Configuration B, worked as a method you can repeat when the transmission changes.
The shaft is a simply supported beam with point loads at the component centrelines. The loads are not in one plane, so split them into the horizontal and vertical planes, solve each as a 2D beam, then recombine.
Why this works: bending in one plane does not care about forces in the perpendicular plane. Each plane obeys its own ΣF = 0 and ΣM = 0, and the stresses only meet again at the cross-section, where the two moments combine as vectors.
Shaft along x from bearing A. A and B are bearings, idealised as simple (pin) supports. T1 and T2 are the track sprockets, C is the driven chain sprocket. D is the drive sprocket on the motor, above and forward of C, with the line of centres at 30° to horizontal. Sprocket sizes are schematic.
Loads orange arrows, applied at the centreline of each sprocket.
Reactions blue arrows, at the bearings A and B.
The chain pulls C along the line towards D, so F_C splits into a forward part (F_C cos30) and an upward part (F_C sin30). The tight bottom run of each track pulls its sprocket rearward, horizontally.
Positive = forward. Every arrow is drawn in its actual direction. Under the FBD, the SFD and BMD share the same x axis (dashed station lines).
Positive = up. Only C loads this plane, with F_C sin30 = 1061N up, so both bearings must pull the shaft down.
Arrow lengths are indicative, the 138N arrow is lengthened so the head stays visible. Read the labels, not the lengths.
Between point loads the shear is constant, so the moment is a straight line whose slope equals that shear. At each point load the shear jumps by the load, and the moment line changes slope. The moment returns to 0 at both pins.
| Plane | A to T1 | T1 to T2 | T2 to C | C to B |
|---|---|---|---|---|
| Horizontal SFD | +574.2N | −145.8N | −865.8N | +972.2N |
| Vertical SFD | −138.4N | −138.4N | −138.4N | +922.6N |
| Moment | A | T1 | T2 | C | B |
|---|---|---|---|---|---|
| Horizontal M_h | 0 | +28.7Nm | −15.0Nm | −58.3Nm | 0 |
| Vertical M_v | 0 | −6.9Nm | −48.4Nm | −55.4Nm | 0 |
The two planes are perpendicular, so the two moments at a section are perpendicular vectors and add by Pythagoras. A shaft does not care which plane it bends in, only how big the total moment is.
| Section | M_h | M_v | M |
|---|---|---|---|
| T1 | 28.7Nm | 6.9Nm | 29.5Nm |
| T2 | 15.0Nm | 48.4Nm | 50.7Nm |
| C (critical) | 58.3Nm | 55.4Nm | 80.4Nm |
Magnitudes only: the signs drop out because the square removes them.
Torque T enters the shaft at C from the chain. It travels towards the track sprockets and leaves through them. At T2 half of it is taken out by that track, and the other half T/2 carries on to T1, where it leaves. So the shaft carries T between T2 and C, T/2 between T1 and T2, and nothing outside the sprockets. A to T1 and C to B carry no torque.
Reactions. Take moments about one bearing, then ΣF for the other, in each plane separately. Confirm both ΣF and ΣM close to 0.
SFD and BMD per plane. Walk along the beam, step the shear at each load, integrate to get the moment, and check it returns to 0 at B.
Resultant M. Combine at each section with M = √(M_h² + M_v²). The largest is at C, 80.4Nm.
Torque. Trace T from where it enters (C) to where it leaves (T1 and T2). The shaft carries T, T/2 or 0 in each segment.
Combined stress at T1, T2 and C.
Pick d so that σ' is below σy/SF at the worst of the three sections. C has the largest M and the full T, so it is the likely governor, but check T2 as well.
Stress concentration. Multiply the nominal stress by Kt at shoulders and keyway ends. The nominal sections above are only the start, the local peak at a shoulder near C can govern the design.
The drive tension sits in the tight bottom run of the track, the run that pushes the machine forward by reacting against the ground. That run pulls on the lower edge of each sprocket in the direction the track wants to travel relative to the snow, which is rearward on the sprocket. The slack top run carries almost nothing, so the net sprocket load is one horizontal force of 720N each, rearward.
Between point loads the shear V is constant. Since dM/dx = V, the moment is a straight line there. A straight line has its extremes at its ends, and the ends of each segment are the load points or the supports. So the maximum can only be at a load point, which is why you only evaluate A, T1, T2, C and B.
In config B, C sits inside the span between the bearings, so the two bearings share the chain load and the moment peaks at C with a modest value. In config A, C is on a 60mm overhang past bearing B. The overhang acts like a lever, the load is carried mostly by the near bearing with a reaction bigger than the load, and the moment at B grows with the overhang length. Bringing C inside the span removes that lever arm.