MICHAEL DURACK is Technology Director at Ultimate Transmissions Pty Ltd in Queensland (Australia).
A scooter delivers over double the mileage of a typical car. However, the scooter's tank to wheel efficiency is less than half that of the car. Ultimate Transmissions looks at why the scooter is so inefficient, its impact on the Indian economy and environment and how the problem may be solved without adding cost, using an innovative continuously variable transmission (CVT).
 Tank to wheel efficiencies of cars, scooters and motorcycles
TANK TO WHEEL – ENERGY CONVERSION EFFICIENCY
Internal Combustion Engines (ICE) convert the chemical energy in fuel to mechanical energy at the engine's crankshaft. The crank drives the transmission, which is connected to the wheels. The relationship of the quantity of chemical energy in the fuel to the quantity of mechanical energy delivered to the wheels is a true measure of the overall drive chain efficiency. The amount of fuel needed to move a vehicle over a fixed journey is a derivative of this conversion efficiency, the vehicle weight, aerodynamic characteristics, rolling resistance and the journey itself. Just because one vehicle can be driven for a longer distance on the same amount of fuel as another vehicle does not necessarily mean that it is more efficient.
Both the engine and transmission waste energy, with the greatest loss being within the engine itself. The overall conversion efficiency is often called the tank to wheel efficiency . The efficiency of the conversion varies from vehicle to vehicle, with the efficiencies of cars, scooters and motorcycles being typically 24 %, 9 % and 12 % respectively .
 Overall drivetrain efficiencies of cars, scooters and motorcycles
Scooters and motorcycles are much less efficient than the car because the efficiency of both the engine and transmission of smaller vehicles is much lower than that of the car. This is not something fundamental to the size of the vehicle's engine or transmission. Instead, it comes about because of the way the bike and scooter transmissions are designed.
The car's engine runs more efficiently than motorcycle and scooter engines because the transmission allows it to run at much lower RPM. ICEs run far more efficiently at low speed and high torques [2,3,4]. They are most efficient within a "sweet spot", where engine speeds are low, power moderate and torque high. They are least efficient when running at low power, high RPM, and low torque. The peak efficiency of a simple gasoline engine can be as high as 27 %. However, this efficiency is only present within a small section of its operation with the efficiency dropping rapidly to below 10 %, when the engine is running at high speed and low power. Unfortunately, for scooters and motorcycles, this inefficient state is precisely where they are designed to operate, as it simplifies the transmission and lowers cost. The effect of advanced transmission technologies, ratio spread and numbers of gears is discussed later in this article.
The rubber belt CVTs used in scooters have very poor mechanical efficiencies, when running at low power, and these are even less efficient when running at low power and high RPM . Some additional losses are associated with the centrifugal clutch, which typically exhibits more slip than the automotive equivalent. Motorcycle transmissions are less efficient than car transmissions, because the quality of the gears is lower, the churning losses within the very small casing are higher and the rotational speeds are typically much higher.
The end result is that a typical scooter's tank to wheel efficiency is less than half that of a typical late model car. A typical small-capacity motorcycle does not perform much better as its transmission, although reasonably efficient, mechanically forces the engine to run at high RPM. This somewhat surprising comparison is borne out by analysing test results for the three vehicles carried out by the Automobile Research Association of India (ARAI).
 MIDC test for cars  WMTC test for motorcycles
ARAI TEST RESULTS
All automobiles are subjected to standardised tests to establish their fuel efficiencies and potential to pollute. These tests are not necessarily intended to perfectly match the real world but they are intended to ensure that when comparisons are made they are done on an 'apples for apples' basis, which is repeatable and legally enforceable. The test cycles describe a hypothetical "journey" to be carried out by a vehicle that involves periods of acceleration, cruising, and stopping over a fixed period and a fixed distance. The tests are carried out on a dynamometer that accurately reproduces the resistance that the vehicles will need to overcome, in order to replicate the journey on a real world road.
In India, the test cycles used by the ARAI to evaluate fuel consumption for private vehicles is the Modified Indian Drive Cycle or MIDC . For small bikes and scooters, it is the Worldwide Harmonised Motorcycle Emissions Certification (WMTC) .
Both these cycles have been developed or derived from Western drive cycles so that they better match conditions in India, but can still be related to the corresponding testing in Europe or the USA. Generally, both cycles involve lower top speeds than are used in the equivalent Western examples.
When an automobile executes the MIDC test, regardless of its size, engine type or style it must follow the journey described with a high level of accuracy, while tailpipe emissions and fuel consumption are being monitored. The mechanical energy delivered to the wheels and required to execute the test can be calculated with a very high level of accuracy, when the vehicle characteristics of gross weight, aerodynamic drag coefficient (CD) and rolling resistance (RRC) are known. Big cars, small cars, motorcycles and trucks are all bound by the three fundamentals of gross weight, aerodynamic drag and rolling resistance, when operating over any of these cycles. It is important to understand that regardless of which test journey (cycle) is being used, the tank to wheel efficiency will remain very similar.
