Jaw coupling alignment checked with a precision straightedge and feeler gauge remains the baseline field technique for low-RPM pump and gearbox sets, because the method uses only a machined straightedge, a taper or leaf-feeler set, and a 0-1 inch outside micrometer, and delivers measurable offset and angularity values rather than a visual "eyeball" judgement [S1][S2][S5].
Two shafts coupled through a jaw coupling must hold offset (parallel) misalignment under roughly 1-3 mils (0.025-0.076 mm) and angular misalignment under 0.5-1.0 mils per inch of coupling diameter, because the elastomer spider in a jaw coupling only absorbs cyclic radial and axial loads when the centreline deviation stays inside the manufacturer's rating [S1][S2].
What the Straightedge-and-Feeler-Gauge Method Actually Measures
A coupling is "correctly aligned" when a straightedge laid across both coupling halves, parallel to each shaft, sits at the same distance from the shaft surface at every point around the circumference, and the axial gap between the two coupling halves is also constant around the full 360 degrees [S5]. Two distinct error modes therefore have to be captured at four clock positions: 12, 3, 6, and 9 o'clock.
Offset misalignment is the radial distance between the two shaft centrelines, reported in mils (0.001 inch) or millimetres; a 10-mil offset means the centrelines are separated by 0.010 inch at the coupling face [S1]. Angular misalignment is reported as mils per inch of coupling diameter, where 1.0 mil/in equals roughly 0.057 degrees, or about 3.4 minutes of arc, and is the gap a feeler gauge has to "make up" between the straightedge and the coupling face across the OD [S1].
Offset misalignment shows up as a 2× running-speed (2× RPM) vibration component, predominantly in the radial direction, because the cyclic bending moment on the coupling and the inboard bearings completes two cycles per shaft revolution; angular misalignment produces 1× and 2× RPM axial vibration, since the shaft ends are being pushed and pulled along the rotation axis [S1]. A tech chasing high axial vibration on a jaw coupling drive is normally looking at angularity, not parallel offset.
Step-by-Step Field Procedure for a Soft-Jaw Coupling
The first mechanical step is to slack the motor feet and bring both coupling halves together by hand so the elastomer spider is fully seated and the gap between the jaws is symmetric, then rotate both shafts to the 12 o'clock position and record the straightedge-to-shaft reading with a feeler or thickness gauge at 12, 3, 6, and 9 o'clock, repeating the rotation to 90, 180, and 270 degrees and tabulating all four numbers to the nearest 0.001 inch [S1][S2][S5].
Vertical alignment is corrected by adding or removing stainless shim stock under all four driver feet by the same thickness, which moves the entire machine parallel to itself and removes vertical offset; horizontal alignment is then corrected with a jackscrew or chain-pull arrangement that slides the driver sideways on its anchor bolts, while a second person watches the feeler readings until 12 o'clock and 6 o'clock match, and 3 o'clock and 9 o'clock match to within roughly 0.001-0.003 inch for a typical 3-6 inch jaw coupling [S1][S3].
Angularity is read directly off the straightedge placed across the OD of the two coupling flanges: if the straightedge touches both flange faces only at the top with a 0.010 inch gap at the bottom of a 4-inch coupling, the angularity is 0.010/4 = 2.5 mils/inch, which is outside the 0.5-1.0 mil/in envelope a jaw coupling spider can swallow without overheating and chunking [S1][S2]. After correction, torque all four foot bolts to the motor manufacturer's value with the coupling spacer in place, then recheck at all four clock positions, because the final tightening sequence can pull a soft-foot machine back out of spec by 1-2 mils [S1].
Where the Method Is Accurate, and Where It Is Not
The straightedge-and-feeler-gauge method is rated as a low-precision technique suitable for soft couplings, slow-speed equipment, and small machines, typically below 1,800 RPM and at tolerances no tighter than about 0.005 inch (5 mils) [S2][S6]. It is the standard procedure recommended for flexible elastomer couplings on centrifugal pumps, where KSB explicitly describes it as the preferred initial check before any dial-indicator or laser work [S5].
Field service shops report using a straightedge and eyeball judgement on small circulators and pumps up to roughly 7.5 HP, and switching to laser alignment in the 10-250 HP range because the laser captures the four corrections (vertical offset, vertical angularity, horizontal offset, horizontal angularity) simultaneously and calculates the foot moves, eliminating the "chasing your tail" loop that hand-feeler methods are prone to [S3]. The straightedge cannot resolve sub-mil parallelism on a rigid flange coupling, and it gives no usable reading on a spacer coupling where the two flange faces are not adjacent, so its real envelope is the small soft-coupled machine, not the API-class process train [S2][S6].
