Appropriate sampling and statistical tool for experiment

good day! we are conducting an experiment on which concrete mixture performs best on certain parameters. the mixtures are 75% cement -25% water, 60% cement-40% water, and 50% cement-50% water. the three mixtures will be subjected to strength test. How do we determine the appropriate number of samples for each mixture so that the strength that will be measured is valid and accurate? What is the appropriate statistical tool that can be used for such experiment?