S,M,G

I will be converting a 2l cdb to 100mm (2.5) and may put a few pics up it will be tried tested in my own car under extreme conditions, closed deck is a standard way of production for most manufactures, open deck is found in a few engines normaly wet liner. Pavlo i think you have done some good research and you will achieve some good results, this is no rocket sience it just needs some time and trial, a good welder should have this sorted in a few days, the welding its self is a few hours work. If a metal shim gasket is to be used you want no more than a 20 ra surface finish if your machine shop can use a diamond PCD tool at about 1200rpm at a slow feed you will be fine, NEVER use any kind of wire ring in the block with metal shim gaskets, these are great with compersit gaskets but thats were it ends. Just a little tip for some of you, even with a closed deck block, metal shim gasket if you are running high boost and rpm,s its only a matter of time before the std bolts will give up. The key is stud and nut conversion with even clamping.
Steve

Pavlo

Stephen,

It's a good job my engine machining and new studs will be done by yourself then

Wet liners?

Paul

Fuzz

wet liners...cylinder liners that are in direct contact with the coolant.

i.e wet

Andy

Fuzz

I shan't tell you about dry liners incase you think I'm on about plastering your owse

Andy

Pavlo

boom boom.

Well, 100mm is onl 1.5mm extra on the rad than 97mm (bored 2.2) bore, sounds possible.

Paul

Pavlo

After Pat's in put, I got to wondering about extending the stroke by 10mm or so. If you added a plate and 2nd gasktet, with a liner that was the full length, I wonder if it would be good enough.

It may be possible to build up with weld, but you would need to use a filler rod of the same alloy to maintain properties. Ether way, getting a custom crank and doing the block work sounds like more trouble than fitting an H6 engine.

Paul

EMS

Paul,

You also need a longer cambelt, if your cilinder heads are moving further apart. I think that´s not easy to get!

The idea of using a longer liner in combination with a plate between block and head is already used by Hartge. (BMW tuning) I worked at Hartge in Germany also during the development of that system. It works well (liquid sealant is used between the block and the plate, because it doesn´t have to cope with combustion pressures)

Is there enough space between the heads and the cars body? Perhaps you need to remove the engine for replacing the spark plugs.

Mark.

Tim W

iTrader: (1)

Paul, can you e-mail a price for those 550 injectors please, I think I may need a set

S,M,G

The wall thickness of the alloy will be the weak link this retains ragidity in the main housing. The liner for 100mm is bigger on the o/d than the 97mm. Paul by going mad on the stroke i think there will be lots of probs on top of the ones already stated you will need an even shorter skirt piston and if you take the liner down much further the ragidity of the main housing will be lessoned.
Steve

Pavlo

Steve,

I was talking about adding the stroke at the top.

So to use 6mm longer stroke say, add 6mm to the height of the block with a plate. Liner trhough plate and block. Add 3mm to the throw of the crank, then add 3mm to the rod length.

This would leave the skirt clearance the same, as the rod length would cancel out the extra crank throw.

New inlet and exhast manifold work would be required

Paul

Tim W

iTrader: (1)

Paul, also note the comment elsewhere about Cam belt lenghts...plus, changing plugs will become even more of a chore if your engine is 12mm wider overall

Pavlo

Changing plugs?

you might just do it, maybe jack the engine to the side, as the clearance would be 6mm reduced.

It's just a thought anyway. Belt wouldn't be a problem, trust me.

Paul

EMS

I forgot about the manifolds, sh*ty boxer engines!

Cambelt is not a problem??? (I am curious!)

Mark.

S,M,G

Paul
I agree with the extra rod lenght skirt clearance will be over come
and will be a good idear as per rod ratio, i guess you are thinking 6mm onto the 79mm stroke. With a 81mm stroke with some rods some machine work is needed for clearance of the cap to the oppersit bore on bdc, not sure without looking further if it will be possible to obtain anough clearance here.
Steve.

Pavlo

I was wondering if 87mm wasn't impossible. BPM run 83 something with their kit.

I was also thinking that running the long stroke on sleeved down bore sizes to. This would mean nice thick liners, and relinering an open deck block with top hat liners might be an easy way, as of course you can make the top plate closed around the liner.

What about machining 6mm off the deck an using a plate/liner combo to close the deck with normal stroke?

