Comparing bananas and lemons - Part 1 : Power

When it comes to motorbikes, there is a lot of partisan attitude, some fanboyism and cliquishness…

Right here we have the Bullet chaps (moi included), the Yezdi/Jawa fans (me included), the RD350 lovers, and the RX boys – Inevitably, there is debate about the relative merits of each and much bragging about top speeds and ¼ mile times.

But, the problem is: How would you go about comparing say a 100cc Hero Honda Splendor with a 1300cc Suzuki Hayabusa? It’s all very well to say that they’re made with different aims and can’t be compared, but actually they can – even though the only common thing about the two is the fact that they’re Japanese designs and 4 strokers.

Looking at the fundamental aspect of a motorbike – the engine – we can define certain parameters that let us place a bike as a point in some multidimensional space and see how it compares to others along different dimensions.

Power

The scientific definition of power is “Work performed per unit time” – no doubt this is the absolute measure of a bike, but then things are also relative – Which is more impressive? An elephant lifting a quarter of its weight, or a slim weightlifter hoisting four times his own weight overhead?

Thus, to measure different classes of bikes we need to normalize the power values.

An engine produces a certain amount of power at a certain RPM with a certain displacement – Ideally, (and this is true to a large extent as you will see later) we should assume that if you double an engines displacement, and maintain the same RPM, you should expect twice the power. The same applies to doubling RPM while keeping the displacement same.
The reason is simply because an engine is essentially an air pump, and the power it produces is purely dependent on how much air flows through it.
 

Thus, the correct way to measure relative power output is to divide the actual value by RPM and displacement.

For example: If a 100cc engine makes 7.5 HP @ 6000 RPM, we should expect a 200cc engine of similar design to develop 15 HP at the same RPM, since it displaces twice the volume as the former in a given interval of time.

Thus one of the first variables we can derive (in no meaningful SI units) is HP/Liter-KRevs - Namely, how many peak horsepower are developed per liter of engine per 1000 RPM – We will call this the K - factor

Let’s look at some famous (mostly from the Indian market) 4 stroke bikes past and present and see where they fare… Note that this data is gleaned from manufactures specs and some of it from reading wheel dyno charts, and adjusting upwards by around 15% to get crankshaft figures.

Japanese bikes

Model`	CC	RPM	HP	K
Hero Honda CD100	100	7500	7.5	10.00
Bajaj Pulsar DTSI 220	220	8500	21	11.23
Honda Karizma ZMR	223	7000	17.6	11.27
Yamaha R 15	150	8500	17	13.33
Honda CBR 250	250	8500	26.4	12.42
Honda CB750	750	8000	67	11.17
Suzuki Hayabusa	1300	12000	190	12.18
HD Sportster 883	883	6000	50	9.45

What can we read from this?

The ubiquitous Hero Honda, seen as the epitome of efficiency in India, is actually close to the bottom of the list – The CBR 250 does 24% better. If you metaphorically grafted 2.5 CD100 engines together and revved it up to 8500 RPM, you would get 7.5 * 2.5 * (8500/7500) = 21.5 HP – The CBR makes 5 HP more! So either more fuel is being burnt, or fuel is being burnt better.

In fact, the Hayabusa is quite astounding! If you grafted 5.2 CBR250 engines together and revved it to 12000 RPM, you would get 26.4 * 5.2 * (12000/8500) = 193.8 HP – Just about 2% more than what the Hayabusa makes on paper – Getting 1300cc to rev at 12000 RPM is no joke, it’s hard rocket science to be able to scale up an engine so well! The Hayabusa is an engineering marvel.

The R15 seems quite amazing, with only 150 cc, it makes almost as much as the 220 Pulsar – Seems like the most advanced engine amongst its peers.

The Harley is nowhere near the rest, even though it is a highly engineered design, but then again it’s a pushrod engine with an extremely under-square design, built for an almost flat torque-curve.

Let’s look at the two popular mid level bikes Pulsar and Karizma – Isn’t it startling that despite all the state of the art technology, fuel injection and what not, a 1970s CB750 of much larger capacity comes within 1% of the figures in terms of our metric?

I tell you sir! They don’t make ‘em like they used to! They just hype ‘em louder.

