Many low-order models for unsteady flows divide the force into circulatory and non-circulatory components. The former is associated with vorticity in the flow field, whilst the latter is often synonymous with the added mass force. Investigating a cylinder sharp-edged gust encounter, at a Reynolds number of 6000, probes the origin of these respective forces. Vorticity residing in the flow field does not only originate from the cylinder but it is also located in the gust shear layers, which delimit the vertical gust velocity from the surrounding quiescent fluid. It is possible to represent the body surface by a vortex sheet where individual components, satisfying the non-through flow condition, originate from different sources. All vorticity external to the body generates a complementary contribution to this surface vortex sheet. A further vortex sheet component is uniquely attributed to linear acceleration of a body and linked to the added mass effect. Finally, a non-circulatory vortex sheet forms due to the induced velocity by the gust vorticity. In Küssner's potential flow gust model this latter vortex sheet contribution is attributed to added mass. However, because the gust encounter is not associated with any body acceleration, the force must rather be linked to the growth and redistribution of this vortex sheet, due to the relative advection of the gust shear layer vorticity. Particle image velocimetry validates this result using a surging and rotating cylinder at gust ratios of 0.5 and 1. Force balance measurements show that the force originating from the rate of change of the non-circulatory gust vortex sheet, unlike in the case for an infinitely thin plate, vastly over-predicts the initial rise in force as the cylinder enters the gust. This is because, when considering the rate of change of the vortex sheet, it is implicitly assumed that all of the vertical momentum of the gust flow inside the region occupied by the cylinder, is lost. The overestimation is a result of the rigid shear layer assumption inherent to the Küssner model. In reality, the gust shear layers deflect, causing a spread of vertical momentum. The deflection of the shear layers can be analytically approximated by removing the contribution due to the rate of change of momentum inside the cylinder. This improves the force prediction but does not fully recover the experimental force measurements during the initial entry into the gust. On a practical level this suggests that for bodies of finite thickness the non-circulatory force cannot be easily calculated, as it is difficult to quantify the effect of the body volume.