fly by wire throttle

Free Open Source Firmware project discussion forum. Post your Free Open Source firmware projects here!
Post Reply
TheRevva
DIP8 - Involved
Posts: 26
Joined: Thu Dec 13, 2012 1:10 am

Re: fly by wire throttle

Post by TheRevva »

Well... The above post has ruined my preconceived idea of how the Fiat Multiair system actually works...
I was under a (perhaps completely incorrect?) assumption that that Multiair system didn't (and I quote) 'SNAP SHUT' the valves at all...
I had surmised / assumed that the system simply altered the effective maximum LIFT of each valve along with the effective opening and closing angles (but still kept a fairly FLUID valve movement.
With me being one of the dinosaur '8 cylinders and pushrods' generation, I was envisaging something equivalent to a dynamically variable ratio rocker arm along with a dynamically variable ratio cam follower.
The variable ratio rocker would affect the maximum lift and the variable ratio cam follower would affect both lift AND opening / closing angles (by effectively increasing / decreasing the camshaft 'base circle')

If I have it all wrong (which is HIGHLY likely), and the vales do indeed 'snap shut', then those valves need to be some SERIOUSLY strong puppies and any associated springs will have to be INSANE to stop valve bouncing!
I guess it's time I took a look at some real pix and associated descriptions.

Footnote:
Actually, now that I think about it a little more, it's not entirely unreasonable to 'factor in' some amount of valve bounce into the equation. Any such 'bounce' would most likely still occur DURING the traditional intake phase and a half decent micro CAN take this into account for its calculations.... Hmmmmmm.... (They'd still need to be STUPIDLY strong valve heads, seats etc)

I _DO_ like this forum.
While it's primary purpose remains the mammoth DIYEFI project, it's certainly forced me to THINK a lot more about all the indirectly associated areas like this!
TheRevva
DIP8 - Involved
Posts: 26
Joined: Thu Dec 13, 2012 1:10 am

Re: fly by wire throttle

Post by TheRevva »

For others who, like me, might be somewhat intrigued, here's a semi decent description of MultiAir / Pattakon that has just HELPED my understanding a lot:
http://www.pattakon.com/pattakonHydro.htm
(And yes, the valves do INDEED seem to 'snap shut' a lot more than with a traditional cam profile.)

Edit (I hit SUBMIT a little prematurely... LOL):
What I don't yet grasp is why they've not been able to completely eliminate the camshaft altogether...
The description of the system in the URL above looks to be using a solenoid operated 'relief valve' on a 'hydaulic push rod' between the cam lobe and the valve.
What's stopping them using another such solenoid on a pressurised fluid supply to OPEN the valve too?

I'm envisaging something not entirely different to a high pressure diesel injector employed to push open the valve!
The next 'logical' step from that would be to eliminate the valve springs altogether (along with their associated losses).
Namely, using hydraulics to push the valve OPEN via one solenoid, and hydraulics to push the valve CLOSED with another.
IMNSHO, the effective 'area under the curve' becomes almost infinitely variable
Furthermore, it'd also suit itself VERY well to an ECU. There would be NO need for a cam position sensor since the ECU effectively _IS_ the cam. (Only the crank angle sensor needs to remain)
Last edited by TheRevva on Sun Jan 06, 2013 9:26 pm, edited 1 time in total.
jtw11
TO220 - Visibile
Posts: 12
Joined: Mon Dec 31, 2012 2:02 pm

Re: fly by wire throttle

Post by jtw11 »

Hi 'TheRevva', the Multiair system is truly brilliant - and in my opinion was completely deserving of the International Engine of the Year Award it received. The system is quite different to what you think - the camshaft does not act directly upon the valves, but rather the camshaft is used to 'drive' oil through channels in head through to a chamber that actuates the valves.

However, there is a solenoid per cylinder that can expand the volume this oil moves through - thus for example, when an early valve closing event is desired - the solenoid opens to expand the available volume, thus pressure is removed from the valve and it is closed by its spring. So yes, you're correct - 'snap shut' was merely my way of describing the valves closing quickly by removing the pressure holding the valve open.

There's a brilliant video on YouTube showing it working. http://www.youtube.com/watch?v=Td9Gz_h7Qpg

So there's no variable ratio rocker, just a variable volume chamber through which the oil forced through by cam motion that acts upon the valve to open it. So the system is still fail-safe in that failure of the electrical solenoids won't lead to a permanently open or closed valve. In concept it's very simple, as with all good things I suppose. In practice, somewhat more intricate :D
jtw11
TO220 - Visibile
Posts: 12
Joined: Mon Dec 31, 2012 2:02 pm

Re: fly by wire throttle

Post by jtw11 »

TheRevva wrote: Edit (I hit SUBMIT a little prematurely... LOL):
What I don't yet grasp is why they've not been able to completely eliminate the camshaft altogether...
The description of the system in the URL above looks to be using a solenoid operated 'relief valve' on a 'hydaulic push rod' between the cam lobe and the valve.
What's stopping them using another such solenoid on a pressurised fluid supply to OPEN the valve too?

