 Open Access
 Total Downloads : 279
 Authors : D. O. Onwuka, C. E. Okere, N. N. Osadebe
 Paper ID : IJERTV3IS10441
 Volume & Issue : Volume 03, Issue 01 (January 2014)
 Published (First Online): 30012014
 ISSN (Online) : 22780181
 Publisher Name : IJERT
 License: This work is licensed under a Creative Commons Attribution 4.0 International License
Poisson’s Ratio of Soilcrete Blocks
D. O. Onwukaa, C. E. Okere b, N. N. Osadebec,
aCivil Engineering, Federal University of Technology, Owerri bCivil Engineering, Federal University of Technology, Owerri cCivil Engineering, University of Nigeria, Nsukka
Abstract
Tropical countries are subject to extreme weather conditions and as such require special building materials to accommodate this. Soilcrete blocks (made with laterite, cement and water) can effectively be used in these regions because of their thermal insulating properties and greater resistance to extreme weather conditions. Poissons ratio, which is required for structural computations, has been ignored to an extent with regards to block moulding technology. In this work, Poissons ratio was determined for the soilcrete blocks produced. Modified regression theory was used to generate a model for prediction of the Poissons ratio of soilcrete blocks. The model was subjected to statistical tests which proved its adequacy.

Introduction
Block can be generally described as a solid mass used in construction. It can be made from a wide variety of materials ranging from binder, water, sand, laterite, coarse aggregates, and clay to admixtures. The constituent materials determine the type of block which includes soilcrete, sandcrete, mud blocks, clay bricks, etc. Soilcrete blocks are made of cement, laterite and water. Sandcrete blocks (made of river sand, cement and water) are used in most places but they are not considered the best for building in tropical countries because of their poor environmental and thermal insulating properties as a result of high degree of porosity. Soilcrete blocks can effectively be used in tropical areas because of their good thermal insulating properties. They are advantageous in hot dry climates where extreme temperature can be moderated inside buildings of compressed stabilized earth blocks [1].
Several researchers have reported that laterite can be used in good quality block production, road and building construction. Boeck et. al., produced cement stabilized laterite blocks using 46% cement [2]. Good laterite blocks were produced from different sites in Kano when laterite was stabilized with 3 to 7% cement [3]. Laterite stabilized with cement was used successfully to produce bricks in Sudan [1]. Aguwa produced laterite cement blocks using 010% cement content by weight of the soil [45]. Alutu and Oghenejobo used 3% to 15% of cement to produce cementstabilised laterite hollow blocks [6]. It is worthy of note here that most of these researchers work revolved around compressive strength of blocks and the cost effectiveness of using laterite in block production. Other properties/characteristics like Poissons ratio have not been handled adequately. Model for prediction of Poissons ratio using mix ratio and vice versa has not been formulated.
Knowledge of Poissons ratio (an elastic constant) is necessary for structural design computations. In all engineering materials, the elongation produced by an axial tensile load in the direction of the force is accompanied by a contraction in any transverse (lateral) direction. The ratio of the lateral contractive strain to axial strain in a material is referred to as Poissons ratio
[7] and it is given as:Âµ = lateral strain/axial strain = l / (1) where l is the lateral strain, is the axial strain and Âµ, the Poissons ratio.
Neville proposed another way of estimating Poissons ratio. It is the ratio of tensile stress at cracking in flexure to compressive stress at cracking in compression specimen [8]. Hence,
Âµ = t / c (2)
Y(n) = izi (n) +
z (n) z (n) (5)
ij
i
j
where Âµ = Poissons ratio
t = tensile stress at cracking in flexure
c = compressive stress at cracking in compression specimen
Ability to predict this vital structural characteristic/ property of blocks is of utmost importance. Modified regression method proposed by Osadebe [9] was used in this work to formulate a model for prediction of
Poissons ratio of soilcrete blocks using specified mix
where 1 i j 4 and n = 1,2,3, 10
Eqn (5) can be put in matrix from as
[Y(n)] = [Z(n) ] {} (6)Rearranging Eqn (6) gives:
{} = [Z (n) ]1 [Y (n)] (7)
i
The actual mix proportions, si(n) and the corresponding fractional portions, z (n) are presented on Tables 1 and 2.
These values of the fractional portions Z(n) were used to
(n)
develop Z(n) matrix and the inverse of Z matrix. The
ratio. The model can also yield all the possible mix ratios for a desired Poissons ratio. The formulation of the regression equation was done from first principles using the socalled absolute volume (mass) as a necessary condition. This principle assumes that the volume (mass) of a mixture is equal to the sum of the absolute volume (mass) of all the constituent components. Osadebe assumed that the response function is continuous and differentiable with respect to its predictors. By making use of Taylors series, the response function could be expanded in the neighborhood of a chosen point. The modified regression theory has been applied successfully with good results by various scholars [914].

