M20 B3x Naturally Aspirated Stroker build past, present and future

[digger]

I’m going to be doing an engine rebuild later this year so here is the journey from 2005/06 until now and beyond

OLD ENGINE HISTORY:

Stock engine 160,000km on the odo


3.1L BUILD 2006

That engine was rebuilt in end of 06 or so. There is nothing new at this point just a recap more than anything since I never actually did a build thread originally on this forum

Bottom End:

- S52b32 89.6mm stroke Crank.
- 11:1 CR 86mm Forged MM Rally Pistons.
- 138mm I-Beam Ultralite Rods.
- 240mm HD clutch assembly (i.e. M5 OEM)
- 12-13lb Stock lightened Flywheel
- Full balance of rotating assembly




I borrowed this pic it is not mine but I used the same pistons, rods etc



Top End:

- Metric Mechanic Complete Cylinder Head .
- Catcams 1300527 inlet lobe, 1300525 exhaust lobe (298/285 advertised duration)

http://www.catcams.co.uk/acatalog/1300525.pdf

http://www.catcams.co.uk/acatalog/1300527.pdf

- Eibach Double valve springs with Factory Retainers
- Nuke Adjustable Cam Gear

I borrowed this pic it is not mine but I used the same head


Intake:

- MM pulse tuned Manifold





- Custom Air filter box based off S14 unit with bigger K&N panel filter
- 63mm Big Bore Throttle Body with Silicon Elbow

Exhaust:

- BTB extractors 1-5/8" primary 26" long, 2" secondary to X-pipe



- Dual 2.25" exhaust from x-pipe back
- Twin resonators and high flow cats
- Twin in twin out straight through muffler

Fuel Injection:

- 24lb/hr injectors 3b regulator with stock pump
- WOLF V500 standalone with MAP

Ignition:

- IE Plug Wires
- NGK BPR6ES plugs
- OEM coil, dizzy and rotor

Best Dyno Runs:



the following posts will detail current developments as well as future ones

[digger]

CURRENT and ONGOING DEVELOPMENTS

This post is for recent and current work with “bolt ons” for the 3.1L engine to prepare for the rebuilt engine


Intake Manifold:

For those of you who have seen this thread

http://www.e30tech.com/forum/showthread.php?t=118357

this is new but it is a copy and paste from that thread

So in preparation for a new engine build later this year I thought it was time to get a few of the bottlenecks sorted. One of them was the intake manifold.

There are no high hp M20 around that persist with the cast OEM manifold it just doesn't have the flow or tuned characteristics to make a lot of hp. i wanted to sort something and my current engine is being used as a test mule.

I had originally thought about jenvey throttles but since RHD already had sorted the linkage, runners etc with proven results I decided to go this path. Also being significantly cheaper and a locally designed product from a guy (Rama) who knows a thing or two about high performance m20’s certainly helped the decision. i also live within an hr of his workshop

The compact throttles give a lot of freedom with regard to runner shape, design etc.

I decided on 42mm throttles because they would suit my new engine which will have more compression, a much better cylinder head, better camshaft and a few other go fast bits.

I would have chosen 40mm for my current spec engine as a 40mm throttle will support 320bhp and I was only making 250-260bhp at best on the standard manifold. I felt 42's should still work well on my old girl though strictly speaking not perfect.

IMO It is nonsense that there are kits are available with 45mm there is probably only a few naturally aspirated m20 in existence that could benefit from throttles that large without sacrificing a good chunk of midrange, this size should not be the norm.

Having done literally hundreds of dyno power runs I understood the importance of a baseline to find out as accurate as possible the benefit from doing any mods.

My baseline is well down on what it made at its best back in 2009/2010, this is due to various reason i wont elaborate on here except to say mostly to do with exhaust changes and an old engine which doesn't have the compression it used to.

Nevertheless I had a baseline with which to do an APPLES to APPLES comparison and the baseline was done only a couple of weeks before i was to convert to ITB to minimise any other chances affecting results.

