The computer aided research of redispersable polymer powder on rapid repair mortar
The computer aided research of redispersable polymer powder
on rapid repair mortar
Jiangwei Ji1，3，Zhang Ge1，3，Xuxun2，3
（1. Shantou SEZ Longhu SCI.& https://www.sodocs.net/doc/0b17787615.html,pany limited, Shangtou city, Guangtong Province 515041; 2. School of material design and technology of Southwest Science and Technology University, Mianyan city, Sichuan Province,621010; 3. United laboratory of Southwest Science and Technology University- Longhu SCI.& https://www.sodocs.net/doc/0b17787615.html,pany limited, Mianyang city, Sichuan Province621010）
Abstract：The uniform design method was applied in the paper, so the experiment scheme which conformed to the restrained condition within a large range could be designed, the experiment times could be reduced, and the scientific and representative of the compound system could be ensured. According to the conventional physical mechanical test of samples with different ratio in the designed experiment scheme, the relationship of the material composition and the property in the ternary compound system which composed of Portland cement–sulpho-aluminous cement–anhydrate was comprehensively studied, and the best ratio of the inorganic adhesive material was finally determined. The influence of the ratio of the main adhesive material of the repairing mortar on the strength and shrinkage property was studied on nonlinear regression analysis by SPSS (Statistical Package for the Social Science), the regression equation of the strength and shrinkage of the Portland cement, Sulpho-aluminous cement and anhydrate was built, and the quantitative study of the influence factor was realized. In the end, the proportion and the experiment was studied on the property of the polymer modified repairing mortar. The result of the uniform design satisfied the practical application according to the test, and the utilization was widespread.
Key Words: ternary compound system, redispersable polymer powder, the uniform design, SPSS soft ware, nonlinear regression analysis
1、preface
As the rapid development of construction in China, the traffic of the road was getting more and more busy, so did the airport runway. Although the repairing of the road and runway started, the traffic cannot be terminated. Because of the restriction, the repairing of the concrete road seems to be hard. So the rapid repairing method was applied to avoid the contradiction of the traffic.
The earliest method to repair the cement concrete road was to remove
the damaged concrete, and cover the road with the same or higher grade cement as the original design. The biggest disadvantage was that the strength development is slow, and the maintenance period and the termination of the traffic is long.In spite of that, the application of the ordinary cement concrete in thin-layer repair may not be compatible. The main problems are as follow: ①the shrinkage of the new pouring concrete was obviously heavier than the old one, so it is easy to get shrinkage-cracking; ②the viscosity of the cement concrete is low itself, so the bond strength with the old concrete is poor;
③the disadvantage of the interface between the new and old concrete make the concrete easy to have fatigue cracks.
For a long time, highway department apply asphalt concrete to repair the damaged cement concrete road, although the method is easy, and the termination is less time consuming, it still have a lot of shortcomings. ①the stiffness is quite different, causing different load transfer; ②if the adhesion between the base and the surface is not appropriate, and the reflection crack does not deal properly, the service life will be very short; ③the evenness and using function of the road is poor, and the color of it is not uniform. ④but we still use asphalt concrete to repair or recapping the concrete road at present if the damaged road surface is large.
2 the main theory of the rapid repairing mortar
In this paper, we make the mixture which is composed of sulpho-aluminous cement, Portland cement and anhydrite in a proper proportion to hydrate. After the hydration, the ettringite that has the expansive characteristic will be formed[1]. At the same time, the repair mortar with rapid set, rapid hardening and low shrinkage performance can be produced in order to shorten the construction time, so the road can be put into service quickly. With the quality of rapid solidification, super early strength and good performance, the rapid repairing material get over the slow solidification and long curing time after the repair of the conventional repairing mortar. It has a wide utilization in repairing the damaged road surface by grouting, maintaining the airport runway, emergency repair for mine tunnel etc. The economic efficiency and the social efficiency are very good.
3 the experiment scheme and result of the basic composition of the ternary compound system
Scientific experiment that can help us recognize and understand the nature
is a main method for scientific research. As an important branch of
mathematical statistics, the appliance of scientific experiment is getting
more and more wide. According to the experiment design, we can arrange
the experiment times properly. We can get more useful information by
applying effective material conditions as much as possible with less
experiment data available to ensure the representative and scientificity of the
data. Designed by Wang Yuan and Fang Kaitai in 1978, the uniform design
is a method which only takes the dots scattered symmetrically within the
experiment range into consideration. The method makes sure how the dots
scatter in the experiment range evenly, so we can get the most information
through less experiment dots.
We had to consider the restraint of which the total percentage of the raw
material was 100%. The orthogonal test can get the best ratio or
technological parameter of the raw material, but the best composition is only
a combination of certain level of different factors, the exact best dot may be
out of those dots. Comparatively, the uniform design is more applicable in
the condition of which the restraints can be ensured to maintain the dots
within the range[2]. The experiment of our subject is the restraint formula
experiment. There are three kinds of raw materials in the compound system,
that is Portland cement, sulfo-aluminous and anhydrite. In our subject, we
want to find out the performance of the ternary compound system to obtain
the best proportion with which the best performance of early strength, high
strength, rapid hardening and smallest shrinkage can be realized. Having
been considered the sulpho-aluminous as the base material, the 1 day
strength and the resistance to the corrosion will be enhanced after the
incorporation of the Portland cement, and ettringite will be more stable.