Methodology
The materials used for this work are Eagle cement brand of Ordinary Portland Cement, laterite, sourced from Ikeduru LGA, river sand from Otamiri river in Imo State and potable water. All the materials conform to British Standard/ specifications [1517].
Here, analytical and experimental procedures were used in formulating a mathematical model for predicting the Poissons ratio of soilcrete blocks. The model is based on the modified regression theory.

Formulation of model based on modified regression theory
The polynomial equation as given by Osadebe [9] is
Y = 1z1 + 2z2 + 3z3 + 12z1z2 + 13z1z3 + 23z2z3 (3) In general, Eqn (3) is given as:
Y = izi+ij zizj (4)
where 1 i j 3
Eqns (3) and (4) are the optimization model equations. Y is the response function at any point of observation, zi, the predictors are the ratios of the actual portions to the quantity of soilcrete (fractional portions) and i are the coefficients of the optimization model equations.
i
Different points of observation will have different responses with different predictors at constant coefficients. At the nth observation point, Y(n) will correspond with Z (n). That is to say that:
solution of Eqn (7) with known Z(n) matrix and Y(n) matrix from laboratory tests gives the unknown constant coefficients i.
Table 1. Values of actual mix proportions and their corresponding fractional portions for a 3 component mixture
N
S1
S2
S3
RESPONSE
Z1
Z2
Z3
1
0.8
1
8
Y1
0.081633
0.102041
0.816327
2
1
1
12.5
Y2
0.068966
0.068966
0.862069
3
1.28
1
16.67
Y3
0.067546
0.05277
0.879683
4
0.9
1
10.25
Y12
0.074074
0.08305
0.843621
5
1.04
1
12.335
Y13
0.072348
0.069565
0.858087
6
1.14
1
14.585
Y23
0.068161
0.059791
0.872048
S1 = Actual water cement ratio Z1 = Fractional water/cement ratio S2 =Actual cement quantity Z2 = Fractional portion of cement S3 = Actual laterite quantity Z3 = Fractional portion of laterite
Table 2. Z(n) matrix for a 3component mixture
Z1
Z2
Z3
Z1Z2
Z1Z3
Z2Z3
0.081633
0.102041
0.816327
0.00833
0.066639
0.083299
0.068966
0.068966
0.862069
0.004756
0.059453
0.059453
0.067546
0.05277
0.879683
0.003564
0.059419
0.046421
0.074074
0.082305
0.843621
0.006097
0.06249
0.069434
0.072348
0.069565
0.858087
0.005033
0.062081
0.059693
0.068161
0.059791
0.872048
0.004075
0.05944
0.05214

Experimental Investigation
The mix proportions from Table 1 were used to measure out the quantities of water (S1), cement (S2), laterite (S3), for production of soilcrete blocks. A total of twelve mix ratios were used to produce thirty six solid blocks that were cured and tested on the 28th day. Six out of the twelve mix ratios were used as control mix ratios to produce eighteen blocks for the confirmation of the adequacy of the mixture design
model. The initial cracking load in flexure was recorded and used to calculate tensile stress at cracking in flexure.
The initial cracking load in compression specimen was recorded and used to calculate compressive stress at cracking in compression specimen. With these two parameters known, Poissons ratio was calculated using
Eqn (2). Three blocks were tested for each point and the average taken as the Poissons ratio of the point.


