Peugeot 8V cylinder to combustion chamber matching.

[Christopher_205Rallye]

Hi Guy,

Would be really good to get your opinion on this.. :D

How important is it that the diameter of the combustion chamber matches that of the cylinder? How does this affect the swirl of the incoming fuel? I if I remember rightly, I have seen that you have matched these exactly in some of you engine builds.

One good reason I could think of , is that it could be that it helps with swirl to have this exact match. But then if left mismatched would this edge cause a region for preignition (maybe burring required)


Perhaps this only concerns combustion chambers that are round and not ice cream shaped for example:



The reason I ask is I am intending to use a cylinder head that has a 75mm diameter of the combustion chamber and a block that has 78.5 diameter of the cylinder. It is possible I can substitute some CR for a match if the net gain might be greater. Also it might give some opportunity to deshroud the outer side of the valves.

You can see the inlet valve is quite shrouded in the picture. And the exhaust for that matter. Some have said that deshrouding the exhaust CAN result in a loss in power also?I read this in a book :?

[Guy Croft]

Hello there Chris.

You wrote: How important is it that the diameter of the combustion chamber matches that of the cylinder? How does this affect the swirl of the incoming fuel? If I remember rightly, I have seen that you have matched these exactly in some of you engine builds.

Well, this is slightly hard to answer straightforwardly as combustion chamber regions are not usually truly circular far less hemispherical but irrespective there must be - above all - a circular region around the chamber that allows the fire ring of the gasket to clamp down (across its whole section) against the cylinder. That feature alone determines how much bigger the combustion chamber can be than the cylinder.

Now, referring to your query, how much smaller the chamber can or should be inside the fire ring, so to speak? Well, the more flat area of combustion chamber in close proximity to the piston at tdc - so the more squish there will be. (See GC V/W - Burn Rate). Often if you mill a head right down the squish bands (regions) get much bigger. Sure, squish improves in-cylinder fuel mixing by vortex generation but is much less important on high speed engines than road cars. It's one of those enhances part-throttle performance things.

Squish helps economy and inhibits detonation. By promoting superior mixing of air/gasoline it gives more complete burn and faster burn rate - no more than that. Too much squish - in other words - too much flat region in the chamber adjacent the piston crown at tdc - can definitely be a bad thing. And squish bands can intrude on symmetric geometry of a chamber and by this alone have an adverse effect on progressive development of the flame front. This is true of your head and many others. Squish is definitely a characteristic that has to be finely tuned. I reduce squish bands but never make them bigger, never mill heads right down for this reason. Mainstream engine manufacturers - in my view - know a lot more about how big the bands need to be than I do.

But if your inlet charge temp is low and the fuel octane is high enough and plugs, fuel/ignition are right, you don't go creating squish bands for the sake if it - because your engine is not going to detonate anyway. Squish is far less vital on 16v than 8v because of the superior charge mixing and exhaust evacuation that takes place with four small valves compared with 2 big ones, and the more compact nature of the chamber.

As for swirl of the incoming fuel, swirl is a method, like squish, used to promote controlled levels of turbulence and mixing, and is not much used on SI engines. (used a lot on older diesels CI engines - often induced by valve masking, but now pretty well overtaken by tumble, perfected by Ricardo years ago, they got a Queen's Award for it). You still see valve offsets to give tangential swirl like yours but I haven't seen an SI with specially-shaped swirl-inducing ports for years, it was always unpopular because of the inlet pumping loss it creates in the head. To a great extent with inline valves you're get get tangential swirl as a by-product.
I have studied your photo and the squish bands visible at lower left and upper right and as an ancillary function create quite a compact 'figure of 8' chamber. You are putting a head designed for a small cylinder (75mm?) on a cylinder of 78.5mm dia right? That's only an intrusion of 1.5mm or so on radius, tiny. That would not bother. You anticipate gains from higher CR and I imagine that yes, the torque will be higher. And though you're going to have 10% or so more by ratio squish area the net amount of squish region is not huge and I don't think it will be an issue.
I would maybe not alter the existing bands or grind out the adjacent overlapping regions of the head. But I am bit lazy with things that I think don't matter, some people love to go mad with things like that but are never quite sure to what they can really atttribute the final power increase.

This discussion would be easier to understand if you could offer a photo of the head you're using with the big bore gasket laid on it, but at this time I think won't shroud either valve. As for exhaust valve working better when it's shrouded, I don't understand the reasoning for that. The power loss from any 'deshrouding' of same - if there even is any shrouding here (which I cannot see) may in fact have come from non-associated cause induced by the 'deshrouding mod itself.

