[Technical] Two Stroke Engine Design

Hello everyone I stumbled across this forum and have read through most of the engine topics and have found the wealth of information and experience here to be very good when it comes to the two stroke engine. I apologize in advance for not knowing any French so I will do my best to use Google translator to make my post bilingual. I am currently a university student studying motorcycle engineering and engine design in Swansea, Wales and will be starting my third and final year in the fall. I am extremely passionate about advancing two-stroke engine technology and have chosen to design a 500cc two-stroke single cylinder engine as my final year project. I have already built a 500cc super mono grand prix bike which can be seen below and would like to complete it by designing my own engine for it.

What I would like to learn more about is the design process if anyone has any experience designing engines from scratch? I have a few different engines and cylinders that I am interesting in taking some of there better qualities from and making one improved cylinder from them. I have recently acquried EngMod2t software which is capable of modeling the gas dynamics of the engine to aid in the design of the engine. I've also got Tom Turner's TSR software which does a similar thing. The models I create should give some information on port configuration, timing, pipe, and intake dimensions. Once some basic parameters are known I can begin to model the cylinder in a CAD program and start designing the engine.

My tentative parameters for the engine:
Bore: 92mm
Stroke: 75.2mm
Capacity:499.90cc
Induction: case reed valve
Guillotine power valves actuated by servo motor
~100hp @ 9500RPM
Fuel Injected with micropump controlling oil mixture

My main concern is getting the transfer port shapes right and flowing in the right direction, getting the transfer port lengths correct (is resonance in the port a big factor), and designing a good flowing crankcase. Has any one flowed a cylinder using the Juante method to look at the scavenge pattern. This is something that could be done in CFD prior to any cylinders being made but I don't know if it will tell the whole story? I have a lot of other questions but I'll leave it at this for now and see if anyone can help or is willing to share information.

Here are a couple of photos of the bike. It currently has a KX500 dirt bike engine in it that has been modified to better suit road racing.

[Vous devez être inscrit et connecté pour voir cette image]

I will be running the first race this summer on carbs and then hopefully switching over to the fuel injection system I've designed for it.

A link to more photos: [Vous devez être inscrit et connecté pour voir ce lien]

Thanks for any help

Paul

Bonjour à tous je suis tombé sur ce forum et j'ai lu dans la plupart des sujets de moteur et nous avons trouvé la richesse des informations et d'expériences ici pour être très bon quand il s'agit de moteur à deux temps. Je m'excuse à l'avance pour ne pas savoir du tout le français donc je ferai de mon mieux pour utiliser Google translator pour rendre mon poste bilingue. Je suis actuellement étudiant à l'université des études d'ingénieur de moto et la conception des moteurs à Swansea, au Pays de Galles et sera de commencer ma troisième et dernière année à l'automne. Je suis extrêmement passionné par les progrès technologiques du moteur à deux temps et nous avons choisi de concevoir une 500cc deux temps moteur monocylindre que mon projet de dernière année. J'ai déjà construit un super-500cc mono Grand Prix moto qui peut être vu ci-dessous et que vous souhaitez le compléter par la conception de mon propre moteur pour elle.

Ce que je voudrais en apprendre davantage sur le processus de conception si quelqu'un a une expérience dans la conception des moteurs à partir de zéro? J'ai quelques moteurs différents et des cylindres que je suis intéressant dans la prise de certains il ya de meilleures qualités et de faire un cylindre amélioration de leur part. J'ai récemment acquried logiciels EngMod2t qui est capable de modéliser la dynamique des gaz du moteur pour aider à la conception du moteur. J'ai aussi un logiciel TSR Tom Turner, qui fait une chose semblable. Les modèles que je crée doit donner quelques informations sur la configuration du port, le calendrier, les tuyaux, et les dimensions d'admission. Une fois que certains paramètres de base sont connus, je peux commencer à modéliser le cylindre dans un programme de CAO et de commencer à concevoir le moteur.

Mes paramètres provisoires pour le moteur:
Alésage: 92mm
Course: 75.2mm
Capacité: 499.90cc
Induction: cas à clapets
Vannes d'alimentation guillotine actionnés par servo-moteur
~ 100 ch @ 9500RPM
Injection d'essence avec le mélange d'huile de contrôle micropompe

Ma principale préoccupation est d'obtenir les formes portuaires de transfert droite et circulant dans la bonne direction, d'obtenir les longueurs de transfert du port correct (est la résonance dans le port un facteur important), et la conception d'un carter bonne fluidité. At-on coulait un cylindre à l'aide de la méthode Juante de regarder le modèle récupération. C'est quelque chose qui pourrait être fait en CFD avant toute cylindres étant faites, mais je ne sais pas si elle sera raconter toute l'histoire? J'ai beaucoup d'autres questions mais je vais m'en tenir à cela pour maintenant et voir si quelqu'un peut aider ou est prêt à partager l'information.

