| Pump concrete construction technology is one of the most frequently used construction method,the construction quality is good or bad ralations major.But at present the construction in some places on the pumping concrete mix design is not accurate enough,in some special area design of mixture ratio still behave,resulting in project accidents.In addition,pumping agent is not panacea,performance in the special environment of pumping agent will have some changes,these problems are urgently needed to study and solve.This paper in order to solve the above problem is the basic starting point, through establishing the evaluation index of the pump to the advantages and disadvantages of the pump as the evaluation standard to match than elements as parameters to solve with the pump range than the various elements.Orthogonal test method to C50 pump sending concrete as the study object to study the C50 pumping concrete slump, extended, collapsed time, pressure bleeding rate, bulk density, 7d split tensile strength and 28 d splitting tensile strength, 7d compressive strength, 28 d compressive strength etc. index of main influence factors and get the changes in the relationship between various factors and indicators estimated interaction between edge average figure are the key factors, and ultimately determine the C50 pump concrete to meet the strength premise achieve optimal pump mix.In addition, this paper uses the statistical software SPSS to give a complete mathematical analysis of the test results.. In regression analysis by fitting the experimental results, obtained the C50 pump pumping concrete slump, expand degree, collapsed time,pressure bleeding rate, bulk density, 7d splitting tensile strength and 28 d splitting tensile strength, 7d compressive strength, 28 d compressive strength of the linear regression formula, which to solve the similar problems in the future provides a fast and simple research method.When using BP neural network prediction, this paper selects the C50 pump sending concrete 28 d compressive strength as the predicting index, through regression equation is obtained for the 28 d compressive strength as the measured value is predicted, the prediction error is small. This shows that the model is feasible. Therefore, the model can also be used to predict other strength grade of pump sending concrete index, according to the pump sending concrete pumpability evaluation system and can evaluate any strength grade pumping concrete pump. |