Lecks an der Ausgangsdichtung des Cycloidalreduktors: Was man vor dem erneuten Austausch überprüfen sollte

A cycloidal reducer weeping oil at the output end is a repair that repeats. The seal gets replaced, the machine runs dry for a few weeks, the weep comes back, and the second replacement is done by someone who has now formed an opinion about seal quality.

The opinion is usually wrong, and it is wrong in a specific way. A lip seal at the output end is the cheapest, most accessible component in the whole drive train, so it is the first one replaced, and it is also the one that fails first when something else is wrong. Replacing it changes the symptom for exactly as long as it takes the underlying cause to damage the new seal.

The productive approach is to treat the leak as evidence and work out what the evidence is consistent with. Housing pressure, shaft condition, Exzentrizität, Orientierung, and the state of the seal itself are the mechanisms that account for most output-end leaks, and they leave different traces. Lubricant incompatibility with the seal compound, a seal fitted with a damaged lip, and contamination under the lip belong on the same list, and they are why a leak that follows the checks below without an obvious finding still needs the seal examined rather than another seal ordered.

What the Seal Actually Does

A radial lip seal works by maintaining a controlled interference between the lip and the shaft surface. The lip is loaded against the shaft by its own elasticity plus a garter spring, and it seals by keeping a thin film of oil in the contact rather than by wiping the shaft dry. That film is supposed to sit around a micrometre or less and to be pumped back inward by the asymmetric lip geometry as the shaft turns.

Three things destroy that arrangement. The lip loses its loading, either through wear, hardening from heat, or chemical attack from the oil. The shaft surface loses its finish, so the lip cannot maintain the film, or gains a groove where the lip has been running, so the contact is on a step rather than a cylinder. The shaft stops running true, so the lip is being flexed once per revolution by an eccentric motion instead of rotating against a concentric surface.

The third is the one that gets missed, and it is the one with the strongest published support.

The Evidence That Eccentricity Drives Leakage

Work in the tribology literature on radial lip seals under dynamic eccentricity points in one direction. Leakage clearance forms mainly during the release phase of the seal's cycle, and both increasing shaft speed and increasing eccentricity raise the probability that it forms, with gap size and leakage rate growing accordingly. Across that body of work, eccentricity comes out as the more influential of the two.

That is a mechanism, not a correlation. An eccentric shaft forces the lip to open and close once per revolution. During the opening phase the contact pressure drops and the film is no longer controlled. If the eccentricity is large enough relative to the lip's ability to follow the shaft, the seal leaks, and it will leak regardless of how good the seal is.

Bench work on seal test rigs that can superimpose reciprocating motion on rotation, and vary shaft-to-bore misalignment and shaft eccentricity, points the same way. Seals run under combined reciprocating and rotating motion fail far sooner than seals run under pure rotation against a hard shaft surface, and the failure mode is lip wear with loss of elasticity rather than any surface defect.

Note what that says about cause and effect. The lip wore out. The wear was not a material defect. It was the result of a motion the seal was never designed to accommodate.

There is also design evidence that runout is a recognized constraint rather than a theory. A published patent for a flexible-lip rotary shaft seal describes a lip expected to encounter significant shaft runout, and incorporates a specific lip geometry to prevent the sealing lip from flaring excessively when that happens. Seal designers build for runout because runout happens.

Each Mechanism Leaves Its Own Trace

The practical value of separating mechanisms is that each one leaves evidence you can collect without disassembling anything. The table below is the shortlist of readings that can be taken with the unit in place, and it covers the mechanisms that are both common and distinguishable from the outside. It does not cover all of them. Lubricant incompatibility with the seal compound, a seal fitted with a damaged lip, and contamination that has worked under the lip leave no signature that can be read from outside the housing, which is why a leak that works through the list without an obvious finding still needs the seal itself examined rather than another seal ordered.

Two of these rows deserve more than a table entry.

The breather is the most overlooked cause

Some oil-lubricated reducers use a breather; others are sealed or use a different pressure-management arrangement. The correct configuration is model-specific. Where the manufacturer specifies a breather, blockage or an incorrect component may contribute to housing pressure and leakage, but neither the presence of a breather nor a leak location alone identifies the cause.

