The connecting rod head — also called the big end — is the larger forked end of a connecting rod that clamps around the crankshaft journal through a removable cap and two bolts, carrying the rod bearing inserts. Charles Algernon Parsons and contemporaries refined the split-cap design through the late 1800s as steam and gas engines pushed loads higher. It transmits piston force into crankshaft rotation while keeping the bearing properly preloaded. Modern forged rod heads in engines like the Cummins ISX run combustion loads above 25,000 lbs per cycle without losing crush.
Connecting Rod Head Interactive Calculator
Vary rod bolt stretch limits and measured elongation to see whether the big-end cap is under-clamped, in range, or near yield.
Equation Used
The calculator checks measured rod bolt elongation against the target stretch band. If the measured stretch is below the minimum, the cap is under-clamped; if it reaches the yield stretch value, the bolt should be discarded.
- Torque-to-stretch inspection is used instead of torque alone.
- Target stretch band is the article range for an ARP 2000 rod bolt.
- A bolt at or beyond the yield stretch limit is treated as discard.
- This check compares elongation only; actual clamp load also depends on bolt stiffness.
How the Connecting Rod Head Actually Works
The rod head is split for a reason — you need to install it around a one-piece crankshaft. The cap bolts back on with two fasteners (sometimes four on heavy-duty diesels), and when those bolts torque to spec, they pull the cap into the rod body across a precision parting face. That joint must transfer load by friction and dowel-fit, not by the bolts working in shear. If the parting face is dirty, nicked, or the cap is installed reversed, the bore goes out of round and the bearing wipes within minutes.
Inside the bore sits a pair of bearing inserts — plain shells, usually a tri-metal or bi-metal construction with a steel back. The bore is machined slightly smaller than the outer diameter of the assembled shells, typically by 0.025 to 0.075 mm. That interference is the bearing crush, and it locks the shells against rotation while ensuring even heat transfer into the rod. Lose crush — through a stretched bolt, a hammered cap, or a re-sized bore that's now too large — and the shells spin. Spun bearing. Engine done.
Rod bolts are the most stressed fasteners in the engine. They cycle from near-zero load at TDC firing to peak tensile load at TDC exhaust as the piston tries to fling itself out the top of the cylinder. Most builders today torque-to-stretch rather than torque-to-spec, measuring rod bolt elongation with a stretch gauge to a target like 0.0055 to 0.0065 inches on an ARP 2000 bolt. If you measure 0.004 inches you're under-clamped and the cap will fret. If you measure 0.008 inches the bolt has yielded and must be discarded — no exceptions.
Key Components
- Rod Body to Cap Parting Face: Either a flat saw-cut surface or a fractured (cracked) surface on modern powdered-metal rods. The fractured face self-locates the cap with zero lateral movement — a flat-cut rod relies on dowel pins or knurled bolt shanks. Surface finish must be Ra 0.8 µm or better; any burr lifts the cap and destroys bearing crush.
- Rod Cap: The removable lower half of the head. Must be installed in its original orientation and matched to its original rod — caps are not interchangeable between rods even within the same engine. Numbering or arrows are stamped at the factory for this reason.
- Rod Bolts: Typically 3/8 inch on a small-block Chevy, 7/16 inch on a big-block, M11 or M12 on heavy diesels. ARP 2000 bolts run 220,000 psi tensile, ARP L19 and Custom Age 625+ run 260,000 to 280,000 psi for high-boost builds. Bolts are single-use after stretch — reusing yielded bolts is the most common cause of rod failure in track-day engines.
- Bearing Inserts (Shells): Two half-shells with locating tangs that sit in machined notches in the rod and cap. Crush range is 0.025 to 0.075 mm. Oil clearance to the journal is 0.025 to 0.064 mm typically — Plastigage is the cheap accurate way to verify this on a rebuild bench.
- Pin-End Bushing (Small End Reference): Not part of the head itself, but the head's geometry depends on the centre-to-centre rod length being correct relative to the small-end pin bore. A bent rod with 0.05 mm of misalignment between the heads will wipe one bearing edge within 500 miles.
Who Uses the Connecting Rod Head
Every reciprocating internal combustion engine on the planet uses a rod head of some form, but the design choices change dramatically with load, RPM, and service interval. Here are the contexts where the head's design directly drives the engine's capability.
- Performance Automotive: Manley Turbo Tuff and Carrillo H-beam rods used in built Toyota 2JZ-GTE engines targeting 1000+ HP, with ARP Custom Age 625+ rod bolts to survive sustained 8000 RPM.
- Heavy-Duty Diesel: Cummins ISX15 cracked-cap powdered-metal rods, designed for 600,000+ mile service life under peak cylinder pressures over 200 bar.
