Winterkorn says Volkswagen developing 10-speed DCT and high-performance diesel; plug-in hybrids offer great potential
U. Delaware and NRG Energy vehicle-to-grid project selling power to PJM

Honda progressing with transverse flux motor for hybrid powertrain

Left: Structure of original TF motor. Right: new TF motor. -32% is the relative reduction in axial length compared to a conventional motor. Takizawa et al. Click to enlarge.

Honda has proposed and is developing a Transverse Flux motor (TF motor) in order to shorten the axial length of the motor for hybrid electric vehicles (HEVs). At the 2013 SAE World Congress, Honda engineers described their progress in improving the new type of three-dimensional magnetic circuit motor.

In contrast to conventional stators composed of a stator core (made from magnetic steel sheet) and winding wires, the TF motor’s stator is composed of a soft magnetic composite (SMC) core and a coil. While reducing axial length and achieving a simple stator architecture comprising only five parts, the new motor raises issues including the need further to improve motor efficiency (currently at 83.7%) and the development of techniques for the manufacture of rectangular wave-shaped coils.

Background. Honda’s front-wheel drive IMA hybrid powertrain system features a motor located between the internal combustion engine and a continuously variable transmission (CVT); as a result, the width of the powertrain increases according to the axial length of the motor. To use the hybrid vehicle system on various vehicle types with different width requirements, the axial length of the motor needs to be shortened.

In an effort to shorten the axial length of the motor, Honda engineers focused on eliminating the coil end, which does not contribute to the generation of torque, by proposing a transverse flux motor, reported in a paper in 2011 (Aoki and Takahashi). The original TF motor featured internal ring-shaped coils and new type of flux pat. The design was later improved to counter magnetic saturation and magnetic “short circuits”.

While the stator of a salient-pole concentrated winding motor, in which each tooth is a separate part, is composed of 110 parts, the TF motor uses a simple stator made by stacking the three 3-dimensional stator cores and two rectangular wave-shaped coils.

However, efficiency in the original design was relatively low (79.1%) due to iron loss characteristics of the soft magnetic composites (SMC) used to make the stator). Too, manufacturing methods for producing rectangular wave-shaped coils from ring-shaped coils was also an issue.

In the new SAE paper, Honda engineers describe methods for reducing iron loss, the development of a method for manufacturing rectangular wave-shaped coils, and the creation and testing of a prototype TF motor that incorporates these elements.

Efficiency. To improve efficiency, the Honda engineers conducted a parametric study on the constituents of the SMC core and manufacturing conditions, and developed SMC specification conditions that reduce iron loss.

Specifically, they needed to reduce the hysteresis loss and the eddy current loss to bring the TF motor iron loss close to their target.

To reduce hysteresis loss, they needed to reduce the coercivity of the iron powder. Principal factors here are the grain boundary of the iron powder and the strain of the iron powder during the compacting process. They therefore sought to increase the grain size through annealing of the iron powder, as well as means of preventing strain by increasing the heat treatment temperature after compaction.

To reduce eddy current loss, the insulation coating of the iron powder must remain intact. The insulation coating can be destroyed during either the compacting or the heat treatment process, and must be able to withstand the heat treatment temperature of 650 °C.

Compacting destroys the MgO coating because the membranes split through contact with each other. This can be prevented by reducing the roughness of the iron powder surface and improving its fluidity during compacting. The engineers therefore examined using a centrifugal mill to reduce the iron powder surface roughness, and increasing the volume of resin binder and using a lubricant to improve iron powder fluidity.

Three-dimension moveable press. Takizawa et al. Click to enlarge.

Manufacturing. Honda engineers developed a coil film that fulfills the requirements for rectangular wave-shaped coil formability and insulation, while also developing winding wires and press-forming processes required to manufacture rectangular wave-shaped coils.

The coil film was produced using an extrusion process that allows a thick film to be made, while PFA (perfluoroalkoxy resin) was selected as the film material due to its elongation properties, which aid forming. For the winding wires process, both lead wires are wound on the outer circumference side, and a three-dimensional movable press-forming method was developed to bend ring-shaped coils into rectangular wave-shaped coils while reducing the coil diameter.

TF motor prototype. Takizawa et al. Click to enlarge.

Prototype. The engineers built a TF motor prototype using the improved SMC core and the new coil-forming method; maximum torque was 140 N·m (103 lb-ft).

Tests on the prototype showed that average motor efficiency when driving in the JC08 mode improved by 4.6 percentage points from 79.1% to 83.7% due to a reduction in iron loss.


  • Takizawa, D., Takahashi, T., Shimizu, H. and Kato, R., “Development of Transverse Flux Motor with Improved Material and Manufacturing Method,” SAE Int. J. Passeng. Cars – Electron. Electr. Syst. 6(1) doi: 10.4271/2013-01-1765

  • Aoki, S. and Takahashi, T., “Development of Compact Transverse Flux Motor with a New Magnetic Circuit Configuration,” SAE Int. J. Engines 4(1):314-322 doi: 10.4271/2011-01-0348



Honda and Nissan have recently been awarded patents for axial motors. Motor efficiency is usually quoted as a peak of say 90% so people believe they are 90% efficient across the whole operating range, which is not so.