Voici quelques photos de la moto. Il dispose actuellement d'un moteur de dirt bike dans KX500-ce qui a été modifiée pour la course conviennent mieux la route.

^^^

Je vais courir la première course cet été sur les glucides et alors on peut espérer de commutation vers le système d'injection de carburant, j'ai conçu pour elle.

Un lien vers d'autres photos: [Vous devez être inscrit et connecté pour voir ce lien]

Merci pour toute aide

Paul

Hello Paul,

Why do you want to make such a short stroke engine?
Square bore and stroke dimensions would be FAR better!
It gives you more port surface, which is power determinating.
And with a short stroke the piston will become very hot.
Personally I would make a slightly long stroke engine.
I cannot see 1 single advantage for a short stroke!

Jan Thiel a écrit:

Hello Paul,

Why do you want to make such a short stroke engine?
Square bore and stroke dimensions would be FAR better!
It gives you more port surface, which is power determinating.
And with a short stroke the piston will become very hot.
Personally I would make a slightly long stroke engine.
I cannot see 1 single advantage for a short stroke!

I believe to get the engine power I want it will have to rev fairly high for a big thumper. With a shorter stroke the motor can spin faster without having a high mean piston speed.

[Vous devez être inscrit et connecté pour voir cette image]

G.P. Blair suggested a good starting point for race engines was a bore stroke ratio of 1.2 in his book so that is another reason I've chosen it to start with. I believe about the max mean piston speed that should be run is around 25m/s. If higher piston speeds can be safely achieved then I would consider going towards a more square ratio for the reasons you've pointed out. Perhaps 100hp can be achieved at a lower RPM in which case the ratio can be squared up as well? The current engine is 86x86 and I intend to make peak power at 8500RPM with it. When I go home for the summer I will be able to dyno it and see how much power it produces which will give me a better feel for what will be needed to get more out of the new engine.

When Mr Jan Thiel says = forget the short stroke option you should read and say 'yes'

then get a plan to use square stroke/bore

Then 100hp is not a reliable option except with super large long rods

Get a search on sabertooth on google , only reliable option IMO
95*95 option ....

Do'nt believe Dr. Blair, he never made a GP-winning engine!
Better keep your 86X86 bore and stroke.
You should easily be able to achieve 120-130 HP with it.
If you do everything right!
Achieving the same MEP as the Aprilia 125 you should even achieve 140HP/8500
at the gearbox output shaft.
But you will not find a big-enough carburetor for it, should be around 65mm!
So fuel injection is the way to go!

PVO a écrit:

G.P. Blair suggested a good starting point for race engines was a bore stroke ratio of 1.2

To starting riders and tuners I keep saying: "Take advice from the guys who are faster than you; not from the ones behind you".
Engines designed by Jan Thiel won 26 constructor's world championships and 25 rider's world championships (or the other way around; I lost count). Engines designed by Blair won zero championships; they did not even win a single GP.

Citation :

I believe about the max mean piston speed that should be run is around 25m/s. If higher piston speeds can be safely achieved then I would consider going towards a more square ratio for the reasons you've pointed out.

That is putting the carriage before the horse, Paul. In a racing two-stroke, rpm is not limited by piston speed but by time.area. It might be a good idea to read a bit more about that subject.

For a 500 cc single you should be aiming for maximum torque @ 7500 rpm, maximum power @ 8000 rpm and max.revs @ 9000 rpm. Revving it any higher will not improve performance; it will only diminish reliability.

Frits Overmars a écrit:

PVO a écrit:

G.P. Blair suggested a good starting point for race engines was a bore stroke ratio of 1.2

To starting riders and tuners I keep saying: "Take advice from the guys who are faster than you; not from the ones behind you".
Engines designed by Jan Thiel won 26 constructor's world championships and 25 rider's world championships (or the other way around; I lost count). Engines designed by Blair won zero championships; they did not even win a single GP.

Citation :

I believe about the max mean piston speed that should be run is around 25m/s. If higher piston speeds can be safely achieved then I would consider going towards a more square ratio for the reasons you've pointed out.

