A stacker-reclaimer gearbox backstop, 6,500 N·m class, was opened at a scheduled overhaul after 26 months. The inner race showed a band of shallow pitting about 8 mm wide, exactly where the sprags sit when locked, with matching crescents on the sprag contact faces. Measured backlash was still within tolerance, the oil was clean, and the bearing had never slipped. By any functional definition the unit was still working — but the pitted band is the first stage of a failure that ends in spalling, sprag skid and, eventually, a runaway belt.
Contact fatigue in sprag backstops is a slow, silent, predictable process. It is also the failure mode that distinguishes a heavy-duty backstop from a catalogue page with a large torque number.
Each sprag touches the inner and outer race on a crowned line. The peak Hertzian pressure is:
pmax = √( w·E′ / (π·R′) )
with w = load per unit contact length, E′ = equivalent elastic modulus (about 113,000 N/mm² for steel on steel) and R′ = equivalent radius of curvature.
Take a sprag carrying 3.0 kN over a 12 mm contact length (w = 250 N/mm), sprag crown radius 6 mm against a 40 mm race radius, so R′ = 5.2 mm:
pmax = √( 250 × 113,000 / (π × 5.2) ) ≈ 1,310 MPa
For heavy-duty units operating near rated torque, with fewer sprags sharing load, 2,000–3,000 MPa is normal. Through-hardened 100Cr6 or case-hardened 20CrMnTi races are specified at 59–62 HRC partly because the allowable contact stress sits only modestly above these values — the design margin lives in hardness and surface quality, not in size.
The dangerous stress is not at the surface where you can see it. Maximum alternating shear stress occurs at roughly 0.5–0.8 times the contact half-width below the surface, at about 0.3 pmax. At 1,310 MPa surface pressure that is roughly 400 MPa of alternating shear at a depth of tens of microns — precisely where non-metallic inclusions, carbide networks and grinding burns live.
| Condition | Sprags sharing load | Load per sprag | pmax (MPa) | Relative life |
|---|---|---|---|---|
| Perfect alignment, rated torque | 18 of 18 | 3.0 kN | 1,310 | 1.0 (reference) |
| 0.3° angular error, uneven engagement | 11 effective | 4.9 kN | 1,680 | ≈ 0.35 |
| Above plus peak torque 1.8× | 11 effective | 8.8 kN | 2,250 | ≈ 0.10 |
| Race hardness 61 → 56 HRC | 18 | 3.0 kN | 1,310 | ≈ 0.5 |
Read the table as a warning: the torque calculation can be perfect and the life can still collapse by a factor of ten through geometry and heat treatment. That is why heavy-duty backstop specifications must include alignment limits and hardness records, not only holding torque.
| Series | Contact-stress character | Where it is the right answer |
|---|---|---|
| BS | Standard sprag; highest stress per element for a given torque | Light-to-medium duty, infrequent locking |
| BSEU | Sealed sprag; same contact mechanics, protected from debris ingress | Dusty/humid sites where oil cleanliness cannot be guaranteed |
| RSBW | High torque density sprag, oil lubricated; excellent film renewal for long continuous freewheeling | Gearbox input shaft duty with controlled transients |
| FRHD | Roller-type, load shared over several elements; lowest stress per element in heavy duty | Mining, bulk terminals, high torque with limited space |
Documentation package to request: race and sprag material grade with hardness traverse (surface and core), case depth, retained austenite content, surface finish Ra of the contact band, nital-etch result for grinding burn, calculated contact stress at rated and peak torque, allowable misalignment, and a contact-pattern photograph from a dismantled sample unit. Any supplier who cannot produce a hardness traverse is guessing.
Q: Is pitting a manufacturing defect or an application problem?
A: Usually an application problem expressed through the surface. Check sprag contact pattern first: even marking with pitting suggests material or lubrication; uneven marking points to geometry.
Q: Can pitting be prevented by better grease?
A: Only partially. Lubricant decides the surface stress concentration; it cannot compensate for a 30 percent hardness deficit or a 50 percent load concentration.
Q: How much does hardness really matter?
A: A few HRC points can halve contact-fatigue life. Hardness is the cheapest reliability feature in a backstop and the easiest one to falsify, so always request the traverse, not the claim.
