Self-studying databases and doubts on PhD

Hi all, hope all is well. I am a graduate student, but not in CS. However, I do program backend. Recently, I've been very interested in understanding Database systems but have been struggling to find the right resources. I've tried googling and reading some, but I'm getting confused by different random sites. Additionally, I generally want to know how people who learn in-depth do it! Here are some of the things I'd like to understand:

- How exactly does an SQL engine organize the tables? What data structure does it use (trees? heaps?)? Does it change to a B-tree when the index is created? How does all of this work?

- What does it mean by an elastic database?

- Completely unrelated, but this is one thing I've been obsessed with recently: How does one choose what to work on in a PhD, especially in the database area? For example, I am interested in query optimization, but what should I research? How does one start? Also, I've seen some profiles of individuals who pursued a different domain for undergrad and master's but then went into a CS PhD. How can one do this? Basically, how do you look for a problem in something you're interested in, and how do you know that it can be solved?

Thank you all! Appreciate your time.