Suppose that the ratio of sulpho-aluminous is bigger than 40%, anhydrite
0-30% and Portland cement 0-30%. Referred to the uniform design method
and thought, and combined the experiment propose and comparison standard
with the project ,we have designed the experiment scheme by uniform
design soft ware as Figure3-1 and Figure3-2 showed.
table3-1 design of raw material ratio and result of performance test
Number EX1 EX2 EX3 EX4 EX5 EX6 EX7 EX8 SAC 93 87.9 84.4 81.5 79.1 76.9 74.9 73 OPC 3.86 1.18 9.89 3.28 15 5.98 20 9.16 Anhydrite 3.11 10.9 5.7 15.2 5.97 17.2 5.15 17.9 Total 99.97 99.98 99.99 99.98 100.07 100.08 100.05 100.06 Water requirement
30.71% 30.14% 32.14% 28.57% 31.00% 30.00% 30.57% 30.71%
of normal
consistency
Initial setting/min 8 7 9 12 4 7 4 6 Final
setting/min
16 15 18 19 9 13 6 18
Compressive strength
/Mpa 1d 52.20 53.10 53.80 52.45 54.65 51.85 52.05 52.45 3d 61.75 64.30 62.65 60.90 63.50 60.50 61.15 62.35 7d 71.00 69.40 67.95 69.40 70.75 66.85 70.75 66.35 28d 73.38 74.35 76.58 72.53 77.93 74.93 78.58 76.73
Flexural strength /Mpa 1d 3.73 2.60 2.37 3.61 2.20 3.82 4.72 5.04 3d 2.41 2.32 1.45 2.88 1.90 2.09 2.86 3.49 7d 2.58 2.25 2.70 3.07 2.39 3.21 1.78 3.56 28d 1.98 2.51 2.07 5.52 2.29 2.97 2.99 5.75
Shrinkage rate/% 1d 0.113% 0.125% 0.103% 0.140% 0.060% 0.118% 0.101% 0.111% 3d 0.156% 0.152% 0.154% 0.181% 0.101% 0.155% 0.130% 0.139% 7d 0.269% 0.189% 0.189% 0.210% 0.128% 0.183% 0.164% 0.175% 28d 0.238%0.238%0.237% 0.265% 0.167%0.236% 0.210% 0.232%
table3-2 design of raw material ratio and result of performance test Number EX9 EX10 EX11 EX12 EX13 EX14 EX15 EX16 SAC 71.2 69.6 66.5 63.8 61.2 55.3 53.2 51.2 OPC 25.2 12.8 16.7 21 25.7 16.3 20.9 25.7 Anhydrite 3.57 17.7 16.7 15.2 13.2 28.3 25.9 23.1 Total 99.97 100.1 99.9 100 100.1 99.9 100 100 Water requirement
of normal
consistancy
31.00% 31.00% 30.29% 30.29% 31.00% 29.29% 29.43% 29.43% Initial setting/min 8 5 4 7 9 6 9 11 Final setting/min 9 14 6 8 10 9 12 14
Compressive strength/Mpa 1d 47.65 53.45 53.10 46.75 45.50 45.15 41.80 43.60 3d 62.05 68.20 64.00 64.20 59.50 60.10 56.80 57.40 7d 70.15 70.70 69.50 67.60 72.95 64.65 68.50 67.80 28d 73.38 75.45 80.85 80.30 85.60 75.13 79.28 82.30
Flexural strength/Mpa 1d 3.73 4.92 3.63 3.30 4.22 4.88 3.63 3.75 3d 2.51 4.20 4.78 3.47 2.79 3.52 2.81 3.19 7d 1.99 2.74 3.05 1.99 1.64 2.88 1.83 1.99 28d 1.98 6.33 5.00 5.76 4.23 6.21 4.89 4.35
Shrinkage rate/% 1d 0.081% 0.036% 0.083% 0.044% 0.028% 0.008% 0.045% 0.026% 3d 0.108% 0.049% 0.104% 0.061% 0.044% 0.031% 0.064% 0.046% 7d 0.136% 0.075% 0.119% 0.116% 0.071% 0.080% 0.108% 0.112% 28d 0.148% 0.131% 0.184% 0.153% 0.112% 0.124% 0.164% 0.176%
4 treatment and analysis of experiment data
It is complicate to deal with the data from uniform design. So we usually apply multi-regression method to deal with the data, and use software to aid
it. In our experiment, we use SPSS statistical analysis software to finish regression analysis. Because there is interaction between the three materials, we introduce non-linear regression analysis method to fit.
