Big Valves

[petert 588]

I finally had a std. head flowed and I've put all the figures together on one graph. The 29.5mm exhaust valves had a nice 3 angle seat cut. The 34.5mm inlets are standard, with that nasty single 45 deg. cut and a straight drop into the bowl, followed by the ridge at the bottom of the seat. The 36.4mm inlet has a nice 3 angle seat, blended into the bowl with no ridge. Maximum lift of the std. cams is approx. 0.360".

 

I'll let you make up your mind.

 

[Dom9 2]

But what would the standard valve curve with nice 3 angle seats and blended throats look like on that graph?

 

Must be very similar to the big valve!!

[Mattsav 1]

Peter,

 

what shape is the big valve used in those tests, i.e tulip shape as std or flat backed.

 

The graph is very similar to what I've found with other people's big valve heads, less flow to 0.200" then it starts to gain.

 

If fitted right a big valve head should shows gains off the seat due to the bigger circumfrence of the valve.

 

I've found gains off the seat by altering the port shape and using the Longmans 35.5mm valves but some 36.5mm flat backed valves I made up flowed very similar to what you have there.

 

I've also found that if you use the right seat angles you can get almost exactly the same gains as in your graph without touching the inlet port at all.

[petert 588]

But what would the standard valve curve with nice 3 angle seats and blended throats look like on that graph?

But that's the trouble with the standard setup. There's not enough seat material to make a 70 deg. bottom cut . The throat goes vertical from the bottom on the 45 deg. seat.

 

I can't recall if they were tulip or flat. Most likely flat though.

Edited October 20, 2004 by petert

[sorrentoaddict 1]

if the above graph is anything to reckon on, than I guess the discussion about

std. valve vs. big valve MI16 should be put to an obvious end.

 

Even if (theoretically) you cut a nice 3-angle seat on the std. valve size MI-16, the flow would (maybe) be similar to, but still somewhat lower than the big valve flow values.

It means that even with std. cams, the benefits of the big valve are there (even if, with std. cams, 80% of this benefit is from the beautiful seat angle shape).

 

And with ANYTHING with at least a little noticable more lift than std. cam, the BV head is sure to make an increased power.

 

2 things that will never be clarified, however, are:

 

1) whether the loss of gas speed could possibly deteriorate the lower-rpm delivery, throttle response, and overall drivability (a marvellous feature on the std. MI-16, as opposed to many std. fuel-injected 16-valvers).

 

2) for out-and-out racing engines the compromises you have to make with limited "cam-timing-playfield" are maybe too big in order to justify the obvious gains to be had in high-valve-opening situations.

 

 

For mildly-to-very-advanced tuned road engines, Mr. Baker seems to be fully right.

(As I don't see any arguments against a BV head, except the obvious downside to it - that only very few people out there have the experience to do it PROPERLY, i.e. to really work and not decrease power, as it can so often happen with ill-executed BV conversions).

 

 

Cheers

[sorrentoaddict 1]

forgot to add this:

 

actually, from this graph it can be seen that the BV will work proportionally better as valve lift increases - but, in reality, it is more of an issue that the max. lift or close-to-max-lift is maintained longer, so what springs to mind is that really longer DURATION is what you are after with a proper BV conversion.

 

And here is the catch IMO: if you use too wild a duration in order to make the obviously very big gains from the BV, then there is even LESS possibility to play with the cam timings, as longer duration means even more likelihood that the intake valves (being MUCH bigger) will "have a meeting" with the exhaust valves at some point.

 

Consequently (I might be fully mistaken anyway), there will be available only a very narrow range of cam timings that will be safe, thereby (most probably) compromising the power delivery at some rpm-range....

 

maybe this is the logic behind the thinking of that gentleman that builds crf450's engine.

 

(this is just a theoretical opinion, as I myself have never "touched" a Big Valve MI-16 head before).

 

 

cheers

[Shevy 4]

Peter,

 

what shape is the big valve used in those tests, i.e tulip shape as std or flat backed.