1) BASLEINE: my baseline is actually a long way from stock but the cast long runner manifold still features ;)
- MM pulse tuned intake, which has tuned extensions and the end of runners are cut up and port internally
- 63mm BBTB with Silicone Elbow
- Custom big filter air box cleaner with larger K&N panel filter see here http://www.e30tech.com/forum/showpos...68&postcount=1
- Custom tuned with MAP. Back in the old days i saw significant gains in midrange and top end by getting rid of my m30afm

Tuned extensions of MM pulse tuned manifold



Ported exit



2) Part DBILAS kit: I bought the plenum off ebay and the manifold from dbilas new
- Dbilas manifold adapter, ported to reshape injector area which has a rather large bump in it. The injector angle was modified to better match port angle and aimed at valve rather than port floor
- RHD 42mm throttles and linkage kit
- Dbilas plenum with silicone reducer couplers knife edged adapter and K&N cone filter.
- Custom tuned on Alpha N (TPS)



3) RHD FULL Length Runners ( 300mm length head face to trumpet)
- CNC manifold (preproduction version), port matched to TB not port matched to head
- RHD 42itb,
- RHD CNC spacer
- RHD trumpet
- Untuned






4) RHD MID Length Runners ( 250mm length head face to trumpet)
- CNC manifold (preproduction version), port matched to TB not port matched to head
- RHD 42itb,
- RHD CNC spacer
- RHD shortened trumpet
- Untuned

You will notice that engine was last tuned with the dbilas setup so (3) and (4) are using the tune developed for the dbilas which is not optimal by any means and there is certainly improvement to be made with the full RHD setup by a simple tune.

For example the 300mm runners leaned out about 0.6 to 0.8 across the board so it is moving more air and adding some fuel will help

Here you can see the difference between the manifold adpaters the RHD one is pretty much perfectly aligned with the port to clear the booster and you can see the spark plug through the runner with throttles open and inlet valve open. so it is a nice straight shot



Here is the dbailas manifold adapter before it was cleaned up with a die grinder to get rid of the hump.





It aims at the roof of the port and would introduce a lot of turbulence in this area which can affect fuel suspension and therefore combustion efficiency not just raw CFM (dry airflow).

The dbilas adapter is 45mm diameter at the start and is only 55mm long so the size is wrong to start and not a lot of room to transition the flow to the 885 port so basically it’s far from ideal. The RHD which is longer at approx. 110mm (IIRC). i port matched to the 42mm throttles (comes 40mm standard) so no step changes in area at all and the transition is gradual to the head.


RESULTS

GRAPH 1: baseline vs dbilas



DBILAS

Torque curve is nice gradual shape but midrange is down alot and this is evident while driving. There are still gains to be had past 5500rpm with peak 10whp topend gain and much bigger useable power band as power doesn’t fall over as fast. You could rev it to 7000-7500rpm and ride the power plateau and be a lot faster but there is room for improvement with no tradeoffs.

Throttle response is good and drivability not even comparable to common plenum single TB, it is miles better with ITB and alphaN. Will idle like a baby and tootle around easily, no grumpiness at all.

Still the smaller throttles can’t change the fact that dbilas hurts the midrange and is inferior to the stock manifold for midrange.

Reason is the rest of the runner volume is still too large and the runners are too short perhaps 200mm. The good thing about the dbilas is that it fits and clears the booster.

The Individual runners got rid of the really annoying dip in the torque curve at 2700-3000rpm on the OEM manifold which is probably due to cross talking nature of the common plenum.

GRAPH 2: #1, #2, #3 & #4



FULL LENGTH RHD (300mm) #3 BLUE

Straightaway using the dbilas tune AFR leaned out 0.6 to 0.8 of a point so definitely moving more air. Despite not retuning it gained 20whp peak (untuned), 30whp at 6000rpm, probably 40-50whp at 7k the power doesn’t fallover anymore and reving it out it is a lot faster due to the bigger area under the curve. Pretty much all the midrange is back then some, the bottom end is as good. There is a small band about 600rpm where stock manifold is better but a custom tune would probably fix that.


