Spiral bevel gears — also called Gleason-type or curved-tooth bevel gears — transmit motion between intersecting shafts, most often at a 90° shaft angle, and are the default choice where high speed and heavy load coincide. The curved tooth trace gives gradual, overlapping engagement rather than the abrupt line contact of a straight bevel gear, which raises the contact ratio, spreads load across the face width and reduces meshing impact. Automotive final drives, construction and agricultural machinery, metallurgical plant and drilling equipment all rely on them.
The same geometry that makes them quiet and strong makes them expensive to produce. The flank is a helicoid generated by a rotating cutter, tooth thickness changes along the face, and the pinion and wheel must be cut with machine settings that are matched to each other. The number of shops able to cut spiral bevel gears remains smaller than the number producing ordinary cylindrical gears, although the spread of CNC bevel generators and imported machine tools has narrowed that gap. Because trial cuts on a generator are slow and costly, numerical simulation of the forming process has become the practical route to a first, pre-cut assessment of tooth shape, contact pattern and dimensional accuracy.
Structure of the forming workflow
A workable forming study proceeds as a chain in which each stage constrains the next: tooth profile selection → comparison of the meshing contact of candidate profiles → selection of the best profile → choice of pressure angle and spiral angle → programming of the generating calculation → forming of the spiral bevel gear solid. General-purpose modelling and simulation software carries the geometric work, while a dedicated gear package supplies the standards-based generation and the contact analysis that a generic CAD kernel cannot.
Two stages deserve emphasis. Profile selection fixes the shape of the flank before any contact calculation is meaningful, and the contact comparison is what converts a set of candidate profiles into a single decision. Only after that decision does it make sense to fix pressure angle and spiral angle, because those two parameters act on a profile that has already been accepted.
Tooth profile selection and contact comparison
Gear profiles fall into three families: involute, circular arc and non-circular curve. The involute form dominates general transmission, circular arc profiles follow in niche heavy-duty applications, and non-circular curves are reserved for non-circular gears with special motion requirements. Within the involute family the generating geometry still admits several variants, and three were examined for the present case.
The first variant was rejected on inspection: its addendum land is noticeably thinner than that of the other two, leaving a narrow top land that is unfavourable for a bevel gear formed at the tooth tip, where the tooth is already narrowest. Because a thin tip weakens the very region that carries the highest sliding velocity, this variant was excluded before meshing analysis, and only the second and third variants were carried into contact simulation.
| Candidate involute profile | Observed feature | Contact pattern (unloaded simulation) | Disposition |
|---|---|---|---|
| Profile 1 | Thin addendum land | Not simulated | Rejected: tip strength and formability |
| Profile 2 | Standard addendum land | Rectangular, evenly spread over the flank | Selected |
| Profile 3 | Standard addendum land | Elliptical, concentrated | Rejected |
The two surviving variants were meshed in the gear simulation package. Profile 2 produced a rectangular contact patch distributed fairly uniformly across the tooth, whereas profile 3 produced an elliptical patch that concentrates the contact into a smaller area of the flank. Since a patch that occupies more of the face width distributes normal load over a longer line of contact and is less sensitive to a given amount of tooth deflection, the rectangular pattern of profile 2 was adopted as the forming profile.
Pressure angle and spiral angle
With the profile fixed, a normal pressure angle of 20° and a mean spiral angle of 35° were prescribed. Both values sit in the conventional range for Gleason-type gearing: 20° balances bending strength at the root against the radial load component, while 35° provides enough overlap for smooth transfer of load without generating excessive axial thrust.
The tooth trace obtained under this pair of values is markedly displaced in space relative to the planar involute from which it starts. The profile is carried onto the pitch cone and swept along the spiral, so that the initially planar involute gradually becomes the curved, tapered flank that characterises a spiral bevel tooth. This transformation of the trace, rather than any change in the basic profile, is what produces the smooth engagement of the finished gear.