Be cautious of 'do this - do that' advice unless back-to-back flowbench or dyno figures are cited and the op displayed in detail pictures.

GC

Some confusing wording re swirl/ports edited by me 01.07.06

[sumplug]

Tell me Guy, what tool do you use to remove the "metal"?

AC.

[Guy Croft]

ATA, Morrison and Garryson carbide burrs and bands, dozens of shapes and sizes, Metabo die grinder and flexi shaft. Some stuff, the really big jobs, I mill out.

GC

[James Bowen]

Guy, could you expand on why too much squish is a bad thing?

I appreciate that you've already noted that it doesn't really help race engines, as opposed to road engines. And the unsymmetric nature, affecting flame progression.

Does it mean that the expanding charge is less able to "push" on that part of the piston crown?

Another question that's related.... Christopher 205rallye's head picture shows a radius on the step down from the squish band region, to the chamber. Something that on the Fiat SOHC head is usually a sharp edge. Is there any merit in producing a radius like that?

Many thanks, James

[Wallace]

I remember something Dave Vizard wrote once - it's possible to have "dead" areas right in the corners of the combustion chamber (be it the head or piston crowns) where the fuel/air won't burn ! So persumably a nicely radiused squish area is better . . .

The dished area on the pistons I've had made up reflect this idea with a 45 degree wall to it - but it probably won't make any real difference in the real world !!!

I'll add a photo when I get a chance . .

And a really cheap way of making up a polishing tool is to wrap insulation tape around a 1/8" silver steel rod - then use "Bostik" to glue on small piece of sanding sheet (use a jubilee clip to hold it all in place). It sets in about 10 minutes - and you can make them in any size you like - up to about 1.5cm across - after that, they get a bit wobbly !

[Guy Croft]

Hi

well, I will try and answer both at the same time and see how we get on

Each engine is different and the impact of radially inward squish motion to induce turbulence is not well understood even at industry level. In this case, referring to Chris' photo what we have for a start is a swirl inlet port - one that causes the mixture to flow tangentially into the cylinder; the mixing this promotes is one good reason for not making the squish bands any bigger - this is by way of saying I doubt that they need to be.

The surface area inside the combustion chamber undergoes convective heat transfer throughout the compression stroke and during firing. The higher velocity of a gas the higher the heat transfer to its surroundings because the heat transfer depends to an extent on the Reynold number. The bigger the area and the more turbulence from squish and swirl - the higher the heat (energy) loss. This is a good argument for never making squish bands any bigger than they come from the factory, especially remembering that turbulence is less necessary on an engine that is going to see more full-throttle operation than a production unit.

Milling down this head does 2 things - it makes the squish bands bigger and it makes the combustion chamber more compact. So we increase the burn rate from CR and squish - but we also increase the heat transfer. Which wins? Only back to back dyno tests on otherwsie identical engines will show and I say again - make the squish bands smaller only to reduce valve shrouding if needed but do not go bigger. I have alluded to this phenomenon on the site with Vauxhall SOHC units.

Where sharp edges are concerned we don't want any sharp edge on the head inside the chamber because it is a potential hot spot for pre-ignition. When I say sharp - I have no more than lightly deburred many heads in that region rather than given them full radius and not seen any adverse effects. Would they be better, eg: on the SOHC Fiat if fully radiused? I cannot think of a reason why.

Regarding asymmetric combusiton chamber, you mentioned ' the charge being less able to push on the piston'; no, it's not this, the pressure acts equally in all directions. Insofar as flame front development is concerned, and 'dead' areas which might be liable to detonate, we're talking (rather generally I grant) about layouts where the ignition takes place on one side of, say a large intruder dome piston and there is a 'race' between the advancing flame front trying to ignite all the mixture on the other side of the dome before its pressure and temperature rise so high it self-combusts on its own with up to 20 x normal firing pressure. But detonation can also occur quite readily even when there is no intrusion like a piston dome, if the spark plug is not central in the chamber ie: the layout is asymmetric - because the flame front has differing distances to travel across the region. Imagine the flame front as an 'eye' - if it cannot see the mixture - it cannot ignite it readily.

