Synchromesh

The problem: meshing two gears that aren't turning the same speed

In a modern manual gearbox every forward gear is permanently in mesh with its partner on the layshaft. This is a constant-mesh design, and it has one awkward consequence: the gear you want to drive through is spinning freely on the output shaft, almost never at the shaft's own speed. To actually transmit torque you have to lock that free gear to the shaft — and the way you lock it is by sliding a toothed collar, the sleeve, so its internal teeth grab a ring of teeth on the side of the gear. Those are the dog teeth.

If the gear and the shaft are turning at different speeds when the dog teeth meet, the teeth clash. They cannot interleave while there is relative rotation; they hammer tooth-against-tooth, chip, and make the unmistakable grinding shriek of a missed shift. The earliest gearboxes — sliding-mesh "crash" boxes — had exactly this problem, and the only cure was driver skill: blip the throttle and double-declutch to manually bring the gear to the right speed before easing it in. Get it wrong and you ground the gears.

Synchromesh is the mechanism that does that speed-matching automatically, in the tenth of a second between when you start moving the lever and when the teeth actually engage. Earl Thompson patented it at Cadillac in 1928; by the 1960s nearly every passenger-car manual used it on all forward gears. The trick is to put a tiny cone clutch in the path, so friction drags the gear to shaft speed first, and a clever blocker ring that physically refuses to let the dog teeth engage until that speed match is complete.

The mechanism, part by part

A single synchronizer assembly sits on the splined output shaft between two gears (it usually serves two adjacent gears — push one way for one, the other way for the other). Its parts:

• Hub. Splined rigidly to the output shaft, so it always turns at shaft speed. It carries the sliding sleeve and three spring-loaded detent keys.
• Sleeve (the slider). Internally toothed, it rides on the hub and slides axially when the shift fork pushes it. Its inner teeth are the half of the dog clutch that belongs to the shaft. The fork groove on its outside is what the shift linkage grabs.
• Blocker ring (baulk ring, synchronizer ring). A brass or steel ring with an internal friction cone and external chamfered teeth. It is the gatekeeper — and the friction surface — at the same time.
• Gear cone and dog teeth. The free gear carries a matching male friction cone and, beyond it, a ring of dog teeth. These are what the sleeve ultimately engages.
• Detent keys and springs. Three keys, sprung outward into a groove in the sleeve, transmit the driver's initial axial nudge to the blocker ring and provide the soft "gate" feel of the lever.

The shift happens in three crisp phases:

1. Presync (key push). The driver moves the lever; the fork pushes the sleeve toward the gear. The detent keys ride along and shove the blocker ring against the gear's friction cone. The cone surfaces touch — a small axial force, a large normal force across the shallow cone angle.
2. Sync (baulking). Friction on the cone generates a torque that drags the gear toward shaft speed. Crucially, that same friction torque rotates the blocker ring by half a tooth pitch relative to the sleeve, so the chamfered teeth of the ring sit squarely in the path of the sleeve's chamfers. The sleeve is now blocked: it physically cannot pass while a speed difference exists, because the indexing torque holding the ring offset is larger than the chamfer can push aside.
3. Engagement. When the speeds equalize, the cone drag torque collapses to zero. There is nothing left to hold the ring offset, so the driver's steady axial push deflects the ring chamfers aside, the sleeve slides through the ring's teeth and onto the gear's dog teeth, and the gear is locked to the shaft. The whole sequence takes 0.1–0.3 s.

The physics: cone torque, blocking, and the angle that makes it work

The synchronizer is, at heart, a cone clutch sized to win a torque race against itself. The friction torque the cone can deliver is

T_cone = (μ · F_a · R_m) / sin(α)

where μ is the cone friction coefficient (≈0.08–0.13 for the brass-on-steel or molybdenum-coated pairs used in gearboxes, running in oil), F_a is the axial force the driver applies through the sleeve, R_m is the mean cone radius, and α is the cone half-angle. Notice the 1/sin(α) term: the shallower the cone, the larger the normal force for a given axial push, and the more torque you get. A typical synchronizer cone half-angle is 6° to 8° — shallow enough to multiply torque strongly, but kept above the self-locking limit (tan α > μ) so the cones reliably separate when you let go. Go too shallow and the cones stick together; go too steep and you lose the torque multiplication.

