Shift points: EPA testing of manual transmissions (why beating MT ratings is easier)

[B440]

Some interesting points:

"The EPA has a specialized company manufacture small batches of consistent fuel, which is 93 octane (cars running 50-state certifications get a slightly different, 91-octane “California” blend)." <--and no mention of 10% ethanol

"The EPA tested the M5 in both 400- and 500-hp modes and found no difference in the amount of fuel used. The demands of the test cycles never call upon all the M5’s horses anyway. And the Honda Insight’s econ mode—activated by a dash button and claimed to improve fuel economy—registered no effect, either. It relaxes throttle response, so the test driver simply compensates with additional throttle to achieve the required speeds. However, GM’s skip-shift device, found in the Corvette among others, irritates by forcing the driver to shift the manual transmission from first gear to fourth at low speeds and was developed precisely to improve fuel economy on the test cycles. It was approved by the EPA, i.e., not considered cheating."

"In 1998, all the major players in heavy-duty diesels (Caterpillar, Cummins, Detroit Diesel, Mack Trucks, Navistar, Renault, and Volvo) were cited for an engine-control strategy that leaned out the air-fuel ratio at steady highway speeds, which boosted fuel economy at the expense of NOx emissions. The EPA slapped those seven manufacturers with what it called the “largest civil penalty in environmental enforcement history,” a total of $83.4 million in fines." <--So this and other posts I've read on here basically lead me to believe that "lean-burning" is an easy and excellent way to save fuel, but raises NOx emissions, so automakers barely implement it?

"Don’t even think of comparing EPA figures with stand*ardized fuel-economy tests from other countries because the test cycles are very different. For example, the European highway rating, called “extra urban,” is higher than the EPA’s by about 30 percent, so a rating on that cycle of, say, 60 mpg, would be closer to 40 in this country. The mainstream press, not realizing the difference, often complains that automakers refuse to bring efficient models here when, in fact, they may not be all that efficient when measured by U.S. standards." <--Euro cars can get 75+mpg. Even dropping that down to U.S. mpgs would still be better than most cars sold in the USA.

[serialk11r]

I'm quite certain they just use the 15-25-40 shift points. My car weighs under 1 ton and has a puzzling 22mpg city rating. The Lotus Elise I think is rated 21 and that thing is another 200 pounds lighter and has a 6 speed transmission with 5% higher revs in 1st gear. Same with the Celica GT-S, which has the same drivetrain more or less (lower mpg rating all around despite more gears and an engine that's within 2% of the same efficiency).

One day I had Torque up on my phone while in heavy traffic and noticed that rolling in 1st gear my fuel economy was typically a bit under 20 mpg, and decreases substantially past 3000rpm. In 2nd, it's in the high 20s, and in 3rd it's in the low 30s. Seeing that the whole city test happens essentially in gears 1-3, basically what's going on is that they're getting maybe mid 20s steady state and then lose some gas mileage to decelerating and accelerating, and you arrive at your low 20s.

If they just shifted at 2500rpm instead (that would be like, 12-20-27) there would be no way to get under 25mpg in this car.

[jamesqf]

Originally Posted by B440

"In 1998, all the major players in heavy-duty diesels (Caterpillar, Cummins, Detroit Diesel, Mack Trucks, Navistar, Renault, and Volvo)..." <--So this and other posts I've read on here basically lead me to believe that "lean-burning" is an easy and excellent way to save fuel, but raises NOx emissions, so automakers barely implement it?

Note that that is for diesel trucks, not gasoline engine automobiles. Only Renault & Volvo even make cars.

The action was in 1998, yet the 2000-2006 Honda Insight implemented lean burn in a gasoline engine, and quite successfully.

[HyperMileQC]

Originally Posted by jamesqf

The action was in 1998, yet the 2000-2006 Honda Insight implemented lean burn in a gasoline engine, and quite successfully.

This is because the Insight is equipped with a special NOx catalyst.

[Blue Angel]

I'm going to drop a line to my buddies at Transport Canada and see if they can find anything on the topic. I never thought to ask about this while I was in the dyno cell at NRCan but I remember wondering about it as there was a Cobalt SS Turbo in line to get tested and it had a 6 speed manual.

[MetroMPG]

Thanks! I submitted a message via the "Contact Us" form on the NRCan site as well. We'll see what happens.

It would be nice to get some current information on this topic.

[MetroMPG]

Re: the question to NRCAN, the Canadian department responsible for fuel economy testing here. I just got a reply with a link to the answer in the U.S. as well:

Hello Darin,

Thank you for your inquiry regarding the manual transmission shift schedule during the federal test procedure. To answer your question, manufacturers provide the speeds at which the manual transmission should be shifted for each gear. The specific details are outlined in the US code of federal regulations in the following section Title 40 part 86.128-79. I have also provided a link to that section. eCFR &mdash; Code of Federal Regulations

Thanks
Chris B.