Paul

pat

The trouble, as I see it, with a 100mm bore is that it's very over-square, the conrod ratio (if going for a 2.5 litre dispalcement) will only further increase the side load on the piston, plus the added mass of a 100mm piston will further increase the stress on the rod (think about the acceleration on the commencement of the induction stroke, I really wouldn't want to rev one of these things too hard, but considering the torque you can get, you wouldn't need to, just fit taller gearing).

The 2 litre has a much more friendly (some would say perfect) conrod ratio, it's also oversquare, but to a lesser extent than the EJ25. The EZ30 is somewhat better with an 89.2mm bore and an 80mm stroke, but there are two main issues that I have with this engine... a) the heads join all three exhaust ports to a single outlet on the head and b) the bore pitch has been significantly reduced (so something must be smaller, either main bearings, big end bearings or crank webs).

With regard to blocks... there are no EJ20 Phase II closed deck blocks, when you order a Phase II with a closed deck block you're actually getting a Phase I closed deck block with a Phase I crank, but Phase II pistons; since the heads are interchangeable, the Phase II heads will drop onto the Phase I block and hey presto you have a Phase II engine with a closed deck block. With regard to EJ22 closed deck blocks, there are officially 19 in the country (in the 16 22B Type UKs and the three 000/400 Prodrive 22Bs), other than that they're all "grey" (either JDM 22Bs or US Legacy / Aussie Liberty). The EJ22 closed deck blocks were made, if memory serves, in the North American foundry and were not on sale in the UK. There are two variants of EJ22 closed deck block, the normal EJ22 closed deck (also found in the 22B) and the EJ22T (very rare and expensive) which has oil squirters and, legend has it, even more "meat" than the EJ22 closed deck, making it as least as strong as, if not stronger than, the EJ20G closed deck block.

I am currently making enquiries as to the feasibility of getting some closed deck blocks cast (ie genuine, "that's how they were made" EJ22s, EJ25s, etc), but I'de need some serious numbers to make it worth while.... group buy anyone ? Already have preliminary pricing on heads (there are some issues with the stock heads that can't be solved by porting etc).

Cheers,

Pat.

S,M,G

Pat
some valid points on 100mm but i dont understand you 100%, if a 100mm bore was to be used then 75mm stroke would not be used and when the piston is desighned for the aplication compression hieght would be the main restriction on rod lenght and we know that an extra 1.5mm on rod lenght even with a 79mm stroke is no problem.
What side loading are you speaking off?, the side of the piston does not come close to the bore, the piston is oval and the thrust side of the piston is on the left faceing from front of the engine
hence offset pins to reduce piston slap. A 97mm piston that is forged is lighter than a 92mm std item and a 100mm is lighter than a stoke 92mm forged sti piston. The stroke and bore is only part of what makes an engine rev just as important is cam lift and port flow and exhaust primary layout not just 4 into 2 or 4 into 1 but what cyl are firing into the secondarys 1 and 4 together 1 and 3 together or evan 1 and 2 together. These are what creat your pulses for example some engines that like to rev will like two close pulses but this will effect low and mid range.These same pulses can effect turbo spool up so on the manifold that we are currently working with we are paying some attension to this along with differant primary diameters for max gas speed.
Steve

Pavlo

Perhaps it was possible due to the small cylinder capacity, but the ford v8 F1 engine had a 2:1 bore/stroke ratio, and reved to 14000rpm

food for though

Paul

pat

Steve,

don't worry aobut not understanding me 100%, most of the time I don't understand myself 100% either A 100mm bore does not preclude the use of a 75mm stroke (to get about 2.35 litres). In many ways this would be favourable...

Stock EJ25 is 79mm stroke, 131.5mm rod, so 1.66:1 conrod ratio. Standard EJ20 is 75mm stroke, 130.5mm rod, so 1.74:1 conod ratio. Given that the EJ25 crank has 2mm more throw and the rod is 1mm longer you could use EJ25 pistons on EJ20 crank with 133.5mm rod to give 1.78:1 conrod ratio, lowering the side-load a little, and reducing peak acceleration, obviously at the expense of displacement.

