1. The Tooth as a Contact Pair
A gear pair transmits power through line contact between tooth flanks, and the contact between two convex rolling surfaces is the most loaded small area in the machine. The designer verifies not only the bending strength of the tooth at its root but the contact stress of the flanks, because the two failure modes are governed by different physics. Bending stress wants to break the tooth; contact stress wants to pit, scuff and wear the surface. This article develops the contact stress verification from the fundamentals of Hertzian contact through the load factors of a real gearbox.
Contact stress, usually written as sigma H, is the maximum compressive stress at the center of the contact zone between the two meshing flanks, and the standard verification compares this stress to the allowable contact stress that the material can survive for the design life. The classic procedure, embodied in the international gear rating standards, calculates the Hertzian contact stress at the pitch point from the transmitted tangential load, the effective radius of curvature of the two flanks, the elastic properties and the face width, then applies a set of factors for the real service: application, dynamic, load distribution and life.
This article walks the full verification: the geometry of the contact, the load and its factors, the material allowable, the safety factor, and the engineering response when the calculation says the teeth will not last.
2. The Hertzian Contact Model
Two cylinders pressed together deform at the contact into a narrow rectangular band, and the pressure across that band is a semi ellipse with its peak at the center. For gear teeth the two flanks at the pitch point behave as two cylinders with radii equal to the radius of curvature of the involute at that point, and the maximum Hertzian contact stress is proportional to the square root of the load per unit face width, divided by the effective radius of curvature, times a material term made of the elastic moduli and Poisson ratios of the two gears. Doubling the load raises the stress by only about forty percent; doubling the face width, the effective curvature or the contact length reduces it.
The effective radius of curvature deserves attention because it changes along the line of action. At the pitch point the two involutes share one sign of curvature, and the effective radius is modest; near the ends of contact the geometry differs. Applying the largest load with the least effective curvature, usually near the pitch point region or the lowest point of single tooth contact, is the design worst case for contact stress, and the standards choose a defined position for the verification. Larger pinions, more teeth and pressure angle all move the effective curvature favorably.
The contact stress is a compressive stress, and like all compressive stresses at a rolling contact it sets up a subsurface shear the material must tolerate. The maximum shear stress lies a small distance below the surface, roughly a tenth to a third of the contact half width, and it drives the pitting crack that starts below the surface and breaks out as a spall. Contact stress verification is therefore a subsurface fatigue calculation as much as a surface one, and the allowable values are set by tests on the same material system under rolling contact.
3. The Load and Its Factors
The nominal contact stress is calculated from the transmitted tangential load, but a real gearbox never transmits a clean steady load, so the verification multiplies the nominal stress by a chain of factors that carry the real world into the calculation. The application factor accounts for the nature of the driven machine and the driver, accepting that an internal combustion engine or a rock crusher adds a shock every revolution. The dynamic factor accounts for the meshing impacts and tooth deflection that raise the instantaneous load above the transmitted average, and it depends on the pitch line speed and the quality of the tooth profile and lead.
The load distribution factor deserves special study, because it expresses the truth that the load is not shared evenly along the face. Misalignment of the shafts, deflection of the housing, helix angle and the crowning of the teeth all concentrate the pressure toward one end of the face or toward one flank, and the factor raises the effective stress accordingly. A pair of perfectly crowned gears running in an infinitely stiff housing plays beautifully with a nominal distribution; the same gears in a deflecting aluminum housing concentrate the load and fail the flanks early. The machinist’s alignment and the housing design are part of the contact stress allowance.
Additional factors modify the life equation: a size factor for large geometry, a surface finish factor for the flank roughness, and a lubrication factor for the regime of the lubricant film. These factors are not decoration; each has a physical origin, and the sum of their overestimates is what the standards deliberately, if conservatively, apply. The verification is a chain, and the weakest factor, the one that moves the result the most, is the one the engineer should quantify with real data rather than accept from a table.
4. The Allowable Contact Stress and Life
The allowable contact stress is established by testing, because the pitting endurance of a gear material, its hardenability, its heat treatment and its residual stress are properties that cannot be calculated from a handbook hardness number alone. The standard reference is a gear test rig, a pair of standard test gears running under controlled load, speed and lubrication until the flank area attains a specified density of pits, and the result is the stress number, the allowable contact stress, of the material system at the reference life. The designer scales that reference value for the actual part and the required life.
