In my plant, the rotary kiln drive is arranged as a single-side edge drive, and the pinion gear is one of the most heavily loaded and most oil-sensitive components in the entire mechanical train. The kiln shell has a nominal diameter of 4.8 m and a length of 72 m. The pinion gear meshes with the large girth gear, and the pinion gear shaft passes through the gear cover on both sides. The original sealing arrangement at each side of the pinion gear and gear cover used a simple gland packing structure. The packing was intended to press against the pinion gear shaft, block the lubricating oil inside the gear cover, and prevent oil from migrating outward along the shaft. In practice, this single packing barrier did not provide reliable sealing. Oil leaked from both sides of the pinion gear along the shaft, and the leakage became especially severe at the low end of the pinion gear during summer operation.
The lubricant used in this drive is an imported high-performance gear oil. Because leakage was continuous and difficult to control, I had to add oil frequently. The annual oil make-up reached approximately 200 kg, and the direct cost of that make-up oil reached approximately 3.2 x 10^4 monetary units per year. This figure does not include labor, cleaning, environmental housekeeping, equipment risk, or the possibility of reduced lubrication film thickness inside the gear mesh. The leakage was not only a cost problem. It was also a reliability problem for the pinion gear, the gear cover, the foundation, and the surrounding operating area.
I analyzed the original sealing structure in detail. The pinion gear sealing system depended on gland packing being compressed against the pinion gear shaft. The packing generated a radial contact pressure, and that contact pressure was expected to be greater than the local oil pressure and capillary force. However, the contact pressure was not stable. After a period of operation, the packing wore, the compression force decreased, and the gap between the packing and the pinion gear shaft increased. Once the gap increased, oil could flow outward along the shaft. Replacing the packing restored the seal for a short time, but the leakage returned after the new packing wore in. This repeated cycle showed that the root cause was not simply poor packing quality. The root cause was the inherent limitation of a single packing seal in a large, vibrating, thermally cycled, and misalignment-prone rotary kiln drive.
The behavior of the original packing seal can be described in a simplified way. The contact pressure produced by the packing is related to the compression force and the contact area:
$$p_c = \frac{F_s}{A_c} = \frac{F_s}{\pi d_s L_p}$$
where \(p_c\) is the contact pressure, \(F_s\) is the effective compression force, \(A_c\) is the contact area, \(d_s\) is the pinion gear shaft diameter, and \(L_p\) is the effective packing length. When the packing wears, the effective compression force decreases. If the wear depth is \(\delta_w\), the contact pressure can be approximated as:
$$p_c(t) \approx p_{c0} – k_w \delta_w(t)$$
Here \(p_{c0}\) is the initial contact pressure, and \(k_w\) is a stiffness coefficient related to the packing material and the gland geometry. The important point is that \(p_c(t)\) falls with time. The sealing capability of the packing therefore falls with time, while the oil pressure and the dynamic excitation from the pinion gear continue to act. This mismatch creates a leakage path.
The leakage path between the packing and the pinion gear shaft can be treated as a narrow annular clearance. For a concentric annular gap, the laminar leakage flow rate can be approximated by:
$$Q_l = \frac{\pi d_s h^3 \Delta p}{12 \mu L}$$
where \(Q_l\) is the leakage flow rate, \(h\) is the radial clearance between the packing and the pinion gear shaft, \(\Delta p\) is the pressure difference across the seal, \(\mu\) is the dynamic viscosity of the oil, and \(L\) is the axial length of the leakage path. The cubic dependence on \(h\) is critical. If the clearance doubles, the leakage flow rate can increase by a factor of eight, all other conditions being equal. This explains why even modest packing wear can produce a visible and persistent oil leak at the pinion gear.