MID LENGTH RHD (250mm) #4

Basically as for 300mm but the 300mm has better midrange 4000-6000rpm. Topend is essentially the same. A higher rpm engine will tend to want a shorter length and 250mm might suit some engines better. Give the engine what it wants

Time for a custom airbox to ensure only cool air gets to the engine since socks are not particularly great…….then a tune.

i recommend anyone looking for performance to get one of these kits. BFYB is there without question

http://www.racehead.com.au/?wpsc_product_category=e30-m20-bmw

As an aside i also made these from fiberglass to replace the dbilas plenum. they curve upwards like the dbilas and at one stage i had planned to make a new plenum that sits on top

they do a reasonable job, nice bottom end and the midrange was back but topend doesnt better the dbilas but its a much better setup to drive than the dbilas.

so the dbilas plenum is certainly hurting the midrange

still these curved ones better the BMW one across the board pretty much

compared to the full length RHD trumpets and CNC manifold the midrange and topend is down quite a bit but dont have a graph. i abandoned the curved fiberglass concept










EXHAUST

Did some exhaust mods a few years back, everything I did went backwards and by a lot….lol



The engine seems really sensitive to the exhaust. Perhaps the exhaust side of the head is not working well (MM hardly do anything to the exhaust side for their head, cam etc due to emission reasons) or perhaps that it is overcammed and the ‘right’ exhaust helps counter some of the side effects/problems ….

Run 1 to run 2 was a catalytic converter change out where the new cats were moved closer and side by side as opposed to staggered

Run1



Run 2




Further investigation into this never found out what happened. I even put back a similar set of cats and used photos to position them as close to where they were but it didn’t do squat. Perhaps it was something to do with the dyno setup on Run1 and as such it is not real or there was another issue at play…..

Run 2 to run 3 was a x-pipe relocated closer to the engine
Exhaust from run 2



Exhaust from run 3 notice the x-pipe is closer to the engine



Run 3 to run 4 was a cam timing change (advanced 8 degrees). i did a cam belt change at the same time and thought maybe the timing shifted a few degrees. while i found good gains the shape of the torque curves shows that this was not the problem. the real problem is the relocation of the x-pipe


I want to redo the exhaust before the new build so will be doling that over coming month or two though not 100% sure on what to do yet

[digger]

NEW ENGINE:

This is where I will update details of the new engine to save people trawling through what might end up as several page thread.

So why the new build?

1) Old engine is very tired lots of blow by and an oil control problem that I think is ring related. The car still start and runs decently though but blows a bit of oil at WOT.

2) Original build was done with not much know how and I think I can do better

What is the purpose of the engine?

- A fast street engine that could be used as a DD, but also be tracked occasionally without getting overtaken by shitboxes on the straight
I did think about an E85 monster with higher compression and crazy cam but this engine

http://performanceforums.com/forums/showthread.php?67286940-NA-M20-with-TK-goodness

shows what can be done. I have no intention for this to be an out and out race engine even if I need to give up 50-60 bhp or so to run it on pump fuel 98RON with streetable cam and lower compression that follows.

i still expect to well and truly spank any M5x naturally aspirated engine and better all but modded S50B32 and S54B32

So what can be done better?

- Better cylinder head
- Better valvetrain
- Better piston design
- A few more cubes
- Better Induction System (See Post 2)
- Better Exhaust (See Post 2)

Pretty much everything..lol. Nothing will be the same

Cylinder Head:

A bigger engine needs to breathe better to make more power. The cylinder head is pretty important in performing this function

The ports need to be sized to suit the engine specs to achieve the correct velocity in a running engine to attain good VE (torque). If you port the head without considering the engine specs, rpm you will turn or expected hp chances are you will end up with ports that are either too large which hurts midrange or even topend (if excessively large) or too small then you choke the top end which limits topend hp. It is a balancing act that really only an experienced engine builder can tell you if you dont have test data.

The ports need to be shaped efficiently for the flow to remain stable to avoid turbulence at 60”+ pressure drop that you will get in a real engine, not just at 28” like the flow bench tells you . Average velocity (cross sectional area) is important but equally as important is low velocity gradients across a cross section particularly since we are talking fuel and air mixture

The combustion chamber needs to be shaped to slow the airflow from simply being dumped into the cylinder which is pressure recovery and it needs to allow an efficient burn.

It’s not all about CFM either but CFM at peak cam lift does correlate well to peak hp potential but it’s not all about peak hp though

The best I can put it is size it for velocity not for flow. Size it correctly and shape it properly the flow will almost look after itself

I did consider using a smaller 731 as many consider the 885 too large but the 2.5L linked to below makes good hp and torque for its size with a stock induction and cam. So the 885 might be a bit to big for a stock 2.5L it is not way too big for my particular application when I am talking 30% more cubic inches and hopefully 90% more power than stock. In either case the ports would be modified to the point where they do not resemble the original factory ports. This leaves the size of the port entry as well as the combustion chamber being the main difference. I would assume that BMW redesigned the 885 to make it better but whether it is better for performance or just emissions I do not know.

