EnginesN54 N55 S55

Crankhub conversions: pinned vs splined vs stock — the honest guide

The N54/N55/S55 crankhub aftermarket, sorted: what actually fails, what pinned and splined conversion kits really demand of the installer, what they cost, and which fix fits which car.

15 min readUpdated 2026-10-20Rapid Racing

Hub styles compared — our schematic, drawn for this guide

Our engine pages carry what the factory service documentation says: the torque-plus-angle law, the special tools, the renewal lists. This page is the layer the factory documents do not touch — the aftermarket crankhub conversions that own the N54/N55/S55 conversation. Nothing here is factory data. It is vendor documentation and community experience, labeled as exactly that: where a claim is a vendor's own testing, we say so; where a number is a forum poll, we say that too. The mechanics, the honest comparison, and the decision tree are ours.

Each conversion path now has its own install guide in this series — the pinned drilling discipline, the splined press-and-broach sequence, and the crank bolt capture with its clocking law. This page decides; those pages wrench.

What the factory built — and what it deliberately didn't

On the N54, N55 and S55, the timing chain sprocket, oil pump drive and vibration damper hub are a stack of separate parts clamped to the crank nose by one central torque-to-yield bolt. The drive through that stack is friction — friction plates and grip discs squeezed by bolt clamp load. There is no key, no spline, no pin between crank and hub. That is not an exaggeration the community invented; it is the design.

Keyed and splined crank noses are the century-old default across the industry. BMW's friction stack is cheaper to manufacture and self-centering on the crank's taper, and under stock loads it holds — the factory documents treat it as a wear-and-renewal item, not a defect. The tell is downstream: when BMW replaced this architecture, the B58 and S58 went to a damper bolted directly to the crank nose with no separate hub to slip. The company never issued a recall or a defect admission for the friction-hub engines; the redesign is the closest thing to a statement that exists.

Two factory laws frame everything on this page:

The central bolt is single-use, torque-plus-angle. A released bolt is stretched and goes in the bin — every engine page in this series repeats it because the factory documents do.
On the N54, the friction plates between the hub and the chain sprocket are mandatory and damage-critical — the factory's own words are that the engine sustains damage without them.

Factory service documentN54 crankhub central bolt — factory torque (11 21 1AZ)Opens in your account's service knowledge — sign-in required.

How it actually fails

Two mechanisms, usually in sequence:

Clamp load is lost. The bolt backs off or was never fully stretched — a reused bolt, an interrupted angle stage, a nose that has been apart before. Friction capacity falls with clamp load. Shops that install these weekly are blunt that this is half the story: a hub that has been apart before can let go on a car that isn't making big power.
A torque spike exceeds what's left of the friction. Hard downshifts, a money shift, a raised rev limit, a launch — the event loads the stack past its remaining capacity and something rotates a few degrees relative to the crank. The chain sprocket is the usual mover; the oil pump drive is rarely the culprit in the field (that observation is a turbo vendor's bench experience, and it matches the failure photos owners post).

The terminology trap: people arrive from keyed-pulley engines expecting a wallowed-keyway story. There is no keyway to wallow here — that is the whole problem. When the stack slips, cam-to-crank correlation is simply lost:

Best case: correlation faults (2A98/2A99-family codes), rough running, a re-time. A leak-down test confirms no valve contact before you re-time and carry on.
Worst case: the shift is large, a valve meets a piston, and you are buying a head or an engine.

The failure-rate question, honestly

There is no controlled dataset. Nobody — not BMW, not any vendor, not any forum — has failures-per-engine-produced with a denominator anyone would defend. What circulates:

A self-selected forum poll on S55 crank hub failure read out around 4%, with posters themselves flagging the selection effect (people who had failures click).
Other community estimates land around 2%, "mostly modified cars."
One owner-club writeup argues under 1% even on tuned cars.
One vendor's reliability dossier says under 10% of S55s produced, rising with tuning — the softest number in circulation and the broadest bracket.

