Shaft alignment controls the relative position of coupled centerlines in an engineering-defined condition after removing mechanical causes that make measurement unstable. Offset, angularity, thermal movement, soft foot, pipe strain, base condition, bolt torque, and runout all participate in the decision. A green screen has little value if the machine moves when the flange connects, if tightening changes geometry, or if the target belongs to another thermal state.
Senior maintenance leadership needs two separate outputs. The first is documented geometric acceptance with planes, sign convention, foot distances, moves, and final values. The second is operating effectiveness after startup, observed through vibration, temperature, sealing, current, and stability. Combining them can close an order with attractive numbers while returning a train that is still loaded by its base or piping.
Define the target condition before uncoupling
The strategy states whether measurement occurs cold and what relative position is required for centerlines to meet in service. Where thermal growth matters, the cold target comes from manufacturer data, an engineering calculation, or validated measurement. Copying compensation from a similar machine ignores bearing height, temperature, materials, anchoring, and movement direction.
The job identifies stationary and movable machines, horizontal and vertical planes, and positive direction. Offset describes parallel separation at a reference plane; angularity describes slope difference. Converting those values into foot moves requires distances from the measurement plane to each support. A report that says only “0.05 mm” leaves future crews without enough geometry to reproduce the result.
Mechanical gates that protect the measurement
Base and hold-down
Before calculating shims, the crew checks cleanliness, corrosion, welds, paint, burrs, baseplate, anchors, and bolt condition. A base that deforms during torque changes position after every attempt. Shims cover the support, remain clean, and avoid unstable stacks. Maximum count, material, and replacement of damaged pieces follow the plant's approved practice.
Soft foot
Each support is tested with a controlled tightening sequence. Lifting one bolt can indicate height difference, angular foot, warped base, external strain, or interaction with another foot. Correction is confirmed after final torque. Compensating soft foot through alignment moves without correcting its source transfers distortion to the casing and makes the result depend on bolt sequence.
Piping and coupling condition
Pipe strain is evaluated under a safe procedure with the train supported and no forced flange fit. Movement observed as the connection is released or remade belongs to the assembly diagnosis. Pulling a mislocated pipe into an aligned pump loads nozzles, bearings, and seals. Shaft runout, coupling fit, and face condition are also checked so the measurement system does not follow a separate mechanical error.
| Gate | MP-47 evidence | Decision |
|---|---|---|
| Base | No crack; paint under right rear foot | Remove paint and clean before shimming |
| Soft foot | 0.18 mm at right rear foot | Correct to 0.04 mm under internal criterion |
| Piping | 0.14 mm vertical move when flange remade | Return adjustment to piping owner |
| Runout | 0.03 mm shaft and 0.02 mm hub | Proceed under approved case criterion |
| Thermal target | Cold vertical offset of +0.06 mm | Load documented target into work package |
Measurement method and data quality
Laser systems or dial indicators can produce valid alignment when mounting, geometry, resolution, competence, and backlash control are adequate. Laser equipment simplifies calculation and documentation but cannot repair a poor base. Dial setups require bracket sag, backlash, and readings at defined positions to be controlled. Selection considers access, coupling diameter and length, available rotation, and the accepted procedure.
Heads or brackets must remain firm. The crew rotates shafts together where required, avoids independent coupling movement that invalidates geometry, and repeats acquisition to test stability. If two sets differ beyond the repeatability criterion, the job returns to mechanical gates. Averaging unstable readings does not turn weak mounting into trustworthy data.
Completed case: MP-47 motor and pump
Didactic data: MP-47 is a centrifugal cooling-water pump driven by a 110 kW motor at 1,780 rpm. After a motor bearing replacement, startup was held because initial measurement found +0.32 mm horizontal offset, -0.18 mm vertical offset, 0.42 mm per 100 mm horizontal angularity, and 0.31 mm per 100 mm vertical angularity. These values demonstrate the method and are not a customer result, universal limit, or commercial claim.
