I trained as an engineer but have often worked as an applied mathematician. I got into this role when a group of geologists were working on a big project and none of them had the necessary math skills so I was assigned to the Team. On several occasions I have worked with geologists, hydrologists, biologists, meteorologists, and more diverse experts to transform their ideas and concepts of processes into mathematical models, often asking, "Does this look like you expect it to?" and "What part doesn't look right?" I chose the methods (FDM, FEM, FIM, etc.), they came up with the test cases, and together we solved some big problems, including the most expensive environmental cleanup ever funded by the USDOE (MMR) and managed the USEPA (PCBs).
In order to understand mathematical concepts, you need to focus on physical geometry, try to convert/formulate it into mathematical codes, modelling or input to output representations. Mathematical Models/equations/functions are nothing but behaves like physical scale/balance for measuring masses or volumes etc.
In general, it takes a lot of time and motivation to understand mathematical concepts. Once you reach that level, you will imagine that everything for recognition, visualization, computation, realization etc need MATHEMATICS.