constantly variable length runner intake. interesting. 1

[robinairtechsupport]

constantly variable length intake runners...


just bouncing ideas around at this point, as i am in no position to begin production on anything at this point, though i may build several prototype units to fool around with. my idea is this... individual throttle body setup with constantly variable length intake runners, to change runner length/volume/tune per rpms, as well as possibly several other factors, such as a/fr, timing, load, and possibly even boost, while driving.

to the best of my knowledge, there are several ways to do this that make sense to me. i believe at least one, if not all have basically been done, by people who are obviously more experienced and intelligent than myself, and seeing as most people on here have intrigued me since i found this forum, i would like some oppinions, as well as some help. as i said, this is probably not an origional iadea, but has probably not been produced for a variety of reasons, probably cost, as well as packaging and reliability concerns. i would like to learn from other people's mistakes. it saves time...

so... hydraulic, electronic, automatronics, mechanical, automotive, fluid dynamics... i am a 20 yr old fabricator with hopes of going to school some day, but these are not my areas of expertise. i work as technical support for a company that does a/c equipment, but on my time i build vehicles. engine and driveline swaps, roll cages, frames, custom exhaust, turbo kits, tuning,
suspension, and r&d for backyard racecars.

a tuned itb intake is a great, but lift changes w/vtec, and optimal length and volume changes w/ rpm.and a fixed length runner system needs to be a compromise between high rpm, low range torque, and idle quality, with the characteristis of one taking away from the others, therefore making tuning into compromization. i like the throttle response, power, and pure coolness factor of the itb setup, but honda, as well asother manufacturers did design and build dual runner intakes, albeit w/ a restrictive butterfly plate setup, a common plenum, and imperfect flow paths, but probably only because of the damn bean counters.this is a great iadea, but not well suited to a race engine. i'm not trying to make a super amount of power, not looking for miracles, just the ability to have the optimal tune at all times, with the least restrictive system possible, to make use of other modifications to that glorified air pump we call a motor. need iadeas, want to share, and explore options.


basically, a conventional cable throttle with mechanically synchronised linkages through a common single cable would be hard, since i am trying to make them[the throttle bodies, not the butterflies] move possibly individually throttle butterfly would be actuated either electronically, with actuators connected to a variable voltage output from the throttle pedal, or mechanically, with a split cable system comparable to what splits brake cables on a bycicle, (1-2, 2-4)giving each assembly adequate travel, wich would be easier, though slightly less acurate. i have the concept figured out pretty much, and i have several actuation systems in mind to change the length of the runners...

the best, least restrictive method of changing the port length i can come up with being to slide a "middle tube", encased between an "inner" tube, being as thin as possible, and chamfered for flow, and an "outer" tube, for strength. the middle tube being attached to the throttle body, and the inner and outer attatched to the head.


object being to account for differences in port design on the head, air requirements, fuel requirements, etc. and possibly monitor throttle position, a/f ratio, intake and exhaust temp, map, (and possibly boost) on each cylinder individually, and feed those values into a program to produce x, x being a specific voltage as per that specific set of variables, specific to that cylinder, with x being fed into a set of actuators to lengthen or shorten the runner,controlled by x.


the throttle body/runner would be at the maximum allowable angle to match the port in the head, addressing flow concerns, while still considering packaging. i figure that with 6 inch(approx) tubes, i could vary effective runner length from (aprox)9-15 inches (about 3" from face of head to top of valve) to make use of that elusive "n/a boost" effect produced by catching the pressure waves at the right time.

one approach being electric/hydraulic, with a hydraulic pump driven by the motor, connected to a series of variable lift solenoids or regulators electronically controlled by x, feeding a system of small hydraulic cylinders with a return line connected through a specified sized "jet" (for lack of a better word.) to let fluid bybass into the return line at low pressure. with controllable flow solenoids, it would be possible to use an electronic Blight springs would hold the tubes in the "closed", or short position, untill pressure was allowed to build, forcing the cylinders to "open" the runners, or make them longer. the cylinders would actuate at least two points on each runner, to eliminate binding, but the system could also be controlled by just hydraulic pressure based on how fast the pump spins by sizing the return orifice to hold back 1 pound of pressure at idle, w/ a 2 pound spring threshold, and then the blockage created would hold back flow and create pressure in relation to pump speed, wich is proportional to rpm...

