For a long time, the apparent similarity between pterosaur skulls and those of skimming birds, such as the black skimmer has led many people to belief that pterosaurs were likely to have fed by skimming. Skimmers fly very low over the water, dragging the bottom part of their beak in the water and snapping up prey on contact.
In their paper titled Did Pterosaurs Feed by Skimming? Physical Modelling and Anatomical Evaluation of an Unusual Feeding Method, Humphries, Bonser, Witton, and Martill use physical models to actually test the feasibility of skimming as a feeding method for pterosaurs.
In short, it couldn't have worked.
The researchers made physical models, both of the beaks of modern skimmers and of the mouths of pterosaurs, and tested them in a water-based drag testing facility. In so doing, they learned that modern skimmers actually burn more energy than we've suspected during their skimming activity, and that no pterosaur larger than about one kilogram could have managed the energetic costs of skimming.
In their closing discussion, the authors point to the dangers involved in using superficial morphology to guess at functional similarity. In addition to the need to physically test things as they did, they also mention that many people who made the superficial comparison between skimmers and pterosaurs ignored just how specially adapted skimmers are to their feeding style, with many skull traits that simply aren't present in pterosaur skeletons.
I'm a big fan of actually going and testing long-standing suppositions, so I really appreciated this paper.