The 3D-printed ITB intake system continues to evolve

[@Lacquerhead-TX]

Check out what Mazda did with the 787B Le Mans car. Sliding trumpets to meet your exact design goals of low rev and high rev optimizations. That could easily be controlled with servos (steppers with feedback to try really hard to reach the commanded position) and some linear guide rail.

[@StandRacing]

Thanks! 787 on the research list

[@TheSunChaserUS]

Thats the way to go, im working on that type of design with a shared plenum for a 4.6 Mod motor

[@maximilianstobinski2469]

Would be interested in that design aswell. Trying to build one myself for a k24 miata.

[@maximilianstobinski2469]

​@@StandRacing This but better: kzbin.info/www/bejne/j2StgIeZoM-NqaMsi=hj5MqrXrvrl6PwOm

[@Xsiondu]

Look into creapunk CLI17 closed loop open source stepper driver.

[@CrazyTony65]

Late to the game here but, take a look at a factory Honda H22-A4 intake, they had the intake runners split horizontally upstream from the connection to the head, the top one curved up and was short, the other stretched out further before turning up to the common plenum. A set of butterflys in the short side forced the air to the long runner until a certain RPM, opening the butterflys, the air naturally takes the shorter route. Also check out a factory Acura RSX-S intake, it's round like yours and has a drum shaped valve to change it from long to short.

[@m_remon]

As a programmer I am familiar with GNU GPL licenses for source code. These types of licenses require any derivative work to be published under the same license as the original work, enforcing the open source philosophy. (Or MIT, APACHE, and others) but I only see them used on source code. The other option I know is creative Commons. Something like CC-BY-SA forces every derivative work to include reference to the original and to be shared/published like the original.

[@StandRacing]

Oh I'll look into that. I too have only seen open source licenses being applied to code, this may be a bit of uncharted territory but the principles should be all the same.

[@zach446]

GPL 2 isn't that bad

[@shijikori]

@@zach446 GPLv2 allows for anyone to make private changes without having to make them available to upstream. That's besides the point though as that license is written for source code that can be used to make software. There might be other licenses existing for other kinds of projects such as the mentioned CC-BY-SA.

[@zach446]

@@shijikori yeah I ain't no lawyer so thanks

[@zuzuzaza98]

@@StandRacing you might want to look at CERN OHL or TAPR OHL instead. It's specifically used for open source hardware, so might be more suited.

[@joncot1812]

I'd recommend setting volumetric flow rate at the cylinder inlet port, not mass flow. With mass flow you're forcing a certain amount of air where the engine is actually displacing volume. I modeled this 12-13 years ago using transient flow and was able to correlate my simulated results with actual on vehicle MAF readings. Calculated volume differential curves based off bore, stroke, and camshaft phasing so it was a sine wave input with cutoof unique to each cyclinder. I was able to measure throttle response of different designs by measuring the time it took to go from near vacuum to steady state mass flow rate.

[@noncog1]

The most interesting comment I've read so far as far as actual useful information

[@warrenlucier5796]

Take a look at the Porsche GT3 design that has ITB's located at the cylinder heads and the single throttle body located at the plenum. This setup actually builds some boost in a naturally asperated engine. Good luck with this project your doing.

[@wrighteously]

That sounds really cool, I'm very unfamiliar with this I'm gonna research it. Cool stuff

[@racinggamer8989]

That sounds awesome! I did a little googling, but couldn't seem to come up with anything. Care to share where to begin?

[@cumbob]

​@@racinggamer8989maybe hpacademy has a video on it

[@HdHype]

Engineering Explained on KZbin has a video on it.@@racinggamer8989

[@johnhunter7244]

​@@racinggamer8989engineering explained has a video on it

[@SartekGaming]

In australia the vz commodore sv6 with ly7 motor at 190kw has variable intake while the le0 175kw model doesn't. I did a bit of a google and came up with this info on a forum thread "The inlet manifold has a variable rotating vane assembly Intake Manifold Runner Control Valve (VIM) which is driven by engine RPM over about 4000RPM(?). This is used to tune the inlet manifold for more efficient airflow rate at higher RPM. Built into the inlet air tract which connects to the throttle body is a "tuned stub" which acts as an input air flow efficiency booster at higher or lower RPM. (Not too certain of the designers thinking here - possibly lower RPM)." as well as "EDIT Did some more digging. 190kw engine. The VIM below 4000 rpm divides the intake plenum into two plenum chambers each feeding left and right banks to get low rpm efficiency. Above 4000rpm the VIM rotates to create a common plenum chamber for all cylinders."

[@BJL2142]

to tune i would data log mass airflow on long runners only, then do the same on the short runner only, overlap MAF readings and you will see your cross over point as where flow rates intersect

[@NigelSwan]

If using map/tps for load (I like tps with map correction) you can do 2 complete fuel and ignition tunes and then switch where torque is the same at the crossover points on a 3D table for it to feel (and be) seamless. Throw in some variable intake and exhaust cams and expect to spend a lot of hours testing and optimising it all together...