 

The graph is very similar to what I've found with other people's big valve heads, less flow to 0.200" then it starts to gain.

 

If fitted right a big valve head should shows gains off the seat due to the bigger circumfrence of the valve.

 

I've found gains off the seat by altering the port shape and using the Longmans 35.5mm valves but some 36.5mm flat backed valves I made up flowed very similar to what you have there.

 

I've also found that if you use the right seat angles you can get almost exactly the same gains as in your graph without touching the inlet port at all.

Matt,

 

Would my B/V head that you fixed, benefit alot from a cam running more lift than the Kent PT1603 I've got fitted, looking at that flow chart it would seem to be the case ?

Also, how easy would it be to test the flow on my head ?

 

Cheers

Kev

[Mattsav 1]

At the moment i've got 2 problems:

 

1) i cant find the paperwork from when I plotted the profile of your kent cams

 

2) My laptop has gone to be fixed, I've stuck the drive into another PC but the flow bench software has flatly refused to open the data files for some reason.

 

If only I was organised......

 

Off the top of my head the 1mm duration of the PT1603's is either 260 deg or 230 deg, 260 wouldn't work with hydraulics so it has to be around 230 degrees (simialr to the 273 deg Catcam Inlet.

I wrote the measurements down it but didn't have time to pay much attention to it, I want to find it and have another look as 230deg@1mm seems to be far to little for a 290deg cam (they were definatley 284 @0.1mm as I double checked incase the dial guage had moved.)

 

As soon as its turns up i'll work out the exact measurements.

 

I obviously couldn't resist flow checking your head and if at some point PissyWorld decide they can fix the laptop then i'll send you the graph's. Its very similar to what Peter_t has, looses flow up to 0.200" then gains. It was something like a 7% gain under the graph up to the lift of the cams but i'll double check when I can.

 

The interesting thing with the mi16's is the poor low lift flow.

The Vauxhall Ecotecs (1.4 & 1.6) manage almost the same flow up to 0.200" but with a pair of 31mm valves.

 

Puma may or may not agree with me (probably not ) but i'd say the tulip shaped valves are a better bet for the mi16.

 

In fact I keep coming across engines where people manufacture flat backed valves but the ports seem to like the a more tulip shaped valve. The saxo VTS heads repeadedly show an increase in flow if you go from a flat backed valve to a more rounded shape due to the angle of the port, whereas the Ecotec prefers a flat backed valve as the ports are at a different angle.

 

Or maybe i'm full of crap who knows

 

The next step for me is to get the car back on the rollers now its got a pair of Catcams in it and then try a head with the 35.5mm valves (as I've got a set)

[smckeown 1]

The next step for me is to get the car back on the rollers now its got a pair of Catcams in it and then try a head with the 35.5mm valves (as I've got a set)

yes half the country's mi16 drivers are waititng big time ...so hurry up

[PumaRacing 2]

There's a general rule that horizontal ports with abrupt short-side bends (Mini, MGB, Pinto) like flat backed valves and the more downdraft the port is the more tulip shaped the valve can be. However, you can make flattish backed valves work very well in downdraft ports if you get certain details of the seat and back angle right and you save a lot of valve weight compared to a tulip valve which shows up as reduced valve spring poundage and consequently reduced parasitic power losses. There's actually a "magic number" for the angle of divergence of tapered sections which leads to best flow and this same number works very well as the angle on the back of many valves.

 

Every valve I design has the port angle taken into account in order to determine the optimum shape for the valve but this shape might not always be the best for flow at a given lift. There are many other factors to be considered. Also the back angle on an inlet valve is always different to that on an exhaust valve.

 

As to reduced flow at low lifts with a bigger valve - back to the flowbench I think because there's no excuse for this to happen if the valve shape and seat shape are correct.

[Shevy 4]

At the moment i've got 2 problems:

 

1) i cant find the paperwork from when I plotted the profile of your kent cams

 

2) My laptop has gone to be fixed, I've stuck the drive into another PC but the flow bench software has flatly refused to open the data files for some reason.

