A knotty ported exhaust problem. 4

[Torquey]

Dear Gentleman,
Seasons greetings! I sent a pair of 4.6L DOHC Ford Cobra heads to be CNC ported. When I got the heads back, to my dismay, the exhaust ports where ported all the way out to the gasket edges. Doesn't this violate a cardinal rule to gasket-match the exhaust ports? Every book I've ever read on porting says "don't do it". Are there any exceptions? Just to give you an idea of size, they went from ~25 X 48mm ovals to 31 X 57mm Texas sized rectangles. Given that this engine is well known for it's problems with reversion, what are my options for fixing this? Will it help to spread out the cam's lobe centers? Should I have a bead welded back into the port to restore a reversion trap? Can a reversion trap be welded into the exhaust manifold? I've heard of fitting a larger diameter manifold to the head but a 57mm (2.25") I.D. primary exhaust pipe would be ludicrous, no? As always, I greatly appreciate any suggestions.
Walter Carter, Palm Springs, California

 

[patprimmer]

I don't quite understand why a particular engine will be more prone than any other once it is modified as the causes are probably changed during modification.

You can reduce reversion by widening the lobe centres, particularly by retarding the inlet so it does not start to open while the piston is still moving up with any speed.

A hot trick (meaning it was probably a waste of time) bac in the 80s was to make an exhaust flange that bolted to the head. It had short stubs that accurately matched the port.

A second flange bolted over that one with the short rad touching the stubs but the log rad being a few mm bigger than the stub so reversion was trapped by it.



The flange should be the shape of the port and the ends of the pipe should be formed to match the flange.

If you drop he idea of ant reversion dams and focus on things that really work, I would make flange that was say 0.5mm below the floor of the port and 1 to 1.5 mm bigger on the roof and sides, The 0.5mm is just so that if you get a bit of misalignment when fitting, they still don't overhang the port. If you can fit dowels to ensure good alignment, you can cut 0.5mm of the above dimensions.

The port is about 17sqcm allowing a bit of a loss at the corner radii

A 2" pipe is about 18sqcm allowing for 1mm wall thickness.

1.875 to 2" primaries seems about right for a high performance twin cam 4.6 litre V8.





Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers for professional engineers

 

[RossABQ]

Were they trying to match a set of Ford Motorsport headers that have this shape of port at the flange?

 

[Torquey]

PatPrimer: So basically all is needed is a 1mm ledge on the low velocity edges. What is the driving theory here? Is this ledge just to get the reverting gasses to tumble? And little is gained by a larger ledge?

RossABQ: If you are familiar with the OEM Ford 3-piece exhaust gasket set, they matched those gaskets exactly (they're more squarish than oval).

 

[dicer]

Opening up that part of the port is what porting is all about. They are opening it up to flow a certain CFM.
Reversion? Well that is what a tuned exhaust system is all about, and of course for only a narrow rpm band, unless you use a dynamicly adjusted exhaust system.
I think what Pat is saying the header flange should be larger than the port then the head area of the port would kinda stick out.
It would be counter productive to CNC port your ex ports and leave material at the exit point.

 

[patprimmer]

I would have thought the top, rof or outside radii was the high velocity side.

I think I also said it was an old "trick" so I expected it did not work. The theory is it works like a poor ratchet with more resistance one way than the other. Good for solids, poor for fluids,

The not mentioned down side is that to get the ledge you need a bigger pipe and therefore reduce velocity and reduce velocity and pulse tuning benefit on scavenging. Reduced velocity is good to point, then counter productive after another point.

Inlet, cam timing and exhaust tuning are all about compromises. to get more at the top you need to lose some off the bottom.

To get a wider usable band you generally need to reduce the intensity of the peak.

You have not outlined your objectives and priorities and you have misinterpreted or put extra spin on advice. YOU need to get this clear in YOUR head then spell it out clearly if you want good advice.

I have not seen your specific head, but I have seen thousands of heads over the years.

Yours sound very much like a typical rectangular port where the 4 walls are straight an at right angles to each other, but have somewhere between about 1/4 to 3/8" rad in the corners. This is really getting down to kindergarten level.




Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers for professional engineers

 

[CNCHP]

With both a dry bench and wetflow technology, we've found it's not productive to match the gasket/port, instead we will use a gasket & header opening that is about 0.050" to 0.080" larger, to create a step. On a wetflow test bench, a larger step on the long side has less effect than the one on the short side: if you must bias the alignment; otherwise center up. We strongly suggest using a means to align the manifold during installation. Many heads we do already have an exh port that is nearly too large, so we take care not to CNC them just for the sake of cosmetics, or they end up too large. We pay close attention to exh gas velocity, and apply ceramic/metallic thermal barrier coating to improve the flow, especially at lower cam lifts, the cermet coating also helps greatly with cyl head heat management. And of course an exh system tuned for the particular motor is very important. Patprimmer is correct, in that one should have a firm "plan" about what will be used upstream and down stream of the head's ports before porting. If you use a "canned" CNC program, then expect to figure out how to deal with reverb as an after thought. Or use a CNC porting shop that is willing to customize the port to fit your plans

 

[Torquey]

CNCHeadporter: Yes, this is what I understand to be true, although I had no idea such a small overlap would be effective. I guess I will fix this by hogging out the ports on a cast iron manifold.

