The JDM Import Problem
The global trade in used Japanese engines — JDM swaps — creates a torrent of kgf·m torque specs arriving in American garages. A single 2JZ-GTE service manual contains approximately 120 torque specifications, from the 0.8 kgf·m oil pump cover bolts to the 75 kgf·m crankshaft pulley bolt. Every single one is printed in kgf·m. The American mechanic who buys a 2JZ crate engine from an importer in California or a container yard in Virginia receives a photocopied manual — often in Japanese, sometimes translated by the importer, sometimes not translated at all. The mechanic must convert every torque value to ft·lb. That is 120 multiplications. 120 opportunities for an arithmetic error. 120 chances to undertorque a connecting rod bolt or overtorque a camshaft cap and crack an aluminum bearing journal.
The conversion itself is simple multiplication: ft·lbf = kgf·m / 0.138254954, or equivalently, ft·lbf = kgf·m × 7.23301385 when going the other direction. The arithmetic is not the problem. The problem is doing it 120 times without a single mistake. A mechanic working from a JDM manual typically writes the converted values in the margin with a pen, working down the page. By the 40th spec, mental fatigue sets in. By the 80th, the mechanic is rounding to the nearest whole number because "close enough." By the 120th, they're eyeballing it. The engine gets assembled. It runs. Six months later, a camshaft bearing cap bolt that was spec'd at 2.0 kgf·m and converted — incorrectly — to 14 ft·lb instead of 14.5 ft·lb has backed out by half a turn. The cam journal ovalizes. The engine develops a tick at 3,000 RPM. The cause is a rounding error made at 11 PM on a Tuesday in a home garage in Ohio. Use the calculator. It converts all 120 specs correctly, every time, without fatigue.
The Difference Between 'Feels Right' and 'Is Right'
Experienced mechanics develop a feel for torque — the resistance a bolt offers as it reaches yield, the subtle change in feedback through the wrench as the clamping load builds. This kinesthetic sense is real and it works for common fasteners: a 10 mm bolt into a cast iron block, tightened with a 3/8-inch ratchet until it "feels tight," is usually within 20% of the spec. It works because the mechanic has tightened ten thousand identical bolts and their hands have calibrated themselves to the feedback. It fails catastrophically for torque-to-yield (TTY) bolts, which are designed to stretch into the plastic deformation region. A TTY bolt at 8 kgf·m (57.9 ft·lb) "feels" completely different from a conventional bolt at the same torque because the TTY bolt's stress-strain curve flattens as it enters yield — the wrench keeps turning with little increase in resistance. The feel method interprets this as "the bolt is stripping" and the mechanic stops early. The bolt is at 70% of its intended preload. The head gasket leaks.
On a JDM engine, TTY bolts are common: the 2JZ-GTE uses TTY cylinder head bolts. The RB26DETT uses TTY main bearing cap bolts. The EJ20 uses TTY case bolts through the block halves. Every one of these fasteners has a specific torque spec in kgf·m — often a two-stage spec with an initial torque followed by an angle turn. The initial torque is a seating value that must be hit precisely because the angle turn starts from that reference point. If the initial torque is off by 10%, the angle turn produces a clamping load that is off by 10% — but in the plastic region, a 10% error in preload is a 30% error in bolt stretch because the stress-strain curve is nearly flat. A manual that says "8 kgf·m then 90°" and a wrench that reads ft·lb require a calculator, not intuition. The feel method is a skill. It is not a substitute for measurement. On TTY bolts, it is not even a useful approximation.
kgf·m = ft·lbf × 0.138254954
where 0.138254954 = 1 / 7.23301385
= (0.3048 m/ft × 0.45359237 kg/lb × 9.80665 m/s²) / 9.80665 m/s²
= 0.3048 × 0.45359237
Every constant in this chain is exact by international treaty.
