Jump to content


Photo
- - - - -

Toyota's non-hybrid Atkinson engines


  • Please log in to reply
69 replies to this topic

#51 Speedman

Speedman
  • Member

  • 53 posts
  • Joined: November 05

Posted 29 April 2014 - 15:32

Hi Manolis

 

Worth mentioning is that the dust of beryllium is more than highly toxic (lug cancer).
 

 

Absolutely correct. This is no problem then the dust occurs only in the manufacture of. Ilmore F1 engine circa 1995 had berrylium.

Very few model engine builders use berylium. The rather engine titanium, high alloy steel or alluminium use rod.

Berrylium is too expensiv, is only for hard core mechanican. Titanium is ok.

 

I know OS Max engines very good. I have many engines of which.

 

OS Max is not a racing engine. Is a very good standard or sport engine.

 

All model dyno-test are not standardized.

Your OS Max dyno chart is a bad example. This is a car engine und not a airplane engine. Car engine are not designed for permanent full-throttle.

And i think os max used (very?) much nitromethane.

 

28'000 RPM are actually to high for the airscrew.  The greater airscrew speed the lower the efficiency of the airscrew.

 

 

Die einzige Annahme, ich tue, ist, dass die Messung auf dem Dyno ist kein Fake (und dass da sowohl die wesentlich kleiner und die wesentlich größer 2-Takter kann das erforderliche Drehmoment Dichte zu erreichen, ein 10cc 2-Takt ist auch in der Lage für eine ähnliche Drehmomentdichte) .
 

 

I know the physical laws, but is only theory. Bigger engine, more friction, more mass forces, more weight. Bigger cylinder bore than problem with nitromethane combustion.  A airplane engine has a air cooling than a race engine. And the air flow is in small nich same as in large, topic reynoldssche number.

A glow nitro engine hasn't a spark ignition but only a glow ignition.

 

It's not so easy as you think. When you can make a better engine, then please make a better engine, you'll be surprised.

Many engineers have tried, but few had success. Buy a 10 cc engine from OPS or a other ABC/AAC  engine on ebay and you can and you can experiment.

Power on dynamometer and power in the heaven are not the same. Many engine has a good power on the dyno, but a bad power in the sky.

The transfer-port and exhaust port-timing is relatively long and so is the resonance range is very small.

Airplane, engine and airscrew must fit together, or you has no chance.

 

Is the fuel mixture a little bit too lean and your engine is in few second scrap metal and is the fuel mixture a little bit too rich and your engine has too less power.

Also you can adjust the fuel mixture in the sky, but this is not so easy as you fink.

 

 

best regards

 

Speedman


Edited by Speedman, 29 April 2014 - 17:05.


Advertisement

#52 manolis

manolis
  • Member

  • 392 posts
  • Joined: May 03

Posted 30 April 2014 - 05:44

Hello Speedman.

 

You write:

"28'000 RPM are actually to high for the airscrew.  The greater airscrew speed the lower the efficiency of the airscrew."

 

You also wrote:

"Do you need a challenge?

Can you construction a simple very light 2-Stroke nitro engine with 10cc and more then 6 HP for RC speed plane?

Conventional special racing 2-Stroke RC-Speed engine are at the limit by circa 5 -5.5 HP

Probably needs more than 28'000 rpm or a many better cylinder filling."

 

I proposed a high revving OPRE tilting ( http://www.pattakon....akonTilting.htm ) revving at more than 35,000rpm having a reduction gearing from the two crankshafts to the prop.

Don't see the 35,000rpm as high.

With less than half piston stroke (oversquare with two opposed pistons) the mean piston speed is quite smaller (and so the reliability), as well as the maximum acceleration.

 

The OS RC engine with 15mm stroke has at 30,500rpm a mean piston speed of 15.25m/sec (not so high) and - as they write - can operate above 40,000rpm (more than 20m/sec).

Are you sure it cannot operate reliably (for "long", what long means for a model airplane) at 30Krpm with full throttle? Where is the problem?

If it can, make a three cylinder and you have 6.8bhp from 9cc.

 

A 10cc single square has 23.35mm bore and 23.35mm stroke. At 28,000rpm the mean piston speed is 21.9m/sec. High but not extreme (the VVA-roller Honda prototype at http://www.pattakon....takonRoller.htm operates reliably at 23m/sec (9,000rpm) mean piston speed).

 

A 10cc oversquare OPRE tilting (say 26mm bore, 10+10=20mm stroke, 10.6cc) has, at 40,000rpm, a mean piston speed of only 13.5m/sec.

