J. J. BROO KS and M.A. AMJAD Department of Civil Engineering University of Leeds Leeds LS 2 9JT, U. K. ABSTRACT

Transcription

1 342 ELASTICITY AND STRENGTH OF CLAY BRICKWORK TEST UNITS J. J. BROO KS and M.A. AMJAD Department of Civil Engineering University of Leeds Leeds LS 2 9JT, U. K. ABSTRACT Compressive strength and modulus of elasticity of various sizes of single leaf walls and piers, constructed from Class B engineering clay bricks and a 1:Y,:4Y, mortar, have been determined. In addition, strength and modulus data have been obtained for brick units, brick core samples and mortar units. lt appears that both strength and modulus of brickwork are independent of concrete platen restraint for height/width ratios up to The accuracy of predicting modulus of brickwork is improved when moduli of brick and mortar are incorporated in composite models. INTRODUCTlON The influence of test machine plate~ restraint on the compressive strength of masonry units is not clear. Platen restraint can cause an apparent increase in strength when the height/width ratio of the test unit is less than 2, but t he types of platen, brick or block and mortar are relevant factors. When the height/width ratio exceeds about 6, the strength of masonry decreases due to the slenderness effect. Platen restraint may affect modulus of elasticity and general stress- strain behaviour as well as strength. Clarification of such effects would be beneficiai when comparing reported test data and when translating resul t s of small units into s t rength etc. of full size masonry members. In the design of masonry, modulus of elasticity is estimated from empirical functions of brick strength (1-3). Frequently, such estimates are not very accurate probably because the confluence of mortar type is not taken into account. AIso, such relationships are restricted to certain types of brick, whereas it would be desirable to have a universal method which is applicable to any types of brick or block and mortar.

2 343 Such an approach is feasible by composite modelling of the two main phases of masonry. Several authors have proposed such models (4-8), but these have not bee n verified in sufficient detail, possibly because of the inconvenience of measuring the moduli of brick and mortar units. The overall aims of thi s investigation are to investigate the stressstrain behaviour through to failure of various masonry units made from a range of bricks, blocks and mortar, and to check the validity o f the composite model approach for predicting the modulus of elasticity of masonry. The present paper presents the initial findings of the research, and is concerned wi t h compressive strength and modulus of elasticity of various sizes of brickwork made from the types of clay brick and mortar. EXPERIMENTAL DETAILS Eleven different sizes of brickwork were constructed in stretcher bond from a Class B engineering perforated clay brick and a 1:Y,:4Y, mortar ; the maximum height of brickwork was 13 courses. For each size, two units were built on reinforced concrete slabs and cured under polythene sheeting for approximately 25 days. The t op surfaces of each unit were then capped with reinforced concrete slabs using high alumina cement mortar. One day before testing at the age o f 28 days, the units were positioned in the test machine and the jack header plate was bedded and levelled by Fondu cement ; Oemec strain and LVOT attachements were then positioned and fixed. One of the aims of the project was to investigate the distribution of strain over each unit by means of Oemec strain gauges. Since the time required for readings at each level of load was lengthy, the effects of creep were minimised by adopting the procedure used for determining the static modulus of elasticity of concrete (BS 1881 : Part 121: 1983), viz. by load cycling the unit twice before taking strain readings at stresses of 10, 20 and 30% of the estimated strength. On completion of the Demec strain measurements, the strain through to failure was recorded on an X-Y plotter using LVOTs. The campressive strength and madulus af elasticity af brick and martar units were alsa determined at the age af 28 days. Campressive strength was measured an full size bricks between bed faces, accarding

