Intake manifold fabrication

[gideon]

Shootingsight - I was puzzled to see CAD drawings of a downdraft manifold when I thought you were planning a sidedraft manifold.  On re-reading your post I saw that you did say downdraft,  so I was hasty and sloppy when I posted.  I'm sorry about that.

But then, part of the reason I made that mistake is it doesn't make sense to me to go to all the trouble of designing and manufacturing a new downdraft manifold that is similar to the original.  Surely it is easier to split the original manifold.  Ok, it's not reversible but so what?  It doesn't sound like you intend to ever put it back to stock, and if you are concerned about resale value then it seems to me that the classic car market likes Weber sidedrafts.  The Hermes style manifold is period correct and I believe it would add value.

Having said that, if you want to do it anyway despite what anyone says then good on ya!  I'd still love to hear what you come up with.

[shootingsight]

I plan on doing both, assuming costs work out.

I do not want to cut the original, in case I screw it up, and now I'm out a manifold.  Since I own a downdraft, I thought I would first make a downdraft, and then later go to a side-draft.  Indeed, I bought the side draft carb and manifold that was in the 'for sale' section, however I also want to make one.  I spent my career in prototyping, albeit for packaging.  So this subject is of interest to me beyond just getting myself a manifold or two.

I added internal runners.  I'll send this to my buddy with the aluminum 3D printer to get a cost estimate.  As stated, for a full-up commercial place in Europe, the price was $2,200 (though it drops to $2,000 if I do three :-).  I found a US based aluminum for $2,200, but if I do 3 it is only $1,100.

Also found another printer in the US who can do it in steel for $1,400.  So progress.

[jbcollier]

Better.  Paired intakes on a four cylinder work against each other so at least some separation makes a huge difference.  That said, you can buy a used manifold to cut up moderately easily.  Apply what you are learning to a proper side draft manifold.  As far as I understand, currently available single side draft manifolds do not have separate runners as it’s easier to cast them that way.  If you could design a single side draft with separate runners that would be well worth the expense — yes, obviously, the separated runners would have to join but the extended separation would make a big difference.

[shootingsight]

[jbcollier]

Now make it clear the header pipes without too much “swan-necking”.   I also would not extend the partition quite so close to the carb flange.

[shootingsight]

I'll 3D print this in plastic to assure fit.

[TurboFource]

Didn't Serge Sleurs 3-D print an intake? As I recall he used heat resistant plastic and tested for compatibility with gasoline.

[shootingsight]

I'm calling BS on Helmholtz resonance in the intake.  I found a cool app for my iPhone that does a frequency sweep with the speaker, and plots the intensity of the response via the microphone.  When a system goes into resonance, the response will peak - referred to as a bode plot.  Think about a trampoline.  If you bend your legs very slowly, you body just moves up/down by the distance you bend your legs, so there is a 1:1 response below the resonant frequency.  Also, if you jiggle your feet up/down very quickly, your body stays stationary.  So there is a 0 response if you move at a frequency higher than resonance.  But in between is a sweet spot, where you kick your feet a little, and the response is magnified.

So, I hooked an amplified computer speaker up to my iPhone, and placed the speaker at the inlet port of my new headers and ran a sweep from 50 Hz to 500Hz.  The strongest resonant peak was at 200 Hz, which is 12,000 cycles/min, which corresponds to 6,000 RPM for a 4 cylinder exhaust.

I have not run the test yet on my intake manifold, but since the runs are MUCH shorter than the exhaust headers, the resonant frequency is going to be a lot higher.  So Formula 1 or a motor cycle might put the intake in resonance, but not likely my little Renault engine.

Here is the plot.

[gideon]

Of course you could have figured out the resonant exhaust frequency by blowing down the pipe like a trumpet player and listening for the note.  If your app is correct is should be about a G3.  Test for equal tuning by doing each of the four pipes.

[jbcollier]

No need for the partition to go as far as you have it:

https://www.mgexp.com/phile/42/457774/DSC_9718.JPG

One last wrinkle.  Pairing 1&2 and 3&4 is not ideal in a four cylinder, inline engine.  Best is 1&4 and 2&3.  This can make for awkward lengths and shapes though:

https://www.lynxauto.com.au/ford-1600-kent-crossflow-1-x-weber-dcoe-manifold.html

[shootingsight]

LOL, I used to be a trumpet player, so actually, blowing on the header was what I first tried.  However without a mouthpiece, it didn't work.  But now that you have thrown down the gauntlet, I will figure out how to mount a mouthpiece.