Of course, we aren’t looking at fuel efficiency (yet), so perhaps this conclusion is flawed; perhaps the bikes that have higher K are burning much more fuel. But nevertheless, if the question is: “How much power can you get out of a liter of engine regardless of fuel consumption?” then the above table shows the answer.

Royal Enfield’s

Model	CC	RPM	HP	K
Diesel bullet	435	3600	7.5	4.79
Std 500	500	4750	22	9.26
Std 350	350	5000	18	10.29
Lightning 535	535	4750	26	10.23
Enfield CL500	500	4750	27.2	11.45
ACE Fireball bullet	535	5750	42	13.65
Hitchcocks 612 stroker	612	5750	45	12.79
Enfield Constellation	700	6250	51	11.66
Enfield Interceptor	750	6750	56	11.06

This had to be in a separate table!

Not much to say here, except that Tom Lyons’ (aka ACE) Fireball kit can take a Bullet well beyond Jap bike territory: The dyno charts for the Fireball charts show about 36 wheel HP, so 42 engine HP is a reasonable estimate.

The Hitchcocks stroker 612 does great too, despite having an even longer stroke, and the CL500 is much better than the old Enfield’s, even considering it has to contend with Euro 4 emissions norms.

Quite surprising that the Connie and the Interceptor of yore do so well for air-cooled long stroke parallel twins.

These numbers are derived from observing several wheel dyno charts, rather than manufacturers specs (which are quite off), 6 HP added to get crankshaft horsepower estimate.

Legendary Classic Brit bikes

Model	CC	RPM	HP	K
BSA Goldstar	500	7000	42	12.00
Velocette Venom	500	6500	41	12.62
Triumph 650 twin	650	6500	46	10.89
Norton 650SS	650	6800	49	11.09
Vincent 1000	998	5500	55	10.02

Just astounding! These were mostly OHV pushrod bikes (except the Norton which was an OHC), ranging from the 40s to the 60s.

The Velocette even beats the Hayabusa in this metric! No doubt it ran on “real gas” but even so – this is a 50s bike with 8.75:1 compression! It still holds the 500 cc 24 hour endurance record at an average of 100 mph+, that too with a stock engine. All the employees of Velocette rode these bikes themselves and the build quality was so good that a stock bike was race ready. A few years after this record, they attempted it again with a souped up 350cc bike, which was going great for hours, unfortunately after 6 hours averaging > 105 mph, the piston blew ( this was not a Velocette manufactured one, but sourced from elsewhere) and the record attempt was thwarted.


The ACE fireball is on par with the Goldie on paper, and that speaks a lot about its potential performance!

I ask you!

Two strokes and four strokes

A four stroke has one power stroke for every two revolutions of the crank, whereas a two stroke has one. That means that a two stoke will (ideally) pump twice as much air as a four stroke of the same displacement at a given RPM.

Thus if we want to measure using the above metric then we need to include a factor of two – This is unfair because usually a 2 stroke of a given displacement makes only about 30 to 50% more than a 4 stroke cousin of the same volume – For e.g. take the Yamaha RX100: It makes 11.5 HP vs. 7.5 for a CD100 – that’s 53% more power, which is impressive, but still not double, even though it displaces twice as much air per unit time!

But fair or not, that’s the metric so let’s look at 2 strokes of all kinds:

Legendary two-strokes from India

Model	CC	RPM	HP	K
Jawa 250	250	4750	12	5.05
Yezdi Roadking	250	4750	16	6.74
Jawa 350	350	5000	23	6.57
Yamaha RX100	100	7500	11.5	7.67
Yamaha RX135	135	7500	14	6.91
Suzuki Shogun	108	8500	13.8	7.52
Yamaha RD350	350	6750	30.5	6.46

The RX 100 is the absolute champ here! Squeezes out every ounce of power from that 100cc, the Shogun does quite similarly. The Roadking has its reputation as a rally and motocross bike and for a 70s design, it is quite impressive.

None come anywhere close to the four stroke bikes, so it’s obvious that if you ride a two stroke, you are throwing a lot of fuel down the exhaust chute…. But then again a two stroke has an exhilarating, almost linearly increasing torque curve all the way up to the top, and who can not be enchanted by the coarse whine of the RX, the scream of the RD350 or the burbling purr of a Jawa?

