the m20 developement thread. Some progress and a new lump.

[HairyScreech]

looks like si13 has been right all along, just been looking at the frictional forces in the 2.8 engines and the side load on the piston that creates these.

using a 130mm rod is the norm. which is a 1.548:1 rod to stroke ratio results in 1518.39n side force in the piston (against the bore wall which results in losses)

using a 135mm rod is a 1.6071:1 rod to stroke and 1461.607n side load.
a 3.885% reduction in friction loss

using 138mm rods gives a ratio of 1.6429:1 and a side load of 1429.78n (thats 10kg side load per piston less every stroke!)
a reduction of 6.197%

a potential 6% increase in efficiency.

these have been done assuming a pressure of 100bar at 15degrees ATDC, now depending on the actual combustion pressure they can be tweeked up and down as the pressure is a linear relationship.
if anyone knows the actual peak pressure angle for the m20 then let me know and ill go over them again as the bigger the angle ATDC the more significant the effect.

the correct length rods really are not to be over looked.

(im going to expand this into an m20 developments thread as it seems the output were getting from the 2.8 strokers is pretty fricking weak. were getting some big losses)

[StuBeeDoo]

So what is the answer?
135mm 'rods and custom pistons?

Does this mean that M20 2.8s aren't worth doing?

Does similar apply to eta 2.7s?

[HairyScreech]

we shall see, i know theres a guy building one with m52 rods and pistons, pocketed for the m20 valves.

the 318 m42 motor uses an 84mm bore and a 138mm rod, i want to find out if theres a way to cannibalise these possibly.....

after all a 130mm rod with a 84mm stroke motor is quite a poor ratio, considering the angle of the rod creates the side load on the bore....

i have a plan, iv been thinking about the losses we get on the m20 and where is will be possible to claw them back, iv got to go build 4l v8 ka stuff up the garage now but ill post again later...

[stonesie]

Very interesting

A 6% increase in efficiency in my engine would be an extra 12.36 bhp at the flywheel, with it's best rr result of 206 that would mean 218.

and for the torque 184.4 + 6% = 195.46..... Worth doing

[e301988325i]

I think the main advantage of the 2.8L conversion is the relatively small cost of donor parts. Custom rods and pistons for 10 brake, whilst worthwhile is going to be around £1200 for a set of 6. . .

Going from 130mm to 135mm rods reduces load by 60N, 135 to 138 only an extra 30N:

Perhaps custom pistons on stock rods wold be the best value, but it's still going to be £100/slug.

I've heard M42 rods are also 10 or 12g lighter than M20 ones and have to handle a higher torque/power output per cylinder than the M20 ones, so this would also decrease losses.

[Ant]

Guys you cant do percentage gains if only it was that simple.

I've built many 2.7's with 130mm rods, and a couple using 135mm rods, they were within 2 Lbs/ft of each other and 1.6 HP, have the dyno sheets somewhere on one of the 7 PC's I use

agreed the longer rods are mathematically "better" no arguing there but in the quest for more, how much £££ are you willing to spend ? and given what I and others have seen are the gains worth the ££ and time ???

I pulled apart an engine using M5x pistons 140mm rods and an 81mm crank, it was internally scrap after less than 40K, thrust lands on the crank were beyond saving and the bores were 1.6mm out of spec on the thrust side, ovalled pistons too, never made more than 200HP either

Easier way forwards is open the bores up IMHO, look at the results from the IE 2.9-3.1 Litre conversions, great but again the outputs are not stacking up against the % increase in capacity.

A typical 10.5:1 CR 81/84 2.7 is 8% bigger in capacity but 32% more powerful as a result, apply this rough calc to the rest of the hybrids and see whats what.

not pi$$ing on anyone's fire here BTW, just being the devils advocate as usual

[daimlerman]

I would also be interested in the 'definative' answer,as I am looking for a winter project but am currently banned from buying any more cars! I reckon that an engine would slip under her radar,though.....

[Morat]

hehehe I can't see much slipping under Jane's rader

[reggid]

Very interesting

A 6% increase in efficiency in my engine would be an extra 12.36 bhp at the flywheel, with it's best rr result of 206 that would mean 218.

and for the torque 184.4 + 6% = 195.46..... Worth doing

Lol 6% increase in efficiency my a$$.........the logic is fundamentally flawed.