The time to synchronize follows from impulse–momentum. The cone torque has to change the gear's angular momentum across the speed gap:

t_sync = (J · Δω) / T_cone

Worked example — a 2nd-gear downshift:
  Reflected inertia of gear + layshaft, J = 0.012 kg·m²
  Speed difference at the synchro, Δω = 250 rad/s   (~2,400 rpm)
  Mean cone radius, R_m = 0.035 m
  Cone half-angle, α = 7°   →   sin α = 0.122
  Friction coefficient, μ = 0.10
  Driver axial force at the sleeve, F_a = 350 N

  T_cone = (0.10 × 350 × 0.035) / 0.122 = 10.0 N·m
  t_sync = (0.012 × 250) / 10.0 = 0.30 s

That 0.30 s is the "feel" of a slightly stubborn second-gear downshift; a fresh single-cone synchro on a small engine is closer to 0.15 s. The energy that disappears into the cone as heat each shift is ½ · J · Δω² — here about 375 J dumped into a brass ring weighing a few grams. Do that thousands of times a day in a city bus and you understand why heavy-duty boxes use bigger, multi-cone synchros and oil cooling.

The blocking action is its own torque balance. The chamfer on the blocker-ring teeth is cut at a blocking angle (commonly 55–65° from the tooth flank). The axial sleeve force, resolved on that chamfer, tries to twist the ring back into alignment with a torque T_index. Synchromesh works only while a speed difference exists if

T_cone > T_index    (blocked — sleeve cannot pass)

The cone torque must dominate the chamfer's un-indexing torque right up until the moment of synchronization. Designers tune the chamfer angle so this inequality holds for any reasonable shift force; the instant Δω → 0 and T_cone → 0, the inequality flips, the chamfer wins, and the sleeve drops through. Get the chamfer angle wrong and you either get a synchro that crunches (sleeve passes too early) or one that baulks forever and never lets you into gear.

Variants: single, double, and triple cones

Because T_cone scales with friction area, a single cone runs out of torque on the hardest shifts — first and second gear, where both the inertia J and the speed gap Δω are largest. The fix is to stack cones in series:

• Single-cone. One friction interface. Standard on the higher gears (3rd, 4th, 5th, 6th) where speed differences are modest. Lightest, lowest inertia, cheapest.
• Double-cone. An intermediate cone ring nests between the gear cone and the blocker ring, giving two friction surfaces in series — roughly double the synchronizing torque for the same axial force. Common on 1st/2nd of mainstream passenger gearboxes.
• Triple-cone. Three friction interfaces, ~3× the torque. Used on 1st/2nd of performance, diesel, and heavy-duty boxes so the driver can downshift quickly without leaning on the lever.

Friction materials have also moved beyond plain brass. Sintered bronze, carbon-fiber-faced cones, and molybdenum thermal-spray coatings raise μ and, more importantly, raise durability and heat capacity. Carbon-faced synchros tolerate higher rubbing speeds and are standard in fast-shifting sports gearboxes.

Synchromesh versus the dog box and the modern alternatives

Property	Synchromesh	Crash / dog box (no synchro)	Electronic speed-match dog clutch
Speed matching	Friction cone, automatic	Driver double-declutches	Motor/engine spins gear to target
Shift skill needed	Low — one clutch press	High — throttle blip + timing	None (computer-managed)
Shift speed	0.1–0.3 s	As fast as you can rev-match	Tens of milliseconds
Parasitic loss / wear	Cone friction, ring wear	Minimal once engaged	Minimal (no friction cone)
Noise on shift	Silent when healthy	Clack/grind unless perfect	Soft clack
Typical use	Road-car manuals, DCTs	Race cars, motorcycles, classics	Multi-speed EV/hybrid, sequential race

The straight-cut dog box survives in motorsport for exactly the reasons synchromesh costs you: no friction cones means no parasitic drag, no inertia to spin up, and a shift that completes the instant the dogs line up — at the price of noise and a driver who must blip perfectly. At the other end, multi-speed EVs increasingly skip the brass entirely: since the traction motor can set any shaft speed in milliseconds, a bare dog clutch with a controller doing the speed-match does in software what synchromesh does in friction.

