Cylinder compression ratio in two stroke engines

I wrote this post a few weeks back, but its currently about 2 pages and only halfway done. Even i dont want to read it, so i'll try to condense the knowledge and answer most of the questions i've been getting lately about combustion dynamics and compression, without boring the heck out of everyone.

Despite the fact that it contains no valves, and could basically be replaced with a flat slab of aluminum, the cylinder head in a two stroke engine has had an incredible amount of research done on it. The head geometry controls of the major factors influencing efficiency (power) of the engine: combustion pressure, and combustion geometry. The burning of fuel and air in the cylinder is basically a very fast chemical reaction. The laws governing speed of a reaction, imply a faster and more thorough burning of the fuel + air in the head, with a) more air velocity and b) more pressure=temperature=decrease in volume.

Quite simply put, the compression ratio creates a higher pressure in the cylinder when the spark plug fires, causing the flame to 'explode' with a lot more force, pushing down on the piston. This creates higher torque because of more force pushing on piston, but also requires more force to push the piston back up and compress the fuel air mix.

We calculate CR by taking the total volume (displacement plus head) and dividing that by the head volume alone. For a 70cc kit with a 7 cc head (stock 50cc) + gaskets ( 1-2 cc) it works out like (70+8)/8= 9.75:1.

PSI can be an indicator of compression ratio, but requires much tricker math to actually figure out CR. Other factors like ring sealing, port dimensions, pipes, intakes, etc, effect the PSI reading.

The 'squish band' affects the 'v' or velocity part of the equation. Air moving quickly has more energy and explodes faster. Imagine having a bowl of vinegar and dumping in baking soda. If you stir the heck out of it, it will fizz up much more violently. That is what is happening as the piston approaches the squish band. The quick decrease in volume will blast all the gasses into the hemispherical chamber around the spark plug at the same time, the violence of moving air speeds the reaction. Jennings and Bell reccomend the 40% of area- 7 deg taper method of sizing a squish band.

Because of the 'faster' flame in a high compression setup, ignition timing will have to be retarded. The ignition triggers and begins igniting the fumes before top dead center, to account for the time it takes to burn the air. When the everything burns faster, the max intensity of the explosion is at the wrong time and pushes the piston down as it comes up (knocking or pinging). This is exacerbated by a pipe, (or supercharger) which increases the pressure in the cylinder.

Sometimes the compression is just too high, the fuel ignightes without any spark, this is also problematic. Going to a higher octane fuel (only necessary in this condiditon) will allow you to continue to run the higher CR and pipe, otherwise you have to drop CR or go to a different pipe.

The 2 stroke gets more complicated also, because there is a huge hole (exhaust port) in the side of the cylinder. So even though on paper the compression ratio for identical 70cc cylinders, with different exhaust timings, will be the same, the holes in the side, and efficiency of exhaust porting will make their actual combustion pressure drastically different.

Finally, what are you looking for with your bike? The mods that mostly effect the final combustion pressure are: More aggressive intake porting= higher compression, more aggressive exhaust porting = lower compression, more radical pipe= higher compression at high rpm's when pipe hits.

If you switch to a higher compression head, and think you are detonating (sounds like rocks rattling in a can... and isn't your crankshaft) try to retard your timing a bit, see if your power improves. Some kits require so many base gaskets to clear a stock head, or small HC, that by the time all that gets put on there they actually have crummy compression, or a lack of bottom end from having all the ports shifted up. When the compression comes back, a bike that was tuned for different setup, can get thrown off. The detonating sound will be nastiest when you are really cramming on the pipe hard. If the timing doesn't fix it, and you're jetted correctly, and running premium gas and synthetic oil, then try putting in more head gaskets to drop compression a bit.

The cylinders that i've already made a lot of heads for (DR hobbit, MK65, and Polini) the compression ratios are just about figured out, but there is a lot of improvement for specific setups. Knowing what to look for makes it a lot easier to fine tune compression with head gaskets, and play with squishin' dynamics.