Really amazing bikes

Model	CC	RPM	HP	K
Maico 760	760	4000	43	7.07
Maico 760 @ peak torque	760	1200	26	14.25
Suzuki RS67	124	16500	41	10.02
Honda Rc 165	250	10500	60	11.43
NSR500	500	19000	240	12.63

Yes, all those numbers are correct! The Maico 760 was a two stroke single, the largest ever made, and perhaps the most “gruntiest” of any bike that I have ever heard of… It makes even a thumper like a Bullet look like a revvy whiner. Stomach churning torque, I say! Look at the K factor at peak torque! That is true “bang” for the buck! Only 6 were ever made…

The other Japanese race bikes are again crazy - the RS67 had 4 cylinders, 12 gears and hit 137 mph – unbelievable!

The NSR 500 is probably the most powerful two-stroke made anytime, anywhere and needs just about 2cc to make a HP!

One thing that’s quite clear to me from real life experience is that manufacturers’ figures are way off reality; we all know how optimistic our speedos are… Dyno results are the most believable, and even they have some contradictions.

Read on to the next part where we talk about power to weight ratios…

Comparing bananas and lemons - Part 2 : Torque

What happens in the hills?

While we have been harping on power all along, some of us are accustomed to grunt and scoff at high revs – which brings us to another metric – torque – To put it very simply, think of your engine connected to a pulley of one meter diameter : The amount of weight it can hoist is the torque.

Torque is the value which can tell you what maximum grade of hill a motorbike can climb – The road is essentially a rope and the rear wheel like a pulley that your bike pulls itself up on.

Most engines have the following power/torque characteristics –

• High revving four strokes – make negligible torque at the lower 1/3^rd of their RPM range, and the difference between peak power and peak torque RPM is less than 25%
• Low revving four strokes – make decent torque starting from the 1/4^th of the RPM range and the difference in peak power and torque RPMs can be up to 50%, at the expense of peak power values
• Two strokes – Rarely make good torque at low end,  and peak power and peak torque RPMs are very close – The Maico 760 is an exception here

Torque defines instantaneous acceleration based on the current gear ratio: Climbing a hill at constant speed is equivalent to accelerating steadily on a plain.

With an engine that has a narrow power band, the gears have to be shifted much more often to ensure that the engine keeps within that range. For a given gearing, a wide power band engine has an overlap of speeds across gears, allowing a climb to happen with less shifting and use of the clutch. There will come a point for a peaky engine, where close to maximum revs in a certain gear are required, but the resulting vehicle speed is too much to handle on the twisties. The low revving bikes have the option of modulating speed with the throttle in the same gear since torque does not fall off too rapidly below peak torque RPM.

Think of the shape of the torque curve, if you are riding at an RPM after the peak, the slope of the curve is downwards, that means a reduction in engine RPM makes the engine push harder – it is a negative feedback effect - and there is an equilibrium found as the bike climbs steeper or flatter slopes. If you drop below peak torque, there is now positive slope in the torque curve and speed falls off rapidly, forcing either a downshift, or wider throttle.

The ratio between peak power RPM and peak torque RPM gives a range of “cruising” speeds for each gear that are in the same proportion – My own thumper for example has 5250/3000 ratio – In practice the peak RPM is closer to 5000, so that’s a 5:3 ratio meaning in 4^th gear, I can range from 75 to 125 KPH with full acceleration potential. If I were on an extremely steep hill where 1^st gear was the only possible option, I could go from around 30 to 50 without needing to up-shift. This is what the definition of drivability is – the engine can pull away cleanly at a range of RPMs

Not to take anything away from the way two-strokes and high revvers deliver torque – They generally have this effect of ever increasing acceleration until peak, and that’s one of the highs of riding such bikes – on the plains!

What makes an engine produce power? Essentially airflow – pump enough air through the engine and you are done – the carb or fuel injector will ensure (mostly) that there is enough fuel in that air, and when all of it combusts you have glorious horse power.

The trouble is that airflow cannot easily optimized across all the RPM ranges – it’s almost as though you can change the power curve, but must preserve the area under it. So you can make it peaky at one end, or flatten it out and reduce the peak.

I do not have enough data, otherwise we could have looked at the integral of the power curve as a real metric of overall performance – An electric motor or steam engine comes to mind as the ideal maximum: Full torque at all RPMs for a motor, and torque descending from peak value to zero linearly across the rev range for a steam engine. That means for both of these, the peak power/peak torque rev range is the entire rev range, so gearing is entirely unnecessary, except for final drive!