[@BJL2142]

@@NigelSwan no not for load, but mass flow, hp= fixed amount of fuel per given kg of air, the maf can output calculated amounts so you can visually see changes/gains in design against others or for figuring out what degree of variable cam timing is best at what rpm, for instance with 50 degrees of variability you can start at 0deg, do a run, change to 10 deg and do another run so on till the 50 deg, overlay their graphs on top of each other and you now know what cam timing is best at all rpms because mass flow will be highest at most efficient rpm ranges, did not mean in regards to tuning and using it for load calculations.

[@FlatlanderGear]

Matthew Gray Gubler and Trevor Moore’s brother explaining Intake airflow to me? Subscribed!

[@Thatdavemarsh]

I do appreciate a video where someone actually iterates through a design and talks through their thought process. I will be interested in your before and after VE tables.

[@944LS]

Good example of “a model is only as good as you make it” and that’s why it’s important to understand the fundamentals!

[@944LS]

Also to add. The mass flow into the system has to equal the mass flow out of the system. The variable that is potentially changing here is the area of the outlet vs the inlet. Mass flow = density of air times velocity of air times area. So air density is constant, you can change the area, so velocity will increase or decrease to conserve mass flow. The simulation may give better intuitive results given you understand how much air is demanded from the engine at a given rpm range, this would be your total mass flow out of the intake, then optimize the intake flow

[@elementaljosh]

This is really some awesome work, I know many OEMs have come up with similar solutions like Ford's IMRC on the mustang mod motors, Mazda's VICS on the NB miatas, and pretty much every ferrari intake from the last 25 years. There's a lot of torque and power to be had by being able to effectively vary intake runner length based on rpm.

[@Hosenhoffen]

BMW 3 stage DISA intake manifold from the N52 is something you can get inspiration from. They use DISA valves to change the flow between long and short intake runners. It's pretty easy to follow too. Yours reminds me a bit of that! Good luck with this build. it's really interesting!

[@n.shiina8798]

or N62's variable intake

[@cumbob]

Vr6 24v is also easy to visualize

[@winha1435]

it has resonance champer too and its combat design. Was n52 direct injection? One way to go would be to model m54/n52 intake as base cut it short to match engine and use same disa flap to controll the intake.

[@Hosenhoffen]

@@winha1435 It was port injected for sure.

[@n.shiina8798]

@@winha1435 nope, N52 uses port injection. DI engines have 3/4/5/7 2nd number on the N engine type (N53/N54/N55/N57)

[@jonnyroy]

My 2000 lexus gs400 v8 had this setup. A vacuum operated valve and solenoid would open a shorter path inside the plenum at higher rpm.

[@Leo99929]

Your harmonics are effected by the length of the runners, so you need to swap out the whole length of the runners by closing one short and opening the other in order to benefit from the pressure waves alignment with the RPM of the engine. You could do this with a change over valve which could be switched between two (or more) connections.

[@_..-.._..-.._]

Exactly! The sonic waves will be wasted bouncing back and forth through the whole system.

[@oscarzt1652]

my Ford ST170 (aka focus SVT in USA) has a 2-position variable runner length inlet manifold which uses long runners up to 6000rpm (ish) and a short runner for 6000rpm to 7300 rev limit. known as IMRC and i think has featured on other fords from the late 90s/early 00s

[@VirtualCryptid]

I definitely think the sliding trumpets are your answer... they have been done on the 787 and a few super bikes... if sliding is too tough...maybe try 2 or 3 sets of interlocking belmouths that are each controlled by a servo... that way you have a TON of adjustability... but all the dyno time will be brutal under each possible configuration

[@Tex777_]

BMW's original approach was called DISA, basically a dual-length manifold with a valve to open and close the intake tracks. I think some of the later iterations had 3 intake lengths. Currently with their virtually all-turbo lineup they have fairly short manifolds and use a valve lift control strategy they term Valvetronic.

[@NigelSwan]

Yep very common on OEM's, and effective.

[@jackoconnell2204]

i know n52 has a 3 stage man

[@stevesloan6775]

There was an Australian ford falcon straight six that used a dual port single throttle body. It also had a continuous 100 degree curved dual inlet runner.

[@cameronschaub6628]

I would add that Toyota's TVIS sounds similar to this as well and that was used on the 4age, I believe yamaha was involved in their intake tuning and it would be cool to see if some of their variable itb design could work on a miata

[@StandRacing]

Toyota TVIS added to research list 👍

[@NigelSwan]

​@@StandRacingtvis blocks one of the 2 intake ports per cylinder on the 16 valve head, so doubles the intake velocity by forcing the cylinder to draw through a single intake valve instead of the 2 available and increasing torque. Slightly different, but same result as a longer runner.

[@stevesloan6775]

When you said “straight plenums all the way out to the guards”. Long inlet runners have their tuning on finite.

[@chilledn52]

Currently running a 3D printed velocity stacks since I don’t have to worry about using a throttle body for MAF. Using “Alpha N” + Variable valve lift for throttle

[@CompproB237]

Congratulations, you discovered multi-length intake runners.

[@heretikmusic7054]

I knew Trevor Moore wasn't actually dead.