In a real pinion gear application, the clearance is not perfectly concentric. Shaft runout, gear mesh forces, kiln shell deflection, support roller alignment, and process instability introduce eccentricity. For an eccentric annular gap, the leakage flow rate increases further. A common approximation is:
$$Q_l = \frac{\pi d_s h^3 \Delta p}{12 \mu L}\left(1 + 1.5 e^2\right)$$
where \(e\) is the relative eccentricity of the pinion gear shaft inside the packing. The term \(1 + 1.5 e^2\) shows that eccentricity amplifies leakage. When the pinion gear vibrates, the clearance is not constant. It varies with time:
$$h(t) = h_0 + e_r \sin(\omega t) + a_v \sin(\omega_v t)$$
Here \(h_0\) is the mean clearance, \(e_r\) is the runout amplitude, \(a_v\) is the vibration amplitude, \(\omega\) is the rotational frequency of the pinion gear, and \(\omega_v\) is a vibration frequency. Since leakage depends on \(h^3\), the time-averaged leakage is larger than the leakage calculated from \(h_0\) alone:
$$Q_{avg} \propto \overline{h^3} = h_0^3 + 3h_0 \overline{\delta^2} + \overline{\delta^3}$$
This relationship explains why vibration at the pinion gear is so damaging to a packing seal. The vibration does not merely loosen the packing. It continuously opens and closes the leakage path, and the opening phase dominates the flow because of the cubic dependence. The result is a pulsing oil loss that appears as a continuous leak along the pinion gear shaft.
Temperature also has a strong effect. The viscosity of gear oil decreases as temperature increases. A simple exponential model is:
$$\mu(T) = \mu_0 e^{-b(T-T_0)}$$
where \(\mu_0\) is the viscosity at reference temperature \(T_0\), \(b\) is the temperature-viscosity coefficient, and \(T\) is the operating temperature. During summer, the ambient temperature is higher, the gear cover temperature is higher, and the oil temperature is higher. As \(\mu\) decreases, the leakage flow rate increases according to the annular gap equation. At the low end of the pinion gear, gravity also favors oil migration toward the seal. The combination of lower viscosity, higher oil mobility, and favorable gravity drainage makes the low-end seal the worst location in summer.
The original leakage problem can be summarized by the following observations:
| Item | Observation | Consequence for the Pinion Gear |
|---|---|---|
| Kiln drive type | Single-side edge drive | High torque and high radial load on the pinion gear |
| Seal type at the pinion gear | Gland packing on both sides of the gear cover | Single barrier with limited wear compensation |
| Leak location | Along the pinion gear shaft on both sides | Oil loss, housekeeping problems, and risk to the pinion gear |
| Worst location | Low end of the pinion gear | Gravity-assisted leakage, especially in summer |
| Lubricant | Imported high-performance gear oil | High replacement cost and supply sensitivity |
| Annual make-up oil | Approximately 200 kg | Direct cost near 3.2 x 10^4 monetary units per year |
| Packing replacement | Frequent replacement | Short improvement, then leakage returns |
| Operating environment | Kiln deflection, process variation, vibration, thermal cycling | Unstable contact pressure and variable clearance at the pinion gear |
I also examined the contribution of various mechanisms to the total leakage. A qualitative ranking helped me decide that adding a labyrinth seal was more effective than repeatedly replacing the packing. The following table shows the mechanism, its effect on the pinion gear seal, and the expected improvement from a labyrinth arrangement.
| Leakage Mechanism | Effect on Original Packing Seal | Expected Effect of Labyrinth Seal at the Pinion Gear |
|---|---|---|
| Direct oil throw from rotating gears | Oil reaches the shaft and then the packing | Oil strikes the moving and stationary rings and drains back |
| Gravity drainage at low end | Oil accumulates near the packing and migrates outward | Collected oil follows the ring surfaces back to the oil pool |
| Packing wear | Contact pressure decreases, clearance increases | Original packing is retained mainly for dust exclusion |
| Shaft runout | Non-uniform clearance and local leakage paths | Large clearances avoid rubbing and tolerate runout |
| Pinion gear vibration | Dynamic opening of the packing contact | Labyrinth cavities dissipate pressure and redirect oil |
| High summer temperature | Lower oil viscosity and higher leakage rate | Oil is captured before it reaches the shaft seal |
| Pressure pulsation | Single barrier exposed to full pressure fluctuation | Multiple cavities reduce pressure step by step |
| Dust ingress | Packing also acts as a dust barrier | Packing remains in place for dust control |
After evaluating the problem, I concluded that the fundamental issue was the lack of a robust oil return path. The pinion gear was throwing oil inside the gear cover, and the packing was expected to stop all of that oil at the shaft. A better approach was to let the oil be thrown, then intercept it, guide it downward, and return it to the oil pool before it could reach the shaft. This is exactly what a moving and stationary labyrinth seal can do when it is arranged around the pinion gear end face and the inside of the gear cover.
