Why Japan Never Let Go of the Kilogram-Force-Meter
Japan adopted the metric system in 1924, but it built its industrial base on the gravitational metric system — the MKS-gravitational variant where the kilogram-force, not the newton, was the base unit of force. This was not a quirk. The gravitational system was intuitive: one kilogram-force is the weight of one kilogram of mass in standard gravity. A mechanic could feel it. A factory worker could calibrate against it. The newton, a physicist's abstraction, offered no such intuition.
JIS (Japanese Industrial Standards) torque specifications were published in kgf·m through the 1980s. JIS B 1083, the standard for tightening torque of threaded fasteners, was issued in kgf·m as late as 1990. The transition to N·m in Japanese engineering was slower than in Europe because Japan's domestic market never demanded it — every Japanese mechanic already understood kgf·m as a physical quantity. Even today, JIS B 1083-1:2022 prints kgf·m as the primary torque spec with N·m in parentheses. The dual-unit approach preserves backward compatibility with every service manual published since the 1960s. Retraining an entire generation of mechanics to think in newtons costs more than any car company is willing to spend, and the risk — a mechanic misreading "118" as "12" on a cylinder head — is one no manufacturer will accept.
The kgf·m's survival is not a Japanese anomaly. German engineering used the identical unit under the name kilopond-meter (kp·m) until the 1970s, when DIN 1319 formally adopted the newton. Older German machine tools, lathes, and presses still carry kp·m torque plates. A German mechanic restoring a 1960s Deckel milling machine faces the same conversion as a Toyota technician. The unit died on paper but lives in every workshop that services pre-transition machinery.
The g₀ Problem: Why 9.80665 and Not 9.81
Standard gravity g₀ was defined by the 3rd CGPM in 1901 as exactly 9.80665 m/s² at 45° latitude at sea level. This is not a measurement of actual gravity anywhere on Earth. Real gravity varies from 9.76 m/s² at the equator to 9.83 m/s² at the poles, depending on latitude, altitude, and local geology. The CGPM fixed g₀ at 9.80665 to approximate gravity at 45° N — halfway between the equator and the pole — and to eliminate the ambiguity that would arise if every national standards laboratory calibrated its kilogram-force against local gravity.
A kgf·m calibrated at the equator against local gravity (g ≈ 9.78 m/s²) would equal 9.78 N·m. The same kgf·m calibrated in Helsinki (g ≈ 9.82 m/s²) would equal 9.82 N·m. Two mechanics working from the same spec would apply different torques. The CGPM's fixation of g₀ eliminated this geographic dependency. Every kgf·m value on this site, and in every modern engineering standard that references the kilogram-force, uses g₀ = 9.80665 m/s². The value is a convention, chosen by committee in 1901, and it has never been revised.
The practical consequence: when a Japanese torque spec says 12.0 kgf·m, it means 12.0 × 9.80665 = 117.6798 N·m. Rounding to 118 N·m is standard practice. The 9.81 shortcut — used for mental arithmetic — introduces an error of about 0.03%, which on a 100 N·m bolt is 0.03 N·m. Most torque wrenches are accurate to ±3%, so the 9.81 approximation is well within instrument tolerance. But there is no engineering reason to use it when the exact value is available.
1 kgf·m = 9.80665 N·m (multiply kgf·m × 9.80665)
Common kgf·m to N·m Conversions
| kgf·m | N·m | Where you'd see this |
|---|---|---|
| 0.5 kgf·m | 4.90 N·m | Fuel pump mounting bolts, small sensor brackets. |
| 1.0 kgf·m | 9.81 N·m | Valve cover bolts, oil pan bolts (small engines). |
| 2.0 kgf·m | 19.6 N·m | Camshaft cap bolts, intake manifold fasteners. |
| 3.0 kgf·m | 29.4 N·m | Water pump bolts, alternator bracket bolts. |
| 5.0 kgf·m | 49.0 N·m | Spindle bearing preload (machine tools), oil filter housing. |
| 8.0 kgf·m | 78.5 N·m | Connecting rod bolts (small-displacement engines). |
| 10.0 kgf·m | 98.1 N·m | Motorcycle axle nuts (Honda, Yamaha, Kawasaki rear wheels). |
| 12.0 kgf·m | 118 N·m | Cylinder head bolts, first-pass torque (Toyota 2JZ, 1JZ, 7M). |
| 15.0 kgf·m | 147 N·m | Main bearing cap bolts, engine mount brackets. |
| 20.0 kgf·m | 196 N·m | Flywheel bolts, pressure plate bolts. |
| 30.0 kgf·m | 294 N·m | Harmonic balancer / crankshaft damper bolts. |
| 50.0 kgf·m | 490 N·m | Crank pulley bolts (large-displacement engines, diesel). |
Worked Examples
Toyota 2JZ-GTE cylinder head bolt — 12 kgf·m → 118 N·m
The 2JZ inline-six, the engine behind the MKIV Supra, specifies cylinder head bolt torque in three stages. The first pass is 12.0 kgf·m (118 N·m). The second pass adds a 90° angle. The third adds another 90°. If you apply only 12 N·m thinking the spec is in newtons, you have undertorqued the head by a factor of 9.8 — the head gasket will leak on first start. This is the mistake every non-Japanese 2JZ builder learns about the hard way.
