37 replies to this topic

[manolis]

At http://www.pattakon.com a new Desmodromic VVA (Desmodromic Variable Valve Actuation system or DVVA) has been added.

In the Lost Motion VVAs (like BMW’s valvetronic and Nissan’s Infinity VVEL) each lift is coupled with one only duration and vice versa.
Every time you pick a lift, you have to compromise with its respective duration.
And every time you pick a duration, you have to compromise with its respective lift.

In the Constant Duration VVAs (like the Honda’s A-VTEC), the duration is the same for all lifts.

In the Desmodromic valve trains, like Ducatti's, there is no need for valve springs.


The pattakon DVVA combines the advantages of all.

Take a look.

Thanks
Manolis Pattakos

[Powersteer]

That looks very similar to Felik's crank.

[F1Champion]

A quick question. The Ducati engine revs very high and arguably produces the most power of all the MotoGP engines. Can the Desmo valve system be used by F1 engines and would it be better than the pneumatic valve system?

[Canuck]

I don't know that your statement is quite accurate about MotoGP. As for F1 adaptation - Ferrari has a Desmo head in their museum as per one of (our) Desmo's photographs. My guess is for whatever reason, the current system was deemed better (or regulated into mandatory).

[manolis]

Originally posted by Canuck
I don't know that your statement is quite accurate about MotoGP. As for F1 adaptation - Ferrari has a Desmo head in their museum as per one of (our) Desmo's photographs. My guess is for whatever reason, the current system was deemed better (or regulated into mandatory).

At long straight parts Ducati is proved the winner. Provided the aerodynamic resistance is similar, only power can make such difference, either it comes from higher revving or from more efficient breathing (probably both).

If you compare the today Desmo of Ducati with the Desmo of Ducati (or Mercedes) some 50 years ago, they are very similar. The worst thing of today Desmo is the elasticity of the restoring rocker arm (a necessarily long L shape member) and the high speed on the periphery of the restoring cam lobe. Lab test say that the valve can make a few jumps before finally closes.

When the two control shafts of the pattakon DVVA are locked, the DVVA behaves as a pure desmodromic system. All quick moving parts (which have to be lightweight) are free from bending loads. And the maximum speed of the roller ROLLING on the outer control shaft is by far lower than the maximum SLIDING speed met in conventional Desmo. And instead of high speed "line contact" on the periphery of two cam lobes, the DVVA uses "crankshaft" ans surface contact (except the yoke roller). This means ability for higher revs than Ducati motoGP Desmo (which is not far from the 19000 rpm rev limit used in F1).

Now suppose that instead of having a throttle valve, the rotation of the inner (constant duration) control shaft is used to control the engine with the DVVA (just like in the Pattakon Honda Civic VVA prototype). This means more free breathing at high revs and at the same time perfect response everywhere, better economy everywhere and low idling - easy starting.

By the way, download and run the http://www.pattakon....mpman/DVVA1.exe . Here the valve lift profiles are more aggressive (racing) than the valve lift profiles of the DVVA.exe .

Thanks
Manolis Pattakos

[Bloggsworth]

Originally posted by F1Champion
A quick question. The Ducati engine revs very high and arguably produces the most power of all the MotoGP engines. Can the Desmo valve system be used by F1 engines and would it be better than the pneumatic valve system?

The "straight eight" Mercedes F1 engines of 1954/5 used desmodromic valves, but they probably peaked at about 1/3 of the revs of even a modern rev-limited engine. They also had very cunning screw together crankshafts to enable them to use roller bearings.

[Joe Bosworth]

F1Champ posts:

"A quick question. The Ducati engine revs very high and arguably produces the most power of all the MotoGP engines. Can the Desmo valve system be used by F1 engines and would it be better than the pneumatic valve system?"

F1 engines are revving to 19000 and in past years ahve exceeded this.

The Moto GP Ducatti is revving to 18000.

It appears that pnuematic closing allows more revs and therefore will ultimately provide the basis for making more power.

Desmo for motorcycles is no doubt preferred over pnuematic because of the simpler ancilaries that add weight and space to the m/c. Not a problem for F1 due to the ancilaries having other uses around the car.

