DirtRider . Net MX, SX, Arena Cross, Off-Road Community
Dirt Rider . Net Text Version Home
Dirt Bike Dirt Bike Dirt Bike Dirt Bike

This is the text version of DirtRider.Net
Click Here for the Full Version


Pages: 1

Injected two strokes

(Click here to view the original thread with full colors/images)

Posted by: Doc

nephron wrote in the Oxygen Meter thread:

Quote:
…what I thought would make it difficult to inject a 2 stroke successfully, particularly in closed loop. Looks like it's possible, though, under the right circumstances--since not only Yamaha, but also Kawasaki (one of their 2 smoke watercraft has DFI), have been doing this quite a while… Do you have any insight into Univ. of Waterloo's FI 2 cycle snowmobiles? I've been unable to find any detailed papers on the Kawi or Waterloo.



The Kawasaki PWC uses the Ficht direct injection system. The presented a paper at the SAE Small Engine Technology Conference in Pisa. The paper is SAE #2001 01 1787.

I expect that Waterloo will have their sled at the SAE Clean Snowmobile Challenge in Jackson Hole in March. I will be judging the written and oral presentations, I can let you know if there is something interesting. Last year’s Clean Snowmobile Challenge written papers were all published by SAE. I do not have the numbers handy, but you can look on the SAE web site: http://www.sae.org/servlets/techtra...&PROD_TYP=PAPER

As far as successful injected two strokes, it has already been done for five years on outboards, it is just a matter of someone doing it on a motorcycle.



Doc


Posted by: Rich Rohrich

Doc - We'd love to hear your feedback on the designs you judge.


On the DFI and EGT subject here's a couple of papers that I found pretty informative.


SAE 1999-01-3288 / JSAE 9938043
A Study of Direct Fuel Injection Two-Stroke Engine for
High Specific Power Output and High Engine Speed
Koji Morikawa and Hideharu Takimoto - Subaru Research Center Co., Ltd.
Taiichiroh Ogi - Fuji Heavy Industries, Ltd.

ABSTRACT
High specific power output two-stroke engine for snowmobile
use was converted to a Direct Fuel Injection (DFI)
engine, in order to achieve lower HC emission by avoiding
fuel short circuiting and to obtain higher specific
power output by increasing induction air and optimizing
A/F. High pressure single fluid fuel system was chosen
because of extremely high fuel delivery rate. The fuel
injector and its location were investigated and optimized
for better mixture formation and lower HC emission.
140kW/L level specific power output, maximum engine
speed of 9000 rpm, and 1/2 the level of HC emissions
were obtained.

SAE # 983072
Measurement of Exhaust Gas Temperatures in a High Performance Two-Stroke Engine

Robert J. Kee, Paul G. O’Reilly and Robert Fleck
The Queen’s University of Belfast
Paul T. McEntee
Optimum Power Technology

Abstract:
The performance characteristics of two-stroke engines are highly dependent upon the gas dynamic wave action in the exhaust system. In a tuned high performance exhaust system, negative suction pulses aid induction of charge into the cylinder, while positive waves aid its retention. The timing of these waves is closely related to the acoustic velocity, and is therefore dependent on the exhaust gas temperature (EGT).

In advanced engine management systems, the control strategy may be tailored to influence the EGT, and to maximize the beneficial influence of the gas dynamics in the exhaust. Therefore, accurate measurement of EGT is required for development purposes, and real-time feedback could potentially be used as an input to the management system. However, accurate measurement of exhaust gas temperature is fraught with difficulties due to a number of sources of error. Steady state errors can arise due to conduction and radiation of heat from the thermocouple junction, while in transient operation, additional errors arise due to the slow responses of the robust thermocouples which are required to survive the harsh environment.

In this study, the influence of the thermocouple geometry on the steady state errors has been estimated, and a technique has been developed to reconstruct mathematically the instantaneous temperature from the output of two dissimilar thermocouples. Results are presented to demonstrate that the reconstructed temperature correlates well with measurements from a fast-response thermocouple.