Here's the text from that link, current as of November 8, 2013, with the manual transmission bits highlighted:

Title 40: Protection of Environment
PART 86—CONTROL OF EMISSIONS FROM NEW AND IN-USE HIGHWAY VEHICLES AND ENGINES
Subpart B—Emission Regulations for 1977 and Later Model Year New Light-Duty Vehicles and New Light-Duty Trucks and New Otto-Cycle Complete Heavy-Duty Vehicles; Test Procedures
§86.128-79 Transmissions.

(a) All test conditions, except as noted, shall be run according to the manufacturer's recommendations to the ultimate purchaser, Provided, That: Such recommendations are representative of what may reasonably be expected to be followed by the ultimate purchaser under in-use conditions.

(b) Vehicles equipped with free wheeling or overdrive, except as noted, shall be tested with these features operated according to the manufacturer's recommendations to the ultimate purchaser.

(c) Idle modes less than one minute in length shall be run with automatic transmissions in “Drive” and the wheels braked; manual transmissions shall be in gear with the clutch disengaged, except for the first idle mode (see §§86.134, 86.136, and 86.137). The first idle mode and idle modes longer than one minute in length may be run with automatic transmissions in “Neutral;” manual transmissions may be in “Neutral” with the clutch engaged (clutch may be disengaged for engine start-up). If an automatic transmission is in “Neutral” during an idle mode, it shall be placed in “Drive” with the wheels braked at least 5 seconds before the end of the idle mode. If a manual transmission is in “Neutral” during an idle mode, it shall be placed in gear with the clutch disengaged at least 5 seconds before the end of the idle mode.

(d) The vehicle shall be driven with minimum accelerator pedal movement to maintain the desired speed.

(e) Accelerations shall be driven smoothly following representative shift speeds and procedures. For manual transmissions, the operator shall release the accelerator pedal during each shift and accomplish the shift with minimum time. If the vehicle cannot accelerate at the specified rate, the vehicle shall be operated at maximum available power until the vehicle speed reaches the value prescribed for that time in the driving schedule.

(f) The deceleration modes shall be run in gear using brakes or accelerator pedal as necessary to maintain the desired speed. Manual transmission vehicles shall have the clutch engaged and shall not change gears from the previous mode. For those modes which decelerate to zero, manual transmission clutches shall be depressed when the speed drops below 15 mph (24.1 km/h), when engine roughness is evident, or when engine stalling is imminent.

(g)(1) In the case of test vehicles equipped with manual transmissions, the transmission shall be shifted in accordance with procedures which are representative of shift patterns that may reasonably be expected to be followed by vehicles in use, in terms of such variables as vehicle speed or percent rated engine speed. At the Administrator's discretion, a test vehicle may also be shifted according to the shift procedures recommended by the manufacturer to the ultimate purchaser, if such procedures differ from those which are reasonably expected to be followed by vehicles in use.

(2) A manufacturer may recommend to the ultimate purchaser shift procedures other than those used in testing by the EPA, Provided that: All shift procedures (including multiple shift speeds) which the manufacturer proposes to supply to the ultimate purchaser are provided to the Administrator as part of the manufacturer's application for certification, or as an amendment to such application, under §86.079-32, §86.079-33, §86.082-34, or §86.1844-01 as applicable.

(h) Downshifting is allowed at the beginning of or during a power mode in accordance with the shift procedure determined in paragraph (g)(1) of this section.

[43 FR 52921, Nov. 14, 1978, as amended at 58 FR 16033, Mar. 24, 1993; 64 FR 23921, May 4, 1999]

[Arragonis]

Whilst all of this research is great, it might also be worth noting the effect of oil etc. - makers have been known to submit cars with lighter engine and gearbox oil for test in the ECE (aka EU EPA) ratings. I would find it hard to accept that US makers / importers do not do the same thing.

My last car an Aygo was in the £20 car tax bracket for CO2 emissions, cars registered 6 months later were £0. The only difference was 0w20 vs 5w20 oil and perhaps several submissions for test until one made it.

[David Philips]

I'd like to hear more manual shift strategies that work to beat EPA numbers. I have a 2000 Miata 6 speed that pretty easily beats the EPA numbers, at least in city/suburban driving in the 30-45mph range. Miatas are not optimized for mileage. The don't pull well at low RPM, they have hair trigger throttles, and they have very low gearing and in particular no long highway gear even in the 6 speed. I use a Scan Guage II and in town what helps me a lot is the ability to run along in a very high gear for the speed and periodically lift the trottle in gear to get the fuel to shut off. I can get 33-36MPG in suburban driving doing this and short shifting. EPA is 24 city/29hwy. At higher speeds with the low gearing the rapid deceleration from the low gearing makes this less useful because you lose too much speed compared to coasting.
A peculiar thing I notice at highway speed is that it gets better mileage around 70MPH than in the 55-65MPH range and this is despite turning 3,500 rpm at 70MPH. Long trips at 70 yield 32-35MPG while trips at 55-65 are typically in the 29-31MPG range. I had an earlier manual shift 626 that behaved the same way at highway speeds.

[Daox]

Pulse and glide. Its work, but it... works.