The cylinder head was done by Tony from Knight Engines in Adelaide, he has done a few m20 recently, most famous the 350whp 2.8L above and also did morerevsm3 impressive 2.5L below

http://www.e30tech.com/forum/showthread.php?t=114967

it was recommended to start with an AMC head as it is less prone to cracking apparently and reputedly having more meat but this would depend on core shift so this is not necessarily a given

out of the box it is shit

Stock AMC Chamber


Stock AMC 885 Inlet Port lol


Stock AMC 885 Inlet Port lol


Stock AMC 885 Exhaust Port lol


Stock AMC 885 vs "stock BMW 885"


But after some love from someone who knows what they are doing scrubs up ok;)









New head inlet flow with 45mm manley valve


New Head Exhaust flow with 36mm manley valve


New head reportedly is nice and quiet and stable with good velocities and also decent % improvement in flow, particularly the exhaust

Some old data I had from a few years back from a different flow bench tested at only 10” so it’s not directly comparable. there are basic correction factors but these are most applicable and accurate to straight pipes where Cd doesnt vary much with pressure difference



MM doesnt live upto its flow claims.....the Alpina exhaust port is quite good.

I plotted out the CFM vs crank angle given the approx cam profile I will be using. I then compared it to both what I have now and a completely stock head and cam.

By looking at the area under the curve It turns out the new head and cam combo as a % have more theoretical peak potential above what I run now with my MM head and catcams cam than what the MM does over stock.

So it could result in a bigger jump in topend than what i got with the original 3.1L build over stock. being only a theoretical potential its alot more complicated than that though.....



Valvetrain:

Looking at the inlet vs exhaust flow % of my current MM head with 70% it doesn't seem logical to use a split cam like I currently use and it would probably be choking the power. By choking I mean increasing pumping losses and also at the end of the exhaust cyle there is still residual gases remaining which take up room and contaminate the fresh fuel air mixture trying to enter. If you look at 4V engines like S14 and S54 they have much higher exhaust flow % so work best with less exhaust duration.

my new head is 78% which probably means there doesnt really need to be any sort of drastic backsplit

I will be using a RHD roller rocker setup with pretty much the same lobes has his 2.95L engine but slightly wider LSA

http://www.e30tech.com/forum/showthread.php?t=108906

to be continued....

[Ray Smoodiver]

Phenomenal amount of knowledge right there. Will be watching intently!

[e30m3s54turbo]

Very interesting work you are doing.

[Ether-D]

Wow. I am only able to understand about 40% of that, but the pics are the effin' business!

[LJ851]

I have been waiting for this thread !

[craz azn]

Really appreciate you posting your findings for us on here on r3v. Thanks Digger, keep up the good work!

[TheTacoMan]

Thats damn impressive, 220 n/a out of a m20 is no small feat.

[pantelones]

Awesome stuff digger! Thank you for sharing!

[The Dark Side of Will]

Who is RHD? That low profile throttle looks like it would be VERY useful for a project of mine.

[whodwho]

http://www.rhdengineering.com.au/

Who is RHD? That low profile throttle looks like it would be VERY useful for a project of mine.

[The Dark Side of Will]

Thanks!

[digger]

Currently working on a plenum/airbox for the ITB to ensure that cool air is fed to the engine. IAT of 65 to 70C in traffic on stinking hot summer day are not great nor is the flow and filtration of foam socks....

ive made a extruded polystyrene mould that fits into the engine bay. just going to wrap it with fibreglass weave cloth and a post curable epoxy resin so it doesn't soften at 50C like most epoxy resins do. i should be able to get the wall thickness quite thin but not like carbon would be
the trumpets go on the inside so it needs a removable lid.

the foam will be dissolved with acetone

im still not 100% certain that all trumpets will fit inside and that there will be enough clearance to the booster and hood with movement of the engine on the mounts so thats why i am not using carbon fibre or anything exotic. i should be able to get it to work though.

[mccor023]

Im so stoked your putting a build together, Ive admired your engine since seeing your sig on e30tech when I first got mine. I look forward to reading through all this.