 3. Another pair of specimens, namely 4 and 5, had the same geometric dimensions. Shear fabric was omitted from the web of specimen 5. Load, cracking and deflection characteristics of the two sets were compared to judge the effect of infill mesh on these aspects.

load Capacity It is significant to note the inability of the infill mesh to improve the load carrying capacity of the composite beam. The ultimate load carried by specimen 2 was 1450kN while that taken by specimen 3

•

1019

was 1900kN. For specimen 2, the load was taken olt at an early stage to prevent distress to the loading rig. Signs of failure were, however, begining to appear at this stage. It could, at best, have taken a load equal to specimen 3 but not more. Similarly, for both specimens 4 and 5, the ultimate failure load was 1900kN. The results were thus quite similar to the first set of specimens and It was concluded that shear mesh in the web did not enhance the strength in the range of geometry covered by these tests.

Deflection There was a significant reduction in the denection of the beam when web reinforcement was added. Figures 3 and 4 show a comparison of deflections of the two sets of specimens.

The effect of shear mesh is clearly visible. The degree to which the presence of mesh affected deflection is quite different in the two cases. In the first case, the ratio of deflection at the same total uniformly distributed load within the elastic range with and without shear mesh Is 1: 1.5 while in the second case it has risen to 1 :5. There was no apparent reason for this dlscrepancy.

1I is obvious lhat lhe steel fabric in lhe web of the beam has a significant contribution in restricling the deflection of lhe beam. This reduction is caused by an even distribution of stresses over the whole area of the beam.

X 100 20

S 2 20

16 ---s;;

16

z:: SpeciIren 3 z: -"" ~12 ~12 u -o Specimen 5 <ti <ti o o -' 8 -'

8

4 4

O O ,

O O 2 4 6 8 10 12 '4 8 12 1'6 20 Deflection (m) Def1ection (nul)

Fig. 3 Midspan Deflection Specimcns 2 and 3 Fig.4 Midspan Deflection Specimens 4 and 5.

Crack Pattern The onset of cracking for both the specimens without shear mesh was similar to those with il. These originated as vertical flexural cracks in midspan regions. For the specimens wilh relnforcement in the infill, lhe cracks were smaller and spread evenly on lhe face of the beam. The specimens wilh plain concrete in lhe infill had concentration of cracks mostly in the midspan region of the beam. The mesh clearly performed an important role in the internai distribution of stress. Thus the cracks were evident outwards and over the supports for the bearns with steel fabric in the web. In the case of the beams wilhout lhis mesh, the cracks were concentrated in the central regions. The crack patterns at failure for ali the beams tested are shown in Figure 5.

1020

b-hr4ff+~fI~Ij'l'-\ I 1 P

DCêlm 1 (mcsh) - 760kN

J~ b'-U~~ K lt i \ L

13C,l :11 2 (m::sh) -1900kN Ilcélm 3(no mesh)-1 900kN

~ J/ibl&"," ~ \~

g- '\\~/l ,/.. \ ' ,tr'/d li I: K\

"r P-.'éun li (Il ~sh ) -1 900kN Ilcam 5(no mesh)-1900kN

;nh",~ II II IJ =::: ~ I (S'·' '11 '~\~ Doam G(mesh) - 800kN

~[z1\~ , eIS' :;~)ú :; 1("1\ \ J?b d Ilcam 7(mcsh) - 2400kN

Figure 5. Crack patterns at failure.

EFFECT OF SPAN

Specimens 1 and 4 were of the same cross-sectional dimensions. Speclmen 1 was 4850mm between centres of supports while the fourth specimen had a span of ?350mm. A comparison of both these specimens was made to study the effect of variation in span on strength as well as deflectlon and crack pattern.