 

If only I was organised......

 

Off the top of my head the 1mm duration of the PT1603's is either 260 deg or 230 deg, 260 wouldn't work with hydraulics so it has to be around 230 degrees (simialr to the 273 deg Catcam Inlet.

I wrote the measurements down it but didn't have time to pay much attention to it, I want to find it and have another look as 230deg@1mm seems to be far to little for a 290deg cam (they were definatley 284 @0.1mm as I double checked incase the dial guage had moved.)

 

As soon as its turns up i'll work out the exact measurements.

 

I obviously couldn't resist flow checking your head and if at some point PissyWorld decide they can fix the laptop then i'll send you the graph's. Its very similar to what Peter_t has, looses flow up to 0.200" then gains. It was something like a 7% gain under the graph up to the lift of the cams but i'll double check when I can.

 

The interesting thing with the mi16's is the poor low lift flow.

The Vauxhall Ecotecs (1.4 & 1.6) manage almost the same flow up to 0.200" but with a pair of 31mm valves.

 

Puma may or may not agree with me (probably not ) but i'd say the tulip shaped valves are a better bet for the mi16.

 

In fact I keep coming across engines where people manufacture flat backed valves but the ports seem to like the a more tulip shaped valve. The saxo VTS heads repeadedly show an increase in flow if you go from a flat backed valve to a more rounded shape due to the angle of the port, whereas the Ecotec prefers a flat backed valve as the ports are at a different angle.

 

Or maybe i'm full of crap who knows

 

The next step for me is to get the car back on the rollers now its got a pair of Catcams in it and then try a head with the 35.5mm valves (as I've got a set)

Excellent, I was meant to ask you when you had the head could you flow test it,but I completely forgot !

 

I look forward to seeing the results.........or maybe not !

 

Once my mapping is finished Matt I'll let you know my R/R results, going by what I got before, if it's not much better I'll be knocking at your door, maybe looking at different cam options or what else can be done to the head/engine.

 

I'm just worried your gonna tell me the flow figures are poor and I'll need to start from scratch again with the head.

 

Cheers

Kev

[Guest stan]

Right! I am no pug expert....I dont even own 1! However, engine tuning and internal combustion engine fundamentals are not Peugeot specific.

The graph that was posted in this thread shows three different valve sizes. This graph is slightly misleading as it shows flow .V. LIFT. As a fair graphical representation it SHOULD show flow .v. lift divided by diameter (L/D). If these flow results were plotted this way, it would be seen that flow tails off at around lift/diameter = 1/4. This is because beyond this lift value, it is the valve that is becoming the restriction rather than the curtain area i.e. if u lift the valve anymore, little more flow will be achieved (this can be proved mathematically but I havent got the symbols! ). This explains the diminishing flow shown on the graph on the previous page. This understanding is the basis of when big valves should be used. Big valves WILL give more power. Power is proportional to airflow through the head. FACT. Therefore bigger valves will give more peak power. However, an appropriate cam must be used i.e. one which allows enough lift for maximum flow to be achieved. This is where geometric limitations come in. An engine that is being built to produce high levels of peak power will/should have an increase in CR and unless funky forged pistons are used, a compromise between valve lift and CR has to be made. In any instance such an engine tuned in this method to produce good peak power will be shocking to drive on the road. lets face it, no matter how much of a rally/race/track-day driver we consider ourselves to be, on the road we still need torque at under ~ 3000-3500 RPM. With these size valves, low gas speed at low engine speed will give poor combustion and inherently poor torque. Also with the hi-spec cam required to get the full potential of large valves there is going to be increased duration (and I dont mean just increased as a matter of course, but due to keeping valve acceleration to a minimum longer duration is required with hi lift). So again at low engine speed/throttle opening the unburnt gas and residual gas levels in the chamber are going to increase, giving further poor combustion and so torque and inevitably power will be reduced.

 

On this basis, the argument for big valves should be down to application. Big valves=poor mid range, or compromise a bit of PEAK power for a lot of mid-range and have a very usable power curve.