That is putting the carriage before the horse, Paul. In a racing two-stroke, rpm is not limited by piston speed but by time.area. It might be a good idea to read a bit more about that subject.

After reading some of Mr. Thiel's posts I now understand that he worked at Aprillia and is quite knowledgeable on the subject. Being new here I simply did not know who he was and agree with you completely and will take his advice very seriously.

Jan Thiel a écrit:

Do'nt believe Dr. Blair, he never made a GP-winning engine!
Better keep your 86X86 bore and stroke.
You should easily be able to achieve 120-130 HP with it.
If you do everything right!
Achieving the same MEP as the Aprilia 125 you should even achieve 140HP/8500
at the gearbox output shaft.
But you will not find a big-enough carburetor for it, should be around 65mm!
So fuel injection is the way to go!

Point taken! So with the bore/stroke established what would be my next step for designing the cylinder? Would I simply get a 125 cylinder and apply the appropriate scale factors to scale it up? I read that when you were at Aprillia you didn't use computer software? Was this because all the programs were that bad or was it simply easier to make the changes and test them?

Hello Paul,

Scaling up a 125 cilinder would be the best idea I think.
We once had a computer software on loan from a world-wide known company.
But exhaust pipes designed with this program were a complete failure!
Flow testing the trasfers proved a bit useless.
So we made many different types and tested them.
The last 125 cilinders were the result of many years of testing.

Jan Thiel a écrit:

But you will not find a big-enough carburetor for it, should be around 65mm!
So fuel injection is the way to go!

what about two 45mm carburetors ?

there is a twin 44 oval carb in the us for the 500CC single class ...

Why is your exhaust so crumpled up?

niels a écrit:

Why is your exhaust so crumpled up?

Fitment issues. Running it straight out would result in ground clearance issues when cornering so sweeping it back and forth was the only way to get it to fit nicely.

PVO a écrit:

niels a écrit:

Why is your exhaust so crumpled up?

Fitment issues. Running it straight out would result in ground clearance issues when cornering so sweeping it back and forth was the only way to get it to fit nicely.

Are those issues caused by your 'up-side down' swingarm? It doesn't look like that's helping.

The bike looks good though!

SEB4LO a écrit:

there is a twin 44 oval carb in the us for the 500CC single class ...

Can you show us a picture Seb? Or a link?

can't find anymore but Calvin @ Cpindinc on TST board should find them

Jan Thiel a écrit:

Scaling up a 125 cilinder would be the best idea I think.

If I could ask an ignorant question about this : I had thought that an issue with larger bore 2-strokes used at sustained high rpm was detonation, because of the difficulty of getting sufficient burn speed across a wide combustion chamber.

Or has squish design moved on enough from when I read/thought about this that it is no longer an issue? I guess max torque at 7500rpm in an 86mm bore should be no worse than a 54x54 motor at 12k rpm... ??

Maybe it's just me, but I think this thread is going to be very interesting

I think that you came to the right place for your project PVO. By the way, why did you chose to do a 500cc? I'm curious because a bigger engine hasn't always made a bike faster, especially with a big displacement per cylinder. A 500cc mono-cylinder 2 stroke is very unusual in road racing...

And when I say "faster" I mean lap times, not pure acceleration.

Edouard B. a écrit:

... why did you chose to do a 500cc? I'm curious because a bigger engine hasn't always made a bike faster, especially with a big displacement per cylinder. A 500cc mono-cylinder 2 stroke is very unusual in road racing...

I know, I raced one 40 years ago because all I could afford was a second-hand motocross engine, around which I built a monocoque frame. Bumpstarting the beast was an art...

Edouard B. a écrit:

And when I say "faster" I'm mean lap times.

It seems that nowadays people use stopwatches and even laptimers. I used a calendar....

Frits Overmars a écrit:

I used a calendar....

If average lap-times should include those which were completed on the recovery truck.... yes, I could have used one

