A rigid coupling is a solid mechanical connector that joins two shafts end-to-end so they rotate as a single shaft. Its critical component is the coupling body — a sleeve, flanged hub pair, or split clamp — that grips both shaft ends through keys, set screws, or clamping force and transfers full torque without flex. We use rigid couplings where shafts must stay perfectly concentric and where any backlash would ruin positioning accuracy. The result is zero lost motion and full torque transfer, which is why CNC ball screws, encoder shafts, and high-precision pump drives almost always run rigid.
Rigid Coupling Interactive Calculator
Vary clamp force, friction, shaft diameter, and applied torque to see whether a split-clamp rigid coupling can transmit torque without slipping.
Equation Used
The calculator estimates the friction torque capacity of a split-clamp rigid coupling. The available torque rises with friction coefficient, total clamp force, and shaft radius. A safety factor above 1 means the calculated grip torque exceeds the applied torque.
- Split-clamp rigid coupling transmits torque by friction grip.
- Clamp force is the effective total normal force on the shaft.
- Shaft and bore are clean and dry with uniform contact pressure.
- Misalignment effects and bolt preload scatter are not included.
The Rigid Coupling in Action
A rigid coupling does one job: lock two shaft ends together so they behave as one continuous shaft. There is no rubber spider, no metal bellows, no slot-and-key sliding interface. Torque enters one end, exits the other, and nothing in between is allowed to move relative to anything else. That sounds simple, but it puts a hard demand on the install: the two shafts have to be concentric to within roughly 0.02 mm TIR (Total Indicated Runout) on a small servo coupling, and parallel offset has to be effectively zero. Miss that target and the coupling will not forgive you — it transmits the misalignment straight into the bearings on either side.
The grip itself comes from one of three strategies. A sleeve or muff coupling slides a hollow cylinder over both shaft ends and locks it down with set screws or a key fitted into a keyway cut along both shafts. A flange coupling uses two hubs, each keyed to its own shaft, bolted face-to-face through a precision-reamed bolt circle — the bolts carry the torque in shear. A clamp or split-muff coupling wraps two halves around the shaft pair and squeezes them together with through-bolts, generating torque transfer through pure friction with no key required. Each strategy has a different failure mode if you get the spec wrong: set screws back out under reversing loads, flange bolts shear if the bolt-hole fit is sloppy, and clamp couplings slip if you under-torque the bolts or get oil on the bore.
Why build it rigid at all when flexible couplings exist? Because any compliance in the drivetrain shows up as backlash, lost motion, or torsional wind-up. On a CNC ball screw driving a 0.005 mm-resolution linear encoder, you cannot afford even half a thousandth of angular slop between motor and screw. Rigid wins there. The trade is that rigid couplings demand alignment work the flexible ones do not.
Key Components
- Coupling Body (Sleeve, Flange Hub, or Split Clamp): The structural element that bridges the two shafts. On a sleeve coupling the bore is typically machined H7 to the shaft's h6 — a clearance of 0 to +21 µm on a 25 mm shaft. The body must be stiff enough that torque does not wind it up and concentric enough that it does not throw the shafts off-axis under load.
- Key and Keyway: A rectangular bar (commonly 6×6 mm on a 25 mm shaft per DIN 6885) sitting in matching slots on shaft and bore. Carries torque in shear. The key must be a snug fit on the sides — typically a P9/N9 fit — but with 0.1-0.3 mm clearance on top so it bears on the flanks, not the crown.
- Set Screws: On keyless or supplementary designs, cup-point set screws bite into the shaft to lock the coupling axially and provide some torque grip. Use two screws at 90° rather than one — single set screws back out under reversing loads. Threadlocker (Loctite 243 or equivalent) is non-optional on any vibrating drive.
- Flange Bolts: On flange couplings, fitted bolts running through reamed holes carry the torque in shear between the two flange faces. Bolt diameter, count, and bolt-circle radius set the torque rating directly. A typical 100 mm flange coupling uses 4 to 6 M10 bolts on an 80 mm PCD.
- Clamping Bolts (Split-Muff Designs): On a clamp-style rigid coupling, multiple bolts pull the two half-shells together around the shafts. Torque transfer is pure friction — no key needed — so bolt preload must be tightly controlled. Under-torque the bolts and the coupling slips silently under load before failing.
Real-World Applications of the Rigid Coupling
Rigid couplings live wherever shafts must run truly co-linear and any compliance is unacceptable. You see them most often on vertical pumps, machine-tool spindles, encoder feedback shafts, and short shaft sections that need to look and behave like a single piece. The choice of rigid over flexible is almost always driven by precision requirements or by the need to maintain axial stiffness — the moment you can tolerate misalignment or want to isolate vibration, you reach for a flexible coupling instead.
- Machine Tools: Coupling a servo motor to a precision ball screw on a Haas VF-2 vertical machining centre, where any backlash between motor encoder and screw would corrupt positioning.
