Why the Contact Pattern Is the Manufacturing Interface
Spiral bevel gears are used where motion must be turned through an intersecting axis under high torque. Their curved teeth engage gradually, giving a larger contact ratio than straight bevel gears and producing smoother, quieter operation with greater load capacity. The same geometry makes quality control difficult: the load-bearing zone is not a simple line but a small, three-dimensional contact pattern whose position, orientation, and size are fixed by the generating machine settings.
A pattern that runs to the toe or heel, touches the addendum, or lies diagonally across the flank concentrates stress, raises transmission error, and shortens service life. For high-precision spiral bevel gears, the acceptance target is a quasi-elliptical pattern centered slightly toward the toe, with smooth transition at both ends and no edge contact. Trial cutting followed by manual blueing is slow and depends heavily on operator judgment. Tooth contact analysis, or TCA, replaces part of that loop by calculating the pattern from cutter geometry and machine kinematics.
The correction problem is therefore not simply “move the pattern.” The pattern position, its angle relative to the pitch cone, and its length and width respond differently to cutter position, wheel position, roll ratio, and cutter radius. A usable correction method must identify which setting controls which pattern feature while preserving the mid-spiral angle and avoiding a new edge-contact condition.
Generating Tooth Surfaces and Simulating Mesh
Generated spiral bevel gears are cut by a rotating cutter head whose cutting edges form a conical generating surface. The machine cradle, cutter, and workpiece motions are represented by a set of coordinate transformations. For a pinion index i = 1 and gear index i = 2, the cutter surface can be written as
$$
\mathbf r_{gi}(u_{gi},\theta_{gi}) =
\begin{bmatrix}
(r_{Gi}\pm u_{gi}\sin\alpha_{gi})\cos\theta_{gi}\\
(r_{Gi}\pm u_{gi}\sin\alpha_{gi})\sin\theta_{gi}\\
-u_{gi}\cos\alpha_{gi}\\
1
\end{bmatrix}
$$
where ugi and θgi are cutter-surface parameters, αgi is the cutter profile angle, and rGi is the cutter tip radius. The upper and lower signs correspond to the concave and convex flanks. The unit normal of the cutter surface follows from the cross product of the two parameter derivatives.
The generating condition is that the cutter surface and the workpiece surface remain tangent along the cutting edge:
$$
\mathbf n_{mi}\cdot\mathbf v_{mi}=0
$$
where nmi is the cutter normal expressed in the machine frame and vmi is the relative velocity at the cutting point. Combining this equation with the coordinate transformation gives the theoretical tooth surface:
$$
\mathbf r_i=\mathbf M_{gi}\mathbf r_{gi},\qquad
\mathbf n_i=\mathbf L_{gi}\mathbf n_{gi}
$$
Here Mgi is the 4×4 transformation from the cutter frame to the workpiece frame, and Lgi is its 3×3 rotation part.
For mesh simulation, the pinion and gear surfaces are mapped into a common coordinate frame. At every contact instant the two surfaces must share a point and a common normal:
$$
\mathbf r_{h1}=\mathbf r_{h2},\qquad
\mathbf n_{h1}=\mathbf n_{h2}
$$
After projection onto the coordinate axes, the unit-normal condition leaves five independent scalar equations for six unknowns. Prescribing the pinion rotation increment reduces the system to a solvable contact-point problem. Repeating the calculation along the path of motion produces a sequence of instantaneous contact points; stacking these paths across the tooth surface produces the simulated contact pattern. The instantaneous contact ellipse is estimated from the principal curvatures of the two surfaces and an assumed elastic approach in the range of \(2.5\times10^{-4}\) to \(6.25\times10^{-4}\) mm.
Parameterizing the Pattern
A visual pattern is difficult to compare when machining parameters change by small amounts. The pattern is therefore projected into the TCA plane and reduced to four boundary points \(A(x_1,y_1)\), \(B(x_2,y_2)\), \(C(x_3,y_3)\), and \(D(x_4,y_4)\). The major and minor dimensions are
$$
L=\sqrt{(x_2-x_1)^2+(y_2-y_1)^2},\qquad
M=\sqrt{(x_3-x_4)^2+(y_3-y_4)^2}
$$
and the orientation relative to the pitch-cone direction is
$$
\zeta=\arctan\!\left(\frac{y_2-y_1}{x_2-x_1}\right)
$$
The midpoint \((x_0,y_0)\) locates the pattern on the flank. An area metric \(S\), formed as \(\pi LM\) in the TCA output, is used for relative size comparison rather than as a physical loaded-contact area. The parameter set \([L,M,S,(x_0,y_0),\zeta]\) gives a compact description of pattern length, width, area, position, and diagonal orientation.