Regards

PS: Agrueably, or at least alledgedly, the latest Honda Moto GP puts out more power than Ducati.

[Ray Bell]

Originally posted by Bloggsworth
The "straight eight" Mercedes F1 engines of 1954/5 used desmodromic valves, but they probably peaked at about 1/3 of the revs of even a modern rev-limited engine. They also had very cunning screw together crankshafts to enable them to use roller bearings.

What must be remembered about these Mercedes engines is that they were never developed...

M-B dropped out prematurely, partly because of the Le Mans disaster, it can be assumed, and partly because they had been so dominant that they'd proved their point.

But developments were under way that could have seen a great deal more come from their engines. They were looking towards 10,000rpm and they were working on variable length inlet tracts at the time they pulled out. You can read a little about this in The Design and Behavior of the Racing Car (Moss & Pomeroy).

As for the cranks, these were an anachronism, it would have to be said. Heavy, complicated, potentially troublesome, they were a throwback to the old days. Ballot in 1919, for instance, had this setup, as did many other of those forerunners to modern engineering.

[Moon Tricky]

I like desmodromic valves, and a version with rollers is also very appealing. But with the variable timing, it is getting very complex...

Is it not plausible to use some sort of electromechanical system? The problem I've heard stated is that it takes too much electrical power to overcome the spring, but if desmodromic valves don't need one, I don't see why we need one here either, as long as the solenoid can close the valves as well as open them.

[desmo]

This old thread might be worth having a look at regarding electromechanical valve actuation.

http://forums.autosp...=&threadid=1723

[manolis]

Originally posted by Moon Tricky
I like desmodromic valves, and a version with rollers is also very appealing. But with the variable timing, it is getting very complex...

Is it not plausible to use some sort of electromechanical system? The problem I've heard stated is that it takes too much electrical power to overcome the spring, but if desmodromic valves don't need one, I don't see why we need one here either, as long as the solenoid can close the valves as well as open them.

Having a Variable Valve Lift / Timing Desmodromic system you get rid of throttle valves, achieving more free breathing at full valve opening / high revs. The engine can provide plenty of torque from very low revs. The starting of the engine is way more easier (at very low lifts you actually have no overlap) and the consumption way lower.

In present motoGP Desmo, for each pair of valves you need: four cam lobes (the restoring ones having large diameter), two opening and two long L shape restoring rocker arms and two long valves. Anything but simple.
The valve lift profile is one and only one for all conditions: either the engine operates at 18000 rpm or it idles at 5000(?) rpm the valve lift, the valve duration and the overlap are exactly the same. To get top power, necessarily the good operation at middle / low revs is sacrificed.
To achieve the best you need different camshafts (i.e. valve lift profile) according the race track, according the specific conditions of the race track (wet / dry), even according the way the driver feels.

In the DVVA case for each pair of valves (which now can be way shorter) you need an eccentric pin, four articulated short/light/strong rods, a yoke roller with a roller pin, and two slow moving control shafts.
Using the constant duration control shaft as the basic control of the engine (i.e. like a conventional “throttle valve” as happens in case of the Honda Civic pattakon VVA prototype), you get rid of the throttle valves (because the intake valves make the throttling), you have torque along the whole rev range, you are less depended from the gear box, you can start the engine easily, you need less fuel etc.
The other control shaft (lost motion) can be rotated only when necessary: for instance, when the rain starts the driver can rotate slightly the lost motion control shaft (the way he now changes for a few clicks its suspension), or for instance when, during the last lap, the driver wants the maximum possible power from his engine.

As you see, the engine can be way better when the valve lift profile can change (as in case of DVVA), and the complication involved is not significant compared to the existing Desmo engines.
In the DVVA you do not need “drive by wire” but direct / instant response to driver’s commands: through a cable the driver rotates directly the constant duration control shaft (as he was rotating, until now, the throttle valves). No time delay. No strange feedback.

The electromagnetic VVA systems cannot, yet, control normal valves at normal engines revs. Read the Fully Variable Valve Actuation thread. Many large companies tried but failed to make a successful electromagnetic VVA for normal (nothing to do with racing revs and loads) use.

