Pretty much, "How does one create his 'own' manifold?"

Today was the first day I stepped into Tig Welding. I'm still in the process of getting better at becoming a fabricator. I have made a thread about "how to become a fabricator?" before, and I thank all the guys on Honda-Tech for leading me to the positive direction.

Now I don't know much about AutoCad nor have I ever taken it before, but in the Program Requirements table it's showing me a Industrial Design/CAD class and Pipe Welding/Fabrication. Both classes which i was unaware of, since I'm only taking a tradeskill(GTAW).

Now I know the AutoCad class would benefit towards making manifolds would it not ? or alteast learning ... since you measure like fluid dynamics, cfm, etc etc. Correct me if I'm wrong please, I'm still a newbie to fabrication. I don't mind getting correct answers either or being flamed, go right at it. I'm planning to learn the easier way or hard way.

To make a long story short, I have no knowledge of Cad all I heard it's a good use computer-aided based designing tool made for not just automotive but a lot of different other things. My primary objective is to use it for automotive.

My question is what other classes would benefit towards making manifolds ?

My goal for this summer is to make my own ramhorn manifold, and I will try to practice as much as I can on stainless steel.

 

your thinking into this with GREAT detail...i myself am a mechanical engineer, where i use AutoCad, Pro-E, Inventor, and Solid Works on a daily basis. there are many more powerful programs than AutoCad, but for the basics that you are looking into, it can be a great tool for you. takes some time to get to know it well enough (i have been using it for over 10 years and am still learning small things here and there)

you need to learn more about basic welding and how to obtain proper penetration of the metal. all this technical stuff you are trying to learn can be beneficial, but not necessary to make a manifold. just find a class, or even somebody to teach you a) how to properly weld, using different materials and thicknesses, and b) how to cut/prepare metal for fabrication. you need to learn the properties of the metal and know what to use and when...

for all of the fluid dynamics you are talking about, if you really want to think about it, you have your material, that is at X thickness and Y ID, you are trying to push Z gas through it. unlikely that you are building something that is going to need that kind of data...most every manifold built on here is out of 304SS 1.5 ID Sch 10 material. there are thousands of ways you can route the pipes, and most every way, so long as the bends are smooth, will flow practically the same

 

Make sure your manifold fits and that it doesn't leak and you will be fine. No need to draft it.

Something I heard from Markuu when I was inquiring about manifolds a while back was hp goals as they relate to tube size.

On my 4 cylinder he asked if I was planning on exceeding 400hp. (100hp per runner in flow)

If below that use 1.25"shed 10
If above that use 1.5" she'd 10 for runners.

Not much to do besides fit some pipe in a space.

Equal length doesn't really do anything beneficial on a turbo manifold so don't waste your time.

 

AutoCAD isn't going to do any fluid simulations for you by itself. Programs like AutoCAD and Solidworks (at least just the base modules) only do 3D modeling. That means that they are great for designing a physical device/system, but any interpretation of how well it will work will have to come from elsewhere. With Solidworks and AutoCAD there are add on modules that will allow you to perform finite element analyses, which basically just model the physical stresses that a part will see. Both vendors also have CFD modules (Computational Fluid Dynamics) as well, which will do the flow modeling that you inquired about. There are also (usually much better, though terribly expensive) standalone CFD apps out there like from CD-Adapco. Keep in mind that these programs and add on modules are usually priced in the thousands, if not tens-of-thousands, of dollars.

 

nice to know about the equal lenght not really needed good input there

 

thats not necessarily true. especially on twin scroll manifolds where exhaust scavenging and exhaust gas pulses come into play.

the whole idea of twin scroll systems is to seperate the cylinders based on the firing order to keep the pulses from interfering with one another. for example if an engines firing order were 1-3-2-4, cylinders 1 and 3 would be paired and the 2 and 4 would be paired together to exit out of one side. so two cylinders are paired to exit each side.

now if we have an equal length manifold where all runners are equal and it is twin scroll in design, cylinder 1's exhaust gases will reach the turbine housing before cylinder 3's will. now if the two are paired together to merge right before the housing that would greatly enhance the flow since if we had an unequal length manifold where cylinder 1's runner was longer, and still firing first the gasses from cylinders 1 and 3 may reach the same point at the same time. so obviously it would make a restriction at the point where they merge since there would be twice as much gas where they merge instead of an equal length where one cylinders ex gasses reaches the merge aka collector before the others does.

the same idea applies for regular turbo setups, it just may not play as big of a role.

this is a good read explaining the idea.
http://www.modified.com/tech/modp-09...n/viewall.html

 

If you want to calculate all there is to know about manifolds I hope you are REALLY good at calculus.

