Installation and alignment

1. Alignment is a machine problem

Alignment is not a coupling problem. It is a machine, foundation and pipework problem that the coupling is asked to absorb the remainder of — and the largest single reason gear couplings fail early is that the remainder was larger than anyone measured.

Be clear about what the published misalignment capacity means: it is a survival limit, not an operating target. Every degree of misalignment makes each tooth slide against its mate and back once per revolution, and the rate at which the flanks wear scales with that sliding. A coupling running at its rated capacity is inside specification and wearing many times faster than one running near zero.

The capacity exists to absorb what the installation cannot control — thermal growth, foundation settlement, pipe loads that change with process conditions. Not to excuse what it can.

2. Soft foot and pipe strain come first

A soft foot is a machine foot that does not sit flat on its support, so tightening its bolt distorts the casing. Any alignment measured afterwards is a measurement of a distorted machine, and the numbers change the moment bolts are slackened, the machine is re-shimmed, or it warms up.

Checking each foot in turn — indicator on the shaft, that foot's bolt loosened and retightened — takes minutes and prevents an alignment that cannot be repeated. Both machines get checked, not only the one being moved.

Pipe strain is the same problem arriving through the flanges. Pipework pulled into position to meet a pump nozzle applies that load to the casing for the whole of its service life, moving the shaft centreline and continuing to move it as the line heats and cools. The test is simple: set indicators on the shaft, slacken the flange bolts, and watch whether anything moves. If it does, the pipework is the alignment problem, and no amount of shimming at the feet will resolve it.

3. Mounting the hubs

This is where irreversible mistakes happen. An interference-fit hub is heated to expand it, driven fully home to the shoulder in one movement, and allowed to cool in position. It cannot be adjusted afterwards without heating it again, and a hub that stopped short of its intended position leaves a coupling that is axially wrong from the start.

Induction or oil-bath heating with a controlled temperature is the correct method. Flame heating is not: it heats unevenly, will readily exceed the temperature at which the hub material is affected, and destroys any prospect of a controlled fit.

The key must be a proper fit in both keyways, with clearance at the top of the keyway rather than at the sides — the bore and keyway practice of whichever standard governs the order sets the tolerances. Check both hubs for correct axial position relative to the shaft ends before the sleeve goes on.

4. The three measurement methods

4.1 Why reverse dial beats rim-and-face

Rim-and-face derives angularity from a face reading, and a face reading moves if the shaft being rotated can float axially. A gear coupling is specifically designed to let the shafts float axially. That is not a fault in the method so much as a mismatch between the method and this coupling type — it matters little on a short, rigid connection and a great deal on a coupling with appreciable separation between hubs.

Reverse dial takes two radial readings instead and derives both offset and angularity from the geometry, so axial float does not enter the result. On dial methods of either kind, bracket sag has to be accounted for: a bracket spanning a coupling deflects under the weight of the indicator, and that deflection appears in the readings as offset that is not there. Measure it once on a length of pipe by rotating the assembly through 180°, and subtract it.

5. What tolerance to work to

Not the coupling's rated capacity — this is the point most often misunderstood. Alignment tolerance is a function of speed, not of what the coupling can survive, because sliding velocity at the flank rises with both misalignment and speed. Tighter tolerances apply at higher speed for the same coupling.

Published alignment tolerance tables express this directly. The practical rule: align to the machinery tolerance for that speed, and treat the coupling's rated capacity as the margin that covers thermal growth and movement in service.

6. The cold offset

Alignment is set cold and the machine runs hot. Casings and pedestals grow as they reach operating temperature, and on a large or hot machine the running centreline can sit measurably above where it was at commissioning.

A train aligned to zero cold is therefore a train misaligned hot — in the condition it spends nearly all its life. The correct approach is a deliberate cold offset in the direction opposite to expected growth, sized either from a growth calculation using casing dimensions and expected temperature rise, or from hot alignment readings taken on the machine itself.

Hot readings are the better source wherever they can be obtained, because they include the effects nobody calculated.

7. Closing out the job

1. Set the axial gap between hub faces to the figure specified for that coupling — the gap accommodates thermal growth of the train as well as assembly tolerance.
2. Fit the sleeve and tighten the flange bolts in a cross pattern to the specified torque. These bolts carry the full transmitted torque and are a rated part of the coupling, not fasteners from the stores bin.
3. Fill with the specified coupling grease before the guard goes on, not after.
4. Record the final cold readings, the shims fitted at each foot, the axial gap and the grease quantity. The next person on this machine needs to know what changed, and will otherwise start from nothing.