Flexural Strength Flexural cracks were initiated in the longer span at a total load of 296kN and in the shorter span at 400kN while the totalloads at elastic limit were 700kN and 1900kN respectively. From sim pie bending theory, the shorter span would requlre a load of 592kN for a stress equivalent to that produced by 296kN in the longer one. Hence, in comparison with the longer span, the cracklng commences In the short span at a lower load than would be expected (400kN compared with 611 kN), suggestlng that relatively higher stresses develop In the shorter span at an equivalent load. As deeper beams deflect, the load is transferred directly to the support by what is commonly known as 'composite archlng action'. This is a comblned structural action consistlng of bending and arching. The upper part acts as an 'arch' in compression transferring the load directly to the support regions while the lawer part of the beam is in flexure as well as in simple tension slnce it acts as a tie between the twa ends of the arch so formed. The camposite structural action set In at an earty stage in shorter spans. It Increased the tensile stresses in the concrete base which resulted in the formatian af cracks at relatively lower load .

•

1021 •

Shear Strength Shear cracks were witnessed only in specimen 4. A total load of 1900kN was taken by this specimen. No shear cracks were noticed in specimen 1. It withstood a maximum totalload of 888kN. AI this load, some brickwork-concrete separation occurred at the horizontal interface. Some cracks were also noticed next to one of the supports. These were due to a combination of shear and direct compressive stresses.

For longe r spans, usually, shear is not criticaI. With shorter spans, the load is transferred directly over the supports resulting in an increase in compressive stresses. A direct comparison of shear stresses could not be made because the longer specimen failed earlier in f1exure. It was, however, evident that, lor smaller spans, there is a greater concentration of stresses not only over the support regions but also in the 'crown' of the arch in the midspan region. Cracks were witnessed at both these locations.

Deflection The span had a very significant effect on the deflection. For the same uniformly distributed load intensity, deflection is proportional to L 4 within the elastic range. The ratio of length was 2:1 so the ratio 01 deflections based on elastic calculations would be 24

= 16:1 . At the same load intensity, the actual deflections recorded for the shorter span were about 1/25th those of the longer one. For shorter spans, lhe load is translerred directly to the supports resulting in lower bending moments and consequently lower qeflections. Figure 6 shows the load-deflection curves for speclmens 1 and 4. At the same load inlensity, the totalload for specimen 1 was double that for specimen 4.

Crack pattern

X 100kN 20 'J

fspecirren 4 16

12

8

4 Specirren 1

o 'i'-~~~~-..,.~-...-~~ DeElection (rrrn) ~ 10 15 20

Figure 6. Load-deflection curves for specimens 1 and 4

The cracks in the case of the longer specimen were spread evenly over the span. The main difference in the crack pattern was the greater extent to which the cracking penetrated upwards in case of specimen 4. There was also a greater compressive stress over the supports which resulted in tensile cracking over the supports at higher loads.

-

1022

COMPOSITE 'ARCHING' ACTION

Some authors have reported [2, 3, 4] that 'arching' action Is present in beams of depth to span ratio of around 0.6. The beams tested had a depth to span ratio much less than this value but arching action was witnessed when they beams were loaded. From the pattern of cracks, the compression strut could be easily identified in ali cases, but it was more clearly vlsible at higher loads. The angle of inclination of struts formed as a result of this 'arching' action and their distances from the edge of the beam were noted for beam. Table 2 gives the distances of the compression struts from the beam edges and their angles of inclination from the horizontal.

TABLE2 Location and incJination of compressive strut

Specimen 2 3 4 5 6 7

Angle(degrees) 45 54.5 55 45 35 45 35

Distance 1.4d .82d .8d .5d .4d d .7d

Note: Distance from pile is given as a ratio of depth d

Purely diagonal tension cracks would be inclined at an angle of nearly 45°. These cracks were formed as a result of the combined effect of diagonal tension and direct compression over the supports.

MISCELLANEOUS PARAMETERS

The tension steel in specimen 2 was 2/3 that of specimen 4. It did not effect the strength but the additional reinforcement seems to have restricted the deffections. For specimens 3 and 5, different percentages of tension steel made little difference to their deffection.

There was a slight difference in the relative brickwork and concrete areas but this did not have a profound effect on the performance of different specimens. Even when there was a significant difference in concrete and mortar strengths, there was no overall variation in strength, deflection ar cracking characteristics.

In some specimens, the load was removed and reapplied to see the effect of repeated loading. At low loads, the performance of beams did not vary with repeated loading but once the loading was near the elastic limit, then each subsequent cycle of loading praduced greater distress to the beam.