[@208Concepts]

The LT5 has a 3 stage intake system thats pretty neat.

[@HAZZA24937]

Porsche made an interesting intake that for a forced introduction car it used the intake pulses to reduce the incoming air density thus cooling it, which allowed them to run more ignition timing and make more power.

[@felixaudet5860]

You can ionized the intake air simply by creating a swirl motion in the intake, especially if in contact with conductive metals such as copper, silver gold or platinum. Silver plated copper would be good price / performance value. You can still get away by using Alcaad, mirror finish aluminum.

[@xtbb1369]

Another direction you can take the dual inlet setup is to divide the cylinders based on the firing order to give the plenum the most amount of time to fill before the next induction so for the 1.8 (1-3-4-2) you could separate 1-4 from 3-2. If you want to try leaning the runners, 15 degrees is the maximum angle air can turn without breaking laminar flow (standard 3 angle valve job is 30-45-60 for the same reason). and dual 3 inch inlet is a bit overkill, a single 3 inch flows about the same as dual sqrt((3^2)/2) = sqrt(4.5) ≈ 2 1/8-1/4 pipe so dual 2.5 inlet would be plenty. Additionally, if you want to flow everything through the stock TB, cut the butterfly shaft in half so only the threaded side remains, should increase the cross sectional area and reduce pumping losses. Hope some of this is helpful, and I can't find your discord.

[@philmariop]

Perhaps you could simplify and accelerate iterations by reducing your study to feeding one cylinder, one intake port. That could help in CFD checks and be easier to print test samples for use on a flow bench. Even the best CFD and FEA tools are cartoons representing what is real. On the far side single testing you could then merge feeding 4 cylinders, ports. Fwiw, check out the section view of the new ish z06 motor. The intake valve is fed by a very sculpted cone transition from the main plenum. Note that the v8 plenums in coyote, modular, or LS sbc use the same circular winding thing to package runner length.

[@ztrxel3374]

this guy is crimanally under rated

[@Giuliana-w1f]

17:10 some sport bikes have something like that. ITBs with short runners (and velocity stacks), going into a chamber with a throttle body, and longer intake runners. At lower rpm/load, the single throttle body is closed, and the air has to go through the longer runners; and at higher rpm with high loads, the throttle body opens and the air can bypass the long runners. I'd say that the main difference to what you did (other than the ITB placement you corrected later), would be that the air enters through every side of the short runners, instead of being a Y split (so it probably has better airflow)

[@tylerspin2417]

Toyota has the ACIS system. It doesn't use another throttle body like in your idea, but a butterfly valve in the plenum that opens (I believe) with vacuum at a certain RPM and and effectively shortens the runners. Similar concept

[@StandRacing]

Toyota ACIS added to research list 👍

[@zuzuzaza98]

Throttle body is basically just a butterfly valve with sensors and other accessories attached to it

[@tylerspin2417]

@@zuzuzaza98 that is a fact.

[@alexhise968]

There are 2 different types that Toyota has listed the same way one on some V engines has a plenum per bank then opens a communication valve allowing acoustics to fill a broader range of rpm. The other version uses a barrel valve to switch between a short or long runner length such as in the 2AR. Pay no attention to Toyotas use of tumble valves.

[@tylerspin2417]

​@@alexhise968 Correct. Even Subaru uses tumble valves.. something I believe is mainly for emissions? Not sure

[@agerrgerra1361]

I was doing some FEA the other day for some machined compressor components, it didn't look like it would flow well until I saw the result showing that the air was moving less than one meter per second, which was nice. However... turns out I'd forgotten to set the fluid to air, it was still set to water and I was using mass flow rate on the inlet... meaning flow was actually hundreds of meters per second. I feel like a small percentage of my time is spent actually setting or performing FEA, most of my time is just spent doubting my results.

[@Finnspin_unicycles]

You should always doubt simulation results, especially for CFD. Most of the simple CFD tools out there are optimized to sacrifice accuracy for numerical stability, so that anyone can just click through a simple program and get results. Maybe alright for a very rough study, but if you really want to optimize parts, you need to pay close attention to all the details in your setup. CFD is a lot more complicated than stress analysis with FEA (you should always doubt those too, but the governing equations for stresses in solids are waaay simpler than those for liquid flow and that shows in how hard the simulations are to properly set up)

[@domoku]

This design looks identical to whats on the Ford falcon 4.0 'Barra" Intake..