Posted by: bwalker

Ski doo has just released a DI 800cc snowmobile engine that injects into the transfer port. I believe this is the Orbitol system. SkiDoo has rights to both the Orbitol and the Ficht system. Check out http://www.snowmobilenews.com/output.cfm?id=102745


Posted by: Bobt250

Really cool stuff, check it out.

http://msnhomepages.talkcity.com/downsizedr/roehr/


http://www.orbeng.com.au/orbital/home/home.htm

Anyone remember or know what ever happened to that experimental Honda that reduced emissions by closing the exhaust port WAY down and turning off the spark plug?

Perhaps one or both of these technologies together will save the two stroke?


Posted by: nephron

Quote:
In this study, the influence of the thermocouple geometry on the steady state errors has been estimated, and a technique has been developed to reconstruct mathematically the instantaneous temperature from the output of two dissimilar thermocouples. Results are presented to demonstrate that the reconstructed temperature correlates well with measurements from a fast-response thermocouple.


...so instead of using real-time measurements (O2 sensing/EG temp, &/or using the fast-response thermocouple), they'll use a "reconstruction" regression formula to approximate fuel/ignition curves as they relate to throttle position or rpm? Is that right, or am I misreading it, Rich?

I don't have any problem with "open" FI. In fact, it's damn near as reliable as carburetion, eh? At least as regards cars--for instance the initial Holley throttle-body FI allowed for O2 feedback/closed loop as an "option" only. I never even bothered much with the feedback circuit, & simply relied on preexisting computation-adjusted fuel curves set by the computer based on throttle position. Of course, it didn't work that well, but was reliable.

Jack


Posted by: Sean Hilbert

Jack,

Ideally, if you are trying to control the air/fuel ratio in the cylinder, you would try to use feeback information in the control system that correlates as closely as possible to that end goal. The O2 sensor is the perfect device since the % Oxygen in a completely burned exhaust stream is directly proportional to the mixture composition, however, as Doc pointed out, its use in two-strokes is limited because of the engine's scavenging characteristics.

The use of exhaust temperature as feedback is mainly aimed at high performance applications where exhaust tuning is an important factor in the engine's overall performance. Since pipe tuning is affected by exhaust gas temperature, the engine's scavenging characteristics can change dramatically over the course of a transient event (load change, speed change)...thus different amounts of fuel may have to be injected for a given throttle position and RPM depending on how the engine arrived at that condition. For example, the amount of fuel injected for an engine running at 50% load, 6000 RPM would probably be different depending on if that condition were achieved by someone riding at a steady speed down the trail, or someone accelerating through the gears -- the reason being that the cylinder is scavenged differently under those two conditions (i.e. amount of fresh air vs. amount of residual exhaust from the previous cycle)...and one of the chief factors in determining that scavenging characteristic is exhaust temperature.

Note that a carb doesn't have the ability to distinguish between the two above transient conditions, and therefore cannot supply optimal jetting. Thus, an injected bike (properly calibrated) should promise very crisp response compared to a carb application.

Sean


Posted by: nephron

Thanks, Sean

Certainly, that aspect (of difficulty) of fuel metering in 2 stroke DFI is understandable, and has obviously been overcome (Yamaha, Kawi, Waterloo...). My question to anyone familiar with those models: what mechanism are they using in these setups to compensate? Are they using feedback data realtime or not?

OK, ok. I guess I could just order the papers above.:o


Posted by: bwalker

Here is a brief description of the Ski-doo system I mentioned above.

Rotax Semi-Direct Injection
On the Rotax 2-TEC 800, two fuel injectors per cylinder pump the optimal amount of atomized fuel into the transfer ports. The increased volume of the dual injectors enables the system to pump the fuel into the transfer port at the last millisecond. At that point, the exhaust ports are closed, preventing unburnt fuel from being released through the exhaust.

