Soft foot is a condition where one or more feet of a motor, pump, or gearbox do not sit flat on the baseplate, so tightening the hold-down bolts bends the machine frame instead of just anchoring it. It sounds like a minor fitting issue, but an uncorrected soft foot silently defeats even a perfect laser alignment job, and it is one of the most common (and most overlooked) root causes of chronic vibration and early bearing failure on rotating equipment.
"Soft foot" is a generic term for machine frame distortion. Anything that causes the casing to twist or bow when a hold-down bolt is tightened (or loosened) qualifies, whether the cause is a machined foot, a bent one, or an outside force. Because most machines have four feet and any three points define a plane, fitting a fourth foot into that same plane requires clean, well-shimmed contact. If it isn't there, torquing the bolts forces the frame to conform to the uneven support, straining the casing and the bearing bores inside it.
This matters because the distortion does not stay in the foot. It travels through the frame into the bearing housings, shifting the internal alignment of the bearing bores and, in turn, the shaft centerline. That is why soft foot must be checked and corrected before you ever start a laser shaft alignment: aligning a distorted frame just locks the distortion in.
Parallel soft foot (also called a "short foot" condition) happens when one foot is machined too short or is missing shims, leaving an even, non-tapered air gap under it. With the other three feet tight, that foot rocks against the diagonally opposite foot across a high ridge formed by the remaining two feet. Torque the bolts down and the frame bows to force contact, exactly like a four-legged chair with one short leg that stops rocking only because the chair itself has flexed.
Angular soft foot occurs when the foot or its support surface is machined or bent on an angle, so only one edge or corner makes contact when the foot is unloaded. An "outside" angled foot touches only at its inner edge and, when tightened, acts like a lever that injects lifting forces into the frame, popping the other feet up. An "inside" angled foot touches only at its outer edge and does the opposite: it pulls the frame down, so loosening adjacent feet shows little or no stress relief. Feeler-gauge readings under an angled foot show a clearly tapered gap (larger on one side, smaller on the other) rather than the even gap of a parallel soft foot. The short-term fix is step-shimming (a stack of shims cut to match the taper); the permanent fix is re-machining the foot or the support pad.
Induced soft foot is not caused by the foot or base at all. It comes from an external force distorting the frame, most often pipe strain from piping that was forced into place before the bolts were tightened, or excessively rigid conduit, structural bracing, or belt and chain loads. It behaves inconsistently and can affect two, three, or all four feet without a clean rocking or tapered pattern. Because the frame may already be under significant stress, never loosen more than one foot at a time when induced soft foot is suspected. The remaining feet may not be strong enough to hold the frame against that external force, and doing so can bend, crack, or break the feet, a safety hazard as well as an equipment one. The only real fix is removing the external force, not shimming around it.
A related nuisance condition, often grouped with the same checks, is "squishy foot": too many shims (commercial precut shims allow any thickness up to 0.150 inch using three or fewer), bent or dirty shims, or debris under a foot. The stack compresses like a leaf spring when the bolt is torqued, so the fastener never fully anchors the foot even though it reads "tight."
When a machine with a soft foot condition is bolted down anyway, the casing and bearing housings deflect internally, shifting the bearing bores out of alignment with each other. That shaft distortion changes the machine's alignment relative to the coupled shaft, producing vibration that accelerates wear in the coupling, bearings, seals, and rotor. The distortion also increases the radial load the bearings must carry, and bearing life is highly sensitive to that load: for ball bearings, expected life varies inversely with the cube of the ratio of design load to actual radial load, so doubling the radial load can cut bearing life roughly eightfold, and tripling it can cut life by around 27 times.
Because the shaft must continuously flex to follow the distorted bearing bore path as it rotates, the bending reverses direction twice per revolution, at twice the running speed. On bearing failure spectra and general vibration analysis, soft foot commonly shows up as an elevated 1x running-speed component, sometimes alongside 2x or 3x running-speed peaks, and on motors it can also raise a 2x line-frequency peak because the distorted stator loses uniform clearance to the rotor. Reviewing severity against ISO 10816-3 vibration severity bands after a soft foot correction is a useful way to confirm the fix actually worked, not just that the readings moved.
The correct sequence, per standard alignment practice, is: clean the mounting area, rough-align the coupling to remove any coupling strain, then check each foot individually with the other three tight.
Once the gap shape confirms the diagnosis, shim (or step-shim, for a tapered gap) the affected foot to fill the measured gap. Good shimming practice keeps the stack simple and predictable:
| Guideline | Value |
|---|---|
| Soft foot acceptance threshold | 0.002 in (0.05 mm) or less |
| Maximum precut shims per foot (normal correction) | 3 to 4 |
| Minimum thickness for all but one shim in a stack | 0.003 in (0.08 mm) or greater |
| Combined thickness of the three thinnest shims | 0.010 in (0.25 mm) or greater |
| Typical precut stainless shim compression | about 0.5 percent to 1 percent of stack thickness |
Precut stainless steel shims compress far less than hand-cut brass shims, which is why they are preferred for anything beyond a temporary fix. After correcting soft foot, recheck all four feet to confirm the problem is actually gone, then proceed with final alignment, which should also be checked against dynamic balancing and coupling condition if vibration persists, since a leftover imbalance or coupling fault can mimic residual soft foot symptoms.
A soft foot correction takes a technician a few extra minutes with a feeler gauge and a shim kit. Missing it means the "aligned" machine is still internally stressed, running with a distorted bearing bore that no amount of coupling adjustment can fix, and heading toward a bearing or seal failure that looks unrelated to alignment work until someone finally checks the feet. Fabrico reads machine condition and OEE from the line and auto-routes a work order the moment a vibration or condition signature drifts, so a soft-foot-driven bearing problem gets caught and dispatched before it becomes an unplanned outage, using computer vision that catches wear patterns sensors alone can miss, built and hosted in the EU with EU data residency, and run under ISO 27001, ISO 20000-1, and ISO 9001. Book a Fabrico demo to see it on your own lines.
Yes. Laser alignment measures shaft-to-shaft position with the bolts tight; it does not by itself reveal that the frame is under stress. A soft foot check is a separate step, done by loosening each foot individually, and should always precede the final alignment reading.
Standard practice treats a shaft movement or laser reading greater than 0.002 inch (about 0.05 mm) at a foot as a soft foot requiring correction. Readings at or below that are generally considered within tolerance.
No. Over-torquing does not eliminate the air gap, it just forces the frame to absorb more stress trying to close it. The gap has to be physically filled with the correct shim thickness (or the foot or base corrected) before the bolt is torqued to spec.
Not necessarily. Because a machine frame is a single connected structure, correcting a rocking (parallel) soft foot often means shimming or adjusting two diagonally opposed feet, not just the one with the highest reading, since the frame relieves stress at more than one location.
Zaloguj
Blog