3 344 to BS 3921: 1974, and between headers. For the modulus of elasticity of brick five different tests were adopted to yield data for the composite models : (a) s ingle brick between b ed faces ; (b) three- stack unbonded brick between bed faces; (c) five- stack unbonded brick between bed faces; (d) single brick between headers ; (e) 50 x 25 mm dia. brick cores between ali three faces. In ali these tests, units were tested in a dry state in order to attach electrical resistance strain gauges, and in tests (b) and (c) the same brick was used as in test (a). The compressive strength of 75 mm mortar cubes was determined according to BS 4551: 1980, while 100 x 100 x 250 mm mortar prisms were used for both modulus of elasticity and compressive strength. DlSCUSSION OF RESULTS Table 1 lists compressive strengths and moduli of elasticity of brickwork and mortar units ; generally, the modulus was determined from the linear part of the stress- strain wave, the limit of proportionality being approximately 30% of the strength. TABLE 1 Test results Test Masonry Height No. size* widthl ratio Masonry strength (MPa) Masonry elasti c modulus (GPa) Mortar cube strength (MPa) Mortar elastic modulus (GPa) 1 3 x lp x lp x lp x lp x 2P x 2P x lp x lp x lp x lp x lw x IVI x lw x lw x 2W x 2\-)

4 345 Table 1 cont'd Test Masonry Height Masonry Maso:1ry Mortal' cube Mortal' No. size* width/ strength elastic strength elastic ratio modulus modulus (MPa) (GPa) (MPa) (GPa) 9 7 x lw x lw x 2W x 2W x 4VJ x 4W NB * number of brieks high x number of bricks wide ; P ~ pier, W = wal1. There was a very large variation in mortar eube strength although mixes had similar eonsistenee; the mean and standard devi ation was 8. 9 ± 4. 3 MPa. For the range of sizes of brickwork investigated, there appears to be no significant influe nce of height/ least lateral dimension ratio on either eompressi ve strength or modulus of elast.i city (see Figs. 1 a nd 2), <U n. ~ ;:: +' Dl) e (!) l. +' Ul (!) >.-< Ul Ui (!) l. c. E o U ao r , 30 95% C.L. o O o r"lean _ s!' ~ J... :0 Ü ~li----2~----~----~---~~---~ Figure 1. Height/l east late,'al cimens lon ratlo Compressive strength o f c1ay brickwork as a function of heighv width ratio although the large variation in mortal' strength could have been a f actor. Consequently, i t may be inferred tha t there is little effect of concrete platen res t raint on strength and e1astieity of briekwork units used in laboratory tests. Confirmation of this finding is difficult as other

5 - 346 C\l ~ o 20, _ 0-8 _.Jl ~ Q..,. L...: e_i _e 8 Mean _... ee 8 2 '0 5 % 5:..:L.>, -'-'.,.., ().,..,.j.j ljj C\l rl Ql 10 e Figure 2. "-' o ljj ;:l rl ;:l 'O o "'" 5 I oi e Pier 'ijall Height/least lateral dimension ratio Modulus of elasticity of clay bri ckwork as a function of height/width ratio investigators' results are sometimes contradictory as far as strength is concerned, and i t should be remembered that the type of platen may be an important factor. However, the non- influence of geometry on modulus of elasticity agrees with that of Lenczner (9). The compressive strengths of the single brick units and core samples are shown in Table 2. There is some anisotropy of strength as indicated TABLE 2 Strength (MPa) of clay brick units and cores Parameter Bed face Header face Stretcher face brick* core brick core brick core f 1ean Standard deviation Nurnber of I samples *BS 3921 : 1974 testo by the cores, the order of strength being bed face> header face> stretcher face. The presence of perforations clearly affected the strength of

