What is each one good at?
The two compare like this (after BTL):
| Clevis joint | Rod end bearing | |
|---|---|---|
| Movement | Single plane | Multi-directional |
| Misalignment | Limited | Accommodates angular misalignment |
| Design | Fork and pin | Spherical ball in a housing |
| Fitting | Pin connection, quick to assemble and take apart | Threaded, male or female, more adjustable |
| Load | High capacity in shear, linear loads and simple pivoting | Dynamic loads and varying directions |
In short: use a rod end where angular misalignment is expected, movement occurs in more than one direction and smooth, precise motion is needed. Use a clevis where movement is limited to a single axis, a simple robust connection is needed and easy assembly and disassembly matter.
What happens if the geometry does not match?
The error goes somewhere. A cylinder is built to push and pull along its centre line. Any bending, twisting or misalignment is absorbed somewhere else, and if the mounting cannot allow angular movement, "somewhere" becomes the rod, the gland bushing and the rod seal. Over time that shows up as uneven seal wear, scratches on the rod and finally a bent rod or gland failure. A plain rod eye, which is a bushing and not a bearing, has little tolerance for angular misalignment; a spherical rod eye is the choice where the linkage moves in compound motion or the mounting is not stiff.
Forcing a clevis into a mechanism that moves out of plane binds the pin and loads the fork in twist. A clevis also has no spherical element to wear, so the pin and the bushing or bore are its only wear points (Q1-12).
How much misalignment can a clevis take?
Very little, and you can estimate how little. Misalignment here means the pin axis is not square to the tang, so the tang tries to tilt on the pin. Two clearances stop it.
- Clearance of the tang on the pin. The tang can tilt until the edges of its bore touch the pin. Let c be the diametral clearance (bore diameter minus pin diameter) and t the tang thickness. The tang tilts until opposite edges of its bore touch the pin, so tan θ = c ÷ t, about c ÷ t for small angles. With c = 0.1 mm and t = 19.6 mm, that is 0.0051 rad, or 0.29°.
- Clearance between the tang and the fork arms. The end of the tang, a distance L from the pin centre, moves sideways by L × tan θ until it meets an arm. With a gap of 0.2 mm between the tang and the arm on the side it tilts toward, and L = 15 mm, that is 0.0133 rad, or 0.76°.
The smaller of the two governs, here 0.29°. These figures are illustrative: your clearances decide the real value. Past that angle the tang bears on the edge of the pin bore, the pin bends and the fork spreads, which is the edge loading that shortens the life of the joint (Q1-05, Q1-12). So a clevis takes a fraction of a degree, not a tolerance to count on. Check the linkage through the full travel for binding (K06-Q1-10) and avoid side load. A spherical rod end also has a limit: the usable angle is smaller than the catalogue angle once a clevis, bolt head or spacer gets in the way (K03-Q1-09).
When is a clevis the wrong mount?
A clevis is usually not the best option when the cylinder will see continuous side load, severe misalignment in more than one plane, or a long stroke with no support. In those cases the mounting concept should be reviewed first. Where a clevis does suit, keep the pin axes at the two ends parallel, so the whole mechanism swings in the plane at right angles to them. Check at both ends of the stroke that the fork clears the bracket and that the pin turns freely. The set-up steps are in K06-Q1-10.
How do you decide?
- Draw the motion. Does the link swing in one plane through the whole stroke? If yes, a clevis suits.
- Check the mounting. Are the two brackets in line, and stiff enough not to move under load? If not, a spherical end absorbs it.
- Think about service. A clevis is quick to take apart. A spherical end has a bearing that wears and can need replacing.
- Check the load. A clevis takes load through the pin in shear, so size the pin (Q1-05, Q1-06).
- If you need both, ask the maker whether a clevis rod end with a spherical bearing is available for your size. See K06-Q1-06 and K03-Q1-03.
Can a clevis cope with pedal-box flex on an off-road or race car?
A pedal arm or throttle lever looks like a clear case for a clevis, because it swings in one plane. But the pedal box is bolted to a bulkhead or a tube frame that flexes, and the driver's foot pushes sideways as well as forward.
Use the method above, with assumed numbers. A tang 8 mm thick and 0.1 mm of diametral clearance on the pin give tan θ = 0.1 ÷ 8, which is about 0.72°. Over a 150 mm pushrod, 0.72° is a sideways movement of only about 1.9 mm before the tang bears on the pin bore, unless the side gap to the fork arms stops it sooner. A pedal box or frame that flexes more than that, or a foot that loads the pedal sideways, will bind the pin and load the fork in twist.
Suspension and steering links move in several planes through large angles as the wheel travels and steers, and a clevis is the wrong end for them. A spherical rod end is the usual choice there. SYZ's rod end misalignment calculator estimates a spherical rod end's misalignment angle, with a suspension example, and its clevis rod ends page covers the one-plane case. Passenger-car suspension parts are the subject of the sister site, syzmachine.com.
Pedal, steering and suspension links are safety parts. Which end is acceptable is for the vehicle maker, or the current rules of the event organiser, to say.