Assuming the calcs for lateral force are correct you have reduced the lateral friction by 6% not the efficiency of the engine so you are not gaining 6% more power/torque.

It is worth noting that the lateral friction is not the only friction component nor is it the highest friction component in the engine.
So by reducing one component of friction by 6% you have not reduced the total friction by 6%......

Secondly while it may reduce by lateral piston loads by 6% that is obviously the peak and at other moments in the cylce the reduction would be much smaller.

It would be good if it was that simple......

[stonesie]

Ok so i was a bit wrong, it would be nice though

[HairyScreech]

yep, i was meaning to post this last night but was down the garage too late, you and ant are both right, the piston friction is not the whole story, its 6% of ~30%.
real gains could be made in the valve train in terms of frictional losses, but for the sake of just using the correct length rods the reduction is worthwhile.
the actual relationship to what comes out the crank at the end is a tricky one, as although its a tiny amount in terms of power lost, its power that once recovered is sent to the crank. (as a fag packet job and ill check properly later its worth about 7nm at the crank, so as ant says about 5lb/ft all else being equal)

the post was really to say that si was right and using the 130mm rods really isnt doing us any favours here, yes as you say those calcs are for peak, and at other moments in the cycle it will be less, but it will still continuously be in the region of 6%, ill run it later as although the numbers may change i have a feeling the % at the end will still be similar.
also as its an L/Rxsinextheta relation ship the later in the stroke your combustion occurs the larger the benefit is, the more advance or the more boost you shove in the greater that effect is going to become.

i posted that as a quicky to get some discussion started as i was surprised in the effect of 8mm on the rod length, iv had it in my mind for a while that a 130 rod is to short for an 84mm stroke, i mean a ford 1.3 is using a 125mm rod with something like a 75mm stroke.

the losses are really what we need to look at in these engines as they just dont give the power a tuned engine with such a theoretically good design should make. (well theoretically but theres obviously something not working out).

i dont want this thread to degenerate into an m20/m5x/m30 slanging match and im quite pleased its got people thinking.
it would be nice if this thread can become a methodical look at the losses and whats stopping us getting better power out of the m20 strokers.

as ant said above the biggest gains seem to be from increasing the capacity, but having run the numbers using build threads iv found with actual independent dyno plots at the end (yes i know dynos vary blah blah) rather than numbers the guy trying to sell you one tells you, the numbers just dont stack up well.

2.5@170hp - 68hp/l
2.5@180hp - 72hp/l (btb chip all the bits you would stick on a 2.8 you gave a toss about)
2.7@190hp - 70.37hp/l (weak head flow as 730 casting)
2.8@206hp - 73.57hp/l
3.1@226hp - 72.9hp/l

now if marchs claim of 220hp from a 2.8 is true (and as he says it was on frizbits rollers that can apparently read high) then hes doing the best with 78.57hp/l, even still theres plenty of room for improvement there and not a lot of 2.8s even get close to this. 205-210 seems the norm.

think its time we sat down and had a proper look at why.

-----------------------------------------------------------------------
edit to add
-----------------------------------------------------------------------

these are my thoughts, im not an engine building master, just an informed logical thinking, willing to learn from the real world, meth drinking, bitch slapping, piston head.


physical capacity
its 100% guaranteed that increasign the capacity is going to give us more power,
the suggestion has been (from bmw no less) that 500cc per chamber is ideal for a road engine, due to thermal losses of a bigger chamber.
now the 2.8 has a nice 84x84 chamber, so its perfectly square already and requires no meat taken out of the block and can use a standard gasket kit. with how prone to cocking up head gaskets these things are is it wise to stick with what works on this end?
is 2.8 the ideal for a home build?
what is the capacity to surface area for all of the well known strokers (ill do this later on as i want to know)

bore to stroke
2.5 = 1.12:1
2.7 = 1.03:1
2.8 = 1:1
3.1 = 0.948:1 (why i think the hp/l drops off on the 3.1 stroker)


rod length
as sugested above theres more advantage to longer rods than first meets the eye, not only will they reduce the side thrust on the piston reducing bore wear, friction loss, heat and side to side piston slap. a longer rod allows a shorter lighter piston, which is easier to start and stop at the top of the bore. a shorter piston will also be less likely to flop out the bottom of the bore at BDC and less likely to collide with the b28TU crank.
the common rod is a 130mm 320 rod, now the options are:
m20b25 rod - 135mm but heavy
m50b25 rod - lighter and 135mm
m52b28 rod - comes with pistons that can be modified to fit and is lighter and 135mm
m42b18 rod - is lighter again, as has been mentioned supports a higher output per cylinder and is 138mm, just need a piston to work.