another option being to control the system electronically, wich will probably be the way i'll go with electronic actuators, controlled directly by x, wich could slide the assembly as well.

please: feedback. thank you. also, please don't patent my idea. its probably not that good, rather expensive, and i'd hate to build one, and have someone confiscate it because someone else patented my project. thoughts, ideas, and information should be shared, discussed, not hoarded. technological advancement is dependant on communication. i don't know anyone who can even remotely understand what i am talking about, let alone the principles behind why it would work. so this is my last hope. am i an idiot? or just smarter than southwestern pennsylvania? i bet im an idiot. let me know.

also, as a final word... something that has always stuck with me: it takes hundreds of engineers thousands of hours to design a car. what makes me think i have a better idea?thats what everyone always tells me, and for the most part, its true... but the thing is, what is the car designed for? oem designs for economy, cost, emissions, fuel efficiency, market trends, etc. i dont care about any of that. loud, smelly, expensive, complicated, high-maintenance, ipmpactical, i dont care, as long as its fast and fun. i build other ppl's junk, b/c i cant afford to buy new. i'd rather spend hours and hours building something than to buy it new. even if thiers works better, i'll cut and weld till i get it right. works for me. thanx... ADI

 

[BrianPetersen]

Formula 1 cars have used something like this for quite some time.

Also, you might want to take a look at the current (2007 model) Yamaha R1 motorcycle engine. It has variable-length intake runners. The Yamaha system is a telescoping inner-and-outer-tube system as you suggest. The Yamaha engine is drive-by-wire, as with most modern systems. Operator gives an input signal to a computer, and the computer gives signals to a pair of stepper motors. One of them opens and closes the main throttles through a small gear-reduction. The other one operates a mechanism that slides the intake runners in and out.

 

[robinairtechsupport]

WELL... GUESS THAT ANSWERS MY QUESTION. THANX, JUST THOUGHT IT WAS A DECENT IDEA. THANX FOR ENLIGHTENING ME, I FEEL KINDA STUPID NOW. SORY TO WASTE UR TIME. THANK YOU.

 

[RickFoley]

The Ford 3.8 liter V6 from the mid 90's is another example. Don't let that stop you. It sound's like you have some good idea's. Keep tinkering. And keep a log book. You may have a problem years from now that you solved today.

 

[SomptingGuy]

This practice is actually banned in F1 now (from last year I think). A shame really, because it was quite neat.

 

[patprimmer]

Individual throttle bodies DO NOT make more power. They do give better response.

The trend in drag racing has been toward individual runners of a common plenum with the air supply to the plenum being throttled. This was to move the restriction of throttle plates away from the restricted areas of the runners to where a very large throttle plate can be used. Of course spool valve type throttles overcome this, but at a cost.

Sliding trumpets will work whether on a throttled plenum or an individual throttle body design. Biggest problems will be air leaks and friction in the sliding mechanism. Individual throttle bodies can be below the slides. A thin rubber bellows on each runner sealed above and below the slides will also cure air leaks.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Ataloss]

Rick Foley said: "The Ford 3.8 liter V6 from the mid 90's is another example."

Does this engine actually have variable length runners or does it simply switch from one fixed length runner to a different fixed length runner at a certain RPM?

Bob

 

[SomptingGuy]

I cannot imagine a 90's production engine with stepper motors on its intake trumpets. It must have been a simple VGIS. I recall driving a thunderbird with a 3.8 V6 from that era - it was a real dog! Many things could have been done before fiddling with continuously variable intake trumpets.