Once again torque should be normalized according to weight, so we have a torque to weight ratio, and again one normalized based on RPM such that making peak torque at lower RPMs gives a higher metric.

Let’s look at the figures:

I have no peak torque data for many of the classic bikes, only the current bikes are shown here

Japanese Bikes

Model	PTQ RPM	KGM	KGM/ton	(KGM/ton)/KRPM
Hero Honda CD100	6000	0.75	4.05	0.68
Pulsar DTSI 220	7000	1.9	8.52	1.22
Karizma ZMR	6000	1.83	7.82	1.30
CBR 250	7000	2.2	9.28	1.33
Suzuki Hayabusa	7000	13.5	42.59	6.08

Enfield’s

Model	PTQ RPM	KGM	KGM/ton	(KGM/Ton)/KRPM
Diesel bullet	2500	1.5	5.77	2.31
Std 500	3000	3.8	14.62	4.87
Std 350	2800	3.2	12.55	4.48
Lightning 535	3000	3.8	14.62	4.87
Enfield CL500	4000	4.1	15.77	3.94
ACE Fireball bullet	3000	5.7	21.92	7.31

Two Strokes (again RPM is factored by 2)

Model	PTQ RPM	KGM	KGM/ton	(KgM/Ton)/2KR
Jawa 250	4500	1.8	8.57	0.95
Yezdi Roadking	4500	2	9.52	1.06
Jawa 350	5000	3.1	12.76	1.28
Yamaha RX100	6500	0.9	5.06	0.39
Yamaha RX135	6500	1	5.62	0.43
Suzuki Shogun	7500	1	5.59	0.37
Yamaha RD350	6500	3.3	15.14	1.16
Maico 760	3600	8.2	43.16	5.99

Look at the figures! Is it a wonder that thumper fans scoff at crotch rockets?  Even the ‘busa, though having great torque, is outdone by the Fireball here.

The Fireball is off the charts, it could do great things off the road, probably has enough torque to hoist itself vertically on a rope at a good speed!

The new age 4 strokers are just not made for relaxed riding – They barely tract as well as the antiquated Jawas, while revving much higher – makes for a whiny washing machine like sound.

The CL500 needs to be ridden in a different style from the older bullets, same goes for the other twin spark and AVL engine ones which also have 4000 RPM peak torque, which is why probably many feel these are not “real” bullets.

The RD 350 is pretty decent, reaching almost to the level of the modern 4 strokers, although nowhere near the Enfield’s.

The Maico looks mighty impressive for a two-stroker, Pity only six of them were ever made.

Next let’s look at the RPM spread metric: 

We would like the ideal spread to be 100% like a choo-choo train – So let’s say the percentage of the RPM range between peak power and torque. We’d also like the difference in actual torque at those two points to be as little as possible, so the ratio between the torque values at peak power and peak torque is another thing we can multiply with the above.

Let’s look at RPM ratios ( The column titled RR ) - A steam engine would have a value of 1.0
PP - Peak power
PTQ  - Peak torque

Model	CC	PP RPM	PTQ RPM	RR	HP	KGM	PP KGM
Hero Honda CD100	100	7500	6000	0.20	7.5	0.75	0.70
Pulsar DTSI 220	220	8500	7000	0.18	21	1.9	1.72
Karizma ZMR	223	7000	6000	0.14	17.6	1.83	1.75
CBR 250	250	8500	7000	0.18	26.4	2.2	2.16
Honda CB750	750	8000	7000	0.13	67	6	5.83
Suzuki Hayabusa	1300	12000	7000	0.42	190	13.5	11.03

Model	CC	PP RPM	PTQ RPM	RR	HP	KGM	PP KGM
Diesel bullet	435	3600	2500	0.31	7.5	1.5	1.45
Std 500	500	5250	3000	0.43	22	3.8	2.92
Std 350	350	5650	2800	0.50	18	3.2	2.22
Lightning 535	535	5250	3000	0.43	26	3.8	3.45
Enfield CL500	500	5250	4000	0.24	27.2	4.1	3.61
ACE Fireball bullet	535	6000	3000	0.50	37	5.7	4.30