Honda is designing this motor for the IMA system which is focussed on the 0-1800 rpm range mainly. It is generally accepted that it is particularly difficult to get high effcy levels at 60Hz excitation levels, consequently all this effort to do with alternative magnetic structures.

It should be pointed out that the standard induction motor efficiency will rise with applied frequency perhaps exceeding 95%. A property of the induction motor that I've noted is that constant copper loss is expended per unit torque i.e. rotations from 1 rpm to 10,000 rpm will incur a constant loss while providing constant torque. It is therefore evident that effcy must therefore improve with rpms. Of course iron loss is to be expected rise with the increasing frequency of excitation in order to gain those rpms however these iron losses can be mitigated with the use of thinner stator laminations as in those 400Hz motors employed in the avation industry.

This new Honda motor is application specific to the company's IMA system and is not going to be suitable as an EV traction motor, but then neither are MG1 and MG2 on the Prius for that matter. However I wouldn't be surprised if it found further use as the generator on later variants of the new 2014 Accord hybrid design.


Seems like the same mind set as with the Wheel-Motors.

Why try to design out normal, well developed, low cost, compact motors that only require simple, cheap shafts/gears to tie them in.

Poor answers looking for problems.


In this case there is a need for a more powerful axial motor to go between the engine and transmission to take up less space.

They have made good advances here, the motor patents I saw from Honda were quite impressive. The market will decide which version from which car maker will be most popular.

Markets don't always pick the technically superior product. For example Beta was technically superior to VHS, but VHS out sold Beta 10 to 1. Perception and inertia are factors beyond technical superiority.


Why not use e-motors developed for in-wheel EVs, when space (thickness) is a problem?


Honda has made a good motor, if it has to be thin.

Step back and decide if and why it really does have to be thin.

Just because it worked well between the engine and transmission for the Insight I, does not mean it is the way to go.

With the right (self imposed) constraints, the full sized Hummer is an excellent design.


Two years ago I saw a huge 'International' full Hummer like Pick-up but it looked at least 20% to 30% larger. I haven't seen another one since.

Was that a good or bad design? It looked much better than a full Hummer with smoother lines. It looked a lot like a Class 8 International heavy truck, from the front.


Years ago Raser turned a Hummer H3 into a hybrid by putting a motor between the transmission and drive shaft.

VIA motors makes stock 3/4 ton pickups and vans into hybrid by putting a motor in the drive line. The more fuel we save the less oil we import, in that case it IS all good.

Roger Pham

Agree with TT, the IMA concept is not an optimal way to make an HEV. Braking energy recuperation is not as good as in the design in which the motor can be disconnected from the engine, and for the same reason, inefficient electric-only mode because the integrated motor must drag the unpowered engine along, having a lot of engine's internal friction.

A small high-speed starter-generator always connected to the engine via gear drive, with a larger motor that can be clutched and unclutched from the engine, would make a more efficient HEV design, like Hyundai is doing right now. Honda is going toward a 2-motor HEV design with the Accord PHEV.


What do you think of using two-phase, two pair of poles induction motor up to 14,000 rpm, instead of three phase one (as used in Tesla Roadster) ?
Actually the first IM, patented by Tesla, were two phase ones. Three phase systems later prevailed because they needed less wires for power transfer to distance. As inverters are here close to motor, that advantage no longer exist.
Two phase motor would require fewer coils (fewer switching elements too), could have smaller diameter, but would need to have higher L/D ratio for the same torque. Its rotor would have smaller angular moment (less load on bearings), but may be more difficult to cool.
I did a search, apparently some 2-phase induction motors are used in the aviation industry.

An interesting invention was announced in October 2012:

A researcher at Indiana University-Purdue University Indianapolis (IUPUI) has invented a new class of power inverter that could lead to cheaper and lighter traction motors and power electronics for EVs.


Mitsubishi Electric Corp has developed a variable frequency/speed liquid cooled compact high efficiency EV motor with an integrated high temperature silicon carbide inverter.



Well, I would have to agree that a quadrature system would meet a constant power minimal torque ripple requirement. And as you imply it could be implemented with just four NPN power devices if each phase winding was center tapped. The downside of powering each half winding alternately would mean a 1/root2 derating compared to a bridge. Notwithstanding there is always that added benefit to have IMs operate to 14000 rpm.

I don't understand why Honda, in the first place, would even consider a 12Kw automotive electric drive specifically within its 0-1500 rpm range. Finally their Accord Hybrid due out this year is using a 125Kw motor operating to 15,635 rpm which will probably be the same size.


