porting a BMC A-series 12G202 (Mini, Sprite etc) & dyno

[SirYun]

Hi.

Guy asked to post some pictures of what I was working on in another topic. Rather than veering further off course I started a new topic. Well here goes.

This is my first time porting what is known as an a-series closed chamber head (I normally work on big bore heads for 1275cc engines) instead of 850cc/998cc/1098cc heads. Much of what I have done so far is described by David Vizard in "tuning the a-series engine" but there are quite a few points that are only described in outline.

The GC-VW features a very similar head btw.

This type of head was used on the cooper 997 and 1098 engines and has slightly bigger valves (still tiny) and a 'squarish' intake runner for the siamesed intake ports. It will get a slightly oversize plasma treated stainless intake valve (same price as the stock ones).

Valve job and bowl:
20/45/70 with a 1.6mm seat which is a tad on the wide side but I prefer seat life and a bit more flow at low valve lift to outright flow.
Guided by many people in the know (Guy and many others) I try to keep the seat angles as discrete and sharp as possible. Which is easier said that done.
Valve back cut to contact face at 30 degrees.

Bowl is blended to the back cut at the long side. SSR goes up pretty much straight with only a fraction of the down cut remaining.

The port is biased (it leans in a bit) toward the cylinder center and I try to keep the sides pretty much flat



The chamber

The closed chamber head has a very deep and distinctive shape to the chamber.

This presents a few problems:

The shape causes quite a bit of shrouding of the valve one if temped to grind away generously to provide ample breathing space.

However, chamber volume is rather large and thus you need to be careful not to cut huge amounts away.
If you cut too much you will have trouble getting the CR up to a useful (in this case 1:10.1) as you will have run out of deck thickness for skimming the head.

As a result the objective is to deshroud the valves as much as possible by cutting away the least amount of metal.

The port

Port runner:

Just cleaning it up and grinding all the surfaces straight. I use a straightedge to check if there are still undulations.

Exhaust:
According to flow bench result published by Vizard the guide boss is a major obstacle to flow from about 0.200 inch lift. I have stared at the port shape (luckily Vizard published cut away images of the ports) for quite a bit and I will try something a bit different.

The casting is quite thick and this allow for some reshaping. My goals is to reshape the guide boss and SSR (which is rather chunky) and the rest of the port into a Laval nozzle shape. I know of one porter who works with this shape but funnily enough he does not want to tell me exactly what he is doing. But I suspect that the idea is to get the exhaust as efficient as possible and use more overlap in the cam but not too much duration..



First thing that needs to be done is to cut away metal in a way that will turn the large and horribly shaped sort side into something that will actually allow air to pass. This entails cutting quite a bit of metal trying to widen the SSR and provide a large as possible radius. Idea being to get a more active SSR with more volume to allow the air to slow down and negotiate the 90 turn at lower valve lift.

Secondly I aim trying to change the discharge pattern from the port and seat to something that is more aimed toward the centre of the cylinder instead of pretty much straight through into the combustion chamber wall.

As I have limited testing equipment at this point I have to rely on some very unscientific methods.

I use the dust from the grinding as tell tales where the flow is going. While very crude, it does tell you that a valve is discharging around the entire or only a portion of its perimeter at a certain valve lift.

By reshaping the SSR, throat and valve boss the pattern of dust when blown from the port has improved from nearly everything in a narrow jet along a small portion the long side to a much more even pattern indicating a more active port with less "dead air".

So far I¢ž¢m a bit disappointed of the air activity at the SSR at low lifts. Even after major reshaping. As this is with a roughly cut seat it might not tell me anything.

Joost

[SirYun]

I figured out how much you can skim in this head by measuring the depth of the oil transfer hole as suggested by Vizard. In my case this adds up to a deck thickness of 4,5mm.
Apparently the minimum thickness is 0.060 inch (1.524mm) so I can take off 1.8-2mm and still have some margin left for future skims if needs be.

I spend half a day hacking away the the head (the joys of cast iron) and my main problem is getting to the guide bosses in the exhaust port. My plan for extensive reshaping is probably asking for trouble so i have reverted to a conventional shape similar to what i use on 1275cc heads. Even getting this shape ground is hard enough as access is very tight.


UPDATE 11/12/2006:

I got the new valves to go with the head. I opted for the maximum size valve that can be had. i.e 30.94mm 1.216 "

these are stainless steel with a plasma treatment so you can use the normal guides as well. (cast iron vs manganese bronze). not expensive and well made with a nice thin profile and waisted stem.

as the valve is quite a bit bigger i worked on the chambers some more

the wall is now cut back on a 0.9inch radius , still a bit tight but after CC'ing the chambers halfway during the process i dare not take too much off in fear of not being able to reach a sufficient CR


I'm now working on getting the ports/pockets/throats matched. quiet a bit of work as i have opted to keep the diameter as small as possible leading to a slightly underhung bowl ( the bowl narrows a bit towards the seat area wich a minimum area about 3mm down from the valve seat). This is pretty much what has been done to the MG factory heads which had a bowl milled in by a form tool and little else..

While the allmost knee jerk reaction is to bin this head and fit a 12G295 head (which was meant to work with a raised crown piston which are no longer available ..thus needing massive skimming often leading to cracked heads). While i can imagine that the massive effort needed makes it a very uneconomical option for professional porters, the shape you get after grinding for many hours is quite similar to the 12g295 but the ports are smaller and the walls are thicker so more tweaking can be done.

According to Mr Calver you can get as much as 120 CFM out of these heads (not measured at 10 cm H2O obviously. probably 25 or 28 cm) if you really work on them, which is pretty close to a (well!) ported 1275 head that will probably get about 125 to 135 CFM witch a lot of effort as well.


hope this is of use to someone

[Guy Croft]

MODEL POST!

Here's a member 'having a go', taking great care to try and get it right, testing his results as best he can and taking the time to present a very nicely detailed and documented article in very good English for us all to enjoy.

I know those heads are not easy to work on. Tiny little regions around the valve throat are a recipe for broken burrs and flexi-shafts...

I found on the little Sprite head (see GC V/W) that, just as Vizard says in his famous book, port mods don't really yield better flow until you start relieving that heart-shape region around the inlet valve. What does it really do apart from shroud the valves? Well I know one thing, the cylinder is not heart shaped so much of it can go..

Well done Joost!

GC

[SirYun]

small update.

the head has since been fitted to the engine along with a new Swiftune Racing sw5 cam (mild road profile with 250 isch timing).

owner reported an extra 20 km/h on the end on the straight at Zandvoort in 3rd instead of 4th .. the Final Drive on this car is a bit long @ 2.95

but it is still not set up properly on a RR.

[SirYun]

witnessed the results of my handywork on the chassis dyno today..

56.9DIN HP @ 5613rpm and 87Nm @3614rpm.

quoted stock output (which was rather optimistic to say the least):
41 hp (31 kW) at 4850 rpm and 52 lbs/ft (71 Nm) at 2750 rpm

I had no hand in the build of the block whatsoever:

shortblock fully stock (and well used)

so the only modifications are a SW5 cam and the head.

furthermore a Hif44 carbarettor on a alloy Minispares manifold and a RC40 exhaust.

nive curves as well.

I will post the curves once i get the images