What is actually defensible, and all we will claim: the mechanism exists in every one of these engines; documented failures cluster on tuned, hard-driven and previously-disassembled cars; N54s are underrepresented relative to S55s in the failure threads, which tracks with the S55's easier path to big torque. A stock car that stays stock and was never apart is a rare case in the threads — rare enough that "awareness, not action" is the honest posture for it.

This page covers N54, N55 and S55 because they share the friction-hub architecture. B58 and S58 have no separate hub to convert — their pages cover the damper bolts and chain drive that replaced all of this.

Path 1 — the factory fix

New OEM hub, new central bolt, the full torque-plus-angle procedure, and — on the N54 — the friction plates inspected and renewed as the factory documents demand. This is what the engine pages in this series exist for; the torque tables, tools and sequences are there.

Engine	Central bolt	Torque
N54	`M16 x 80`, replace, light oil	100 Nm (73.8 lb-ft) + 360°
N55	`M18 x 108`, replace, light oil	100 Nm (73.8 lb-ft) + 270°
S55	`M18 x 108`, replace, light oil	100 Nm (73.8 lb-ft) + 270°

One parts wrinkle worth knowing, community-documented (a parts-supercession question answered by real owners on real build threads, not by BMW): from roughly the 2009 model year on, the N54 moved to an updated hub/sprocket set that carried into N55 and S55 — hub 11237593701, sprocket 11317502180, crank sprocket with grip disc 11418091535, washer 11237557561, bolt 11211439395. Early cars keep the old-design parts, which remain available. Verify against your VIN before ordering; the community consensus is that the update is a refinement of the same friction architecture, not a mechanical lock.

When the factory fix is enough: a stock-refresh or a car that stays near stock power, any car already apart for chain or seal work where the bolt must come out anyway, and any owner whose honest answer is "I'm not drilling my crank." Done correctly — new bolt, full angle, clean friction surfaces — it restores the design as delivered.

Path 2 — the pinned hub

One-piece conversion: the separate hub and timing sprocket become a single machined body, and dowel pins are driven through the hub into holes drilled in the crank nose. The pins, not friction, carry the load. This is the oldest aftermarket answer and the one most big-power builds run; the market has moved from two pins to four.

What the kits are: a one-piece hub/timing gear, drill bit and drill jig (the fixture bolts to the crank snout and guides the bit), pins, a new bolt, and hardware. Some add a keyed oil pump gear and a crank seal guard.

What the install actually demands — this is the honest part the product pages compress:

The oil pan comes off. Pinned installs are done with the pan dropped so chips can be controlled from below, which means subframe work and an engine support cradle.
You drill a hardened crank nose in the car. Carbide bits, cutting oil, slow speed, mild pressure — the supplied bits are hard and brittle and snap if side-loaded. A dull or rushed bit doesn't stop cutting; it rubs, wanders and bell-mouths the bore, and the hub is only as good as its holes.
The jig fit is a real variable. Vendor install notes themselves record an N54 where the fixture sat with play and had to be foil-shimmed to keep the holes on-center. Read that sentence again with your own crank in mind.
Chips are the enemy. Blow out each hole with the jig still installed, shop vac running, never upward; vendors point at the internal lip where debris collects and tell you to wipe it and check with a finger. That paragraph exists in their instructions because engines have been lost to it.
Depth control is a tape mark on the bit. The pins must seat without bottoming out.

The engineering tradeoff, stated fairly: the crank nose is surface-hardened from the factory. Drilling cuts through that case, leaving bore walls softer than the surface around them, and adds stress risers in a high-cycle-fatigue zone. That is the entire no-drill argument, and it is legitimate — yet thousands of pinned hubs run in the field without a pattern of crank-nose failures emerging. The unanswered question belongs to the vendors, not the critics: in ten manufacturers' product pages surveyed by an independent shop, not one publishes the dowel pin's material or hardness — and the pins are the load path. Too hard and a pin can brinell its bore oval; too soft and it peens over. Buy from whoever will answer that question in writing.