The signature is suggestive rather than conclusive. The unit does not leak when cold and leaks after running for a while, and the seal that was replaced fails again, because the pressure, if that is what is happening, is still there. An audible release on opening a filler cap is sometimes described as confirmation, but it is a clue rather than proof: it says the housing was above atmospheric pressure at that moment, and it says nothing about how far the seal has already been affected or what else is contributing.

The corresponding check costs no parts: with the unit stopped, cooled, and depressurised, and following the approved procedure for that model, remove the breather and confirm it is open, sauber, and the correct element for this model. A blocked or incorrect breather is a finding to compare with the model documentation. Any pressure measurement or corrective test must use the specified point and approved method. A change after approved pressure correction is consistent with pressure contributing; it does not establish that pressure was the sole cause or exclude a damaged seal, wear, material incompatibility, or incorrect installation.

Orientation changes the load on the lip

A reducer designed for horizontal mounting and installed vertically is a different seal application. In a vertical arrangement the output seal often sits low, where oil is continuously present against the lip rather than splashed against it intermittently. Static head pressure acts on the lip continuously, and the lip is being loaded in a direction and manner it was not designed for.

Vertical installation also changes the lubrication of the seal itself. A lip seal relies on a small amount of oil reaching the contact to stay cool and to maintain its film. At the top end of a vertical shaft, that supply can be marginal; at the bottom end, it can be excessive.

If a unit has been reoriented, the correct response is to check whether that orientation is within the manufacturer's permitted mounting positions and whether the oil fill level specified for that orientation has been used. Filling to the horizontal-mount mark on a vertically mounted unit overfills it, and the excess has to go somewhere.

The Bearing Question That Comes First

Because eccentricity is the strongest documented mechanism, the bearing condition is the first thing to establish, not the last.

Output bearing wear produces radial play, and radial play at the bearing becomes eccentric motion at the seal track, multiplied by the distance from the bearing to the seal. A few hundredths of a millimeter of bearing play can be multiplied at the seal track depending on the overhang. That is enough to matter, given that the lip is designed to follow a shaft with a runout measured in hundredths of a millimeter.

The measurement is straightforward in principle. With the unit stopped and the shaft free to rotate, mount a dial indicator on the housing and read the shaft at the seal track while turning by hand. Then load the shaft radially, in the direction the belt or coupling pulls, and read again. The first reading is the total indicated runout at that point. The change produced by the radial load mixes bearing play with elastic deflection of the shaft, movement in the mounting, and any clearance in the assembly, so it is an indicator of how much the shaft can move rather than a direct reading of bearing internal clearance. Where the model's documentation specifies a procedure for checking bearing play, that procedure is the one that gives a number comparable with a limit.

If the total runout is within the seal manufacturer's shaft requirement and the shaft movement under radial load is negligible, the seal becomes the leading candidate. Even then the other paths are not closed: Oberflächenoberfläche, a groove under the lip, lubricant compatibility with the compound, and contamination are separate from runout and have to be read separately. If either reading is out, a new seal is a consumable, not a repair.

The Shaft Surface Is the Second Thing

A shaft that has been running with a lip seal for years develops a wear track. The lip polishes the surface and, over time, cuts a shallow groove. A new seal installed over that groove has its lip sitting in a depression with a sharp shoulder on each side, and the lip cannot maintain a uniform film across the step.

This is why the same seal location leaks repeatedly after replacement. Measuring the shaft diameter at the seal track against the diameter outside it, and checking surface roughness against the seal manufacturer's specification, settles it. Where a groove exists, options include a wear sleeve that presents a fresh surface, repositioning the seal to run on an unworn section, or replacing the shaft. Selecting a different seal compound does not help.

Scoring is different from grooving and more serious: it comes from a hard particle trapped under the lip, or from assembly damage. A scored shaft destroys new seals quickly, and it will do so regardless of seal quality.

Lubrication Configuration Has to Be Confirmed, Not Assumed

Cycloidal reducers are supplied with either oil bath or grease lubrication depending on model and application, and the two have entirely different sealing requirements. A grease-lubricated unit does not have the same internal pressure behavior as an oil-filled one, and it does not tolerate the same oil level.