head flow
the heads on these are not the best design granted there not the worst in the world but theres real world improvements to be had.
the alpina spec seems to be a good starting point, as they show that theres plenty of meat to be taken out of the exhaust ports for a good gain.
im going to pick up a head and section it then play with it on the flow bench i have access to in order to find what can be gained where.
are larger valves worth considering on these heads? i know theres a valve clash issue around TDC.
is it worth finding some valves from another car that are either the same size or a little larger and have wasted stems? wasted stems can account for a lot of extra flow, especially if a head has issues with flow on the short side of the port as it allows more to pass across the back of the head and into the chamber.

current thoughts im having are:
alpina spec on the exhaust, short side of the intake port, clean up the guide area, smooth the seat to port transitions. use flow bench to determine whats working the best. pay attention to port velocity at low depression.


Head Camber Design
at the moment we have the normal bmw swirl head, which has been shown can be taken out to fully hemispherical by alpina.
as im going to be getting my hands on a scrap head iv got no worries about routing a cylinder out, seeing how much volume is gained/cr lost and seeing the effect on flow from the greater area surrounding the valves.
this alone is known to improve things by gettting the chamber as close to spherical as possible, this leaves us with a shape as close to the shape of the flame front as practical and improves det resistance and thus allows higher CR and ignition advance, both or which improve efficiency.


Pistons
we are still using the standard 325 pistons, now there good but there giving us sub 10:1 compression on a standard head, with a hemi head its going to be awful.
i know a guy is using m52 pistons which have enough meat to pocket for the valves and give a 13:1 on a standard head (he has skimmed a few mm off to get 12:1, plenty of meat on them apparently). only issue here is they need to be used with the m52 135mm rods.

if we settle on 2.8 then its a 84mm bore were after and ideally something that would work with the m42b18 rods as there 138mm, cheep and ideal.
we also want something thats going to bring the cr up to :10.5:1 on the hemi head.
pocketing to pistons is a given if there going in with the 2v head. but thats no real drama.


Windage losses
Now in the m20 as standard there is no control of crank windage at all, no scraper and no tray.
theres real advantages to both.
worth looking into the available scrapers and modifying a m50/52 windage tray onto it?
iv heard of 5-10% from the fitment of both to other large capacity engines. even if its only 4% then were still onto a winner as the will reduce surge up the block on left turns and prevent oil frothing. there also cheep, ~£45 delivered from america and a sh m52 tray tacked on = cheep win?


now nothing above is to radical, im confident i can cary out most of the stuff myself, and none of it is expensive stuff to have done.
im happy to do the porting myself, with a superflow flowbench to back it up.
im also happy to sacrafice a head to do it and the hemi chamber.

i have access to machine tools (mill, lath, surface grinder etc) to pocket the pistons and skim them if need be.

perhaps we can get a new spec together than can be built at home and results in a bit more win? (that will show the 24v coonts)

[e301988325i]

have you read this???

http://www.e30zone.net/modules.php?name ... 64&start=0

Overall it says BMW did a very good job on the M20 head!

BMW's talk of 500cc being best is marketing B.S. 500cc can be achieved in infinte combinations of bore and stroke, the shorter the stroke compared to cylinder capacity the greater the thermal efficiency of the cylinder, BMW aren't using notably short strokes and big bores so go figure. . .

You're talking about bhp/L, which in NA engines is increased by high revs (cam) or raising the compression ratio. Let's forget high revs, if people wanted that then Honda have that base covered. I'd suggest that by doing strokers, people actually want more useable mid range torque.