[@hotdognr12323]

I believe Ferrari among others have used a system quite similar to this on their v8's

[@r3bproduction]

Do we have any calculations of resonance tuning and what cams/exhaust/runner length/taper etc? I love the documentation and effort but feel once its on the car running there will be something overlooked providing negative or zero gains. When you run another simulation blocking the shorter runner you very well might find its actually providing less volumetric flow to the port choking it out. You can kinda see it in the plot towards the bottom having the really dark below with almost 0 velocity even though the shorter runner is wide open. Like wise having having uniform radius of the runner might help or hurt the flow as opposed to a straight runner. Q= A x V where area/volume are porportional but there are few other factors and you can look up the theory for nozzel flow where an increase in velocity is a decrease in pressure. Less pressure is less volume and less volume is less power along with some ideal gas laws with density/temperature of the fluid. Its like trying to fill a bucket with your finger on the end of the hose making the water have an increased velocity but very little volume vs leaving the hose open will have more volume but less velocity. Im not sure were your getting 6192 cubic in per second when ball park a 1.6 engine @ 75-85% Volumetric efficiency will have about 168cfm or 4838 cubic in/second. On the same site you used and for a straight circular pipe not a curved oval cross section would have 163.55ft/s velocity which is about right for production cars. Ignoring the intake design to be able to pull in 200+ ft/s you'd need some pretty good internal work and that isnt just whole power band but the 163.55ft/s is at 7000 RPM. I'd look into as well with cams/exhaust plays a roll and what volumetric effiency and flow rate really means instead of looking at vector/velocity plots which are good visuals but provide no real data on flow or performance. Start with the engine displacement, port length/size of intake exhaust, choose cams and bounce back and forth what resonate freq you're aiming for increase intake length, change exhaust, increase/decrease cam duration or degrees. The volume flow your using as well is a pretty big asumption off of pipe diamater which an oval cross section has different pros/cons. To get the exact amount of volume of air you'd need to do a little math on the engines bore and use a pressure drop once intake valve opens pulling in a vaccum and then just below atmospheric plenum pressure rushes to fill the engine bore. Think of a vaccum, it doesnt suck air in, there is a pressure drop inside and the outside air wants to move inside the unit to fill up the volume of the container. Might need to check that second intake design on the drawing board and the intake CFD and really question are they really any different in practice/theory with your observation of flow.

[@StandRacing]

I got 6192ci/s from 1.8L at 8k RPM, didn't apply an efficiency percentage. I have a 4 into 2 exhaust that I'm going to modify to the intake once these questions you've posed have answers. The way I see it, the chamber will be able to resonate at two different frequencies (and even a third that's not shown in the video, there's a bypass I envisioned that I showed in an earlier video). I think that's the next major step for me is understanding how to calculate resonance and tune with it.

[@JamesAnderson-fw5ew]

Alot of oem manufacturers use a butterfly to change the flow from a short runner to a long runner depending on map. I think it would be far easier/simpler to do. You could get a vacuum operated butterfly thing from a lexus gs300 or is300 to control the butterfly and the rest could be 3d printed

[@Google_Is_Evil]

The reason you want a variable length intake is to profit from the air volume in the individual inlet tracts resonance/speed while pulling in air. Using a single long header per cylinder and then combining the short header part with a single throttle valve will seriously mess with the whole resonance trick. Formula one had actual "slide trombone" inlet headers with flat slide throttle valves. This was possible because they had insane high RPMs and a narrow power band.

[@rgsv2248]

Twin plenum set up, with one for higher revband wave reflection on late 90/early 2000 Ferrari V12 ;) 550/575/456 etc... Keep in mind movig air = lower pressure, Keep in mind lager inner radius = better, can trick this with oval flowing area see Renault sport engine intake design

[@flymetothemoon5138]

This has been done over the years by many manufacturers, Ford VIS on the Scorpio Cosworth for example

[@BJL2142]

might be worth building a method of measuring flow consistently, Maf sensor on the front and a pump on the backside, because of the internals not being entirely separated you would need to group the runners as a common output, setup the motor on the blower so that you can read volts and amps and run it the same on all tests, this way you can see improvements in Real-time not just simulated, because with a given amount of power a given amount of work can be done and the Maf will show you the differences.

[@maxbgsl2348]

You definetly got to check out 1-dimensional gas flow calculation possibilities for that thing you wish to implement, as it is first of all acoustics of tubes and plenums system. For use with engine simulations, there is an awesome open source program called Lotus Engine Simulation.

[@burford2008]

I think possibly switching the short runners into more of a straight shot into the intake ports with the longer runners coming in at an angle might help performance.

[@BJL2142]

would be easy to test, make two profiles and run the motor to redline with both only using the short runners while logging maf, overlay the graphs and your answer is there.

[@Leo99929]

This is a two length inlet runner concept. Formula Student cars have experimented with continually variable length inlet runners by actuating the telescopic intake horns length. Single telescoping limits it to 50%-100% length range, triple: 33%-100%, etc. So it's diminishing returns for increased complexity. There's also the required change in diameter that might change your airflow undesirably if not factored in.

[@Sharpened_Spoon]

I love the scroll design. Is there a reason you couldn’t have a gate between inner and outer that is vacuum controlled or servo controlled based off engine parameters and input? Also is there a potential benefit to having a waste gate type door mid way through the scroll that feeds back from the inner to the outer; this could allow air to keep circular momentum when ITB closes and easily flow when it opens again.

[@Google_Is_Evil]

Air will go faster at lower pressures. Make sure you don't create a pressure difference that is so high that you aren't actually getting more air in the cylinder. There is a reason why MAF sensors output weight per time unit. The more air weight you get in, the more fuel you can burn and the more power you get per combustion. This is tricky to wrap your head around so it helps to step back and look at the principles of combustion engines and verify you aren't tricking yourself.