Sensitivity of Pattern Position, Shape, and Size
Position along the face width. The radial cutter position is the primary control. Within the TCA parameterization, moving the radial cutter position from negative to positive shifts the pattern midpoint from the large end toward the small end. The horizontal midpoint coordinate is more sensitive to this setting than the vertical coordinate, so a large face-width error is best corrected with the radial cutter position. The setting is also tied to the mid-spiral angle:
$$
S_{d1}=\sqrt{L_m^2+r_d^2-2L_m r_d\sin\beta_m}
$$
where \(L_m\) is the mid-cone distance, \(r_d\) is the cutter radius, and \(\beta_m\) is the mid-spiral angle. An offset change \(\Delta E\) can be converted into an equivalent radial cutter correction by
$$
\Delta S_{d1}=\frac{\Delta E\,r_d\cos\beta_m}{S_{d1}}
$$
Position along the tooth height. The horizontal wheel position mainly controls the root-to-tip location. Moving the horizontal wheel position from forward, negative, to backward, positive, shifts the pattern midpoint from the small end toward the large end. The vertical midpoint coordinate is more sensitive to this setting. On the pinion concave flank the pattern moves from dedendum toward addendum; on the gear convex flank the same setting moves it from addendum toward dedendum. Bed position follows the coupled relations
$$
\Delta X_{P1}=-\Delta P,\qquad
\Delta X_{B1}=-X_{P1}\sin\delta_1
$$
where \(X_{P1}\) is the pinion horizontal wheel position, \(X_{B1}\) is the bed position, and \(\delta_1\) is the pinion pitch-cone angle.
Diagonal shape. The roll ratio controls the angle between the pattern major axis and the face width. Increasing the roll ratio increases the angle \(\zeta\), and an excessive angle produces internal diagonal contact. The correction is coupled to the horizontal wheel position through
$$
\Delta j=\frac{j\cos\delta_i}{L_m}\Delta X_{P1}
$$
where \(j\) is the roll ratio. In practice, the roll ratio is used after the face-width and tooth-height positions are approximately established, because changing it also alters the contact path.
Pattern size. The cutter radius mainly controls the major dimension: increasing the cutter radius reduces \(L\). The vertical wheel position mainly controls the minor dimension: increasing the vertical wheel position reduces \(M\). Both settings shift the mid-spiral angle, so the radial cutter position must be compensated. For a vertical wheel change, the compensation follows
$$
\Delta S_{d1}=\Delta E_{m1}\sin q_1
$$
where \(E_{m1}\) is the vertical wheel position and \(q_1\) is the pinion angular cutter position. The radial cutter position also affects the area metric non-monotonically: as it increases from negative to positive, the area first rises and then falls, with the largest area occurring at a relatively small change. Area adjustment therefore requires a final TCA check rather than an open-loop setting change.
Case: Correcting a 28/33-Tooth High-Precision Pair
The correction procedure was examined on a high-precision spiral bevel pair with 28 pinion teeth and 33 gear teeth. The pinion is left-hand, the gear is right-hand, and the shaft angle is 90°. The large-end module is 5.7 mm, the face width is 37 mm, the normal pressure angle is 20°, and the mid-spiral angle is 35°.
| Item | Pinion | Gear |
|---|---|---|
| Number of teeth | 28 | 33 |
| Hand | Left | Right |
| Shaft angle (°) | 90 | 90 |
| Large-end module (mm) | 5.7 | 5.7 |
| Face width (mm) | 37 | 37 |
| Outer cone distance (mm) | 123.343 | 123.343 |
| Mid-cone distance (mm) | 104.841 | 104.841 |
| Inner cone distance (mm) | 86.34 | 86.34 |
| Normal pressure angle (°) | 20 | 20 |
| Mid-spiral angle (°) | 35 | 35 |
| Pitch-cone angle (°) | 40.314 | 49.686 |
| Face-cone angle (°) | 43.317 | 52.18 |
| Root-cone angle (°) | 37.82 | 46.683 |
| Addendum (mm) | 4.422 | 3.493 |
| Dedendum (mm) | 4.564 | 5.495 |
| Pitch diameter (mm) | 135.66 | 159.885 |
The initial pinion settings are listed in Table 2. TCA produced the pattern metrics in Table 3. Both working flanks—pinion concave and gear convex—ran close to the addendum, and the pattern ends terminated abruptly without a transition zone. The major-axis angles were \(10.12^\circ\) on the pinion and \(-9.86^\circ\) on the gear, indicating a diagonal-contact tendency that is unfavorable for stable transmission. The pinion pattern was also narrow, with \(M=2.9703\) mm.
