fly by wire throttle

[manousos]

During 'outgoing air control' which crank angles are the fuel and spark?

It depends on the engine.
Direct injection? Port injection?

The PatAir can operate with all the available modes / strategies ("Ingoing Air Control") of the MultiAir of FIAT and with an infinity of additional "Outgoing Air Control" modes / strategies.
The image bellow shows the difference in the hardware of the two systems.



The selection of the optimum mode is based on the operational conditions.

If necessary, different cylinders can operate on different "mode". All it takes is the proper "timing" of the electromagnetic solenoid valves by the ECU.

Flexibility: A PatAir four inline engine is easy to operate with one cylinder deactivated, with another cylinder running at full load, with another cylinder running at medium load according the "Ingoing Air Control" cycle and with another cylinder running at medium load according the "Outgoing Air Control" cycle. It is also easy to swap, a few dozens of times per second, the above modes among the cylinders.


The difference during the cranking process (and the "stop - start") between an “Outgoing Air Control” system (PatAir) and an “Ingoing Air Control” system (MultiAir of Fiat) is also significant.

During the cranking process, the MultiAir (and the conventional engine) needs a strong torque (energy) to make the first “compression”. If the torque provided by the starter is not adequate to compress the full charge into one cylinder, the engine cannot rotate. The starter has to be able to provide to the engine the power for a “full load” compression. Despite the progress in the battery technology, everyone has experienced an "empty" battery at cranking.

The “Outgoing Air Control” / PatAir is an efficient “de-compressor”. The starter rotates the engine without seeing significant resistance (a dozen times less energy-torque than the original MultiAir engine), because the intake valves stay open during the compression cycle (they close a few degrees before the TDC) and allow the mixture to return back into the intake manifold, just like as a de-compressor does.

The cranking becomes so easy (light load with “Outgoing Air Control”) that you need neither the ring gearwheel on the flywheel periphery, nor a conventional starter with pinion. A modified electric generator (that operates as an electric motor, too) is all it needs.

This is more important for cars equipped with a “start-stop” system: every time the engine stops, the cylinders cannot help filling up with mixture, which must be compressed at the next start.

The easy cranking of the “Outgoing Air Control” PatAir is significant / crucial in other cases: Imagine for instance, a powerful outboard engine on a boat and a discharged battery beside it, or a big single cylinder motorcycle engine with empty battery. With the “Outgoing Air Control” / PatAir the manual cranking is easy.

Some would not call that a loss. And would say that the process of heating the air of that 1/8 partially-filled cylinder is a heat recovery process. In that the compression pressure is higher due to pre-combustion heat recovery, even though the cylinder is only partially filled. It is a novel method of heat recovery.

This "novel method of heat recovery" consumes "expensive" mechanical energy ("expensive" because this mechanical energy was generated at a small "brake thermal efficiency" due to the light load operation).
There are others more fuel efficient ways to do it.
If you need hotter air at 30 degrees before the combustion TDC, you can pass the air around the exhaust pipes and use the "Outgoing Air Control" mode.

Thanks
Manousos Pattakos

[TheRevva]

It is not the pressure of the hydraulic circuit that opens and closes the intake valves in the MultiAir of FIAT and in the PatAir and HyDesmo of pattakon. The oil pump supplies oil in a relatively low pressure to just fill the oil chamber during the period the valve remains closed. Then the trapped (into the oil chamber) oil behaves as a “solid” / "steel" pushrod untill the electromagnetic solenoid release valve is actuated by the ECU, the hydraulic "push rod" collapses and the intake valve - under the action of the valve spring - closes independently of the cam.