 

[MikeHalloran]

I couldn't find a picture of it, but I think there's a BMW engine that has continuously variable length runners, formed by a grooved drum and a shell. The drum rotates under control of the ECU. I think I saw it in the SAE house rag, and I had the impression it was in production, or coming very soon.




Mike Halloran
Pembroke Pines, FL, USA

 

[robinairtechsupport]

the ford 3.8 did have an intake with two seperate fixed length paths, so did many other manufacturers. i am not sure about the f1 cars, but the yamaha's system, though revolutionsry, is not exactly what i am describing. it kind of switches between two fixed lengths as well, also, i cant see how that method wouldn't hurt flow w/ the step down in length produced by basically creating a leak. i'm sure there's reasons, as this is a high dollar sport bike, and i'm a kid w/ a welder, but in my ideal setup,i would have telescoping runners, connected to a throttle body, with a fixed air horn. the runners telescope, with the middle and inner tubes being as close in diameter as possible, for flow, to have a constantly variable runner volume, basically adjustable on the fly. other systems, including the r1 change between two fixed lengths. another iadea would be slide throttle plates, instead of conventional butterflies, for better flow. the seals would be another area of concern. bellows would be functional, and maybe its just me, but i've never liked the looks of bellows, kinda like its trying to hide something. i'd prefer seals simmilar to cam or crank seals, but they would need lubrication, and probably wouldn't last long, as that is not the way thet are designed to move. it may be better, however, to fix the location of the throttle body, and make the air horns move, effectively doing the same thing, and if a slide throttle was used, instead of butterfly, it wouldn't make much difference, but you would still need something to create the proper turbulence in the air path. moving the air horn would work to change the effective volume of the runner. sounds like bmw moves the air horns, and yes, i understand the ir intake design, w/ common plenum, but regardless of trends in drag racing, thats not what i am trying to do. thanx for all the comments, and please keep them coming. please understand, yamaha, bmw, and formula 1 have spent millions on r&d. im talking about making a part that i thought of while i was bored at work, for less than $200.00 when i make the intake, i'll post pics. 'd like to use clear acryllic for the tubes, it'll look cool, but how well will it stand up to the heat? any other suggestions? once again, thank you for your time,

 

[BrianPetersen]

The R1 makes the leakage issue irrelevant by containing the entire length-changing apparatus within the bike's airbox (air filter housing a.k.a. plenum, for the car guys).

I'm pretty sure the R1's system is capable of being infinitely variable within the limits imposed by the mechanical parts, even if the ECU programming might not (yet) be taking advantage of that.

 

[robinairtechsupport]

not leak in the sense of outside air, but regarding flow. air horns are shaped the way they are for a reason, b/c that shape is conducive to max flow. if u put a gap between the air horn and the runner, even if it is infinitely and progressively variable, u introduce flow from the sides, creating turbulence. if u extend the runner, or the air horn w/o creating that gap, u keep the flow as smooth and controlled as possible. i guess what the r1 does, like u said, is enclose the tbs creating a plenum effect. what i am trying to avoid, by using itb's, is theplenum effect, b/c of the way a 4 cyl is designed, one cyl robs air from the next. i'm not really speaking of the type of system in the r1, although it's an awesome setup. lots of respect to all u engineers, i've learned a lot by studying that design, and it's pretty simmilar to my design, but i'm trying to extract every last bit of power out of a high revving, de-stroked motor, capable of moving a lot of air, very fast. probably a honda d16z6 destroked to a 1.3, with itb's, and a free flowing head and exhaust. high rpm race only motor, but w/ that type of motor, i need to make the best bottom end power possible. thank you, mr. peterson, b/c u made me realise that it's not really a design flaw, but probably makes the bike a lot more streetable. i appreciate all the coments. thank you

-A. D. IADANZA

 

[RickFoley]

That is correct. The 3.8 liter Ford V6 was sequential (trumpet like). You may be looking for something more trombone-like. Infinitely variable. One moving part. Minimum un-metered air leakage. Possibly vacuum actuated. Porsche used this set up one the new 911. They insist that is was only to capture the sound of the old RS throttle bodies at full song. I'm not sure of any published power curves, both with and without, though.

