Aaron;
Usually most of us are looking for increases in horsepower and torque, which usually means opening up breathing. But, I have a different kind of performance question: there are situations (parade duty for example) where the goals are rather different: Low speed smoothnesss and torque and the ability to motor along at low revs and smoothly accelerate when needed are important. What have you got for us - I looked at the dyno curves for hurricane flow and like the evidence, but as there was no torque data especially in the 1000 to 2000 rpm range - curiousity is killing the cat - does Hurricane Flow, or for that matter freeing up the breathing (whether exhaust or intake), risk impairing the ability to plonk along?

Aaron;
Usually most of us are looking for increases in horsepower and torque, which usually means opening up breathing. But, I have a different kind of performance question: there are situations (parade duty for example) where the goals are rather different: Low speed smoothnesss and torque and the ability to motor along at low revs and smoothly accelerate when needed are important. What have you got for us - I looked at the dyno curves for hurricane flow and like the evidence, but as there was no torque data especially in the 1000 to 2000 rpm range - curiousity is killing the cat - does Hurricane Flow, or for that matter freeing up the breathing (whether exhaust or intake), risk impairing the ability to plonk along?

What a great QUESTION....

Looking forward to Aaron's response....

I will go out on a limb and say that for those purposes, stock is better...the less done, the better...

That DOESN'T mean I still wouldn't recommend a Stage 1, but certainly NOT anything more than air cleaner and exhaust, and appropriate tuning...

I would even bet that there is some fine-tuning that could be done on a dyno to get the response you want...

The more specific the need, the easier to find your end-result....

I don't mean to jump into Aaron's thread, but ya might want to look at lowering your gearing via your drive sprocket (if you have a factory 1200).

A 27 tooth sprocket from an 883 would give you better gearing and lugging ability for parade-type duties.

That is a great question, and I appreciate you asking it.

Let's start at the beginning, torque and horsepower, and define these terms.The mistake a lot of people make is to treat them as separate entities; they're not separate at all. Torque is a component of horsepower, and the other component is rpm. Kind of like a margarita is made of tequila and lime juice. There's a fixed mathematical relationship between the three of them. If you know two of them, you can readily calculate the third.

Torque is basically how hard the motor is pushing, conversely rpm is how fast it's pushing. Horsepower then is the combination of how hard you're pushing and how fast you're pushing. Horsepower is literally torque times rpm (divided by 5252 but that's a nit, just a scaling factor to relate the number to a horse)

When a person says they want a lot of torque, nine times out of ten what he's really saying is that he wants power composed of low rpm and high torque. It's entirely possible to build a motor that has terrible torque at low rpm and tremendous torque at high rpm, I've done it and I've got charts in the dyno area to prove it. And technically that meets the request for a lot of torque, no? But that's not what the person really means.

So anyway, I think what you're really saying is how do we make a lot of power at low rpm. Power without rpm has to come from torque. So let's define torque. Torque really boils down to three things:

1) cylinder pressure
2) mechanical advantage of the cylinder pressure on the crank
3) gear reduction between the crank and the rear wheel

Let's work backwards on these, starting at number 3. Every time we apply gear reduction, we increase torque and decrease rpm by equal factors. Since one goes up and the other goes down by the same factor, power remains the same, it's just got a different makeup. So let's say we've got 10ftlbs at 1000rpm at the crank and 10:1 gear reduction through the primary, transmission, and final drive combined. The rear wheel will have 10 * 10 = 100 ft-lbs and 1000/10 = 100 rpm. Power (torque times rpm) is the same at either place, because the gear reduction changed the torque and rpm equal factors.

Now say we gear it 20:1 instead. What happens? Torque at the rear wheel becomes 10 x 20 = 200 ft-lbs. Rpm at the rear wheel becomes 1000/20 = 50 rpm. We're pushing twice as hard as the 10:1 gearing, but only going half as fast. To get back to 100 rear wheel rpm, we need to raise the motor's rpm to 2000. If torque doesn't drop as we raise the rpm, we're back to 100 rear wheel rpm and still have 200 ft-lbs at the rear wheel.

RedRider hit the nail on the head. You want more rear wheel torque, make the gearing deeper. But you'll have to turn the motor more rpm to get back to the same rear wheel rpm. Hmmm.

Okay, now item #2, mechanical advantage of the cylinder pressure on the crank. The most obvious source of this is the stroke. The stroke is basically how far away from the crank centerline the crankpin is located (i.e. where the rod is attached). And the farther away it is, the more leverage it has. But bore also gives more leverage, because cylinder pressure is managed in pounds pres square inch (psi). Make more square inches at the same PSI, and you have more pressure on the connecting rod.

So this is another big opportunity for increasing torque, more stroke and/or bore. Or "displacement" for short.

The biggeee, though, and I'm sure the one you're most interested in, is item number 1: cylinder pressure. How do we get more cylinder pressure? And do it at low rpm?

'scuse me, I gotta go for a bit, more later, stay tuned.

How about short duration cams, Cams with wide lobe separation, or both?