The motor was the movable machine. Distance from coupling plane to front feet was 310 mm and to rear feet 760 mm. The cold target included +0.06 mm vertical offset from the train's engineering record. Inspection found 0.18 mm soft foot at the right rear support and 0.14 mm vertical movement when piping was remade. The team suspended movement calculations until both causes were treated.
Paint was removed, the support was inspected, and one full-size shim replaced two deformed leaves. Soft foot fell to 0.04 mm after torque. The piping crew corrected support and spool position; reconnection then moved the pump by 0.02 mm, within the case's internal criterion. Only after those gates passed did the crew acquire the set used to calculate moves.
Post-gate acquisition used by the calculation
After soft-foot correction and reduction of piping movement, the repeated set recorded at the coupling plane: horizontal O = +0.059 mm and A = +0.0422 mm per 100 mm with target T = 0; vertical O = +0.056 mm and A = -0.0311 mm per 100 mm with thermal target T = +0.060 mm. The convention placed x = 0 at the coupling, positive x toward motor feet, horizontal positive to the right, and vertical positive upward.
Each foot correction used C(x) = -[(O - T) + A × x], with A expressed in mm/mm. Negative horizontal C means a left move; positive vertical C means adding shim. Therefore, CH(310) = -[0.059 + 0.000422 × 310] = -0.190 mm and CH(760) = -0.380 mm. Vertically, CV(310) = -[(0.056 - 0.060) - 0.000311 × 310] = +0.100 mm and CV(760) = +0.240 mm after operational rounding.
Calculated moves and controlled execution
| Motor plane and support | Calculated move | Recorded execution | Final reading |
|---|---|---|---|
| Vertical, front feet | Add 0.10 mm | One full 0.10 mm stainless shim | Raw offset +0.058 mm; angular 0.003 mm/100 mm |
| Vertical, rear feet | Add 0.24 mm | 0.25 mm followed by measured adjustment | Cold thermal target retained |
| Horizontal, front feet | Move 0.19 mm left | Controlled move with jacking bolts | Offset +0.008 mm |
| Horizontal, rear feet | Move 0.38 mm left | Measured move without striking the foot | Angular 0.004 mm/100 mm |
The moves are educational and depend on the stated geometry. A calculated value does not authorize one blind displacement. After each material move, bolts are seated under the defined sequence, position is remeasured, and the next correction uses the current state. Progressive convergence reduces overshoot, excessive shim stacks, and unnecessary lateral force.
Final cold condition was +0.008 mm horizontal offset and +0.004 mm per 100 mm horizontal angularity. Vertically, raw offset was +0.058 mm against the +0.060 mm thermal target, leaving -0.002 mm residual error, with +0.003 mm per 100 mm angularity. A repeat after torque differed by no more than 0.01 mm. Acceptance cited the train's internal criterion and thermal target. No tolerance was presented as a universal rule based only on speed.
Sensitivity, uncertainty, and technical sign-off
The reviewer checked how one millimeter of distance-entry error would affect calculated foot movement and confirmed the instrument file matched physical dimensions. Ambient temperature, bracket stability, coupling backlash, and available rotation were recorded. A technically acceptable result needs enough context to explain why the crew trusts the number, not only the number itself.
Engineering sign-off covered the cold target, soft-foot result, piping displacement, final torque state, and coupling reassembly. Operations owned startup conditions, while inspection owned post-start measurements. This division prevented a technician from accepting an undocumented thermal target or an operator from changing a mechanical criterion during startup pressure.
Decision and passage into SAP PM
The technical decision was to correct soft foot and pipe strain before final alignment, accept cold geometry under the train criterion, and revise the bearing-replacement task list. Base inspection, pipe-strain check, and thermal-target record now precede foot movement. The change prevents the next crew from treating alignment as an isolated final adjustment.
The original notification retained the post-maintenance symptom and initial readings. The order contained separate operations for base inspection, support correction, piping adjustment, alignment, coupling assembly, guard installation, startup, and verification. Distances, sign convention, measurement report, and final values were attached under governance. A structured maintenance work order preserves this sequence.