 

Craig

[sorrentoaddict 1]

this provokes some thoughts again:

 

if this 1/4 ratio of lift / valve head dia. is applied to the std. XU9-J4 34.6 mm inlet valve head size, it seems that anything above 8,65mm lift is practically useless.

 

Having in mind the std. 9.20 mm lift on this engine, and having in mind many reports that we have seen about the fact that duration is much more beneficial to this engine than simply lift increase, it might be true.

 

Someone (called petert, if I remember correctly) has even developed a cam grind for this engine that actually lifts the int. valve LESS than std., with allegedly noticable effects, so this might be very true.

It might also support the concept of Mr. Baker (PUMA RACING) that BIG VALVES (if properly executed) are the best way of making more output out of this engine.

 

What would be still not so clear, however, is how does the extreme 12-13mm valve lift (used by Longman in BTCC and (sort of) french tuners PIPO Moteurs in 306 Maxi et al.), translate into cca. 300 BHP with the std. valve size ?

Obviously, one would guess from the many answers that it is the CR that conditions this, what with it in these engines being much higher than what is normally used on the MI-16 even in advanced fast-road guise.

 

But then again, the necessity to play a LOT with the camshaft timing (and the obviously necessary very high lift) in order to establish proper gas cycle when you have a very high CR engine - might be the reason why a std. int. valve size is used at the first place - actually the limitation of the valve timing adjustment that is brought by the Big Valve concept, might put the engineers away - as they have calculated that there is more to be had with proper valve timing at those levels of rpm.

 

For a street-going engine, where 300 BHP is not really the goal, the BV option should be more beneficial in terms of power-per-dollar at the end of the day - whether it really creates a significant loss of power in some part of the rev range, is another discussion altogether -- and as I have never made a BV-head MI-16 myself, I leave to others to comment.

 

And anyway, there is this issue remaining about whether the power increase of the BV is maybe simply due to the fact that on a std-seat-size MI head it is not possible to cut a 100% proper 3-angle seat, and on a bigger valve seat it is possible.

 

 

Cheers

[Mattsav 1]

If you look at the flow graph posted by peter you'll se the flow is still increasing way past the lift of the std cams.

 

I've only tested the inlet flow to 13mm but it still hadn't levelled off

[Robsbc 0 1 Cars]

Think Longman used 35.5mm on their mi16 engines...Well they do for road going mi16 conversions..

[petert 588]

The two cams I have, both regrinds, have only 0.350" and 0.360" lift. While the inlet port does flow more past these points, I believe there's little point in doing so for Joe Average. 260cfm is a lot of flow, my Stage II grind makes 215-220hp on TB's. The other important point is that lifting the same as stock doesn't put strain on the springs and cam lobes. How many people put in big cams, don't measure the open height, then whinge about broken springs?

 

If you look at the exhaust port, also exceptional flow figures, you'll notice that it goes DEAD FLAT after the std. lift point. So it's absolutely useless lifting it any higher than that.

Edited November 7, 2004 by petert

[base-1 17]

The other important point is that lifting the same as stock doesn't put strain on the springs and cam lobes.

really? surely that depends on the cam profile, if it is opening and closing the valve alot faster/more aggressively then it will strain the springs etc

[PumaRacing 2]

The two cams I have, both regrinds, have only 0.350" and 0.360" lift. While the inlet port does flow more past these points, I believe there's little point in doing so for Joe Average. 260cfm is a lot of flow, my Stage II grind makes 215-220hp on TB's. The other important point is that lifting the same as stock doesn't put strain on the springs and cam lobes. How many people put in big cams, don't measure the open height, then whinge about broken springs?

 

If you look at the exhaust port, also exceptional flow figures, you'll notice that it goes DEAD FLAT after the std. lift point. So it's absolutely useless lifting it any higher than that.