Mid-winter update

As indicated by my previous post, i've been busy as heck. I'm trying to get a bunch of projects finished up over my 'quarter break' coming up at the end of the month here, because next quarter i'm headed back to MSOE more-fuller-time with 2 or 3 classes, and it looks like i'm headed back to my other day job too, so my moped time will be shut down for about 3 months.
I've got three major projects I'd like to get into 'production' state by then, so I can sit back and coast, but its put my personal bikes on hold. I figured, in the spirit of Terry Dean's helming of MotoMatic Blog, i'd cruise around the shop today with a camera and take some teaser photos of what i've been up to.

Vespa cylinders!




100SL2 yamaha project, for a customer, pretty rad bike





Ewwww... four stroke... my CL100 racebike for the Lake Erie Loop this spring. thats got a big 'in progress' sticker on it.




What's this all about?


Exiting things to come..

Busy

Holy cow, this blog is going down in flames. I'm sorry, i've been too busy to post, but there are a lot of fancy things in the works. Sometimes it seems like the amount of real mopeds v. internet mopeds is inversely proportional.

Ciao FAIL

On wednesday, i was bored, and still frustrated with not having fixed the leakage issue on the ciao, so Alex, Joel, Jimmy and I tore it apart again (I can get the whole engine out in about 7 minutes now, most of that is removing the pipe) And I was finally able to figure out what was going on. In addition to the Decomp valve still leaking quite a bit, It appeared that when I used the face mill on the cylinder, It cut unevenly where it caught on the edge of the broken off fin. I should have broken or ground that fin off to start with, but you live and learn, eh.

Back in the machine shop, I took the cylinder and lined it up in the lathe, clamping the cylinder skirt. I was in a bit of a hurry to finish it up, and forgot my rule about doing this sort of thing... you have to run a piston inside the skirt or it will break the skirt off. I just had to take a few .001's off, so i wasn't too concerned, but when i started up the mill, some boner had left the feed engaged and right away the tool ground into the face of the cylinder and did this:


What a buzzkill.

I also went ahead and took the head down a couple .001's, which was probably a bad idea because now the extra tall piston i made is hitting, so i'll have to machine that stupid thing down again.


Yeesh, look at that nasty decomp, no wonder it was leaking, carbon and gunk all over...


Oh well, thats mopeds, should have left well enough alone when I had the chance. It looks like the cylinder will be fine with the skirt broken off, but if it eats it, I'm on the lookout for a new stock vespa cylinder. Anybody got one they'd like to get rid of? I've got a parts grande lined up, but i'm pretty sure thats a 10mm pin bike, so i'd rather not back track and rebuild another whole engine.

The good news is, break-in appears to be going well. I was concerned about re-using the old rings, but i cleaned them up with a little sand paper, and re-honed, and it looks like they are sealing really well. Given the wierd compression issues, i was worried they hadn't sealed, but this pic says it all:


The transfers also seem to be pretty well matched, the burn pattern on the piston is a little bit dark, but i'm 99% sure thats because of the wicked air leak causing things to run a bit hot.


You can see on the piston the hot spot from the decomp valve leaking (also perhaps because of the wierd head shape around the decomp) and there is a big black mark where the head wasn't sealing that well. Hopefully cleaning the decomp again and fixing the sealing issue will take care of my shitty low-end and maybe even 'fix' the misfire situation.

Mysteries of the universe. of mopeds.

Recently I discovered the library.

Rediscovered, is more like it really, I used to spend a lot of time in Eau Claire's LE Phillips Memorial Library reading and perusing all sorts of interesting things, but since coming to MSOE and signing my life over to the forces of homework, lab reports, and other such nonsense, I haven't really had time to waste on fiction. My dynamometer research got me back in the MSOE library where I found the amazing and mysterious TJ section. Titles like 'The design and development of 2-stroke engines' have completely rocked my world and dominated the last few weeks with knowledge I had, until lately, just grasped at.