I'm envisaging something not entirely different to a high pressure diesel injector employed to push open the valve!
The next 'logical' step from that would be to eliminate the valve springs altogether (along with their associated losses).
Namely, using hydraulics to push the valve OPEN via one solenoid, and hydraulics to push the valve CLOSED with another.
IMNSHO, the effective 'area under the curve' becomes almost infinitely variable
Furthermore, it'd also suit itself VERY well to an ECU. There would be NO need for a cam position sensor since the ECU effectively _IS_ the cam. (Only the crank angle sensor needs to remain)
The primary reason why the camshaft remains, is think of the oil pressures we are talking about here to open the valves. As far as I'm aware, and clearly as far as the manufacturers are aware, there are no solenoids powerful or reliable enough to generate the kind of pressures the camshaft will here. I suppose looking at it from another point of view, how often do you hear of camshaft failures compared to electrical component failures? Fleet reliability would suffer somewhat I'd imagine without a camshaft.

It has been done, there's a web blog somewhere by a bloke who replaced mechanical actuation of the valves with normal solenoids on a GX140 engine, but then again - you can open those valves with the springs attached with your pinky it's such a tiny engine. I'm sure somebody did it at the uni I'm at too... I'll have a look around this week.

The losses associated with driving a camshaft and springs are relatively small - if you think about compressing the springs only, the losses are huge. However, don't forget the work is returned to the camshaft when the valve closes and the spring pushes against the camshaft. The only actual losses are the frictional losses in the bearings, heat generated in the springs and the tiny amount of energy absorbed in the elastic deformation of valvetrain components.

Positive closing systems aren't particularly necessary to be honest - in the old days when metallurgy hadn't come very far, and valve spring failures were common due to fatigue, pneumatic springs were invented to prevent those failures. Then the desmodromic systems came along that allowed very aggressive cam profiles, as old springs couldn't be made stiff enough without even more premature fatigue failures. These days, metallurgy has come a very long way indeed - and it wouldn't suprise me at all if we see the reintroduction of 'normal' metal valve springs in Formula 1 come 2014 with the reduced rev limit.
TheRevva
DIP8 - Involved
Posts: 26
Joined: Thu Dec 13, 2012 1:10 am

Re: fly by wire throttle

Post by TheRevva »

I hear ya loud and clear...
However, I was thinking of the solenoids in question being more similar to a fuel injector rather than acting as the compressor itself.
i.e. The solenoid simply acts as a 'switch' to open an orifice to a high pressure supply pump just as any traditional fuel injector does.
The effective pressures involved would no doubt be significantly higher that the 3-6 bar pressures we see in traditional fuel rails, but they're still easily within reach of what CAN be achieved.
Furthermore, the solenoids can be 'driven' in either direction thereby giving a three-way setup per solenoid (and I'd envisage two solenoid per valve)
TOP solenoid:
1: Energised in one direction, it would permit high pressure fluid to push the valve OPEN
2: Energised in the OTHER direction it would permit the relief of the 'stored' pressure holding the valve open
3: When de-energised, the valve would 'maintain' the valve at the current position (subject to hydraulic seal leakages)
BOTTOM solenoid:
1: Energised in one direction, it would permit high pressure fluid to push the valve CLOSED (emulating the valve return spring)
2: Energised in the OTHER direction it would permit the relief of the 'stored' pressure holding the valve closed
3: When de-energised, the valve would attempt to 'maintain' the valve at the current position (subject to hydraulic seal leakages)

It's still a year or two away, but I cannot see it as being beyond current technological ability!
Any such system could / should be able to offer TOTAL control of every aspect of poppet valve operation.
Naturally, the 'risks' associated with a systemic failure on any such system would be potentially catastrophic... Especially if the design called for an interference engine.

I'm betting we'll be seeing a LOT more advances in poppet valve operation technology over the next few years and my 'crystal ball' includes visions of future engines without ANY camshaft at all.
Furthermore, I also envisage poppet valve based engines that dynamically switch between partial 2-stroke and normal 4-stroke operation with such systems dependent upon instantaneous requirements.
jtw11
TO220 - Visibile
Posts: 12
Joined: Mon Dec 31, 2012 2:02 pm

Re: fly by wire throttle

Post by jtw11 »

A thought-provoking post indeed!