 

The two crankshafts of the OPRE tilting operate inside the pistons. There are no crankcases. 

In comparison, a single crankshaft having three times larger stroke needs a big crankcase and has a bigger front area (aerodynamic resistance).

 

Compare also the passageways of the mixture. In the OS engine you have to pass it through a hole (7mm diameter) in the crankshaft.

 

I see a lot of advantages, without significant drawbacks. If I had to built an RC engine to win the competition, it would be an OPRE tilting.

 

I asked where the real problem of the "challenge" is.

Until now I didn't see a straight answer.

Please focus on the problem you see is the most critical, the most difficult to be solved.

 

Thanks

Manolis Pattakos



#53 Kelpiecross

Kelpiecross
  • Member

  • 907 posts
  • Joined: October 10

Posted 30 April 2014 - 06:12


Speedy - I had a quick look at the rules for F3D Pylon Racing and they appear to allow for twin-engine 'planes. Generally speaking both twin-engined real and model planes are faster and generally perform better than single-engined of the same engine capacity - presumably because of the "propeller disc efficiency" (more disc area is better apparently) and fact that two smaller engines will make more power than one big one.

These very high RPM 2-strokes are very tricky to handle - so getting two to operate at the same time could be tricky - two engines would sound good though.

#54 Speedman

Speedman
  • Member

  • 53 posts
  • Joined: November 05

Posted 30 April 2014 - 10:03

Hi Manoli

 

 

 

The OS RC engine with 15mm stroke has at 30,500rpm a mean piston speed of 15.25m/sec (not so high) and - as they write - can operate above 40,000rpm (more than 20m/sec).

Are you sure it cannot operate reliably (for "long", what long means for a model airplane) at 30Krpm with full throttle? Where is the problem?

If it can, make a three cylinder and you have 6.8bhp from 9cc.

 

 

I haven't wirite the piston speed are a problem. The piston speed in model engines what never a problem. I habe write the problem up.

 

Sorry but your assumptions are more theory than practice. A 3 cylinder engine is too heavy, you need 3 reconance exhaust and you need  lots of space.

Aerodynamic is more importent then engine power by 400 km/h. And is it very difficult 3-cylilnder engine optimally adjust. And a uniform cooling of all cylinders is very very importent. Proper racing glow engine are very sensitive.

 

It is not certain whether a 10cc engine with 35'000rpm  a good concept is. 35'000 rpm a bad for a good airscrew  efficiency. And a reduction gearbox is too heavy.

 

 

 

I see a lot of advantages, without significant drawbacks. If I had to built an RC engine to win the competition, it would be an OPRE tilting.

 

 

Maybe, but I only believe what I see.

Please buy a proper race nitro engine and you will see, ist not so easy as you think. Theory and practice are not the same.

Or all  F3S builder are Idiots and you a genius. I don't think  F3S builder are Idiote, but practitioners.

 

 

These very high RPM 2-strokes are very tricky to handle - so getting two to operate at the same time could be tricky - two engines would sound good though.

 

Thats right.

 

Speedy - I had a quick look at the rules for F3D Pylon Racing and they appear to allow for twin-engine 'planes.

 

 

F3D is a whole other world, but a real F3D pilot build a plane with one engine. The most winners had planes with one engine.

 

 

best regards

speedman.


Edited by Speedman, 30 April 2014 - 15:46.


#55 malbear

malbear
  • Member

  • 243 posts
  • Joined: September 02

Posted 04 May 2014 - 09:33

prius_atkinson_450.jpgjust as I suspected an offset bore to crank



#56 malbear

malbear
  • Member

  • 243 posts
  • Joined: September 02

Posted 04 May 2014 - 09:36

atkinson.gif



#57 manolis

manolis
  • Member

  • 392 posts
  • Joined: May 03

Posted 04 May 2014 - 14:52

Hello malbear.

I suppose the engine and the timing plot of your posts are from the Toyota Prius.

The timing diagram is not good for a conventional non-hybrid engine wherein the operation at very light load is quite common.

If, without using a throttle valve, you want to further reduce the load (urban cycle, for instance), you have to keep the intake valves open for several more degrees after the BDC.
I.e. you have to further retard the intake opening.
But this spoils the cycle at, and a little after, the overlap TDC.
The piston moves away from the overlap TDC with the intake valves still closed.


The solution is to have a VVA of variable duration, like the PatAir:

mito_cam_lifts.gif

at http://www.pattakon....ttakonHydro.htm

The angle wherein the intake valves open is optimized. The angle wherein the intake valves close can vary a lot (you can keep the intake valves open even at the "combustion TDC" to de-activate a cylinder.