A stacker-reclaimer gearbox backstop, 6,500 N·m class, was opened at a scheduled overhaul after 26 months. The inner race showed a band of shallow pitting about 8 mm wide, exactly where the sprags sit when locked, with matching crescents on the sprag contact faces. Measured backlash was still within tolerance, the oil was clean, and the bearing had never slipped. By any functional definition the unit was still working — but the pitted band is the first stage of a failure that ends in spalling, sprag skid and, eventually, a runaway belt.
Contact fatigue in sprag backstops is a slow, silent, predictable process. It is also the failure mode that distinguishes a heavy-duty backstop from a catalogue page with a large torque number.
Each sprag touches the inner and outer race on a crowned line. The peak Hertzian pressure is:
pmax = √( w·E′ / (π·R′) )
with w = load per unit contact length, E′ = equivalent elastic modulus (about 113,000 N/mm² for steel on steel) and R′ = equivalent radius of curvature.
Take a sprag carrying 3.0 kN over a 12 mm contact length (w = 250 N/mm), sprag crown radius 6 mm against a 40 mm race radius, so R′ = 5.2 mm:
pmax = √( 250 × 113,000 / (π × 5.2) ) ≈ 1,310 MPa
For heavy-duty units operating near rated torque, with fewer sprags sharing load, 2,000–3,000 MPa is normal. Through-hardened 100Cr6 or case-hardened 20CrMnTi races are specified at 59–62 HRC partly because the allowable contact stress sits only modestly above these values — the design margin lives in hardness and surface quality, not in size.
The dangerous stress is not at the surface where you can see it. Maximum alternating shear stress occurs at roughly 0.5–0.8 times the contact half-width below the surface, at about 0.3 pmax. At 1,310 MPa surface pressure that is roughly 400 MPa of alternating shear at a depth of tens of microns — precisely where non-metallic inclusions, carbide networks and grinding burns live.
| Condition | Sprags sharing load | Load per sprag | pmax (MPa) | Relative life |
|---|---|---|---|---|
| Perfect alignment, rated torque | 18 of 18 | 3.0 kN | 1,310 | 1.0 (reference) |
| 0.3° angular error, uneven engagement | 11 effective | 4.9 kN | 1,680 | ≈ 0.35 |
| Above plus peak torque 1.8× | 11 effective | 8.8 kN | 2,250 | ≈ 0.10 |
| Race hardness 61 → 56 HRC | 18 | 3.0 kN | 1,310 | ≈ 0.5 |
Read the table as a warning: the torque calculation can be perfect and the life can still collapse by a factor of ten through geometry and heat treatment. That is why heavy-duty backstop specifications must include alignment limits and hardness records, not only holding torque.
| Series | Contact-stress character | Where it is the right answer |
|---|---|---|
| BS | Standard sprag; highest stress per element for a given torque | Light-to-medium duty, infrequent locking |
| BSEU | Sealed sprag; same contact mechanics, protected from debris ingress | Dusty/humid sites where oil cleanliness cannot be guaranteed |
| RSBW | High torque density sprag, oil lubricated; excellent film renewal for long continuous freewheeling | Gearbox input shaft duty with controlled transients |
| FRHD | Roller-type, load shared over several elements; lowest stress per element in heavy duty | Mining, bulk terminals, high torque with limited space |
Documentation package to request: race and sprag material grade with hardness traverse (surface and core), case depth, retained austenite content, surface finish Ra of the contact band, nital-etch result for grinding burn, calculated contact stress at rated and peak torque, allowable misalignment, and a contact-pattern photograph from a dismantled sample unit. Any supplier who cannot produce a hardness traverse is guessing.
Q: Is pitting a manufacturing defect or an application problem?
A: Usually an application problem expressed through the surface. Check sprag contact pattern first: even marking with pitting suggests material or lubrication; uneven marking points to geometry.
Q: Can pitting be prevented by better grease?
A: Only partially. Lubricant decides the surface stress concentration; it cannot compensate for a 30 percent hardness deficit or a 50 percent load concentration.
Q: How much does hardness really matter?
A: A few HRC points can halve contact-fatigue life. Hardness is the cheapest reliability feature in a backstop and the easiest one to falsify, so always request the traverse, not the claim.