Results and Discussions
The experimental values of Poissons ratios of the soilcrete blocks are presented on Table 3 while the replication variances of the test result are presented on Table 4.
Table 3. Experimental values of Poissons ratio of soilcrete blocks
Exp. No
Mix ratios
(w/c: cement: laterite)
Repli Cates
Initial Cracking Load in Flexure
(KN)
Tensile Stress at Cracking in Flexure
t (N/mm2)
Initial Cracking Load in Compression
(KN)
Compressive Stress at Cracking in Flexure
c (N/mm2)
Poissons Ratio
Âµ = t/c
Average Poissons Ratio
Âµ
1
0.8:1:8
A
15.5
0.230
80
1.185
0.194
0.174
B
17.5
0.259
90
1.333
0.194
C
16.0
0.237
120
1.778
0.133
2
1:1:12.5
A
2.5
0.037
20
0.296
0.125
0.135
B
3.5
0.052
25
0.370
0.141
C
2.8
0.041
20
0.296
0.139
3
1.28:1:16.67
A
2.5
0.037
20
0.296
0.125
0.110
B
2.3
0.034
40
0.593
0.057
C
3.0
0.044
20
0.296
0.149
4
0.9:1:10.25
A
3.5
0.052
30
0.444
0.117
0.095
B
3.0
0.044
35
0.518
0.085
C
2.9
0.043
35
0.518
0.083
5
1.04:1:12.335
A
2.1
0.031
30
0.444
0.070
0.074
B
4.0
0.059
50
0.741
0.080
C
2.9
0.043
40
0.593
0.073
6
1.14:1:14.585
A
2.2
0.033
20
0.296
0.111
0.091
B
2.5
0.037
40
0.593
0.062
C
3.0
0.044
30
0.444
0.099
7
1.09:1:13.46
A
2.5
0.037
20
0.296
0.125
0.099
B
2.0
0.030
20
0.296
0.101
C
2.1
0.031
30
0.444
0.070
8
1.02:1:12.417
A
2.1
0.031
30
0.444
0.070
0.105
B
2.1
0.031
20
0.296
0.105
C
2.8
0.041
20
0.296
0.139
9
0.866:1:9.485
A
2.5
0.037
50
0.741
0.050
0.054
B
2.5
0.037
50
0.741
0.050
C
2.8
0.041
45
0.667
0.061
10
1.0924:1:13.8761
A
4.0
0.059
40
0.593
0.099
0.079
B
2.5
0.037
/td>
45
0.667
0.055
C
3.8
0.056
45
0.667
0.084
11
1.052:1:12.818
A
4.0
0.059
30
0.444
0.133
0.092
B
2.5
0.037
30
0.444
0.083
C
3.8
0.031
35
0.518
0.060
12
1.1:1:13.685
A
2.0
0.030
20
0.296
0.101
0.098
B
2.0
0.030
20
0.296
0.101
C
2.3
0.034
25
0.370
0.092
Table 4. Poissons ratio test results and replication variance
Expt.
No.
Replicates
Response Yi
Response
Symbol
Yi
Y
Y 2 i
S 2
i
1
1A
1B
1C
0.194
0.194
0.133
Y1
0.521
0.174
0.093
0.0012
2
2A
2B
2C
0.125
0.141
0.139
Y2
0.405
0.135
0.055
0.000
3
3A
3B
3C
0.125
0.057
0.149
Y3
0.331
0.110
0.041
0.002
4
4A
4B
4C
0.117
0.085
0.083
Y12
0.285
0.095
0.028
0.000
5
5A
5B
5C
0.070
0.080
0.073
Y13
0.223
0.074
0.017
0.000
6
6A
6B
6C
0.111
0.062
0.099
Y23
0.272
0.091
0.026
0.001
Control
7
7A
7B
7C
0.125
0.101
0.070
C1
0.296
0.099
0.031
0.001
8
8A
8B
8C
0.070
0.105
0.139
C2
0.314
0.105
0.035
0.0012
9
9A
9B
9C
0.050
0.050
0.061
C3
0.161
0.054
0.0087
0.000
10
10A
10B
10C
0.099
0.055
0.084
C4
0.238
0.079
0.020
0.001
11
11A
11B
11C
0.133
0.083
0.060
C5
0.276
0.092
0.0282
0.001
12
12A
12B
12C
0.101
0.101
0.092
C6
0.294
0.098
0.0289
0.000
0.0084
Legend: y = y/n
S 2= [1/(n1)]{y 2 [1/n(y )2]} where 1in
y i i
yi = the responses
y = the mean of responses for each control point
n = control points, n1 = degree of freedom
Considering all the design points, number of degrees of freedom,
Ve = (Ni1) (8)
where 1i 12
Ve = 12 1=11
Replication variance,
S 2 = 1/ V S 2 (9)
where is the estimated standard deviation or error,
y e i
y
S 2 = 0.0084/11 = 0.0007636
i
where S 2 is the variance at each point
y
Replication error, Sy = S 2 (10)
= 0.0007636 = 0.028
This replication error value was used below to determine the tstatistics values for the model.

