1. The Field Problem: Two Backstops, One Doing All the Work

A dual-drive incline conveyor with two 250 kW drives, each gearbox fitted with its own backstop, survived a power failure — but the right-hand backstop was destroyed and the left one was untouched. Both units were identical, both had been selected on the same calculation, and both passed their factory tests. What differed was the engagement angle: the right-hand unit locked about 1.3° earlier, and in the milliseconds before the left unit caught, the right unit absorbed the entire reverse load of the loaded belt. It failed; the left unit then held the belt alone and survived because the transient had already been spent.

Multiple backstops do not share load because they are the same size. They share load only if their engagement is coordinated and the structure between them is compliant enough to redistribute, and this must be verified — not assumed.

2. Mechanism: Why Load Sharing Is Not Automatic

2.1 Every backstop has an engagement angle

A sprag backstop does not hold at zero rotation. It must rotate a small amount — typically 0.3° to 2° at the shaft — before the sprags are fully wedged. Two units will never have identical values; even from the same batch, manufacturing tolerance, sprag seating and lubrication give a spread of several tenths of a degree.

2.2 The first unit takes everything

On reversal, the belt moves until the unit with the smallest engagement angle locks. From that instant it carries 100 percent of the reverse load until the structure stretches enough to bring the second unit's sprags into contact. The load transfer mechanism is elastic: belt elongation, shaft torsion, coupling clearance and housing deflection.

2.3 Dual-drive adds two more asymmetries

• Different tension side. In a dual-drive head arrangement the two gearboxes see different belt tensions, so their torque contributions during driving are already unequal, and reverse load follows the same split.
• Control asymmetry. Two VFDs rarely decelerate identically; the drive that stops first hands the load to the other before the backstop has even engaged.
• Structural path. If one torque arm reacts into a stiff frame and the other into a flexible section, the second unit is effectively softer and sees less load.

3. Worked Example: Quantifying the Split

Two backstops, each rated 8 kN·m. Total reverse load at the belt 40 kN, split ideally 20/20 kN. Effective reverse-path stiffness of each drive train 2,000 N/mm. Engagement offset between the units corresponds to 8 mm of belt take-up.

An 8 mm take-up — well within normal belt elasticity and sprag engagement tolerance — converts a 50/50 design into a 70/30 reality and overloads the first unit by 40 percent before any shock factor is applied. Add the transient peak of 1.8× and the first unit sees 250 percent of its intended share. This is the mechanism behind the field failure described above, and it is entirely calculable at the design stage.

4. Engineering Solutions

• Prefer one backstop where physically possible. If a single unit can be mounted on the common drive or head pulley shaft, all the load-sharing uncertainty disappears at zero cost.
• If two are unavoidable, size each for 100 percent of reverse torque. Full redundancy is the only assumption-free approach; anything less requires a verified load-sharing test.
• Match the pair. Order units from the same production batch, ask for the measured engagement angle of each unit, and pair the closest two.
• Reduce the offset's effect. Increase system compliance deliberately (for example, mount both on low-speed shafts where the belt and structure compliance dominate) or reduce reverse speed so transfer occurs at low dynamic load.
• Coordinate the drives. Set identical deceleration ramps and stagger the backstop engagement by control, not by luck.
• Verify with instrumentation. Strain-gauge both torque arms during a commissioning reverse-load test; the ratio of measured reactions is the only trustworthy evidence of load sharing.

5. Series Selection and Verification

Ask the supplier for: the measured engagement angle (not the nominal range) of each unit supplied, the torque-versus-angle characteristic, the torque arm reaction force at rated and peak torque, whether the units are batch-matched, and a recommended commissioning reverse-load test procedure. In multi-backstop systems, ask additionally for the supplier's written position on load sharing and whether they warrant individual units for full reverse torque.

6. FAQ

Q: We fitted two identical backstops — is that not redundancy?
A: Only if each is rated for the full reverse torque. Two half-rated units that engage at different angles give you one overloaded unit and one spare.

Q: Can I adjust the engagement angle in the field?
A: Not by adjusting the backstop. You can influence the split by the mounting compliance and by pairing matched units, but the engagement angle itself is set by the design.

Q: Does a bigger engagement angle difference matter more at high or low speed?
A: More at high reverse speed, because the take-up happens dynamically and adds shock. Slow, controlled reversal is far gentler on multi-backstop systems.

Q: What is the minimum evidence I should demand?
A: A commissioning reverse-load test with instrumented torque arms, recorded reaction forces on both units, and the calculated split. Without it, load sharing is an assumption.