Fuel economy is also increased and emissions reduced thanks to the smart nature of the system. Only one injector per cylinder operates at idle and low RPM. As the RPM and load increase, the MPEM (Multi Purpose Electronics Module) central computer instructs the second injectors to function.

The MPEM also optimizes engine performance. Sensors monitor ambient temperature, altitude, engine temperature, throttle position and RPM, then makes adjustments to optimize performance. The computer can adjust injection variables, ignition timing and exhaust valve (e-R.A.V.E.) operating timing.


Posted by: BRush

Interesting stuff. I’m assuming the Rotax system monitors ambient pressure and not altitude directly? Was there any mention of the capability of the controller? I’m curious about what frame rate is required to run a system like this (frame rate = the basic scheduling period of the controller encompassing one complete input-compute-output cycle).


Posted by: Nverfastenuf

Ski Doo has not yet released the DI 800(snowmobile) to the public. Polaris has the DI on their 800 watercraft and it is working well.


Posted by: bwalker

Quote:
Ski Doo has not yet released the DI 800(snowmobile) to the public

Yes they did.I cut and pasted the release in my post above.


Posted by: nephron

Quote:
The MPEM also optimizes engine performance. Sensors monitor ambient temperature, altitude, engine temperature, throttle position and RPM, then makes adjustments to optimize performance. The computer can adjust injection variables, ignition timing and exhaust valve (e-R.A.V.E.) operating timing


...so, looks open (to the most degree), with the only afferent/loop variable being the ambient motor temp. My guess is there are no functional EGT/O2 feedback systems in operation TODAY on popular watercraft or snowmobile.

Hopefully, I didn't miss another post describing just that monster...:o


Posted by: bwalker

Quote:
My guess is there are no functional EGT/O2 feedback systems in operation TODAY on popular watercraft or snowmobile.

Yes there is. Artic cat, Polaris and I think Skiddo on there 440cc race sleds all use a egt probe mounted on the fat section of the pipe to adjust timeing in order to keep pipe temp consistent. I believe they also link this egt info with a tps on the carb and the electronic powervalve servo. These sleds are all non direct injected though. BTW the mpem system that skiddo uses works very well. Sleds with the mpem are very consistent performance wise, which cannot be said of some of the others.


Posted by: nephron

Thanks, bwalker.

So, when you figure out how to inject your AF--I take first dibs on copying the design for my KX.


Posted by: bwalker

You know with the sled mfgs doind the leg work on this direct injection technology we may see it on bikes. I think the main hurdle to bike use would be to engineer throtle response into the sytem. That would take some serious r&d work, not to mention very high speed electronics.


Posted by: Sean Hilbert

Throttle response is part of it, however, crisp operation over the entire speed / load range is also a biggie. While sled engines are constrained in their speed / load range by the clutch (and boats by their prop characteristic), motorcycle engines have few bounds. It will take a few years after the high performance sled stuff arrives to see the first off-road applications.

Processor speed is also important, but Moore's law is firmly in force, so that shouldn't be a big issue.


Posted by: KnobShredder

I'm skeptical about Ski-Doo's 800 SDI system, if only because Bimota tried a similiar approach with the BB500 and lost. In Bimota's system they injected into a transfer port, which in turn deflected off the piston dome, giving even less time for the fuel to atomize in the combustion chamber, versus spraying it right into the combustion chamber. The engine actually ran well at high rpms, but was completely unstable at lower revs. Apparently Ski-Doo has found a way to overcome whatever it was that troubled the Bimota. Yes the outboard manufacturers DI engines are going great guns, but the one thing that is missing on those outboards is high rpm's - essential to a racing MX machine. I believe the Bimota's injectors allowed the engine to reach 10,200rpm, and this was in 1997. A few weeks back i had read a post by Eric Gorr where he said that the small droplets required for proper atomization wouldn't provide the cooling that the piston crown required at high rpms. I'm certain that with each year the technology will allow higher and higher rev ceilings, through better injectors, piston coatings etc.
In answer to an earlier post, that Honda system you were referring to was the EXP-2, which utilized Active Radical combustion. There are several websites with articles detailing how the system worked. From what i gather, the engine had greatly lowered emissions, and fuel mileage comparable to it's 4strokes, though they were never able to solve the unburned fuel problem inherant to conventional 2strokes, so the project was shelved.