Getting the compression ratio up would surely be the main aim. what are these 2.8L m20's banging away at, 9.5:1 if that??? IIRC wouldn't going up a whole point to 10.5:1 produce

With the M52b28 and 130mm rods, the piston is actaully 0.5mm lower at TDC, so that's an easy 0.5mm skim, how much valve/piston clearance is there in a standard engine, how much can the standard pistons be pocketed?

An interesting read here although it says going from 9:1 to 10:1 would only net 3%

http://www.popularhotrodding.com/tech/0 ... index.html

[HairyScreech]

yeah i have seen the head flow thread, theres lots of good stuff on there, but no numbers really, no talk of which valve he is dealing with on the graphs, just comparative flow to other heads.

they did a good job, but its been said that a further 10% is available in the head. and there is also talk of improvements that can be made to the head on there. most of which follow conventional porting thinking.

if we can find even 3% of that then its worth a further 6hp to a 2.8 (ish i know you cant go by percentages)

i also want to see how much the alpina style hemi head is worth when combined with a modern comp ratio and ignition system.

iv just won an organ donor 2.5 on ebay for £51 (no manifolds or accessories but if got those from a mates car) so let the ghetto super stroker project commence.

this is more about what we can learn than bleeding max power out of the engine. agreed torque is just as and in many ways more so than peak hp. afterall torque is the push that makes you move, hp is just a function of how often you are pushed at one specific operating speed.

if all im left with at the end is a 2.8 making 230hp +220lb/ft and has cost me the best part of £1k then ill be happy. even if it makes nothing above a conventional 2.8 build (unlikely) then the learning involved will still make it worthwhile.
if we only make 210hp but gain 2mpg then ill also be happy. it will pay for its self quick once it goes in the daily.

[e301988325i]

This will be interesting, a 2.8L stroker is planned in the back of my head, but there's a lot to do first!

I would like to know how much could be skimmed off the head in terms of water channels or the like, with combustion chamber modifications and the piston being 0.5mm lower then I'm hoping for quite a lot!

[stonesie]

Mine had 0.5mm taken off the block to return the squish height to standard, iirc that worked out to a compression of about 9.7:1

[HairyScreech]

well i already know of a cracked 325 head back home then i can section in both directions, i can split it into multiple bits if anyone wants to know anything specific.

hemi-ing the head will lower the cr a bit, but untill i know how much meat there is there i cant say by how much, and si has already put the head back on his alpenis.

might pop into work over the next few days off and route a head chamber out to hemi on the cracked one. its a test i want to do anyway and i have 5 cylinders to play with on the flow bench.

[daimlerman]

Folklore tells us that the 885 casting is thinner in places than the less effiencent 731,be interested to see a photo of both sectioned,across the chamber and above the exhaust ports,if that is possible.
Hairy,are you planning on marketing improved heads as time goes by?
No doubt in my mind that Alpina knew their stuff,the head mods that we have seen via Simon and Trevor's photo's show a modded chamber and exhaust port,it's knowing just how much to take out and where that is the skilled bit! I would expect that Alpina's specialy cast piston would also be needed to get the very best from this kind of work.
At the end of the day,guys like myself,working in our back passage,have to make the best of the available parts,and zoners like stonesie and goosie seem to have done just that.
So if a knowledgeable guy such as yourself can come up with something better,we are all interested.

[HairyScreech]

i have a 731 head thats of no use as well, consider it sectioned (it was fitted to a 325 engine by a mate of mine when the 885 head cracked, on a early high comp motor no less, yes it is possible but requires the chambers cuting back on the swirl bit al la alpina hemi.)

although i would be surprised if the 885 is actually any thinner, their totally different castings so i expect there the same thickness but with the ports enlarged.

[HairyScreech]

right, mate dropped off the 885 head earlier today, its well and truly cracked so its a prime sacrifice to the gods of flow and VE. cracked clean through two bearing journals, possible to spot with the naked eye without cleaning the head up first.

will try and get pictures in the morning, first jobs are to section it, measure a standard chamber and hemi and measure a chamber.
measure the valves and see what can be done in the way of wasted valves/multi angle seats.

