With the MotoGP teams switching back and forth between "big bang" and "screamer" firing orders, there is a lot of discussion about the relative advantages and disadvantages of the two engine configurations. On the one hand, Ducati has elected to switch back to the screamer configuration, with a great deal of success, while Kawasaki is working hard on its own version of the screamer, their task made more complicated by the fact that the Green Machine is an inline four. In the other camp, Yamaha are utterly convinced of the merits of the big bang engine, and have declared that they will not use a screamer again.
And there are those who believe there are even more options that this. An Australian physicist and motorcycle racer called Sean McConnell sent us a piece exploring some of the rationale behind the big bang, and offering a suggestion on how to get the benefits of both firing orders. We hope you enjoy it.
The Big Bang vs. Screamer Debate in MotoGP
by Sean McConnell
I haven't really had too much to say of late of my racing activity, except that I'm selling my bike and plan to buy something bigger, but this issue of screamer style engine vs. big bang in MotoGP is something that has piqued my interest of late, see this article on Crash.net for a bit of a background, that site contains many such articles on the topic that can be found with a simple search. Here's my two cents on the issue.
For the uninitiated a screamer style engine is one that:
"Has an equal amount of time between the firing order of each cylinder during the rotation of the crankshaft"
For a four cylinder engine, a screamer configuration would see one cylinder firing every 180 degrees of rotation on the crank.
A big bang styled engine is one that:
"has a very short interval of time between the firing order of each cylinder, and an extended break until each cylinder fires again."
It takes 720 degrees of crankshaft rotation for a four stroke engine to complete the well known cycle of intake, compression, ignition and exhaust. In a four cylinder, MotoGP big bang engine, all the cylinders fire within a very small angular range of each other, my guess would be within 90 degrees.
The guiding philosophy being that in a big bang engine, the "relaxation time" is long enough between the time when the engine is applying force to the road (the first 90 or so degrees of crankshaft rotation, or ignition) and the time when it's going through the non-force applying functions of a four stroke engine (exhaust intake and compression), that the remaining 630 degrees allow the tyre to re-grip, such that if too much force is applied during that 90 degrees of ignition and the tyre begins to slide, there is still 630 degrees of non force application. This provides a buffer for the rider to not be sent over the handlebars.
Changing up through the gears means that force is applied to the wheel with decreasing angular frequency. For example, if there is a one to one relationship (we'll call this first gear) between the application of force to the wheel and the engine, then in the big bang style engine cited above, force will be applied to the wheel through the angles 0 to 45 degrees, then no force between 45 to 180 degrees, then force from 180 to 225 degrees then no force from 225 to 360 degrees, thus completing one full rotation of the wheel. As the gearing increases, the angular frequency decreases, if there is a 2 to one relationship (we'll call this 2nd gear) then force is applied between 0 and 22.5 degrees, no force from 22.5 to 90 degrees and so on.
So we have seen how the angular frequency decreases with gearing, shortening the time the tyre has to re-grip the road if too much force is applied during ignition. As wheel speed increases, the time it takes to complete one full rotation naturally decreases. So there are two things that shorten the time a wheel has to re-grip the road if too much force is applied during ignition, wheel speed and gearing. The higher the speed and the higher the gear, the less "buffer" a rider has during a slide.
Until now we have only focussed on a big bang engine, for a screamer engine, the situation is certainly far worse. A big bang engine can provide far more "buffer" as speed and gear increases than can a screamer. The screamer engine effectively is applying force constantly to the wheel, there is never any (useable) relaxation time. When Honda switched the NSR500 to a screamer in 1997, riders we're throwing themselves off all the time, only the skill of the rider could compensate for such a nasty peice of machinery, one of the reasons Mick Doohan won an incredible 12 of 15 races that year.
But there is a way for a screamer to compete with the big bang in this battle to squeeze in some sort of relaxation time for the tyre, should there be a sudden loss of grip. The solution comes from, as indeed the entire development of motorcycles, the bicycle. In time trials, it is not uncommon for cyclists to employ an elliptical front sprocket (a bicycle mechanic would call it a chainring). The reason being that an elliptical sprocket can smooth out the application of force to the tyre by changing the torque profile during the rotation of the crankshaft. On a bicycle, the main application of force is to push down on the pedal, the rest of the rotation of the crankshaft, you are simply waiting until your leg is in position to re-apply that downward force. The elliptical front sprocket drastically shortens this waiting time between the application of force. The torque profile changes as your leg reaches the bottom of the stroke, the gearing decreases at the bottom of the stroke, pedaling becomes easier, and your leg moves quicker through this period of non-force application, significantly extending the time that force is applied to the tyre.
For a MotoGP machine, the philosophy is the same, but applied in reverse. Where an elliptical sprocket is used to smooth out the force application on a bicycle, it can be used to roughen it up on a motorcycle. What I'm suggesting is that an elliptical rear sprocket would change the torque profile as the wheel rotates, this is exactly what a big bang engine does, except now a screamer can play the same game. At zero degrees of wheel rotation, with an elliptical gear, whose semi major axis is vertical, the engine would "see" a large sprocket, and consequently the wheel would be easier to turn. At 90 degrees of wheel rotation, the semi minor axis would be vertical and the engine would "see" a small sprocket and the wheel would be harder to turn. What this means is that when the semi minor axis is at the vertical, this is the point at which it is most difficult for the engine to continue to over-apply force, and cause the wheel to continue sliding, allowing the rider a buffer that he doesn't get with a circular sprocket.
The added advantage of an elliptical gear is that it also eliminates the problem talked about earlier of decreasing the time between the application of force with increasing gear. Because it is the final drive sprocket, the application of force is greatest at 0 and 180 degrees (semi major axis at the vertical) and least at 90 and 270 degrees (semi minor axis at vertical) no matter what gear the bike is in. So in this case, the only thing acting to shorten the time between the application of force is naturally the increasing wheel speed, as described above.
Finally, there is the other argument for using the big bang engine, and it's an argument I can totally agree with. You can read about it in an article by Julian Ryder over on Superbikeplanet.com. It is the idea that a big bang configuration is able to somehow improve the quality of the "signal" that a rider receives from the rear tyre, so that they can better understand what the rear tyre is saying to them. Furusawa talks about a connection between throttle and tyre in terms of harmonics, and for me the explanation is simple. The philosophy is that with a big bang, and as I speculate an elliptical sprocket, you can apply some sort of input, such as the application of throttle to a tyre, and be given in return a useable amount of time to determine what that input has done to affect the motorcycle. It's like a feedback system, but with a screamer engine, the input is constant and permanent, there is no time to "hear" anything other than the input signal, you never get the chance to "hear" the feedback signal.
One potential problem for an elliptical rear sprocket is that it would cause the chain to flop around a bit more than usual, although I'm sure a simple mechanism could be devised to take the slack out of the chain, similar to a bicycle derailleur. In saying that however, the eccentricity of the sprocket would not need to be terribly much to achieve the desired effect, and a "derailleur" of sorts mightn't be necessary, and besides that, if you've ever been to a race and watched a chain driven motorcycle, you'd know how much the chain flops about anyway. It is not clear to me why none of the teams that have ever used a screamer style engine did not or have not tried this, as from an engineering perspective, it is the simplest solution to the problem. And according to Occam's razor, the simplest solution is usually the best.