Sorry for not responding sooner. I spent the weekend down at the drag strip (and it's taken me quite some time to recover from the celebrating after the fastest ever 1/4 mile pass in New Zealand)
Thanks again for explaining the MultiAir et al system. (Actually, I already had a pretty decent understanding by reading the relevant external URLs, but I'm sure others might be interested!(
The 'theoretical' engine I was talking about is / was truly devoid of ANY camshaft.
A separate (perhaps standalone?) hydraulic pump would operate the valves in a HyDesmo style setup. i.e. Actively open and close the valves with hydraulic pressure + associated solenoids under the control of the ECU.
I'd still envisage some WEAK valve springs purely as a 'safety system' to pull the valve closed. I emphasize that they're VERY weak!
The entire camshaft subsystem including timing chain 'disappears'. (This would probably necessitate a redesign of the primary oil pumping system since many of the engines I've seen run the oil pump from a gear on the camshaft).
Perhaps I should re-stress at this point that my thought process is centered around conversion of a traditional 'dinosaur' V8 with 2 valves per cylinder, single camshaft etc. My 'background' is with drag racing, and the class rules for group 1 mandate 2 valves per cylinder. Those same rules do NOT say that you have to run a camshaft at all, and nor do they say you have to run the engine as a 4-stroke!!! <Grins> Under group 1 drag racing rules, the mandatory ECU that would be necessary is forbidden! However, the requisite ECU _is_ allowed within group 2 racing. (And along with items such as the ECU, you're also allowed to run a more 'modern' technology engine with > 2 valves per cylinder...)
IMNSHO, with such an 'infinitely variable' electronic camshaft and being able to dynamically switch between 2-stroke and 4-stroke operation, I strongly suspect some amazing things can be achieved!
Perhaps most importantly of all is that it's remotely achievable!

[MotoFab]

It depends on the engine.
Direct injection? Port injection?

I was just wondering is all. In the example you gave, PatAir 'outgoing air control' had the compression stroke reversing 85% of the intake charge back out of the cylinder, backwards through an open intake valve, and back into the manifold, with the intake valve closing 30° BTDC on the compression stroke. I understand there is a purpose for doing that, but fueling and sparking that situation seemed like it would be interesting to hear about.

Some would not call that a loss. And would say that the process of heating the air of that 1/8 partially-filled cylinder is a heat recovery process. In that the compression pressure is higher due to pre-combustion heat recovery, even though the cylinder is only partially filled. It is a novel method of heat recovery.

This "novel method of heat recovery" consumes "expensive" mechanical energy ("expensive" because this mechanical energy was generated at a small "brake thermal efficiency" due to the light load operation).

Those same people would say that the mechanical energy from pulling against a closed valve is recovered on the compression stroke. And that energy conservation occurs in all cases, regardless whether or not the air in the cylinder absorbs waste heat from the surrounding cylinder walls during the process.

There are other more fuel efficient ways to do it.
If you need hotter air at 30 degrees before the combustion TDC, you can pass the air around the exhaust pipes and use the "Outgoing Air Control" mode.

That would seem to defeat the purpose. I mean, heating the air before it enters the cylinder decreases the intake density 'before the fact' say, and defeats the positive effect of recovering heat into the 1/8 partial intake charge 'after' the valves are closed, and before compression. As it does on the Fiat MultiAir system.


It's clear your PatAir system has some advantages, and I wish you good success in licensing the technology to the major manufacturers!

[manousos]

The entire camshaft subsystem including timing chain 'disappears'. (This would probably necessitate a redesign of the primary oil pumping system since many of the engines I've seen run the oil pump from a gear on the camshaft).
Perhaps I should re-stress at this point that my thought process is centered around conversion of a traditional 'dinosaur' V8 with 2 valves per cylinder, single camshaft etc. My 'background' is with drag racing, and the class rules for group 1 mandate 2 valves per cylinder. Those same rules do NOT say that you have to run a camshaft at all, and nor do they say you have to run the engine as a 4-stroke!!! <Grins> Under group 1 drag racing rules, the mandatory ECU that would be necessary is forbidden! However, the requisite ECU _is_ allowed within group 2 racing. (And along with items such as the ECU, you're also allowed to run a more 'modern' technology engine with > 2 valves per cylinder...)
IMNSHO, with such an 'infinitely variable' electronic camshaft and being able to dynamically switch between 2-stroke and 4-stroke operation, I strongly suspect some amazing things can be achieved!
Perhaps most importantly of all is that it's remotely achievable!