I decided to use the opportunity of a large girth gear and pinion gear flip to modify the seal. I did not change the original packing structure. Instead, I added a moving ring and a stationary ring made from 4 mm thin steel plate. The stationary ring was welded to the inside of the gear cover. The moving ring was divided into three segments and bolted to the end face of the pinion gear. The moving ring and stationary ring formed a labyrinth clearance. The radial clearance was set to 15 mm, and the axial clearance was set to 20 mm. These clearances were intentionally generous. They prevented contact between the moving ring and the stationary ring, even when the pinion gear experienced vibration, thermal expansion, or small alignment changes.
The design requirements for the new seal were as follows:
| Requirement | Target Value | Reason for the Target |
|---|---|---|
| Moving ring material | 4 mm steel plate | Light, weldable, formable, and sufficient for oil deflection |
| Stationary ring material | 4 mm steel plate | Compatible with welding inside the gear cover |
| Moving ring construction | Three segments | Allows installation around the pinion gear without full disassembly |
| Moving ring fixing | Bolted to the pinion gear end face | Rotates with the pinion gear and throws oil outward |
| Stationary ring fixing | Welded inside the gear cover | Remains fixed and forms the labyrinth cavities |
| Radial clearance | 15 mm | Prevents rubbing under runout and vibration |
| Axial clearance | 20 mm | Prevents axial contact under thermal growth and thrust variation |
| Original packing | Retained | Acts as a dust seal after the labyrinth takes over oil control |
| Oil return | Gravity return to oil pool | Keeps oil away from the pinion gear shaft |
The new arrangement works as a labyrinth seal around the pinion gear. The moving ring rotates with the pinion gear. Any oil thrown by the rotating gear mesh or by the pinion gear itself collides with the moving ring or the stationary ring. After collision, the oil loses most of its axial velocity. It then falls vertically along the ring surface instead of continuing along the shaft. The stationary ring and the gear cover form a collection path. The oil returns to the oil pool by gravity. Because the oil is intercepted before it reaches the packing, the original packing no longer has to withstand the full oil flow. The packing still performs a useful function as a dust seal, but it is no longer the primary oil seal.
The labyrinth principle can be described by pressure staging. Each cavity between the moving and stationary rings creates a local pressure drop. For a single throttling clearance, the pressure drop can be approximated as:
$$\Delta p_i = \frac{\xi_i \rho v_i^2}{2}$$
where \(\Delta p_i\) is the pressure drop at stage \(i\), \(\xi_i\) is the local loss coefficient, \(\rho\) is the oil density, and \(v_i\) is the local oil velocity. For a labyrinth with \(N\) effective stages, the total pressure difference is distributed over the stages:
$$\Delta p_{total} = \sum_{i=1}^{N} \Delta p_i$$
If the stages are similar, the pressure difference across each stage is reduced approximately as:
$$\Delta p_i \approx \frac{\Delta p_{total}}{N}$$
The oil that enters the labyrinth therefore loses pressure in steps. At the same time, the rotating ring imposes a circumferential velocity on the oil. The centrifugal effect helps to keep the oil moving outward and downward, away from the shaft. A simplified centrifugal acceleration at the ring surface is:
$$a_c = \omega^2 r$$
where \(\omega\) is the angular velocity of the pinion gear and \(r\) is the local radius. The oil film experiences this acceleration and tends to follow the rotating surfaces. The combination of pressure staging, centrifugal separation, and gravity drainage creates a strong barrier without any rubbing contact.
The return flow along the ring surface can be estimated by a thin-film gravity drainage equation. For a film of thickness \(h_f\) flowing down an inclined surface of width \(b\), the volumetric return flow is approximately:
$$Q_r = \frac{\rho g \sin\theta b h_f^3}{3\mu}$$
where \(\theta\) is the inclination angle of the drainage path. For effective oil control, the return flow capacity should be greater than the oil reaching the seal region:
$$Q_r \ge Q_{captured} + Q_{residual}$$
where \(Q_{captured}\) is the oil intercepted by the labyrinth and \(Q_{residual}\) is the small amount of oil that may still reach the packing. The generous clearances and the downward slopes of the fabricated rings help ensure that this condition is satisfied. The return flow does not require a pump. It uses gravity and the natural geometry of the gear cover.
