Two-Stroke Engine Tuning: Maximizing Performance

[Averagesupernova]

Ok. It's a crude drawing but I think it will get the point across. I've drawn it as you would view it from the end of the crank. The blue is the piston and rod, the white represents the bore and block, the green is the crankshaft. The red is the valve. It isn't really a 'spool valve'. I decided that there is no need to run the scavenging air through the valve. Do it as I described in my first post. The little circle on the left represents the exhaust port. The lower circle on the right represents the fresh air only port for scavenging. The upper circle on the right is the air-fuel charge port. It is connected to the slide valve. The diagonal lines that are red represent a spring that pushes down on the end of the valve. The little horizontal lines are sealing rings. The valve is basically a straight rod with sealing ring grooves cut into it. The bottom end rides on a cam on the crank shaft. There is a LOW spot on the cam that causes the valve to drop when the crank gets to the right spot in its rotation. The upper sealing rings drop below the port in the cylinder bore opening up the port to the pressurized air. This would be about the moment that the injector fires. The injector is the pointy thing on the right side of the valve bore. The diagonal line at the top of the valve bore is a throttle valve. The main valve which is drawn in red and air-fuel port are actually drawn 90 degrees off of where they would actually be on the engine. The valve naturally needs to be arranged in-line with the crank.
-
Now for the shortcomings of this. This particular engine is not going to be able to be a real high reving engine because of the spring. The valve will float if it revs too high. This valve pops up and down TWICE as often as a valve in a 4 stroke engine. In order to prevent this it will have to be run with a connecting rod, wrist pin, and extra crank journal. That could be made to work ok I guess, but then the valve moves constantly throughout the cycle which means if the valve is to act quickly at the right moment it needs to move a fair distance. The port always needs to have sealing rings below it to prevent combustion pressure from blowing by down between the valve and the bore. The injector also needs to have sealing rings below it to prevent combustion pressure from reaching it. Sooooooo, if the valve were to be run off of a crank journal instead of a cam like I have it drawn, the injector will have to be higher up in the valve bore to guarantee that it always has sealing rings below it. I don't like it as high as it is in the bore and I certainly wouldn't want it any higher. I would prefer from a fuel delivery standpoint that it is directly across from the port.

 

[Ayrity]

very cool SUPERnova haha. I think it might be easier to just use an electronic valve than messing with all that stuff, at least for me rigging an engine up, and keeping weight down. I planned on adding a battery for an electric start anyway...

again i think we both have the same basic idea, ill draw out mine tomorrow sometime and show you.

 

[Ayrity]

Sorry I havnt drawn out my idea yet, busy school schedule... I am getting to it.
just another quick idea btw, how about making a mini super charger for the engine by moding a smog pump off of a car?

 

[Ayrity]

ok so I've acquired a small B+S 4 stroke engine (3hp) and I now think I am going to try to convert it into a 2 stroke and try to almost double my power. My plan so far is to get a smog (AIR) pump from a car at the local junkyard, and hook that up to a pulley on the crank somplace so as to pump air into the intake, like a mini supercharger. I will then calculate the amount of air needed to fully scavenge the exhaust out (a bit more volume than the displacement i would guess) and get the right ratio for the pulley wheel so the air pump will provide the correct volume of air to the engine.

Step 2 will be to create an intake tube with 2 paths that then lead back into one path, there would be 1 line that branched into 2 lines with a y splitter, than back together with another y splitter.

there will be a valve in the first Y split that will be electronically controlled. in the top path the carb will be split in. this will allow the air path to either go PUMP-->CARB-->ENGINE or PUMP-->ENGINE. this will let me use just air to scavenge the exhaust out rather than risk losing fuel. (im pretty sure the carb would have to put out a rather rich ratio)

the spark plug will have to fire twice as often, the second one being exactly 180 degrees off from when it fires now. I am pretty sure it uses a magnet on the flywheel so if this is right, it is probably sparking at that time anyway (during the exhaust stroke of the 4 stroke engine design). so i think that's covered.