In order to combine the fuel efficiency of the Atkinson cycle with plenty of power, you need conventional valve modes, too.

So, a simple VVT (variable valve timing, or phaser) is not the solution for a non-hybrid car operating in the Atkinson cycle.

A good VVA (Variable Valve Actuation system) combined with a good VCR (Variable Compression Ratio) is the solution.

Thanks
Manolis Pattakos

#58 Canuck

Canuck
  • Member

  • 1,664 posts
  • Joined: March 05

Posted 04 May 2014 - 16:56

As someone who worked extensively with Beryllium based materials, they are quite usable, quite workable and quite machinable. As with anything, appropriate precautions are easily undertaken. In the case of beryllium, don't machine (cut, turn, mill, grind, etc.) dry. Always wet machining.

#59 Kelpiecross

Kelpiecross
  • Member

  • 907 posts
  • Joined: October 10

Posted 05 May 2014 - 06:54

prius_atkinson_450.jpgjust as I suspected an offset bore to crank


And if you made the offset variable you could have VCR as well.

Advertisement

#60 manolis

manolis
  • Member

  • 392 posts
  • Joined: May 03

Posted 05 May 2014 - 12:59

And if you made the offset variable you could have VCR as well.

Hello Kelpiecross.

If, in order to have Variable Compression Ratio (VCR), you move the crankshaft basic bearing (and, necessarily, the flywheel and the gearbox), it is better to move them along the cylinder axes (in the other direction, even a big change of the offset causes just a small variation of the compression ratio; and a big offset brings more problems than it solves).

Thanks
Manolis Pattakos

#61 Kelpiecross

Kelpiecross
  • Member

  • 907 posts
  • Joined: October 10

Posted 05 May 2014 - 13:27

Hello Kelpiecross.

If, in order to have Variable Compression Ratio (VCR), you move the crankshaft basic bearing (and, necessarily, the flywheel and the gearbox), it is better to move them along the cylinder axes (in the other direction, even a big change of the offset causes just a small variation of the compression ratio; and a big offset brings more problems than it solves).

Thanks
Manolis Pattakos


I didn't claim that it was a good idea - just that it would have a VCR effect.

#62 Powersteer

Powersteer
  • Member

  • 2,460 posts
  • Joined: September 00

Posted 06 May 2014 - 06:51

atkinson.gif

Bit curious with part or light throttle application. On light throttle, the cylinder would have vacuum or negative pressure cause by throttle butterfly blocking most of intake upon intake valve closing, if I were to assume this, pumping losses during the compression stroke should not be much of a hindrance towards inefficiency and the Atkinson Toyota style would not have much of a point? Again, assuming, it might even draw in more air as vacuum is still happening post bottom dead center. Will vacuum on light or mid throttle reach neutral pressure way beyond valve closing? It may be less efficient on light throttle. Another method of Atkinson is simply to lengthen the combustion stroke with exhaust valve opening advanced(crank cycle or open later).

 

Another possible way is to have no throttle and let the intake valve open all the way towards close to TDC having a really short compression stroke prior to combustion. This will take away all the energy needed for vacuum and compression. Have the engine breath and let the intake valve close and direct fuel injection determine engine output. Swirl can also be harnessed with low valve lift. In fact, let all that energy harness swirl instead.

 

:cool:


Edited by Powersteer, 06 May 2014 - 07:14.


#63 manolis

manolis
  • Member

  • 392 posts
  • Joined: May 03

Posted 06 May 2014 - 10:55

Bit curious with part or light throttle application. On light throttle, the cylinder would have vacuum or negative pressure cause by throttle butterfly blocking most of intake upon intake valve closing, if I were to assume this, pumping losses during the compression stroke should not be much of a hindrance towards inefficiency and the Atkinson Toyota style would not have much of a point? Again, assuming, it might even draw in more air as vacuum is still happening post bottom dead center. Will vacuum on light or mid throttle reach neutral pressure way beyond valve closing? It may be less efficient on light throttle. Another method of Atkinson is simply to lengthen the combustion stroke with exhaust valve opening advanced(crank cycle or open later).
Another possible way is to have no throttle and let the intake valve open all the way towards close to TDC having a really short compression stroke prior to combustion. This will take away all the energy needed for vacuum and compression. Have the engine breath and let the intake valve close and direct fuel injection determine engine output. Swirl can also be harnessed with low valve lift. In fact, let all that energy harness swirl instead.
:cool:

Hello Powersteer.