The axial and radial clearances in the new pinion gear seal can be checked against thermal growth. The thermal expansion of a diameter is:
$$\Delta d = \alpha d \Delta T$$
For a representative steel diameter near 750 mm, a coefficient of thermal expansion \(\alpha = 12 \times 10^{-6} \text{ K}^{-1}\), and a temperature change \(\Delta T = 60 \text{ K}\), the diameter change is:
$$\Delta d = 12 \times 10^{-6} \times 750 \times 60 = 0.54 \text{ mm}$$
This value is small compared with the 15 mm radial clearance. Similarly, the axial thermal growth is small compared with the 20 mm axial clearance. The clearances therefore provide a large margin against contact. The design condition can be written as:
$$\delta_r > \Delta r_{thermal} + \Delta r_{runout} + \Delta r_{deflection} + \Delta r_{vibration}$$
$$\delta_a > \Delta a_{thermal} + \Delta a_{runout} + \Delta a_{deflection} + \Delta a_{vibration}$$
In my case, the 15 mm radial clearance and 20 mm axial clearance were chosen to be much larger than the sum of these effects. This was intentional. A labyrinth seal at the pinion gear should not be designed as a tight clearance seal. It should be designed as a robust oil interception and drainage system that tolerates the real motion of a rotary kiln drive. The clearances ensure that the moving ring never rubs the stationary ring, so there is no wear, no sparking risk, and no additional vibration excitation.
The manufacturing and installation work was planned around the existing maintenance window. The moving ring and stationary ring were fabricated from 4 mm steel plate. The moving ring was divided into three segments so that it could be installed around the pinion gear end face without removing the entire shaft assembly. The stationary ring was welded inside the gear cover. The original packing gland was left in place. The work sequence was as follows:
| Step | Activity | Key Check |
|---|---|---|
| 1 | Isolate the kiln drive and lock out all energy sources | Zero mechanical and electrical energy confirmed |
| 2 | Remove the gear cover and expose the pinion gear seal area | Safe access and support for the pinion gear |
| 3 | Measure the pinion gear shaft, end face, and gear cover bore | Record actual dimensions before fabrication |
| 4 | Fabricate the stationary ring from 4 mm steel plate | Fit to the gear cover and match the oil return path |
| 5 | Fabricate the moving ring in three segments | Segments match the pinion gear end face and bolt pattern |
| 6 | Weld the stationary ring inside the gear cover | Continuous weld, no interference with the pinion gear |
| 7 | Install the moving ring segments on the pinion gear end face | Radial clearance 15 mm and axial clearance 20 mm |
| 8 | Check runout of the moving ring with a dial indicator | Runout within allowable tolerance |
| 9 | Reassemble the gear cover and restore the oil system | No contact and no trapped foreign material |
| 10 | Start the kiln and monitor the pinion gear seal | No leakage, no abnormal temperature, no vibration increase |
During installation, I paid special attention to the original packing. The packing was not removed. It was cleaned and adjusted so that it still provided a light dust seal around the pinion gear shaft. The labyrinth rings became the primary oil-control device. This division of labor is important. If the packing is too tight, it can generate heat and wear on the pinion gear shaft. If the packing is too loose, it may allow dust to enter. With the labyrinth in place, the packing can be set with a moderate preload because it no longer has to stop the full oil flow. The packing therefore lasts longer and requires less frequent replacement.
After the modification, I monitored the pinion gear seal under different operating conditions. The observations covered normal load, high load, startup, shutdown, and hot summer operation. The leak at the low end of the pinion gear disappeared. The pinion gear shaft remained dry. Oil was found returning to the oil pool along the fabricated rings, which confirmed that the labyrinth was capturing oil and guiding it back rather than allowing it to reach the packing. The original packing remained in place and continued to act as a dust barrier. There was no rubbing between the moving ring and stationary ring, and no abnormal temperature rise was detected at the pinion gear bearing or seal area.