Agree with you on piston ovality and taper, but I would disagree that they do no touch the bore, one only needs to have a look at the actual wear on a piston to realise that they do actually touch (even if they shouldn't.. it is essentially a hydrodynamic bearing surface but it isn't perfect, and it doesn't have the luxury of copper-lead surface ) In any case, the shorter the rod, with regard to the stroke, the greater the angle of the conrod to the centre line of the bore. Given that the only force vector is coaxial with the rod, on any rod which is not infinitely long, there MUST be an element of side load, and the normal reaction to this side load will push the piston toward the cylinder liner... ie the piston is being pushed into the liner, as well as down it. The problem is that this side load happens somewhat down the bore (where you're trying to maintain high cylinder pressure to get good torque), so at a point where the bore has no support (like the main bearing webs at the bottom or the head / closed deck at the top), it is being subjected to high lateral rather than axial stress, which will cause bore distortion, blow-by etc etc.

I also agree about the weights of the pistons... a forged item is lighter, for the same strength, than a cast item. But a 92mm forged piston is lighter than either a 97 or 100mm forged piston :P

With regard to the periphery of the engine, sure there are many things that will affect how an engine will perform, my comments were merely regarding the effects of engine geometry on the performance and longevity. One often ovelooked thing is port flow at low valve lift, this can drastically alter the output of an engine, but many tuners are obsessed with flow at maximum lift, without giving regard to flow turbulence (swirl, tumble etc) which will greatly affect mixture preparation. At the end of the day, the aim is to get as much homogeneous mixture into the cylinder as is possible, with sufficient intertia to allow a quick and complete burn.

Fuel injection has one major disadvantage over carburetion; the fuel delivery is discontinuous, it happens in discrete bursts; the mixture is therefore naturally less homogeneous than that supplied from a carburettor. This can be an advantage in lean burn, stratified charge applications but is a real nightmare in maximum effort engines. This is partly why a setup with two injectors per cylinder will be much more friendly to drive than a setup with one large injector...atomisation is better.

With regard to header design, sure you can mess about with the actual organisation of the collectors to provide different characteristics. On a normally aspirated engine this can have a huge effect on reversion, overscavenging etc, but in all this one is relying on a step change in characteristic impedance of a wave transmission line to send a negative pressure wave back up the exhaust, and timing it to arrive at an exhaust port just before it closes. The trouble with this is that it may arrive too early of too late at different rotational speed, and thus it's a trade off. On a turbocharged application, the step change in characteristic impedance is somewhat muddied by the presence of the turbine, and hence the design goals will also change. IMVHO, one should not rely on pulse tuning in turbocharged applications, rather the exhaust manifold design should be such that the maximum amount of time for pressure reduction is available for a given pair of cylinders... on an EJ series that would mean primaries on 1&2, 3&4, secondaries not joining until they get to the turbo (ie two up pipes). At very high RPM the 4 into 1 design gives a greater volume (and hence "padding") for exhaust has pulses and may have an advantage over the 4->2->1 design.

Turbos respond better to less frequent "violent shoves" rather than a continuous pressure driving them.... in the same way as a PWM motor drive controller will give more torque for a given RMS voltage than actually applying a DC voltage equal to the RMS; it's this torque that will spool the turbo up. Obviously gas speed is of some importance but it's worth bearing in mind that the actual gas speed going through the turbine will be dictated by the turbine housing, but that what the housing can do with a series of phuts from the engine will depend somewhat on how the phuts are delivered to it. If you make your tubes too large then there will be more "buffering" and hence the driving impulses to the turbo will be reduced in peak amplitude and it won't spool as well. If you make them too small then you'll have very high gas velocity and excellent "shoves" but they'll choke at high outputs. The key is to chose a combination that won't provide too much buffering, which still providing enough isolation between firing-sequence-adjacent cylinders, and also not so small as to limit ultimate gas flow.. in essence as small as you can get away with without impeding flow, long enough to isolate cylinders, but not so long as to increase buffering again. It's a game of balancing compromises...

Cheers,

Pat.

Andy.F

I know we are talking different duties here but one of the most successful road/competition engines 'ever' The 350 Small Block Chevy has a ratio of 1.63:1
Its big brother the 400 has a 1.5:1 and was considered marginal for high mileages (200K) due to excessive bore wear caused by side loading.
'Strokers' are common place using the 400 crank with 350 rods, this gives a ratio of 1.425:1 !!! This is with a 4" diameter piston on a 4" (101.6mm) stroke. They run them at anything up to 9500rpm !

