I've been building up another Ryobi 31cc motor over the last couple months. This one is an early long shaft, although not the earliest. The crankcase has a bolt on fan shroud, a single transfer port runner on each side, a single ring piston, and an exhaust port with the web in the middle of it. I've cobbled this one up from left over parts, so this motor never before ran with all these parts together.

I will not go through all the details of each individual change and effect for the mods I made as most of that has been covered previously. I will only highlight significant findings of the project.

Piston Ring:

The previous posts clearly outlined the need for good compression and pinpointed the stock single ring piston setup as being inadequate (the later two ring piston was deemed to be just fine and in no need of modification). I purchased a piston ring from Frank Bowman <[email protected]>. Interestingly enough the new piston ring showed no performance improvement from the stock ring, BUT......read on.

Using an average of three consecutive runs to wide open throttle. (Stock Ryobi 31cc in every respect except for 1/2" exhaust stack in stock muffler. APC 18x8W prop.)

Stock piston ring: 6640 rpm, ambient temp. 66 deg F
Bowman piston ring: 6630 rpm, amb. temp. 67 deg F

In both configurations the piston ring was not broken in to the cylinder. In each case they had run no more than 20 minutes prior to taking the data, and although I did not do leak down tests it was clear there was blow-by with both rings. So, the test is inconclusive.

A good tight piston ring seal is always good for optimum compression and performance, and reducing the ring end gap from around .070 inch in the stock configuration to .003 inch or so with an improved piston ring has to be good for performance. In order to do the test accurately one should have two identical motors fully broken in to test for the effect.

After installing the new ring I noticed some strange behavior in the operation of my motor. The performance seemed to significantly deteriorate during the 10-15 minutes of operation each time I ran it. It took me a while to find the problem but I discovered that although the piston ring seemed good a free in the ring groove there was one point in the circumference of the piston that was too tight. My guess is when the piston was cold everything seemed ok but during operation as the piston got hot it expanded just enough to reduce the groove/ring clearance and seize the ring. The Bowman ring is ever so slightly thicker than the stock ring. So the lesson is, when fitting a new ring make sure the ring groove is shiny aluminum color on all sides all the way around the piston before installing the ring, but also be sure not to machine off piston material and increase the piston to ring groove clearance.

After cleaning the ring groove the performance of the Bowman ring was much improved, although I had already made some other mods to the motor so a back to back test of the Bowman ring vs. the stock ring after I fixed the seizing problem was not possible.

After several more runs I noticed that the first data run of the motor (out of 3) produced better results (by 60 rpm) than the last run. There was still a heat effect. This may by normal but so as to be sure I did not still have a ring seizing problem I reduced the ring thickness by about .0003 inch just by using 600 grit wet/dry paper lubed with WD-40 placed on a sheet of glass. I don't have the measurement equipment to actually measure the ring side clearance but it's well under .002 as I couldn't even get a .002 feeler gage started in the clearance.

Port Timing:

In post #294 I outlined my experience of modifying the exhaust port of the later two runner per side transfer port cylinder. I made the same measurements for this early cylinder and found some interesting results. I checked the port timing and mean effective port area of the cylinder according to the methods set forth by Jennings. I also include a comparison of this older cylinder with the newer one that I worked up previously.

Stock, before modifications:

Single runner: Exh 150deg dur: Phys area, .27sqin: Mean eff area, .21 sqin :: Trans 124deg dur: Phys area, .29 sqin: Mean eff area, .26sqin

Two runner: Exh 128deg dur: Phys area, .25sqin: Mean eff area, .16 sqin :: Trans 106deg dur: Phys area, .24 sqin: Mean eff area, .19 sqin

It is most interesting to note that although the physical port areas of the two models are not too dissimilar, being 7% and 17% smaller for the exhaust and transfer ports respectively on the two runner cylinder, the mean effective areas are 24% and 27% smaller on the two runner cylinder. This obviously is caused by the much lower duration of both port timings on the newer two runner per side transfer port motor.

Time Area:

To take this study to conclusion I calculated the Time-Area parameters outlined in Gordon Jennings book. These calculations were for an engine speed of 7200 rpm for the single runner cylinder and 7500 for the two runner cylinder. To be perfectly consistent they should be the same speed but the difference in the outcome is small.

Single runner: Exh .000150 sec-sq cm/cubic cm: Trans .000154 sec-sq cm/cu cm

Two runner: Exh .000093 sec-sq cm/cu cm: Trans .000091 sec-sq/cu cm

The recommended range is for Exh .00014 - .00015: Trans .00008 - .00010 sec-sq cm/cu cm

What this tells us is that for the early single transfer port cylinder the exhaust port is just the right size and the transfer port(s) is bigger than necessary. For the later two runner per side transfer port cylinder the exhaust port is significantly too small (in mean effective area) and the transfer ports are just the right size. This explains why the performance improvement resulting from porting the exhaust on the later two runner per side cylinder was so dramatic.

Gordon Jennings cautions that these recommended port time-area calculations assume that a proper exhaust system (read expansion chamber) is installed to effectively scavenge the cylinder. An ideal expansion chamber design for this motor is about 3 feet long and not too practical although not impossible.

I actually did remove the center web, on the early single runner transfer port cylinder, just as an experiment and I found no performance improvement (using the minimally modified stock muffler).

I've done some testing of a DIY canister system for the Ryobi 31cc and have gotten some very positive results. I'll report them in another post.

Thanks to combatpilot for some parts to make up this motor.

diceco


Elnino, I'm interested in your exhaust system. Do you have any dimensions for it, ie. total length, pipe and cone diameters, tapers, etc. Any data empirical or anecdotal on the effect of the pipe on performance? Cool boat!