I did put the speaker on different ports.  To my pleasure, they all came out close to 200 Hz, so it is a tuned system.

As to what ports get paired, I'll look at that.  Yes, it is messy, but the benefit of 3D printing is being able to do things that are not ordinarily possible.  Are there any twin barrel side-draft carbs with vertically aligned ports?  Alternately, are there itty-bitty carbs that would allow a dual .... or even 4 carb arrangement?  Perhaps motorcycle carbs?  Read the bottom of this page - four motorcycle carbs is a Spitfire setup .... I may have to follow this further. http://www.triumphspitfire.com/carbchoices.html

Alternately, alternately, people have spoken of Mega-Squirt as a computer/controller that allows you to mix up fuel injectors, temp sensors, mass flow sensors, O2 sensors, and tie it all together to get a fuel injection system.  A quick look at throttle bodies suggests they are only about $50 each.

[shootingsight]

Ha! I was right.  Helmholtz is nonsense on intakes.

The goal on intakes is to develop ram flow.  Once your piston reaches Bottom Dead Center, the suction stops.  So your goal is to make sure the flow in the intake manifold has enough momentum that the air keeps flowing, packing in extra fuel/air before the valve closes.  This is what justifies long/thin runners over a larger pipe.

So as air flows, the momentum of the incoming air is m x V, mass x velocity.  So if the intake manifold runner were the same diameter as the piston (and pretending for a second that there were no flow resistance in the valve - ignore that for the sake of the example), the flow in the intake manifold would have the same velocity as the piston going down.  However if the pipe had half the cross sectional area, the air in the pipe would have to move twice as fast ... and if it had 1/4 the area, 4x, and so on.  However there comes a point where (as I said above), drag through the pipe is a 4th order factor of diameter.  So as you go smaller, the drag goes up.  So there comes a point where the extra drag from a smaller pipe is less than the benefit of the extra velocity ... and therefore extra momentum in packing the piston with extra air.  I'm going to have to ponder, and I'm not sure it is easy, to figure out where that crossover is, but it is the momentum that counts, and based on comments, it seems 1 round 2" pipe is not as good as a 2" pipe with a center divider.

That brings me to the second half of the revelation.  Momentum does not work as well if there are flow restrictions.  Going around a corner is a flow restriction.  This is why down-draft carbs are not as good as side-draft.  In side draft, you get a straight shot, so you can develop better straight line momentum of the gas.  So when JBCollier told me the manifolds should not have too much 'swan necking', that was why.  Bends in swan necks slow you down.

SO now I'm thinking four straight pipes with four single barrel motorcycle carbs.  No bends.  Gotta figure out how to size carbs, but a cycle of 800cc going at 12,000 RPM is the same as a 1,600cc at 6,000 RPM in flow rate.  I have to think aabout pipe diameter though.

[shootingsight]

Well that ran full circle .... the Mikuni website describes the PHH series carbs they used to make for automotive use, but with fuel injection growth, the market has died off and they are discontinued .... they recommend going to DCOE carbs from Weber as a close substitute ....

[jbcollier]

Straight runners won't work as there will be clearance issues.  The ports curve in side the head as well.  Definitely keep curves to a minimum and make sure the curves you do have are open and  have smooth transitions.

Yes, you can use four motorcycle carbs.  Mikuni has flat slides meant for 4 stroke engines.  They flow very, very well indeed.  Much better than a Weber or Dell sidedraft.

This is fun BUT it is pointless to have massive flow capability when the engine itself can't take advantage of it.  Porting the head, fitting larger valves, higher compression, more cam overlap and lift, etc.

[shootingsight]

I hear you.  Without engine tuning, you get to a point of diminishing returns.  I'm not looking to do a total rework on a Renault engine, but if I can get the 80 for the 20, I'm good.  I am also curious if new manufacturing techniques, like 3D printing can make something that was never before possible, so this is not really changing science, but just optimizing it based on possibilities.