I'm getting there! Sorry for the break. Wife came home. Like to spend the evenings with her.

Please keep in mind that entire books have been written about this subject. Lots of them. In no way shape or form can I do the subject justice here, all I can really do is sort of a reader's digest version.

So back to cylinder pressure. When we talk about cylinder pressure, we're really talking about a pressure curve. The curve can build a lot of pressure early in the cycle, or it can spread it out, with less peak pressure but over a longer period of time. In the ideal world, we generate a curve that matches the mechanical advantage curve of the rod on the crank as defined by the degrees of rotation and rod length. That way we get maximum benefit from the pressure. In the real world, this is not so easy to achieve, and in fact the easy things to do tend to cause a narrower pressure peak and make managing detonation more problematic. Plus we have a varying load. Pushing the piston out of the way has a hell of a lot to do with regulating the pressure in the chamber. A piston that won't move out of the way easily, because the motor is under a lot of load, will see higher peak pressures that are difficult to manage.

Without doing this subject justice, let me just interject and say that the best method on a normally aspirated motor is to optimize cylinder fill, burn efficiently, and time the exhaust open point such that we take full advantage of the pressure.

The last of those is pretty self explanatory. The exhaust valve opens as the piston is on it's way down on the power stroke. The idea is to open it late enough that we take full advantage of the cylinder pressure, but early enough to efficiently scavenge the exhaust (cylinder pressure is useful for giving the exhaust a head start, and also induces a pressure wave in the exhaust that can later reflect back and be helpful for cylinder fill). Of course, the optimum point to open the exhaust valve is going to be different at different rpm's, because as rpm goes higher, there's less time to get the scavenging done, and therefore opening the exhaust earlier is beneficial. But too early just dumps power straight out the exhaust. An early exhaust open point is the reason a big cammed motor has a lumpy idle. So there's one opportunity for improved low speed performance, delay the exhaust open point.

An efficient burn is accomplished by minimizing the separation of the air & fuel, flame front obstructions, and pockets of trapped fuel. Properly sized & shaped chambers that generate good turbulence are a key part of this, and a whole discussion in themselves. And suffice to say that there's lots of controversy on this subject. For example, a "hemi" type chamber trades off chamber turbulence for maximum valve unshrouding, and there are people who believe this trade-off is helpful.

The holy grail, though, and what motor guys work the hardest on, is cylinder fill. And there are a bunch of things that affect it. Some of these things are static in nature, but most are dynamic, i.e. they're a function of how fast things are moving.

More later ... my time is finite ...

Back to cylinder fill. It's useful to think of the intake cycle as three distinct phases:

Phase 1) Early in the cycle is heavily influenced by the exhaust system. The stroke just prior to the intake stroke is the exhaust stroke. The piston is coming up on the exhaust stroke, exhaust is rushing out the exhaust valve, and as the piston approaches top dead center, the exhaust valve is closing. And the intake valve actually starts opening during those closing stages of the exhaust stroke. So we have this window in time when both valves are slightly open, known as "overlap".

The escaping exhaust has a suction effect, this is known as inertia scavenging. It's more pronounced at higher rpm due to the increased velocities (the same amount of exhaust has to move through the same size hole in less time, so an increase in velocity is the only way that can happen). Since both valves are open, the intake manifold and carburetor will see this suction also. It can actually start the intake charge moving before the piston even starts going down, which greatly improves cylinder fill. But in our low rpm scenario, inertia scavenging isn't particularly strong. That can play hell with the following intake stroke because with insufficient exhaust port velocity, it can actually suck exhaust back in (the exhaust port is under more pressure than the intake port so where do you think it's going to suck from?). This is the reason why high overlap and low rpm rarely mix well.

The other thing going on here though is that the pipe has positive and negative pressure waves travelling up and down it. When we opened that exhaust valve awhile back, we created a wave, and if we can get that wave to arrive back at the chamber in negative form during this window of time that both valves are open, it'll suck on the carburetor. Another form of exhaust augmented intake flow.

The problem at low rpm, though, is that it's a relatively long time between when we open the intake valve (thus initiating these pressure waves) and overlap (when we need a negative pressure wave to arrive). The waves travel at a constant speed in the pipe, about the speed of sound, regardless of the rpm of the motor. So to get a wave to show up during overlap at low rpm, we need a *really* long pipe. The alternative is to catch a specific reflection off of a shorter pipe, but the wave weakens with each trip up and down the pipe. So in the real world, where pipe lengths are limited, it's difficult to get significant augmentation from pressure waves at low rpm. Baffling to alter the timing and width of the pressure waves can be useful, but they inherently cause pumping losses. But it's all we've got.

The point here really is that the exhaust system has a profound effect on cylinder fill. It can help tremendously, and likewise it can hurt tremendously, just depends on what type of wave arrives at the chamber (positive or negative) during overlap at the rpm of interest. To reduce the effects of the pipe, reduce the amount of overlap in the cams. The factory has done this for years with the "D" cams, which have only 4 degrees of overlap. They know they have to ship the bike with a poor exhaust system, so they cut out the overlap to reduce it's effects.