COMPARISON OF VALUES OF SHEAR STRENGTHS - US ANO UK COOES

Concrete and Brickwork Both the codes [5, 7] evaluate the shear strength provided by brickwork (or concrete) and shear reinforcement separately. The American Cade [7] suggests the following expression for calculating the value of characteristic shear strength of concrete.

Vc = 1.9y1f'c + 2500~uWVud (1)

•

1023

Vud/Mu depends on the geometric properties of the beams and not on the extent of loading because the load W cancels out as is evident from the expressions for Vu and Mu. This is logical since the value of shear strength of a beam should not change with a change in loading. The expressions for Vu and Mu are:

Vu = [0.5 - (150+d)/L]W

where the diameter of the supporting pile is 150 mm and

Mu = [O.5(d + 75) - (d + 150)2/(2L)]W

The same will be the case in any other type of loading. As an alternative to equation (1), the code suggests that the value of Vc may be approximated as 2y' f'c. The maximum value in both cases is restricted to 3.5 y' f'c. The values of f'c has to be taken in psi units in ali these cases.

The value suggested by the British code (CP11 O) is 0.35 N/mm2. Consider a very low value of concrete characteristic stress of 30 N/mm2 (4350 psi). For this type of concrete, the US Code would suggest a value of 2 y'4350 = 120 psi or 0.8 N/mm2 and a maximum value of 3.5y' 4350 = 230.8 psi or 1.59 N/mm2

. Thus there is a considerable difference in the characteristic shear stress values suggested by the two codes. There is some sense in relating the shear stress to the characteristic compressive stress since, in shear, it is the diagonal tension or compression which causes failure. The revised British Code BS 8110 (table 3.9) does take into consideration the positive effect of tension steel and the decrease in beam depth but the effect of concrete strength is not taken into account. The range of values for design concrete shear stress is from 0.34 N/mm2 to 1.22 N/mm2. These values still seem conservative and are much less than the design concrete shear stresses suggested by the American Code.

Shear Steel Similar expressions are used by both the codes to evaluate the contribution of web reinforcement towards shear. The British Codes i.e. CP110 and BS 8110, both give the same expression, namely:

Asv = bvsv(v-vc)/(0.87fyv) (2)

In this expression, (v-vc) is the stress in the shear steel, vs. Thus, rearranging the expression in terms ofvs,

vs = Q.87AsyflllL. bvsv

where fyv is not to be taken greater than 0.87fy. Then in terms of totalload and steel yield stress, the expression for shear strength of steel becomes;

Vs = O.75Asvfy'bvd bvsv

or

Vs = Q.7SAsyfyd sv

The equivalent expression suggested by the American Code is;

Vs=~ Sv

(3)

(4)

Thus a factor of safety of 0.75 for the steel shear stress exists in the British Code whereas there is no factor of safety in the American Code. The values of the shear strengths calculated using both

1024

codes were compared with maxlmum load taken by the specimens prior to any slgns of shear cracking and the results are glven In table 3. In calculating the shear strength using the Amerlcan Code, f'c was taken as 21 N/mm2 which was the characteristic strength of brickwork, the weakest of ali the elements.

TABLE3 Comparison of Shear strengths (kN)

Specimen 2 3 4 5 6

US Code 353 367 279 353 264 367

UK Code 136 139 73 136 69 141

Experimental 316 394 699 403 403 292

Note: The experimental values are the rnaximum loads taken by the specimen in shear and not the failure loads.

COMPARISON OF FLEXURAL STRESSES - US ANO UK COOES

Figure 7 shows the equivalent stress blocks and strain dlagrams used for design in the American and British codes. The value of f3 in Figure 7 is 0.85 for characteristic concrete stress of upto 4000 psi and dereases by 0.05 for each additional 1000 psi subject to a minimum of 0.65.

Ecu=0.003 0.8Sfe Ecu=0.003S 0.67feu

Tnse lZ T I: x

1 , , I

I I I I

I , I

I

I

Asfy ,

L-.Asfy Strain Stress Strain Stress

(a) US Code (b) UI< Code

Figure 7: Dcsign strain diagrams and equivalent stress blocks - US and UK codcs .

