I think the "Cornucopia" model is a great example of "blind" design that works a lot better in theory/fluid dynamics simulations than it would in actual usage.

I think the "Cornucopia" model is a great example of "blind" design that works a lot better in theory/fluid dynamics simulations than it would in actual usage. I would expect a significant percentage of users to find the "hole" design... uninviting, and as a result stand way farther back than the simulation assumed. (Exhibit 1: https://news.ycombinator.com/item?id=43669248 https://news.ycombinator.com/item?id=43669248) The best way to validate such findings would likely be 3D-printing single use versions of different urinals, mounting them at the student pub for one Friday evening, and monitoring usage amounts + the amount of urine on the floor mats (compared with a regular urinal next to it to account for different inebriation levels). This also avoids identifying models that men prefer/avoid if given the choice. (Also a great way to advertise that "90% of patrons preferred the new urinals" because you'll get that effect from the novelty value alone).