The springboard of uniform design is to build the multivariate optimizing model. The data of the experiment can be dealt with by screening variable, such as stepwise regression. Normally, the quadratic polynomial can be founded by the result of uniform design. If there are m factors X1，X2，……X m , and the observation index is Y, the quadratic index is :
Y=B0+ΣB i X i+ΣB ii X i+ΣB ij X i X j+ε
In the equation: B0、{Bi}、{Bii}、{Bij}—regression parameter,(i ε— random error The quadratic polynomial was built up by the appliance of SPSS software. In the non-linear regression analysis, the parameter and the model fraction had been set. The output of the linear regression data was showed as table 4-1 to table 4-4[4]. Table 4-1 Regression equation of compressive strength Parameter Estimates Parameter Estimate Std. Error 95% Confidence Interval Lower Bound Upper Bound B0 -0.199 1,176.154 -2,878.144 2,877.746 B1 0.717 11.752 -28.039 29.472 B2 -0.025 11.510 -28.188 28.139 B3 -2.179 12.193 -32.014 27.657 b11 0.010 0.000 0.010 0.010 b22 0.480 5.722 -13.522 14.482 b33 0.000 0.000 0.000 0.000 b12 0.010 0.020 -0.040 0.059 b13 0.034 0.012 0.005 0.063 b23 0.076 0.029 0.005 0.147 Table4-2 Regression equation of compressive strength ANOVA(a) Source Sum of Squares df Mean Squares Regression 95,849.642 10 9,584.964 Residual 30.526 6 5.088 Uncorrected Total 95,880.168 16 Corrected Total 198.212 15 Table4-3 Regression equation of shrinkage rate Parameter Estimates Parameter Estimate Std. Error 95% Confidence Interval Lower Bound Upper Bound b0 1.129 23.650 -56.742 58.999 b1 -0.029 0.237 -0.609 0.552 b2 -0.023 0.247 -0.626 0.581 b3 -0.072 0.935 -2.361 2.216 b11 0.926 0.679 -0.735 2.587 b22 0.982 0.334 0.165 1.799 b33 1.018 0.238 0.435 1.600 b12 0.000 0.004 -0.010 0.011 b13 0.000 0.006 -0.015 0.015 b23 0.000 0.005 -0.013 0.014 Table4-4 Regression equation of shrinkage rate ANOVA(a) Source Sum of Squares df Mean Squares Regression 0.591 10 0.059 Residual 0.012 6 0.002 Uncorrected Total 0.603 16 Corrected Total 0.035 15 The analysis table and formula of the variance of compressive strength and shrinkage rate: Decision coefficient=1-residual square sum /correction sum of square sum , we can get two model decision coefficients which are 0.854 and 0.834. With the SPSS software, the compression equation and shrinkage rate equation are: Y1=-0.199+0.717X1-0.025X2-2.179X3+0.01X12+0.48X22+0.01X1X2+0.0 34X1X3+0.076X2X3 Y2=1.129-0.029X1-0.023X2-0.072X3+0.926X12 +0.982X22+1.018X32 Linear Regression Predicted Values com pre ss ion st ren gth c om R -S Figue1 the standard regression predictive dots of the sample of 28 days compression strength From the figure1, there was the linearship between the standard predictive value and the actual value, and the slope is 1. The dots fluctuated from the line, but not vialently. We concluded that the fit between the model equation and objective function was very well. At the same time, the regression model expressed the influence of the objective variant on the factors well. 5 Experiment of rapid repair mortar There are three kinds of cementicious materials in rapid repair mortar with best ratio. Considering the fluidity and strength of the system, the total percentage of the cementicious material was 45%. We applied the best water PCE reducer to better workability and condense the system, so the early strength and high strength could be gained. Besides, the appliance of VINNAPAS 5011 redispersable polymer powder could enhance the cohesive strength between the repair mortar and the interface, and the cohesive to the old concrete would be last long. The basic ratio is as follow: cementicious material 45%, sand 53%, early strength agent0.2%, water reducer 0.15%, retarder 0.05%, enhancer 2%. The result showed as table 4-1. accord with the application requirement. 6 Conclusion （1）The rapid repair mortar with good property can be made by the appliance of ternary compound system composed of Portland cement, sulpho-aluminate and anhydrite, so the excellent social efficiency and the economy efficiency of the road ,such as rapid repair, durability, and rapid open to traffic, will be attained. （2）With the uniform design , the best model can be found quickly, so the experiment times can be reduced and the expenditure can be reduced. （3）With SPSS to fit the uniform experiment data, the regression equation which can fit well can be founded. The influence of model variant on factors will be more clear and quantitative, so it is a good guiding role for the study of real product. reference [1] Sheng Wei, Huang Wenxi, Min Panrong,《cement technology》，Wuhan Technology University Publishing House 2000.8:318-320 [2] Fang Kaitai, Ma Changxing, Design of Orthogonal test and Uniform Design [M].Beijing Science Publishing House.2001:144一152 [3] Fu Xinghua, Hou Wenping. The Study on the Improvement of the Property of Sulpho-aluninate [J].Cement Technology2001.2：10-16 [4] Zhang Wentong.SPSS11 The Statistical Analysis, Beijing Hope Electric Publishing House 2002.6：140-145