Thanks for the reply.
At these two links you can find something on designing inverters for two phase motors.

WrightSpeed demonstrated earlier two speed clutchless shifting, with ultra-fast synchronization, using IM. I think it was a very similar IM to the one in Tesla Motors (the founder worked for Tesla Motors, then left), AC propulsion, or perhaps a Fukuta one.

In 2011 there was similar video on his site, with some info on rpm, and shift times (it was below 0.1 sec), now only as video on YT:

Smaller diameter rotor makes easier quick synchronization of motor's rpm, as moment of inertia is proportional to square of diameter.
So it's another potential advantage of 2-phase 14,000 rpm IMs - for use in clutchless multi-speed transmissions, in designs where IM length can be longer.

Honda may need those 12 kW motors for mass hybridization of some small engined cars, used in Japan and in many Asian countries (say sub 1.2L engines).


Alex, clutches and two-speed transmissions probably work for WrightSpeed's business plan. If I remember more than 30 years ago Eaton Corp was working on a similar idea and abandoned it. Their team had problems getting the hydraulic actuators to work consistently.

Much later Tesla tried the same thing, it failed, and Musk couldn't even get Magna onboard to build that type of system. There are easier ways to go.

Tesla's problem was not difficult to solve but it was a case where the design team needed someone with a better grasp of the fundamentals. In essence they needed a motor with a larger magnetic circuit that would be capable of driving to 60 mph in four sec but with the motor turning at only 7000rpm rather than 14000 rpm.

Then for those speeds exceeding 60 mph and approaching 125 mph they needed a Prius style voltage upconverter of around 100Kw. This would allow the controller to preserve the V/F ratio going forward.

Acceleration would continue at constant power all the way to top speed and there would be an inverse torque drop off during this period but not as drastic as there would have been if the V/F had to suffer through the lack of an upconverter. Tesla didn't take this route, however. They chose a friendlier top gear ratio, moved 650 amp transistors up to 850 amp and did something to the motor on which there were no specific details.

I think that progress is made when moving away from mechanical systems. Linamar's twin motor rear drive unit described on GCC this month is a case in point. Despite dispensing with a differential and a hollow shaft motor, the agregate power of two smaller motors could be 26% greater than one purely on considerations of surface/volume ratios alone.

Thanks for the links BTW


I found saved text from the Wrightspeed site, accompanying the video (same video as the one at YT, link from my previous post). The old link no longer works, so I'll copy the text, it may interest you.
BTW I don't know what type of actuator Wrightspeed use. I personally like the Vocis system w/ torque fill, so no ultra-fast actuator is needed, at least for passenger cars. Reportedly actuators in AMTs using very small electric motors are cheaper that hydraulic ones.

The text:

Wrightspeed demonstrates software controlled Clutch-less shifting

Wrightspeed's Powertrains move the complexity from mechanical systems into electronic and software systems, making them lighter, cheaper, and more efficient. Clutchless gear shifting is a good example of this:

Traditional multi-speed transmissions use clutches (synchro rings, multi-disc wet clutches, twin-clutch arrangements) to achieve synchronization before engagement; this makes them, heavy, expensive, and less efficient. But with electric motors, it becomes possible to control the motor speed so precisely, and change it so quickly, that the shifter dog-clutches can be engaged without clashing. The sync function that used to be performed by mechanical means has been shifted into software control of electronics, driving the electric motor with precision. The system is therefore lighter, cheaper, and more efficient. Wrightspeed's control software weighs nothing, costs nothing to manufacture, doesn't wear out, and uses the electronics that are already present to drive the motor.

The Wrightspeed GTD is shown here, on a dynamometer, simulating acceleration from stop to first gear, shifting to second gear, back down to first gear and then a full torque stop. The shifting is too quiet to hear. The sound produced here is the motor accelerating and decelerating at maximum torque. The GTD jumps when the motor changes speed due to torque reaction (torque reaction can be observed under the hood of a conventional car, when accelerated hard in neutral).

// Video:

Motor Sync Time: 80 milliseconds

Simulated Truck Acceleration, Full Torque Stop
1. Motor is accelerated to 20,700 rpm in first gear
Upshift (Shift from 1st gear to 2nd gear)
2. Motor torque is reduced to zero
3. Shift actuator is moved to neutral position (mid-stroke)
4. Motor speed is synchronized to the correct speed for second gear, 9000 rpm (80 ms)
5. Shift actuator moves to second gear
6. Torque is reapplied to maintain "vehicle speed"
Downshift (Shift from 2nd gear to 1st gear)
7. Motor torque is reduced to zero
8. Shift actuator is moved to neutral position (mid-stroke)
9. Motor speed is synchronized to the correct speed for first gear, 20,700 rpm (80 ms)
10. Shift actuator moves to first gear
Full-Torque Traction Drive Stop
11. Motor comes to full stop

The comments to this entry are closed.