- Marine and Stationary Power: Caterpillar 3500 series rod assemblies in tugboat propulsion and prime-power generation, with serrated parting faces for dimensional repeatability across hundreds of rebuilds.
- Aviation Piston Engines: Lycoming O-360 forged steel rods in Cessna 172 trainers — inspected for yield at every overhaul under FAA AD compliance.
- Vintage Engine Restoration: Babbitt-poured rod heads in Model T Ford engines and Fairbanks-Morse hit-and-miss engines, where the bearing material is cast directly into the rod head and hand-scraped to fit the journal.
- Motorsport: Pankl titanium rods in Formula 1 power units, where head geometry is optimized for minimum reciprocating mass while surviving 15,000 RPM.
The Formula Behind the Connecting Rod Head
The number that decides whether a rod head survives or grenades is the peak tensile load on the rod bolts at TDC of the exhaust stroke — the moment the piston reaches the top with no combustion pressure pushing it back down, so inertia tries to throw it through the valve cover. At low RPM (idle, 800 RPM) the bolt sees almost nothing. At nominal cruise RPM the bolt is well within fatigue limits. At redline the bolt approaches its stretch yield, and that's where you decide whether stock fasteners survive or whether you need ARP. The formula gives you the inertial load each bolt must clamp against.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Fbolt | Peak tensile load per rod bolt at TDC exhaust | N | lbf |
| mrecip | Reciprocating mass (piston + pin + rings + small-end of rod) | kg | lb |
| r | Crank throw radius (half the stroke) | m | in |
| ω | Crankshaft angular velocity | rad/s | rad/s |
| n | Rod length / crank radius ratio (typically 1.6 to 2.0) | dimensionless | dimensionless |
Worked Example: Connecting Rod Head in a built Honda K24A2 four-cylinder for time attack
You're spec'ing rod bolts for a built Honda K24A2 destined for a time-attack Civic. Reciprocating mass per cylinder comes to 0.62 kg (forged JE piston, tool-steel pin, rings, and rod small end). Stroke is 99 mm, so crank radius r = 0.0495 m. Rod length is 152 mm, giving n = 152 / 49.5 = 3.07. You need to know what each rod bolt sees at idle (800 RPM), at the typical track operating range (7500 RPM), and at the planned 9200 RPM rev limit, then choose between OEM bolts and ARP 2000 upgrades.
Given
- mrecip = 0.62 kg
- r = 0.0495 m
- n = 3.07 dimensionless
- RPM range = 800 / 7500 / 9200 RPM
Solution
Step 1 — convert the nominal track RPM (7500) to angular velocity:
Step 2 — compute the inertial term, which is split between the two rod bolts (hence the divide-by-2):
That's well within the working load of an OEM Honda rod bolt rated around 18,000 N preload. Comfortable margin.
Step 3 — at idle (800 RPM), ω = 83.8 rad/s, and force scales with ω2, so:
Effectively nothing — the bolt is just holding the cap on. Bearing fatigue at idle is not a concern.
Step 4 — at the 9200 RPM rev limit, ω = 963.5 rad/s:
Now you're at the OEM bolt's preload limit — meaning every rev cycle is dragging the bolt close to its yield point. This is exactly the scenario where guys lose engines on the third hot lap. ARP 2000 bolts at 220,000 psi tensile add roughly 40% headroom and that's the reason every serious K24 build uses them above 8500 RPM.
Result
Each rod bolt sees roughly 12,540 N (2,820 lbf) of peak tensile load at the nominal 7500 RPM track condition. That's a comfortable 30% margin on OEM fasteners — the engine will run all season. The range tells the real story though: at 800 RPM idle the load is essentially zero (143 N), at 7500 RPM it's manageable, and at 9200 RPM it climbs to 18,870 N which is right at OEM yield — the sweet spot for a stock-bolt build is below 8500 RPM, and above that you're buying ARP. If you measure rod knock or find bearing copper on the magnetic drain plug despite calculations saying you have margin, the usual suspects are: (1) rod bolts that were torqued without measuring stretch, leaving them under-preloaded so the cap frets at TDC, (2) a rod cap installed 180° reversed which puts the bearing tangs out of register, or (3) reused rod bolts from a previous build that have already yielded once and now stretch unpredictably.