[@BenjaminHolcomb-kg1wv]

I believe in the mid 90s Honda had something like this on their c30 and c27 engines, (the engine that is in the nsx, and the 1997 v6 accords) VVIS/VRIS

[@Thatdavemarsh]

Neat work. My gut suggests that the plenum pipe may be undersized to feed the middle cylinders or the last one if feeding from one end (consider the volume and diameter of the stock unit).

[@NeoIsrafil]

Thank you! Things like this should be open source, because 3d printing and home robotic construction should be the thing that gives freedom to create and mod back to the people. ❤ The biggest problem you're gonna have is the PCM.. I guarantee it isn't set up to let you change your VE tables on the fly or have a switch to a second set of tables... Which means you'll need a custom PCM

[@noncog1]

I think he's likely already in the standalone realm, so it's just setting it up in the tables at this point

[@scottcasiowatcher3501]

Mitsubishi made an intake manifold for the 4G63 called the cyclone with two runners per cylinder (one long, one short) with a valve keeping the shorter runners closed at low load/rpm.

[@scotthatch4548]

You need to look up offenhauser 360 intake manifold.... it devided primary and secondary sides of the carb to increase airspeed ... each had its own runner ...there was a bit of a odd transition because the secondary air was at a dead stop and had to get up to speed when the secondary side opened

[@StandRacing]

Thanks for the response 👍

[@seanb250]

I can help you again in the math department for airflow 👍 Taper of the runner can be 1.5/2 degrees, then over your length will give you the diameter of the bell mouth. Airspeed wise on an engine we don’t really want peak airspeed past 640fps, this will occur in the head port itself at the minimum cross sectional area and localised points like the peak of the short turn radius. Airspeed is dictated by a pressure differential created by the piston accelerating down the bore, peak pressure differential should be less than 3.4psi to keep airspeeds under 640fps. I suggest testing your manifold in cfd at the industry standard of flowbenches which is 28”water column (1.01055psi) then you can calculate efficiency, the maximum average at that test pressure is 350fps, localised areas within the cross sectional area can be higher but the average can’t be higher than 350fps, that is 100% efficiency. Then you can use the cfm demand calculations I gave you last time to see if it all lines up. Look up the Ford Australia broad band manifold for our in-line 6’s they are a dual length runner manifold

[@StandRacing]

That's great info. I imagined that there is probably a max useable airspeed so I'm grateful you've confirmed that. Using 28" water column in CFD would mean I plug in 1.01055psi at the pressure inlet side? At the outlet side I use the CFM demanded flow rate? (215cfm) 👍Ford Australia broad band manifold added to the research list

[@seanb250]

@@StandRacing in regards to what to plug into cfd software I don’t know, I personally haven’t designed a manifold from scratch, I just port and modify existing manifolds and confirm airspeeds and map velocities with my flowbench. I would image that you would have atmospheric pressure at the inlet and then 1.01055psi less at the end of the runner, that is the pressure differential, air moves from high to low pressure. One thing I want to point out that occurs with real life physical manifolds as well, is that people port and modify them, then flow test them by themselves, not bolted to a cylinder head. This is not the right way to test a manifold, the cylinder head port dictates how the runner behaves, the runner also dictates how the port can move air, they are actually 1 and need to be tested as 1. When you flow test a head and measure the airspeeds at the opening they are not all even around the entire surface area, in fact the spread can be very wide, so when you bolt a manifold on, the manifold runner is not being drawn apon evenly on its entire surface area either. The issue I see with cfd is that I’m not sure you can replicate that pressure differential at the cylinder head side of the runner, plus that end of the runner is discharged into nothing, into an open space, so it’s not representative of a port and you will find that the results in real life bolted to a head to be very different. I am enjoying this project of yours tho 100%, the idea of straight runners is ultimately the way to go but as other manufacturers have done for packaging reasons the long curved runners have been done successfully. The broad band manifold I mentioned looks like a snail shell along the top, this is the long runner section, inside the manifold there is a row of butterflies on a shaft that are solenoid operated to open at a certain rpm point to open up the short runners, the manifold was used on Australian Ford Falcon SOHC in-line 6’s from about 1995 to 2008

[@julianjennings4638]

Expansion rate : 1.618. For runner taper, convolute spiral ratio. And horn tapers. For every 10mm in lengt of horn expand by 1.618.

[@julianjennings4638]

Check David Visard. 45 + years of scientific research and development.

[@getahanddown]

Total noob here but my 2c: Intakes are air cooled, you can get away with a lot if you insulate from the head eg bakelite gasket to short alu' section that has intake taps for your MAP reading, another bakelite gasket then plastic. Very likely to be OK. I run a 3D printed engine mount mod next to my exhaust / behind the radiator and it's sweet. You could also put gentle bends on the long intakes and have them going to an accumlator like a 4 to 1 exhaust collector. They keep up flow by timing flow in a circle, I'd bet the same thing + constant taper and smooth turns will add up to max velocity

[@stevesloan6775]

You’d almost need another multi throttle body setup run by a computer 😊🤜🏼🤛🏼🇦🇺🍀😎

[@sphericalsphere]

Check out the BMW DIVA intake. They used it in a V8 but there's no reason it wouldn't work here. They got full variable runner length by turning an inner cylinder

[@alexzgolak8820]

Did you look at hemi intake? Either the 6.4 truck or car or the 5.7 truck intake? The factory made a dual runner intake for those.