Thanks
Manolis Pattakos

[Bill Sherwood]

Originally posted by manolis
Having a Variable Valve Lift / Timing Desmodromic system you get rid of throttle valves, achieving more free breathing at full valve opening / high revs.

I'll have to partly disagree there - If the throttle butterfly is restrictive at all, then all you have to do is make the inlet a touch larger to compensate.
Whilst there's certainly a slight advantage to having a totally clear inlet path, in the vast majority of engines it's not worth the extra cost (and often complexity) over having a simple set of butterfly throttles.

[Moon Tricky]

Originally posted by desmo
This old thread might be worth having a look at regarding electromechanical valve actuation.

http://forums.autosp...=&threadid=1723

It's interesting reading, although that's from way back in 2000, surely there must have been a few developments in the last 7 years?

[Moon Tricky]

Originally posted by Bill Sherwood
I'll have to partly disagree there - If the throttle butterfly is restrictive at all, then all you have to do is make the inlet a touch larger to compensate.

I'm not sure I understand this. Surely the air is going to struggle to get through a small hole, irrespective of the size of any other hole it has already been through.

[phantom II]

On this BB a long time ago, was a picture of a DOHC L4 head, maybe Alpha.
The cam lobes were about 2" wide and each camshaft slid in extra long bearings back and forth with hydraulic actuation from the front end like an aircraft constant speed propeller hub.The lobes had complex shapes that varied timing and lift depending on camshaft position. There was no throttle.

Originally posted by manolis



As you see, the engine can be way better when the valve lift profile can change (as in case of DVVA), and the complication involved is not significant compared to the existing Desmo engines.
In the DVVA you do not need “drive by wire” but direct / instant response to driver’s commands: through a cable the driver rotates directly the constant duration control shaft (as he was rotating, until now, the throttle valves). No time delay. No strange feedback.

The electromagnetic VVA systems cannot, yet, control normal valves at normal engines revs. Read the Fully Variable Valve Actuation thread. Many large companies tried but failed to make a successful electromagnetic VVA for normal (nothing to do with racing revs and loads) use.

Thanks
Manolis Pattakos

[cheapracer]

Originally posted by Moon Tricky


I'm not sure I understand this.

Must be us simple Australians!

If your hole is the right size in the first instance but then you put in one of those horrible butterflly thingys that reduces airflow by say 5%, then make the bloody hole 5% bigger to compensate!

[Moon Tricky]

Originally posted by cheapracer


Must be us simple Australians!

If your hole is the right size in the first instance but then you put in one of those horrible butterflly thingys that reduces airflow by say 5%, then make the bloody hole 5% bigger to compensate!

No, I still don't get how you can get more air through a small hole by putting a big hole in front of it, even if the big hole is the size of the whole world.

[cheapracer]

Originally posted by Moon Tricky


No, I still don't get how you can get more air through a small hole by putting a big hole in front of it, even if the big hole is the size of the whole world.

Yes, you are correct but who said the next hole was in fact smaller? I can list a number of cylinder heads that the inlet port area is greater the butterfly/venturi area.

Bills point and my support were in regard to a specific sentance - "Having a Variable Valve Lift / Timing Desmodromic system you get rid of throttle valves, achieving more free breathing at full valve opening / high revs" - maybe better to use the lack of throttle body/butterflies as an offset cost argument?

[Moon Tricky]

I happened across Wikipedia's article on steam engines the other day and the animation there made me wonder...
http://en.wikipedia....e_in_action.gif

Forget the wheel. If the whole thing were small, and the main piston were connected to an engine valve, and the control arm was connected to a solenoid, it would take a lot less electrical power than having the solenoid directly opening the valve, and the thing could be pressurised from the exhaust. Would that work? Or has it been tried?