Model	CC	PP RPM	PTQ RPM	RR	HP	KGM	PP KGM
Jawa 250	250	4750	4500	0.05	12	1.8	1.76
Yezdi Roadking	250	4750	4500	0.05	16	2	2.35
Jawa 350	350	5250	5000	0.05	23	3.1	3.05
Yamaha RX100	100	7500	6500	0.13	11.5	0.9	1.07
Yamaha RX135	135	7500	6500	0.13	14	1	1.30
Suzuki Shogun	108	8500	7500	0.12	13.8	1	1.13
Yamaha RD350	350	6750	6500	0.04	30.5	3.3	3.15
Maico 760	760	4000	3600	0.10	43	8.2	7.49

The numbers speak for themselves – no doubt all these engines have usable torque for well below peak, but the fact remains that below that RPM the torque curve slopes upwards, so the ability to cruise is absent in that lower rev range.

The Hayabusa impresses again, with almost 50% of the rev range spread out between PTQ and PP!

The Enfield’s are all superlative when compared to the commuter bikes, and the Std 350 is still the king in terms of tractability! The fireball does as good on account of spreading the rev range upwards to 6000 revs

The CL500 (and all AVL and UCE engined bullets), have opted for a more rev happy engine, sacrificing that thumpy nature, by tuning it towards a later peak torque and better higher end breathing – most likely with higher valve overlap.

Here’s a simple test to determine true peak torque RPM on a bike : On a straight road, accelerate well past the estimated PTQ RPM, and then slowly back off the throttle -  the RPM keeps dropping until there’s a point where the speed of the bike stabilizes. If you go below the actual PTQ RPM, the bike will slow rapidly and ask for more throttle to maintain speed. So essentially find the slowest “cruising” RPM – you can feel the top of the torque curve.

It’s not quite that you cannot cruise at any given speed on any given bike, it’s just that if you are on the downward sloping part of the torque curve, the engine governs itself due to a negative feedback effect – if it encounters more load, the revs drop and there is more torque, and vice versa. This happens without any throttle change. Having as large a range as possible where this happens makes the ride “relaxed”.

Let’s now incorporate the ratios between the two torque values and see what we get - we get the ratio between the torque values at peak torque and peak power and multiply it by the previous RPM ratio value.

Model	Torque metric
Hero Honda CD100	0.19
Pulsar DTSI 220	0.16
Karizma ZMR	0.14
CBR 250	0.17
Honda CB750	0.12
Suzuki Hayabusa	0.34

Model	Torque metric
Diesel bullet	0.30
Std 500	0.33
Std 350	0.35
Lightning 535	0.39
Enfield CL500	0.21
ACE Fireball bullet	0.38

Model	Torque metric
Jawa 250	0.05
Yezdi Roadking	0.06
Jawa 350	0.05
Yamaha RX100	0.16
Yamaha RX135	0.17
Suzuki Shogun	0.13
Yamaha RD350	0.04
Maico 760	0.09

The metric here is

(KGM@PP / KGM@PTQ) * ((PP – PTQ) / PP)

Meaning the RPM spread times the ratio between the torques – sort of describes how flat the curve is and how wide the “plateau” is.

All said and done, the Hayabusa is more or less like an Enfield in this metric! Eminently rideable.

The UCE bullet will do fine as long as you ride at revs, and not in “Uncle” mode.

I’m a bit surprised that the L535 tops this metric – I’m sure both the CL 500 and the Fireball have a torque curve well above it in most areas, but it’s all about the plateau shape…

Among the two strokes, the RX’s are pretty awesome, figuring higher than the DTSI and the ZMR – Who said two strokes are always twitchy and peaky?

The whole point is unless you are racing and have umpteen gears, and are willing to shift every time your speed increases by 10% to 15%, having extremely peaky power curves does not make for easy riding.

I’ve seen with my own eyes a few days back, ZMRs, DTSIs, CBRs spinning wheels and stopping on an off road track. They have to be revved high in 1^st to get the torque to climb, but that needs 80% or more or so of the 1^st gear top speed. That is surely not a slow enough wheel speed to start off up a slippery slope smoothly, so the wheels spin hopelessly and much use of the clutch is necessary.

I shudder to think how a two stroke would have fared!

My own beast had enough low end to chug up along that slope at 1200 to 2000 RPM, barely above idle in 1^st gear, at almost fully closed throttle.

Now let's move on to everyone's favorite yarn - speed