The full drilling discipline — jig fit and foil shims, depth stops, the carbide law, the chip protocol, the hold-the-hub torque method — is the pinned install guide's job.

Path 3 — the splined hub, no drilling

The modern conversion inverts the machining problem: instead of removing crank material, the hub carries precision-machined splines harder than the crank, and is pressed on like a stock hub — the bolt torque drives the splines into the nose, broaching their own teeth. A slip event only drives the teeth deeper, the same way a splined extractor grips. Because nothing is drilled, there is no chip control problem, no jig fit problem, and no oil-pan-off requirement to drill (vendors still recommend dropping the pan for the job; it is simply no longer the drilling that demands it).

The honesty ledger for splined hubs:

It is a vendor's proprietary design. One shop dominates this category; their published testing (a hydraulic rig their final hub broke before slipping, a shop race car at 8K RPM and 900+ WHP) is their own claim, not independent verification.
The hardness story is published and checkable, at least. The hub is 440C tool steel at roughly HRC 60; the vendor has published their own Rockwell testing of the crank's hub area around HRC 23 and a stock hub too soft to read on the scale. Hardness numbers from the seller of the harder part — but they are numbers, which is more than the pin market offers.
The bolt capture is required, not optional. Vendors state the splined hub is not intended to run without a crank bolt capture — because the second failure path (bolt back-off) is untouched by the hub itself. Budget for the pair.
The crank nose must be undamaged. A hub that already slipped can leave burrs or scoring; the install guide is explicit that the nose must be in good shape to accept the splines.
It is reversible-ish. The broached splines stay in the crank; a stock hub can go back on afterwards, aligned to the existing teeth. No holes, no permanent modification beyond the impressions.

The step-by-step — the old-bolt press, the 75 ft-lb stage, the factory angle, the capture companion — lives in the splined install guide.

The companion fix — the crank bolt capture

Worth its own line in your thinking: one of the two failure paths is the bolt, and a capture plate (bolts to the damper, captures the crank bolt head so it cannot rotate back) addresses exactly that, on a stock or converted hub. Vendors pair captures with splined hubs as a system; on a budget or a mild build, a capture on a fresh factory hub is the cheap tier of the aftermarket — it does nothing for a friction-stack overload, everything for bolt back-off.

What a capture can and cannot fix, the clocking problem, and the kit torque specs are the bolt-capture install guide's territory.

The honest comparison

Prices are the vendors' own listed street prices for the hub kit itself (checked October 2026 — verify before you buy; several kits ship in several configurations). Labor dwarfs parts on every row except the first.

Option	Crank machining	Parts cost (kit)	Install beyond a stock hub job	DIY-able?
Factory fix	None	OEM hub + bolt + plates — parts money, source-dependent	None — it is the factory job	Yes, with the engine-page tools and torques
Bolt capture	None	Low three figures	Damper-off access; modest	Yes — the easiest tier
Pinned one-piece	Drill 2–4 holes in the crank nose	~$1,000–1,700 (Insane 2-pin $999; VAC 4-pin ~$1,100; SSR expanded kit $1,687 with gaskets)	Pan off, subframe work, jig drilling, chip control	Honest answer: a skilled-shop job
Splined (no-drill)	None — hub broaches its own splines	~$734–890 (VTT configurations; capture required)	Press-on with the old bolt, then new bolt + factory angle; timing tools mandatory	Yes for the experienced; vendors still say qualified tech

No option is engine-out on the N54 or N55 by nature — the hub comes off in the car during the chain job (the S55 pages carry the engine-out caveats that apply to S55 hub work specifically). The "crank-out question" that circulates for splined conversions is a misreading: nothing in these kits needs the crank on a lathe. The machining is either drilled-in-place (pinned) or done by the hub itself (splined).