Published work on cycloidal drives notes that grease lubrication is often chosen to reduce the need for disassembly and relubrication during service, but that starved conditions can develop in operation, and that the inlet oil film thickness affects the central film, friction coefficient, and friction power loss. That is a reminder that the lubrication state of a cycloidal drive is a variable, not a constant.

Which lubricant, what quantity, and which fill level for which mounting orientation are properties of the specific model. They are stated in that model's documentation and they do not transfer between models or between sizes in the same series. A leak investigation that assumes a fill level, rather than looking it up, is investigating the wrong machine.

Working Through It in Order

The efficient sequence for an output-end leak on a cycloidal reducer runs from the cheapest check to the most invasive.

1. Verify the lubricant and the fill level for this model, this size, and this mounting orientation. From the model's documentation. Correct any deviation before going further.
2. Check the breather. With the unit stopped, cooled, and depressurized, remove and inspect it, and confirm it is open and the correct element for this model. If internal pressure is suspected, establish it from the model's specified measuring point or an approved method rather than by loosening a fastener in service.
3. Measure shaft runout at the seal track and bearing radial play. Establish whether the shaft is running true.
4. Inspect the shaft surface at the seal track for groove depth, Punktewertung, and roughness.
5. Confirm the installed orientation is permitted and that the oil level corresponds to that orientation.
6. Only now assess the seal itself, including whether the correct seal type, Material, and dimensions were used for the oil and the temperature.
7. After any seal replacement, recheck steps 3 and 4. If the runout or the surface is out of specification, the replacement will fail again, and doing it a third time will produce the same result.

The order matters because the cheap, fast actions are at the top and the expensive, invasive ones are at the bottom, and because the diagnosis is only valid if the easy causes have been excluded first.

Where the sequence ends with a replacement rather than a repair, the unit this maps onto is a Zykloidaler Radreduzer, and the information worth carrying into that conversation is the model, the mounting orientation, the lubricant type, photographs of the leak, the repair history, the runout and play readings, and the output shaft connection.

Before the Next Seal Comes Off

If this unit has already had a seal replaced at the output end, the next action is not another replacement. It is one hour with a dial indicator.

Mount the indicator on the housing, indicate the shaft at the seal track, and turn the shaft by hand through a full revolution. That reading is the total indicated runout: Geometrie, Lagerzustand, and whatever the shaft itself contributes. Then press the shaft radially, in the direction the belt or coupling actually pulls it, and read again. The change between loaded and unloaded tells you how much the shaft can move, which spans bearing play plus deflection and assembly clearance. Two numbers, one instrument, no disassembly.

The two numbers narrow the field without closing it. If the total runout is within the seal manufacturer's shaft requirement and the shaft movement under radial load is negligible, the seal becomes the leading candidate and the question turns to whether the right seal was specified for this oil at this temperature, with the surface under the lip still to be read. If runout is out but shaft movement is not, the shaft itself or a bend is the more likely cause, and a new seal will not last long because the lip is being flexed once per revolution. If both are out, the bearing is where the repair starts, because a bearing that lets the shaft orbit will keep working on every seal installed after it.

Begin with the model documentation: confirm whether the unit is vented, sealed, or uses another pressure-management design. Inspect only as directed, with the unit stopped, isoliert, cooled, and depressurised where required. If an approved pressure correction changes the leak, pressure may have contributed; continue to assess the seal, shaft surface, Installation, lubricant compatibility, and contamination against the applicable model limits.

Initial observations can help organize the investigation, but allowable shaft runout, lubricant quantity, seal requirements, and housing pressure arrangements must be checked against the applicable model documentation before confirming a root cause.

Where the readings clear the shaft and the bearing, the seal becomes the leading candidate, and what remains is a specification question that the sixth step deals with: right seal type, right compound for this oil at this temperature, right installation. Leading candidate is as far as the evidence goes. Readings that come back clean do not rule out a lubricant that is attacking the compound, an installation that damaged the lip, or contamination already under it, because none of those leave a trace that a dial indicator can pick up. Where the readings do not come back clean, the seal is a consumable and the repair starts somewhere else, and the conversation worth having is about the shaft, the bearing, and the surface under the lip rather than about another seal. The contact page takes that conversation with the readings attached.