Common ft·lbf to kgf·m Conversions
| ft·lbf | kgf·m | Where you'd see this |
|---|---|---|
| 5 ft·lbf | 0.6913 kgf·m | Bicycle stem bolt, small M6 fastener on a motorcycle |
| 10 ft·lbf | 1.383 kgf·m | Valve cover bolt on a small-block Chevrolet — check against JDM equivalent |
| 25 ft·lbf | 3.456 kgf·m | Cylinder head bolt first pass on many 4-cylinder engines |
| 35 ft·lbf | 4.839 kgf·m | Subaru EJ20 engine case bolt (block half mating bolt) |
| 55 ft·lbf | 7.604 kgf·m | Nissan RB26DETT oil pump drive bolt |
| 80 ft·lbf | 11.06 kgf·m | Brake caliper bracket bolt — common to both US and JDM brake systems |
| 90 ft·lbf | 12.44 kgf·m | Passenger car lug nut — the most frequently torqued automotive fastener |
| 120 ft·lbf | 16.59 kgf·m | Truck lug nut, axle nut on a half-ton pickup |
| 165 ft·lbf | 22.81 kgf·m | Harley-Davidson compensator sprocket nut (Twin Cam engines) |
| 200 ft·lbf | 27.65 kgf·m | Commercial truck lug nut, large diesel engine flywheel bolt |
Worked Examples
Nissan RB26 oil pump bolt at 55 ft·lb → 7.60 kgf·m
55 × 0.138254954 = 7.604 kgf·m. The Nissan RB26DETT oil pump is driven directly off the crankshaft nose. The drive bolt — a single M12 fastener — is spec'd at 7.6 to 8.0 kgf·m in the Japanese factory manual. Many English-language forum write-ups and translated shop manuals round this to "7.8 kgf·m." But converting 55 ft·lb gives 7.60 kgf·m — which lands at the bottom of the spec range, not the middle. Convert 58 ft·lb and you get 8.02 kgf·m — the top of the range. The 3 ft·lb difference between "mid-range" and "bottom of range" is the width of the acceptable torque band. An RB26 oil pump bolt that walks loose spins the oil pump drive off the crankshaft flats. Oil pressure drops to zero. The engine seizes — usually at high RPM under boost, where the bearing loads are highest. The RB26 is a $6,000 engine on the used market. Its oil pump drive bolt is one fastener. Get the conversion right.
Subaru EJ20 case bolt at 35 ft·lb → 4.84 kgf·m
35 × 0.138254954 = 4.839 kgf·m. The Subaru EJ20 boxer engine uses a horizontally-split crankcase — two aluminum case halves bolted together through the main bearing webs. The case bolts are M10 fasteners spec'd at 4.8 to 5.0 kgf·m. A mechanic building an EJ20 for a WRX swap in a Colorado garage reads a translated manual that says "5.0 kgf·m." Converting backward: 5.0 / 0.138254954 = 36.2 ft·lb. The mechanic sets the wrench to 35 ft·lb thinking "close enough." That's 4.84 kgf·m — at the bottom edge of the spec. The EJ20's case bolt preload determines the main bearing bore diameter because the case halves clamp around the bearing shells. A 0.16 kgf·m shortfall per bolt produces a main bearing clearance that is wider by a few microns. At 7,000 RPM with 20 psi of boost, the crankshaft deflects into that clearance. The bearing wears asymmetrically. After 30,000 miles, the engine develops a main bearing knock. The failure gets blamed on "Subaru rod knock." It was the conversion.
1-ton truck lug nut at 150 ft·lb → 20.7 kgf·m
150 × 0.138254954 = 20.74 kgf·m. A Ford F-350 lug nut spec is 150 ft·lb — a value burned into the muscle memory of every tire shop technician in America. When that same truck gets exported to a country that uses kgf·m torque wrenches in commercial shops — common in parts of Southeast Asia and the Middle East where Japanese-spec tools dominate the market — the spec must be converted. 150 ft·lb becomes 20.7 kgf·m. A technician who rounds to 21 kgf·m sets the wrench 1.4% high — 152 ft·lb equivalent. On a 14 mm wheel stud with a 1.5 mm thread pitch, that's well within the clamping margin. But a technician who uses "20 even" — because round numbers are easier — applies 145 ft·lb equivalent, a 3.5% undertorque. The stud doesn't fail. The wheel doesn't fall off. But over 100,000 miles of thermal cycling from brake heat, that 5 ft·lb shortfall on every stud is the difference between a wheel that stays seated and one that develops a low-frequency vibration at 65 mph that no tire balance can fix. The conversion matters because the number on the wrench is the only thing between the wheel and the hub.
Harley-Davidson compensator nut at 165 ft·lb → 22.8 kgf·m
165 × 0.138254954 = 22.81 kgf·m. The compensator sprocket nut on a Harley-Davidson Twin Cam 103 engine is torqued to 155–165 ft·lb depending on the model year. It is a left-hand-thread fastener that sees reversing torque pulses from the crankshaft every 45 degrees of rotation. If it loosens, the compensator sprocket walks forward on the crankshaft splines and destroys the primary drive. The factory service manual specifies the torque in ft·lb. Harley-Davidson engines are exported worldwide, including to Japan, where they have a dedicated enthusiast following. A Japanese Harley shop receives a US-spec manual and converts: 165 × 0.1383 = 22.8 kgf·m. The wrench gets set. The nut gets torqued. The compensator holds. The conversion works. But a shop that converts lazily — 165 × 0.14 = 23.1 kgf·m — has overtorqued the nut by 1.3%. On a left-hand-thread fastener that is already preloaded to 90% of its proof load, 1.3% overtorque pushes the bolt into the plastic region on every torque cycle from engine firing pulses. The nut doesn't fail on the first ride. It fails on the 400th, at 3,500 RPM on a highway on-ramp, when the thread finally yields and the compensator walks the nut the rest of the way off. The cost of the conversion error: a primary drive rebuild. The cost of using the exact factor: zero.
