Rotary kilns and calciners · the sleeve or stand-off that carries the seal surface on a hot shell, and the bellows or membrane that lets the kiln grow and shift without opening the seal
The end of a rotary kiln is where a turning, hot, moving shell meets a fixed hood, and the seal between them has to keep outside air out while the shell grows in length, rises as it heats and wobbles as it turns. Two parts carry that duty: a sleeve or stand-off structure on the shell that holds the sealing surface away from the heat, and a bellows, membrane or spring assembly on the stationary side that lets everything move. When either is sized for the cold kiln only, the seal is crushed on one side and open on the other once the kiln is up to temperature.
Thermal sleeves, stand-off rings and expansion elements at the feed and discharge ends of rotary kilns, calciners, incinerator kilns and indirect-fired drums: the cylinder or stand-off that attaches the rotating seal surface to the hot shell, cooling of the annulus beneath it, bellows and flexible membranes that absorb axial growth and centreline shift, and the leaf, graphite face, packed and labyrinth seals they carry, with their guide rollers. Conveyor rolls for float glass lehrs are covered on the Float glass annealing lehr roller page and the bearings and axial float at the ends of lehr and hot-conveyor rolls on the End bearing assembly for annealing lehr and hot-conveyor rolls page. The welded shell itself, its ovality on the tyres and the straightness of its axis are covered on the Rotating drum shell page, complete rotary dryers on the Drying drum (rotary drum dryer) page, and girth-gear drives on the Drive system page.
01
Drexler's 1990 report on the Process Experimental Pilot Plant (PREPP) kiln at the Idaho National Engineering Laboratory is a measured case of end seals designed for a cold kiln. The incinerator kiln was about 8.5 ft (2.6 m) in diameter and 25 ft (7.6 m) long, ran about 0.5 in water (1 mm Hg) below room pressure, and was to reach 980 °C. [1]
Its original primary and secondary seals were rotating drums attached to the kiln shell through a cylindrical thermal sleeve, running inside stuffing boxes with five rings of packing 19 mm square in section and 10 m long; cooling air was ducted into the annulus between shell and sleeve to keep the seals below the 232 °C shell temperature, and metal bellows sealed between the end structures and the seal support rings. [1] On heat-up the kiln could not reach its operating temperature because of air leakage through the end seals, and drive motor current rose. [1]
The engineering evaluation found that the drum seal could take the 41 mm (1.6 in) axial motion from the kiln's float on its rollers and the shell's thermal expansion, but not a vertical shift of the centre of rotation. The support rollers bear on the tyre about 1.25 m below the kiln centreline, and thermal growth lifted the centreline about 2.5 mm above its cold position; with the stuffing box rigidly fixed at the centreline, the rising drum crushed the packing at the top and opened a gap at the bottom. [1] The report also flagged the sleeve joint itself: differential expansion between the hot shell and the much cooler sleeve, seal drag torque, drum weight and cooling-air thermal cycles made fatigue failure of the sleeve-to-shell joint very probable. [1]
A temporary fix let the stuffing boxes float on brass shims and slotted bolts and added torque rods that resisted drag without restraining vertical or axial movement; it cut kiln drive amperage by 20 % and allowed a year of test operation. [1]
Ask the kiln supplier for the axial movement at each end (float plus thermal growth), the rise of the centreline between cold and hot, the shell temperature at the seal location, and the eccentricity of the rotating seal surface.
02
The replacement design in the same report moves the sealing surface onto a smooth-faced rotating disc fixed to a steel stand-off structure welded to the kiln shell, bedded in high-temperature silicone RTV. The stand-off and bedding provide support, thermal insulation, accommodation of differential expansion and seal continuity; with the shell measured at up to 232 °C, the disc temperature was calculated at 116 °C. [1]
The report is not consistent on axial travel: it describes the original seal as taking 41 mm of axial motion, while the replacement membrane gives a total axial motion capability of 38 mm at each seal assembly, and its abstract speaks of a 1.5 in axial shift. [1] A buyer working from these figures should confirm which value applies to the kiln in hand. Editorial calculation: the 2.5 mm centreline rise that destroyed the original packing is one tenth of the 25 mm radial allowance designed into the replacement, and the calculated disc temperature is half the measured shell temperature. [1]
Flexible membranes of fibreglass fabric impregnated with a high-temperature material join the two seal housings and connect the primary housing to the firing hood; the outer one takes kiln float, shell expansion between inlet and outlet seals and wobble of the disc. The housings hang from overhead rods that let them swing axially with the shell but keep their elevation at the disc's centre of rotation, and horizontal tie rods take the drag torque. [1]
| PREPP replacement seal item | Published value |
|---|---|
| Radial displacement of the disc allowed without losing seal continuity | 25 mm (1.0 in) in any direction [1] |
| Total axial motion capability per seal assembly (membrane) | 38 mm (1.5 in) [1] |
| Rubbing faces finished wider than the bar track | About 50 mm (2.0 in) [1] |
| Graphite-carbon seal bars | 36 per ring, 25.4 mm square, 280 mm long, 10° taper at each end, lap-jointed [1] |
| Face pressure | 0.03 MPa (5 psi) nominal from 72 preload springs, reduced as seal-air pressure rises [1] |
| Slip at each lap joint from differential expansion | About 0.3 mm [1] |
| Wear allowance and estimated bar life | 16 mm; up to 10 years under PREPP conditions [1] |
| Design life between major services | 10,000 operating hours [1] |
For the sleeve or stand-off, ask how it is attached to the shell, what temperature the seal surface will run at, how differential expansion between shell and sleeve is relieved, and how the joint is protected from cyclic and drag loads.
