Pretty much all of our physics is built on the basis that energy is conserved. It can be “changed” into something else – mass, linear momentum, angular momentum, … – but it can neither be created nor destroyed. This applies to the universe as a whole. Does it apply to a system (which is the word we often use to refer to the part of the universe that we are investigating)?
That can seem to depend on our view of that system. We commonly refer to friction as a non-conservative force. By this we mean that if friction is present energy changes form: from momentum to heat or sound, for example. Aren’t heat and sound are forms of energy too? That depends on your view of the system. For purposes of ease of calculation you may choose not to include all possible forms of energy in your model.
We use conservation principles, together with constitutive equations, to derive Governing Differential Equations that are “solved” by computational methods such as CFD. Most computational methods do not conserve energy precisely: they may do so in an average sense or at selected locations, and even that only to within a convergence limit. This is fine for most engineering designs.