| Setting | Concave, initial | Concave, corrected | Convex, initial | Convex, corrected |
|---|---|---|---|---|
| Radial cutter position \(S_d\) (mm) | 97.8676 | 98.2244 | 111.142 | 110.7405 |
| Angular cutter position \(q_1\) (°) | 63.7692 | 66.6537 | 63.7692 | 61.3159 |
| Bed position \(X_B\) (mm) | 0.55 | 0.3312 | −0.65 | −0.2157 |
| Roll ratio \(j\) | 2.2 | 2.3163 | 2.3 | 2.1836 |
| Horizontal wheel position \(X_P\) (mm) | 2.5068 | 2.9081 | −2.4762 | −2.8842 |
| Vertical wheel position \(E_m\) (mm) | −0.4456 | −0.5515 | 0.5289 | 0.6207 |
| Cutter radius \(r_d\) (mm) | 114.3 | 113.5475 | 114.3 | 114.9689 |
The corrected setting set moved both patterns away from the addendum and introduced a visible transition at the ends. The pinion pattern changed to \(L=15.5496\) mm and \(M=5.5747\) mm, while the gear pattern became \(L=16.8383\) mm and \(M=4.9006\) mm. The pinion midpoint moved to \((102.4529,-0.5043)\) mm and the gear midpoint to \((103.0009,-0.5322)\) mm. More importantly, the diagonal angle dropped to \(2.56^\circ\) on the pinion and \(-2.79^\circ\) on the gear. The area metric rose from 169.4197 to 272.1915 on the pinion and from 245.6795 to 259.1057 on the gear, so the contact zone became broader and better supported without running to the tooth edges.
| Flank and condition | \(L\) (mm) | \(M\) (mm) | \(S\) (area metric) | Midpoint \((x_0,y_0)\) (mm) | \(\zeta\) to pitch-cone direction (°) |
|---|---|---|---|---|---|
| Pinion concave, initial | 18.1647 | 2.9703 | 169.4197 | (104.6581, 1.1127) | 10.12 |
| Gear convex, initial | 18.1716 | 4.3057 | 245.6795 | (104.7635, −1.0245) | −9.86 |
| Pinion concave, corrected | 15.5496 | 5.5747 | 272.1915 | (102.4529, −0.5043) | 2.56 |
| Gear convex, corrected | 16.8383 | 4.9006 | 259.1057 | (103.0009, −0.5322) | −2.79 |
The \(S\) column is the \(\pi LM\) area metric used by the TCA comparison; it is not treated as a physical loaded-contact area.
Rolling Inspection and Agreement with TCA
The simulated correction was checked by rolling inspection, which is the practical test of engineering interchangeability for a assembled gear pair. The test used a CNC-modified Y9550 bevel gear rolling inspection machine. The pinion and gear teeth were coated, the pair was mounted according to the required running position, and the machine was run until the mesh was stable and free of abnormal noise. The drive was then stopped and the working-flank pattern was photographed and measured.
The inspected pattern was located near the center of the tooth, slightly toward the small end. Its length occupied about 50–60% of the face width, and its width occupied about 60–70% of the tooth height. These proportions are consistent with the intended contact zone. The TCA pattern was quasi-elliptical, while the rolling-inspection pattern appeared date-pit-shaped. Grinding error, mounting deviation, and subjective boundary judgment during pattern transfer can account for part of this shape difference. The position, overall size, and diagonal orientation nevertheless agreed closely enough to support the correction rules.
The comparison also shows the proper role of TCA. It is a deterministic guide for setting the generating parameters, not a replacement for final inspection. The calculated pattern is based on nominal tooth surfaces and a small elastic approach; the rolling test includes the actual manufactured surfaces, assembly condition, and operator judgment. For high-precision spiral bevel gears, the two checks should be used together.
Practical Correction Sequence and Constraints
The sensitivity results point to a practical correction order. First, establish the face-width position and mid-spiral angle with the radial cutter position. Second, set the tooth-height position with the horizontal wheel position and bed position. Third, remove excessive diagonal orientation with the roll ratio. Fourth, adjust the major and minor dimensions with the cutter radius and vertical wheel position, then compensate the radial cutter position to hold the spiral angle. Finally, recheck the area metric and confirm the pattern by rolling inspection.
This order matters because the settings interact. A change made to correct length can shift the pattern toward the toe or alter the spiral angle; a change made to correct the diagonal can alter the contact area. The correction is therefore a constrained multi-parameter adjustment, not a sequence of independent knobs.
The method also has limits. Unloaded TCA does not predict the load-distributed pattern or the transmission-error spectrum under torque. The settings are specific to the tooth geometry, cutter head, and machine configuration. Heat-treatment distortion, grinding variation, and fixture error remain outside the nominal TCA model. Rolling inspection therefore remains the final acceptance step, especially for spiral bevel gears used in vehicle, construction-machinery, and aircraft drivetrains where edge contact and noise are critical.