[@LordCogordo]

You know a guy its passionate for his topic when he says: "Here me out" I studied chemistry, all the aerodynamics that i know are from yt videos so take this as a ideas/question to a professional more than a solution per se. First i would work more on the distribution between the cylinders of the new intake, but its fking awesome, how it would react if you put a turbo there? Check it out with simulations! You can 3d print them to test performance, as long as you dont use it in a race i think that you can try modelling one after other and running it on a dyno or some similar way to check if its better or no. As for the plenum i think that they should enter from the left bottom corner at a 30/45° angle to aim the air directly to the cylinders and get rid of that 90° angle that you mention

[@enriquealvarez4702]

When working on b-series honda. The aftermarket intake market had optimized runner length and plenum size. I would start using their design and modify it for bp application.

[@BJL2142]

according to what? id like to see designs with flow data and hp figures because that sounds like bro science claim.

[@brianrobards8393]

Are you modeling with a fixed flow rate? You need to consider the flow rate versus valve events, the benefits seen from variable length runners is by taking advantage of the wave bouncing off the closed intake valve and getting a suction event timed at the same time the next intake valve opens. This is much more complex than just a fixed flow rate problem.

[@Aspire705]

Also check out Nissan's VLIM (Variable Length Intake Manifold)/Dual Plane Intake Manifold in their VQ30DE-K in the 2000-2001 Maxima.

[@StandRacing]

Added 👍

[@TheFrancisco170]

Keep it up man, very detailed

[@stevesloan6775]

It would be nice if you had a main airbox through put channel, and your dual inlet systems both syphoned from the same car speed moving volume. 😊

[@pauliusaleliunas2841]

If you have not already visited Garage 4age YT channel there is a nice video about ITB intakes and runners "Best Intake Runner Shootout - Velocity stacks - Trumpets - Dyno tested on ITB's - 200whp 4AGE"

[@flyfaen1]

Volvo had a system similar to this on the B5254F 5 cylinder (since you are adding a secondary throttle-plates anyway) Basicly the short runners were nearly straight with a valve that closed them off at low revs, where the cylinders then drew air through the long and curved runners from the same plenum volume (which had no valve), over 4000rpm the valves opened and the air took the shortcut through the short runners. It worked, acted kind of "V-tec'y", but became obsolete when they switched to VVT (first on intake only) as advancing the intake cam on the lower revs where it was "overcammed" basicly means earlier IVC, and thus more trapped air (far more than acoustics would do) than what would have been possible at low revs with a fixed cam. I'm sorry to say, but you are very late in the game. Toyota, Mazda, BMW nearly all else has had some form of active or passive two- or three-mode intake-resonance system before VVT became mainstream.

[@StandRacing]

I'll take a look at that one too 👍 I may be "late" as you say, but the game of advancement has no beginning or end, only a direction.

[@jvsyoutube3298]

telescoping trumpets outside the itb's, maybe using stepper motor, tuned for max hp on the dyno rollers, i want to see

[@kevinpaynter]

Your MAP sensor on the final drawing has to be in the plenum, otherwise you wont get proper tuning. Pulsing in each runner messes with the MAP. Pulsing in each runner is why a lot of people/race teams dont use ITB and MAP. You need to build a plenum/air box with the map mounted there to get usable readings for tuning. This is why ITB cars are so rough and jurkey at lower RPM.

[@HAZZA24937]

4AG silvertop and probably black top ITBs are very similar spacing to the Mazda BP engine.

[@ToasterTR]

you might try putting laminar flow guides a little bit up from the ITBs or behind the cone, effects should* include a much smoother flow to the cylinders and better combustion. not sure how much of a difference it will make in the final product since i dont think solidworks could simulate it properly but knock yourself out if you want to. search for a 2007-2011 crown victoria throttle body gasket on your browser of choice and you will find exactly what im referring to

[@avocadoarms358]

If you manage to make an intake that increases the hp na, you’re on to the winner of all bolt on power

[@julianjennings4638]

The Mazda 24 hour LeMans car used variable length runners with an electric actuator. Mapped to engine load and rpm

[@krap101]

It seems like where you're going wrong is looking at velocity by itself is only giving you part of the picture. You should also be looking at total pressure loss and mass flow rate.

[@Thatdavemarsh]

With pulsed flow, alternating in the tubes, is this steady state analysis valid?

[@Leo99929]

Velocity doesn't matter. Mass flow rate does. You need to either ask the software for that metric, or the average density at the boundary so that you can calculate it from the cross section and flow velocity.