Funny you should mention the two/four stroke switching, I'm in my final few months of my Bachelors at the moment - a couple of the researchers here did that a few years back - http://www.brighton.ac.uk/shrl/projects/2-4sight.php

The engine is on show in the main entrance, I'll snap a shot tomorrow! Two of the three guys listed on that page, Nick Miche & Steven Begg are my supervisors for my dissertation as we speak.

Here's another paper on it http://www.ricardo.com/Global/IA/What-W ... 0study.pdf
Last edited by jtw11 on Sun Jan 06, 2013 11:09 pm, edited 1 time in total.
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: fly by wire throttle

Post by Fred »

jtw11 wrote:I'm in my final few months of my Bachelors at the moment
Way to lose all respect from MotoFab! You're clearly too young and inexperienced to so much as consider turning the key to start an ICE, let alone explain to him how they work.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
jtw11
TO220 - Visibile
Posts: 12
Joined: Mon Dec 31, 2012 2:02 pm

Re: fly by wire throttle

Post by jtw11 »

Fred wrote:
jtw11 wrote:I'm in my final few months of my Bachelors at the moment
Way to lose all respect from MotoFab! You're clearly too young and inexperienced to so much as consider turning the key to start an ICE, let alone explain to him how they work.
Haha, I shouldn't worry - I've got used to the "he's too young to know anything" train of thought by now, although it soon becomes clear who's cut out to manage good teams when age is put to one side... And like I say, I've been fortunate to work with some very good people indeed, once you've earned their respect, all works harmoniously.

Nothing better than watching your teams fire breathing effort cross the line in first place...
TheRevva
DIP8 - Involved
Posts: 26
Joined: Thu Dec 13, 2012 1:10 am

Re: fly by wire throttle

Post by TheRevva »

jtw11 wrote:A thought-provoking post indeed!

Funny you should mention the two/four stroke switching, I'm in my final few months of my Bachelors at the moment - a couple of the researchers here did that a few years back - http://www.brighton.ac.uk/shrl/projects/2-4sight.php

The engine is on show in the main entrance, I'll snap a shot tomorrow! Two of the three guys listed on that page, Nick Miche & Steven Begg are my supervisors for my dissertation as we speak.

Here's another paper on it http://www.ricardo.com/Global/IA/What-W ... 0study.pdf
To directly quote an old TV Advert here in NZ.... "BUGGER!!!"
I guess it was just 'far too obvious' that someone else hadn't already thought of it AND done it too... LOL

Does this mean I'm going to have to concentrate on the 'one-stroke' engine design instead??? (Using a 'double ended' piston).
Oh hang on, I think someone already beat me on THAT one too (with a Diesel no less!)
manousos
TO220 - Visibile
Posts: 7
Joined: Thu Jan 10, 2013 5:35 am

Re: fly by wire throttle

Post by manousos »

TheRevva wrote:For others who, like me, might be somewhat intrigued, here's a semi decent description of MultiAir / Pattakon that has just HELPED my understanding a lot:
http://www.pattakon.com/pattakonHydro.htm
(And yes, the valves do INDEED seem to 'snap shut' a lot more than with a traditional cam profile.)

Edit (I hit SUBMIT a little prematurely... LOL):
What I don't yet grasp is why they've not been able to completely eliminate the camshaft altogether...
The description of the system in the URL above looks to be using a solenoid operated 'relief valve' on a 'hydaulic push rod' between the cam lobe and the valve.
What's stopping them using another such solenoid on a pressurised fluid supply to OPEN the valve too?

I'm envisaging something not entirely different to a high pressure diesel injector employed to push open the valve!
The next 'logical' step from that would be to eliminate the valve springs altogether (along with their associated losses).
Namely, using hydraulics to push the valve OPEN via one solenoid, and hydraulics to push the valve CLOSED with another.
IMNSHO, the effective 'area under the curve' becomes almost infinitely variable
Furthermore, it'd also suit itself VERY well to an ECU. There would be NO need for a cam position sensor since the ECU effectively _IS_ the cam. (Only the crank angle sensor needs to remain)
Quote from http://www.pattakon.com/pattakonHydro.htm :

The MultiAir of Fiat works according the "Ingoing Air Control": The sooner the intake valve closes, the more "ingoing air" is kept from entering the cylinder during the suction. The later the intake valve closes, the heavier the load.

The MultiAir mechanism:
An "oil push rod" is interposed between the valve and the cam.
The cam pushes the "oil push rod" and the "oil push rod" pushes the valve.
At the right moment a solenoid valve opens, the "oil push rod" collapses and the valve closes under the restoring action of the valve spring.
By a "hydraulic braking mechanism" the landing of the valve on the valve seat becomes acceptably smooth.