In the upcoming weeks, I'll be spending as much free time as i can spare reading and distilling all this crazy knowledge, for you, my lovely readers. I just had the bomb of knowledge dropped on me in a big way, not to mention a lot of orders for machine work, and some interest in some of my other lesser-discussed projects, so it might take me awhile to dig out, but trust me, it will be well worth it.

Also, in the world of the ciao... massive head leakage, i think it has a lot to do with the decomp valve which i suspected was leaky-weeky, top speed so far has been a shy over 42, but low end is miserable, and compression is very low. I'll have this written up also as i get a 'round tuit'

Moped tuner/machinist for hire.

After a few recent conversations, It has occurred to me that I'm not the best self-promoter or businessman. Its apparently unclear here and other places I post my knowledge, that I do moped work for-hire. I spent the summer doing repair work trying to pay the bills, and since I'm still under-employed, I've got a lot of time to hang out in the workshop and make things faster.

My standard shop rate is $30/hr, for just about anything. Fabrication, welding, custom parts, adapters, manifolds, exhausts, if you've read the blog you know what I can do.

In addition to that, some services I commonly perform have flat-rate pricing:

Cylinder head milling: $25/ on-center spark plugs, $35/ off center spark plugs, $30/ Garelli and Minarelli heads with recessed sealing faces. All heads are custom machined for a specific bore and compression ratio, with 2-angle squish band.

From Machined Heads

Stock-cylinder porting: $50, you specify performance characteristics, I compute ideal port map. Jobs requiring additional machining extra.

Dellorto SHA or Bing carbureator boring, up to 7/8" (15.8mm): $20, plus any time spent cleaning/disassembling carb. This is performed to a mirror finish with a boring reamer, I will also remove and modify dellorto emulsion tubes.

Case-matching: $40 + gaskets. Shown is metrakit E50 with additional transfer porting.

From Metrakit Nate B

My gmail address works well for contacting me. I turn stuff around pretty quick, usually a week or less, but it depends on how busy I am. Email and discuss things with me, everything is custom, so Its my priority that its done just right for your build.

Super Stock Ciao

The lowly Vespa Ciao is one of the most elegantly designed motor vehicles ever produced. Truly the 'volkswagen beetle' of mopeds, its simple, rugged, compact, and designed for cheap mass production. Its no wonder the tuning scene in Europe has gone completely bonkers for these bikes. The parts, knowledge, and resources available for these dwarfs even the mighty puch. Being knocked-off in India doesn't hurt either, making this the most produced, cheapest, and simplest moped powerplant ever.

Ever since I picked up this low-mile rigid Ciao, I knew it was a vehicle with massive potential. TreatsHQ declared 2008 the year of the Ciao with a flood of vespa parts, and I was itching, but other projects took center stage, and it got shipped all over Wisconsin as my beat-around loaner bike. Two years came and went until about a month ago when I decided to set it up for my girlfriend, because what's better than pretty girls on Ciaos? Pretty girls on ridiculously fast Ciaos! (especially when they are stock and embarassing kitted bikes)

The Ciao's massive potential lies in its unique induction system. Rotary-valve induction is different from piston-port and reed-valve induction, in that the opening and closing of the intake valve is controlled by a mechanical valve, usually a disc on the end of the crankshaft, that opens and closes with the rotation of the engine. The rotary valve offers the low flow-resistance of a piston port, with the asymmetrical timing of the reed valve, for a 'best of both worlds' situation. The main drawback to rotary valve induction is the extra cost of the mechanical hardware which forms a tight seal, and extra complexity. Vespa engineers, however, solved this problem by incorporating a recessed area into the crankshaft web, and placing the port on the back of the crankcase where it would be covered and uncovered by the rotating crank web. Genius!

Seeing as this is the only moped made with this style of induction (aside from a few rare Kriedler models) I wanted to see what could be done in the way of unleashing the potential of the rotary valve. Rather than piston ports which just offer two dimensions of tuning (duration and area- time/area) and reed valves which are pretty boring (bigger=faster, yawn!) the rotary valve allows for tuning of duration, advance, area, and in the specific case of the vespa there is a lot of room for improvement. I started the tuning process by researching and studying the rotary valve system itself, and based the entire performance build around the idea that tuning for certain intake characteristics could yield high torque across a wide powerband, with maximum power in the mid-upper mid range. Two years gave me a lot of time to study and think about the improvements I wanted to make, and my research had already yielded a pretty conclusive game plan before I removed the first screw.