 

calculate all you want, but most of the time, manifold design will be dictated by space constraints and placement. those are the first factors you need to consider, THEN start thinking about making the most efficient design. things to focus on: least amount of bends, try to avoid sharp bends, WG routing as smooth and 'in stream' as possible and so forth.

 

This manifold worked JUST as well as any others I have used.


No additional lag, noone cares about equal length besides some kids that read it online.

Scavenging doesnt apply to a turbo manifold because its a pressurized system.

 

very nice

 

that was a custom lovefab manifold that i welded a vband to for my setup. cant take credit for it all.

 

Thanks a lot for your input on fabricating manifolds. I am actually planning to fab up my own manifold coming up this summer.

As for this Spring: Wanted to start off by constructing a stainless steel downpipe for my 5bolt sc34 that I have. "that-guy" mentioned something about a 304SS material, I couldn't find any information on it but there are different other random 304 stainless steel materials I pulled up on the web. I couldn't happen to find anything 304SS related, is this like the only "recommended" material that's up for grabs when it comes to fabricating an actual durable "good-looking" manifold ?

 

The most common metal used for building manifolds is SCH10 and SCH40 1.5" "weld-els". They are 304 stainless, and are actually weldable pipe fittings (like something you'd use in industrial plumbing). They're thick, strong, and fairly cheap. (I think Acestainless sells them for like $4 a piece)

Thin wall tubing is also usable, you have 304, 316, and the expensive 321, but it is more expensive and you have to know what you're doing with your welds. Any imperfections or non-full penetration joints will likely crack later on.

I don't know all of the ins and outs of each stainless grade, you could easily look that up online somewhere and get a full written out detail of each though, or I'm sure someone may chime in here too.

 

Proof?

How is an N/A setup NOT pressurized? Are you saying that the manifold is in complete vacuum?

Regardless of that answer, why does wave theory become false in a boost application?

 

An N/A system has minimal pressure at the port outlet from the head and less than 1psi assuming you have a cat converter installed. A boosted setup will have significant pressure at the port outlet and within the manifold (higher pre turbo exhaust pressure compared to the boost pressures you are trying to build).

Wave theory would still apply, but it would not have the same effect at all.

Think of wave theory in different material density (which this is, but in gas form). Assume temperature has no effect on the discussion to follow to keep the theory of the systems being compared similar. A wave transfers much faster in a more dense material and the waveform becomes much smaller and tighter in comparison to the large wavey waveforms of that in a less dense material or a material under less pressure. this means you have less effect on your turbo manifold design from the wave form in comparison to a non-turbo system.

Also my previous comments are based on real world experience, knowledge of others and do not rely on science. Science is good to predict but real world applications will generally vary from expected outcomes.

Cheers,
Matt

 

So an N/A setup still has pressure. For this discussion the value is moot, unless you would like to bring it up.

c = (E / ρ)1/2

The above equation is the speed of sound. Rho equals density, E is bulk modulus of elasticity. When the density becomes larger, the speed of sound slows down.

If the exhaust gas is treated as ideal, then: c = (k R T)1/2


Where R is the gas constant and T is temperature (which are the same in our examples). So the speed of sound is the same in the boosted setup and the N/A setup. So that means that E, has changed with the pressure of the gas.

Then it registers that this is an isentropic process so, E = k*p (k = specific heat ratio).

If the setup has changed and the resultant output from the engine is 'same', what was it from? It had an equal length manifold before and now it doesnt. It makes the same power though. This doesnt prove X was the reason why it makes the same(or essentially the unequal length didnt make more power per $), it just says Y could have been the factor.

Im not speaking of gas dynamics with inertia, Im speaking of sound waves through these 'different' mediums.

If the length the sound wave travels in the 'equal' length manifold is not built directly for an rpm where the waves would be useful [as in the pressure wave return to the valve face is not in the lower values (~5 or less?)], then its not going to affect in a manner some might think it should.

 

having a bit of computer know how can make things a little easier too
Virtual Rapid Header Design

 

For every manifold, the features (merge collector angle and length, equal length runners, minimal bends. etc...) will only be realized based on power level and turbo sizing, and the characteristics of the engine.

Equal length does matter for setups that are running a big cams, and boost pressure / exhaust pressure ratio is higher (1:0.9 or less). This is also directly related to engine RPM; obviously, to sustain good power at high RPM's will require big cams and good velocity at the intake/exhaust ports. In this scenario, your typical exhaust manifold is now behaving very similar to an N/A header, and scavenging takes effect. That's why our modern "efficient" turbo manifolds are looking more and more like an N/A header nowadays.