When the vehicle described in , (a petrol-powered compact SUV) performs the MIDC test, it requires 1.12 kWh of mechanical energy to be provided to its drive wheels. This energy can be calculated very accurately, when the vehicle characteristics are known and are accurate. The test covers a distance of 10.647 km at an average speed of 32.48 km/h. It consumes 0.48 l of fuel, which is an overall mileage of 22.2 km/l or 4.5 l/100km.
Unleaded petrol contains around 32.4 MJ (9.5 kWh) of thermal or chemical energy per litre. 0.48 l (the amount consumed in the test) holds 4.56 kWh of energy. The tank to wheel efficiency is energy out/ energy in or 1.12/ 4.56 = 24.4 %. The tank to wheel efficiency already predicted in the efficiency analysis matches very closely to this tested evaluation.
 MIDC test specifications  WMTC test specifications
A small scooter, with a 125 cc engine and the specification set out in , carrying out the WMTC (part 1 reduced) test requires 0.068 kWh of energy to be delivered to the back wheel in order to execute the test. It covers a distance of 3.937 km at an average speed of 23.6 km/h. It consumes 0.079 l of fuel or 50 km/l or 2 l/100 km.
The energy in the fuel consumed to complete the test is 0.75 kWh. The tank to wheel efficiency is energy out/ energy in or 0.068/ 0.75 = 9 %, also predicted using an analysis of engine and transmission efficiencies.
This difference is not being caused by any difference in how the testing is carried out. It is just a simple reflection of a fact that small scooters are extremely inefficient, when compared to a modern car. Motorcycles are more efficient, with a typical tank to wheel efficiency of 12 % but are still way behind the tank to wheel efficiency of cars.
The analysis of the tank to wheel efficiencies can also be expressed by looking at the actual wheel energy that is extracted out of one litre of fuel by a car, scooter and bike, when executing these cycles, .
 Actual wheel energy that is extracted out of one litre of fuel by a car, scooter and bike, when executing these cycles
The extra mileage should not be seen in any way as indicating that the different vehicles have similar or equivalent efficiencies, it is only an indication of how physically small motorcycles and scooters are, in comparison with cars. It is the relative tank to wheel efficiency that is the only true measure of the performance of the vehicles drivetrain.
If scooters were operating at the same energy conversion efficiency as the car, they would travel over 130 km on one litre and a bike would travel 160 km. However, without fuel injection, stop-start, and variable valve timing, this overly optimistic target will not be possible. An improvement in mileage of 100 % for scooters and 70 % for bikes is an achievable target, without any changes to the engine itself. Although engine downsizing may be possible with the improvements in performance, it is considered that the current 100 cc to 125 cc engine size will remain the most suitable for the majority of two-wheeler buyers in India.
The solution lies not with improvements in the engines themselves but simply with improvements in the transmissions design. The gains in fuel conversion efficiencies of modern four-wheelers over the last 20 years have predominantly come from improvements in the transmission . Most of this improvement came from increasing the number of gears, which enabled an increase in ratio spread. Some came from improvement in transmission efficiencies and others from technologies such as variable valve timing, start stop, and fuel injection.
At the same time that the automobile's efficiencies improved, the responsiveness and drivability of the vehicles also improved without increasing the base model's costs. The improvement in fuel efficiencies caused a typical North American vehicle to move from 20 m/gl to 40 m/gl, when comparing vehicles of similar weight and power. The current fleet average target for the US Environmental Protection Authority (EPA) is 54 m/gl, which they aim to achieve by 2025.
It is important to remember that while these improvements were made in energy conversion efficiency, private vehicles, particularly in the US, have increased in size and power by a factor of almost two. This has meant that the average mileage of a North American car has only improved by around 30 % in real terms.
The typical scooter transmission, using the rubber belt CVT, has a ratio spread of less than 3, while the four-speed motorcycle transmission is around 4. A modern automobile has a six speed transmission and a ratio spread of more than 5. Most automatics are now fitted with 7 gears and a ratio spread of at least 7, with some going up to 10 gears and a ratio spread of 10.
Ultimate Transmissions has developed a CVT that has a ratio spread of 9.5, which can simply replace the rubber belt without adding cost, or requiring significant modification of overall design. The transmission uses double rollers in place of the more conventional single rollers. The efficiency of this CVT is much higher than the earlier types .
The transmission design is based on a traction drive CVT called a DFTV-CVT . This uses hard steel rollers clamped between hard steel discs and by rotating the rollers, different ratios are delivered in much the same way as they are delivered in the rubber belt. However, the traction drive is much more efficient than the rubber belt, particularly at low power, and its ratio spread is more than double. It exhibits a very large power density .