Valve Modifications:
Starting with the rotary valve, the geometries of the parts themselves can be improved for flow, specific opening and closing characterisitcs, etc. Looking at the assembly in 3-d, the passage of air from the carb inlet to the cylinder is hampered at any number of places, most likely due to manufacturing considerations. Right away the inlet 'manifold' nipple that the carb clamps onto is much too small to flow enough air. The rotary valve is very efficient, and capable of extremely high intake velocities, so the carbureation can remain relatively conservative, but there is no reason not to maximize the size of the intake tract. The largest I could bore the front half of the Dellorto SHA 12:10 carb out to was close to 14 mm. To maintain the integrity of the nipple inlet, i went to a more conservative 13-something for the nipple, and to match, the back-half of the carb. This was reamed far enough in, that the rest of the porting in the intake port could be smoothed for minimal flow disturbance.

The 'pencil grinder' which is a tiny little pneumatic 1/8" shaft grinder I bought at harbor freight, is invaluable for this stuff. I could actually stick the whole grinder right in the intake hole for the finish work. I left things a little rough because I really dont care that much about making it look good. A slight bit of surface roughness can actually improve flow characteristics if you are dealing with a convoluted shape, but to be honest, I'm just too lazy for all that tiny grinder bit junk. As you can see, I've opened the port up just about as big as i can safely go. The edges are round, because anytime you flow through a square tube, you basically have no flow in the corners anyhow, and this will make the transition easier and prevent any weird eddy-current crap. The primary concern is not breaking the seal made by the crankshaft web on the ground smooth surface around the port. Opening it up 'all the way' like this is only going to change your intake timing by a matter of a couple degrees and this engine is already absurdly conservative in porting.



Because the crankshaft serves as the rotary valve, opening and closing this port, the modifications made to it, directly translate into the modifications made to the port timing in the cycle. To calculate this, in degrees, just measure the diameter, dummy. I used a paper strip wrapped around the crank web, then i double-checked using the diameter, because the paper couldn't be tight or it would shortcut the recessed area of the crank web. Simple geometry will get you into radians of port timing, then degrees if you prefer. I'm not going to share all this, because I think this might be a competitive bike in stock cylinder racing, and I want a leg up on y'all, but if you really want to copy me you can probably figure it out from the picture. I also added some 'flow improvements' to the open area, sortof like knife-edging but in multiple dimensions.




Man grinding on that hardened steel is a real pain-in-the-butt. I still have slivers all up in my fingers. Brutal. The results are well worth it however, the modifications to the intake tract like this can easily out-flow a smallish 4-petal reed. Just think about how much flow area there is while this port is open! Compare that to four little slits 1/8 by 1/2 inches. Not to mention the fact that no energy of the incoming air is being lost opening the valve. We're talking about serious flow here, from a 13mm carb. Exciting!

To finish everything up, I ground out the transfer ports to match the modifications to the cylinder (which we'll talk about later) and to ensure smooth flow between the back of the crank-case and the transfer ports. Its hard to imagine how this all works in three-dimentional space, but I think I was able to clean everything up for a small improvement. I ported the case about as big as I felt comfortable, so if this engine ever does end up with a kit, it wont have to be rebuilt.

Disassembling and reassembling these cases is the nicest piece of cake in the entire moped tuning world. We're talking easier-than-legos easy. Heated bearings, dropped them on, heated cases, plopped it all together with a nice brown-paper-bag gasket and a few dabs of sealant. I splurged for the nicer bearings and seals from my friends at Allied Bearing(still cheaper than the moped shops) because I have a feeling this engine will have the ever-loving piss revved out of it, and I want it to be reliable for many years of blasting.

