The paper studies a class of equations that describe how certain fluids with memory behave over time. Specifically, these are viscoelastic fluids, meaning materials that behave partly like a liquid and partly like an elastic solid, such as some polymers or biological fluids. The equations are "nonlocal in time," which means the current state of the fluid depends not just on what is happening right now but on its entire past history. The diffusion process built into these equations is of a type called Levy diffusion, which can capture irregular or jumpy behavior rather than the smooth spreading described by ordinary heat-like equations.
The first main result establishes that a well-behaved, classical solution to the linear version of these equations actually exists, provided the memory kernels (the mathematical objects encoding how past states influence the present) satisfy a property called complete positivity. The authors then examine what happens as a key parameter controlling the diffusion becomes concentrated near a specific value. They show that in this limit the solutions converge, uniformly and at a quantifiable rate, to solutions of a simpler but related equation called the generalized Rayleigh-Stokes equation, which is a known model for certain viscoelastic flows. This provides a rigorous connection between the more general nonlocal framework and a more classical model.
The second main result extends the analysis to a nonlinear version of the problem, where additional complexity is introduced through nonlinear terms. Here the authors work with a weaker notion of solution, called a mild solution, and prove both that such solutions exist and that they exhibit similar limiting behavior as before. The technical tools used throughout rely on two key ideas: the subordination principle, which relates solutions of different equations through probabilistic representations involving random processes, and relaxation function theory, which provides a structured way to handle the memory effects encoded in the equations.