[Ninja2b]

You could also move to barrel throttles if you wanted to avoid the flow restriction caused by a butterfly. We designed one for our formula student car, and for a given inlet diameter the barrel throttle improved flow area by 12% over the butterfly as I recall. I also recall it was worth an extra 3-4% in terms of top end power. Either way, it was definitely a lot simpler than a VVT system

[Moon Tricky]

Originally posted by Ninja2b
You could also move to barrel throttles if you wanted to avoid the flow restriction caused by a butterfly. We designed one for our formula student car, and for a given inlet diameter the barrel throttle improved flow area by 12% over the butterfly as I recall. I also recall it was worth an extra 3-4% in terms of top end power. Either way, it was definitely a lot simpler than a VVT system

It is surely a good idea, but how does it work? I've never heard of barrel throttles before and the internet isn't helping (although it can sell me one).

[Ben Wilson]

The ones I've seen have basically looked like a ball valve.

[Ninja2b]

A ball valve isn't too far from what we had, but instead of a ball it was a cylinder which rotated, supported by a bearing at each end. The cylinder axis was perpendicular to the inlet axis. The cylinder had a hole bored through it the same diameter as the inlet, so when it was fully open there was no impendance to the flow. CFD showed that at part throttle the butterfly was better, but that was a simple 'back of the envelope' style CFD calc, not sure how much faith to put in it. And besides, poor flow at part throttle wasn't a huge concern for us.

[Greg Locock]

What sort of silliness is this? In what way does a barrel throttle cause less restriction than a butterfly? Given what a thottle does, and all that?

[NRoshier]

I had a conversation some time ago with Roger Duckworth on his masters thesis on port barrel throttles: http://www.patentsto...escription.html
He stated clearly that there was no power advantage over a well designed butterfly system and that the butterflies were effectively invisible at lengths of more than 10 port diameters at the butterfly.
The advantages of PBT was as stated in the link above and in essence were do with part throttle economy.

[Ninja2b]

Originally posted by Greg Locock
What sort of silliness is this? In what way does a barrel throttle cause less restriction than a butterfly? Given what a thottle does, and all that?

Its a simple matter really. For a given inlet size, a fully open throttle butterfly has less flow area than a fully open barrel throttle. I will have to check out the link that NRoshier posted, but like I said earlier - on the dyno we noticed a top end power gain when we switched to a barrel throttle.

[McGuire]

Originally posted by Ninja2b


Its a simple matter really. For a given inlet size, a fully open throttle butterfly has less flow area than a fully open barrel throttle. I will have to check out the link that NRoshier posted, but like I said earlier - on the dyno we noticed a top end power gain when we switched to a barrel throttle.

They key being "for a given inlet size." If you can't get the required flow capacity at WOT with your butterfly valve, why not just bolt on a slightly larger butterfly valve? It will still be lighter, simpler, cheaper, less bulky overall and more reliable than a barrel valve of the same effective flow capacity.

The student series are interesting to watch in that there seems to be a considerable amount of annually reinventing the wheel. Not knocking it, I guess that's one of the lessons of the exercise. The time, resources and $$$ spent developing the barrel valve -- where could that have been spent more effectively?

[dosco]

Originally posted by McGuire
They key being "for a given inlet size." If you can't get the required flow capacity at WOT with your butterfly valve, why not just bolt on a slightly larger butterfly valve? It will still be lighter, simpler, cheaper, less bulky overall and more reliable than a barrel valve of the same effective flow capacity.

Indeed.

The student series are interesting to watch in that there seems to be a considerable amount of annually reinventing the wheel. Not knocking it, I guess that's one of the lessons of the exercise. The time, resources and $$$ spent developing the barrel valve -- where could that have been spent more effectively?

Heh. Ignorance is bliss, no?

Sounds like a "trade study" is in order. Problem is that not many people doing them.

[manolis]

A good VVA is, by far, better than the state of the art conventional valve train.

Honda S2000 has, for years, the top power concentration naturally aspirating production engine. It provides 120 PS/lit at 8600 rpm (i.e. at 24m/sec mean piston speed). No other production engine approaches this number. And it is equipped with the simplest VVA, the VTEC.


Schrick sells camshafts for S2000 Honda. I copy from Schrick’s catalog:

Intake Valves:
308 crank degrees duration, 12.5mm valve lift, 52 deg BTDC valve opens, 76 deg ABDC valve closes, 4.4mm valve lift at TDC, top valve lift at 102 deg ATDC.

