Anti-Shudder Test Methods for Transmission Fluid: What a “Meets the Specification” Claim Does Not Prove
Last verified: 2026-07-27
Procurement documentation often treats "anti-shudder tested" as one claim, backed by one number, from one test. It is not. Anti-shudder performance is measured by at least six distinct, separately-numbered standards issued by two different bodies, on hardware that produces a friction coefficient which changes depending on the clutch material it was run against — which means a bare "meets anti-shudder requirements" line on a data sheet is, by itself, close to unverifiable.
That is not a reason to distrust the fluid. It is a reason to know what question to ask next.
Why does a transmission fluid shudder at all?
Shudder is a stick-slip vibration. It shows up when a torque-converter lock-up clutch, or a wet clutch inside a DCT or CVT, is running in partial slip — neither fully locked nor fully open, the state every automatic transmission passes through on every lock-up event.
What decides whether that partial-slip moment is smooth or juddery is a single sign: the direction of the friction-coefficient-versus-sliding-velocity gradient, written dμ/dv. If friction rises as slip speed rises, the system is self-damping and nothing shudders. If friction falls as slip speed rises, small disturbances feed on themselves and grow — the shudder a driver feels as a shiver through the pedal or the console.
The friction-modifier package in an ATF or CVT fluid exists for exactly one job: hold that gradient positive for the fluid's service life. It degrades through three separate mechanisms — thermal breakdown at the sliding interface, oxidative degradation of the additive package itself, and mechanical wear of the friction-modifier film under continuous partial-slip contact. A fluid that ships with a positive gradient can still lose it in service.
That is what the methods below exist to catch, and it is where a bare pass/fail label stops carrying enough information.
What does "SAE No. 2" actually name?
This is where the confusion most easily starts: "SAE No. 2" is not a test. It is the name of an apparatus lineage.
SAE J286, "SAE No. 2 Clutch Friction Test Machine Guidelines," defines minimum equipment requirements for that machine — not a complete procedure, and not a pass/fail threshold. J286 has been carried as Stabilized since its 2012 revision, meaning it is no longer being actively updated.
Everything a supplier actually reports — friction coefficients, durability cycles, power-step results — comes from a separate procedure run on that machine, and there are three of them, each with its own number and its own revision history. A document saying only "tested on an SAE No. 2 machine" has told you the brand of the scale, not the weight.
One correction belongs here, because it circulates as a genuine false lead in sourcing conversations: SAE J2360 is a gear-lubricant standard — "Automotive Gear Lubricants for Commercial and Military Use" — with nothing to do with clutch friction or anti-shudder performance. If it appears in a fluid's anti-shudder documentation, that documentation is citing the wrong standard.
Which SAE No. 2 procedure is actually being cited?
Three procedures run on the J286 machine, and they answer three different questions.
SAE J2488, "SAE No. 2 Friction Test Machine 6000 rpm Stepped Power Test," evaluates friction behaviour across a stepped power sequence; its current revision dates from July 2019. SAE J2489, "SAE No. 2 Friction Test Machine Durability Test," first issued in 2000 and last revised in March 2012, evaluates how the wet-friction system performs as a function of the number of engagement cycles it has been through — a wear-and-aging story rather than a single-point measurement. SAE J2490, "SAE No. 2 Friction Test Machine µPVT Test," measures midpoint, endpoint and breakaway friction coefficients, and the ratios between them, across evaluation levels that vary speed, pressure and temperature at constant inertia. Its most recent revision is dated 10 September 2025.
That last date is worth sitting with. The base machine guideline has not moved since 2012, while one of the procedures that runs on it was revised in September 2025. A supplier statement that a fluid was "tested per SAE No. 2 methodology" says nothing about which of the three procedures was used, and nothing about whether it was the current revision or one several years stale. Neither omission is necessarily deceptive. Both leave a real gap in what the buyer knows.
How does the JASO family differ from the SAE family?
JASO's standards sit on different hardware, and they come from JSAE — a separate issuing body with its own numbering and its own test rig.
JASO M349, "Road vehicles — Test method for anti-shudder performance of automatic transmission fluids," runs on a Low-Velocity Friction Apparatus (LVFA). It screens whether the friction-versus-slip-speed curve keeps the positive slope described above: the core anti-shudder question, asked directly.
There is a real limit on what a citation of M349 can promise a reader outside Japan. JSAE's own published English standards list carries the 2012 English edition of M349 with a flag whose legend reads that the Japanese edition has been revised and only the older edition is available in English. A document reading "meets JASO M349" may therefore be referencing a text that is no longer the current one, and an English-language reader has no straightforward way to check which edition was actually used. Standards resellers sometimes attach a specific later year to M349; that year appears nowhere in JSAE's own list, so it is not repeated here as fact.