The application of the MultiAir / TwinAir in mass production engines proves that the state-of-the-art solenoid valves and the hydraulic system do operate reliably and efficiently.


The plot below gives the difference between the various modes / strategies of the MultiAir of FIAT:

Image

Quote from http://www.pattakon.com/pattakonKeyAdv.htm#pattair :

The PatAir, based on very similar hardware (only the duration of the camshaft needs to change), works according:
either the "Ingoing Air Control" of Fiat,
or according the "Outgoing Air Control" of pattakon wherein the load is controllably increased by preventing more "Outgoing Air" from leaving the cylinder. The sooner the intake valve closes after BDC, during the compression stroke, the heavier the load.
The thermodynamic cycles of the "Ingoing Air Control" (Fiat) and of the "Outgoing Air Control" (pattakon) are different.
The "Outgoing Air Control" cycle avoids not only the underpressure, under part load, into the intake manifold (as the throttle-less VVAs, like the Fiat MultiAir and the BMW valvetronic, do) but it also avoids the underpressure into the cylinder by avoiding the expansion of the charge before the compression.
By minimizing the pumping-loss, by avoiding the consumption of mechanical-energy to just expand and warm the charge, by keeping alive the turbulence and swirl during combustion and by improving the mixture homogeny the "Outgoing Air Control" minimizes the mechanical-energy loss and optimizes the combustion.
The PatAir is an evolution of the MultiAir because it can operate not only according the infinite available modes of Fiat MutliAir cycle, but also according the infinite modes of pattakon "Outgoing Air Control" cycle.



The animation below was made to expalin the difference between the MultiAir and the PatAir:

Image

(at http://www.pattakon.com/hydro_files/PatAir.exe there is a controllable windows exe animation that further explains how things really work).


If the "Outgoing Air Control" seems complicated, please imagine the following:

a) In an empty cup someone adds water by a spoon and stops when 1/4 of the capacity of the cup is filled (Ingoing “Water” Control).

b) In en empty cup someone adds water by a spoon to completely fill the cup, then he pours out of the cup the 3/4 of the water by the same spoon (Outgoing “Water” Control).

Aren’t the above two “cycles” operative and clear?
They look similar. However, when the cylinder takes the place of the cup, and the air takes the place of the water, and the intake valve takes the place of the spoon, the resulting “Ingoing Air Control” cycle and “Outgoing Air Control” cycle differ, from the thermodynamic viewpoint (as explains the description and the Fig 3), as the day from the night.

The plot below is from the first PatAir prototype (adds an infinity of "unlimited" Miller cycles).

Image

At operation with a very little quantity of fresh air, say 10% of the capacity of the cylinder, the “Ingoing Air Control” mode runs as a refrigerator (or heat-pump) working for cooling purposes and indicating negative indicated power output, whilst the same engine running according the “Outgoing Air Control” would indicate its best indicating fuel efficiency. Against the common sense, at “extremely light loads” the “Ingoing Air Control” provides negative indicating power, whereas the “Outgoing Air Control” mode delivers its best indicating efficiency.


What makes the MultiAir of FIAT so good?

There are several VVA systems (electromagnetic, hydraulic, mechanical-hydraulic etc) that, theoretically, provide full control over the intake valve lift, allowing the engine to operate according any mode.

In practice (i.e. in the mass production) the only successful system is the MultiAir of FIAT. It is based on the simplest and cheapest hydro-mechanical mechanism that intergrades (built-in) the indispensable safety (the valve lift envelope makes the valve-piston-collision as impossible as it is in the best conventional engines).

The direct control of the ECU over the solenoid valves makes the MultiAir engines the first successful true-digitally-controlled engines in mass production. The ECU, based on the feedback, varies the time each solenoid valve is triggered to open (and thereby the moment the intake valve closes), which is an amazingly flexible and an exceptionally accurate control (impossible for the best mechanical systems).

The internal combustion engine turns into a true “digitally controlled device”.

The evolution from the inflexible mechanical valvetronic of BMW (regarded as the state-of-the-art VVA a couple of years ago) to the electro-mechanical-hydraulic MultiAir of FIAT is comparable to the evolution from the “old centrifugal spark advance control” to the modern electronic ignition systems. However they insist using it the wrong way.



TheRevva wrote: "What's stopping them using another such solenoid on a pressurised fluid supply to OPEN the valve too?"

It is more easy than this. You don't need a second solenoid electromagnetic valve. Take a look at the HyDesm (Hydraulic Desmodromic) at http://www.pattakon.com/pattakonHyDesmo.htm

Image

Thanks
Manousos Pattakos
Post Reply