PDA

View Full Version : Intake manifold design



mitch_04
10-22-2013, 02:28 PM
I've always thought it'd be interesting to build my own intake manifold. I have a 3.5L inline 5 out of a '05 Colorado that maybe my test engine. I've noticed most sheet metal intakes are built with a rectangular main section and the runners are cylinders. Is there a reason for this?

Also, why don't people manufacture an intake manifold similar to a reverse header? As in having equal length runners meeting up at a collector, then having the throttle body mount onto the collector?

Would love to get a good thread going on this!

sam 74
10-22-2013, 07:07 PM
there are some intakes out there like you described, some are custom builds and some are mass produced, mainly tunnel rams or large single plane intakes. i've run both and find more power in the single plane over the tunnel ram for my engine, but for cool factor you can't beat the tunnel ram for NA applications.

e129745
10-23-2013, 01:12 AM
Fluid Mechanics and its various sub-disciplines like fluid dynamics along with several of its sub-disciplines like aerodynamics and hydrodynamics is(are) a branch(s) of physics that is considerably complicated. It contains an amazing amount of compromises that complicate the endeavors of even the most financially backed entities. Consider running an engine at a static RPM, load and ambient air temperature. Even then it is very complicated, but now consider dynamic RPM and loads (forget temperature/density variances) and the magnitude of complexity goes up dramatically. Pressure differentials caused by opening/closing valves, piston motion, pressure reversion or reverse pressure waves, turns in the plenum, cross-sectional area, velocity differentials, temperature differentials, kinetic energy differences between fluid (gasoline) and gas (air), surface textures, etc, etc.
I love this stuff, but I have come to understand that limitations are very soon realized when looking for a simple answer to such topics.

astroracer
10-23-2013, 02:07 AM
I am not schooled in any of this "dynamics" stuff but I do have enough of a back ground in design and developement to know how this stuff works...
Take, as an example, a simple throttle body spacer for GM's TBI unit. I have one (a smooth bore) on my '95 pick up and it did make a noticable improvement. NOW take that throttle body spacer and add a helix (and a lot of hype) to the diameter of the bore to "spin" the incoming air column... Is this going to work? No. Why? Because of fluid dynamics. The air charge wants to flow in a linear direction (straight). Introducing a very short section of bore and trying to "change the direction" of the air column will generally create a restriction as the air tries to flow "over" the disruption in a straight line. Also fluid (air) will slow down the closer it comes to the outer wall of the bore. This creates laminations in the airflow which further reduces the total flow. Also consider the diameter of the bore. IF the inside diameter of the helix is reduced to put it IN the main flow stream (so it can react with that stream) you are further reducing the total flow by reducuing the bore diameter. If the helix is outside that flow stream (same size as the bore in the throttle body) it will have little effect on the column of air flowing past it other then to reduce some flow because of the lamination effect.
It's the same hype they use to sell those "Turbinator" air cleaners... "Spins the incoming air for better milage and power".... What does the air do as soon as it hits the carb or throttle body blades? It straightens back into a linear flow... All those spinner air cleaners do is block flow...
As you can probably tell this is a bit of a peeve with me... People fall for and buy this junk everyday and it kind of pisses me off... :)
Mark

Schwartz Performance
10-23-2013, 04:30 AM
I was at a junkyard recently and noticed a tubular intake manifold. It was a 1.8l Toyota corolla.
While this doesn't answer any questions you have, I too had thought about "why not" for a tubular intake like this. :)

-Dale

mitch_04
10-23-2013, 06:45 AM
e129745, I do understand that it is an incredibly complex concept. However, you could say the same thing about most anything we do. Adding turbos, changing camshafts, cylinder heads, etc etc.. all these things can be made to be insanely complex. We experiment using the best of our knowledge. I would imagine that many of the successful sheet metal manifolds were made by people using a basic understanding of airflow and a lack of aftermarket support. I hope I'm not coming off as brash, but I'm not trying to have a production grade part for GM. Just trying to create something that flows better than an original intake. Especially when they put "mufflers" on the factory intake to keep the "whoosh" sound down.

Schwartz, that's the sort of thing I'm thinking about! I'm curious whether or not it could work... I do have a friend who is an engineer that just moved close by, might have to give him a ring. Or just build one someday, the works case scenario is it is ineffective and I'd have to go back to the original.

Astro, I remember when the Tornados came out. Never made much sense to me, but there is a sucker born every day! The spacer is a sound concept, sort of like a tiny tunnel ram.

