Agwuocha, Ifeanyi M.2026-03-252026-03-252020-01Agwuocha, I. M. (2020). Formulation of matlab based computer program for mixture experiments, using Osadebe's Regression Model [Unpublished Master's Thesis]. Federal University of Technology, Owerri, Nigeriahttps://repository.futo.edu.ng/handle/20.500.14562/2480This thesis is for the award of Master of Engineering (M.ENG.) in Civil Engineering (Structures)This work “Formulation of a matlab based computer program for mixture experiments, using Osadebe’s regression model” formulated a computer program for analysing mixture experiments based on the Osadebe’s regression model. The computer program developed on Matlab (2015) was called MIX-PRE. It comprises five graphical user interfaces and they are; about user interface, model predicting user interface, model validation user interface, graph validation user interface and the result user interface. The predicting ability of MIX-PRE was tested by comparing results obtained from it to those from previous studies on Osadebe’s Regression model carried out by Anya in 2015 and Okere in 2014. The percentage difference between the results from MIX-PRE and from the previous studies generally showed a less than 4% variation in the results. The highest and lowest values of the percentage differences obtained for the coefficient of regression between MIX-PRE and Anya were 1.79% and -1.43% respectively. While, that between MIX-PRE and Okere were 1.95% and -0.92% respectively. Statistical F and T-tests were used to validate MIX-PRE. The percentage difference in the F values between results obtained from Anya and Okere were respectively compared to that generated by MIX-PRE and their values were 0.02056% and 0.004826%. Similarly, T values obtained from the works of the two authors were compared to that of MIX-PRE and the percentage difference obtained were 0.0001664% for Anya and 0.003733% for Okere. The accuracy of MIX-PRE was further tested using results obtained from experiments carried outin the laboratory. In the second part of this study, a laboratory experiment to determine the compressive strength of concrete using red lump stone as coarse aggregate was carried out. A total of 45 cubes of size 150mm were produced in the laboratory from ten (10) randomly selected mix-ratios. Optimum compressive strength obtained after 28 days of curing by immersion was 20.33N/mm2 and the corresponding mix ratio was 0.657:1.00:0.914:2.286 (water: cement: river sharp sand: red lump stone). F and T-tests at 95% confidence level were then used to test the adequacy of the model predicting ability. F and T values calculated were 1.00293 and 1.735 respectively. While, F and T values from the statistical tables were 9.1172 and 2.7765 respectively. Since the values obtained from calculations were all less than that from the tables, the null hypothesis was accepted, that is, the model showed no significant difference between the experimental results and model prediction. Anova test was also used to further validate MIX-PRE. An anova calculated value of 67.748 from MIX-PRE was found to be greater than the statistical table F value of 17.4434. Therefore, the formulated program (MIX-PRE) will be of great importance in the reduction of the time and resources needed in meeting up with the requirement of the mix design process in concrete production since, there will no longer be the need for trial mix designs in the laboratory.enAttribution-NonCommercial-ShareAlike 4.0 InternationalMix-Prematlabgraphical user interfaceOsadebe regression modelred lump stoneDepartment of Civil Engineeringcompressive strengthMaster’s Thesis