Determination of Osadebes mathematical model for Poissons ratio of soilcrete blocks
Substituting the values of Y(n) from test results (given in Tables 3 and 4) into Eqn (7) gives the values of the coefficients, as:
1 = 8790.55199, 2 = 1164.1268, 3 = 27.3908, 4 = – 16111.1579, 5 = 9787.0468, 6 = 866.3843
Substituting the values of these coefficients, into Eqn
(3) yields:
Y = 8790.55199Z1 + 1164.1268Z2 + 27.3908Z3
16111.1579Z4 9787.0468Z5
866.3843Z6 (11)
Eqn (11) is the Osadebes mathematical model for optimisation of Poissons ratio of soilcrete block based on 28day strength.
3.1.1 Test of adequacy of Osadebes model for Poissons ratio of soilcrete blocks
t is the tstatistics,
n is the number of parallel observations at every point Sy is the replication error
ai and aij are coefficients while i and j are pure components
ai = Xi(2Xi1) aij = 4XiXj
Yobs = Y(observed) = Experimental results Ypre = Y(predicted) = Predicted results
Using Eqns (12), (13), (14), the students ttest computations are presented on Table 5.
Tvalue from table
For a significant level, = 0.05, t/l(ve) = t 0.05/6(5) = t 0.01(5) = 3.365. The tvalue is obtained from standard t statistics table.
This value is greater than any of the tvalues obtained by calculation (as shown in Table 5). Therefore, we accept the Null hypothesis. Hence the model equation is adequate.
(ii) Fisher Test
For this test, the parameter y, is evaluated using the following equation:
y = Y/n (15)
where Y is the response and n the number of responses. The Fisher test computations are presented on Table 6.
Using variance, S2 = [1/(n1)][ (Yy)2] and y = Y/n for 1in, S2 and S2 are calculated as follows:
The model equation was tested for adequacy against the S2
(obs)
(pre)
2
controlled experimental results. The statistical hypothesis for this mathematical model is as follows: Null Hypothesis (H0): There is no significant difference
(obs) = 0.001763/5 = 0.0003526 and S (pre) =
0.000695/5 = 0.000139
The fisher test statistics is given by:
F = S 2/ S 2 (16)
between the experimental and the theoretically 1 2
expected results at an level of 0.5.
where S 2 is the larger of the two variances.
Alternative Hypothesis (H1): There is a significant
difference between the experimental and theoretically
1
1
Hence, S 2
= 0.0003526 and S 2
= 0.000139
2
expected results at an level of 0.05.
The students ttest and fisher test statistics were used for this test. The expected values (Ypredicted) for the test control points were obtained by substituting the values of Z1 from (Table 2) into the model equation i.e. Eqn (11). These values were compared with the experimental result (Yobserved) given in (Table 3).
(i) Students ttest
For this test, the parameters y, and t are evaluated using the following equations respectively
Y = Y(observd) – Y(predicted) (12)
= ( 2 + a 2) (13)
Therefore, F = 0.0003526/ 0.000139 = 2.54
From standard Fisher table, F0.95(5,5) = 5.1 which is higher than the calculated Fvalue. Hence the regression equation is adequate.
i ij
t = yn / (Sy1+ ) (14)
Table 5. Tstatistics test computations for Osadebes Poissons ratio model
N
CN
i
j
i
ij
2
i
ij2
Y(observed)
y(predicted)
Y
t
1
2
0.125
0.25
0.01562
0.0625
1
C1
1
2
3
3
0.125
0.125
0.5
0.5
0.01562
0.01562
0.25
0.25
3
–
0
–
0
–
0.04686
0.5625
0.6094
0.099
0.075
0.024
1.170
Similarly
2
–
–
–
–
–
–
0.6094
0.105
0.079
0.026
1.268
3
–
–
–
–
–
–
0.899
0.054
0.101
0.047
2.110
4
–
–
–
–
–
–
0.8476
0.079
0.096
0.017
0.774
5
–
–
–
–
–
–
0.640
0.092
0.073
0.019
0.918
6
–
–
–
–
–
–
0.6208
0.098
0.078
0.020
0.972
Table 6. Fstatistics test computations for Osadebes Poissons ratio model
Response
Symbol
Y(observed)
Y(predicted)
Y(obs) y(obs)
Y(pre)y(pre)
(Y(obs) y(obs))2
(Y(pre) y(pre))2
C1
0.099
0.075
0.011167
0.00867
0.000125
7.51E05
C2
0.105
0.079
0.017167
0.00467
0.000295
2.18E05
C3
0.054
0.101
0.03383
0.017333
0.001145
0.0003
C4
0.079
0.096
0.00883
0.012333
7.8E05
0.000152
C5
0.092
0.073
0.004167
0.01067
1.74E05
0.000114
C6
0.098
0.078
0.010167
0.00567
0.000103
3.21E05
0.527
0.502
0.001763
0.000695
y(obs)=0.087833
y(pre)=0.083667
Legend: y =Y/n
where Y is the response and n the number of responses.


