Experimental Study On Limit Of Stone-Powder Of Gravel-Crushed Proto-Machine-Made Sand For Concrete

Natural sand is a local resources, non-renewable for a short time and not suitable for long-distance transport, which make the contradiction between supply and demand caused by sand intensified.The development of new sand source—proto-machine-made sand is imperative. However, limits of stone powder content in proto-machine-made sand in current standard are still disputed and restrict application of proto-machine-made sand.With the gravel-crushed proto-machine-made sand for research object, on the basis of analyzing characteristics of proto-machine-made sand and stone powder, effects of stone powder on properties of concrete are studied and limits of stone powder content in gravel-crushed proto-machine-made sand for concrete are discussed in this dissertation.Main characteristic parameters of proto-machine-made sand and stone powder are statistically investigated through analyzing sample.The investigation reveals the interrelation between grading of proto-machine-made sand and accumulating sieve residue of 0.6mm sieve. Effects of stone powder on cement mortar strength are researched in aspects of macro and micro, stone powder is not entirely inert mineral admixtures, increasing appropriately stone powder content in a certain range can make cement strength improved, especially for early mortar strength.The effects of stone powder on properties of concretes are researched. Stone powder can make the workability of concrete improved. Comparing with the cube compressive strength, the axial compressive strength is influenced more greatly by the increase of stone powder content in a certain range . Comparing with the lower strength grade of concrete ,the higher strength grade of concrete is influenced more greatly by the increase of stone powder content in a certain range. However, in general speaking, the effect of stone powder content on concrete compressive strength so small that is considered unnecessarily in the actual engineering application. Stone powder content has an effect on the tensile strength, especially for the higher strength grade of concrete. The flexural strength has the same basic change rule as the tensile strength according to the content of stone powder. The elastic modulus is reduced with the increasing of the stone powder. According to these effects,suitable contents of stone powder in gravel-crushed proto-machine-made sand for concrete is suggested around 10% .Method for proportion design of concrete with proto-machine-made sand containing micro fines is proposed, the values of practical parameters in formula are obtained for calculating mixed strength of concrete, and the beneficial range of water to cement ratio is proposed. Meanwhile, the reasonable water content, sand ratio and content of stone powder are also suggested.Basing on the research results about prediction of concrete strength and incorporating the principle of Artificial Neural Networks, This dissertation adopts individually self-adaption learning ratio algorithm of additional momentum and L-M algorithm to train networks. After piloting calculation and comparing with forecasted results, L-M algorithm is accepted finally,and concrete srength model of gravel-crushed proto-machine-made sand is established. Comparing with forecasted results of the multivariant lineal regression model,the model can meet engineering needs, which also has others advantages like reasonable structure,quick convergence and high accuracy. Moreover, BP network can control forecast error within 2 percent and avoid the problem that the forecast of concrete strength has biggish deflection.