- Vertical Pumping: Joining shaft sections on a Goulds VIT vertical turbine pump for municipal water supply, where 6 m of line shaft runs from a surface motor down to a submerged impeller through multiple rigid sleeve couplings.
- Marine Propulsion: Connecting tail shaft sections between the gearbox and propeller shaft on a tugboat fitted with a Caterpillar 3512 engine, using flanged rigid couplings rated for full continuous torque.
- Industrial Mixing: Coupling a SEW-Eurodrive helical gearbox to an agitator shaft on a stainless reactor in pharmaceutical production, where a clamp-style rigid coupling allows disassembly for shaft removal without disturbing the gearbox.
- Test Stands and Dynamometers: Linking a torque transducer to the unit under test on a HBM T40B inline torque sensor rig, where rigid coupling is mandatory to avoid the torsional ringing that flexible couplings introduce into the measurement.
- Conveyor Drives: Joining drive-roller stub shafts on long conveyor systems in distribution centres, where multiple short rigid sleeve couplings transmit torque through bolted flanges along the line.
The Formula Behind the Rigid Coupling
The basic question on any rigid coupling is whether the shaft itself can carry the torque you intend to push through it. The shaft is almost always the weak link — the coupling body and bolts get sized after the shaft. At the low end of the typical industrial range (10-50 Nm on a 20 mm shaft) you have huge margin and almost any catalogue coupling works. In the middle of the range (200-500 Nm on a 35 mm shaft) you start caring about key shear stress and bolt preload. At the high end (2000+ Nm on a 60 mm shaft) you are running close to material limits and the choice between key-driven and keyless friction couplings starts to matter. The formula below gives shaft torque capacity from torsional shear stress.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| T | Allowable torque the solid shaft can transmit | N·m | lb·ft |
| τallow | Allowable shear stress of the shaft material (typically 40-60 MPa for general-purpose steel shafting after safety factor) | Pa (N/m²) | psi |
| d | Solid shaft diameter at the coupling bore | m | in |
Worked Example: Rigid Coupling in a packaging line conveyor drive
A packaging line integrator in Eindhoven is joining the output shaft of a Nord SK 9032 helical gear reducer to the head pulley of a 12 m carton accumulation conveyor using a sleeve-type rigid coupling. The shaft is 35 mm diameter, C45 medium-carbon steel, with τ<sub>allow</sub> taken as 50 MPa after a safety factor of 3. The integrator wants to confirm the shaft can handle the drive torque and understand where the operating sweet spot sits across the line's load range.
Given
- d = 35 mm (0.035 m)
- τallow = 50 MPa (50 × 10⁶ Pa)
- Operating torque range = 150-450 N·m
Solution
Step 1 — compute the maximum torque the 35 mm shaft can carry at the allowable shear stress:
That 421 N·m is the shaft's continuous capacity. Now look at the actual operating range.
Step 2 — at the low end of the line's load range, 150 N·m (light cartons, half-full conveyor):
At 36% utilisation the shaft is loafing. The coupling bolts feel almost nothing and you would not detect any wind-up by hand. This is comfortable cruising territory and the conveyor will run quietly with no measurable shaft deflection.
Step 3 — at the nominal design torque of 300 N·m (full conveyor, normal product mix):
This is the sweet spot — enough margin to absorb startup spikes and the occasional jam-clearance event without yielding the shaft, but using the steel efficiently. A 35 mm shaft is correctly sized here.
Step 4 — at the high end of the operating range, 450 N·m (jam recovery, reversing pulse):
You are slightly over the allowable. The shaft will not fail instantly — the safety factor of 3 against ultimate is still intact — but if 450 N·m happens repeatedly you will eventually fatigue the keyway. Either size up to 40 mm shaft (which gives Tmax ≈ 628 N·m) or limit the drive's torque output through the VFD to 400 N·m.
Result
The 35 mm C45 shaft handles 421 N·m at the chosen allowable stress, which sits comfortably above the 300 N·m nominal operating torque. At 150 N·m the shaft runs at 36% utilisation — quiet, no detectable wind-up. At 300 N·m nominal it sits at 71%, the engineering sweet spot. At 450 N·m peak it tips slightly over allowable at 107%, which is fine occasionally but bad as a daily condition. If you measure premature wear or hear a periodic clunk on startup, check the keyway first — a worn key with 0.2 mm side clearance lets the shaft hammer the keyway under reversing loads, which fatigues the corner of the slot. The second culprit is set-screw back-out: cup-point screws without threadlocker walk loose within 200 hours on any reversing drive. Third, check the bore-to-shaft fit — if the coupling bore is worn beyond H8 the entire shaft assembly will run eccentric and chew the gearbox output bearing.