Conclusion

Poissons ratio of soilcrete blocks was determined

Modified regression theory proposed by Osadebe was used to generate a model for prediction of Poissons ratio of soilcrete blocks

The efficacy of the model was proved using students t test and fisher test

The model can predict the Poissons ratio of soilcrete blocks if the mix ratio is specified and vice versa

References

E.A. Adam, Compressed stabilised earth blocks manufactured in Sudan, A publication for UNESCO, 2001. [Online]. Available from: http://unesdoc.unesco.org.

L. Boeck, K.P.R. Chaudhuri, H.R. Aggarwal, Sandcrete blocks for buildings: A detailed study on mix compositions, strengths and their costs, The Nigerian Engineer, 38 (1), 2000.

R.H. Aggarwal, and D.S. Holmes, Soil for low cost housing: In tropical soils of Nigeria in engineering practice, A.A. Balkema Publishers, 1983.

J.I. Aguwa, Study of compressive strength of laterite cement mixes as building material, AU Journal of Technology, Assumption University of Thailand, 13(2), 2009, 114120.

J.I. Aguwa, Performance of lateritecement blocks as walling units in relation to sandcrete blocks, Leornado Electronic Journal of Practices and Technologies, 9(16), 2010, 189200.

O.E. Alutu, and A.E. Oghenejobo, Strength, durability and cost effectiveness of cementstabilised laterite hollow blocks, Quarterly Journal of Engineering Geology and hydrogeology, 39(1), 2006, 65 72.

D.O. Onwuka, Strength of engineering materials, WEBSmedia Communications, Nigeria, 2001.

A.M. Neville, Properties of concrete. 3rd ed., Pitman, London, 1981.

N.N. Osadebe, Generalised mathematical modelling of compressive strength of normal concrete as a multi variant function of the properties of its constituent components, A paper delivered at the College of Engineering, University of Nigeria, Nsukka, 2003.

O.S. Ogah, A mathematical model for optimisation of strength of concrete: A case study for shear modulus of Rise Husk Ash Concrete, Journal of Industrial Engineering International, 5(9), 2009, 7684.

D.O. Onwuka, C.E. Okere, , J.I . Arimanwa, S.U. Onwuka, Prediction of concrete mix ratios using modified regression theory, Computational methods in Civil Engineering. Vol. 2, No. 1, 2011, pp. 95 107.

C.E. Okere, D.O. Onwuka, N.N. Osadebe, Mathematical model for optimisation of modulus of rupture of concrete using Osadebes regression theory,
Research Inventy: International Journal of Engineering and Science. Vol. 2, issue 5, 2013.

B.O. Mama, and N.N. Osadebe, Comparative analysis of two mathematical models for prediction of compressive strength of sandcrete blocks using alluvial deposits, Nigerian Journal of Technology, Vol.30 (3), 2011.

D.O. Onwuka, , C.E. Okere, S.U. Onwuka, Optimisation of concrete mix cost using modified regression theory, Nigerian Journal of Technology. Vol. 32, No. 2, 2013, pp. 211216.

British Standards Institution, BS 12, Specification for Portland cement, 1992.

British Standard Institution, BS 882, Specification for aggregates from natural sources for concrete, 1992

British Standards Institution, BS EN 1008 Mixing water for concrete Specification for sampling, testing and assessing the suitability of water, including water recovered from processes in the concrete industry, as mixing water for concrete, 2002.