Comparison against the three baseline methods commonly used on motor-pump skids:
Straightedge and feeler gauge: cost under 50 USD in tooling, accuracy about 0.005 inch, single-plane only, requires two people, ideal for jaw and grid couplings below 1,800 RPM [S1][S2][S5].
Reverse-indicator (rim-and-face) dial indicator: cost roughly 500-2,000 USD for a used setup, accuracy around 0.001-0.002 inch, two-plane simultaneous, the workhorse of the 1970s-1990s alignment trade [S3].
Laser shaft alignment: cost 3,000-12,000 USD for a modern single-laser unit, accuracy 0.0005 inch or better, all four corrections in real time, dominant on 10 HP and up, on rigid couplings, and on any train where thermal growth compensation has to be entered as a known offset [S1][S3].
Soft Foot, Thermal Growth, and the Two Failure Modes That Fool the Gauge
Soft foot, a condition where one of the four driver feet does not sit flat on its shim stack and instead rocks when the opposing bolt is torqued, will generate a repeatable but false misalignment reading on a straightedge and feeler gauge, so the first mechanical check on any soft-jaw train is a rock test: loosen each foot bolt one at a time and confirm a 0.001-0.003 inch shim change does not move the indicator, which proves the machine is not standing on a corner [S1][S3].
Thermal growth is the second trap: a cold-aligned pump and motor at ambient temperature will not be aligned at operating temperature, because the pump casing grows toward the suction flange, the motor frame grows upward, and the driver centreline rotates on the order of 2-6 mils on a typical 1,500 RPM process pump [S1][S5]. Good practice is to set the cold alignment deliberately off by the predicted growth vector, so the train lands inside tolerance once both machines reach steady state, a step that a simple straightedge measurement cannot validate on its own and that has to be carried on the alignment sheet as a documented offset [S1][S5].
For hot-water pumps specifically, KSB flags the manufacturer's thermal-deformation instructions as a hard requirement, because the pump case can grow asymmetrically along the shaft axis and drive a cold-aligned coupling into a 5-10 mil angular error at temperature, which then chases the spider and the bearing housing [S5]. Pinning the pump and driver to the common baseplate only after the alignment is finalised and rechecked is the standard step that prevents the alignment from drifting as the anchor bolts are torqued [S5].
Jaw-Coupling-Specific Tolerance Numbers and What Exceeding Them Does
A typical L-series or spider-style jaw coupling is rated for roughly 1 degree of angular misalignment, but good practice targets 10-25% of that rating, not the maximum, because cyclic loading at the rated limit shortens the elastomer life and pushes heat into the hub [S1]. As a working number, the field target is offset under roughly 0.003 inch and angularity under 0.5 mil/in on a 3-6 inch jaw coupling, with the exact figure taken from the coupling OEM's data sheet, not from a generic alignment table [S1][S2].
Operating above the rated misalignment window typically shows up first as spider wear on the lobe tips, with a measurable temperature rise of 10-20 F on the coupling housing, then as bearing-race spalling on the inboard bearings, and finally as a fatigue crack in the hub at the keyway [S1]. On a disc coupling or a gear coupling the failure signature is different, but the upstream cause, alignment that drifted above the rated envelope, is the same; that is why documenting the four clock readings on the alignment sheet is non-negotiable in any predictive maintenance program [S1][S2].
One widely-quoted engineering reference point: a soft-jaw coupling aligned inside roughly 1/3 of its rated misalignment capacity will typically run 5-10 times longer between spider replacements than the same coupling aligned to its rated limit, because the bending strain in the elastomer scales with the cube of the deflection [S1]. That ratio is why a 15-minute extra shimming step with a feeler gauge is a much cheaper reliability move than the laser it eventually replaces.
Signals Worth Tracking Over the Next Two Quarters
Watch for updated OEM tolerance sheets from the major elastomer-jaw coupling lines (Lovejoy, Martin, Rexnord, KTR), which have been re-issuing installation manuals in 2025-2026 with tighter cold-to-hot alignment offset tables tied to specific frame sizes. Watch also for the next round of ISO 1940 balance-grade updates and any movement on API 686 recommended practice, both of which feed back into the alignment-target percentages that procurement specs cite on pump and motor data sheets. [S1]
Related analysis: ISO 6336 vs AGMA 2001: Diverging Gear Load-Capacity Calculations.