 DFTV-CVT uses hard steel rollers clamped between hard steel discs
This transmission is lighter than the rubber belt transmission and similar to a four-speed motorcycle transmission . It improves the acceleration, hill climbing ability and top speed. It is lighter in weight than a belt CVT and removes almost 10 kg of un-sprung weight from the back wheel. A scooter can be designed taking advantage of the un-sprung weight reduction using tyres designed for low rolling resistance and lower inertia, not improved ride. The DFTV-CVT is more durable than the rubber belt and its performance does not deteriorate over time as does the rubber belt.
 DFTV-CVT is much more efficient than the rubber belt CVT, particularly at low power, and its ratio spread is more than double. It exhibits a very large power density, is lighter than rubber belt transmission and improves acceleration, hill climbing ability and top speed
The DFTV transmission will increase the tank to wheel efficiency from 9 % to more than 18 % by simultaneously increasing the engine operating efficiency to 21 % and the mechanical efficiency of the CVT to 87 %. This enables a scooter to ride over 100 km on one litre of fuel and a bike with lower rolling resistance tyres reach 110 km, while retaining the automatic functionality.
The annual savings that can be delivered to the owner of a scooter are equal to around 10 % of the capital cost of the scooter. They would typically offset around 30 % of the bank repayments or allow the scooter to be sold for 25 to 30 % more (approximately ' 15,000) to customers, who believe the story or are environmentally aware. The effect of Indian buyer reaction to the cost of fuel efficiency is discussed in .
ENVIRONMENT AND THE ECONOMY
India is currently producing over 16 mn two-wheeled vehicles per annum and the number is increasing at more than 5 % every year. Globally, the number of bikes and scooters using internal combustion engines is around double this.
It is highly likely that as many as 200 mn new two-wheelers will be produced and will be plying India's roads by the year 2025. Using current trends as an indicator, the split between motorcycles with manually operated transmissions and automatic scooters will be around 50:50. An improvement in the efficiency of this fleet will have a profound and beneficial effect on the global environment, the Indian economy, and individual Indian consumers.
This bike and scooter fleet will consume 36 mn tonne of petrol if current tank to wheel efficiencies are maintained. They will emit 83 mn tonne of CO2 every year. An improvement in fuel efficiency of 30 km/l of this fleet will reduce annual fuel consumption by 13.5 mn tonne and reduce CO2 emissions by 31 mn tonne.
India is currently responsible for the creation of 6 % of the global production of CO2, emitting close to 2,000 mn tonne annually. An improvement in overall efficiency of its two-wheeler fleet by 2025 would see a reduction in CO2 emissions of almost 1.55 % of its current CO2 footprint. India currently imports 4 mn barrels of crude oil per day or 240 mn tonne per annum. The improvement in efficiency would reduce its need to import crude oil by over 6 % of today's imports.
Indian consumers are very experienced at identifying value for money and within the automotive field, demonstrate a willingness to pay for the future benefits of fuel saving . This will give the OEMs an opportunity to generate higher profits out of fuel efficient vehicles and to recover the investment costs associated with funding the change from an inefficient product to an efficient one.
 Chen D.W., Lee D.W., Sung C.K. An Experimental study on transmission efficiency of a rubber belt CVT. Mechanism and design Theory Vol 33. No 4. Pp. 351-363, 199. Department of Power Mechanical Engineering National Tsing Hua University Hsinchu Taiwan 300 Republic of China.
 Goering C., Stone M., Smith D., Turnquist P., Engine Performance Measures, Chapter 2 in Off-Road Vehicle engineering Principles 19-36 St Joseph Mich. ASAE American Society of agricultural Engineers,
 Myer J., Engine Modelling of an Internal Combustion Engine with Twin Independent Cam Phasing. Ohio State University Thesis 2007.
 Moawad, A. and Rousseau, A., "Impact of Transmission Technologies on Fuel Efficiency to Support 2017-2025 CAFE Regulations," SAE Technical Paper 2014-01-1082, 2014, doi:10.4271/2014-01-1082.
 Greiner, J., Grumbach, M., Dick, A., and Sasse, C., "Advancement in NVH- and Fuel-Saving Transmission and Driveline Technologies," SAE Technical Paper 2015-01-1087, 2015, doi:10.4271/2015-01-1087.
 De Novellis L., Carbone G., and Mangialardi L. Traction efficiency performance of the double roller full-toroidal variator; a comparison with half and full-toroidal drives. ASME, Journal of mechanical design, 2012, vol.134; 071005-1 – 071005-14.
 Walker P., Durack J., and Durack M. Laboratory testing of a new form of toroidal CVT. FISITA Conference June 2014.
 Chugh R., Cropper M., NarainU., The cost of fuel economy in the Indian passenger vehicle market Department of Economics, University of Maryland, College Park, 3105 Tydings Hall, MD 20742, USA Resources for the Future, USA World Bank.
(ATR hasn't independently verified the test results as claimed in this article)