Exhaust Valves:
300 crank deg duration, 12.2mm valve lift, 76 deg BBDC valve opens, 48 deg ATDC valve closes, 3.8mm valve lift at TDC, top valve lift at 102 deg ATDC.

Such a valve lift profile cannot work properly at medium / low revs and partial loads. To bypass this problem, the S2000 engine has, for medium / low revs, another mode of operation (it is a VTEC engine, i.e. it is a two step VVA) as follows:

Intake Valves:
244 crank degrees duration, 7.15mm valve lift, 6 deg BTDC valve opens, 58 deg ABDC valve closes, 0.46mm valve lift at TDC, top valve lift at 116 deg ATDC.

Exhaust Valves:
244 crank deg duration, 6.93mm valve lift, 58 deg BBDC valve opens, 6 deg ATDC valve closes, 0.5mm valve lift at TDC, top valve lift at 116 deg ATDC.

When the S2000 engine revs between 5500 to 9000 rpm, the wild cam lobes activate the intake and exhaust valves (it is the ‘high rpm’ mode), and when the engine of S2000 revs below 5500 rpm, the valves are actuated by the mild cam lobes (it is the ‘low rpm’ mode).

At the soft mode, the engine is driver friendly and torquie, the consumption is reasonable, the emissions are acceptable, and so on. When the driver decides so, he uses revs above the VTEC point (5500 rpm) and the engine is transformed into a true racing engine.

Spot on the difference of the valve lift profile (lifts, duration, overlap, etc) used in normal conditions compared to the “wild” valve lift profile.

For top power it is necessary extreme overlap (at TDC the intake valves are open 4.4mm and the exhaust 3.8mm), extreme valve lifts (12.5 and 12.2mm) and extreme valve duration (308 and 300 deg).
But if you keep this valve lift profile at medium / low revs and partial loads (for instance below 4000 rpm), the torque provided (as well as the emissions, the consumption, the response etc) is worse than the torque provided with the low rpm cam lobes!.
Practice proves that with a valve lift less than 60% of the wild mode valve lift (7.15mm/6.93mm instead of 12.5mm/12.2mm), with way less duration and with way softer overlap (0.4mm/0.5mm intake/exhaust valves open at TDC compared to the 4.4mm/3..8mm intake/exhaust valves open at TDC), the engine provides (at medium low revs) more torque, is more fuel efficient and true driver/environment friendly.

Is it possible, using the best ITB system – either with conventional throttle valves or with spherical or cylindrical or … throttle valves - to use only the high rpm mode of S2000 and still have an engine capable for both, racing AND normal use?


Now think the case of the S2000 Honda engine with the DVVA (desmodromic variable valve actuation system) on it.
Instead of having just two modes of operation and a step between them, there are infinite (continuously variable) modes of operation, with a great number of them being wilder than the high rpm mode of S2000 (to milk even more power), and with a great number of them softer than the low rpm mode of S2000. There are modes with less than 0.1mm maximum valve lift and modes with zero or negative overlap (an idling at 300 rpm is possible and saves a lot of fuel and emissions in town traffic).
The original plenum can change (at no cost) to a free flow ITB, there is no need for throttle valves (because the intake valves make the throttling), etc, etc.

As the valve springs are not necessary any more (the DVVA holds the valves during opening AND during closing), the space they occupy can be used to improve the intake and exhaust systems.

The cost of the valve springs plus the cost of a decent ITB system is more than the cost of the DVVA.

The complication and weight of the existing VTEC system of S2000 (three pieces heavy rocker arms, pin lockers, hydraulic control, piping and relevant valves, double injection and ignition tables etc, etc) is not less than the DVVA system.

And the rev limit of the DVVA is way higher.
Showing the www.pattakon.com/vvar/OnBoard/A1.MOV video (4MB, QT format) to a group of Honda VTEC funs, the question was: ‘where is the VTEC point?’. A true VVA has infinite VTEC points, and there are no noticeable steps in its operation.

