بهينه سازي طرح اختلاط بتن خودتراكم اليافي حاوي پلي پروپلين با استفاده از الگوريتم هاي فرا ابتكاري ژنتيك و جستجوي كلاغ

Optimization of Mixture Proportions of Self-compacted Fiber Reinforced Concrete incorporating Polypropylene using Genetic and Crow search Algorithms

اﻳﻦ ﻣﻄﺎﻟﻌﻪ ﺑﻪ ﺗﺤﻘﻴﻖ در ارﺗﺒﺎط ﺑﺎ ﻃﺮﺣﻲ ﺑﻬﻴﻨﻪ ﺑﺘﻦ ﺧﻮدﺗﺮاﻛﻢ اﻟﻴﺎﻓﻲ ﺣﺎوي ﭘﻠﻲ ﭘﺮوﭘﻴﻠﻦ ﺑﺮاي ﺳﺎﺧﺖ در ﺻﻨﻌﺖ ﺳﺎﺧﺘﻤﺎن ﻣﻲ ﭘﺮدازد. ﺑﺘﻦ ﻣﺎده اي ﺗﺮﻛﻴﺒﻲ اﺳﺖ ﻛﻪ از ﻣﻮاد ﺳﻴﻤﺎﻧﻲ ، آب، ﺳﻨﮕﺪاﻧﻪ و ﻣﻮاد اﻓﺰودﻧﻲ ﺗﺸﻜﻴﻞ ﺷﺪه اﺳﺖ ﻛﻪ ﻣﺠﻤﻮﻋﻪ اﻳﻦ ﻣﻮاد ﻋﻨﺎﺻﺮ اﺻﻠﻲ ﺗﺸﻜﻴﻞ دﻫﻨﺪه ﺑﺘﻦ ﻣﻲ ﺑﺎﺷﻨﺪ. ﺑﺎ ﺗﻮﺳﻌﻪ اﺳﺘﻔﺎده از اﻳﻦ ﻣﻮاد در ﻣﻬﻨﺪﺳﻲ ﺳﺎﺧﺖ وﺳﺎزﻣﺨﻠﻮطﻫﺎي ﺑﺘﻨﻲ ﺑﻴﺸﺘﺮﻳﻦ ﻣﻮارد اﺳﺘﻔﺎده را در ﺻﻨﻌﺖ ﺳﺎﺧﺖ و ﺳﺎز در ﺳﺮﺗﺎﺳﺮ دﻧﻴﺎ داﺷﺘﻪ اﺳﺖ. ﻣﻘﺎوﻣﺖ ﻓﺸﺎري ﻣﻌﻢ ﺗﺮﻳﻦ ﻣﺸﺨﺼﻪ در ﺑﺘﻦ ﻣﻲ ﺑﺎﺷﺪ. ﺑﺎ اﻳﻦ ﺣﺎل ﻧﻘﺎﻳﺼﻲ در ﺑﺘﻦ ﻣﺎﻧﻨﺪ ﺗﺮك در ﺑﺘﻦ، ﭼﻐﺮﻣﮕﻲ ﭘﺎﻳﻴﻦ و ﻣﻘﺎوﻣﺖ ﻛﺸﺸﻲ وﺟﻮد داﺷﺘﻪ ﻛﻪ ﻧﻴﺎز ﺑﻪ ﺗﻘﻮﻳﺖ ﺑﺘﻦ در ﺑﺮاﺑﺮ اﻳﻦ ﻋﻴﺐ را ﺑﺮﺟﺴﺘﻪ ﻛﺮده اﺳﺖ. در ﺳﺎﻟﻴﺎن اﺧﻴﺮ اﺳﺘﻔﺎده از ﺑﺘﻦ ﻣﺤﺘﻮي اﻟﻴﺎف ﺑﻌﻨﻮان ﻳﻚ ﻣﺎدهي ﺳﺎﺧﺘﻤﺎﻧﻲ ﻣﻬﻢ و ﺑﺎ ﺧﻮاص ﻣﻜﺎﻧﻴﻜﻲ ﻣﻨﺎﺳﺐ ﺟﻬﺖ ﺳﺎﺧﺖ و ﺳﺎز اﻧﻮاع ﺳﺎزهﻫﺎ اﺳﺘﻔﺎده ﻣﻲﺷﻮد. ﻫﺪف اﺻﻠﻲ در اﻳﻦ ﺗﺤﻘﻴﻖ، ﻃﺮاﺣﻲ ﺑﺘﻦ ﺗﻮاﻧﻤﻨﺪ ﺧﻮدﺗﺮاﻛﻢ اﻟﻴﺎﻓﻲ ﺑﺎ اﺳﺘﻔﺎده از اﻟﮕﻮرﻳﺘﻢﻫﺎي ﻓﺮا اﺑﺘﻜﺎري ﺑﺎ ﭘﻴﺎده ﺳﺎزي در ﻧﺮم اﻓﺰار ﻣﺘﻠﺐ ﻣﻲ ﺑﺎﺷﺪ. ﺑﺮاي ﺑﻬﻴﻨﻪ ﺳﺎزي ﻣﺒﺘﻨﻲ ﺑﺮ راﻫﻜﺎرﻫﺎي ﻓﺮا اﺑﺘﻜﺎري، اﻟﮕﻮرﻳﺘﻢ ﺟﺴﺘﺠﻮي ﻛﻼغ )CSA( و اﻟﮕﻮرﻳﺘﻢ ژﻧﺘﻴﻚ )GA( ﺑﻪ ﻋﻨﻮان راﻫﻜﺎرﭘﺮدازﺷﻲ ﻣﺤﺎﺳﺒﺎﺗﻲ ﺗﻮﺳﻌﻪ داده داده ﺷﺪه اﺳﺖ. ﺑﺮاي اﻳﻦ ﻣﻨﻈﻮر، 67 ﻃﺮح اﺧﺘﻼط ﺑﺘﻦ ﺧﻮدﺗﺮاﻛﻢ اﻟﻴﺎﻓﻲ ﺷﺎﻣﻞ آب 137/2 – 195 ﻛﻴﻠﻮﮔﺮم ﺑﺮ ﻣﺘﺮﻣﻜﻌﺐ(، ﺳﻴﻤﺎن 325/5 – 520 ﻛﻴﻠﻮﮔﺮم ﺑﺮ ﻣﺘﺮﻣﻜﻌﺐ(، درﺷﺖ داﻧﻪ 722 – 920 ﻛﻴﻠﻮﮔﺮم ﺑﺮ ﻣﺘﺮﻣﻜﻌﺐ(، رﻳﺰداﻧﻪ 804/9 – 960 ﻛﻴﻠﻮﮔﺮم ﺑﺮ ﻣﺘﺮﻣﻜﻌﺐ(، ﻧﺎﻧﻮﺳﻴﻠﻴﺲ 0 – 49/6 ﻛﻴﻠﻮﮔﺮم ﺑﺮ ﻣﺘﺮﻣﻜﻌﺐ(، درﺻﺪ ﺣﺠﻤﻲ اﻟﻴﺎف 0 – 0/9 درﺻﺪ(،ﭘﻮدرﺳﻨﮓ آﻫﻚ 0 – 288/9 ﻛﻴﻠﻮﮔﺮم ﺑﺮ ﻣﺘﺮﻣﻜﻌﺐ( و ﻓﻮق روان ﻛﻨﻨﺪه 1/75 – 10/5 ﻛﻴﻠﻮﮔﺮم ﺑﺮ ﻣﺘﺮﻣﻜﻌﺐ( ﺑﺮاي ﻃﺮاﺣﻲ ﻣﺨﻠﻮط ﺑﻬﻴﻨﻪ ﻣﻮرد اﺳﺘﻔﺎده ﻗﺮار ﮔﺮﻓﺖ. در اﻳﻦ ﻣﻄﺎﻟﻌﻪ ﺑﺮاي ﻓﺮﻣﻮﻟﻪ ﺷﺪن ﻣﺴﺌﻠﻪ ﺑﻬﻴﻨﻪ ﺳﺎزي، ﺗﺎﺑﻊ ﻫﺪف ﻣﻘﺎوﻣﺖ ﻓﺸﺎري ﺑﺘﻦ ﺑﺮﭘﺎﻳﻪ روش رﮔﺮﺳﻴﻮن ﭼﻨﺪﮔﺎﻧﻪ ﺧﻄﻲ ﺗﻮﺳﻌﻪ داده ﺷﺪ. ﻫﻤﭽﻨﻴﻦ ﻗﻴﺪﻫﺎي ﺑﺮرﺳﻲ ﺷﺪه در اﻳﻦ ﻣﻄﺎﻟﻌﻪ ﻧﺴﺒﺖ ﻣﻘﺎدﻳﺮ ﻃﺮح اﺧﺘﻼط و ﺣﺠﻢ ﻣﻄﻠﻖ ﻣﻘﺎدﻳﺮ ﻃﺮح اﺧﺘﻼط ﺑﺮاي ﻃﺮاﺣﻲ ﻣﺨﻠﻮﻃﻲ ﺑﺎ ﻣﻘﺎوﻣﺖ ﺑﻬﻴﻨﻪ و ﻣﻘﺮون ﺑﻪ ﺻﺮﻓﻪ ﺑﻪ ﻋﻨﻮان ﻣﺤﺪودﻳﺖ ﻫﺎي ﺗﻜﻨﻮﻟﻮژﻳﻜﻲ از ﻓﺎﻛﺘﻮرﻫﺎي آزﻣﺎﻳﺸﮕﺎﻫﻲ ﺗﻮﻟﻴﺪ ﺑﺘﻦ ﻣﻮرد ﺗﻮﺟﻪ ﻗﺮار ﮔﺮﻓﺘﻪ اﺳﺖ. ﭘﻴﺎده ﺳﺎزي اﻟﮕﻮرﻳﺘﻤﻴﻚ روشﻫﺎي ﻓﺮا اﺑﺘﻜﺎري در ﻣﺤﺪوده 30 – 88/7 ﻣﮕﺎﭘﺎﺳﻜﺎل ﺗﺎ رﺳﻴﺪن ﺑﻪ ﻣﺨﻠﻮط ﺑﺎ ﻣﻘﺎدﻳﺮ ﺑﻬﻴﻨﻪ اداﻣﻪ ﭘﻴﺪا ﻛﺮده و در ﻧﻬﺎﻳﺖ 5 ﻧﻤﻮﻧﻪ از ﻣﺨﻠﻮط ﺑﻬﻴﻨﻪ ﺗﻮﺳﻌﻪ داده ﺷﺪه ﺑﺎ اﺳﺘﻔﺎده از اﻟﮕﻮرﻳﺘﻢﻫﺎي CSA و GA ﺟﻬﺖ ﺑﺮرﺳﻲ ﻗﺎﺑﻠﻴﺖ و ﺑﻬﺮه وري اﻟﮕﻮرﻳﺘﻢﻫﺎ ﮔﺰارش ﮔﺮدﻳﺪ. ﻧﺘﺎﻳﺞ اراﺋﻪ ﺷﺪه در اﻳﻦ ﻣﻄﺎﻟﻌﻪ ﻧﺸﺎن داده اﺳﺖ ﻛﻪ ﻋﻤﻠﻜﺮد اﻟﮕﻮرﻳﺘﻢ CSA ﺑﺎ ﻣﺤﺪوده ﺧﻄﺎي ﻣﻴﺎﻧﮕﻴﻦ 3/38 – 14/49 درﺻﺪ در ﻣﻘﺎﻳﺴﻪ ﺑﺎ اﻟﮕﻮرﻳﺘﻢ GA ﺑﺎ ﻣﺤﺪوده ﺧﻄﺎي ﻣﻴﺎﻧﮕﻴﻦ 7/95 – 15/52 ﻧﺘﺎﻳﺞ ﻗﺎﺑﻞ ﺗﻮﺟﻪ در دﻗﺖ و ﻫﻤﮕﺮاﻳﻲ ﺟﻮابﻫﺎ اراﺋﻪ ﻧﻤﻮده اﺳﺖ. از اﻳﻦ ﻫﻤﭽﻨﻴﻦ، اﻟﮕﻮرﻳﺘﻢﻫﺎي ﻣﻮرد اﺳﺘﻔﺎده ﺑﻪ ﻋﻨﻮان اﺑﺰار ﻗﺎﺑﻞ اﻃﻤﻴﻨﺎن در ﺣﻞ ﻣﺴﺎﻳﻞ ﺑﻬﻴﻨﻪ ﺳﺎزي در ﻣﺴﺎﻳﻞ ﻣﻬﻨﺪﺳﻲ ﺑﻮﻳﮋه ﺗﻜﻨﻮﻟﻮژي ﺑﺘﻦ ﻗﺎﺑﻞ ﺗﻮﺟﻪ ﻣﻲ ﺑﺎﺷﺪ.

The utilization of concrete Incorporating with fibers is one of the proper issues of construction industry in last years. The main focus of this research to design a high performance self-compacted fiber reinforced concrete (SCFRC) by using an evolutionary algorithm, which is implemented in MATLAB. Crow Search Algorithm (CSA) and Genetic Algorithm (GA) are statistical ways which are developed by optimization based meta-heuristic