A few thoughts:

During the 2-stroke operation you need faster opening and closing of the valves (the valve needs to open a little after the middle stroke and needs to close sooner than 180 crank degrees later). This increases the inertia loads and makes difficult the landing of the valves on their valve seats (bouncing).
If this is difficult for normal revs (say 6000 rpm), what about racing revs?

And you need to actuate the exhaust valves, too, and this is harder (with several bars inside the cylinder the moment the exhaust valves start opening, the hydraulic system needs even higher pressure).

On the other hand, a "loop" scavenged 2-stroke has necessarily small "valve-time" area as compared to a "through (uniflow) scavenged" 2-stroke,

like the PatPortLess at http://www.pattakon.com/pattakonPatPortLess.htm having 8 valves per cylinder:



and like the PatMar engine at http://www.pattakon.com/pattakonPatMar.htm having 2 valves per cylinder:



Both, the PatMar and the PatPortLess have true 4-stroke: lubrication and specific lube consumption and scuffing resistance.



MotoFab wrote:

That would seem to defeat the purpose. I mean, heating the air before it enters the cylinder decreases the intake density 'before the fact' say, and defeats the positive effect of recovering heat into the 1/8 partial intake charge 'after' the valves are closed, and before compression. As it does on the Fiat MultiAir system.

The same is happening during the light load MultiAir operation. You pay mechanical energy to expand the air inside the cylinder.
During the delay at the BDC (the long piston dwell is at the BDC) the air temperature rises and the pressure inside the cylinder increases.
During the following compression, the piston "compresses" hotter air of higher temperature, i.e. less mechanical energy returns to the crankshaft-flywheel.

If, for some reason, you need hotter air in the cylinder, you can pre-heat it (without spending expensive mechanical energy) and use the "Outgoing Air Control".

Thanks
Manousos Pattakos

[Fred]

A reply to TheRevva's post (Mon Jan 14, 2013 8:18 pm) and to MotoFab's post (Mon Jan 14, 2013 11:50 pm) was posted a couple of days ago.

Sorry for the delay. I deleted the above message as it added no value.

[jtw11]

A pressure/volume diagram does not show the amount of power required to move air through a motor. A PV graph shows pressure, not airflow. The intuition is that a PV graph is a graph of 'voltage', and since a PV graph doesn't show the airflow 'current', the 'power' consumed in moving the air cannot be derived.

Well here I'm afraid you're mistaken, since pressures are directly related to airflow - that's exactly what the lower loop of a pV diagram shows - as for the electrical analogy, I never think it works.

I suggest a read of a number of papers;

http://v-scheiner.brunel.ac.uk/bitstrea ... lltext.pdf

The improvement in BSFC primarily came
from the absence of pumping losses of throttling at part
load.

Therefore, changing
residual fraction by adjusting the valve phasing can be
an effective means of controlling the engine load
during the CAI combustion operation, resulting in the
throttleless engine operation and hence reduced
pumping losses.

http://www.brighton.ac.uk/shrl/publicat ... 1-0750.pdf

Through the use of very dilute mixtures an HCCI engine could operate unthrottled at part load - reducing pumping losses - as the diesel engine does.

Perhaps so far in our discussion there has been a misunderstanding in the definition of a 'closed' throttle. Certain I've been referring to a closed throttle to be idle throttle demand, not fully closed. I suppose looking at the concept of a CLOSED, i.e. 0deg throttle, yes there are no pumping losses since we're not pumping past a throttle, but given we're talking about part load pumping losses at low throttle angles, that point is invalid.

To TheRevva -

Those same rules do NOT say that you have to run a camshaft at all, and nor do they say you have to run the engine as a 4-stroke!!!

Every rule book I've seen this side of the pond specifies 4 stroke, shame!

[TheRevva]

<Snipped somewhat>

If this is difficult for normal revs (say 6000 rpm), what about racing revs?

Maybe I'm being a little naive here, but it's my 'understanding' that the primary issue at 'race revs' is valve springs (and thus the ensuing valve float).
IMNSHO, use of a Hy-Desmo type system (albeit completely camless), in conjunction with a fairly powerful ECU to control the hydraulic 'switching', we would largely eliminate the traditional issue at higher RPM.
I am NOT in any way saying it would be trivial to accomplish it, but it's less 'extreme' from a development perspective than some of other options.
Specifically, I would suggest that such a system could be 'retrofitted' on existing head designs with comparatively little changes to the remainder of the physical engine.

To TheRevva -

Those same rules do NOT say that you have to run a camshaft at all, and nor do they say you have to run the engine as a 4-stroke!!!

Every rule book I've seen this side of the pond specifies 4 stroke, shame!

I guess that just makes it more 'interesting' for the NZ dragracing scene?
Here's a direct excerpt from a rulebook on allowable engines in our 'Competition Dragster' category:

Engine(s): Must be internal combustion automotive type. A maximum of two engines may be used Any modification may be made to the engine(s) and they may be in any location...

At the moment, I don't have my ANDRA (Australian) dragracing rulebook in front of me, but I seem to recall they are equally 'open' to such inventive excursions from 'the norm'.
It certainly would make things 'interesting' to say the least... LOL

The sole 'obstacle' is the ECU itself. Whilst EFI and electronic ignition control is explicitly classified as 'Category 1' in the rulebook (and thus is 'allowed'), any category 3 electronics is explicitly forbidden. Category 3 is defined as any 'interactive controlling device'. The intent is to ban any devices such as "traction control or fly-by-wire".
The easiest way to overcome that is to simply start protesting all the existing vehicles that possess factory VVT etc. I suspect the rule will quietly 'disappear' when it finally becomes obvious that a decent number of their 'customers' are forbidden from racing! A surprisingly LARGE percentage of vehicles produced this millennium explicitly possess category 3 electronics in defiance of the current rules and are therefore technically illegal.

I'd envisage a dual-stage approach to development.
Stage 1 would be replication of the camshaft function in a PURE 4-stroke mode. (Albeit with infinitely variable valve timing under ECU control)
Stage 2 would involve the dynamic switching between 2-stroke and 4-stroke

[manousos]

Maybe I'm being a little naive here, but it's my 'understanding' that the primary issue at 'race revs' is valve springs (and thus the ensuing valve float).

In order to increase the intake valve lift from the original 10.5mm to 12mm in the "pattakon Honda VTEC B16A2" prototype engine (details at http://www.pattakon.com/pattakonVtec.htm and at http://www.pattakon.com/pattakonRoller.htm ) and in order to shift the red line from 8000 rpm to 9000 rpm (wherein the underneath engine - crankshaft, connecting rods, pistons, block - is at its limit), we had to replace the original valve springs with egg-shape TODA ones.

The modified B16A2 engine operates with valve lift from 0.15mm (idling) to 12mm (peak power / 9000 rpm):



The conventional throttle valve is not necessary any longer: the intake valves make the "throttling" (ideal for efficient and cheap ITB).



No more than 2 - 3mm of intake valve lift is necessary for driving in the town traffic (the engine is torquie and driver friendly as the softer family car engines).
On the open road, the same engine turns to a racing engine.
But the engine is still based on valve springs to restore the valves.


The next reasonable step is to completely remove the restoring valve springs.
The mechanical Desmodromic VVA ( or DVVA ) at http://www.pattakon.com/pattakonDesmo.htm needs not restoring valve springs at all.
During operation you can select (by rotating the control shafts) the desirable valve duration and the desirable valve lift like:



The DVVA can approach the available valve lift profiles of any existing valve train (conventional, VVA or Desmodromic). So, there is no reason for not being (the DVVA) more fuel efficient and for not providing top power.



Rid of valve springs, rid of unnecessary loads (like the restoring force from the valve springs at medium and low revs), rid of heavy quick moving parts, rid of sliding friction etc, the reliable rev limit of the engine is no longer set by the valve train but by the underneath mechanism (crankshaft, connecting rods, pistons and block).


The four-stroke DVVA is ready for political and racing use.