In the http://forums.autosp...&threadid=95192 thread it is explained, between others, the improvement a VVA brings at partial loads and low revs.
Take a look at the following plot.
It is taken during road test of the full version of the B16A2 pattakon VVA prototype. .
The blue curve is the air handled per cycle, while the red curve is the air handled per time unit.
The curves show also the potential of the engine above the present rev limit (9000 rpm).
Even in case the peak power is the only that matters, a good VVA has no rival.


The low rpm mode and the high rpm mode of S2000 engine show one more thing: the significance of the exhaust valve lift profile on the engine behavior .
Most auto makers sacrifice the exhaust valves control for the sake of production cost. Honda does not.
It is quite different to have a VVT system to phase the exhaust cam, than to have a VVA to control exhaust valve lift profile. Just think the case the low rpm exhaust cam lobes of S2000 are omitted and the wild exhaust cam lobes are properly phased in order to close the exhaust valves not at 48 deg ATDC, but at 6 deg ATDC – as is the case with the soft cam lobes – and think that this way the exhaust valves open at 72+42=114 deg BBDC, i.e. at 66 deg ATDC!!.

Many auto makers, like BMW, Nissan, Honda and Toyota, move to the mechanical VVAs. There is no other way to pass the next emission regulations and they know it. BMW has it VVA (valvetronic) in mass production for years. And the top model of Nissan (G37 infiniti) uses Nissan’s VVEL VVA into its cylinder head, providing (compared to the previous infiniti) more power, better efficiency, more torque, better response etc etc.

The question is not if a good VVA is better than the state of the art conventional valve train. It is far better.
The question is if the VVA is applicable in reasonable cost and without side effects.

I hope things are more clear now and that the discussion will be focused on the VVA systems and the DVVA.

Thanks
Manolis Pattakos

[Moon Tricky]

Ok how about a valve that's basically a barrel with a hole drilled through the other way, i.e. perpendicular to the axis. The axis is perpendicular to the flow and the barrel rotates on this axis. So when the valve is fully open, it's as if it weren't even there. It shouldn't be too heavy and certainly isn't complex.

I'll do a picture...

Originally posted by manolis
A good VVA is, by far, better than the state of the art conventional valve train.

I agree, at least, if you can warrant the complexity. Combine with variable compression ratio for the best efficiency at any load.

[phantom II]

Give me a break. Here is a case for the desperate need of a LS 2 transplant with only an 80lb weight gain.. I have a 505 HP Corvette. On a bad day I get 18mpg. This would be a good day for a Honda S2000 where as on a good day, I get 30mpg at 80mph. Your bad days are 14mpg. This is disgusting. For 08, Honda will lower the RPM once more and get more toque in a desperate attempt to make it drivable. Two really awful sports cars come from Honda: NSX and a S2000. Watch how great they perform against Elise and Boxter, Suburu, Corvette, etc. at Solo 2 events. Engines and suspensions are the culprits. The clutch and the engine will never make it to 100 000 miles. A bicycle will beat you if you don't rev the **** out of it, FFS.

Originally posted by manolis
A good VVA is, by far, better than the state of the art conventional valve train.

Honda S2000 has, for years, the top power concentration naturally aspirating production engine. It provides 120 PS/lit at 8600 rpm (i.e. at 24m/sec mean piston speed). No other production engine approaches this number. And it is equipped with the simplest VVA, the VTEC.


The question is not if a good VVA is better than the state of the art conventional valve train. It is far better.
The question is if the VVA is applicable in reasonable cost and without side effects.

I hope things are more clear now and that the discussion will be focused on the VVA systems and the DVVA.

Thanks
Manolis Pattakos

[Ninja2b]

Originally posted by McGuire
They key being "for a given inlet size." If you can't get the required flow capacity at WOT with your butterfly valve, why not just bolt on a slightly larger butterfly valve? It will still be lighter, simpler, cheaper, less bulky overall and more reliable than a barrel valve of the same effective flow capacity.

I'll not deny that there is a LOT of solving problems that don't exist in the FS competitions, something I have critiscised our team for myself. But the barrel throttle was due to our inlet diameter effectively being fixed by the very expensive carbon fibre inlet we had made and were wanting to re-use. Believe it or not, it was simpler, faster and cheaper to make a barrel throttle ourselves than to buy a bigger buttefly and adapt out inlet system to fit it. It only took us about an hour or two to design (it's a simple component!) and it was CNCed pretty sharpish.