The improvement can be expressed quantitatively. Before the modification, the annual oil make-up was approximately 200 kg. After the modification, the make-up oil was reduced to a small amount associated with normal consumption and inspection. For illustration, if the after-modification make-up is 20 kg per year, the oil saving is:
$$m_s = m_b – m_a = 200 – 20 = 180 \text{ kg/year}$$
The direct cost saving depends on the oil price. If the effective oil cost is \(c_o\), the annual saving is:
$$S_o = m_s c_o$$
Using the pre-modification cost of 3.2 x 10^4 monetary units for 200 kg, the effective oil cost is:
$$c_o = \frac{3.2 \times 10^4}{200} = 160 \text{ monetary units/kg}$$
The annual oil cost saving is therefore:
$$S_o = 180 \times 160 = 2.88 \times 10^4 \text{ monetary units/year}$$
If the after-modification make-up is lower, for example 10 kg per year, the saving becomes:
$$S_o = (200 – 10) \times 160 = 3.04 \times 10^4 \text{ monetary units/year}$$
Additional savings come from reduced packing replacement, reduced cleaning labor, reduced environmental risk, and lower probability of oil starvation in the pinion gear mesh. A simple total saving model can be written as:
$$S_{total} = S_o + S_p + S_l + S_d$$
where \(S_o\) is the oil saving, \(S_p\) is the packing replacement saving, \(S_l\) is the labor and cleaning saving, and \(S_d\) is the avoided downtime and risk cost. Even when only the oil saving is counted, the modification pays back quickly because the fabrication work uses ordinary 4 mm steel plate and simple bolting and welding.
| Performance Indicator | Before Modification | After Modification | Change |
|---|---|---|---|
| Visible oil leakage at the pinion gear | Continuous, especially at the low end in summer | No visible leakage | Eliminated |
| Annual oil make-up | Approximately 200 kg | Approximately 10 to 20 kg | 90% to 95% reduction |
| Annual oil cost | Approximately 3.2 x 10^4 monetary units | Approximately 0.16 x 10^4 to 0.32 x 10^4 monetary units | Approximately 2.88 x 10^4 to 3.04 x 10^4 monetary units saved |
| Packing replacement frequency | Frequent, with leakage returning after a short time | Rare, because packing is mainly a dust seal | Reduced maintenance |
| Pinion gear shaft condition | Oil-wet, contamination risk | Dry and clean | Improved |
| Oil return to sump | Poor, oil escaped along the shaft | Good, oil follows the labyrinth rings back to the pool | Improved |
| Dust protection | Packing provided both oil and dust control | Packing remains for dust control | Maintained |
| Contact wear between seal parts | Packing wore against the shaft | Moving and stationary rings never touch | Eliminated |
| Summer performance | Worst leakage period | Stable, no seasonal leakage peak | Improved |
The improved seal also changed the maintenance strategy for the pinion gear. Before the modification, maintenance was reactive. When leakage became visible, the packing was replaced, and the area was cleaned. After the modification, maintenance became preventive and inspection-based. The labyrinth rings are simple steel components. They do not require lubrication, and they do not wear against each other because of the 15 mm radial and 20 mm axial clearances. The inspection items are therefore visual and dimensional. I check for oil accumulation, loose bolts, weld cracks, deformation, and clearance changes during scheduled kiln shutdowns.
