Optimizing compressed Earth blocks mix design incorporating rice straw and cement using artificial neural network

• Main Authors: Gapuz, Emerson O., Ongpeng, Jason Maximino C.
• Format: text
• Published: Animo Repository 2017
• Subjects: Earth construction; Self-organizing maps; Neural networks (Computer science); Civil Engineering
• Online Access: https://animorepository.dlsu.edu.ph/faculty_research/1902
• Institution: De La Salle University

Compressed earth blocks (CEB) in construction is an alternative way of promoting sustainable construction building materials. Compressed earth blocks are used as replacement to concrete masonry wall units. It has many advantages in terms of material cost, thermal properties, embodied energy, sound and fire proofing. In this study, production of 250 CEBs was made with dimension of 290mm×140mm ×100mm. With the available data, Self-Organizing Map (SOM) toolbox was used to classify results according to the parameters that have similarities. The groupings that were classified through SOM were observed, analysed and was related to the compressive strengths of CEBs. Two Self Organizing Map (SOM) Models were derived in the study. These are Model E and Model F. Model E and Model F contains 2 input parameter. Model E has 4 classifications: Group A and C classified CEBs that are above the strength requirement of Philippine National Standard (PNS). Group B and D clustered the CEBs with the lowest compressive strength value. Overall, Model E clustered the groups into similar characteristics. Model E showed that CEB with 10% cement and above with any fiber content conforms to the requirement of PNS under TYPE 2 CHB. Model F has also 4 classifications: Group A and C classified CEBs that are above the strength requirement of PNS. Group B and D clustered the CEBs with the lowest compressive strength value. Model F showed that any fiber content can be used in combination with 10% or more Cement to achieve the requirement of PNS Type 2 CHB. © 2017 IEEE.