Luleå uni did one with a 'hollow barrel' as Moon Tricky suggested. Not sure if they are still using it.

I wonder if the presence of the restrictor pretty close downstream of the throttle has a significant effect on how the airflow over the butterfly develops? Maybe I'll do a quick CFD next week to have a squiz at it.... anyways this is all horribly OT!

[NRoshier]

The barrel throttle can have driveability issues at part throttle, so for your advantage to be 'real' with your given area being fixed, I would assume you do not use a great deal of part throttle and are mostly WOT. I investigated a 'squeeze' throttle arrangement which was using a flexible section of manifold that was then squeezed closed to restrict area. It performed wonderfully on a flow bench but I never tried it on a car. I'm sure it would work and would take considerable skill to make reliable. I am not sure there was any specific advantage, but it was an interesting intellectual exercise.
Phantom, my understanding is that Solo2 events are relatively low speed autocross events? So I am not sure what I am comparing. I would be amazed if a tripling of the engine capacity (LS2 is 6lt?) would not result in much more torque. But I will disagree with you about the NSX, I think that for such an old design it has been a great car for honda.

[Moon Tricky]

Seeing as it's going a bit off-topic, I think I'll take the opportunity to start a new thread on the subject. "Alternatives to the butterfly valve"

[manolis]

Phantom II wrote:
Give me a break. Here is a case for the desperate need of a LS 2 transplant with only an 80lb weight gain.. I have a 505 HP Corvette. On a bad day I get 18mpg. This would be a good day for a Honda S2000 where as on a good day, I get 30mpg at 80mph. Your bad days are 14mpg. This is disgusting. For 08, Honda will lower the RPM once more and get more toque in a desperate attempt to make it drivable. Two really awful sports cars come from Honda: NSX and a S2000. Watch how great they perform against Elise and Boxter, Suburu, Corvette, etc. at Solo 2 events. Engines and suspensions are the culprits. The clutch and the engine will never make it to 100 000 miles. A bicycle will beat you if you don't rev the **** out of it, FFS.


The same thing we say.
To achieve the top power concentration, Honda sacrificed the linear operation of the engine of S2000.
This happens because the VTEC system has only two modes of operation (high rpm mode and low rpm mode) and these two modes have to cover a really long rev range (from 1000 to 9000 rpm).
Necessarily there is a big gap between these two modes, and the driver has to decide either to go with the soft mode (good mileage) either to go like driving a racing engine. It is a “switchable” engine, having two discrete faces. Smooth or racing. No intermediate way.
The problem, according auto magazine journalists, was even worse in case of the Toyota Celica Vvtli (another two mode VVA system “very similar” to VTEC). Many customers of Toyota complained for dangerous lack of torque at specific revs and, some hundreds of revs higher, for too much torque, like the old two cycle engines.

Imagine a VTEC or Vvtli with a thousand or a million modes of operation, all available instantly to the driver.
This is what a good VVA can offer.
This is the way to increase even more the power concentration of S2000 engine, having at the same time, a true drivable engine along the whole rev range.
The plot below is the air handled per cycle (blue curve) and the air handled per time unit (red curve) of the B16A full version pattakon prototype. The car moves slowly, with third gear, at a steep uphill road. There is plenty of torque from 750 rpm.
.
It is the same engine that revs at 9000 rpm (look at the plot in my previous post) and seems to have the potential (at least as regards its breathing system) for even higher revs.

This is the key point of a good VVA: to optimise the breathing of the engine at all revs and every load.

Thanks
Manolis Pattakos

[phantom II]

This is Alphas solution. The camshaft spins at constant RPM, yet, by moving it back and forth hydraulically, the follower traces a very different profile on the lobe, changing lift, duration, LCA and timing. Both cams are controlled independantly.
My 7 ltr Corvette Z06 has a single NVR(non variable timing) cam in block. It produces well over 500hp@7000 rpm and gets 80% of its torque at 1500rpm which is nearly 80mph in 6th gear and 30mpg. 7000rpm in 5th gets me 200mph. How about them apples?