[@TheRickyJohn]

Correct me if I am wrong, but long tube runners are for low end torque. If you have a race car you would rarely keep the car under 5k rpm, hence why everyone just runs ITBs on an NA car with the exception of specific class specific restrictions (some require factory manifolds etc)

[@StandRacing]

You are correct. But, that gap between rarely and never can make a substantial difference. The biggest difference being gear selection. Shifting takes time, so if you're maxing out 2nd gear out of a turn because you downshifted during braking, but keeping it in 3rd gear puts you at 4K RPM at turn in, that extra strength down low will make it very likely that not downshifting will be faster there.

[@alexhise968]

Another thing to consider is some racecars are starting to use CVTs and be hybrid version tuning for 2 primary bsfcs is now necessary for winning long race events

[@user-oz5yk9bm5c]

13:15 over 100% is realistic, honda F20C heads have a VE of 112%. Air is compressable and thats why achieving over 100% with modern tech is not out of the realm of possibility.

[@LukeAnthony-v3r]

There is an ideal length from the back of the valve to the throttle plate in the ITB. research that. it will change where you put the ITB throttle plates. depends on rpm of the engine.

[@gp8666]

i love variable flow intakes

[@nightmarejr]

Few car companies attempted this already. Comments have pointed some out

[@HAZZA24937]

I really hope the street version will take into account RHD MX5s they have the brake booster limiting space a bit.

[@energetictv9340]

Id say as a tuner that will work. How ever itb are very sensitive throttle responce and idle tuning can be a little bit of a pain. Your drive by wire that opens when you want wont work on most controllers but id say using maxxecu race it is very possible to do what you need. Not all standalones support.

[@StandRacing]

I decided to can that particular multi TB design to focus on the broadband 4x ITB streetable version. You can see it more in the more recent videos

[@pennysgarage]

Man, I think you just made a redesigned dual mode intake. Those already exist.

[@oikkuoek]

As for the 2-valve design, the junction before the ITB creates a turbulent low pressure area just above the valve. This could cause heaps of problems, up to backfire. As for the curved design, Air speeds up naturally when it's pushed/pulled through a loop. Shape of the loop effects on the velocity, where tightening curve accelerates the flow. Focusing on the outside curve inlet, if you'd trumpet the deviding wall towards the inner opening, this would create a vortex into that junction. This vortex would then drive more air into the outside curve, as well as keep the bleed into the inner trumpet minimal, while accelerating the inner trumpet laminar flow velocity. By playing with this trumpet bend, you could tune the flow between the outer and inner curves.

[@StandRacing]

That's an interesting note about the turbulent area. How far away from the head would this area need to be to give the air enough time to clean up?

[@oikkuoek]

@@StandRacing Basic velocity stack length math applies, but on the 2-valve system, there's always less air in the stack/port, due to the vortex created by the junction. The secondary valve is only creating more mess before the stack decreasing the overall flow. The double curve airbox has much more potential. The inlet to the airbox needs to come middle instead of side to make the load to each runner as even as possible. The round hole should spread like an old vacuum cleaner nozzle making the entering gate to the box the whole width of the box. Playing with the gate height effects the overall performance of the box. Get the gate height and the outside curve trumpet in tune, and you should see positive pressure @ the ITB end of the box.

[@oikkuoek]

@@StandRacing And one more thing about the Double Curve airbox, Due to it's functioning nature of acting as a passive, resonance based pressure chamber, it can't have any filtering membrane. every type of filter causes too much restriction for the resonance to function. A suggestion to fix that: Dust collector box. 100mm tube straight down from the middle into a 250-300mm ID bucket coated with thick grease. 10mm radius trumpet at the end, 50mm from the bottom. exit air into the DC airbox comes from around the pipe continuing 100mm through the MAF sensor. Second line of dust collecting is the gate into the airbox. This could be lubed with crankcase vent via catchcan breather. A row of 4mm nozzles at the apex of the inlet gate would form a curtain of greasy air to clean the intake charge from any particles. Just make the bottom of the gate removable for cleaning. Directing the crankcase gas back into the intake may cause knocking in extreme conditions, but if E85 is used, this issue should solve by the fuel.

[@StandRacing]

That sounds like a novel idea. After two engines melted down I vowed to ban all oil vapors from entering the intake charge in any racecars though. It sounds like it would work, but how awfully grimy is the entire inside of the intake going be after a 4-hour race?

[@oikkuoek]

@@StandRacing The state of grime has four factors: SIze of catchcan.The vapours should pass a crude filter, a layer of coarse steel wool for example. Height of the dust collector. Taller the bucket, lighter the trapped particle. Engine finish and breaking in period. If you finish the bores finer and cycle the engine/bottom end before you start it, you will have less blowby and more power. Engine tune and used rev range. If you run the setup "fat" and compensate the torque loss by overtiming it, you will stress the components to their wear limits. This results to more particles in the oil, more heat in the oil, less lubrication, more wear, more heat expansion and death. To fix this, you'll need tighter balance in rotating assembly, smoother hone and larger diameter in oil passages. That last part is a bit ticky, and needs to tuned with oil grade and system type in mind (dry or wet sump, dry sump can use thicker oil with fuel separator and cooler. If you fit ticker oil in wet sump, you'll lose power in oil slush. Slush management helps, but doesn't fix all of it. In wet sump, it's either slush or starvation.) Playing with oil grade allows you to play with bearing tolerances and piston to bore tolerances. There is power and rev range hidden in these too. All and all, the vapour circulation itself is very rarely the actual root cause. If it were, there should be more regular ICE's on fire on the streets than there's EV's. But, if the extra hydrocarbons gives a headache, one can always mold the combustion chamber to burn those too. This may smooth out low end, but frees up the top, where this setup likes to live anyways.

