I will help you, but this experiment is not a JMP data table and it was not designed with JMP as far as I can tell.
I imported the spreadsheet into a JMP data table. I converted the percentages ( 0-100 ) to proportions ( 0-1 ). I added the necessary column properties to the factors. Then I gave up. I made the design with JMP Custom Design and copied your values over. I can now fit the model. The two-level components support only the pure blend estimates.
What have you done or tried on your own so far? What difficulty were you having?