solutions. A total of 67 concrete mixtures were considered by varying the levels of key factors affecting concrete strength of concrete, namely, water content (137.2-195 kg/m3), cement content (325.5-520 kg/m3), coarse aggregate content (722-920 kg/m3), fine aggregate content (804.9-960 kg/m3), nano silica content (0-49.6 kg/m3),percentage of volumetric of fibers (0-0.9 %), lime stone powder content (0-288.9 kg/m3) and superplasticizer content (1.75-10.5 kg/m3) were developed to design optimized mixture proportions. The objective function called maximizing concrete strength was formulated as an optimization problem on the basis of Multiple Linear Regression (MLR) method. The constrains including ratio of mixture proportions and absolute volume of mixture design were utilized to obtain an optimal-strength and cost-effective design. The concrete technological constraints were identified as the factors of experimental design for concrete production. The evolutionary implementation of results reached incorporating mixture proportions having strengths in range of 30 - 88.7 MPa. Five numerical examples for optimum mixture design of SCFRC were considered to evaluate the capability and efficiency of CSA and GA algorithm. These results were compared and concluded that CSA (3.38-14.49 % of mean error) performed better than GA (7.95-15.52 % of mean error) for this application. Also, the proposed evolutionary CSA and GA algorithms are found to be reliable and robustness tools to solve and optimize engineering and concrete technological problem.