Which one for which car

Stock, staying stock, never apart: nothing. Read the failure section, sleep fine. If the bolt ever comes out for other work, a fresh bolt at full spec — or a capture plate if you want the belt-and-suspenders version — is the whole ask.
Tuned street car, stock rev limiter: minimum a bolt capture; ideally a conversion timed with any front-of-engine work you're already doing — the marginal cost rides along with a chain or seal job you'd pay for anyway.
Raised rev limiter, big torque, hard launches, track: a one-piece mechanical lock, full stop. Pinned or splined is an install-philosophy choice — drill the crank and carry the load on pins, or press on a hub that cuts its own teeth and lock the bolt. Pick by who does your work: a shop fluent with the drill jig makes pinned routine; a no-drill kit removes the two install risks (chips, jig fit) that scare careful people.
Single-turbo / big-power build: the conversion is part of the build list, not a later decision — every serious build guide for these platforms treats it that way, and the failure threads are full of people who sequenced it wrong.
It already slipped: stop. Leak-down before you crank it again. If the valves survived, you re-time, convert while it's apart, and inspect the crank nose — a damaged nose accepts neither pins nor splines happily.

While you're in there

Both vendor install guides (pinned and splined) say the same things, and they match the factory procedures our engine pages carry:

Front crank seal — "now is the time" is the pinned guide's own phrasing; the hub is out and the seal bore is right there.
Chain guides, tensioner, rails — age items on all three engines; if the front is apart at high mileage, this is the moment (the valvetrain and timing refresh page covers the judgment).
Oil pan gasket and hardware — the pan is off on a pinned install regardless; the one-time hardware gets renewed with it.
VANOS bolts — one-time-use wherever they're touched; the engine pages carry the specifics.
S55 oil-chain tensioner piston — the splined vendor's guide warns it can pop its bore during oil sprocket work and is very hard to reseat; have the replacement on hand.
Head studs — a different job with the same "while it's apart" arithmetic; if the head is coming off anyway, the HCR head stud guide sequences that work.

Where the factory numbers live

This page deliberately carries no factory torque tables of its own beyond the central-bolt law — the per-engine pages hold the full tables, tools and honest step summaries:

N54 — the original friction-plate crankhub
N55 — the March 2012 procedure split
S55 — the grip-disc law and the 600 Nm+ breakaway
B58 / S58 — why this whole subject doesn't exist there
The overview — the field guide to all five engines
Valvetrain and timing refresh — the guides-and-chain judgment call

Every factory number on this page traces to the factory document ids listed in the sources. Everything else is vendor documentation or community experience, labeled inline — where a claim is a vendor's own testing, it says so; where a number is a poll, it says that. If a number isn't in either category, we didn't print it.

References — the research shelf

The vendors' own product pages and install documents, and the independent shop writeup this guide leans on hardest:

Factory sources

Questions

Does a stock N54 or S55 crankhub really need upgrading?

On a stock, untuned, sensibly driven car, documented failures are rare — every number in circulation is a forum poll or a vendor estimate, not a controlled dataset. The risk climbs with tuning, raised rev limiters, hard launches and downshifts, and with any hub that has been apart before without the full torque-plus-angle procedure. Awareness for a stock car; action for a tuned one.

Is drilling the crankshaft for pins safe?

It is a real engineering tradeoff, not a settled question. The crank nose is surface-hardened from the factory; drilling cuts through that case, leaves bore walls softer than the surface around them, and adds stress risers in a fatigue zone — the entire argument the no-drill camp makes. Thousands of pinned hubs are in the field without a pattern of crank failures, but no pin kit vendor publishes pin material or hardness, which is the single most load-critical spec in the design. Skilled install, or don't.

Pinned or splined — which is better?

Both convert the drive from friction to a mechanical lock. Pinned kits drill the crank nose and drive dowels through the hub into it — the most direct load path, the most demanding install (oil pan off, jig drilling a hardened surface, chip control). Splined hubs press on like a stock hub and broach their own splines into the nose — no drilling, but the hub must be harder than the crank to cut, and vendors pair them with a crank-bolt capture as a required companion. Pick by install philosophy and by who is doing the work, not by forum loyalty.