| Inspection Item | Method | Acceptance Criterion | Action if Not Acceptable |
|---|---|---|---|
| Moving ring bolts | Visual and torque check | No looseness, specified torque maintained | Retorque or replace bolts |
| Moving ring runout | Dial indicator | Within design tolerance | Adjust or repair segment |
| Radial clearance | Feeler gauge or direct measurement | Not less than 15 mm design target | Adjust or replace ring |
| Axial clearance | Feeler gauge or direct measurement | Not less than 20 mm design target | Adjust or replace ring |
| Stationary ring weld | Visual and dye penetrant if required | No cracks or undercut | Repair weld |
| Oil return path | Visual inspection during operation | Oil drains freely to the oil pool | Clean or modify drainage |
| Original packing | Visual and compression check | Light dust seal, no excessive heat | Adjust or replace packing |
| Pinion gear shaft surface | Visual and contact check | No scoring or oil tracking | Polish or investigate alignment |
| Gear cover interior | Visual inspection | No excessive oil accumulation | Check drainage and oil level |
The success of this modification comes from a simple principle: do not fight the oil at the shaft. Intercept it before it reaches the shaft. The pinion gear rotates and throws oil. That energy is not a problem if the oil is directed back to the sump. The original packing tried to hold the oil at the shaft, where the environment is most difficult: small clearances, high surface speed, vibration, misalignment, and thermal variation. The new labyrinth seal works farther away from the shaft, where there is more space, larger clearances, and a natural drainage path. This is why the new seal solved the leakage problem without requiring a complicated control system or a tight-tolerance mechanical seal.
The moving ring and stationary ring also reduce the pressure pulsation that reaches the packing. In the original design, the packing was exposed to the full fluctuating pressure inside the gear cover. In the new design, the labyrinth cavities damp the pressure fluctuation and spread the pressure drop over multiple stages. The oil that reaches the packing is therefore at a much lower pressure and much lower flow rate. The packing can perform its remaining dust-sealing function without being overloaded by oil. This is a key reason why the original packing lasts longer after the modification.
The vibration tolerance of the design is another advantage. A contact seal such as packing requires continuous contact with the pinion gear shaft. Any shaft motion changes the contact pressure and creates wear. A labyrinth seal with 15 mm radial clearance and 20 mm axial clearance does not require contact. It tolerates shaft motion and gear cover motion. In a rotary kiln, the pinion gear is affected by kiln shell deflection, riding ring slip, support roller alignment, gear mesh variation, and process changes. A non-contacting labyrinth is therefore a better match to the actual behavior of the pinion gear than a contacting packing seal. The clearances must be large enough to avoid contact, but the geometry must still guide oil back to the sump. The 15 mm and 20 mm values achieved this balance in my application.
I also considered the effect of oil level and oil flow rate. If the oil level inside the gear cover is too high, the labyrinth may be flooded, and oil may be carried over the rings. If the oil level is too low, the pinion gear mesh may not receive enough lubrication. The correct oil level is therefore important. After the modification, I verified that the oil level was within the specified range and that the oil return path was not restricted. The labyrinth is not a substitute for correct oil level control. It is an enhancement that keeps the oil inside the gear cover when the oil level is correct.
The oil return path can be modeled as a hydraulic resistance network. The captured oil flow \(Q_c\) divides into a return flow \(Q_r\) and a residual flow \(Q_{res}\) that may reach the packing:
$$Q_c = Q_r + Q_{res}$$
The return flow is driven by gravity:
$$Q_r = \frac{\Delta p_g}{R_r}$$
where \(\Delta p_g\) is the gravity head available for drainage and \(R_r\) is the hydraulic resistance of the return path. To minimize \(Q_{res}\), the return resistance \(R_r\) should be low. This means the return path should be short, steep, and free of obstructions. In the modification, the rings were arranged so that oil falls onto a downward surface and then flows to the oil pool. The original gear cover was not modified in a way that traps oil. This simple detail is important for the long-term success of the pinion gear seal.
The axial clearance of 20 mm provides another benefit. It allows the moving ring to move slightly relative to the stationary ring without contact. This accommodates thermal expansion of the pinion gear shaft, axial thrust variation, and small assembly tolerance. It also makes the seal easier to install. The moving ring segments can be bolted to the pinion gear end face without requiring precise shimming to a tight axial gap. A 20 mm axial clearance is large enough to absorb practical installation variation and small enough to maintain the labyrinth effect. The 15 mm radial clearance performs the same function in the radial direction.
The new seal does not rely on the packing to stop oil. This changes the failure mode of the system. In the original design, packing wear directly caused oil leakage. In the new design, even if the packing wears, the labyrinth continues to intercept and return the oil. The packing becomes a secondary dust barrier. The system therefore has redundancy. If the packing is damaged or incorrectly adjusted, the pinion gear shaft may remain dry because the labyrinth is still working. This redundancy is valuable in a continuous cement production line, where unexpected leakage can create environmental and safety issues.