the hardest problem to solve will be control of the intake and exhaust valves. they run on a cam shaft with lobes and everything like in a car engine. I have to be able to modify these valves to open and close twice as often... and maybe even modify when they do so. so far I have 2 ideas: the first is to take off the cam shaft timeing gear and replace it with another smaller gear, half as big=twice as often and run a chain to bridge the gap, or maybe find new gears for both the crank and cam shaft that makes the right gear ratio and mesh.

so far that is the idea, any help especially on the technical side of things would be GREATLY appreciated thank you everybody.

 

[NateTG]

If I were trying to do this my approach would probably be:
Convert to electronic ignition.
Convert to electronic valves.
Add driven air system (supercharger)
Add driven air scavenging system.
Use reprogramming of ignition/vales to switch between 2 and 4 stroke.

You could, very reasonably, get away with doubling the timing on the sparks, but the valve behavior is likely to be something that you're going to have to explore, and, quite frankly, you'll have an easier time of that with electronic valve control.

If you're good with a lathe, you could replace the camshaft.

Of course,

 

[Ayrity]

i do have access to a lathe here at my mechE lab, although i don't see how you can make lobes with a lathe... they arent circular all the way around. and also what would I use as an electronic valve? a solenoid or something?

 

[Cliff_J]

Remember the electronic valves in the link earlier were moving 2mm which is about 10x too short and the valve weight is substantially lower as well. You will be fighting a bad economies of scale problem on the wrong side of exponential problems.

With a B&S you do have a very inexpensive and plentiful platform to work with. But if you only wanted to double the HP, you could easily fit it with an OHV setup and seal off the flathead portion of the chamber, increase the CR, and fit it with adequate carberation for that increased power level. Basically, leapfrog it from 1940 to 1990.

 

[Ayrity]

would an overhead valve design really double the power output? where would i find that sort of a retro fit head? this could be useful info or a useful thing to do no matter what else i decide to do to the engine, thanks! and i also assume by increased compression ratio you mean a supercharger or something of the sort. on a side note, do you guys think that idea for using a smog pump is worth looking into as a mini supercharger?

 

[NateTG]

i do have access to a lathe here at my mechE lab, although i don't see how you can make lobes with a lathe... they arent circular all the way around. and also what would I use as an electronic valve? a solenoid or something?

I'm not sure it's an ideal cam profile, but one way is to use a 4 jawed chuck, (or some other excentric chuck) and split the lobes of the cam into circular sections, and then a similar set up on the cylindrical grinder.

If it's available, and you're familiar with it, CNC milling is also a reasonable option for creating a camshaft.

The EVIC guy was using solenoid controlled poppet valves. One of his friends was apparently using solenoid controlled rotary valves. Cliff's comment that the characteristics of the valves are non-ideal is certainly appropos, but if you're doing some sort of MechE experiment, that's less of an issue.

Similarly, you could bootstrap (or even run) with an external source of compressed air using regulators and whatnot to see what kind of flow you're going to want.

 

[Ayrity]

ya i think that's a good idea, the use of the compressed air to test with. I don't know why everyone is always saying that we don't have the ability to make electric valves work. the way i see it, i have 3 options: 1) use solenoid controlled poppet valves 2) solenoid controlled slide valves just attached in line to the port holes. 3) use an electric motor with a big thread or ramp wrapped around it and use that like a cam lobe, have the neck of the valve ride on it and as the electric motor spins (at a selected speed) it would make the valve open and close at different speeds.

 

[Ayrity]

ok so i can't find ANY sort of solenoid poppet valve... hmm maybe that will limit my choices

 

[NateTG]

ok so i can't find ANY sort of solenoid poppet valve... hmm maybe that will limit my choices

The EVIC guy rolled his own.

Hmm, how many cylinders does your motor have?