Engineering Context
The JDM engine swap is not a niche hobby. It is a global industrial phenomenon. Every year, an estimated 200,000 used Japanese engines are exported — pulled from cars scrapped under Japan's shaken vehicle inspection system, which makes it economically irrational to keep a car older than about 10 years. These engines, with 40,000 to 60,000 miles on them, are in better condition than a 150,000-mile US engine, and they sell for a fraction of the cost of a rebuild. They arrive in containers at the ports of Los Angeles, Houston, Savannah, and Newark. They end up in the engine bays of everything from Nissan 240SXs to Toyota Supras to homebuilt kit cars. Every single one of those engines was designed and documented in kgf·m. Every single one of those engines will be assembled by an American mechanic with ft·lb torque wrenches. The conversion runs at least once per bolt, per engine. At 120 bolts per engine and 200,000 engines per year, that's 24 million conversions annually — most of them performed on a phone calculator in a garage with grease on the screen. The problem of unofficial manual translations compounds the conversion risk. A PDF manual translated from Japanese to English by an enthusiast in Malaysia, hosted on a forum in New Zealand, and downloaded by a mechanic in Texas has passed through three hands before the torque values reach the wrench. Each hand is an opportunity for a rounding decision, a unit confusion, or a simple typo. The only defense is to verify every torque value against the original Japanese numerical tables — the numbers that remain Arabic regardless of the surrounding language — and run the conversion yourself. Trust the original digits. Convert everything. This page is the tool for that job. For the reverse direction — converting from the JDM manual's kgf·m to the wrench's ft·lb — see kgf·m to ft·lb. When the build crosses into pure metric territory, ft·lb to N·m and N·m to ft·lb handle the US-to-SI pair. For Japanese engines where the manual was updated to N·m in later printings, the kgf·m to N·m page confirms whether a value was correctly converted or just copied from the older edition. And for the small fasteners that live below the ft·lb range — the M6 valve cover bolts, the throttle body screws — in·lb to ft·lb covers the sub-division that a standard 3/8-inch torque wrench cannot resolve. The entire torque unit ecosystem traces back to the same treaty constants. The only variable is which two units you need to bridge.
More: kgf·m to ft·lb · ft·lb to N·m · kgf·m to N·m · N·m to ft·lb · Torque Guide
Related Unit Converters
Frequently Asked Questions
Why doesn't Japan just switch to N·m like the rest of the metric world?
Japan is switching. The Japanese Industrial Standard (JIS Z 8203) formally recommends N·m for torque, and every new-vehicle service manual from Toyota, Nissan, and Honda now prints torque in N·m. But the switch began in the 1990s, and JDM engines from the golden era — the 2JZ, RB26, SR20, 4G63, EJ20 — were designed and documented in kgf·m. Those manuals will never be reissued. The Japanese engineering education system taught kgf·m until at least the 1980s because it matched the gravitational metric system that preceded SI adoption. The transition is generational: engineers who trained in kgf·m are retiring, and their manuals remain in circulation. The aftermarket sustains the kgf·m legacy because those are the engines people still build, swap, and race.
How do I know if a translated manual converted the torque values correctly?
Find a known reference value and multiply backward. The RB26DETT factory manual specifies cylinder head bolts at 12 kgf·m initial torque — exactly 86.8 ft·lb. If your English translation says "87 ft·lb," the translator converted and rounded correctly. If it says "80 ft·lb," someone used an approximate factor. If it says "12 ft·lb," the translator copied the number without converting — a distressingly common error in unofficial PDFs. Cross-check three values from different sections of the manual: one large (crank pulley), one medium (head bolt), one small (valve cover). If all three are plausible converted ft·lb values, the translation is probably trustworthy. If any one of them looks like a direct copy of the kgf·m number, assume none of them were converted and do the math yourself.
What's the difference between a JDM manual spec and a USDM manual spec for the same engine?
For engines sold in both markets — the 2JZ-GTE in the Toyota Supra, the 4G63 in the Mitsubishi Eclipse — the JDM manual prints torque in clean kgf·m values (8.0, 12.0, 2.0) and the USDM manual prints converted ft·lb values rounded to whole or half numbers (58, 87, 14). The rounding introduces small discrepancies. The 2JZ head bolt first pass is 8.0 kgf·m = 57.86 ft·lb, which Toyota USA rounds to 58 ft·lb — a 0.14 ft·lb overtorque, invisible. The camshaft cap bolt at 2.0 kgf·m = 14.47 ft·lb rounds to 14 ft·lb — a 0.47 ft·lb undertorque on a small bolt. Neither error is dangerous individually, but the USDM manual is always a rounded version of the JDM original. Converting directly from the JDM kgf·m value with the exact factor gives a more precise result than reading the pre-rounded USDM number. Best practice: obtain the JDM manual for the original digits, convert them yourself, and use the USDM manual only as a reference check.