A very interesting thread. Jan your acomplishments are very impressive. I agree with you 100% that a square or slightly undersquare (longer stroke) motor is the best design to take advantage of cylinderwall port area. I am interested in the areas that you made the most gains in transfer design. I have a large number of cylinders on the shelf that now seem to be scrap because the dyno results show loss of power after the hours spent shaping the ports to what was thought to be optimum. My dyno is a very accurate Schenck model W70 controlled by Dynomax 2010 software so I can accurately measure and repeat runs. I can measure torque, rpm, hp fuel flow air flow coolant in/out temp/ egts bsfc a/fuel ratios.
The engine I am developing is 65 mm bore 65.8 mm stroke twin cylinder with a transverse rv shaft that controls both intake ports. We rev the engine to 10500. I would like to rev higher but the cvt we use looses efficiency above that. The port timing I am most successful with to date is 80 degrees atdc exhaust opening, ( main and auxillary exhaust all open the same) 111 degrees rear transfer opening and 113 degrees opening on the main and auxillary transfers. The best power for the rv timing is 150 btdc opening and 105 atdc closing. The engine requires a variable inlet tiiming arrangement that closes at 80 degrees atdc to start with the late closing. We are mandated to use 34 mm carbs so that is why (according to Frits) is why we need the realy late closing.
My question to you is can you comment on the port timing and give me some indication as to what area of the transfer angles and timing I should concentrate on to find gains. Thanks for your much appreciated wisdom./
Howard..

niels a écrit:

PVO a écrit:

niels a écrit:

Why is your exhaust so crumpled up?

Fitment issues. Running it straight out would result in ground clearance issues when cornering so sweeping it back and forth was the only way to get it to fit nicely.

Are those issues caused by your 'up-side down' swingarm? It doesn't look like that's helping.

The bike looks good though!

Thanks

Probably to some extent yes. The project originated as "I'll just make a frame and use a swingarm from another manufacturer" which progressively turned into "I'll fabricate nearly every last piece on the bike". Had I known at the time the Honda Swingarm I was going to use wasn't going to work I would have redesigned part of the frame and changed the position of the shock. The linkage layout I chose ultimately resulted in the swingarm being "upside down".

Just like this engine design project I want to take on designing and building the bike was a learning activity which has resulted in a wealth of knowledge. Unfortunately, I did not for see everything that would impose a problem and there are already some design changes I'd like to make to the bike and I've not even had it out for its first track day! My other problem is I designed the bike at university and didn't have all the components I was going to use at hand so the CAD model was never completed in detail prior to the start of construction.

Frits Overmars a écrit:

Edouard B. a écrit:

... why did you chose to do a 500cc? I'm curious because a bigger engine hasn't always made a bike faster, especially with a big displacement per cylinder. A 500cc mono-cylinder 2 stroke is very unusual in road racing...

I know, I raced one 40 years ago because all I could afford was a second-hand motocross engine, around which I built a monocoque frame. Bumpstarting the beast was an art...

Edouard B. a écrit:

And when I say "faster" I'm mean lap times.

It seems that nowadays people use stopwatches and even laptimers. I used a calendar....

Frits is partly right. I had a KX500 dirt bike prior to the super mono so I knew a bit about the engine and costs could be kept down in theory. The other factor was I wanted a bike that could be competitive in multiple classes. The club I race in allows singles of any displacement, two or four stroke to compete in the ultralight class which primarily has Ducati Monster 620s and 450cc dirt bikes in it so it should be easy to be competitive. The next class is lightweight (SV650s and ER6s) where I should still be competitive and the final class I want to run is middleweight with the 600cc four-strokes. Since grand prix racing seems to be dead in America for me to make a trip to the track to run one race would not be worth it.

From a development cost point of view a single cylinder is the cheapest to run and make. My dream is to see road going two strokes again one day and if I can help develop the engines to make that happen then I would be pretty happy. Once I've got a single cylinder that I'm happy with making a twin and or scaling the displacement shouldn't be too hard.

Frits what balance factor did you run in your 500 single? I'm having trouble with engine vibration making it unpleasant to ride and my engine design professor and myself have decided to try and balance the crank to 100% forward balance to change the plane of vibration to the horizontal in hopes of the vibrations being less felt. If that doesn't work I've designed a counter balance I think I can fit into the engine however that will be a lot more machine work and I'm not sure if I can get enough weight on the counter balancer. Forgive me if it has been covered elsewhere but what percent of mass did the RSA 125 counter balancer cancel out? I've read 30-50% is enough but have no first hand knowledge.

Jan Thiel a écrit:

Hello Paul,

Scaling up a 125 cilinder would be the best idea I think.
We once had a computer software on loan from a world-wide known company.
But exhaust pipes designed with this program were a complete failure!
Flow testing the trasfers proved a bit useless.
So we made many different types and tested them.
The last 125 cilinders were the result of many years of testing.

Hi Jan,

Is there a road going 125 or 250 twin you would recommend I start with as I doubt I'll get my hands on a RSA 125 or some of the karting replicas? Perhaps any 125 will do for modeling purposes since it is unlikely I will get the ports right on my first go?