Photo credit: Don Williams / UltimateMotorcycling.com
“...until you can put numbers on your problem, you are not yet at the beginnings of a science....” - Prof. Gordon P. Blair
“We've tried nothin' and we're all out of ideas.” - Ned Flanders' beatnik mother on The Simpsons.
Ask anyone who has ridden an electric trials bike – they deliver power differently than an internal combustion engine (ICE) trials bike. This difference again caused me to ask, what makes a good trials motor? I say again because my curiosity was first piqued when I stumbled upon a 1985 SAE paper entiteld Two-Stroke Cycle Engine with Flow Induction Corrected at the Intake and Transfer. The paper was written by Manuel Sanz who was a consultant to Montesa. A separate section will describe the paper in greater detail, but suffice it to say it was important to me for two reasons:
It tried to explain in engineering terms what expert trials riders consider desirable engine characteristics.
It was the only time I had ever seen dyno curves published for anything other than full-throttle operation.
Before Sanz could get to the meat of his topic, he first had to define in engineering terms things that trials experts expressed with their particular vernacular – words like “touch”, “swiftness” and “stretching.” Although, Sanz does admit their lingo is not perfectly consistent.
Note that the paper was written in the era of non-stop trials and at a time when the clutch was used differently (if at all) than it is today. And, unlike today's hydraulic clutches, in 1985 the clutch was operated by a cable.
The author's first language is not English, and some of his writing is difficult for me to follow (but scrutiny helps develop a better understanding). I'll be quoting some of his definitions below and adding my interpretation and commentary.
Touch: “Having the wheel in one's fist.” Today we call this throttle connection. “Those curves corresponding to partial throttle openings must have a maximum located at the lowest velocity possible, and must fall rapidly for higher velocities.” That is, a falling torque curve is desirable. “The greater the 'touch' of the engine is, the more the behavior of the vehicle depends on the accelerator control and the less on terrain features or the inertia of its moving parts.”
Swiftness: “The curve corresponding to full throttle must rise very rapidly from the minimum usable velocity and must maintain high values up to the neighborhood of maximum velocity.”
Stretching: “The whole of the characteristic torque curves must display a high dynamic range of velocity.” Having a wide dynamic range implies the difference in vehicle speed from slowest to fastest usable in a single gear should be large.
“To avoid a sudden loss of adherence causing the driving wheel to end up in a fiasco due to an uncontrolled run-away of the engine, the engine should work in the range of stability of its torque characteristics. That is to say, in that zone of the torque-rpm plot for the parametric curves, a constant position of the accelerator must produce a negative slope..” Translation: When holding the throttle constant, a falling torque curve makes it easier to control the motorcycle.
Sanz views inertia as a double-edged sword, helping to extend the dynamic range into low RPMs (Stretching), but also negatively impacting the Swiftness of the engine.
To get a feeling for the characteristics of successful trials power plants, I compared ones for which I could find dyno curves and gearing data. They represent a nice range of eras and technologies:
1983 Montesa Cota 242 (air-cooled 2T, 242cc, 5-speed, with cable-operated clutch)
1990 Aprilia Climber (Rotax 250cc 2T, 249cc, 6-speed)
2001 Montesa Cota 315R (liquid-cooled 2T, 249cc, 5-speed)
2005 Montesa 4RT (liquid-cooled 4T, 249cc, 5-speed)
2015 EM 5.7 (5kW nominal electric, no clutch)
2020 GasGas TXT 300 (liquid-cooled 2T, 297cc, 6-speed)
2021 EM ePure Race (6 kw nominal electric)
Driving Force! Unfortunately, driving force can't be determined solely based on the motor itself. But assuming the rear wheel has a standard diameter, it's simply a matter of also considering gearing. Motorcycle gearing is always set up to provide a decrease in speed and an increase in torque.
Many readers will not be interested in the gory details, so I have summarized some of my findings below. However, for those who are interested, there will be subpages that provide methods, results, and more.
The EM 5.7 has a lower top speed than claimed. According to the dyno curves, at 48 volts, the Golden Motor HPM48-5000B won't spin faster than about 4500 rpm. I have measured a maximum of only about 4000 rpm. This equates to a top speed of around 22 mph, as geared (9T front / 57T rear). This is a tradeoff to achieve a reasonable driving force with its single-speed transmission.
The EM ePure Race would need to rev to about 8000 rpm to achieve its claimed top speed (43 mph). Dyno curves, again at 48 volts, for the Dana TM4 IPM33-200 top out at about 6000 rpm. The ePure Race uses a nominal 56-volt battery. This would permit about a 16% increase in speed to 7000 rpm (about 37 mph). I have been able to achieve over 39 mph according to the Cycle Analyst in Green mode. Field weakening would probably be required to get the motor to spin faster than 7000 rpm.
The driving force for the ePure is initially considerably greater than that of the 5.7 But they are about the same at a ground speed of 16 mph. For more on this, examine the driving force curves in the spreadsheet.
The GasGas TXT 300 dyno chart leaves me with more questions than answers. The YouTube dyno run shows missing tachometer markings at 3000 and 9000 rpm. It would seem the dyno chart somehow got stretched by 2000 rpm. It also exhibits a very long torque peak, which would be great if true, but is quite dissimilar to the other ICE trials engines in that regard. I'm not sure how/why the dyno chart got corrupted, but I don't think it's accurate. The bike is geared to do 94 mph at its maximum indicated engine speed of about 10,500 rpm. If we assume a maximum engine speed of 8500 rpm, the ground speed would be 76 mph. This is more in keeping with other trials bikes. (The Montesa 4RT is geared for 73 mph. The Montesa 315R is geared for 75 mph. The Aprilia Climber is good for 70 mph.)
The ePure Race's flywheel has a much lower MoI than the ICE flywheels I have calculated. This could explain why Ryan Young could not get the bike to behave in the manner he expected (this was mentioned in my write-up on the clutch. So it's probably not that the clutch's operation is deficient, it's that there was insufficient energy in the flywheel.
The flat torque curve for electric motors is impressive at low RPMs. But this may not be the best thing for trials. As Sanz expressed in SAE 850184, having a falling torque curve at a steady throttle makes it easier to control the bike.
The fact that an electric motor can make maximum torque at zero speed is probably good for trials. Again, according to Sanz, “The curve corresponding to full throttle must rise very rapidly from the minimum usable velocity....” I noticed this is the goal of ICE trials engines when I reverse-engineered my OSSA. The ignition timing at idle is about 20° BTDC and reaches a maximum advance of about 30° very quickly and at any throttle opening. I called this “snap,” or the ability to make power instantly.
Although the torque produced by the electric motors is generally greater than the competing ICE bikes, the electric's taller overall gearing usually results in a lower driving force.
Having a single speed, the electric trials bikes exhibit another desirable quality according to Sanz, that of “stretching” or the ability of a single gear to encompass a large range of speeds.
No consideration has been given to part-throttle operation. This is a huge complicating factor. Although it would be possible to generate curves for the electric motors, there would be little comparative data for the ICE bikes. One factor in favor of the electric motor is that the part-throttle behavior could be more easily tailored to the rider's needs. In its most basic form, we see this in the availability of different power modes.