The Gomacsys VCR operates in a limited Atkinson cycle.
Depending on the compression ratio, the expansion piston stroke can be longer than the compression piston stroke (into limits).

With a throttle valve the pumping loss increases:

PumpingLoss.gif

In the Prius the engine operates only at medium / heavy load, with the surplus energy strored in batteries for later use.
In the non-hybrid cars, there is problem. A VVT is not adequate.

In the following plot:

PatAir2.gif

(it is the PatAir VVA at http://www.pattakon....ttakonHydro.htm )

the engine can operate without a throttle in all the modes you describe.

You don't even need a VVT system. You don't even need a second camshaft for the exhaust.

The ECU triggers the moment the intake valves disengage from the cam and close.

For each load you have several options (early intake valve closing, late intake valve closing, multilift etc). The ECU desides which is the best for the existing operational conditions (for instance, during the warming-up period the "Ingoing Air Contro" may be prefferable, and when the cooling temperature is high, the ECU uses the "Outgoing Air Contro"). The more the variability of the VVA, the better.
The deactivation of cylinders is also an easy option.

Thanks
Manolis Pattakos

Edited by manolis, 06 May 2014 - 11:01.


#64 Powersteer

Powersteer
  • Member

  • 2,460 posts
  • Joined: September 00

Posted 06 May 2014 - 20:18

thats very interesting. the out going in going air control graph for piston position and valve lift with the lines being cylinder pressure? and why patair? i think you have got other continuously variable valve control far more sophisticated than patair which is similar to fiats. so which would be more efficient? longer duration breathing the engine or short valve lift during mild to moderate engine demands? longer duration would certainly allow swirl would it and i think swirl during low engine speeds is quite important. what if the negine was turbocharged?

 

:cool:



#65 manolis

manolis
  • Member

  • 392 posts
  • Joined: May 03

Posted 07 May 2014 - 13:52

thats very interesting. the out going in going air control graph for piston position and valve lift with the lines being cylinder pressure? and why patair? i think you have got other continuously variable valve control far more sophisticated than patair which is similar to fiats. so which would be more efficient? longer duration breathing the engine or short valve lift during mild to moderate engine demands? longer duration would certainly allow swirl would it and i think swirl during low engine speeds is quite important. what if the negine was turbocharged?


Hello Powersteer.

The plots at the top show the valve lift vs the crank angle and are only indicative (the real think is at http://www.pattakon....o_cam_lifts.gif that shows only the “cam lift”, which defines the “valve lift” envelope).

The plots at bottom are also indicative. They are P-V diagrams. The horizontal axis is the volume, the vertical axis is the pressure. The “area” below the curves gives the related energy (in the animated P-V plot of my last post, the yellow area is proportional to the energy consumed “before the actual compression” of the charge).


The two P-V plots show a significant difference between the throttle-less “ingoing” air control at light load and the “throttle-less” “ingoing air control at light load, too.

In the “ingoing air control” after the closing of the valve it follows an expansion of the trapped air, heating of the expanded air by the cylinder walls and then compression of the warm air. Mechanical energy is consumed / lost.

In the “outgoing air control” the intake valves stay wide open until a good part of the entered to the cylinder air / mixture to return back to the intake manifold by the upwards moving piston. The energy loss is minimized.


The double-line curve in the valve lift vs. crankshaft angle plots is the “valve lift envelope”. The valve lift is always inside the “envelope” even during a malfunction. This is vital for the safe operation of a valve train: “valve – piston” collision cannot happen (unless in the case of a broken timing belt, as in the conventional valve trains).


The control is electronic and independent in each cylinder; nothing to do with the mechanical-“electronic” control of the valvetronic of BMW which was regarded as the “state-of-the-art” VVA (before the presentation of the MultiAir of FIAT – INA). The true electronic control of the MultiAir / PatAir offers an incredible flexibility (not possible in the mechanical systems).


The MultiAir can operate only according the “Ingoing Air Control”.

The PatAir operates according the “Ingoing Air Control” and according the “Outgoing Air control” (the ECU decides, based on the operational conditions, which is the best for the instance). It provides all the modes the MultiAir can operate, and an infinity of additional highly efficient modes (unlimited Miller / Atkinson cycle).
Entering and leaving the cylinder a few times before the combustion, the mixture (case of indirect injection) in the "Outgoing Air Control" mode becomes "perfectly" homogeneous.
Regarding the swirl just before the ignition, the “Outgoing Air Control” appears better. In the “Ingoing Air Control” the air / mixture trapped in the cylinder has plenty of time to “calm down” / fade the initial swirl.