I had came across a discussion on home-built intakes onetime but can't find it for the life of me now. Had a great discussion on airflow.

astroracer
10-23-2013, 06:51 AM
e129745, I do understand that it is an incredibly complex concept. However, you could say the same thing about most anything we do. Adding turbos, changing camshafts, cylinder heads, etc etc.. all these things can be made to be insanely complex. We experiment using the best of our knowledge. I would imagine that many of the successful sheet metal manifolds were made by people using a basic understanding of airflow and a lack of aftermarket support. I hope I'm not coming off as brash, but I'm not trying to have a production grade part for GM. Just trying to create something that flows better than an original intake. Especially when they put "mufflers" on the factory intake to keep the "whoosh" sound down.

Schwartz, that's the sort of thing I'm thinking about! I'm curious whether or not it could work... I do have a friend who is an engineer that just moved close by, might have to give him a ring. Or just build one someday, the works case scenario is it is ineffective and I'd have to go back to the original.

Astro, I remember when the Tornados came out. Never made much sense to me, but there is a sucker born every day! The spacer is a sound concept, sort of like a tiny tunnel ram.

I had came across a discussion on home-built intakes onetime but can't find it for the life of me now. Had a great discussion on airflow.

That's it! Tornado... I couldn't think of the real name. I have always called them Turbinators... The spacers do work, I can attest to that. The smooth bore spacers will do the job though. Those Helix things bring nothing to the table over the smooth bores.
Mark

SShep71
10-23-2013, 07:23 AM
There are MANY aspects of the design of an intake manifold, it is similiar to intake port design. Its all in what you want the part to do. The over all design (single plane, tunnel ram, tubular, etc) all have good points and bad points. It basically comes down to three things, Volumetric Efficiency, Runner Velocity, and CFM. If you just want to put an intake together for the cool points and the experience its not a problem, but like any system the whoole system is only as good as its weakest link. To properly build an intake you can start at the cam, or start at the heads. The flow characeristics of the intake have to MATCH/COMPLIMENT the flow characteristics of the heads, and the capabilities of the cam. You can do all the computer modeling you want to reguarding flow characteristics but until you put your intake on a flow bench you will never understand how it actually works. The reverse header you referanced is exactly what those tunnel rame intakes are essentailly. Each individual port is a runner and the collector is the large center section. It is designed like this to keep the volume and flow consistant over each port.

mitch_04
10-23-2013, 01:14 PM
I'm curious why people use a rectangular "collector" vs tubular, seems like it impede airflow. I did find the thread I had been looking for though!

http://www.theturboforums.com/threads/361799-Intake-manifold-design

SShep71
10-23-2013, 01:42 PM
The ports on the heads are rectangular, so it just makes sense to stay rectangular. There are several sheetmetal intakes out there that have round ports that transition to rectangular where the port mates with the intake manifold flange. If anything the rectangular design should assist with airflow. As ASTRORACER mentioned, you do not want the air to spin while it is going through the port. Its a matter of physics. If you want to witness it there is a old experiment where if you take two 2liter soda bottles and you glue the caps together with a hole in them fill one bottle then turn the hole thing upside down the fluid inside will start to spin and the top bottle empties into the bottom bottle rather quickly. If you add something like a flat plate to impede the spin effect in the bottle with the fluid in it the rate the top bottle empties into the bottom bottle in increased drastically. Its all about an object (air) maintaining its direction and velocity unchanged. When a change is applied to one factor it translates into a direct change in another factor. That is the most simple way I can think to explain it.

astroracer
10-23-2013, 01:45 PM
This is all going back to the flow characteristics of a rectangular x-sect as compared to a round x-sect. Go back to my first post where I was talking about "laminar" airflow. The rectangular x-sect will actually flow more air then a comparatively sized round x-sect because the corners give the air column a place to do it's slow, laminar thing that is outboard of the actual fast moving air column. This "lost space" in the rectangular tube actually helps increase the flow on any given air column. Most round tubes, to flow the same numbers as a rectangular tube, would have to be almost half again as big as the rectangular tube. That is why you see the rectangular tubes on all of the high-end fabricated intake manifolds. These guys don't do stuff "just because", there is a lot of testing and tuning that goes into these intakes and they do what they do for a reason...

mitch_04
10-23-2013, 02:52 PM
I should point out a few things. First, my recent "theory" thread are largely due to the fact that I just got out of nasal surgery and I am basically bed-ridden for a few days. Next, the intakes I have been thinking of are mostly on inline engines, not the V8 style, although airflow is airflow.