The lift at which a valve stops flowing is not relevant in determining an optimum cam profile which should always lift past this point so that the valve spends as much time at or above this point as possible. In fact if the cam lobe acceleration and peak velocity stay the same or higher than standard then any increase in duration must lead to an increase in lift. If not then the acceleration or velocity has fallen which means the profile is wasting lift needlessly and is now a "bad" profile in that it could be better designed to have less duration, more lift and the same area under the flow curve but with better low rpm tractability.

 

"Strains" is a rather nebulous word. The forces acting on a cam lobe are determined by the acceleration generated by the lobe times the mass of the valve train it is opening. If the accelerations don't increase then nor will the forces but the lift will still increase quite safely with extra duration. Newton's Laws of Motion apply just as much to cam lobes as they do to anything else.

 

Certainly any valve spring must be checked for coil bind but OE setups rarely go very close to this point, not least just to have margin in hand for manufacturing tolerances.

 

As a general rule, any cam with more duration than std but no extra lift is badly designed although reprofiled cams often have to take this route because of lack of material to play with on the lobe.

[Dom9 2]

The lift at which a valve stops flowing is not relevant in determining an optimum cam profile which should always lift past this point so that the valve spends as much time at or above this point as possible. In fact if the cam lobe acceleration and peak velocity stay the same or higher than standard then any increase in duration must lead to an increase in lift. If not then the acceleration or velocity has fallen which means the profile is wasting lift needlessly and is now a "bad" profile in that it could be better designed to have less duration, more lift and the same area under the flow curve but with better low rpm tractability.

 

Ok, this makes sense to me in terms of 'performance cams' and the difference between these and standard cams in terms of area under the lift vs. duration graph...

 

Would it also be useful then for performance cam manufacturers to quote something like 'duration above standard OE manufacturer cam lift' additionally to the standard 0.1mm (or is it 1mm?) duration data?

 

e.g. If a performance cam had a max lift of 11.5mm and a standard cam 9mm, I guess we would have an even better idea of how it will perform if they said 'it has a duration of 30° above OE manufacturer lift'... Or is that pointless or overkill?

[petert 588]

"Strains" is a rather nebulous word.

 

As a general rule, any cam with more duration than std but no extra lift is badly designed although reprofiled cams often have to take this route because of lack of material to play with on the lobe.

 

Sorry, I used one word to describe two different problems. I knew I'd get picked on. The point that needs to made is that the Mi16 has very small lobe base circles and small diameter buckets, compared to say an 8V. This means the rate of accel'n rate of the lobe is far more, with the tendancy to "smack" open the bucket, with excessive lift and/or badly designed profiles.

 

I don't agree with your comments about lift and duration. The dyno figures we're getting from regrinds with std. lift speak for themselves.

 

What's the point opening past the maximum flow point if you can do the same with less lift? Sure the duration at maximum flow is important but if you can do it with less lift, why not? Opening the spring more is a waste of energy.

 

I really can't see the point of high lift cams in Mi16's destined for fast road use. When you open the valve another 2-3mm, that just wasted spring pressure and a formula for spring bind and early lobe/bucket wear.

 

If it's a 250hp, track only car then yes. But when you get reliabley get 215-220hp using a std. exhaust cam and a regrind on the inlet, why bother? (btw. that's with big inlets too!)

Edited November 9, 2004 by petert

[taffycrook 2]

Peak lift is such a small amount of time, that it makes sense to lift beyond this point, but if the power expected requires less flow than max possible flow then it will have no effect.

 

However will this have the effect of reducing max possible VE and reducing the torque available which is more important than outright power.

[Guest stan]

Peak lift is such a small amount of time, that it makes sense to lift beyond this point, but if the power expected requires less flow than max possible flow then it will have no effect.

agreed, you need max flow for longest poss time so lifting beyond L/D = 1/4 achieves this.

 

i dont get what you mean by: "but if the power expected requires less flow than max possible flow then it will have no effect".....if u have more flow than the flow required for 'X' BHP, then u will have more BHP.....

[petert 588]

To give you an idea of how good an Mi16 head is (I know you already know), have a look at the flow data from some of these aluminium small block V8 Chev racing heads, then compare it back to the graph I posted earlier.