I wouldn't think 1.66:1 is a problem really

S,M,G

Pat
Some interesting thoughts there and in keeping with my developments on the manifold, when you said about side loadings, i did not think you were talking about skirt to bore contact and of course this will ocour and this is the thrust side of the piston
most of the bore wear is caused by the rings and will be at the top of the bore and an over rich mixture will cause excessive bore wear. You may or may not of heard of nippon rings these are what i use on my venolia pistons, most rings are square section and as the piston rocks in its travel it rides on the edges of the rings creating the wear said about in the bore,it needs a longer bedding in process and opens a door for blow by these types of ring also need a 25-35 ra bore finish to bed in without glazing the bore if the bore does glaze by to finer finish it will smoke and breath heavily (blow by). These nippon rings have a radiuosed top ring so as the piston rocks the ring is always on two flats and a good seal is achieved these rings need a 14-16 ra bore finish and bed in very fast and are very durable if a to course bore finish is used the rings will be worn out with in 50 miles. Believe it or not but by using a 14 ra finish on a 1000cc race engine we found 4bhp and at the end of the season there was still only 10% blow by. Some of the top uk piston manufactures also use these rings and the suprising thing is, this is not mentioned in the boring instructions.I agree totaly about what you say regarding bore expasion under presure but this is far harder to overcome, this is why i use heavy duty diesel liners with a top flange the liners are almost as thick as some cast iron blocks them selfs, but the grade of cast iron used in a liner is better than a cast block which is why they are used in the first place. You may here of people saying dont liner a turbo engine because the differance in expansion and the conditions that ocour in the cyl the liner will come out, and it will. This is why i use the top flange the bigger top flange section being bored into the block and the head gasket holding the top of the flanged liner in place. Subaru get round this by casting the liner into the block and a section of alloy sits above the cast in liner but under an extreme condition this can distort or melt away causing engine or gasket failure.
Steve

S,M,G

97mm lined block top flange liner

part of the machineing process
Block with higher quality insert machined in

Steve.

pat

Steve,

One of the biggest problems you'll find with making an exhaust manifold for the EJ series engine is that the tubes tend to be rather long, at least from the passenger side. I think it's about 38 inches, and ideally you wouldn't want to be much longer than 20 inches. But the physical packaging restrictions will obviously dictate what you can and cannot do. This is one of the reasons I decided to go twin turbo on my new engine, it allows the exhaust manifold (or to be more precise, exhaust manifolds) to be shorter, reducing the problem of heat loss over a long length, and reduces the "buffering" effect. I have the "luxury" of getting an exhaust pulse every 120 degrees of crank rotation (per side) making it feasible to have two separate manifolds, on a 4 cylinder I think the pulses would be too far apart at 180 degrees. It may be worth having a look at a JUN-style manifold which has two up-pipes, one either side of the engine, joining above the 'box and entering the turbo, with the downpipe going left before going down... it of a custom job, and will definitely lose the scooby off beat exhaust note, but with a wastegated twin scroll turbo (such as the STi VIII or RX7 Turbo 2 Hitach job) you'de get low volume and good heat retention.

Agree about the bore wear coming from the rings... the piston is much softer than the bore liner and of course will wear rather than the bore, but of course as the skirt wears, the thing will get slappy, and then it's all down hill... hmmm, sounds rather like the MY97-98 UK engines on No. 4

Interesting note regarding the piston rings; there are of course many many different types available, including the rounded type you mention... Guess it's a variation on the bevelled type

Bore expansion is very difficult to overcome, agreed, and there have been many attempts over the years to overcome it... from installing little bracing stays in the water jacket, to the STi VII block which has additional support ridges, and also with different liner materials, guages and grades... any time you add material you're faced with a dilemma, on the one hand you want to improve the physical strength, but on the other you don't want to impair the ability of the thing to shed heat to the water jacket, and it seems that most things that are pretty hard to deform, are also quite difficult to get a decent heat flux through. Hence my affinity to smaller bores, I guess... for a given gauge of material, the smaller the bore the more difficult it is to deform, that coupled to the fact that you'de be running closer clearances and that the deformation is generally a percentage, a smaller bore will deform less in absolute terms.

