Cyclone Butterfly Activation Point

I'm wondering what RPMs everyone is activing the butterflies in their Cyclone manifolds at.

I've seen an airflow chart that shows the long runners start to plateau off and the short runners passing them by at ~3800 RPMs. I've also seen a dyno chart (16G equipped car) overlaying a run done with just the long runners open and a run done with the long and short runners open. This suggested an ideal changeover point of ~4300 RPMs. I remember reading in the archives where Jeff mentioned the JDM ECUs opened the butterflies at ~4600 RPMs. I'm assuming this was done for a little better mileage and torque as overall horsepower wasn't the ultimate goal.

Just looking for data points right now as I'll probably have to tweak and experiment with the changeover point to fit my particular setup.

BTW, this is hooking up the manifold the "correct" way with an adjustable RPM switch, Hobbs switch and FPR solenoid.

 

[JNR]

I believe most people are opening them ~4,100rpm when using Jeff's (Keydiver) chip and it seems to be a good point...I haven't hooked mine up yet, but that seemed to be the concensus...

Otherwise, people were opening them with the Hobbs switch, which was boost dependent; I cannot remember though what people were opening them up at?

 

I was thinking of starting at ~4200 RPMs and then tweaking from there. Just wondering what everyone else's experiences are.

 

[powerplay]

I am using Jeff's chip and it is at 4100 rpms.

 

Can anybody think of a good reason that these should not be triggered by an airflow variable rather than an RPM point?

I think the best reason is probably that it keeps them from having to flutter all the time. Other than that, It seams that airflow is the key. RPM alone makes the assumption that all you are really interested in is WOT.

 

that's because you are.

vacuum sucks the valve to the small runner position, so as soon as you lose vacuum the big runners would open, you would probably be right if you needed a certain amount of boost to open the bigger ones, but that's not the case and you will render the small runners useless when hooked up straight to the mani since vacumm disapears damn quick when you floor it.

airflow has no bearing on the valve.

 

[Polish]

I think Lucian is talking about putting a code into the ECU to open the runners at a given airflow as opposed to RPM only.

 

Different airflow can be reached at different rpm points and driving conditions with different turbo's. it might open to soon on big turbo car, too late on a stock turbo car, and anywhere in between, unless you tailor it to each specific car i think rpm is the way to go since you only benefit from them in daily driving anyway and 1st gear on a drag run (you'll be above 4100rpm after takeoff anyway).

 

[dime2nr]

would it be possible to use something like apexi vafc and using it to control the point beyond what is the chip set point

 

Alright, I see where you guys are coming from. I guess I'm talking, in a perfect world, there is a mass flow rate through those runners where choke begins to occur, who cares what RPM and what throttle %, if there is enough flow, the car could benefit from the extra runner area. The opposite is also true. Above 4100 RPM there are operating points that don't need the runners. But the more I think about these runners trying to flap open and closed with the throttle plate, I am agreeing that the best control strategy is light switch on at ~4100 RPM... Even that control strategy could probably be mellowed out further with some hysteresis (god I hope that is the right term). The fuel trims use it and most thermo-switch cooling fans have it. The fan might come on at 180* but it doesn't shut back off until 175*. That is 5* hysteresis.


Second thoughts:
If the turbo waste gate can function being pneumatically controlled, I think the cyclone could keep up with the throttle too. Imagine this scenario, racing, banging gears... Road/drag, transmission ratios are more than close enough to keep shift points above 4100 trigger point. The turbo is loosing shaft speed between shifts and must recover. Wouldn't there be benefit to spooling the car on the small runners? Small runners until turbo recovers. I don't think it would help the turbo so much as reinforce the torque curve until the turbo is re-spooled. Another scary thought, the solenoid could be fed varying duty cycles and perhaps the cyclone could be 50% open or gradually opened. It would appear to take so much control strategy though. I would like to just do the experiments and back it all up with data but I'm a sloppy scientist, an armchair one.

Oh, my eyes are so much bigger than my stomach!

Other guy:
I don't see why you couldn't get piggy backs involved. But to run the cyclones, you are probably performing a code change anyway, so just get it set to your liking in the code. I know, I know, you want to optimize. Well that can be done from a single run with the runners open and a single run with them tied closed. That will give you the two curves to find the crossover point. Just like VTEC, the crossover point should be smooth. If there is a big jump in torque at that point, you f*ed up and rode the primaries too long, a dip, and you jumped on the secondaries too soon.

 

[joemoto]

The velocity of the intake charge, regardless of the PSI, is directly related to the ideal runner length. Therefore, RPM is a very reliable measure of the intake charge's velocity and to my understanding is better then Boost level. The size the turbo, creating the boost level, does not matter aside from of course the temp of the intake charge. Regardless of temp, it is still moving as fast as the RPM dictates to fill each cylinder.

