16v head differences - port designs revisited

lol, blackberry "curve". I presume curve applies to the shape of the lens...and not the correct curvature either, optics are not the strong point of the unit shall we say!!

 

As for some of the other points...

There can indeed be a big difference depending on the bore adapter size used. If you look at the specs the other company use a 4 inch (100mm) bore adapter. When I swapped from an 82.5mm adapter to a 93mm (Ford Pinto) I gained 9cfm...but the head had'nt changed one bit.

I'm fairly confident our bench gives real world and consistent readings...going by the others shown maybe it's calibrated a bit low...which suits me...that means our heads actually flow more than they show.

A std 8v GTi head show's 71-73 cfm on average on our bench...when Dave Walker tested one many years agoon his Superflow 110, it made 75-76cfm. The industry std for a race Ford crossflow head is 103-105cfm...the last one I tested made 103. I've also tested many Ford Pinto heads...my own and others, all making between 108-118 cfm depending on valve size...Dave Walker used to claim 115 cfm from 1.75 inlets, with the best 1.8 inlet versions making 120cfm.

Head flow measured in any form is'nt the be all and end all, the same as dyno and rolling road testing only show numbers...they does'nt tell how an engine "delivers" it's power...some punch way above they're weight compaired to the numbers they show...but mine has certainly helped development.

I think the main point here is the ABF appears to flow better than the earlier heads...which I took to be the idea behind the thread...trying to find the possible differences between head types. The ABF looks to have a better/longer short side turn compaired to the KR's flatter short turn, but the 9A is similar to the ABF in that area.

I'll dig out the 9A I tested...not sure what type it is...but yes the type numbers should maybe be quoted as well where differences are possible.

Exhaust flow to come...but my money is on the KR making mid to high 70's cfm.

 

Yr forgiven...mine is'nt much better..

 

It would be good to use head casting numbers Jason as the head fitments overlap between codes.

KR mainly got 027 103 373 and 051 103 373

6A and 9A got 051 103 373 and 051 103 373 D

ABF got 051 103 373 D

When I spoke to CNC heads a few years ago they were unaware there were different castings!! You'd machine into the waterway of an inlet port on an 051 head if you programmed the mill for an 027.

 

Hmm...need we say more...tho in fairness there are variations on many heads...3 different types of Ford Pinto for example in the short turn area alone...thats one of the limitations of CNC porting.

Yep...I'll dig out the casting numbers..someone can put them in a chart or something perhaps...the KR is an 027 for sure...I may have another one somewhere...be good if that's an 051 then.

 

This has raised a few questions in my old noggin box. Concerning porting, in particular with the 'D' ABF and late 9A head, what are the recommended improvements?

Inlet port matching
Knife-edging the center in the inlet ports
Leave or remove the hump between the ports, or put a sharper edge on it?
Exhaust port match.
Exhaust port knife edge between the ports?
What about that point where the fuel injector inlet meets the center of the ports?
Mirror finish or rough inlets?
Mirror finish or rough exhaust ports?
Mirror finish or rough combustion chamber?

 

Manifolds generally match well. I've seen mismatched dth bodies help make 190hp on a tame engine. The same mismatch on the flowbench shows no flow change.

Not worth worrying about...losses and gains seen through the flow curve when tried during testing..1-2cfm best gain...similar losses.

Hump?...as in the port split as above?

Exhaust port outlet is more than big enough...exhaust manifold ports should ideally be bigger so there's a step into the exhaust port to discourage reverse flow.

If the exhaust ports are sized correctly during porting it'll result in a natural thin edge...but no need to try and specificly aim for it.

The injector cutouts?...obviously needs to be left when the std plenum / injectors are still being used, but a decent designed inlet manifold should block it off for dcoe carb or t/b fitment.

No need for any polished or mirror finish...never EVER on the inlet ports and largely pointless on chambers and ex ports...in fact some carbon build up can help with a thermal barrier.

 

No idea why there's an angry face at the top of the above post...bloody phone!

 

We always look at the 051 103 373 D for ABF / 9A ….they originate from one drawing,
Model makers are maybe a few more in the process and for sure castings are made by more than one Foundry to guarantee a supply chain that keep working at all times
( this apply to all car parts)
I spotted casted in numbers where the water flange fits, 22 , NNE, or something else there….must have to do with the casting mould they use, there might be little differences ….?

A flow bench is only a tool that tells you basicly as much as measuring a camshaft with a Verniers-calliper
Lets say it delivers a number that does not guarantee that the head with highest number will make the most power…contrary the hype created…but it will perform maybe 90 -95 % ?
It does not tell something about Tumble and Swirl, gas reversion, valve sealing, guides, detailing to reduce detonation and so on ….and here comes experience from the head porter/ engine builder into play… you could even take the piston used in the low to medium valve lift range into account and what cam profile/ timing is used … its a system and always a compromise ....