[HairyScreech]

right got the head the other day as above and spent a few hours on it yesterday, iv got it into a position to begin the first tests.

i was to flow it:
standard
standard ports hemispherical chamber
ports modified chamber standard
ports modified hemispherical chamber

then from that work out what can be done to make it better and use the last cylinder to do a final design sort of thing.

ill be using a port matched 2.5 inlet throughout the tests unless it seems that its restricting or seriously screwing with the results, if thats the case then ill have to do some inlet fannymould work before flowing.
regardless i think i will be flowing the standard and final cylinder with the inlet and with a trumpet to see how much there is to be gained from correctly sized and tuned throttles. (in flow terms).

anyway onto the pictures.

[HairyScreech]

first up, the organ donor. a pre-facelift 885 casting thats cracked clean through 2 cam journals. spent a bit of time cleaning it up to give myself an easier time later on.






im a little curious as to how it ended up cracked, as it seems to be quite a good one other than the massive cracking. bit of corrosion around number 6s water jacket but nothing to make me think its been abused. still will see what the jackets like when i section 6.

next up, standard chamber standard ports.
came out to ~41cc


next job was a hemi chamber like the alpina head, wasnt a very delicate job tbh, just smashed a load of alu off with a burr and then smoothed it all out.
end result is this


nice smooth hemi shape, the little lumps and bumps the air has to flow over at low lift removed and the whole thing basically made as smooth and spherical as possible.
end result is ~45/46cc cant be more exact than that as at the end it began to leak where i had nicked the valve seat. ill lap it back before i flow it and that should help.
havent worked out the effect on CR yet.


next job was the find chop the thing out in a similar style to the alpina porting job.
the aim being really to produce something similar to the best factory head available. going by the alpina job i know im not going to break into a water or oil jacket, so i could safely take this much out without having the head sectioned yet and it will give me an idea of the best starting point. once iv got the thing sectioned then ill give it a more aggressive porting job but for now the main things have been to:

-smooth and transitions in the casting most noticeable are where the casting meets the bore for the vlaves, around the guide, and the seat/port interfaces
-reduce the radius on the short side of the port to reduce the corner the air has to turn around
-open up the top of the ports to make the ports a little bit steeper
-open the exhaust right out to the 34mm manifold hole like on the alpina heads
-radius/smooth the end of the valve guide boss

the ports shown in the comparison shots are the ones for the hemi cylinder, but the ports are still standard on that one, it just happens i worked along the head so their next to each other.

bottom of the valve guides and valve bore to port transition




intake ports


exhaust ports


looking down the exhaust ports (shots taken at the same angle)



same with the intake ports



comparison shots of each




and a little cheep shot.

[daimlerman]

Interesting work....beats making a sodding doorlock funtion correctly!
Of course,the spark plug position is far from ideal,if it were central at the top of the chamber you would do even better.
'Proper' hemi-heads that immediately spring to my mind are the Triumph Speed twin by Edward Turner,developed into the Daimler V8's in the late '50's.

[HairyScreech]

yes, it is a bit of an issue, as it means a difference in the distance the flame front travels, which is increasing the detonation risk over at the far side of the chamber, im not 100% if its fod damage, damage caused in transit/storage or if the damage on number 1 at the far side of the chamber is detonation damage. this heads certainly had a chequered life that much i know so it could be anything.

i think a trick could be to either
1 use a longer nosed plug so that it protrudes more towards the centre.

2 key the plug so that it faces the centre of the chamber

or 3 use a rim fire plug that dosent spark from the centre pin to the electrode vertically but sparks from the centre pin to the rim horizontally, and make sure it has only one electrode so that it always fires the spark towards the centre of the chamber.
therefore not actually moving the plug at all but moving the spark 5mm or so towards the centre of the chamber.

[Rav335uk]

Lovely.

[jimmyspeed]

good work!!!!!!!

am slowly getting the parts together for my 2.8 build and would be cool to do some porting for some extra speed

what have you been using to do the shaping a dremmel or air drill???