6 347 full-size bricks, especially between headers for which the strength was only 14 per cent of that between bed faces. e l asticity data are given in Table 3; Corresponding modulus of the moduli of cores follows the TABLE 3 Modulus of elasticity (GPa) of brick units and cores Parameter Bed face Header face St retcher face l-brick* 3-brick 5-brick core brick core brick core Mean Standard deviation Number of samples I I *BS 3921: 1974 test same pattern as that of strength, i.e. there is some degree of anisotropy. For t he full size bricks, there were difficulties in determining t he modulus of the middle brick in three- and five - stack unbonded tests b e - cause the stress- strain curve was non- linear. The reasons were attribut ed to the uneveness and incomplete contact area of the bed faces, despite careful preparation by grinding beforehand. The moduli of these tests (secant modulus of 3 MPa) were greater than that of either the single brick test or the core test, the moduli of the latter tests being simi lar. This finding is encouraging since it implies that modulus can be relate d t o 'standard' brick strength testo The modulus of elasticity of mortar (see Table 1) was 6.5 GPa with a standard deviation at 2.2 GPa. The values were obtained from prisms whose strengths were the same as the mortar cubes, a result which might s eem surprising because of the platen effect but the prisms were tested in a dry state which has the effect of increasing strength. As before, this finding is encouraging because there is a possiblity of relating the modulus of mortar directly with the 'standard' cube strength. The average experimental data of this investigation were used to compare the various methods of predicting elasticity, as prcposed by

7 different investigators. For those methods based on brick unit strength, the accuracies were: Plowman (1), +28%, Sinha and Hendry (2) -24% and Lenczner (3) +41%. Using the composite model approach, an improvement in accuracy is achieved, viz. Ameny et alo (4) - 15%, Ameny et alo (5) +6%, Jessop et ai. (6) +18%, Sahlin (7) +12%. The model used by Brooks (8) was developed for general application as follows: b y. e ~ + r1y. (e + 1). Ewy Eby ~ + Em Am H E m where Ewy b y e H A w Ab A m E by Em ~ modulus of masonry under axial loading; height of brick or block unit; number of courses; height of masonry; cross- sectional area of masonry; cross- sectional area of bricks or blocks; cross-sectional area of vertical mortar joints; modulus of bricks or blocks (between bed faces); modulus of mortar; thickness of mortar bed joints The above expression yielded a precicted modulus to within 1% of the average measured modulus. CONCLUSI ONS Based on the initial findings of this investigation, the conclusions are: 1. For the range of clay brickwork geometries of this investigation, compressive strength and modulus elasticity of brickwork are independent of concrete platen restraint, as measured by the height/least lateral dimension ratios between 1 and Tests on core samples of brick unit showed anisotropy. Strength and modulus decreased in the order testing between bed faces, h3ader faces and stretcher faces. The presence of perforations greatly affects the anisotropy of full- size brick units. 3. The modulus measured by the 'standard' strength test for brick units is similar to the 'unrestrained' modulus as given by the core sample testo

8 Composite models yield more accurate estimates of modulus of brickwork than empirical equations based on brick unit strength. REFERENCES 1. Plowman, J.M., The modulus of elasticity of brickwork. Pr oc. Brit. Ceram. Soc., 1965, 4, pp Sinha, B.? and Hendry, A. ~., The effect o f brickwork bond on the l oadbearing capacity o f model brick wa l ls. Proc. Brit. Ceram. Soc., 1968, 11, pp Lenczner, O., The effect of strength and geometry on the elastic and creep properties o f masonry members. Proc. Amer. Masonry Conf., Boulder, Co l o rado, 1978, pp Ameny, P., Loove, R. E. and Jessop, E. L., Strength, elastic and creep properties of concrete. lnt. J. Mas. Constr., 1980, 1, 1, pp Ameny, P., Loove, R. E. and Shrive, N., Prediction of elastic behaviour of masonry. lnt. J. ~l as. Constr., 1983, 3, 1, pp Jessop, E. J., Shrive, N. G. and England, G. L., Elastic and creep properties of masonry. Proc. N. Amer. Mas. Conf., Boulder, Colorado, 1978, pp Sahlim, S., Structura1 Masonry, Prentice - Hall Inc., New Jersey, Brooks, J. J., Composite models for predicting e1astic and 1ong- term movements in b r ickwork wa11s. Proc. Brit. Mas. Soc., 1986, 1, pp Le nczner, O., Brickwork - guide to creep. Structura1 C1ay Products Ltd., SCP17, 1977, 26 pp. ACKNOWLEDGEMENT The authors acknowledge George Armi tage & Sons PLC who supp1ied the bri cks.