Manifold design: Does boost mean you can get lazy?

[child9]

So I've heard a lot of people say in various communities that when you turbo a vehicle and are building the manifold for it, things like the collector smoothness and runner lengths don't matter as much as they do for an NA vehicle.

This contradicts what I believe, but I was curious if anyone on here strongly agrees with that and why.

I think that runner length, smoothness of merges and collectors, and other factors all mean much much more when you start adding boost. For example, friction and drag caused by wind resistance increases as velocity squared. I think this also applies to the gasses in the manifold. If you are doubling the amount of air molecules being pushed through the engine by running 1 bar of boost, any disturbances in the flow of these gasses through the manifold will be exponentially increased...I suspect as velocity squared also. The density of the fluid has been doubled and I believe the relationship there to be exponential, not linear.

Does anyone know any different? Thoughts?

 

[eflores]

You are correct. Friction creates heat. Restriction creates friction. I run about 2.75 bar in my pickup and at those pressures it is crucial to maintain a cool dense charge or......you're just blowing hot air. Make it wide make it smooth and keep the sharp bends to a minimum.

 

[Boostwerks.com]

Couple things to keep in mind...

In a perfect world, your correct. However, compromises must be made for packaging, reliability and cost concerns. Good luck fitting an equal length runner turbo manifold on the 4.0L motor in any jeep. not saying it's impossible, but the other concerns come into light (cost, runner strength, heat control, packaging of other components...ect).

I'm not simply defending my log manifold design either. It's just easy to "build" the perfect setup in terms of an idea, but when you actually start building it everything else you might not have thought of starts rearing it's ugly head.

I will say though that tuned equal length runners are in fact more important for N/A applications as the scavenging can play an integral role in making power, especially with lumpy cams. Equal length runner's are less important with a turbo application due to the fact that turbos work primarily off of heat and pressure. In race applications where turbos are very large and see lower back pressures, equal length runners do help make power. The individual cylinder pulses are more effeciently applied to the turbine wheel maximizing flow and reducing heat. However, this is not at all important to a street application.

It is good practice in any header or turbo manifold fabrication to die grind any rough edges and try to maximize the line of sight in the collector. I like to think of the header/manifold as simply an extension of the cylinder head, and thus any good porting practices should carry to the manifolds as well.

Some of my previous work...

 

[child9]

I was hoping you'd chime in Bryson, I know you have experience with this. I really hope you didn't feel the need to defend the log design you make for the 4.0, because I certainly wasn't attacking it. It's good, strong, and effective in both creating boost with quick spool and with cost...all things that are important to jeep freaks. I don't know if it's obvious, but the manifold I built for my 3.0L 2JZ is FAR from tuned, equal length runners or anything like that. I also have a small turbo and I know my motor is gonna choke at the higher rpms. I'm hoping to lower the shift points without damaging bearings...it's an experiment. I did port the SHIT out of my collectors though.

This only reinforces your previous and critical statement regarding real-world applications, and I couldn't agree more. It's one thing to design a good system, but it's a whole other thing to be given a bunch of restraints and limitations and STILL be able to build a good system. Emissions, climate controls, and dash light indicators are often skipped because of the difficulty in keeping them as well. Some of those are REQUIRED by law, so I find it part of the challenge, as the laws are part of the restraints we are given when we decide to re-engineer something.

Part of the reason for starting this thread is to discuss a lot of the factors that affect turbo-powered engines in the Jeep community, since I believe (and hope) that more and more turbo Jeep applications will be coming out, thanks to Bryson and others like us. There is a lot of misinformation that I've seen on various sites that I would hate to hear people repeating in this community, mostly just because I want to see sweet, well-built Jeeps. We gotta keep it crisp. Besides, the better you build it, the more you can beat the crap out of it. I also don't know everything so I value the input and experiences that I get to read here. Thanks everyone, and keep up the good work!

 

[5-90]

No, you can't. The guiding principles will change somewhat, but there are still guiding principles to be followed.

For instance, a tuned equal-length header is useful for a NA engine, when you're trying to maximise output in a given crankshaft speed range. There is an optimax primary tube diameter, an optimax collector diameter, an optimax long tube diameter, length guidelines, ...

For a turbocharged manifold, you're just fine with a large-diameter short log manifold - the more you can do to retain heat in the exhaust pulses (therefore energy,) the better. Then, you're better off with a large-diameter p[ipe after the turbo - the exhaust pulses are no longer coherent, and the optimax rules are largely out the window.

For a supercharged engine, you still get benefit from exhaust tuning - but intake "tuning" is basically trying to maintain laminar airflow with as few restrictions as possible (as for turbocharged engines) - it's just that you don't have a device in the exhaust train to screw with pulses there - just increased pulse density and slightly increased pulse duation due to the increased airflow through the engine.

You can't really get lazy - you just change the rules you need to follow.