PatPrimer: The bone stock 4-valve Cobra motor is a flat out pig from a dead start. So much so that you must juice the gas to keep it from stalling. From 3700-6800 rpm, the car is wonderful. I really don't mind that the car is a low end pig but I keep that in mind as I cannot afford to take very much more from the low end. My pide-piper is this fellow who plays with 4-valve Honda bikes, which look, for all the world, exactly like a 4-valve Cobra (at least when looking down the intake ports). This fellow has empirical evidence that you get more power across the spectrum by building plastic nozzles just behind the valve guides. He has found that the desired nozzle size is a certain percentage smaller than the area of the inlet valves. However, according to this fellow and many others, it is unwise to gasket-match the exhaust. Although most modifiers will shorten the intake runners, I have purposely left them long in hopes that they will help with the low-end sluggishness problem. In addition, I have raised the compression to around 10:1 and plan to stave the ensuing detonation with water injection. Actually, I have a lot in common with the Honda bike racers... they like smooth progressive power at higher rpm and so do I. Those are my plans/goals. I am no quarter mile guy.

 

[patprimmer]

I agree with all CNC says, although I do not attribute the gains to anti reversion.

A step facing away from flow direction on the inside rad tends to tumble the air and slow it. This helps it follow the curve.

His examples to back up his points are flow bench data. Flow bench data is steady state so reversion is not possible. Also wet flow is hardly relevant to exhaust side.

His point about velocity is very important and minimising changes of velocity within the port and maintaining maximum velocity at the required airflow and matching that to the inlet side airflow and velocity then having appropriate cam timing for the speed range required will give the broadest power band.

The ways to reduce reversion are to maximise velocity and keep cam duration short and specially at overlap.

If a normally aspirated engine develops good power to 7000 rpm you must expect it to be flat below 3500 to 4000 unless it has fancy valves in the inlet to adjust the manifold to different speeds. That is why the factories are spending a lot on these style manifolds and Vtec on more engines. They would not go to that much effort if there was an easier answer.

Even though off the power band, a modern EFI engine should pull smoothly if not strongly from idle.

If you have to juice the gas (whatever that means on an EFI engine with no accelerator pump) I think you have a clutch control problem and/or are over geared for the car weight.



Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers for professional engineers

 

[CNCHP]

Thank you for pointing out the limitations of "steady state" flow bench not being "real world" test - wish more folks would realize this... IMhO, its just a tool for comparative measuring, and even then one must be very careful when relating "improved" results to what actually works well on a running engine. Many of the things we once thought was "fast" according to a dry bench have been proven counter-productive when tested on a wet bench...or dyno While at first glance, it would seem a wet flow bench is not relevant for exh flow - however when used to flow backward, thru a clear plastic copy of the exh header, cyl, and even inlet manifold runner insome cases, with the ultraviolet lighting, it is very helpful to actually be able to see the air's flow characteristics. It can be quite enlightening and often dispells what was once thought to be "fast" ... Its still not "real world, but its a very useful tool...more so than a dry bench.
In addition to the same cam "tricks" Pat recommends, we will often use different lobe profiles for each cylinder, depending on that particular cylinder's manifold/port length and reverb effects...more or less treating a V8, for example, as eight single cylinder engines, with each "cam" being different, depending on the inlet and exh path parameters for that cyl. On multi-valve engine we will use different exh valve seat angles to reduce reverb effect in the combustion chamber and intake ports/manifold (this is something the wetflow bench let us "stumble upon"). Best regards & happy New Year!

 

[patprimmer]

Thanks for the explanation of the wet flowing the exhaust and the UV light trick. I take it you use a fluoro dye to react to the UV.

I have seen pics of a running V8 with carburettor and a clear plastic inlet manifold. I was astounded by the maelstrom of fuel inside the manifold. It looked like surf breaking over a reef in a wild storm.

Thee things are simply tools and best results are obtained when the strengths and limitations are clearly understood.

I can't imagine the development cost of individual profiles for ever cylinder in a V8. I don't suppose you use a splined cam shaft with individual lobes assembled on the spline for easy changes during development.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers for professional engineers

 

[marcusaurailius]

Tourqey, what you might be experiancing is a phenomonon!! called ...stand off,,, .
the fuel is not going into the engine due to the air/fuel hitting ...things ... in its path,.
this could be your problem when you put your foot down,
big W,O,T, but no flow in to engine,.
try some trumpetss, or a baffle plate,.
if you running engine magement, the you may not have a big enough ...fuel enrichment... for the snap throttle opennings,
this is very pronounced at low revs ,wide open throttle[wot] but not so much so at higer revs
hope this is some good to you
Marcus