Cylinder Porting

To say the stock porting is restrictive is an understatement. I took a port tracing to start with, and was very confused by the results. The port map i traced showed the bottom of the exhaust port was 46 mm from the deck. A lot of very slow stock bikes (cough, Batavus, cough) have a ridiculously high deck height, making the piston come a few mm short of the deck at TDC. This drops the compression dramatically and gives you that nice speed restriction Iowa tuners and 'Scared Mommy' magazine are always raving about without having to make pesky tooling changes. I figured this is what was going on, and the wear pattern from the rings seemed to agree, coming well short of the top, so I proceeded to calculate my new port map assuming the bottom of the exhaust was BDC. Once I got everything together, however, I realized the real reason for the discrepancy was due to the piston coming up and covering 3 mm of the exhaust port at BDC.

This is huge

For 3 mm of the already tiny exhaust port to be covered at BDC is shocking to say the least. Correcting this massive error alone could account for an additional 5-10 mph on a stock bike. Its the same thing as raising a stock exhaust port 3 mm, not to mention the transfer ports, which are level with the exhaust on the bottom. To correct this error, I machined a whole 3.25mm off the top of the stock piston, along the edges, leaving more material in the top of the crown to effectively create a 'squish band' between the stock head and the modified piston.
This will drop the edge of the piston to the proper placement for the exhaust port timing, but will cause the piston to rest a full 4 mm below the deck height. I solved this problem with an iron fist. More accurately, the facing mill, which I have now nicknamed the 'iron fist' as it has proven itself to be an invaluable and dangerously aggressive tool for dealing with these sorts of things.
As you can see, the casting wasn't machined on center, but the milling was done true to the cylinder base face, so it should be accurate, not that it really matters all that much for this. Four millimeters is quite a bit, and what you see left of the top fin, is thin enough to slice tomatoes... which reminds me of an infomercial.

Once the cylinder and piston agreed with my theoretical port map, I did a little math and came up with a port map to highlight the strengths of this little machine. Since this is all for my girlfriend to have a nice, cute, reliable bike she can ride with the gang, I didn't want to make it too crazy, but it would have to keep up with the rest of the Cranks at 35-40 mph. The Ciao is so light, and geared so low, that torque isn't as critical as mid to top end power, plus the Tecno Circuit is a pretty torquey pipe. I also have a lot of faith in the stock single-speed dry clutch. Its simple and easy to get/make parts for, so I figure I can sacrifice some low end if I lighten/tighten up that clutch. The rotary valve was designed, along with the pipe, and the cylinder, to start pulling hard around 2-3k rpm, and hang on up to 8-9k. I went pretty huge with the exhaust and transfers, because I know that pipe will keep the low end manageable, and I wanted to have a little bit of power on the top end when the pipe is starting to poop out. I'll probably keep stock gearing for daily riding, but the plain ol' pulley makes it really easy for me to alter the 'gearing' with a short trip to the machine shop, so I might play with that a bit depending on what the GF likes.

The transfers on this engine leave a lot to be desired, and present a serious challenge for the tuner, because there just isn't any room around them to open them up conventionally. Everything has to stay symmetrical for proper scavenging, and that means getting into the two interfering studs next to the exhaust and the one transfer. To lay things out, the studs were drawn into the port map, and I started measuring. The exhaust port is off-center in relation to the transfers. For a moderate build, it could be widened 4 mm on one side (away from the stud, thank god) and probably make 90% of the performance gain my build is going to see. To push it a bit farther, I went even bigger with the exhaust, and bigger with the transfers, shifting them all away from the offending stud near the exhaust. The Transfers couldn't go out much, but they had a few mm to go in, by removing the absurdly thick cylinder skirt and making them 'finger' or fully open ports. I like to be precise about these things, so I took the cylinder into the machine shop and roughed out the ports. I also added a 3rd transfer 'boost' port in between the two gudgeon pins opposite the exhaust. This is directly adjacent to the third stud, so it couldn't be very deep away from the piston, but the boost port doesn't need to be very large to have a huge impact directing and energizing the air out the exhaust port. This should really liven up the 5k-7k rpm range where the pipe is starting to come off its powerband, and give a lot better torque on the bottom before the pipe is hitting. This is the cylinder back from the machine shop, rough, but the important stuff is all there.