I don't know...? ...sounds good to me

Joe

 

Quote:
Can anybody think of a good reason that these should not be triggered by an airflow variable rather than an RPM point?


I'll be doing something similar to that by running through an adjustable MSD RPM switch and a boost-referenced Hobbs switch. One of the easier ways I could think of killing two birds with one stone.

 

Quote:
The velocity of the intake charge, regardless of the PSI, is directly related to the ideal runner length. Therefore, RPM is a very reliable measure of the intake charge velocity and to my understanding is better then Boost level. The size the turbo, creating the boost level, does not matter aside from of course the temp of the intake charge. Regardless of temp, it is still moving as fast as the RPM dictates to fill each cylinder.

I don't know...? ...sounds good to me

Joe



Joe, very good of you to point this out. Fluid flow, even for the simplest of static flow situations in a pipe is something that I consider impossible to grasp.

So, I'll just fall back on my simulation tools. I ran a simulation of a single cylinder engine with the freeware lotus engine simulation tool and found that throttled to 66% of total power on this particular example engine, average inlet runner velocity only changed by about 10%. That would be like dropping down 1000RPM on the unthrottled engine. A summary is in order:

unthrottled engine
7000RPM 29.6m/s avg. inlet runner velocity
6000RPM 27m/s avg. inlet runner velocity
5000RPM 24m/s
throttled to 66% of peak torque (reduce throttle area to 40% max)
7000RPM 27m/s avg. inlet runner velocity

5000RPM 20m/s avg. inlet runner velocity

When you close the throttle, there is a lot more going on than simply lowering the plenum pressure. The fact that our engine has a medium plenum is going to help this though, to even out the nastiness of throttling I suppose.

I think I could prove this by leaving a throttle out of the system and trying to reduce peak torque by 33% just by changing the inlet static pressure (never mind exhaust)

Ok I'm doing that......

Yep, now the velocity remains nearly constant while the torque has dropped of by 33%.
So, I am going to agree with you, RPM is perhaps the best way to trigger the cyclone.

One thing that comes to mind is what happens when the engine is spun to 2x its red line? Does the velocity continue to increase with RPM? The simulation suggests that the velocity increase is kind of a logarithmically ddecreasing function. When you reach the point that the engine is no longer breathing enough to overcome its own friction, kind of a self imposed red line, the velocity is basically not increasing with rpm. (this simulation doesn't predict any mechanical failures or valve float or anything of that matter).

You yourself have pointed out that velocity isn't really changing, it is dictated, from what I can tell, by the rate of change of volume in the cylinder, the piston velocity for the most part.

What is changing is the density of this gas by way of pressure. Would you not say that this has repercussions at the fluid to walls boundary [layer]? (viscosity effects) and what about particle momentum.

Two gases flowing through the same pipe with the same velocity, one dense, one half as dense (throttling), will require different energy to change directions through a bend will they not? Those are basic things which I think have to be addressed. I need a "working model 2d" for fluid flow to gimp me through this stuff.

This is depressing, I am never going to feel confident about fluids.

 

Quote:
Imagine this scenario, racing, banging gears... Road/drag, transmission ratios are more than close enough to keep shift points above 4100 trigger point. The turbo is loosing shaft speed between shifts and must recover. Wouldn't there be benefit to spooling the car on the small runners? Small runners until turbo recovers.


I've got to imagine that you'd have to be running one basketball-sized turbo to not build enough boost above 4100 RPMs (with both runners feeding the engine) to get the car scooting again. And if you were going with a snail of that magnitude you surely wouldn't want to be using a Cyclone manifold. My reasoning for going with this manifold is fun around-town driving and possibly a tick better mileage. Went through the "I'm building a 700 HP drag car" phase a few years back and now just want something enjoyable to drive. God, I'm getting old... /ubbthreads/images/graemlins/wink.gif

 

[cheekychimp]

This is off topic a bit but quite closely related I'd imagine. Snow Performance offers variable injection point controllers to activate alcohol/water injection at set airflow or boost referenced points.

Any advantage to setting the injection poinr based on airflow as against boost or is that ONLY relevant to NA vehicles?

 

Personally I use the WI to eliminate/reduce knock, so boost probably be the best activation. maybe if you run a setup to save gas, airflow would be a better choice.

 

You could most certinatly program one of the outputs on a stand alone to open the secondary's at a given load point rather then a defined rpm. You just pick a load point to open and a load point to close and when ever the ecu calculates the air flow it can decide to open or close the secondary runner control.

On a side note, I usually set my secondary runner control and V-tec on my customers cars to come on a little late. This gives you a very defined activation point that you can feel and most importantly IT MAKES THE CAR FUN TO DRIVE. Doing this will not slow the car down mostly b/c when shifting in a drag/road race car you typically stay above this opening point.