It's as easy to overstate as understate the effect of intrusions, you need experience to judge; the reality is that even 10 deg after TDC only about 1/2 of the charge is burned and max pressure doesn't occur until about 16 deg after. The entire burn process occupies something like 28-90 deg of crank rotation, so what you think may be an adverse intrusion when examining the piston at TDC - may in fact not be an intrusion at all.

The 8v Peugeot heads and pistons are very effective designs and very resistant to detonation, the most common cause of knock I know of with them is over-advanced ignition from the MBT (max brake timing) value - overheating everything in the cylinder, under-lean fuelling, overheating intake air or too high CR for given fuel octane. The same is true of all the 16v layouts I know of.

Hope this sheds some light on a few things.

GC

[Christopher_205Rallye]

Hi Guy,

Many thanks for taking teh time to answer in such detail. I read your reply and examine your photos with great interest. Its really helping me to understand a bit better about squish, swirl etc..

This discussion would be easier to understand if you could offer a photo of the head you're using with the big bore gasket laid on it,

Click on picture for a large image:

Closeup with value in



Closeup with valve out


Combustion chamber with inlet valve in


Combustion Chamber valve out

Well, the more flat area of combustion chamber in close proximity to the piston at tdc - so the more squish there will be.

Block (I know it looks nasty). Covered in oil and a unfortunatley a bit of dirt..Deck height is zero!! This means there will be a 1.5 mm gap between the head and block on the overlapping region. Valve cutouts are obviously for the 16 valves (born in a 106 GTi 16v)

Often if you mill a head right down the squish bands (regions) get much bigger. Sure, squish improves in-cylinder fuel mixing by vortex generation but is much less important on high speed engines than road cars. It's one of those enhances part-throttle performance things.

The plan was not to skim the head (only a very think layer maybe). Because I calculate that with the longer stroke and zero deck height the CR will be about 12:1 with a standard gasket.

I would maybe not alter the existing bands or grind out the adjacent overlapping regions of the head.

Ok.....But anyway this was my original idea. Since the cylinder was now bigger it would give me the opportunity to free up the space between the valve and the edge of the combustion chamber, and still keep the contour of the figure of 8 . Just an idea..

You are putting a head designed for a small cylinder (75mm?) on a cylinder of 78.5mm dia right?

Yes that is right :D But I just measured everything again and hopefully I don't confuse matters :?
Diameter cylinder 78.5 mm (1.6)
Diameter head gasket 79.5mm
Thickness head gasket 1.5mm
Diameter combustion chamber 76mm

Hopefully I have explained a little better this time :roll:

As for exhaust valve working better when it's shrouded, I don't understand the reasoning for that. The power loss from any 'deshrouding' of same - if there even is any shrouding here (which I cannot see) may in fact have come from non-associated cause induced by the 'deshrouding mod itself.

Be cautious of 'do this - do that' advice unless back-to-back flowbench or dyno figures are cited and the op displayed in detail pictures.

Yes you are of course right. It comes from a book but there is no evidence, I have to take the authors word for it. Hope its ok I quote here..The book is over 20 years old now. But actually an interesting read :shock: Just for your info..

"The exhaust valve requires very little deshrouding at all. I generally work to a figure of 60% valve lift for the radius between the valve head and the chamber wall. It is a mistake to exceed this figure as the exhaust valve flow very well partially shrouded in fact in seems to enjoy being shrouded. I remember when I was able to gain 4% bhp on a VW by shrouding the valve a little"

A. Graham Bell, Performance Tuning in Theory and Practice

Regards

Christopher

[James Bowen]

Guy,

Many thanks for the detailed explanation. Your insights into the subject of CR, and Squish, have really helped my understanding, and hopefully will result in a better "managed" head, and compression chamber, on my next effort.

Incidently, just to show how valid this is... My SOHC Fiat 1500, produces max power at 30 deg advance. Standard is I believe 34 deg. Sufficent to say that Compression is raised to 10.8:1, and squish is appreciably more than standard. Something that will be done better next time.

Regards James

[Wallace]

A bit dealyed - but this is one of the pistons for my engine. The top of the piston will just about be level with the top of the block at TDC - this will hopefully give a nice squish area.

The dish is to give the required CR - 7.5 to 1 - and the 45 degree angle to the walls is to avoid and dead areas.

[Guy Croft]

Wallace hi

may I request sight of your stroke, head vol and bore size and bowl volume for my CR archive? PM, Email or here as you wish.

Thanks

GC

[Wallace]

Will do - the paerworks in Stamford - I'll dig it out.