 

[Ayrity]

just the one. I've been doing searches and i really need to figure out the psi and temperature that this engine will be putting out. anyone have an easy way of figuring that out? it is a 127cc engine and I am planning to have a little bit of blowing from that supercharger so i figure the pressure can't be more than 700psi at its greatest. am I way off here? thanks for all the help

 

[sid_galt]

Isn't a solenoid valve implementation going to be tough?
I mean it brings along tons of issues with it like the high power required and valve seating and things like that. Besides it will likely end up being costlier than normal valve procedures.

 

[Ayrity]

well to allow for an easier valve timing this is an option to go in, and it would also take the stress off of the engine somewhat. I am open to ideas though!

 

[sid_galt]

You mean you want variable valve timing in your engine?

 

[Ayrity]

well, yes, because with such a different engine set up, I am pretty sure I am going to have to mess with the valve timing a bit. and plus with an electronic valve set up, I can use a pretty simple circuit to do a derivation and give me the optimum timing and have it do it on its own, granted i give it the right sensors which i would need to adjust it myself anyway.

 

[NateTG]

Well, solenoids are an obvious choice (as used by the EVIC guy). I haven't seen the engine, but it should be feasible to place a solenoid so that it replaces the cam lift on the rocker arm. Leaving the return spring in place can work, or you could use a double-acting solenoid.

Another option is to hook servo (or stepper) motors to a cam of some kind.
Conventional cams should work, but the "skew disk circular cam" at http://www.ducati.com/bikes/techcafe.jhtml?artID=5&detail=article&part=technical also looks interesting.

Edit: Having missed your earlier post -- if it's a single cylinder engine with a cam shaft (rather than push rods) consider taking the timing belt/chain off of the crank shaft and hooking it to a servo motor.

 

[NateTG]

just the one. I've been doing searches and i really need to figure out the psi and temperature that this engine will be putting out. anyone have an easy way of figuring that out? it is a 127cc engine and I am planning to have a little bit of blowing from that supercharger so i figure the pressure can't be more than 700psi at its greatest. am I way off here? thanks for all the help

Heat is going to be an issue for you when you're running 2-stroke.

If you know the typical operating temperature for the thing, you could use something like a tempilstick, or an infrared thermometer to monitor temperature during test runs. If you don't know the operating temperature, you can get a pretty good idea of what it's going to be by running the motor for a while.

Regarding figuring out pressure:
You can use a compression tester to figure out what you're running at currently. From there the natural gas law will give you a decent prediction on what sort of pressure you'll be getting from supercharging.

Noteably: You may be seeing lower rather than higher pressures when you're running as a 2-stroke - especially at high RPM - because you've got to scavenge and charge the cylinder in a single stroke, you're likely to see lower compression. Of course, this does depend on how much supercharging you're getting.

 

[Cliff_J]

Ok, an OHV from a small Honda or something like it has been done before, a roommate in college had done so at his high school when they competed in the high mileage contest.

There is a simple reason you don't see solenoid activated valves, and its not because everyone is some Luddite who thinks cams are cool because they're old school.

You need to accelerate a mass, let's say 100g, a certain distance and in a certain time. Let's say we have 6000 RPM, that's 100 revs per sec. For a 4-stroke engine, that's 50 times per second or 20msec period. Now it needs to open and close, so just the opening needs to happen in 10msec. We don't want a trianglular wave vavle action though, we want a sharp ramp rate and lots of vavle hang-time to get good airflow, a regular camshaft is similar to a sine wave but with the electronic valve we could get a near square wave right? Sure, if the motor operated at 60RPM.

A typical solenoid has the property of a varying magnetic field strength where it will be strongest in the activated position and weakest when first activated, which could be of benefit since the valve spring will increase its tension as its compressed, but our ramp rate is now compromised.

Please, do the math and figure out what force is needed to accelerate 100g a distance of 12mm in 10msec or less. Now the spring will need the same force, so double it to get the total force. Then locate a solenoid that can pull that force and complete its cycle in 10msec or less.

