Coupling Alignment Tolerance by Speed

ScopeWhether those numbers are good enough depends almost entirely on SPEED, and the tolerance tightens far faster with rpm than intuition suggests. Alignment tolerance is a statement about the cyclic bending the coupling and the shaft ends see once per revolution: double the speed and the same misalignment produces the same deflection twice as often, so the allowable misalignment falls roughly in proportion. A 0.005 in offset that is fine at 900 rpm is a failure at 3,600. The widely used field tables reflect that, -- the short-coupling table VibrAlign publishes (reproduced by NE Spintech) gives an acceptable offset of 4.0 mils at 1,200 rpm, 3.0 at 1,800 and 1.5 at 3,600 (excellent 2.5, 2.0 and 1.0), and acceptable angularity of 0.8, 0.5 and 0.3 mils per inch (excellent 0.5, 0.3 and 0.2) -- but NO TABLE IS SHIPPED HERE, because the values differ between references and between plant standards and a table copied into a calculator is a table that goes stale silently. Two things this catches. A coupling manufacturer's published capability is NOT a tolerance: couplings can accommodate far more misalignment than the bearings and seals behind them will tolerate, and the machine sets the limit, not the coupling. And for a spacer coupling the meaningful quantity is the SLOPE across the spacer rather than an offset at a plane, so a long spacer legitimately allows a large end-to-end offset while the slope stays tight -- judging a spacer coupling by its end-to-end offset condemns alignments that are fine. The value of running the check rather than eyeballing the readings is that at 3,600 rpm half a thousandth is the difference between inside and outside tolerance, and the same reading at 1,200 rpm would sit comfortably in the excellent column. This compares magnitudes against tolerances the user enters. It does not select the tolerance, which depends on the reference, the plant standard, and the machine; it does not evaluate the coupling type's own limits, the bearings, or the seals; and it says nothing about whether a passing alignment will stay passing, which is thermal growth, pipe strain, and soft foot rather than arithmetic. The machine manufacturer's alignment specification, the plant's precision maintenance standard, and the coupling manufacturer's data govern.

Formula or tablemeasured offset and angularity expressed as a percent of the acceptable and excellent tolerances entered for the operating speed; for a spacer coupling the judged quantity is the SLOPE across the spacer, = end-to-end offset / spacer length.

Source and editionThe speed-dependent alignment tolerance comparison, with the tolerance values ENTERED rather than shipped because they differ between references and plant standards. Tolerance tightens roughly in inverse proportion to speed because the misalignment is cycled once per revolution. A coupling manufacturer's published capability is NOT a tolerance. The machine manufacturer's alignment specification and the plant's precision maintenance standard govern.

Free accessA ratio against tolerances the user enters from their own adopted standard; no tolerance table is reproduced.

What governsEstimate. AHJ and licensed professional govern.

API field namesrpm, measured_offset_in, measured_angularity_mils_per_in, offset_excellent_in, offset_acceptable_in, angularity_excellent_mils_per_in, angularity_acceptable_mils_per_in, spacer_length_in, spacer_end_offset_in, offset_pct_of_acceptable, angularity_pct_of_acceptable, offset_over_by_in, offset_pct_of_excellent

• No tolerance table is shipped values differ between references and plant standardsprecision maintenance practice
• A coupling capability is not a tolerance the bearings and seals set the limit, not the couplingthe coupling manufacturer's data
• Passing now is not staying passing thermal growth, pipe strain and soft foot decide thatmillwright practice