 

Combustion cycle modelling

Here's one tool available to well funded researchers - -

www.youtube.com/watch?v=WpMifop7fyu

(if it fails to materialise try entering the address on your favourite browser - - )

 

can't watch that one..

 

[YOUTUBE]WpMifop7fyu[/YOUTUBE]

 

Broken for me too.

 

HPR and LeftcoastTigger have touched on very import aspects of port design and that is the effect on charge motion in cylinder.
This is why I asked the question earlier.

I do agree, like all non intrusive measurement tools, the flow bench numbers used by the aftermarket to assess cylinder head porting work can vary and in reality are only stating "how much port flow" there is in a port at valve lifts from a given baseline.
I would expect if testers were to provide aftermarket flow bench data all the standard VW 16v heads i.e. 027103373E, 051103373 and 051103373D, even though the numerical values would vary, based different measurement approaches, the graphical profiles at different lifts should follow a pattern.

When a cylinder head is designed the ports will be developed for the best charge motion in cylinder.
This process, using a flowbench in association with some form of particulate analyser, studies the behaviour of the charge motion at different gas velocities.
This data can be fed in CAE analysis to further tweak the port, which can be validated once again on the flow bench/ particle analyser.
You want to best gas velocity that ultimately influences the most uniform mixture distribution in the cylinder just before combustion to achieve optimised combustion - all things being equal of course.

Some very experienced tuners/head porters can achieve the above on certain modified engines without the tools.
Most of the times engines that have benefited from such expertise do have other dress-kit enhancements to minimise pumping loss in and out of the system i.e. ITBs, tuned exhaust systems, dry sumps, gear reduction FEAD for alt and water pump, low friction pistons etc.

I have used another illustration from Brain's cutways to further describe what happens to charge motion in the EA827 16v engine. This looks at the port shape from a side profile perpendicular to the cam drive gear rather than a "plan" view of the roof and floor.
The labelled up illustration attempts to represent a segment in time for an intake stroke.




OEMs know from CAE analysis observing the flow coefficient to tumble correlations, as the piston travels up the bore and the inlet valve closes, the charge motion vortex, in cylinder speeds up and compresses from a round to elliptical. On the outside of this vortex are the yet-to-be-mixed fuel droplets. As charge tumbles these droplets get distributed within the centre of vortex forming a homogeneous mixture.
Your aim is when the piston approaches the compression stroke the mixture formation does not break up thus going on to have combustion that evenly consumes the fuel air mix.
Too fast port velocity = breaking up of the vortex mixture charge prior to combustion = uneven combustion, hot spots, knock
Too slow port velocity = uneven mix due to weak vortex and poor combustion, high fuel consumption.

The difference in the valves, as shown by Neal in particular the former 051103373D units from the ABF/ADL//ACE/Later9A engine, would to reduce the vortices between the stem and the valve head and improve the flow and direction of the charge into the cylinder.
This would be most definitely the as a result of CAE simulation models to improve the later head casting.
The inner port radius shape is designed, modelled and finally cast to keep majority of charge flow from the shrouded area of the valve head which would have negative consequences for the in-cylinder charge motion vortex.
The CNC area that is milled in the port casting for seat placement is usually an area that could be mismatched slighty to the modelling work. However when the other segments of the OEM AIS are added to the engine, the effect of this mismatch is negated by the reduced inlet velocity from the full intake system, meaning the engine can achieve the auto and mini map power points for homologation /cert meaning this is less of a concern to the process.
NB This may not be the case if the engine is used for competitive purposes though.

The link by LeftcoasTigger, does not show if clicked but you can view the animation by doing as he said.

PS I am not a combustion analyst but need to understand the above to carry out some of the tasks I am delegated in my day role. So if anyone reads this who has more in depth knowledge about this process please feel free to correct me and chime in.

 

No, i did what he said but still cannot view it

 

Found it using Google. It seems to be case sensitive (Google won't be) and has a full stop =>youtu.be

Try this:

http://youtu.be/WpMifop7fYU

[YOUTUBE]WpMifop7fYU[/YOUTUBE]

 

Thanks Chris

 

The animation is good but its not very relevant to our engines as that engine is using HCCI combustion. That's basically compression ignition using direct injection in a gasoline engine and the flow structure is very different to that used in normal port injection, spark ignition engines.

 

Nice, cheers for the help.

By the 'hump' I mean the ridge on the floor of the port. It can be seen fairly clearly on the inlet pic I posted near the start. I was wondering if it helps swirl or if it should be removed.

From what I've read it's difficult to nail down what is best for them when porting but I think for mine a general clean up and removal of the casting marks and a slight smoothing and widening of the inlet ports would be what I need.

 

Yr welcome...

Ah at the very leading edge...just ignore it...no issues with that.

If your only going to do minimal work, a small increase in throat diameter with due regard to having a 3 angle seat cut...have the valves back-cut...and leave it there. Smoothing anything wont gain anything unless a decent amount of material is removed...which take you into proper porting.