[@JohnBrown-cn8xg]

You definitely shouldnt “let the air figure it out” at the end if you want to achieve an even distribution of air-pressure over the four different cylinder intakes. If you want to do it the easy way, you could lengthen the way the air has to take in order to reach the first cylinder so it matches the distance needed to reach the furthest cylinder. If you dont want the added material weight, cost and room you could insure the even distribution of flow by giving the pipes reaching the cylinders different diameters. (Although there would be some nasty math involved in that solution)

[@LMBLNCR]

try a center entry for the plenum instead of the side entry, configure your design for flowing from a center port and see if you can increase the flow rate in doing so, might require mirroring the end opposite to the normal entry for the center port design. i'm also curious if there is a scavenging-esque method to the madness for designing an intake manifold similar to an exhaust, like if a triple y design for the flow would do anything to aid it, i gather it would disrupt the flow in reverse but i'm still curious!

[@sublunacy]

non linear flow towards the runners will net a freakish tune that will be non linear. i just went thru that on mine because i had a vacuum hose inlet in a bad area on the crosspipe, it was doing that to me. i believe this to be fact.

[@Leo99929]

Your intake throttle bodies shouldn't get too hot apart from heat soak as they're actively air cooled when in use. Hypothetically high temp 3D print should withstand the temps, but maybe not the negative pressures if you happen to use a flow restriction plate. Resin 3D printing has some high temp options.

[@chriserlandson6316]

vw cavd/cth engines have a change over flap design that might be good to look into for the swap from supercharger to turbo air.

[@MNPontiac]

do u know flow rate and cam specs of this engine so you get the harmonics right. if you using harmonics and wave pulses that is how you can get over 100% efficiencies at certain rpm ranges. multiple runner lengths have been used by many manufacturers. Ford did it on the old sho Taurus yamaha engines. it had intake manifold butterflies that would change into a shorter runner at high rpm

[@EXTRA---LARGE---FARVA]

This is brilliant. I think you're going to have to go through a lot of trial and error with diameters and lengths. Maybe make the design very modular. Then swap runners on the dyno to test if a smaller diameter will increase velocity the way you want, for example. I'm sure you can send CAD files to a Chinese source for prototyping, aluminum casting, and machining. I'm not sure how licensing will work for this, but MIT and Apache is typically the go-to for Python amd is fairly straight forward. I'm a lot more familiar with Python than automotive R&D practices. Good luck! I'm excited to see how this goes.

[@ericsmidt5612]

hi! awesome content! why don't you take a look to VR6 variable intake for some inspiration. They developed something alike

[@TheDonMariusz]

Have You thought maybe about design similar to slide trumpet (trombone)?

[@greatestevar]

TPU will def work for engine parts

[@raulgomez1420]

bmw use the disa valve, that in concept it is somthing similar to what you are trying to do. It comes un m54 engines like the 330i e46

[@charliemaybe]

Ive seen adjustable intake tubes on a couple race cars where depending on rpm and load, the intake runner length would change to optimize the power potential. I dont know if that is something you want to do, but it is an idea. It would make tuning a nightmare though

[@zslemp]

would a roller barrel throttle body be able to have less volume and more flow for the dual runner set up and not create as much vacuum? also if your designing your own system why limit yourself to a butterfly type throttle body?

[@wschoenborn92]

Are you Trevor from WKUK son??

[@user-oz5yk9bm5c]

when it comes to licensing the GPL v3 license seems to be best for this project, i just dont know if it is applicable outside of programming

[@Finnspin_unicycles]

CFD = Customer Fooling Device, you can fool others or yourself easily.. That calculator gives you bulk velocity, so the velocity that the air is moving at for a given flowrate if it were flowing in a uniform profile. But fluid in a pipe does not have a uniform profile, so yes, you are going to see some faster moving air in the middle of the pipe (and at the walls speed is 0, with some profile inbetween).. Certainly not over 100% efficient, that's just what happens when fluid flows through a pipe. I like the ITB+ conventional throttle body idea you present at the end. Tricky to know if it would work in this way, but there have been similar systems that switch from long runners to short runners on some cars. They typically used a blade on each runner instead of a single throttle body like you are proposing to "switch" all of them, but maybe one throttle works. (My worry would be that there could be some funky resonance going on in that plenum with the short runners.)

[@aethervvav1658]

This is a really cool idea

[@StandRacing]

Thank you!

[@sublunacy]

furthermore you should be testing it with tubing leading up to the manifold inlet.

[@StandRacing]

Agreed. From what I've learned about the impact of small changes, I'm curious now how much difference that can make all other things being equal.