 

http://www.protopline.com/racingaluminumsbc.htm

 

Note they were also tested at 28 in, so you can make a direct comparison.

 

and what about these Honda heads at 0.500" lift?

 

http://www.protopline.com/v-tec.htm

Edited November 20, 2004 by petert

[Martin@PRD 0]

ok after reading all of this thread about bvh and cams etc, what clylinder head work would you recommend for my cam ive just bought below

 

kent PT1604 Supersports R Hydrolic deg' 316 lift 10.84 p/band 3500-7500

 

i am currently using 45 webbers, lumenition 3d ignition map (hall effect), wizard of nos contoller up to 200bhp (90-100bhp max i would use), copper head gasket, forged pistons and conrods (next couple of months), .9kg lighter flywheel, arp bolt though out engine.

 

i will be going to solid bucket and d/springs if i come to a point were i want to rev over 8k which i doubt i will with nos, lmao.

 

ps how do you convert hydro buckets to the solid type?

 

cheers

[hiflow 1]

I have done an extensive session on a Superflow flow bench & found that the larger inlet valves in the Mi16 are not really worth it.

 

Here is a table of our findings.

 

A INLET PORT 34.6mm (std) valves/ported with guides in

B EX PORT 29.6mm (std) valves/ ported with guides in

C EX PORT 30.9mm O/s valves /HFH Profile throats / guides in

D EX PORT 30.9mm O/s valves /HFH Profile throats/ guides in + HFH ported manifold

E EX PORT 30.9mm O/s valves /HFH Profile/ guides in + mi16 std manifold

F INLET PORT 34.1mm (U/s) valves, port cleaned up / seats blended into throats

G INLET PORT 35.5mm O/s valves , port cleaned up. Seats blended into throats

H EX PORT 31.5mm O/s valves /HFH Profile/ guides in

 

 

INLET PORTS

 

lift 0.100" 0.200" 0.300" 0.400"

 

Std Inlet 43.4 82 108.2 122.6

A 44 79 108.3 126.1

F 28.8 69.6 102.3 126.7

G 43.3 80.1 110.9 131

As you can see, the larger valves hurt flow below .200" lift (when compared to std valve). If you went for larger valves than 35.5mm, this figure would be worse. You can also see, that its only above .300" lift that the larger valves just start to work. Its is cleary evident that the ports are to big & the 1905cc bottom end is to small for the head port Dia. In my view larger inlet valves are not required on a std plenium. you will only rep there full potensial when running carbs/TB. I will also add that the larger valves did not flow any better over 0.440" of lift, so mega lift cam are not the quick fix answer. The valve spec on valve G , are 6mm stem with a wasted stem, leaving a 5mm stem protuding into the port.

 

EXHAUST PORTS

Std Ex 23 55.5 86.3 101

B 38 74.2 93.8 105.7

C 45.4 87.3 113.1 125.4

H 43 91.1 111 125.7

 

As you can see, fitting oversize valves + changing the port profile give over 20% improvment in flow over the std setup. Again, as you can see in H, valve are getting so big , that valve shrouding is hurting flow at initail valve lift.

 

 

Exhaust Manifold

D 45.4 83.6 101.2 109.6

E 44.4 78.7 93.2 98.4

 

With the above, we have bolted the two manifold onto 'C' configuartion head . Note that the ported head with the HFH manifold fitted flows better than just ported head.

 

 

As the most of our customers are running 1905cc engine , therefore all out initail dyno testing will be done using this block. After this we aim to repeat the above using 2088cc & 2149cc bottom ends.

 

Please note the superflow has its our PC programme linked into the machine, the PC simlpy instructs you which plugs to remove maintain the correct vaccum throught the test.

 

Im my view the larger bottom end will, will improve the low down flow figures on the inlets ports, BUT only time will tell.

 

All prices of valves/maifolds/heads have just been added to our website

 

 

 

Andy

Edited August 13, 2005 by hiflow