Phase 2) Through the meat of the intake cycle, the piston pulls on the intake port. It pulls in a non-linear fashion, though, as piston movement is governed by the rod and crank geometry. It actually takes a pretty fast ramp (depending on the lift) to get to full flow by the time the piston is pulling it's hardest. So fast ramp cams tend to be beneficial to us because they help maintain a constant velocity through the intake stroke.

Speaking of velocity, it has a big effect on cylinder fill. Until you reach mach .55 (690fps), cylinder fill goes up as the square of the velocity. So for example, if you increase velocity 10%, cylinder fill goes up 33% (1.1 cubed = 1.33). But once you get past 690fps, it actually starts taking more energy to accelerate the charge from a dead stop (at the beginning of the intake cycle) than you get from the additional cylinder fill. This effect is known as "inertia block".

And of course, the velocity varies with rpm. So if we size the intake system for optimum velocity at one rpm, and then change the rpm, it's not sized properly anymore.

There's a whole ton of science to this thing that I'm not going into, with respect to choke points, limiting port velocities, valve l/d ratios and discharge coefficients, etc. Suffice to say that it can all be calculated. But my real point is that for a real low rpm performer, you'd want to optimize port and valve sizes around it.

Another thing that should be mentioned here is flow losses in the ports. Sharp corners, ridges, and poor shapes cause turbulence and resultant flow losses. To the extent the shape of the port can be improved and the turbulence inducing problems corrected within the constraints of the proper port size for the rpm of interest, cylinder fill will improve. That's true at any rpm you target.

Phase 3) After the piston rounds the corner and starts back up, the intake valve is still hanging open. We don't actually close that valve until the piston is well on it's way up. Why? Because the intake charge has inertia, and the cylinder is still filling despite the fact that the space is getting smaller. Close the valve too early and we terminate that filling process too soon. Likewise, if we close the door too late, the rising piston may push some of the intake charge back out the intake valve, which also hurts cylinder fill. So our intake close timing is critical, and once again, it's tied closely to the velocity of the intake charge which varies greatly with the rpm.

The intake close timing also marks the point where we start compressing the intake charge. Close it later, and compression will drop, necessitating a higher compression ratio to build a given level of compression. Close it earlier and and compression will rise, necessitating a lower compression ratio to build a given level of compression.

So anyway, at the low intake charge velocities associated with low rpm running, earlier intake close points coupled with correspondingly lower compression ratios are required. The ideal point, though, needs to be calculated with the intake charge velocity in mind.

So the short answer here boils down to:

1) deeper gearing
2) more displacement
3) efficient ports (i.e. minimal losses)
4) smaller ports and valves
5) cam timing with a late exhaust open event, modest overlap, and an early intake close event.
6) An exhaust system that generates a negative pressure wave that arrives during overlap at low rpm.

I didn't go into intake system pressure waves at all but it's not as big as all these things.

Aaron
Thanks for the tutorial. My translation of your remarks is that intake wave effect is comparatively less important than the 6 factors listed in your summary. Thus, the Hurricane Flow would appear unlikely to impair parade duty performance. And by extension - things like intake manifold resonators would be unlikely to figure significantly into low rpm running. Of the list in your last post - it would appear the most cost effective are likely to be pulley (deeper gearing) or exhaust which generates a negative pressure wave that arrives during overlap at low rpm. Is there any dyno data out there on exhausts or cams in the 1000 to 2000 rpm range with the stock 04-05 heads and cams? All the dyno stuff I can find for sportsters seems to start no lower the 2000 or 2500 rpm.

Another excellent question (your conclusions are correct BTW). Dyno pulls below 2000rpm are pretty hard to do, even getting them to start at 2000 is difficult, and getting enough repeatability down there to make the results meaningful is next to impossible. Imagine riding your bike in 4th or 5th gear and then slowing down to 1000rpm and then trying to give it full throttle, pretty much the same thing. Even starting at 2000 rpm I often have to roll the throttle on gradually (depending on the motor) and of course that's not a real repeatable process. Makes comparisons or drawing any conclusions real difficult.

It might be possible in first gear, I've never tried it to be honest. I'll give it a shot next time I get the chance. But a pull like that will be over in a hurry and it's not a good representative picture of the performance through the main portion of the rpm range.

Aaron;
Thanks for the really informative answers! There is no dyno nearby for motorcycles and the only one in reasonable distance is a large inertia dynojet for cars. Your explanation about the difficulty with reproducible pulls from low rpm in high gears makes sense particularly having seen the size of the steel roller the car has to spin on the dynojet. But, a 1000 rpm roll on in first gear is the kind of stuff that happens during parade duty, so when you try a first gear pull on the dyno tell us what you learn in the 1000 to 2500 rpm range. (I don't think a first gear pull needs to go to redline probably just to 2500 or at max 3000rpm - because, as you point out, the 2000 or 2500 rpm to redline range gets covered by the more conventional 4th gear dyno run.) I suppose a variable load dyno like the Dynapack brand that are available for cars would be another way to get some data.