Does JASO M348 measure the same thing as JASO M349?
No — and this is the pairing almost nothing public gets right, because the two designations are one digit apart and both belong to JSAE.
JASO M348:2012, "Friction characteristics test method for automatic transmission fluids," runs on an SAE No. 2-type machine rather than the LVFA, and evaluates transmittable torque capacity and shifting-property durability across repeated engagement cycles. That is a durability and torque-capacity question. M349's LVFA screen is a shudder-tendency question. A fluid can be documented against one without ever having been run against the other, and a line reading simply "JASO tested" does not say which.
Can a passing screen predict long-term field durability?
Not reliably on its own, and this is documented rather than speculative. SAE Technical Paper 2007-01-1974, "Test Methods for Determining Anti-Shudder Durability of Automatic Transmission Fluids," reports that some fluids passing an LVFA-only JASO M349 screen do not perform well when evaluated by a method that adds high-energy engagements and oxidative aging.
The reason tracks back to the three degradation mechanisms already named. A fresh-oil LVFA pass shows the additive package works when it is new. It does not show how that package holds up after the cycling and heat a transmission delivers over years of service. A durability claim needs a durability protocol behind it — cycle-count testing, or an equivalent aged-oil evaluation — not a single fresh-fluid screen presented as though it settles the question.
Why isn't a friction number printed on a technical data sheet?
A TDS or COA is built to report fluid-only, batch-verifiable bulk properties: viscosity, flash point, pour point — numbers belonging to the fluid alone, re-measurable from any sample.
A friction coefficient cannot be reported the same way, because it is not that kind of number.
A friction coefficient is not a property of a fluid; it is a joint property of the fluid and the specific clutch friction material and surface finish it was run against. Change the friction material and the same fluid produces a different curve. That is exactly why OEMs run their own friction and durability rigs on their own clutch hardware before issuing an approval, rather than accepting a generic supplier figure — the number only means something in the system that generated it, and a fluid-only data sheet has no system to attach it to. Publishing one figure without naming the friction material behind it would misstate what that figure proves.
This is where the distinction set out in the meets-specification versus OEM-approved analysis turns concrete. An OEM licensing programme means the friction and durability testing was actually run against that OEM's own hardware, under a dated, checkable licence. A self-declared "meets" statement does not require that any of these procedures were ever run at all.
What should a buyer request instead of a number?
Not a friction coefficient. Ask for the paperwork that makes a coefficient mean something.
- The OEM licence or approval reference itself, checked against the OEM's own channel rather than accepted from a printed certificate alone.
- A written statement naming the specific method used — "evaluated per JASO M349", or the specific SAE No. 2 procedure — together with the revision, and a compliance or pass statement. This is a reasonable request even where full numeric results stay proprietary.
- Confirmation of which clutch friction material the evaluation ran against. The result does not transfer to a different friction system, so this line decides whether the evaluation is even relevant to the transmission in question.
- Whether the claim rests on fresh oil or on a cycling or aged-oil durability protocol. A durability claim requires the latter; a fresh-oil screen is a starting point, not an endpoint.
None of these four requires a supplier to disclose proprietary formulation data. All four separate a documented claim from an assumed one.
Which standard answers which question?
| Standard | Issuing body | What it evaluates | What it does NOT tell you |
|---|---|---|---|
| SAE J286 | SAE | Minimum equipment requirements for the SAE No. 2 friction test machine | Any pass/fail result — it defines the apparatus, not a procedure |
| SAE J2488 | SAE | Stepped-power friction behaviour on the SAE No. 2 machine | Durability across engagement cycles |
| SAE J2489 | SAE | Wet-friction-system durability as a function of engagement-cycle count | A friction coefficient mapped across operating conditions |
| SAE J2490 | SAE | Midpoint, endpoint and breakaway friction coefficients and their ratios, across speed, pressure and temperature levels | Cycle-count durability — it maps friction, it does not age the fluid |
| JASO M349 | JSAE | Anti-shudder tendency, via the slope of the friction-versus-slip-speed curve, on an LVFA | Torque capacity or shifting-property durability |
| JASO M348:2012 | JSAE | Transmittable torque capacity and shifting-property durability over repeated engagement cycles, on an SAE No. 2-type machine | Anti-shudder tendency specifically — that is M349's question |
Six standards, two bodies, one apparatus family, and no single number that carries all of it. For the wider ATF, CVT and DCT specification landscape these methods sit inside, see the transmission fluid specifications guide; suppliers listing under these specifications can be reached through the transmission fluids category, where the method and revision questions above belong in the quotation request itself.