The exhaust nipple was also reamed out as large as I felt I could go before compromising the integrity of the metal. I'm kindof touchy about exhausts, because of bad experiences with my Sachs, and dont like to take chances. At this point I still didnt have the pipe so I couldn't match them, but it turned out to be very close.

Once things were roughed out in the machine shop, it was grinder time. A lot of it. Probably about 5 hours total behind the grinder cleaning up the transfers, booster, and cases to make sure everything matches.

Still pretty rough

Getting a lot better

Finished.. with a custom-made base gasket, felt rubber for maximum sealage.

Whew, thank god for photo montage, that would have taken all day

After finishing up the cylinder, and cleaning everything up really thoroughly, the head was bolted back on and i did a quick compression test. I didn't take any photos of the embarassment, but I think my face mill left things a bit too rough, because I had to use a modified puch aluminum gasket and a tiny bit of goop to hold it in place. Ugh, i hate gasket goop, oh well, things seal great now. Torquing those 3 7mm nuts is very touchy, if they are just the slightest bit off they are real prone to leaky nonsense. Who uses three studs? Whatever.

Frame and other such necessary junk.
As much as I hate to admit it, moped engines actually have to go into frames at some point to realize their true potential. I guess that part just doesn't interest me much. Once the engine was back together, and the electronics were re-installed, I slapped it back in the bike. The wiring was a mess from a previous owner, but I had to redo most of it anyways because full-rigid wasn't going to cut it for a daily driver. As much as I love the ciao's aesthetics, either springer front or rigid, they are both almost un-rideable on Milwaukee's war-torn roads. After doubling the speed, I didn't feel very good about putting my girlfriend on something even I had trouble controlling, so I started trying to find different forks. I had a pair of maxi, garelli, and hobbit forks laying around, all of which I tried. The Hobbit forks seemed to be the best fit for size, looks, etc, and they have the same threads as the top nut from the garelli forks which uses the same 'outside' bearing as the vespa forks. Bingo. The tube was the perfect length for the ciao, but not big enough to fit the post style handlebars. Blah blah, boring stuff. Wiring, handlebars, brackets for the headlight. Not done here, but you can see it coming together.

Running with the pipe on.

First start went perfectly. Fired on the first kick, jetting was correct for my basement. No funny business. I adjusted the timing a bit, initially I retarded it a bit from stock because of the porting, but after running it a bit, it seems like stock timing works best. The cold streak we've been having has prevented me from getting out and test-running it, but it pulls really hard on the stand... cant wait until I finish my brake dyno. More results and pictures of finished bike to come!

Update:

First ride on the bike was kinda lousy actually. It was on a very cold day at the beginning of January, and the jetting was way too lean, no power whatsoever, and super boggy. I brought it back in and adjusted the timing, which for some reason was way off, possibly didn't press the foot down enough when gapping. I made a quick hi-flow air filter to richen it up.

The finished bike came out pretty clean, I had to make some headlight brackets.


Here it is all done.



I'm really serious about clean engine bays, wire routing, cables, etc.



So, fast forward a couple weeks, the nice weather yesterday was just too much, I finally got out again on the bike and got a chance to wind her out a bit on lincoln memorial drive. The results are sortof dissapointing. Only 43 mph top speed on the flat, but it picks up to 46 on the down hill (a bit too fast for a rigid if you ask me). I think the compression is too low. I never did actually turn down the head, so i might still have to do that. Also the questionable rings I re-used might be weak. Also the warm weather threw my jetting off again, and now its too rich.

I'm really not happy with the low end either, before the pipe hits, it has no power whatsoever. I know I need to tune up the clutch quite a bit, but there should still be more torque than there is. Correct jetting, finer timing adjustments, and some serious clutch tuning will all be conducted as soon as i have some time. Stay posted for more awesome.