In case you want to operate the spring-less DVVA engine as a two stroke for racing use, all you need is to change the transmission ratio from the crankshaft to the "secondary crankshaft" (or crank-cam-shaft) from 2:1 to 1:1.
During operation you can vary the valve lift and the valve duration as you like.

For high revving applications (racing) the mechanical full control over the valve motion (the mechanism holds positively the valve during the valve opening AND during the valve closing) seems more promising than any hydraulic control. Doesn't it?

Thanks
Manousos Pattakos

[TheRevva]

I am pretty sure I understand all the basics of the Pattakon valving system... Seriously!
And yes, it _IS_ a seriously innovative design.
However, it still uses a camshaft (albeit of a seriously radical shape) and, at least in my mind, the inclusion of _ANY_ camshaft must, by necessity, impose certain restrictions upon any given engine.

What I would love to hear (from anyone) is a list of reasons why it's impossible to implement a true 'camshaft-in-silicon' solution.
The Hy-Desmo valving system has all-but-proven to me that it's feasible to both open and close the valves using hydraulics along with a suitable electrically operated hydraulic solenoid.
In my simplistic view of the existing Hy-Desmo descriptions, the primary purpose of the camshaft is to supply the mechanical force to drive a local hydraulic pump.
I cannot see why someone couldn't supply a totally external hydraulic source to completely eliminate the camshaft.
I would envisage this 'external' hydraulic source to be a combination of a crank driven pump as well as a small electric pump.
The electric pump would provide primary hydraulic pressure during initial cranking / starting with the crank driven pump guaranteeing sufficient pressure once running. (The two pumps would act as 'backup' for each other too!)

Since the entire system is under direct control of the ECU, it would be logical to include some basic 'self checks'.
Once the ignition is switched on, the electric pump brings up hydraulic pressure and then proceeds to verift that it can both open, and close each valve. (Simple limit switches on each valve!)
Naturally, there MIGHT be some valves that can not open at all due to interference with a corresponding piston being positioned at TDC.
If the pre-crank testing is unable to CLOSE every valve in the engine, then the system remains inoperative (illuminating a traditional CEL for example).
If that test 'passes' (by getting positive feedback that every valve is closed), the ECU would then 'arm' the starter motor solenoid to enable traditional engine cranking.
Once the ECU (via the crank angle sensor) was able to determine the current crank angle (eg: with a 60-2 trigger wheel?), it could then enable valve, injector and spark operation and the electric hydraulic pump could be 'retired'

There would be some obvious 'safety interlocks' designed in to the system.
For example, if the hydraulic pressure falls too low during operation, the electric hydraulic pump would be switched back on and a 'limp-home' mode engaged.
Each of the electrically operated hydraulic solenoids would probably have a decent return spring to engage positive valve closing hydraulic pressure. (i.e. Failsafe to a 'valve closed' position upon electrical failure)
If the limit switch on a valve continually indicated 'valve closed' when it should actually be open, that cylinder would be 'deactivated' and limp home mode engaged.

The ECU, being in TOTAL control of valve operation, could thus dynamically switch between 2-stroke and 4-stroke at will.
Furthermore, it could even selectively disable injection, ignition, and certain aspects of valve operation on any given cylinder(s) to enable total deactivation (and thereby some potentially impressive fuel savings?)

The above post from Manousos has also highlighted that the total valve control also obviates the need for a throttle butterfly too! (Yet another potential cost saving to offset the additional costs of valving and hydraulic pumping).
The downside being that a Speed Density based system would be impractical as there would be negligible 'manifold vacuum' to speak of. (Any such system has the manifold at effectively WOT 100% of the time!)
However, I suspect a decent MAF based system should be more than adequate in any such system? An Alpha-N type system might also prove to be workable?

I _must_ be missing something truly fundamental here.
I don't see the additional hardware costs as being prohibitively more than current engine technology (and as mentioned above, there are SOME savings that help to offset them!)
It all seems so easy that I cannot believe _someone_ hasn't already done it and it's about to reach mass production?

[Fred]

Speed density doesn't apply when you're directly controlling in-cylinder pressure through specific valve actuation profiles.