The design of swirl and tumble into cylinder heads

[gijim]

I've been investigating how I can improve the heads on my Northstar (cadillac 4.6L 32 valve V8) engine. Obviously one of the first things that comes to mind is porting. But one of the engineers that worked on the northstar had something different to say about the subject:

The issue at hand is the fact that port fuel injected engines have a very stratified charge. All the fuel is injected for that cylinder event onto the back of the intake valve so that the heat of the valve can help vaporize it. When the valve opens the first bit of the charge rushing into the cylinder contains ALL the fuel for that event and the remaining 90 percent of the "charge" rushing in is nothing but air. Compare this to a carbureted engine that has a relatively homogenous charge since the carb is constantly adding fuel to the incoming air. All the charge in the intake manifold and ports consists of fuel and air. With the port fuel injected engines all the fuel is injected in one shot of the injector and only a small part of the charge going into the cylinder has all the fuel. This system requires extensive incylinder motion to mix the charge inside the cylinder as it is compressed before the spark plug ignites it. The intake ports are designed for swirl and tumble which, along with the squish areas in the chamber, thoroughly mix the charge so that it is no longer stratified. Random grinding and porting of the intake ports, while looking good on the flow bench, will often cause a power loss as it almost always takes away the swirl and tumble features built into the head making the engine very A/F sensitive. As an example, the production 2000 model year engine can run happily at full power with ratios as rich as 10:1. No problem. The Pace Car mod'd heads would rich misfire if run rucher than 12.5:1 indicating severely non-homogeneous charge levels. Not good. You cannot run the engine that lean for more than a few seconds without risking preignition and/or piston dome overtemp. It works OK for the 15 second pull on the speed shop dyno but would be sure death to the engine if you did a late night top speed run.

So obviously 9/10 shops that do headwork probably wouldn't know that having heads that flow more could still make less power than stock if the heads destroyed the swirl/tumble built into the head design. That makes me leary.

Then what really piqued my interest was a post on a forum about the 3.4 DOHC V6 found in GM W-bodies. Stock they're rated at 215hp, and their heads flow quite well for the power they make, 264cfm on the intakes at stock lift. (As compared to the 218cfm on the intake at stock lift for the northstar that makes 300hp)

A performance shop working with a wet flow bench ported these heads, which in combination with a rebuilt and balanced bottom end, created 260whp, or about 300 crank HP, in other words, 80 more than stock. Supposedly they did little to change the flow, but ONLY worked on the swirl and tumble. They obviously won't release any information about what they did, and I don't blame them. That's a monumental find.

Now, my question is how I can apply this information to essentially the same engine, with 2 extra cylinders. Swirl rpm, tumble rpm, and flow cfm can all be measured. So the question becomes, how do you determine what is optimal? Both of these engines were released early-mid 90's, so their development started in the 80s, when they probably didn't have the tools of today. Computer modeling of flow, etc.

So, any ideas?

I was browsing through an article with specs on 350 heads,

http://herningg.com/projects/350heads.html

and the swirl rates of 4500rpm, with a 4" bore would mean a 1000 G acceleration acting on the fuel suspension. Is this not huge? Would that not centrifuge the fuel to the bore wall in a hurry? I have no experience at all here, so any information is helpful. Thanks guys!!

 

[Greenlight]

Several prominent engine builders disagree with the need for swirl and/or tumble at engine speeds above 3-4k rpm.

They (Reher-Morrison, etc.)indicate that at high rpms the fuel/air mixture becomes highly turbulent due to the squish areas of the chamber and heads with zero swirl/tumble can/do outperform heads with substantial swirl/tumble.

They do, however, agree that swirl (on 2 valve engines) and tumble (on 4 valve engines) is very important at engine speeds in the 1.5-3k rpm range when piston velocity is slower.

Also, engines with good combustion chamber designs/dome shape can tolerate A/F ratios in the 13.5:1 range (BSFC numbers in the 0.36 range) without detonation concerns.

Larry Widmer (Energy Dynamics, aka Endyn) was one of the first engine builders to explore the concept of swirl/tumble in the early/mid 80s. Check his website for more information.

The "craze" lasted for almost two decades. It wasn't until recently when several serious engine builders conducted detailed studies and back to back dyno test were the theories of swirl/tumble in Hi-Perf. engines proven to be wrong.

I'll try and find the articles and post links to them.

 

[patprimmer]

Swirl and tumble have a nice ring to them. Unfortunately this adds credibility with reduced need for data.

There have been several previous long threads on this subject. A site search by you might gain more than trying to incite us to repeat ourselves.

If people want to work on my parts and won't explain what they are going to do and why, they never get to see my money.

Increased airflow in the heads increases power potential of the head. many other factors must be matched to it to yeild that potential, compression ratio, A/F ratio, spark timing, quench or squish, chamber shape, displacment, bore to stroke ratio, rod to stroke ratio, fuel quality, air temperature, sump design, fuel distribution, air flow distribution, surface area to displacment ratio, surface temperature of chamber, oil temperature, internal friction, etc etc etc

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[gijim]

They do, however, agree that swirl (on 2 valve engines) and tumble (on 4 valve engines) is very important at engine speeds in the 1.5-3k rpm range when piston velocity is slower.

That's exactly where the 3.4 DOHC picked up some significant torque. Their 4 valve design makes for a weak bottom end stock, but with the porting, it just made for a nearly flat torque curve. It was very surprising.

Now I also want to just separate myself from any "lean burn" crowd. I'm not interested in running 18:1 AFR at WOT. I was reading a bit on Larry Widmer's site, and one of the things he mentioned was the EGT being surprisingly low. I'm sure a lot of you have read Allen Cline's article on detonation and preignition:

http://www.streetrodstuff.com/Articles/Engine/Detonation/

Detonation will actually cause EGTs to drop. This behavior has fooled a lot of people because they will watch the EGT and think that it is in a low enough range to be safe, the only reason it is low is because the engine is detonating.

I can't trust Mr. Widmer's design because it's antiquated, and frankly, he spent no time at all on the dyno. However, that rather long quote in my original post is from Allen Cline, who was a design engineer for the Northstar engine. I trust him because he's run the Northstar at wide open throttle for HOURS at a time. They have to be run at 10:1 so the piston crowns don't melt, he was telling me how at 12.5:1 they wouldn't last 2 minutes at maximum power.

At any rate, Mr. Cline was saying that when the 2000 northstar's heads were ported for more flow (and the swirl and tumble characteristics were destroyed), that the engine made less power, and was on the verge of a rich misfire at the AFR needed to keep the pistons from melting. So obviously swirl and tumble are both important. I guess lacking any mathematical analysis, I'll just have to measure swirl and tumble rpm at different lifts for a stock head, and try to reproduce that in the ported heads... Nevermind that I wouldn't know the first place to start......