Honda CBR600RR rear axle nut — 10 kgf·m → 98.1 N·m
Honda's middleweight supersport specifies the rear axle nut at 10 kgf·m. A mechanic using a European-market torque wrench set to 98 N·m is within 0.1 N·m of the target. But a mechanic who misreads "10" as newton-meters will undertorque the axle by 9.8× — a dangerous condition for a motorcycle that sees 14,000 RPM and 160 mph. The error is silent until the rear wheel walks in the swingarm.
Japanese lathe spindle bearing preload — 5 kgf·m → 49.0 N·m
Precision machine tool spindles manufactured in Japan — Mori Seiki, Okuma, Makino — often specify bearing preload in kgf·m in their setup documentation. A 5 kgf·m preload converts to 49 N·m. Over-preloading by using 5 N·m (a 10× under-torque) leaves the spindle loose and inaccurate; using 50 kgf·m (thinking "50" is the newton value) crushes the bearing races. Spindle preload tolerances are typically ±0.5 kgf·m — about ±5 N·m — so the conversion must be exact.
Keihin carburetor float bowl screws — 0.2 kgf·m → 1.96 N·m
Small-engine carburetors — Keihin, Mikuni, Dell'Orto — use tiny M4 and M5 fasteners torqued to fractional kgf·m values. A Keihin CV carb on a Honda motorcycle specifies float bowl screws at 0.2 kgf·m (1.96 N·m). This is below the range of most click-type torque wrenches and close to the limit of what a mechanic can feel by hand. Over-torquing these screws to even 0.5 kgf·m (4.9 N·m) strips the zinc casting. The conversion from kgf·m to N·m at this scale forces the mechanic to use a beam-type or digital torque wrench capable of single-digit newton-meter resolution.
Engineering Context
The kgf·m-to-N·m bridge exists wherever Japanese-engineered machinery meets non-Japanese service tools. JIS B 1083 (tightening torque for threaded fasteners) was published in kgf·m through 1990 and dual-printed in kgf·m/N·m thereafter; JIS B 1180 covers hexagon bolts and screws with torque tables in both units. Japanese automotive engineering — Toyota, Honda, Nissan, Mazda, Subaru — continues to dual-spec torque values in service literature. The German DIN transition from kp·m to N·m under DIN 1319 occurred in the 1970s, but pre-transition machine tools (Deckel, Maho, Gildemeister lathes) still carry kp·m torque plates that require conversion. In machine tool engineering, spindle bearing preload specifications in kgf·m appear on Japanese CNC equipment exported globally, and misreading the unit when setting up a Mori Seiki or Okuma machining center can destroy a spindle that costs more than the car the mechanic drove to work. The conversion factor 9.80665 is embedded in every torque wrench calibration performed to ISO 6789, which requires traceability to the newton — meaning even a "kgf·m" wrench is ultimately calibrated against the SI definition of force.
More: N·m to kgf·m · kgf·m to ft·lb · ft·lb to kgf·m · ft·lb to N·m · N·m to ft·lb
Related Unit Converters
Frequently Asked Questions
Is kgf·m the same as kg·m? What about m·kg?
Technically no. kg·m is ambiguous — it could mean kilogram-mass times meter, which is not a torque quantity. kgf·m explicitly means kilogram-force times meter. In Japanese engineering documents, "kg·m" almost always means kgf·m — the "f" is dropped in common usage but implied by context. No engine produces torque in kilogram-mass-meters; the "f" is a formality that engineers omit. "m·kg" is the same quantity with the units written in reverse order, common in older German documents (m·kp). All three notations — kgf·m, kg·m, kp·m — refer to the identical torque value: multiply by 9.80665 to get N·m.
Why does my Japanese torque wrench say "kpm"?
"kpm" is kilopond-meter, the German name for kilogram-force-meter. One kilopond equals one kilogram-force exactly — the pond was the gravitational unit of force in pre-SI German engineering. Postwar Japan adopted German engineering conventions heavily, and many Japanese torque wrenches manufactured from the 1970s through the 1990s carry the German "kpm" marking rather than "kgf·m." The conversion is identical: 1 kpm = 9.80665 N·m. If your wrench says "m·kp" (meter-kilopond), it is the same unit written in German word order rather than English word order — the torque value is unchanged.
Does 1 kgf·m equal 9.81 or 9.80665 N·m?
9.80665 exactly, by the 1901 CGPM definition of standard gravity g₀. The value 9.81 is a two-decimal-place approximation used for quick mental arithmetic, accurate to within 0.03%. For torque-critical fasteners — cylinder head bolts, connecting rods, main bearing caps — use 9.80665. The 0.03% error on a 100 N·m bolt is 0.03 N·m, well within the ±2–4% accuracy of most click-type torque wrenches. But there is no engineering reason to approximate when the exact value is a single keystroke away. The value 9.8 is an additional digit worse and should not be used for any torque calculation where the result matters.