Parameter set and geometric consistency
The gear pair studied is a 2.5:1 reduction with a 90° shaft angle, a normal module of 3.5 mm and a face width of 20 mm. Both members share the same module, pressure angle, spiral angle, addendum coefficient and clearance coefficient, which is a prerequisite for conjugate action. The addendum coefficient of 0.85 and clearance coefficient of 0.188 are deliberately smaller than the values used for general-purpose cylindrical gears; short addenda of this kind are standard practice in bevel gearing, where tooth depth is kept tight to limit sliding at the toe and to avoid interference at the heel.
| Parameter | Pinion | Wheel |
|---|---|---|
| Number of teeth z | 12 | 30 |
| Normal module mn (mm) | 3.5 | 3.5 |
| Normal pressure angle (º) | 20 | 20 |
| Spiral angle at mid-face (º) | 35 | 35 |
| Pitch cone angle (º) | 21.801 4 | 68.198 6 |
| Face cone angle (º) | 26.634 | 70.715 |
| Root cone angle (º) | 19.285 | 63.366 |
| Addendum coefficient | 0.85 | 0.85 |
| Clearance coefficient | 0.188 | 0.188 |
| Face width b (mm) | 20 | 20 |
| Shaft angle (º) | 90 | |
The pitch cone angles follow directly from the tooth numbers for a right-angle drive:
$$\delta_1=\arctan\frac{z_1}{z_2}=\arctan\frac{12}{30}=21.8014^\circ,\qquad \delta_2=90^\circ-\delta_1=68.1986^\circ$$
The tabulated cone angles are internally consistent with that result. The pinion face cone angle plus the wheel root cone angle gives 26.634º + 63.366º = 90.000º, and the pinion root cone angle plus the wheel face cone angle gives 19.285º + 70.715º = 90.000º. The two members therefore share conjugate cone elements at the specified shaft angle, which is the geometric condition the forming calculation must reproduce rather than approximate.
Numerical calculation chain
The generating calculation was programmed in MATLAB, taking the parameter set above as input. The first output is the pair of base circle curves. Their centres do not coincide: the pinion and the wheel sit on different pitch cones, so the base circles from which their involute flanks are unwrapped differ both in position and in curvature. Treating the two members with a single base circle — a shortcut sometimes taken when a generic CAD package is used alone — is not admissible here.
From the base circle curves the calculation proceeds to the tooth profile curve of the 30-tooth wheel, then to its two-dimensional tooth contour, and finally to the three-dimensional solid. The contour stage is where the accumulated effect of module, cone angles, addendum coefficient and spiral angle becomes visible as a closed, manufacturable outline; the solid stage converts that outline into the geometry a generator would be programmed to cut.
The division of labour between the two software tools is the practical point of the method. A dedicated gear package is efficient at standards-based generation and contact analysis, but it depends on accurate parameter entry and on its own internal generation routines; when the target geometry departs from the built-in defaults, the package alone struggles to return precise results. MATLAB contributes the high-precision numerical evaluation of the base circle and profile curves, and its output is transferred into the gear package as point data from which the solid is built. The combined route yields a wheel whose contour and dimensional accuracy correspond to the specified duty, and whose tooth shape is complex enough that neither tool would produce it alone.
What the workflow delivers
The value of this kind of simulation lies upstream of the machine tool. Contact pattern and tooth contour can be examined, compared and revised before any blank is cut, so that the number of trial cuts on a generator — and the associated consumption of tool life and machine time — is reduced. Because the calculation is parametric, changes in module, tooth numbers, pressure angle or spiral angle can be propagated through the whole chain and re-evaluated quickly, which makes the approach usable for process planning and for parameter optimisation rather than for a single fixed design.
The limits should be stated as clearly as the benefits. The simulation represents the unloaded, nominal geometry: it does not by itself account for cutter blade geometry, machine tool settings, heat treatment distortion or elastic deflection of the pair under load, and an unloaded contact pattern is not the same as the pattern observed on a running drive. Results of this kind therefore function as a well-founded starting point for cutting trials and for subsequent loaded tooth contact analysis, not as a replacement for either. Within that role, the chain from profile selection through contact comparison, angle selection and numerical generation to a solid model gives a coherent, checkable basis for improving the forming process, optimising parameters and raising the overall performance of the finished spiral bevel gear.