 

[samplesecond]

Do a bit of poking around for the Le Mans winner-Mazda 787b. Its a 4 rotor wankel with pneumatic actuators that variate the length of the intake runners.

 

[wgknestrick]

I've seen vids of the 787B (as mentioned above) of the VLR system sliding up and down on youtube. You should try to find those.

The control system for this is going to be hard to do, because I think it would be difficult to keep up with the engine RPMs. You will obviously have to use a power signal from the alternator to get RPMs as an input, but the actuation is going to be costly to get components that can hold up to 200F working environment.

This is certainly a cool and challenging project. Good luck.

 

[patprimmer]

Why do you need the signal from the alternator to get RPM. If it is a non toothed belt drive as most are, belt slip will certainly seriously effect accuracy.

Crank shafts, ignition firing and camshafts seem to rotate in accurate proportion to engine speed.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[robinairtechsupport]

a signal from the alternator? i dont believe that would give u rpm. maybe the coil, distributor, crank position, cam position etc. as far as belt slip, most timing belts have teeth, serpentine and v belts dont, also, cam gears, timing, etc., while proportionate to engine speed, are adjustable, hence the invention of the timing light. they are adjustable to make the engine run better, wich is what i'm trying to do. if i really wanted to be accurate, i'd catch the signal from the crank position, or crank angle sensor, and use that to reference rpm, as well as where what cyl was on what stroke. i'd use a signal from the coil or distributor for actual timing. i don't see how the alternator would work. at all. ever. in any way. period. 200 degrees would be heaven if thats all the hotter it got in my bay. coolant temp is like 190, and the exhaust manifold kinda hangs out right behind the rad, no heat shield, it gets pretty hot. i could get the wiring to hold up, isolate the brains, and insulate the stepper motors, or use mechanical linkages w/ stepper motors. i've researched stepper motors a bit. cool stuff. thanx for the suggestion. can't find the vlr on youtube, u got a link? well, i'll do more research. thanx, ADI

 

[BrianPetersen]

On the 2007 R1, the original-equipment ECU that operates the variable runner mechanism (along with everything else) has no trouble pulling the engine speed signal from the same crank sensor that it uses for ignition and fuel-injection timing, and it has no trouble operating the stepper motor very quickly to "keep up with engine RPM" ...

The throttle bodies on that engine (and every other bike engine) are right next to the engine, separated only by a rubber flap as a (sort-of) heat shield, and rubber "manifolds" as insulators between the cylinder head and the throttle bodies. The "under-hood" environment in that application is the same for the intake runner stepper-motor as it is for the main throttle drive-by-wire stepper motor; sure it all has to be designed for the environment, but it's not rocket science to do that ...

 

[bradrs]

On the R1, that is acceptable because all of those are controlled by the ECU, and the ECU gets the cam/crank data. If you were to add a second device on to the cam/crank sensor line, the loading of the line can adversely affect the signal conditioning to the stock ECU. Care has to be taken whenever messing with signals on OEM magnetic reluctance sensors like most bikes use.

If this is on a car, with the common hall effect cam/crank sensor, you could probably do it without any adverse effect. But realize that any noise you accidentally inject into the system would affect vehicle operation.

Reading the signal off the ignition has its own drawbacks. It won't necessarily read properly at rev limit, or when a 2 step launch limit is activated. It could also affect the driving of the coils, depending on your ignition setup.

Reading off an OEM tach signal would probably be the most consistent, require less signal conditioning, and be less likely to have an adverse effect on vehicle operation.

 

[IceStationZebra]

I concur with MikeHalloran, this was done by someone. I remember a big write up in "Engine Technology International". It might have been MB. ISZ