Originally posted by manolis



Imagine a VTEC or Vvtli with a thousand or a million modes of operation, all available instantly to the driver.
This is what a good VVA can offer.
This is the way to increase even more the power concentration of S2000 engine, having at the same time, a true drivable engine along the whole rev range.

This is the key point of a good VVA: to optimise the breathing of the engine at all revs and every load.

Thanks
Manolis Pattakos

[Moon Tricky]

Originally posted by phantom II
This is Alphas solution. The camshaft spins at constant RPM, yet, by moving it back and forth hydraulically, the follower traces a very different profile on the lobe, changing lift, duration, LCA and timing.

It is a very simple and clever solution, but does it not require the cam followers to be a funny shape? Either tapered to a point at the top, or sloping towards one side?

[manolis]

Phantom II wrote:
“This is Alphas solution. The camshaft spins at constant RPM, yet, by moving it back and forth hydraulically, the follower traces a very different profile on the lobe, changing lift, duration, LCA and timing. Both cams are controlled independantly.”


At first look the conical lobes seem as a good solution for a VVA.
But there are problems in this architecture.
Yamaha and Suzuki tried to solve some of them, according their patent applications in US-PTO.

Problems?

The valve lift cannot be zero, so the available valve lift range starts from a medium lift and changes continuously to the maximum lift, making the throttle valve absolutely necessary (so the advantages in torque and fuel consumption – from pumping loss at partial loads - the state of the art VVAs offer are gone).

The contact of the valve tappet with the conical lobe is ‘point contact’ instead of line contact. There have been proposed some intermediate mechanisms between the conical cam lobe and the valve, but none has been proved too good, yet. The kind of contact is crusial for the rev limit.

The thermal expansion does not allow accurate control on all cylinders. A typical camshaft is about 400mm long, made from steel, while the cylinder head is from aluminium. An increase of the temperature for 100 degrees centigrade causes an about 0.5 mm offset of the cam lobes of the last cylinder in regard to the cam lobes of the first cylinder (in this kind of VVA the control of the valve lift profile is based exclusively on the linear displacement of the cam lobes compared to the valve). This is crusial in case of very short lifts and in case of cold operation.

For each valve lift there is one and only one valve duration and vice versa. And the architecture does not allow the free combination of them.
Etc, etc.


Compare the case to the Pattakon FVVA or of the Pattakon DVVA.
For each valve lift there is an infinity of available valve durations and vice versa. The thermal expansion has not significant effect.
The valve lifts start from zero (i.e. valve deactivation and ability for operation without throttle valve) and vary continuously to the maximum.
The engine is rid of throttle valve, the partial load consumption is lower and the intake manifold changes to a free flow ITB without cost.
The torque can be better than true flat.
Especially in the case of the DVVA, the engine is also rid of valve springs and restoring springs (for the VVA mechanism), and instead of a camshaft, a crankshaft (shaft with eccentric pins) is used. The DVVA engine rid of valve springs is capable for way higher revs and so for way higher power concentration.




Phantom II wrote:
“My 7 ltr Corvette Z06 has a single NVR(non variable timing) cam in block. It produces well over 500hp@7000 rpm and gets 80% of its torque at 1500rpm which is nearly 80mph in 6th gear and 30mpg. 7000rpm in 5th gets me 200mph. How about them apples?”

A 7 lit engine with 500hp@7000 rpm means an engine with 500/(7*1.4*7)=7.3Kp*m/lit torque at 7000 rpm.
An engine providing only 7.3 Kp*m/lit torque, is nothing special.
With a good VVA on it, you could easily achieve some 9 to10 Kp*m/lit torque at 7000 rpm, which means some 650PS at 7000 rpm, and this combined with better mileage and better torque and response at medium low revs.
It would be very useful to know how much fuel per hour at idling your engine consumes, as well as its bore and stroke.

Think that what a VVA offers is additional valve lift profiles to operate the engine. A good VVA cannot but improve the operation of an engine.

Thanks
Manolis Pattakos