03
FEECO, a US maker of rotary dryers and kilns, groups its seals by relative leakage, and points to the bellows type where longitudinal expansion is large.
FEECO lists what it needs to specify a replacement seal, including shell temperature at the seal area, maximum material temperature, drum speed and outside diameter, axial expansion of the shell at the seal location, shell material, which end, breeching type and opening, exposed length available for the sealing surface, and suction or pressure. It gives lead times of 1–4 weeks for new leaf seal plates and 4–10 weeks for a newly designed seal system. [3] Its signs of a failing seal are missing, worn or drooping leaves, material spilling back through the seal, chattering or vibration near the breechings and unexplained rises in fuel use. [3]
Guide rollers matter as much as the seal faces. On a 45 in (1.14 m) test replica of a John Zink double-faced kiln seal at the Savannah River Site, the seal plates rode on cam followers spaced 60° apart on the rim of a rotor turning with the kiln. Followers were found up to 3.5° off perpendicular because their mounting holes were poorly machined; high contact forces flaked metal off the rotor and weight-bearing followers backed out of their threads. After re-fitting with machined spacers and nuts the deviations were 0.0–0.3°, inside the follower maker's recommended 0.5°. [2] The follower maker's notes in the same report give 1.5° misalignment tolerance for a crowned follower with a 24 in crown radius, a sealed operating limit of 250 °F (275 °F intermittent) and a surface hardness of Rockwell C 58 to 60. [2]
| FEECO seal type | Relative leakage | Construction and expansion |
|---|---|---|
| Labyrinth | ≫ 5.0 % [3] | Close clearance between rotating seal ring and stationary housing; a flexible wrap can be added [3] |
| Single leaf | ≪ 3.0 % to 5.0 % [3] | Overlapping spring-steel plates mounted on the housing, riding on a wear ring on the shell [3] |
| Double leaf | ≪ 1.0 % to 3.0 %; about 25 % of a single leaf [3] | Two leaf seals with a ceramic fibre blanket between them [3] |
| Bellows | ≪ 1.0 % [3] | Graphite or bronze face seal on a flat rotating surface; corrugated bellows take significant longitudinal expansion, guided and supported by cam rollers [3] |
State the seal type and leakage class wanted, whether the kiln runs under suction or pressure, the axial expansion at the seal, and how guide rollers or followers will be aligned and kept within their temperature limit.
04
Paste it into your enquiry below or into an email to a maker, then replace every bracket.
Kiln: [type; outside diameter; length; speed; direct or indirect firing] End: [feed or discharge; breeching type and opening] Temperatures: [shell at seal location; maximum material; hood gas] Movement: [axial expansion at seal; axial float; centreline rise cold to hot; runout] Pressure: [suction or positive; any purge or seal air] Seal: [labyrinth, leaf, double leaf, bellows, graphite face; leakage target] Sleeve: [attachment to shell; cooling of annulus; exposed length available] Atmosphere: [corrosive or hazardous gases; dust]
Manufacturer profiles associated with Thermal Expansion Sleeve.
Explore company information and manufacturing capabilities, then discuss the scope of your project with a relevant supplier.
At the PREPP pilot kiln the seal took 41 mm of axial motion, but thermal growth lifted the kiln centreline about 2.5 mm while the stuffing box stayed fixed, crushing the packing at the top and opening a gap at the bottom. [1]
It carries the sealing surface away from the hot shell; in the PREPP replacement design a stand-off with RTV bedding held the seal disc at a calculated 116 °C on a shell measured at up to 232 °C. [1]
FEECO gives relative leakage of ≪ 1.0 % for its bellows seal, ≪ 1.0 % to 3.0 % for a double leaf seal, ≪ 3.0 % to 5.0 % for a single leaf and ≫ 5.0 % for a labyrinth seal. [3]
In the Savannah River test the follower maker recommended straightness within 0.5°; followers found up to 3.5° off damaged the rotor until re-fitted to 0.0–0.3°. [2]
Sources checked 2026-10-13. The movement figures and the replacement sleeve and seal design come from an Idaho National Engineering Laboratory report on the PREPP incinerator kiln; the guide-roller findings from a Westinghouse Savannah River Company test report; seal types, leakage classes and ordering data from FEECO's seal pages.
Ask for use case, specification boundaries, supplier type, and RFQ preparation information for this product.
Compare manufacturer profiles with relevant product and process capability.