Astro, I read a book by David Vizard that spoke quite a bit about airflow and some of what you are saying is "linking up" with what I've read. Although it's on 2 different parts of the engine, it's still having the same affect. It was a book about cylinder head porting (although it hardly touched on the porting) and design. He had a few interesting ways of measuring airflow in cylinder heads that could probably be translated to an intake manifold.

Something tells me I need to make more engineering friends...kind of tough in a town of 1300 people 2 hours from anywhere! Does anyone have any good links to websites that could help me out with understanding fluid dynamics and the like?

e129745
10-26-2013, 01:22 AM
Hey mitch 04, hope you recover soon and well. Not enjoyable to be forced down, nor does it feel good to be shut down...I so dislike communicating when not face to face, things can come across in ways not intended. I did not want you to feel belittled by my previous response, I want to share that there is a lot going on with this simple thing and that is exactly what makes it worthwhile to pursue the questions. I have enjoyed realizing and accepting how intricate the world is and frankly I'd be disappointed if it were not so.
I came across some stuff quite awhile ago that looked pretty interesting and on target, however it was primarily focused on head port design/modifications. It does address some points of interest though. Might be interesting for you and would help pass the time. The guys name was David Vizard and he wrote some technical articles that were found on GoFastNews.com. I think it originated in the UK. Been awhile but I did some searching and ended up finding several of these articles in various archives. I don't think I found all that he wrote for them, but I did save the ones I found and contemplated the info. as I was building my currant engine and it caught my attention.
Best wishes,
Patrick

mitch_04
10-26-2013, 06:54 AM
I have David Vizards book on Porting and Modifying Cylinder Heads. That guy is very obviously intelligent but, to me, his writing style left much to be desired. He was huge into the Mini Cooper engines, and his book is where I got my (limited) knowledge of how air flows. I really wish I could hang out with the guy, he seems to know so much but his book just seemed like he was defending his knowledge the entire time. I understand how hard it is to get your point across without coming of brash online. I struggle with that myself. There is just so many knowledgeable people on this forum, I try to get good discussions going to learn from.

e129745
10-26-2013, 09:26 AM
I have David Vizards book on Porting and Modifying Cylinder Heads. That guy is very obviously intelligent but, to me, his writing style left much to be desired. He was huge into the Mini Cooper engines, and his book is where I got my (limited) knowledge of how air flows. I really wish I could hang out with the guy, he seems to know so much but his book just seemed like he was defending his knowledge the entire time. I understand how hard it is to get your point across without coming of brash online. I struggle with that myself. There is just so many knowledgeable people on this forum, I try to get good discussions going to learn from.

Absolutely agree on all those points, Hope this takes off.

wizbang
10-29-2013, 06:45 PM
I’d be willing to toss in my two pennies on the issue.

While I would wholeheartedly agree that the true science of engine intake manifold design can get pretty overwhelming if considering all the factors that come into play. From my experience for most systems, while the number of all related factors are many, the number of strong/major factors tend to be more manageable. Most of the minor factors are typically not worth the effort to take into account or can not be accounted for in the design for practicality sake. For example, one factor that can not be practical take into account is the sonic pressure waves generated with each opening of the intake valve. If you wanted to take full advantage of the sonic pulses to assist with the induction into the cylinder at the valve port, you would want to use the 1st reflection pulse. Therefore you would want each cylinder’s intake runner length to the main plenum to be long enough for each negative pulse generated to reach the runner opening, reflect off the runner/plenum opening, thereby inverting to a positive pulse, returning to the intake valve and reach the intake port just as the valve closes. A rough calculation of this length to be tuned for the engine running at say 5,000 RPM and assuming the intake air temperature is 150F, speed of air roughly 1000 ft/sec, intake valve is open for 180 of crank rotation, the ideal runner length would be somewhere around 6.75 feet. This is obviously not very practical and doing so would likely be ignoring other factors that more strongly influence intake performance, namely frictional losses of the intake runner. To be practical, one might choose to move to higher order reflections. However, you are likely near the 6th order or higher before the timing is right and the pulse strength at these higher orders have decayed considerably. Bottom line, I would not fret over designing runner length to account for sonic pulse tuning.