I also evaluated the effect of the modification on gear mesh lubrication. The labyrinth seal does not reduce the oil supply to the gear mesh. It only changes the path of oil after it has been thrown from the mesh. The oil pool remains the same, and the oil circulation inside the gear cover remains active. The moving ring rotates with the pinion gear and does not block the meshing zone. The stationary ring is welded inside the gear cover at a location that does not interfere with the large girth gear or the pinion gear teeth. This is why the modification can be made without changing the gear mesh geometry or the lubrication strategy.
The material selection was intentionally simple. A 4 mm steel plate is easy to cut, bend, weld, and bolt. It is strong enough to withstand oil impact and vibration. It is not intended to be a pressure boundary. It is an oil deflection and drainage component. The moving ring experiences centrifugal load, but the load is small because the ring is lightweight and close to the pinion gear end face. The bolted connection is sufficient when the bolts are properly torqued and locked. The stationary ring experiences oil impact and vibration, but it is welded to the gear cover, which is a stiff structure. The 4 mm thickness therefore provides a good balance between weight, stiffness, and fabricability.
For similar applications, I recommend the following design checks before adopting a labyrinth seal at the pinion gear:
| Design Check | Purpose | Typical Consideration |
|---|---|---|
| Available axial space | Ensure moving and stationary rings fit inside the gear cover | Allow at least 20 mm axial clearance plus assembly tolerance |
| Available radial space | Ensure no interference with the pinion gear and cover | Allow at least 15 mm radial clearance |
| Oil throw direction | Confirm where oil leaves the gear mesh | Place the moving ring to intercept the main oil path |
| Drainage path | Ensure captured oil returns to the oil pool | Avoid pockets, steps, and blocked drains |
| Pinion gear runout | Set clearance to avoid rubbing | Measure shaft and end face runout |
| Thermal growth | Confirm clearances remain positive at operating temperature | Use thermal expansion formulas |
| Vibration level | Estimate dynamic displacement at the pinion gear | Use vibration data if available |
| Packing condition | Retain packing as a dust seal | Do not over-tighten the packing |
| Fastener locking | Prevent bolts from loosening | Use lock washers, thread locking, or tack welds as appropriate |
| Inspection access | Allow future inspection and maintenance | Provide removable covers if possible |
The improvement has been in service long enough to demonstrate its effectiveness in both normal and summer conditions. The low-end leakage that was previously the most serious problem has not returned. The oil make-up interval has been extended significantly. The area around the pinion gear is cleaner. The risk of oil contamination on the foundation and surrounding equipment is reduced. The original packing continues to provide dust protection, but it is no longer the limiting component for oil sealing. The pinion gear seal now behaves as a system: a labyrinth for oil control, a packing for dust control, and a gravity return path for oil recovery.
The economic effect can be summarized with a simple annual savings equation that includes oil, packing, labor, and risk:
$$S_{annual} = (m_b – m_a)c_o + (n_b – n_a)c_p + (h_b – h_a)c_h + \Delta R$$
where \(m_b\) and \(m_a\) are the before and after oil make-up masses, \(c_o\) is the oil unit cost, \(n_b\) and \(n_a\) are the before and after packing replacement counts, \(c_p\) is the packing replacement cost, \(h_b\) and \(h_a\) are the before and after labor hours, \(c_h\) is the labor rate, and \(\Delta R\) is the reduction in risk cost. In my case, the oil term alone is dominant. The packing and labor terms add further benefit. The risk reduction is difficult to quantify but important because oil leakage at a pinion gear can lead to slipping, housekeeping problems, and accelerated wear if the oil level falls.
