The BMW M20 engine

[Erland Cox]

One thing I noticed on the rolling road were I test is that the losses seem to be constant to the rear wheel speed.
And they go up a lot with speed, from 30hp at 150 km/h to 60hp+ at 200 km/h.
But we had the cars tied down against both the rollers.

Erland

[1jzracing]

I think the power is completely achievable, however dyno dynamics in shootout mode is a bit of a happy number setting because it uses quite a fast ramp up so engines with a light reciprocating mass often come up a bit generous. Interested to see the results on a mainline dyno. I have a healthy m20 too with the best of 200rwkw (265Hp) on a fairly straight forward 2.7 build with lots of attention where it counts but it only has 731 head with 42mm valves etc. K/Engines Your customer might be interested in this roller cam conversion, I currently have the head off mine and the first prototype is going in over the next week or so. 1st cam we designed has 315 seat, 255 @ 050 net, 12.7 net lift and 106c/l , its a bit smaller than you guys use but the roller allows faster opening so the lift profile is much fatter in the upper areas so should spread power well. Had the head on the flow bench the other night too, files are still at work so will post up soon

2012-11-07_10-01-08_83.jpg

2012-02-12_16-27-04_503.jpg

[1jzracing]

sorry screetch been a bit of a hijack, at least its still all m20 and back on to some port flows

picked up flow graph today, dont know how to post up so its visible maybe someone can help
Port flow for my engine to make 265rwHp is 180 cfm @ 12mm lift so it shows big flow isnt everything, I also had a bit of a play with a stock head when i flowed it.
have to ignore the 1mm lift point as the flow was too small for the sensors to pick up a stable signal

m20 731 port graphs.JPG

Purple- is the race head 42mm SS race valves quite a bit of port work done by banyard
yellow- stock 731 std 40mm valve 140sh cfm
orange- stock 731 with 42mm race valve 45/65 seat cut and only the lower cut lightly blended in
red- with throat opened out a little more, went harsh low lift flow dropped off badly (this is what most head reco shops would have done)
green- some attention to the transition and ssr, now things really start to come good.
blue- more of a full pocket port tidied up around the guide etc. not a lot of metal removed and its already matching the race head for max flow.

most interesting thing i took away from the afternoon playtime is how sensitive adding a bigger valve is if its not done well. I havnt done a lot of flow bench work but i think about 30% gain is pretty significant!

[Erland Cox]

I did a 80 M20B20 head that I put in a 42mm valve in just to see what I could get.
A guy that drives rally wanted to know if I could better his head so I Pipemaxed his 11,5mm valve lift.
The port is 34mm wide and 32mm high and the flow values at 28" from 2 to 16mm are:

49
85
122
153
166
172
177
178

Hopefully with 45:s it could make 250 engine hp with a head like that.

Erland

[KnightEngines]

I'll just leave these here for now:

[Erland Cox]

Very good looking but you can make some gains by keeping sharp demarcations between the lower valve cuts on the intake side.
60-75 and maybe 85 degrees. Never grind into the ridge under the 75 degree angle.
It helps to turn the flow and to reintroduce fuel into the air.

Erland

[KnightEngines]

The angles are there, just sanded over - fuel shear comes from the angle change, not the sharp edge.
Sharp edge seats work well on typical V8 ports where the MCSA is at the pinch & the air has to turn hard into the bowl & slow down - fuel tends to drop out more readily in ports like that.
With high & straight ports you can use a lot more velocity & don't have to slow the air to make it turn as much - fuel does not fall out as readily & you do not require a sharp edge valve job, in this case the blended valve job will make more power.

Valve job is 35/45/65/80 & works very well on these heads.

You can see the long 65 cut - finishes just before the bottom of the seat insert, if you look you can see where it meets the 80 cut.
Ridge under the seat is long gone - 45mm intake valve.

No problem turning the air - port is dead stable out past .750" lift - varies less than 1cfm on the flowcom & sounds very nice.

Head flows 211cfm with the manifold bolted up, a bit less than the last one but this is an injected street engine & won't turn as hard.

[KnightEngines]

Intake flow:

Lift - bare head - with manifold
.1 - 55 - 54
.2 - 113 - 111.5
.3 - 161 - 159.5
.4 - 193.5 - 191
.45 - 197.5 - 196
.5 - 202 - 201.5
.55 - 206 - 206
.6 - 209.5 - 208
.65 - 212 - 210.5
.7 - 212 - 211
.75 - 212 - 211.5

[KnightEngines]

1JZracing - you know the owner of this one, he's gonna use one of your roller cams.
What lobes do you have now? - I'm thinking mid 260's duration with a small backsplit, have to run some numbers but maybe 265/262 on 106 with .550" intake lift & around .530" exhaust.
Head can take up to .600" lift, maybe a tad more, so no need to be shy with the cam.
Is Banyard still working? - someone told me he opted out? - I've talked with Mark a few times, not a dumb guy

[1jzracing]

Knight engines- ill have to give you a call sometime be good to chat, Banyard is still going as far as i know havnt caught up for a while, I have found some quite handsome gains over the head he did with a bit of persistance and bigger valves. Ive spent months of playing around after work when i get a chance and even with 44 valve i got to similar cfm and some good numbers up to that too. No matter what i did a 36 exhaust valve was always better than a 37 by quite a lot so thats my final valve size 44 and 36, my head is only a factory 731 casting tested with a 12mm radiused entry plate and no manifold, Its 1mm lift points but switched to inches and true 28 inch depression, Flow Quick system.

0.0787 45.5
0.1181 69.8
0.1575 90.3
0.1969 114.1
0.2362 135.3
0.2756 154.9
0.3150 172.9
0.3543 186.4
0.3937 196.0
0.4331 202.9
0.4724 207.7
0.5118 211.9

The 1st roller cam profile has quite large ramps but high velocities its about 110 to 115 seat duration depending on the lash and 257 duration at net .050" total lift is net .500 and C/L is 106. It would be no problem to stretch lift to at least .550, it opens so quick it almost has dwell at full lift so we could easily add more lift over the nose and keep the lower part the same.

What sort of seat duration did the race engine have?? My feeling is that the seat duration on mine is a bit much for the 11.3 compression it was running and thats too low for 310+ seat duration would this thinking be correct?? How badly does large seat duration affect the performance potential?? What does everyone else think about that relationship? and at what distance off the seat does the engine actually "see" as the open/closing point??

Installed we made good power 240+RwHp but the much smaller 288 cam made pretty much the same power within 3-4Hp from 5500 to 8000 and unsurprisingly the roller was worse below 5500. Mind you this comparison was made with a tired engine and a lot of miles between dyno tests so not a true back to back. New eng will have better head more comp and more cubes so hoping components will all compliment a bit better

[1jzracing]

some 731 pics

[KnightEngines]

Improved production racing round at Phillip Island recently, starting from rear of field, goes arright:

http://www.youtube.com/watch?v=bthhTBNbgg4#t=246

[Cubic_Cleveland]

Nice work! Sounds good too!

[KnightEngines]

Another one done, this one has behive springs, RHD roller cam, rocker arms & intake manifold

[Erland Cox]

Really nice looking camshaft and rockers.

Erland