Posted by: bwalker

Quote:
A few weeks back i had read a post by Eric Gorr where he said that the small droplets required for proper atomization wouldn't provide the cooling that the piston crown required at high rpms

This was a problem with the first DI sytems on outboards, but i beleive it has since been fixed.


Posted by: Daniniowa

Yamaha is using an oxygen sensor in their DI freshwater outboards. I think the sensor has a short life problem in the salt water applications, and thus is not used in the salt water 2-stroke outboards.

Mercury Marine Racing is getting 6500 RPM out of a 250 hp 3.0L DFI V-6, and 9000 RPM out of their 2.5L FI Drag racing outboard.

The droplet size and excess heat is a constant struggle. It is being managed, but improvements do continue to improve the cooling of the pistons.


Posted by: bwalker

I thought it might be a good time to bring this thread to the top. The Doo sytem mentioned above is in production now and works quit well and meets 06 standards. Polaris also has a simular sytem that is being released based AVL & Schrick inc technology. The AVL sytem uses two injectors positoned at the top of the intake boost port.
Lastly, LONG LIVE THE TWO STROKE!


Posted by: bwalker

Here is a schematic of the AVL sytem I was talking about.


Posted by: Kav

Doc: I saw your KTM at DW’03, that is a very impressive bike.

I’ve been wondering during recent road trips how is the amount of air going into the engine measured? Is it measured directly, or inferred from data gathered from other sensors? And if it’s measured directly is it measured at the reeds or in the intake port of the cylinder?


Posted by: bwalker

A MAF sensor maybe?


Posted by: Doc

Unfortunately, you have to map the flow rate from other sensors because a lot of the air that goes in goes right through. With 4-strokes it is a lot simpler because almost all the air that goes in is trapped in the cylinder. My 2-stroke engines are mapped RPM vs. TPS, with correction maps for atmospheric pressure and temperature.


Posted by: nephron

Doc, forgive my ignorance--but wouldn't MAP (in a way) account for the flow 'waste' of a 2 stroke?

Also, your current setup looks pretty similar to Ficht. How are your 2 strokes running?


Posted by: Folken83[MdW]

I'm an italian student of mechanical engineering at Univesity of Pisa;I'm writing a thesis on gasoline direct injection in two stroke engines to take degree.I would be very interested to your pubblications on this argument, therefore I was asking myself if kindly you could send those pubblications to me.Sorry if I give you any trouble.
Thank You Very Much.


Posted by: cujet

On another note, there are more and more direct gasoline injection engines available in auto's now. Maybe some of these components will be an easy way to make the DI 2 stroke work well. Volkswagen and Mitsubishi both make DI gasoline engines.

Chris


Posted by: Folken83[MdW]

My email is romano83mdw@hotmail.it Thank You Very Much


Posted by: dklink2000

This is a link to an RZ forum I frequent. He has a FI rz 350. For those of you that don't know it was a two stroke street bike yamaha put out about 20 years ago. It is basically a banshee ATV motor with power valves. I won't pretend to understand how he did it. Enjoy!

http://www.rzrd500.com/phpBB2/viewtopic.php?t=1453


Posted by: Joshtee

http://www.evinrude.com/en-US/Campaigns/Emissions.htm



This is the text version of DirtRider.Net
Click Here for the Full Version

Text Version Home





vBulletin Copyright ©2000 - 2008, Jelsoft Enterprises Limited.
vB Easy Archive Final ©2000 - 2008 - Created by Stefan "Xenon" Kaeser