The life scaling follows a classic fatigue law: the allowable contact stress decreases as the required life grows, so a gear specified for a million cycles can carry a higher flanks stress than a gear specified for a hundred million. The slope of the life line differs for the finite and the infinite life regions, and the designer verifies against the right branch. For a gear to run as long as its machine, the infinite life branch is used and the material is chosen so its endurance limit is not exceeded; for a duty with a defined overhaul life, the finite branch gives a lighter but honest design.
Heat treatment changes the allowable more than any other design choice. A through hardened carbon steel flank has a modest allowable contact stress; a case hardened and ground flank, carbonitrided and polished, reaches several times that value because the hard case resists both the surface pitting and the indentation that seeds it. The production route, the case depth relative to the effective radius of curvature, the core hardness and the residual stress from grinding, is therefore part of the material specification, and the verification’s allowable value is chosen only after the manufacturing process is fixed.
5. Safety Factor and the Verification Statement
The verification concludes by forming the quotient of the allowable contact stress and the calculated contact stress, the safety factor, and comparing it to the minimum the application demands. A safety factor above one means the flanks survive the nominal calculation, and the design rule adds a margin for the uncertainty in the factors, the scatter in the material and the impossibility of perfect alignment in service. The required margin is a project decision, written into the specification, not a number pulled from a table without context.
| Design variable | Effect on contact stress | Practical lever |
|---|---|---|
| Center distance | increases effective radius of curvature | larger module, more teeth |
| Face width | reduces load per unit width | wider gears, stiffer housing |
| Pressure angle | changes curvature and load sharing | profile shift, larger angle |
| Hardness and case | raises the allowable stress | case hardening, grinding |
| Lubricant and finish | protects the flank film | correct viscosity, honed finish |
The numbers in the verification statement should tell a story: the calculated stress, the allowable for the chosen material at the required life, the safety factor, and the governing factor that limits the result. When the safety factor is too low, the engineer has a menu of moves in that table, and the right move is the one that addresses the governing term, larger effective radius for the curvature, wider face for the load distribution, or a harder flank when the allowable is the bottleneck. Raising a non governing factor grows cost without improving the verdict.
6. The Verification Procedure
- Fix the gear geometry: module, teeth, pressure angle, helix and face width
- Calculate the transmitted tangential force from torque and pitch radius
- Determine the effective radius of curvature at the design point
- Select the material, heat treatment and case depth for the flanks
- Choose the standard load factors from the real service duty
- Calculate the nominal and factored contact stress
- Form the safety factor against the life scaled allowable
- Check the flank against pitting, scuffing and wear in parallel
Contact stress verification sits in the product development loop alongside bending, scuffing and wear, and it decides to a large degree the cost of a gearbox, because the flank quality and material that pass the contact check are the expensive parts of the gear. A verified contact design is not a gear that simply fits its envelope; it is a pair of surfaces engineered to survive ten million cycles of rolling contact while the rest of the machine ages around them.
A final judgement from practice: the contact stress that fails in service is rarely the one the spreadsheet calculated, because in service the alignment moved, the load spiked and the lubrication film thinned. The verified design leaves margin for those three, and the longest lasting gearboxes are built with alignment and lubrication treated with the same seriousness as the mathematics of the flank.
7. Common Verification Pitfalls
Three mistakes account for most failed flank verifications and most early pitting in service. The first is using the pitch circle pressure instead of the operating pressure angle, which overstates the effective curvature and understates the stress for profile shifted gears. The second is applying the rated torque without the application and dynamic factors, which turns a gear that survives its nominal life into one that pits after a week on a demand without mercy. The third is verifying the pinion and the gear with the same allowable, when a case hardened pinion meshing with a through hardened gear carries most of the risk on the softer member.
The reported numbers need care with units and with the position of the check. Contact stress varies along the line of action, and different standards check at different points, the pitch point in the classical calculation, the lowest point of single tooth contact in the conservative branch, and each yields a different sigma H for the same pair. The verification must state which point and which standard it used, so that the safety factor can be compared with the factory acceptance or the field experience that produced the allowable.
When the safety factor comes out marginal, the cheapest levers are usually profile shift, a modest increase in center distance or a change in the lubrication specification, and the most expensive is changing the material and case. The verification is a design conversation, not a rubber stamp: it should be run early with nominal values, repeated when the geometry is fixed, and finally verified when the manufacturing route and the duty cycle are known, because the contact stress verdict is a summary of the whole engineering of the mesh.
