Cylinder head and exhaust flow

I thought maybe we could start a new thread about this subject that was not model or head type specific. We presently have Mark getting his QV heads ported, and Russ is trying to build a better exhaust system for a Mondial and he even has a thread over in the V-12 area on the same discussion.

I am presently working to create and build an air flow bench so I can watch and understand the airflow rates of a 2V 308. But the amount of material to read and understand can become quite absorbing. The more you read and learn, the more your eyes become glassed over. I didnt intend to make it any more complicated than I thought I needed, but once I started reading I found that to do it correctly you really need to know what your supposed to be watching. Air is wierd stuff.

For example, lets discuss exhaust flow and how it pertains to engine performance. When you begin, you assume its like flowing air, that just like the intake you need things smooth and well shaped. But start reading and you soon realise there are many other factors.

I have read some articles claiming that as the exhaust is leaving the chamber, a shock wave, approching, or possibly exceeding the speed of sound (1100 feet per second) precedes the main flow charge as it travels down the primary tube. This rapidly exiting shock wave "pulls" a vacuum that helps to draw the main gas flow from the chamber. There is a point during the overlap phase where pressure differentials between the exhaust port and intake port can exceed 6 psi. I dont know what that equates to in W.C. yet, but some articles I have read are eluding to the idea that the exhaust flow creates enough vacuum draw through the intake, that the downward speed of the piston is negligable to filling the cylinder. That the pressure differential created by the exhaust flow is what allows the cylinder to fill to a volume exceeding it capacity, actually "boosting" the engine.

Additionally, I have been reading that to find proper air flow, and to observe combustion chamber turbulence, some are flowing cylinder heads with pressure differentials exceeding 50 inches W.C.. That in some case its been seen that real life airspeeds in the intake port can cause seperation of the air flow and cause it to stall. There is discussion that fuel added to the airflow does not behave the same way as dry air, and that in some cases porting can cause a dramatic drop in flow once the heads are on the car, because the fuel in the air totally destroys the flow and causes the flow to stall.

For these reasons, I am questioning if cylinder head flow shouldnt be tested at super high flow rates, something more approaching real life WOT at full RPM, and if the flow shouldnt be tested at the exhaust port with a piston at the correct height of overlap in the cylinder and with the air entering the intake, carb and all, to mimic the true valve overlap phase. Secondly, I am trying to find a way to introduce a non flamable fluid mixture that would mimic fuel density and behavior as well as a vacuum source that wont be destroyed by it. I guess what I am questioning is if its possible to create enough flow at the exhaust port to reach realistic engine speeds.

 

So what your really telling me is that I should direct my energies into beer making? Oh wait. Probably cant make good beer if your drinking it a lot I suppose, eh?

 

I wanted to discuss Ferrari's 308 headers. Presently I am blessed with both a stock US set, as well as a euro set. The euro set being far better condition have taken to residence in the driver, and the crummy holy US set are resting up in the garage attic waiting for me to figure out what to do with them.

Both sets have what to me appear to be a problem, although the present phrase probably applies : that everything a reasonably intelligent person should intuitively believe to be right, will probably be totally wrong.

Each header, all four of them, have the prescribed 4 into 2 into 1 setup. What bothers me is that at the outlet of all these headers, the outlet is a straight section of pipe that extends several inches into the cone of the collector. So in my mind I am picturing this nice well behaved exhaust flowing out into the collector, and it being gently ramped down by the cone, but here is this stove pipe sticking into the cone screwing up that nice gentle flow.

And I cant tell you guys how many times I started to pick up tools to go after them to cut that pipe out and smooth it out inside. I havnt simply because I told myself I dont yet know why Ferrari did it that way. Was it done because the company making them was cheap and didnt care, or was there a philosphy behind it that we dont yet understand?

When I get to a point where I can start blowing some air, I want to first, simpy see what realigning the manifolds will do to the airflow. But when it comes to the header, I really dont yet know what to do. I could use the US pair and cut the inner pipe out after flowing it and see what it does, but it seems right now that I dont know what that would tell me.

Does anyone know what Ferrari did with other models of thier cars headers?

 

But regarding the headers themselves, are any other models stovepiped like these 308's ?

My thinking has me strongly questioning why Ferrari did any of these things. They didnt build some of the greatest racing engines by not having a clue how to build a performance engine. If anyone back in the early 1970's would have known how to make a cylinder head, one couldnt imagine Ferrari not being at least close to the top.

I understand what your saying about the port size, but why then didnt they correct it? Why didnt they just use the 208 heads, or whatever. I look through the parts from the Ferrari website and there are miriad part numbers and cam part numbers, so they obviously made many different castings. The Michelotto cars were in the 77 and up range, yet Ferrari kept putting the large port 2V heads on the 308 until the QV.

Is it more possible they wanted the cars detuned for some reason? Could they have upstaged the bigger cars from Ferrari had they been massaged better? Was it simply to pass emissions? Who knows. I just cant imagine these stovepiped header collectors are doing anything but robbing power. But I dont know enough yet.

 

I was reading some literature regarding CNC machined ported heads. Something along the line of it first needing to be created from a hand created port. The CNC can make absolutely precision cuts and make every port in a particular head exactly the same time and time again, although some ports are actually impossible for a CNC machine to get into and machine all surfaces. But at any rate, if the port they build the program off of isnt a well flowing port, you will just have a head with a bunch of equally exact ports that dont flow very good.

 

I still havnt gone out and bought a decent camera, and its not worth using the one I have, so no pics.

I finally crawled up in the garage and dug out the rear US header to look it over. I have to admit my ignorance as I never really looked it over very well before. But compared to the "euro" header that I have on the driver, its "appears" to be far better made. I had somewhere along the way taken comments people had made to concur the US header was crap and the euro header was superior. But I now believe the reverse to be true.


While the euro header has wrinkles in all the bends, the US header appears to be mandrel bent, but without those tell tale reliefs you see in some pipes. More like it was hydraulically formed? Looking inside at the collector end, the pipes fit tightly to the walls, and the collector flanges are deformed around the primary tubes for a tight fit, the euro's are just a sheet metal can welded around a pipe. The exit pipe is short, and fit snug to the walls of the cone, not at all like I had seen on the euro header. Again, the collector cone is actually deformed to fit around the exit pipe. The euro header had the exit pipe extending several inches inside the cone, and it was nowhere near the walls of the cone. Think of a hose crammed into the neck of a funnel, and standing up an inch or two inside the funnel bowl. Wouldnt flow very well would it?

On the rear header, from what I can measure, the primary tubes appear to be 34mm ID, the exit pipe is 46mm ID. Primaries vary in length from about 19 to 22 inches. I will have to remove the rear header off the car to get measurments from the euro header, but I would assume simular sizes. I havnt measured it yet, but I believe the front header uses the same 46 mm ID pipe for the two long secondaries. Guess I have to go up there and dig some more, lol.

 

With further rummaging I retrieved the front US header. Same construction, except the secondaries have that tell tale rounding in the inside bend from a mandrel, but still no wrinkles anywhere. The secondaries are approximately 47mm OD, exit pipe equal size to rear header. Because all the pipe (the rear header has a broken primary that allows me to measure wall thickness there as well as each end) appears to be 2mm wall thickness, I would assume the ID of the secondaries to be about 43mm.