When to Use a Rigid Coupling and When Not To
Rigid couplings are not the default choice for every shaft connection — they are the default only when alignment is guaranteed and zero compliance is needed. The realistic competition is jaw couplings (with elastomer spider) for general motion control and disc or bellows couplings for precision servo work that still needs misalignment compensation. Here is how the three stack up on the dimensions readers actually search.
| Property | Rigid Coupling | Jaw Coupling (Elastomer Spider) | Disc/Bellows Coupling |
|---|---|---|---|
| Backlash / lost motion | Zero | 0.1-0.5° depending on spider hardness | Zero (similar to rigid) |
| Allowable parallel misalignment | ≤ 0.02 mm (effectively zero) | 0.1-0.4 mm | 0.1-0.2 mm |
| Allowable angular misalignment | ≈ 0° | 1-2° | 0.5-1° |
| Torque capacity for given OD | Highest — full shaft torque | Lower — limited by spider material | Medium — limited by disc/bellows |
| Torsional stiffness | Maximum (effectively infinite) | Low (compliant spider) | High but finite |
| Cost (typical 25 mm bore) | $15-60 USD | $25-90 USD | $120-400 USD |
| Service life under aligned load | Indefinite (no wear elements) | Spider replacement every 2-5 years | 10+ years if not overloaded |
| Best application fit | Precision ball screws, encoder shafts, vertical pump line shafts | General motor-to-pump/fan, mild misalignment | Servo drives needing precision plus misalignment tolerance |
Frequently Asked Questions About Rigid Coupling
Bearing heat on the coupling-side end of a rigid drive almost always means residual misalignment that the coupling is forcing into the bearing rather than absorbing. A flexible coupling would have flexed and shed that load — a rigid one cannot, so the radial component shows up as bearing preload.
Put a dial indicator on the motor shaft and rotate by hand. If TIR exceeds 0.02 mm at the coupling face, you have parallel offset that the rigid coupling is faithfully transmitting into the motor bearing. Realign with shims at the motor mount until TIR drops below 0.02 mm and the temperature rise will follow it down within an hour of running.
The shaft formula gives steady-state shear capacity. It says nothing about the grip mechanism between coupling and shaft. If you specified a clamp-style rigid coupling, slipping means bolt preload is below the friction threshold needed to carry your peak torque, not your nominal.
Clamp couplings rate friction torque at a specific bolt torque — typically the manufacturer publishes a value like 25 N·m bolt torque gives 200 N·m transferable. A shock load of 2-3× nominal needs a coupling rated for that peak, not for nominal. Either step up coupling size, switch to a keyed sleeve coupling, or add a torque-limiter upstream.
Pick a flange coupling when you need to disconnect the shafts without pulling either shaft axially. On a long line shaft — a vertical turbine pump or a marine prop shaft — you cannot slide a sleeve off; there is nowhere for it to go. A flange splits in the middle and lets you separate two assemblies that are each axially constrained.
The other reason is sheer torque capacity. A flange coupling with 6 fitted M16 bolts on a 150 mm PCD carries far more torque than any sleeve coupling of similar OD because the bolts work in shear at a long lever arm.
No — and this is one of the most common misuses we see. A rigid coupling assumes both shafts are independently supported by their own bearings, with the coupling spanning a short gap between them. If one side is a long cantilever with no outboard bearing, the coupling becomes a structural joint carrying bending loads it was never designed for.
The symptom is a shaft that whips at speed and a coupling that cracks at the keyway corner within months. Either add an outboard pillow block to support the driven shaft, or switch to a flexible coupling that tolerates the angular deflection a cantilever produces.
Critical to the point of being the single biggest variable in real-world service life. Standard practice is H7 bore over h6 shaft — a clearance fit of roughly 0 to +21 µm on a 25 mm shaft. Loosen that to H8/h8 (up to 60 µm clearance) and the shaft will run eccentric inside the coupling, the key takes hammering, and you get measurable runout downstream.
If a coupling has been removed and re-fitted several times, measure the bore. We have seen 25 mm bores opened up to 25.08 mm by repeated installation — at that point the part is scrap regardless of how the outside looks.
Sheared flange bolts almost always mean the bolt holes are not reamed — they are drilled clearance holes. In that case the bolts carry torque in shear across whatever clearance exists, which means under load only one or two bolts actually engage at a time, instead of all six sharing the load.
Proper flange couplings use fitted bolts running through reamed holes (typically H7) with a transition or interference fit between bolt body and hole. Check your bolts — if they slide through the hole with finger pressure, you have clearance bolts in a flange that needs fitted bolts. Replace with reamer-fit body bolts and the shearing stops.
Yes — and this catches people out on test stands and dynamometer rigs. A rigid coupling adds almost no compliance, so the torsional natural frequency of the system is set entirely by the shafts and the rotor inertias. Swap a flexible coupling for a rigid one and the system natural frequency typically jumps by a factor of 3-10×.
If you are running at a speed that excites a multiple of that new frequency, you will see torsional resonance — measurable as periodic torque spikes on a transducer or audible as a beat tone. The fix is either to shift operating speed away from the resonance or to deliberately add a tuned flexible element. Rigid is not always the right answer on a test rig for this reason.
References & Further Reading
- Wikipedia contributors. Coupling. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