To be mentioned that as the load increases the two modes (Ingoing and Outgoing Air Control) get progressively similar (at full load there is no difference between them).

Both systems fit with turbo-charging.

Both (MultiAir and PatAir) are for normal revs (say till 7,000 rpm).

A good VVA needs not to be over-sophisticated, at least not as regards its mechanism.
The simpler the mechanism and the less the necessary construction accuracy, the better.
The direct control by an electronic control unit (ECU, based on feedback) seems to show the way.
Comparing the expensive valvetronic of BMW with the cheap MultiAir of Fiat is like comparing the old mechanical spark ignition systems (those with the "centrifugal" spark advance) with the modern Electronic Ignition Systems.


The HyDesmo at http://www.pattakon....akonHyDesmo.htm is a better choice; it can also operate at higher revs. It needs not restoring valve springs (Desmodromic Hydraulic)

For sport / racing engines the mechanical VVA’s have their own advantages.

Thanks
Manolis Pattakos

#66 Powersteer

Powersteer
  • Member

  • 2,460 posts
  • Joined: September 00

Posted 19 May 2014 - 00:54

324.png

I fully see the pressure changes now.

 

324.pngSeems the 5-Stroke engine utilizes a central piston for heat energy recapture with 26% cylinder pressure. If mechanical losses and added weight are smaller than it would be good on small cars although turbocharging would have less energy to spool, its concentrated charge creating a lot of heat might reap even larger rewards with that central piston 

 

:cool:


Edited by Powersteer, 19 May 2014 - 01:14.


#67 malbear

malbear
  • Member

  • 243 posts
  • Joined: September 02

Posted 01 June 2014 - 06:47

http://www.mechadyne...ts/new-products

 

This company does a variable valve tech



#68 indigoid

indigoid
  • Member

  • 384 posts
  • Joined: March 04

Posted 01 June 2014 - 12:13



 

Seems the 5-Stroke engine utilizes a central piston for heat energy recapture with 26% cylinder pressure. If mechanical losses and added weight are smaller than it would be good on small cars although turbocharging would have less energy to spool, its concentrated charge creating a lot of heat might reap even larger rewards with that central piston 

 

Difficult to achieve a nice engine balance, do you think? BMW's inline 2-cyl bike engines have a third rod with no piston attached as a weird sort of balancer, more info here. Seems to work. I've ridden a few of those bikes and they certainly don't vibrate much. They do totally lack character, but that's not a technical problem :)



#69 Powersteer

Powersteer
  • Member

  • 2,460 posts
  • Joined: September 00

Posted 01 June 2014 - 13:27

It would have a balance like an in-line 4 if they can get the central system balanced with the outer two piston. Thing about that 5 stroke engine is, in true Atkinson stroked engine with a mechanically much longer combustion stroke, the same can be achieve but even more efficiently doing away with an entire cylinder, cylinder head and crank allocated for it.

#70 manolis

manolis
  • Member

  • 392 posts
  • Joined: May 03

Posted 26 June 2014 - 09:21

http://www.mechadyne...ts/new-products
This company does a variable valve tech

Hello Malbear.

The US8,776,749 patent was granted today to the HyDesmo VVA (presented at http://www.pattakon....akonHyDesmo.htm ) .

Can you compare the available modes of the VVA systems of Mechadyne with those of the HyDesmo of pattakon?
Or the control "mechanisms" for switching between different valve lift profiles?
(the ECU of the HyDesmo just triggers the moment of the valve opening and the moment of the valve closing, independently in each cylinder; for instance, advancing the opening of the intake valve of the 2nd cylinder by 0.00001 seconds, and retarding the closing of the same intake valve by 0.000025 sec (easy for a computer), during the next cycle of the engine this intake valve will remain open for another 0,000035 seconds. Interesting?).


Quote from "Engine Technology International" (June 2014):

"Ford reveals 140ps 1-litre:
New editions, featuring a 140ps version of the manufacturer’s award-winning EcoBoost engine, will be the world’s most powerful 1-liter production cars, boasting more power per liter than the Bugatti Veyron and Ferrari 458"

Imagine the same 140bhp/lit small engine of Ford having the HyDesmo in its cylinder head and operating - when desired - in the fuel efficient Atkinson / Miller cycle.

Thanks
Manolis Pattakos

Edited by manolis, 26 June 2014 - 09:26.