Choosing the Connecting Rod Head: Pros and Cons
Rod head construction breaks into three families that builders actually choose between: forged I-beam, forged H-beam, and powdered-metal cracked-cap. Here's how they compare on the dimensions that drive purchasing decisions.
| Property | Forged H-Beam Rod Head | Forged I-Beam Rod Head | Powdered-Metal Cracked Rod Head |
|---|---|---|---|
| Peak RPM capability (typical street/race build) | 9000-11,000 RPM | 8000-9500 RPM | 7000-7500 RPM (OEM-limited) |
| Tensile load capacity per bolt | Up to 25,000 N with ARP Custom Age | Up to 20,000 N with ARP 2000 | 12,000-15,000 N OEM rated |
| Cost per set (4-cyl) | $650-1,200 USD | $450-800 USD | $0 (factory) to $250 aftermarket |
| Service life under rated load | 500-1000 hours track use | 300-700 hours track use | 300,000+ miles street use |
| Resizing tolerance after rebuild | ±0.013 mm achievable | ±0.013 mm achievable | Cannot be resized — fractured face won't realign if bore is altered |
| Best application fit | High-boost, high-RPM forced induction | Naturally aspirated and moderate boost | Stock and mild OEM rebuilds |
Frequently Asked Questions About Connecting Rod Head
Torque-to-spec is unreliable on rod bolts because friction in the threads and under the bolt head varies by 20-30% depending on lubricant, surface finish, and how many times the bolt has been used. Two bolts at the same indicated torque can be at different actual preloads, which means the cap pulls down unevenly and the bore distorts.
Switch to torque-to-stretch with a rod bolt stretch gauge — measure the bolt's free length, torque it down, and verify elongation against the manufacturer's spec. ARP publishes target stretch for every bolt they sell. If two bolts on the same rod show different stretch at the same torque, one of them is galling or the threads are dirty.
No. Rod bolts are torque-to-yield in most modern engines, meaning the factory installation procedure deliberately stretches them past the elastic limit to maximize clamping force. Once stretched, the bolt's stress-strain curve is permanently altered and it will yield further at lower loads on reuse.
Even ARP bolts, which are reusable in theory, should be replaced if they've been through more than 5 torque cycles or if measured stretch on retorque exceeds the as-new spec by more than 0.0005 inches. Rod bolts are the cheapest insurance in any engine build — buy new ones.
The deciding factor is peak cylinder pressure, not horsepower. H-beam rods carry higher compressive loads better because the H cross-section resists buckling under the column load that turbocharged combustion produces at TDC. I-beam rods are lighter and survive tensile loads (the inertial pull at TDC exhaust) just as well, but they're more prone to bend under high boost.
Rule of thumb: if your build will see more than 25 psi of boost or peak cylinder pressures above 150 bar, go H-beam. For naturally aspirated or low-boost builds, I-beam saves reciprocating mass and lets the engine rev faster.
You're almost certainly missing crush. The bore can measure correctly while the bearing shells lack the interference fit they need to lock in place. Check three things in order: (1) confirm the bearing shell thickness against spec — counterfeit bearings from low-cost suppliers run 0.02-0.05 mm undersize, (2) verify the rod and cap are matched and oriented correctly, since installing a cap from a different rod will shift the parting face, and (3) check that the bearing locating tangs sit cleanly in their notches without lifting the shell.
If all three check out and crush is still absent, the rod bore has been resized too large during a previous rebuild and the rod is scrap.
Manufacturing economics, not engineering compromise. Powdered-metal forming presses near-net-shape rods in one operation, then a controlled fracture splits the cap along a microscopically irregular surface that self-aligns on reassembly with zero machining. No saw cut, no dowel pins, no precision grinding of the parting face. The process is faster and cheaper at OEM volumes.
The trade-off is that you cannot resize a cracked-cap rod — once the bore wears or distorts, the fractured face cannot be realigned to a new bore. Forged rods can be resized 2-3 times in their service life, which is why race programs and rebuilders prefer them despite the higher upfront cost.
Probably not yet — but it's a warning. Cold-start rod knock that quiets as oil pressure builds usually means rod bearing oil clearance is at the high end of the spec range (0.05-0.064 mm) and the journal is dropping onto the bearing during the brief moment before pressurized oil arrives. The bearing material is taking a small impact every cold start.
Pull the pan, plastigage one or two rods, and check actual clearance. If you measure above 0.075 mm you're past spec and the rod head bore has likely gone out of round — replace the bearings and inspect the rod for resizing. If clearance is in spec, the cause is more likely a worn oil pump or a clogged pickup screen delaying prime.
Industry standard is Ra 0.8 µm or better on the parting face. A nick or burr of even 0.05 mm acts as a fulcrum that lifts one side of the cap when bolts are torqued, which simultaneously distorts the bore (kills crush on one side) and concentrates clamp load at a single line of contact (frets the joint).
The fix is to surface the parting face flat on a precision rod resizer, then resize the bore to spec. Trying to file or stone the nick out by hand changes the rod centre-to-centre length and creates a worse problem than the original burr.
References & Further Reading
- Wikipedia contributors. Connecting rod. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