Of the major factors, there are basic principles of fluid dynamics that will help with the design without a whole lot of effort. I would think some of these are:

1. Do not change air velocity sharply throughout the runners. Keep the air velocity consistent or increasing gradually towards the intake valve. This means maintaining a constant or gradually decreasing cross-sectional area for the runner.
2. The air in the intake runners will be in turbulent flow or more realistically, in plug flow. Therefore, gentle curves will not be detrimental or diminish air flow potential.
3. Eliminate sharp edges at runner openings to the main plenum to minimize entrance effects that increase frictional losses.
4. The shape of the runner is not likely as important as not changing the shape too quickly.

In my opinion, the greatest opportunity probably lies in understanding the effects on the dynamic pressure pulses as related to runner cross-sectional area and runner length. By dynamic pressure pulses I mean the pressure pulse created at the intake valve after the valve closes and the air in the intake track is still moving. I believe many people call this the ramming effect. The strength of which is determined by air velocity and runner length, for any given cam timing and engine speed. I believe some good designed experimentation with these factors could lend some good dividends.

For plenum design, I always felt designing the shape of the plenum for constant velocity as air is delivered to each runner would provide the best chance of equalizing the volumetric efficiency (VE) of each individual cylinder. To me, this is much more important on multi-port fuel injected engines then it ever did on carbureted or throttle body injected engines. This is because there is no practical way for most tuners to accurately adjust fueling between cylinders if the VE between cylinders is significantly large. This will result in some cylinders running rich and other running lean. If the VE varies too greatly between cylinders, then engine’s potential power production is reduced due to non-optimal fuel ratio in each cylinder and the chance of any lean running cylinder to suffer detonation and damage is greater.

I guess my 2 has expired.

MAP
11-12-2013, 02:55 PM
Greetings,

Very quickly:
1) A cylindrical runner offers minimal impediment to any swirl component of motion. A rectangular runner offers small resistance, while a runner stuffed with straws would abate all but laminar, one-dimensional flow. It would also flow with enormous drag...
2) The ram-tuning effect is mostly a consequence of non-linear acoustics and the "wave-steepening" phenomenon. This is the main reason why port and runner sizes are usually made as small as possible with respect to a given volume velocity target: we want to maximize linear velocity so as to delve further into acoustical non-linearity. The flip side of high linear velocity is a high Reynolds number, and therefore a high propensity toward turbulent flow. That's why smooth contours and low curvature along the flow path are beneficial.

Best,
MAP

iadr
11-24-2013, 11:31 AM
I was at a junkyard recently and noticed a tubular intake manifold. It was a 1.8l Toyota corolla.
While this doesn't answer any questions you have, I too had thought about "why not" for a tubular intake like this. :)

-Dale

Don't have to look that far. I made a manifold for an early 70's Volvo from the tubular Olds Quad4 HO (first gen W41, Quad 4 IIRC?).
Turned it upside down, build transitions to the volvo flange. Hasn't fired yet, but should boost low-mid range. Very Long runners.

Kenny
11-25-2013, 10:32 AM
You can reach your target with rectangular or round. I would recommend a good software such as the pipemax pro to get you in the ballpark. Plenum volume is very important as well. A couple of things to consider is 1) wave tuning, and whether you want to use 1st,3rd, or whatever, and 2) runner taper. If you do 1st wave tuning, you will end up with a shorter runner with less cross section that does not make as much peak power at the tuning point as it's counterpart, but will make more power above and below it's tuning range. So the next option is a longer runner that will need to have larger cross section due to restriction, again it will make more power at it's tuning peak, but will have less above and below the targeted tuning range. The same holds true with taper vs non taper runners. On some early testing, I found that a taper runner had a very small gain in the very top end, but my previous non-taper runner held almost the same peak power, but held it from 2500 rpm lower- I'm talking table top flat. The end result was that the straight runner made a far wider powerband= more average power= better performance on anything that did not have a cvt trans.

The suggested starting point would be 1-7/8 ID x 8.5" long runners......That should get you a good powerband from 6500-8500 with reasonable torque. There are quite a few tricks to getting power out of a 4v motor. Using Bosch ev12 injectors with a dual spray pattern will be worth about 2 hp per cyl because it doesn't go through the continuous cycle of wetting the septum.......

If you do a valve job on the intake side (regular 5 angle), use a 75 cutter to bring the 60 cut to .040-.060 wide. This will really clean up the seat to port transition if you aren't doing any porting, and still flows better if you are. DO NOT fall into the trap of people saying you don't need much lift on a 4v motor. I don't know how deeply you want to dive into this engine, but there are tons of tricks.....