So, this seemed really fun to me until I got stuck. THE TASK is about an object with mass m, moving in a basic (2D) coordinate system. It is attached to origo (0, 0) by a "rope" with constant length r=5. In position P0(-5, 0) it has the velocity v0=[0, -10]. Hence, the object moves around origo in a circular loop, and the force on m from the rope shall be refered to as C. There is also another (constant) force: F=[10m, 0] (m being the mass: all calculations should be done without units). F could represent a gravitational pull, to ease the comparison to a standard loop. By the terms, the mechanical energy is conserved. First, you are asked to find the acceleration (expressed as a vector) in points (-5, 0), (0, -5), (5, 0), (0, 5). This is easily done by Newton's 2nd law: ΣF=ma⇒ either m[±v2/r, 0] (in points (-5, 0) and (5, 0), because ΣF=centripetal force=mv2/2) or m[10, ±v2/r] (in the other two points, because ΣFx=F and ΣFy=centripetal force). FINALLY: You are now asked to find a function a(α)=[ax, ay]. - a is the acceleration vector of m. - α is the angle, determining the position of m. Preferably, you want this to be the angle from the positive x-axis to the "rope" (counter clockwise). - The vector function shall describe the acceleration vector of m at any position in the loop. Good luck Notes: It seems that the reason why the second task is hard is because F interferes with Cx. In comparison: in the first task, F has either everything or nothing to do with C. C does of course consist of the pull from the rope, and in most positions also some push/pull from F. I can hardly explain the issue any better (at least in English), but you may anyhow find that the task is quite difficult. An idea of mine is to define the x-/ y-axes along the velocity vector and the C vector, so that when α changes, F only rotates around m. In this redefined coordinate system: ΣFx will then depend only on the angle, and will simply* be equal to Fx (*still difficult to cover the entire 360° movement with one function). ΣFy will always equal C, again equal to mv2/r - and a function of the size v may be possible to work out. With this x-/y-perspective I suppose you will result in a vector function for a, which would then have to be modified to "rotate" back to work with the x-/y-axis definition in the original task. I am inexperienced with the English physics terms, hence the poor or incorrect use of terminology. Please inform me on potential for improvement!