 

[sid_galt]

In that case, these thesis about electronic valve actuation may help you.

innovexpo.itee.uq.edu.au/2001/projects/s366921/thesis.PDF
legend.me.uiuc.edu/~astubbs/acc_2001.pdf

a few papers
www.seas.ucla.edu/~ttsao/Publications/ACC02_ChunTai.pdf
http://ghost.engin.umich.edu/SAE2002_01_1106.pdf
http://www.math.rutgers.edu/~sontag/FTP_DIR/03cdc-papers-refs-eds/WeA08.1.pdf

 

[brewnog]

Regarding Cliff's post, you should also consider reliability. As soon as anything goes wrong with your solenoid or controlling system, or anything in between, there's a good risk of killing your engine when the valves hit the piston.

 

[Ayrity]

i have class soon, and will respond in full more later, but just to touch on your post brewnog, luckily the B+S engine I have is pretty low compression, and if i mod the engine to be OHV i can choose the height of my head and valves over the piston, and with a big enough blower, that won't be a problem.

 

[Metro]

Hola, Ayrity

Joined PF so I could get in on your discussion regarding 2-cycles and injection.

I own an ancient 1973 Yamaha scooter called a U7e. It looks like a moped <em>sans</em> pedals. It's a single-cylinder, 80cc, and for reasons I can't figure out, I'm quite fond of it. Maybe "character" is of more value than little traits like "reliability"?

The pollution issue is really getting up my nose, as it were. There are tweaks--using different oils, keeping the machine in tune, etc. But the best available solution seems to be adding fuel injection, ideally direct injection. According to the miles of what I've read out here on the weerd wibe wob, properly tuned injection can bring a 2-cycle down to California ULEV status (probably not THIS one, but any improvement is an improvement).

Since it seems no-one's doing commercial conversion, I am left with experimentation on my own. It'll be slow going, but I intend to try and create a bolt-on system similar to the U of Colorado mini-taxi conversion gadget pictured here:http://www.envirofit.org/technology/kit.php

I don't think it's actually as complex as it looks. I hope it ain't.

Anyway, since you seemed to have questions along the same line, I thought I'd quit lurking and chime in on the chorus.

Recent thought: Your bike engine might be small enough to make use of scavenged parts from a scooter salvage yard. Many late models from Honda, Aprilla, et al, come with injection systems.

Re exhaust valves: howabout a simple valve held closed with a spring and opened by chamber pressure ... Or an adaptation of the leaf valves many 2-strokes use for intake control?

 

[fredric21]

Your idea for a valve in the exhaust in the form of a fenestrated disc is a poor one I'm afraid.
You would have major problems with both sealing, friction and lubrication. You would also have to deal with uneven heat distribution within the disc itself, not to mention the problem of carbon build-up (always a problem in conventional two-strokes).
Ad to this the increased weight of your drive mechanism, which would require some form of dwell period and things don't look too good.
Why not have a look at the Yamaha power valve system for ideas, instead!

 

[gear head 460]

i've been reading on this string. why are you worried about the waist of gasses out the exhaust. if the gasses doesn't suspend fuel. direct injection! much like a 2 stroke detroit diesel. supercharged and with direct injection, there is no waist of fuel on the return trip of the piston. the spray of fuel starts as soon as the piston rings pass by the exhaust port. and the inrush of air is pushed in by forced induction system. supercharger, and that only, a supercharger can start the engine with cranking rotation. a turbo can't start the combustion process. unless you pre-spool the shaft before firing the combustion process. breathing the cylinder with air. this would make the engine the best a 2 stroke could be. the less parts the better. the more simple the parts the better. the fewer reciprocal parts the better. kiss, keep it simple stupid. there would be no wasted oil mix in the fuel. sealed crankcase, much like a fore stroke. this is the best of the best in 2 stroke ideas. and it can be multi-fuel operated. given a varied emission output. just a thought.

