Spin - "… this property of the electron has no classical analog, and, as will soon become evident, it must be treated by somewhat abstract methods." [Gasiorowicz]
Well here is a model that is classical momentum applied in a non-classical space. There is very little abstraction.
The problem of spin is that it acts like angular momentum but nothing is spinning. Nothing is spinning in even hyperspace. Luckily, there is a definition of angular momentum that applies for object moving in a straight line, l = r x p. Now the short answer is that momentum traveling in a straight line has a value for angular momentum, even in hyperspace dimensions.
In normal space, r is a variable distance. A particle's angular momentum varies depending on how far away you are from an object at its closest point on its trajectory. The angular momentum is a different value if one is a different distance at the closest point on a trajectory.
When dealing with distances that are very tiny, quantum theory with p*x=h applies. Both x and r are distances so that l and h have the same unit of measure. When a particle travels along a short hyperspace dimension, it has a fixed momentum of p. Now this p is the same value as p. If one tries to have another particle interact with it, the interaction can only occur is in units of 1/2 h. This is the minimum interaction of momentum with this particle. Any interaction occurs at a distance r so that r x p = 1/2 h. An interaction acts as a transfer of angular momentum because the vector of r is perpendicular to the vector of p. The spin of 1/2 h is not inherent to a particle, only p is. But an interaction can only occur in set quantum values so that spin is very consistent and seems to be a property of a particle. It really is a result of a combination of hyperspace dimensions perpendicular to normal space and quantum theory. This last is similar to the original theoretical work on spin. PAM Dirac combined special relativity and quantum theory. One result was antimatter another was spin. HyperSpace is an extension of curved space-time so that a model for spin should require quantum theory and relativity.
What does spin look like for an electron? If we start with a balloon model as a magnified version of hyperspace one can pick out a couple of the properties that is associated with 1/2 spin. An electron has a classical dimension and a fine structure constant that is 1/137 the size of the classical dimension. This can be viewed as a balloon that is 137 times longer than it is around. One takes the two ends of the balloon and brings the ends together to form a very skinny donut. Now the ant that we imagine traveling on this donut shape wants to take the longest path before returning to its starting point. This gives it the smallest p that fits on this shape. There are a couple of paths one could imagine, take the outer edge and circle it along the long axis, this is a distance of 137 if the diameter of the small dimension is 1. Another possibility is to circle along the short dimension but slightly move along the long dimension. This unfortunately is actually requires quite a lot of p along the short dimension though p is a lot less along the long dimension. The solution that has less momentum than the long dimension is to spiral 1/2 turn along the short dimension after travelling along the long dimension once. One doesn't return to the starting location until the long axis has been circled twice. The real result is a slowly twisting spiral that slowly precesses about the short axis. A full wave mechanic solution is needed, a boundary value problem on a curved surface isn't in any of my textbooks. Or a minimum action calculation.