Synthetic rock samples made of cement and silica sand were mixed to target strengths for the erosion experiments.
The setting time for the erodible samples was the main time-limiting factor of the experimental schedule. The samples were poured in the drum, and had to set before any other step that involved rotating the drum was taken. To decrease the time needed for the setting of cement, we used Type III Portland cement instead of the more common Type I Portland cement. The samples were poured into wooden forms that were 60x60x10 cm. The samples were mixed, dry ingredients first, then addition of water to the point of saturation in a cement mixer, then poured into the main form and smaller shoebox-sized forms for Brazilian Tensile Strength samples.
Sample strength evolves with time as the cement cures, so we tested the same sample at multiple times, ideally near the date of the erosion test. The main factors influencing the strength of the sample were the ratio of sand to cement and the time since pouring (add Figure S1).
To perform the Brazilian Tensile strength test, a cylinder of 4 cm (check) in diameter and ~10cm long was formed with a drill. Two different machines performed the test over the course of the study, an analog machine at San Francisco State and a digital machine at the Richmond Field Station. The test applies a compressive force in the vertical axis along the length of the sample until it fails in tension, creating 2 half-circle cylinders (add Fig S2). Thin bamboo stickers were placed at the contact points to focus the pressure.
Although there is some variation in the strength average, the different sand:cement ratios were able to hit a certain strength range.
New samples were made once the old one eroded approximately 5 cm below the flume depth.