 

[fredric21]

Yes, direct fuel injection would indeed appear to be the answer. The problem is, however, that at high speeds, there is very little time available after exhaust closing to allow sufficient fuel to be injected before the point of ignition, (typically less than 2msec.) in a conventional, piston ported, two-stroke engine.
Earlier injection would only result in short-circuiting of the charge, hence fuel lost to the exhaust.

 

[gear head 460]

ok, then multiple high flow injectors in the head, should cover the fuel needs and a balanced air fuel mix across the piston, with a centrally located spark gap plug. may cover the needs.

 

[gear head 460]

i never saw a bore size. so this may be a problem. with size limiting factors for extra injectors. so how many can be mounted in a given space without detrimental head strength loss.

 

[fredric21]

It's the actual time available for injection that's the problem. I have discussed this with Siemens technical staff.

 

[HowlerMonkey]

The problem with 2 strokes is scavenging and overlap.

Direct injection does away with the issue of overlap allowing unburnt fuel/air being passed through without being combusted by only adding fuel after the ports are closed.

Direct injection also helps scavenging because it allows manufacturers to achieve much higher percentage of scavenging because they are no longer worried about passing fuel along with the air through the engine unburnt.

This allows for port or valve timing schemes that are more favorable for power production.

For gasoline engines, Goliath cars used direct injection as far back as 1952 to clean up their two-strokes.

OMC made outboards with direct injected two strokes in 1996 but you have to wonder why so many outboard makers have an ever increasing percentage of 4-stroke engines in their lineup every year.

 

[fredric21]

If it were only so simple, then we would have Saab two-strokes, Trabants, Warburgs, and NSU's dominating the Automotive market, but we don't, because direct injection is not the magic wand for the genre.
We don't have any two-stroke motorcycles being manufactured currently, not even for racing.

 

[gear head 460]

if its a time issue, wouldn't the addition of more injectors put the proper fuel mix into the cylinder. or is this because it wouldn't have time to spread the fuel about the area fast enough to mix properly. even using fast spraying high flow injectors. but it that's still the problem. so how about the use of exhaust valves like in the detroit diesels. like the v6 71 run in semi-tractors. but with less stroke. for faster running engines. and less compression for gas or run a higher compression for propane or other high no knock fuels. you would defeat the issue of time to have a good spread of fuel across the cyl. by using a valve train. but going back to having a rpm limit and issues with a 4 stroke. but power would still be in abundance of power. having a power strokes on every return trip to tdc. and with the forced induction system. you can still get a clean cyl. every time too. just calculate the time to clean out the cyl, and pressure flow factors of the forced induction system, then there you go. because a drtroit diesel has the fastest running diesel out there to date. and as you know diesels don't turn up that high in rpm's. but detroits do. over 3800 rpm's, more than a cummings or cat ever thought about.

 

[gear head 460]

more 4 stokes, is because of epa reg's that all, but put a valve train on the exhaust side and a forced induction system. this will clean the old burnt fuel, for the new air charge. and there you go. with the time to have direct injection.

 

[fredric21]

If only life was that simple for the two-stroke! The problem with poppet exhaust valves is two-fold. a) the tremendous amount of heat that needs to be dealt with at high bmep's, due to the extra power strokes, and b) inertia loading on the valve train, limiting rpm.

 

[gear head 460]

ok, did i say poppet valves. i said valves. just like those used in a 4 stroke engine. research the detroit engine. from and inline 2 to a v24 they used mushroom valves just as those in all cars & trucks and alike. using a cam, rockers, springs and push rods. only on the exhaust side, every thing else is normal two stroke, but a sealed crankcase. no carter influence involved on the induction system. being a forced induction. you can use any fuel type on it being a direct injection system. like i stated. look up the detroit diesel engine. then get back to me on this and say it won't work. because it has already worked before. http://www.expertdiesel.com/71series.htm this is just one lead to what I'm stating.