| Cost or Benefit Category | Before Modification | After Modification | Annual Effect |
|---|---|---|---|
| Oil make-up quantity | 200 kg/year | 10 to 20 kg/year | 180 to 190 kg/year saved |
| Oil cost | 3.2 x 10^4 monetary units/year | 0.16 x 10^4 to 0.32 x 10^4 monetary units/year | 2.88 x 10^4 to 3.04 x 10^4 monetary units/year saved |
| Packing replacement | Frequent | Rare | Maintenance cost reduced |
| Cleaning labor | Frequent oil cleanup | Minimal cleanup | Labor reduced |
| Environmental risk | Oil loss to surroundings | Oil contained | Risk reduced |
| Pinion gear lubrication | Oil level unstable | Oil level stable | Reliability improved |
| Production risk | Potential unplanned intervention | Lower intervention probability | Availability improved |
From an engineering standpoint, the most important lesson is that the pinion gear seal should be designed around the actual oil flow, not only around the shaft clearance. The original packing seal focused on the shaft. The new labyrinth seal focuses on the oil. It catches the oil where it is thrown, redirects it along a controlled surface, and returns it to the oil pool. The pinion gear shaft is no longer the primary leakage path. The packing is no longer the primary seal. The system is more tolerant of wear, vibration, thermal change, and misalignment. This is why the modification solved a problem that repeated packing replacement could not solve.
I also observed that the modification did not require a change in the lubricant. The same imported gear oil continued to be used. The oil viscosity and additive package remained suitable for the gear mesh. The labyrinth seal is compatible with the existing oil. It does not introduce a new material that could react with the oil. The steel rings are painted or oil-coated on non-contact surfaces as appropriate, and the drainage surfaces are smooth enough to promote oil return. The original packing material remains compatible with the oil and the operating temperature. This compatibility is important for long-term reliability.
The installation tolerances were checked after the first heating cycle. Thermal expansion can change clearances, and vibration can cause bolts to settle. I therefore recommend a follow-up inspection after the first period of operation. The inspection should check bolt torque, ring runout, clearance, weld condition, and oil return. If the clearances remain positive and the oil return path is clear, the seal will continue to perform. If any bolt loosens, it should be retightened and locked. If any weld shows a crack, it should be repaired. If oil accumulates inside the cover, the drainage path should be cleaned or improved. These steps are simple and can be done during normal shutdowns.
The modification is also scalable to other rotary kilns with similar pinion gear sealing problems. The same principle can be applied to different kiln diameters, gear widths, and oil flow rates. The key parameters are the available space, the oil throw pattern, the shaft runout, the vibration level, and the drainage path. The radial and axial clearances should be selected from measurements and operating data. A larger kiln may require larger clearances and a more robust ring structure. A smaller kiln may require smaller clearances. The general design logic remains the same: intercept the oil before it reaches the pinion gear shaft, guide it back to the oil pool, and leave the packing as a dust seal.
In my first-person assessment, this modification has been one of the most cost-effective improvements on the rotary kiln drive. It required only simple steel plate fabrication, welding, and bolting. It did not require a new oil system, a new lubricant, or a complex control device. It used the existing gear cover and the existing packing. It converted a chronic oil leak into a controlled oil return. The pinion gear seal now performs reliably in summer, when the original seal was weakest. The oil make-up cost has fallen dramatically, and the maintenance workload has decreased. The pinion gear area remains clean and dry. The original packing continues to protect against dust, while the labyrinth protects against oil loss. This combination is simple, robust, and well matched to the operating conditions of a rotary kiln pinion gear.
For future work, I would add a periodic oil balance check. By measuring oil added and oil drained, I can confirm that the labyrinth is returning oil efficiently. I would also record pinion gear vibration and temperature during the hottest part of the year. If vibration increases, I would inspect the ring clearances and the pinion gear alignment. If temperature increases, I would verify the oil viscosity and the oil level. These checks are not required for the basic function of the seal, but they help maintain long-term reliability. The pinion gear seal is now a system that can be monitored, not a consumable that must be replaced repeatedly.
The final result is clear. The original single packing seal allowed oil to leak along the pinion gear shaft. The improved labyrinth seal, made from 4 mm steel plate with a moving ring bolted to the pinion gear end face and a stationary ring welded inside the gear cover, captures the thrown oil and returns it to the oil pool. The radial clearance of 15 mm and the axial clearance of 20 mm prevent contact and tolerate vibration and thermal movement. The original packing remains as a dust seal. The pinion gear shaft stays dry, the oil level stays stable, and the annual oil make-up is greatly reduced. This is a practical, low-cost, and reliable solution for pinion gear seal leakage in a cement rotary kiln.