[HairyScreech]

big bosch power drill with a few burrs and grinding bits, not ideal and ill find something better (like the die grinders at work) when im doing something other than a scrap head.

my goal here is to build a spec or formula for a better 2.8, im trying to avoid full on custom work and parts, with the intention that anyone at home could then build one.

the main reason your told not to port at home is most dont know where to take material out from and what effect there having on flow without a bench, but if theres a specification to follow then you can just copy that and apply it to your head, anyone can grind some material out if they are told where to cut.


need some info on rod lengths, piston heights that fit an 84mm bore and the comp ratio they will produce now really.

first thoughts are:
135mm e30 2.5 rods with pistons skimmed down to near flat to raise comp and prevent them hitting the head with the 3mm extra stroke.

m50/52 rods and pistons, 135mm and again pistons modded as an American guy has done.

140mm rods from somewhere with pistons from somewhere else?

anyone got any ideas?

[Ant]

135mm e30 2.5 rods with pistons skimmed down to near flat to raise comp and prevent them hitting the head with the 3mm extra stroke.

Will leave the centre bowl of the combustion chamber very thin dude, also you'd need to replicate the squish band profile and the added issue is the top piston ring is going to be closer to the combustion chamber than ideal.

Just a few pence from me

[HairyScreech]

your couple of pence are always welcome.

both good points, i have about 12 spare m20b25 pistons here, my thoughts were just to bring the outer edge down and leave the bowl as untouched as possible, iv got some that are bin fodder anyway so it might be no loss to try one out.

the ring land is an issue, im not sure how far down the piston they are to begin with though so thats a bit of a make or break issue to that idea.

standard squish pattern? not sure if thats going to be so much of a problem anymore, the head will be hemi, so the standard squish will be wrong anyway, thats all something im going to have to look at.

personally i dont think the skimmed m20 piston is a go-er, its just something to try (the pistons are frag anyway )

you havent got any frag rods and pistons about to fiddle with have you?

http://www.e30zone.net/modules.php?name ... c&t=171406

[HairyScreech]

heads flowed with some interesting results.

[Jhonno]

Which aaaaaare... ?

[HairyScreech]

coming, sorry, im still fiddling with them a bit,

long and short of it is

standard flows very well, peaks at 85% of potential on the inlet and 79.95% on the exhaust, meaning that its hard to improve on that much, its doing its job well for a road car. the intake also seems to suddenly go turbulent at 14.7mm lift and flow dropped quite a bit. (edit, but that dose mean theres 15% and 20% more available. which means its still worth attacking.)

to lift the valve to 35% of its area on the intake requires 14.7mm of lift!
so i need to make the results fit to the real world yet.

it would appear that the light porting work i did had more of an effect on the inlet, with a gain of about 4% around the 5-8mm lift mark and less either side of that.
exhaust port didnt have much of a response, a gain but only significant below 2mm lift (about 4%) and then only a couple of % after that.
the porting work gave peaks of 89.4% potential and 80.9% of potential respectively.


the hemi head did some strange things.
it showed as much of a flow gain as the porting work on the inlet, right up to about 7mm lift when it suddenly tapered off, ending up 10cfm behind by 10mm lift.
the exhaust seemed to do similar but held on up until 9mm.

ill keep messing and post more as i figure it out.
more thought required and more suggestions welcome.

[Simon13]

its all in the rod angle!

[HairyScreech]

very much so, and if my calcs are right then its a good couple of degrees difference at just 15 degrees crank angle.

a 140mm m40 rod may be even better, and dave up in scotland said he had some, just have to wait for him to get back from lemons 24........

im going to take another look at those pics of your alpina head, there may be something iv missed that will improve the flow results even further. with these kind of things the devil is in the detail.

[daimlerman]

Interesting,may just be that the only 'real world' improvement will be a little polishing and sorting out rod/piston compatability.
Have you tried larger valves,yet?
731 chamber shape seems to work well with eta flat pistons,gives good low down torque and there is room for at least 325 sized valves....but....bigger valves seem to loose the low down torque at the expense of more bhp above the 4000rpm mark,any scientific answer here,please?
Just thinking out loud on this one,would 731 head with eta pistons suit any of the available rods for a torque laden 2.8?

[DanThe]

Longer rods are indeed the way to go, but as the M20 uses great lumps of material for pistons your going to have to go custom, which sort of defeats the purpose? M50/M52 pistons are much shorter/lighter but they are also a lot thinner, if you fancy machining the wasted space of valve pockets out you might run into problems.

Ive got some 140mm M52 rods here if you need them