Field confirmation listed removed and installed shims, torque, found condition, labor, accountable people, and test. History distinguished mechanical completion from engineering acceptance. Maintenance planning and control closed operations after checking material, report, and startup release. SAP PM remained the work system of record, not the instrument that calculated alignment.
Startup and effectiveness window
Startup occurred with the guard installed, area released, and operations present. After 30 minutes at stable condition, axial velocity at the motor bearing fell from 6.1 to 2.2 mm/s at the comparable point. Bearing temperatures remained within baseline range and the seal showed no leakage. This response supports effectiveness but does not prove that all initial vibration was caused only by misalignment.
Within 24 hours, inspection confirmed hold-down, noise, seal, and temperature. At seven days, it repeated vibration at the same positions with load recorded. At 30 days, it reviewed trend, accessible shim condition, and any piping intervention. A new flange opening restarts the external-strain question; passage of calendar time alone cannot preserve geometric acceptance.
Program measures include alignments accepted on the first stable set, orders with soft foot documented, interventions requiring piping correction, 30-day recurrence, and delay caused by unplanned prerequisites. Measuring duration alone can reward skipped checks. Rework and operating stability balance productivity with technical quality.
Limits and neighboring methods
Vibration analysis observes dynamic response and other excitation sources but does not replace measured geometry. Balancing treats mass distribution. Modal analysis investigates resonance. Base and piping verification address external force. When vibration remains high with stable alignment, investigation continues without manipulating shims merely to chase a symptom.
Flexible couplings accommodate operating displacement within specification, but that capability does not permit arbitrary installation. Inaccessible shafts, fluid-film bearings, large trains, gearboxes, or complex thermal movement may require dedicated methods and engineering. Manufacturer criteria, drawings, applicable standards, and approved plant practice prevail over generic tolerance tables.
PM Run's role in the alignment flow
PM Run is a planning, mobility, and execution layer over SAP PM. It can support order scheduling, carry operations to the field, and return notes, measurements, permitted images, PDFs, and confirmations. It is not a laser aligner, dial indicator, sensor, automatic geometry calculator, IoT system, or predictive diagnosis. Calculation, acceptance, and release belong to qualified people.
A useful demonstration begins with a real order, its operations, competence, window, and criteria. Review maintenance software integrated with SAP PM and assess how planning, mobile feedback, and history can sustain alignment discipline without replacing engineering.
Control coupling assembly and future boundary changes
Coupling assembly is part of the accepted state. The crew records spacer orientation, fastener condition, torque sequence, lubrication where applicable, and guard clearance. A geometry file captured before a forced spacer or incorrectly seated hub does not represent the returned machine. The final repeat occurs after every assembly action capable of moving a shaft or casing.
Acceptance validity ends when later work changes feet, base, coupling, piping, or machine position. The task strategy defines which interventions require a full realignment and which need a targeted confirmation. Seal replacement, flange opening, foundation repair, or movement for access cannot inherit an old report without checking the boundary that was disturbed.
Use recurrence to improve the maintenance standard
History compares the previous accepted geometry, reason for opening, and new measured state. Repeated displacement after work on the same line points toward support, assembly sequence, piping restraint, or thermal assumptions. Treating every event as routine alignment would normalize rework. Engineering instead opens the appropriate cause investigation and updates the task list only after evidence supports the change.
A periodic audit samples closed orders for physical dimensions, stable repeats, documented soft foot, piping result, thermal target authority, and post-start exposure. It also checks whether technicians used the same sign convention as the calculation file. These controls find silent data-quality failures that a final tolerance flag alone cannot reveal.
Technical references
- U.S. Department of Energy, Operations and Maintenance Best Practices Guide, a neutral source for maintenance practices.
- NASA Reliability-Centered Maintenance Guide, a public basis for function and consequence.
- SAP Help Portal, Maintenance Order, official documentation for the planning and execution object.
