How to Calculate Belt Drive Power, Tension and Pulley Sizing

Before computing tensions, establish the belt velocity and the design power — the service-factor-adjusted load the belt must transmit. Belt velocity v is set by the driver pulley diameter D₁ and driver speed n₁. Higher belt velocity means more power for the same tension, but above roughly 25 m/s centrifugal effects dominate and the belt begins to lift off the pulley.

The service factor Ks (≥ 1.0) accounts for load variability: 1.0 for smooth, steady power (fans, centrifugal pumps); up to 1.5–2.0 for shock or reversing loads (reciprocating compressors, heavy conveyors). The effective pull Fe — the net tangential belt force doing useful work — follows directly from design power and belt velocity. Ks is applied to the nominal power before calculating Fe, so the resulting tensions already carry the service margin.

For an open belt drive the wrap angle on the smaller (driver) pulley α₁ is always less than 180°. As the speed ratio D₂/D₁ increases the wrap angle on the small pulley decreases, which reduces the maximum transmissible tension ratio before slip. Industry practice flags a warning when α₁ < 120° (≈ 2.09 rad) and recommends adding a tensioner idler to restore contact arc.

The belt pitch length L is fixed by the pulley diameters and centre distance. It determines both the number of standard belt lengths to order and the bend frequency — how many times per second each cross-section of belt is flexed — which drives fatigue life. A longer centre distance reduces bend frequency and extends belt life, but adds weight, vibration sensitivity and space.

The Euler–Eytelwein capstan equation gives the maximum ratio of tight-side tension T₁ to slack-side tension T₂ that a friction belt can sustain without slipping. Once the effective pull Fe (= T₁ − T₂) is known from the power requirement, T₁ and T₂ follow directly from this ratio. The slack side must remain positive (T₂ > 0); if it goes negative the belt is too loose and will either sag or skip over the pulley — increase the centre distance or add a tensioner.

At higher belt speeds, centrifugal effects generate an additional tension component Tc = m_belt · v² (where m_belt is the belt mass per unit length). Centrifugal tension increases T₁ without increasing the transmitted force — it is a dead load that both belt sides carry equally, raising fatigue stress without adding power capacity. The shaft bearing load equals T₁ + T₂ (conservative, straight-line assumption) and must be carried into the shaft and bearing design.

The driven pulley diameter D₂ is set by the desired speed ratio: D₂ = D₁ · (n₁/n₂). In practice you select D₁ and D₂ from the standard sheave-size series for the chosen belt section (A, B, C, D for classical V-belts; SPZ, SPA, SPB, SPC for narrow-section). The minimum recommended pitch diameter for each V-belt section ensures the bending stress in the belt does not become excessive — running a V-belt on an undersized pulley is one of the most common causes of premature fatigue failure.

ISO 4184 / DIN 7753 tabulates rated power per belt as a function of small-pulley diameter and belt speed. The number of belts required is N_belts = P_design / (rated power per belt × correction factors). The MechanixCalc belt calculator performs the tension, wrap angle and belt-length calculation and also flags whether the chosen V-belt section and pulley size meet the ISO 4184 section selection guidance.