Failure modes and trade-offs

• Worn cone / glazed friction surface. The cone's micro-grooves wear smooth and oil can no longer be squeezed out of the interface, so μ drops and the cone can't generate its synchronizing torque. Symptom: crunch into gear, worst when cold and thick-oiled. Most common on 2nd gear, which sees the most frequent large speed changes.
• Rounded blocker-ring chamfers. If the blocking chamfers wear round, the indexing inequality T_cone > T_index fails early and the sleeve passes before synchronization completes — a partial crunch even though the cone may still grip.
• Wrong gear oil. GL-5 hypoid oils carry aggressive sulphur-phosphorus extreme-pressure additives that chemically attack yellow-metal (brass) baulk rings. Synchromesh boxes specify GL-4; using GL-5 can destroy the synchros in months. This is a genuine, frequently-made maintenance error.
• Thermal overload. Repeated hard shifts (city buses, aggressive downshifting) dump heat into a small ring faster than the oil can carry it away; the ring discolors, distorts, and loses grip. The cure is multi-cone synchros and sometimes dedicated synchro cooling.
• Inertia and shift effort. The fundamental trade-off: the bigger and multi-coned the synchro, the more rotating inertia it adds and the heavier the box. Designers balance shift quality against weight and the parasitic drag of more friction surfaces.
• Detent spring fatigue. Weak keys give a vague, notchy gate feel and can let the sleeve creep, causing pop-out of gear under load.

Where synchromesh shows up

• Every modern road-car manual transmission. All forward gears synchronized; reverse is frequently still unsynchronized (hence the occasional reverse clunk), because reverse is selected at rest where speeds are already near zero.
• Dual-clutch transmissions (DCT). A DCT is two manual gearboxes sharing an output; each uses synchromesh to pre-select the next ratio on its idle shaft so the clutch swap is seamless. Volkswagen DSG, Porsche PDK, and Getrag/Ford PowerShift all rely on it.
• Heavy commercial gearboxes. Trucks historically used unsynchronized "crash" boxes (skilled drivers double-declutch), but modern truck transmissions and the lower gears of many buses use heavy multi-cone synchros.
• Tractors and off-highway. Synchromesh on the road-speed ranges, sliding-mesh or dog engagement on creeper/PTO ranges selected at rest.
• Some multi-speed EV/hybrid transmissions. A handful of two- and three-speed EV gearboxes use synchronizers, though many move to electronically speed-matched dog clutches.

Common pitfalls when reasoning about synchromesh

• Thinking the synchro transmits the drive torque. It does not — the cone clutch only equalizes speed. Once synchronized, the dog teeth carry all the torque; the cone goes slack. Sizing the cone for drive torque rather than synchronizing torque is a beginner error.
• Forgetting the self-locking limit. The cone half-angle must satisfy tan α > μ or the cones weld together and won't release. The 6–8° range is a deliberate margin above this limit.
• Confusing the friction torque with the blocking torque. They are different mechanisms acting at once: the cone friction does the speed-matching, the chamfered ring teeth do the gatekeeping. A healthy synchro needs both correct.
• Assuming faster oil helps. Lower-viscosity oil reduces drag but can starve the cone of the boundary film it needs; synchro friction is a designed tribological pair, not a place to second-guess the oil spec.
• Treating reverse like a forward gear. Unsynchronized reverse must be selected at rest; forcing it while the input is still spinning is what produces the grind, not a fault.