Nice piccies of the blocks Interesting piston design too, somewhat more closely matched to the squish areas than the standard ones... I'm looking at something similar at the moment, but rather than trying to match the piston to the head, I'm completely altering the combustion space geometry... I think one of the problems with the Subaru head is that at TDC the combustion space is too irregular, also having altered the combustion space, and drilled holes in the head to check for residual material thickness, it became apparent that the standard Subaru heads "squish" areas (they don't match the piston so they're aren't) are a long long way away from the water jacket, which will lead to irregular thermal gradients in the head. My test geometry de-shrouds the valves, while at the same time restoring some sort of normality to material gauge. A lot of squish area is "removed" (in reality it is relocated, as the new combustion space will need special pistons to function). They'll end up something like the following picture, but obviously this isn't a Subaru piston, it's just for illustration purposes :



We're gonna cut a Subaru head up this week, to join the collection, next to the genuine RS500 head Should get a few more insights as to what can and cannot be done; suffice to say we already know where to put the Optrand sensors (have a look at http://www.optrand.com/17nfo.htm for more info on these puppies)

Cheers,

Pat.

S,M,G

Pat
I like your way of thinking and im sure you will have some great results that looks like a N/A piston very similer to a N/A cossie.
I have found with heads ive cut up that minimum wall thickness is about 4mm.
Keep up the good work.
Steve

S,M,G

N/A cossie

Steve.

pat

Steve,

the trouble, as always, is that even if the theory is fine, actually turning it into reality and then making it work isn't exactly easy, and involves some degree of trial and error. The number of nuances in head design are probably somewhat beyond me... suffice to say that the new combustion space design I'm looking at constructing will not work with normal cams, or any cam profile you can get off-the-shelf; there should be NO valve reliefs on the piston (other than perhaps the slight amount the valve heads are proud of the head when closed) and obviously this requirement is at odds with valve overlap etc.

The example piston is either N/A or low pressure turbocharged application, and although I've not seen an N/A cossie piston, it does seem like it mirrors the cossie head reasonably well

Which heads have you cut up? EJ20G, EJ20K, EJ205 ? There seem to be some casting differences between each of these, the ones I was messing around with are EJ20G heads, but I've got some EJ20K ones to play about with too I couldn't find any areas of the combustion chamber roof as thin as 4mm, but of course there may be areas in the intake or exhaust ports that are quite thin... I'm concentrating on the combustion space, I've yet to pluck up the courage to think about port design

Ditto re: keeping up the good work

Cheers,

Pat.

S,M,G

Pat
I have not cut up any ej heads yet but have cut up many other veriants, 4mm seems to be a min normality and as you say the port is the thinest area, one thing i have learnt with valve pockets is allow pocket depth of lift @ tdc plus what ever you have between your coils @ fill lift on the v/springs as a matter of safty.
Steve

harbering1

a phase 2 closed deck





[Edited by harbering1 - 1/2/2003 12:24:26 PM]

Fuzz

I think the first block will cost me a "little" more than £250..

Parts required :

1 Large Tig welder

haven't done any TIG since college either...oh the fun of trying to remember.

all other work cost free

P.s I think TIG is the way to go and NOT plasma.
a lot less heat is produced with the TIG setup, I only want it getting truly hot when in the oven and evenly distributed...
wonder what Mums going to think If I shove a block in her kitchen oven for a couple of days
On another note, I collared the machinist (who popped in to get something, silly man , he doesn't start back until Monday) not a problem to machine anything like the block I mentioned, time consuming, setting up the Jig is the biggest problem (trying to clock up the deck to be parallel with the bore)
aslo a note to mention, was that NO machining is worth doing AT ALL until all heat treatment is finished. (he made a big point about this) reason I'm saying, is someone above mentioned lots of small skims etc between treatments (yes it was refuted later on, just re-itterating (sp?))

Andy

edit to add : not sure her oven goes up to 350 degrees anyway, that would be one well cooked turkey

[Edited by Fuzz - 1/2/2003 8:20:21 PM]

R19KET

Pascal,

IMO, and depending on how much power you want to try and achieve, I would ask the welder to redo some of the areas, I've highlighted a few.

These areas look a little suspect, and the first places to go, if the block is highly stressed.

Edited to say that I would double check the block, and head is 100% flat, and it may be worth getting a finer finish. It looks like the fly cutter has left a very course finish, and a "friend" recently had the same problem, and the head gaskets leaked !!!

Mark.

[Edited by R19KET - 1/3/2003 12:21:59 AM]