•

1025 .. The US Code has an equivalent stress block of greater area as well as greater lever arm for

concretes of low compressive strength. Thus, for this type of concrete, the American Code suggests higher flexural strengths than the British Code. As the strength of concrete increases, the depth of the stress block reduces and thus for high strength concretes, the area of equivalent stress block for US Code is smaller than that suggested by the British Code. However, there is also an increase in lever arm z and the net effect is that at higher concrete strengths, both the codes give similar values. The values of flexural strength calculated using both the codes are compared with each other and with the experimental results in table 4.

TABLE4 Comparison of bending strengths (kN)

Specimen 2 3 4 5 6

UKCode 301 503 503 623 623 423

US Code 305 515 515 630 630 429

Experimental 570 767 1630 1843 1700 570

It is evident that the bending strengths calculated on the basis of elastic theory are significantly lower than the experimental results. The ratio of calculated to experimental values is 1:3 for shorter specimens (2, 3, 4 and 5) and 1:2 and 1 :1.5 for (Ionger) specimens 1 and 6 respectively. Composite structural action enhances the flexural load bearing characteristics. This improvement in performance is more evident in shorter beams, where 'arching' occurs more readily than in longer ones where, its effect is slight.

CONTINUOll5 BEAM5

Specimen 7 was tested as a continuous beam of two equal spans of 2500 mm span each. The cross-section was similar to the other beams except that two 16 mm diameter high yield steel bars were placed in the top portion of the beam over the central support to cater for negative bending momento

The crack pattem for this beam is included in Figure 5. The load bearing characteristics of the beam were greatly improved as a result of the continuity of spans. The most severe cracking was witnessed over the central support. The deflection was much lower than the simply supported spans confirming the usual advantages of structural continuity.

CONCLU510N5

The tests proved that the structural performance of beams can be greatly improved by proper combination of brickwork and reinforced concrete. From a study of the experimental results, the following conclusions were drawn;

1. The mesh in the concrete infill does not add significantly to the strength of the beam. In the tests performed, both the flexural and shear characteristics of beams were unaffected by the addition of shear fabric in the infill.

2. The shear mesh helps to reduce cracking and deflections by an even distribution of stresses. The deflections in case of beams with shear mesh were reduced by a factor of 1.5 to 5 when compared to similar specimens without mesh.

1026

3. Composite 'arching' action is present in beams of depth to span ratio much lower than 0.6 although it was more evident in shorter beams than in longer ones. The composite action creates tensile stresses in the base of the beam, which acts as a tie. The distribution of stresses direct to the supports also results in high compressive stresses in these areas which can cause crushing and splitting of brickwork. Tensile stresses were also witnessed in the top central regions ('crown') of shorter beams.

4. A new approach is needed to cater for composite structural action as also the composite interaction of different materiais.

5. Continuous beams are structurally advantageous to simply supported ones. The criticai stresses in this case were noticed over the central support.

REFERENCES 1. ABRAHMIAN, S. (1986) Reinforced pile capping beams. MSc project report, Department of Civil Engineering, The City University, London 1986.

2. DAVIES, S. R. and HENDRY, A. W.(1986) Reinforced masonry beams. PROCEEDINGS OF THE BRITISH MASONRY SOCIETY: NO 1 NOV. 1986. pp 73-76.

3. SMITH, B. S. KHAN, M. A. H. and WICKENS, H. G. (1978) Tests on wall-beam structures. PROCEEDINGS OF THE BRITISH CERAMIC SOCIETY NO 27. December 1978. Load bearing brickwork (6) . pp 289-304.

4. WOOD, R. H. and SIMMS, L. G. (1969) A tentative design method for the composite action of heavily loaded brick panel walls supported on reinforced concrete beams. BUILDING RESEARCH STATION,1969. CURRENT PAPER 26/69.

5. Building Code Requirements for Reinforced Concrete. American Concrete Institute. Committee 318-83. 1983.

6. Code of Practice for the Structural use of Concrete CP110 Part 1. Design, materiais and workmanship. British Standards Institution. 1972.

7. Structural use of Concrete. Code of Practice for design and construction. British Standards Institution. BS 8110 Part 1. 1985.

•