2 cycle engines (& dyno)

[RG500]

I have a couple of questions about 2 cycle engines:

1) Can someone further explain “time-area” to me in regard to the 2 cycle engine porting? I do grasp the basic concept of time-area values and I do have a formula to calculate the time-area values for a 2 cycle engine however there is still quite a bit of gray area involved that I would like to clear up before I get too far into this latest little project of mine.

2) My engine is a ’86 Suzuki RG500, a 4 cylinder 2 cycle square configuration as in it has 2 cylinders in the front and 2 cylinders in the rear. I would like to make up a small engine dynamometer for testing my changes to this engine however I would like to test and change only 1 cylinder instead of changing and testing all 4 cylinders at the same time. I do have 3 spare cylinders, 2 cylinder heads, a complete spare ignition system, several carburetors and I am in the process of obtaining a crankcase and a crankshaft at this time.
What I would like to do is to make a small engine dynamometer that I could make in my garage to check what changes I’ve made in the output of the engine by changing only 1 cylinder and without having to change all 4 cylinders then take the bike to the track for testing. I really don’t care about the real world horsepower numbers, only if a change is in a positive direction or not so I am looking for ideas on how to gauge that change with reasonable accuracy.

Any thoughts and ideas will be greatly appreciated.

Thanks in advance,
RG500

 

[Cliff_J]

I think porting is more of an artform, maybe even a 'black art' since the proficient practitioners are many times going to want to protect the secrets that put them in the winners circle. I find it so odd the dichotomy of some large corporation with countless engineering resources and high-tech CFD on one end versus some guy with a die grinder on the other end and they're both able to get somewhat similar results.

While looking around to see what you're asking about, I ran into this website and it seems to have quite a bit of information.
http://www.hpt-sport.com/direct.htm

I had a 250cc Tri-Z that had been raced before I bought it. It had some big odd expansion chamber and massive porting (as replacing 2 pistons in one year reinforced) where this thing had a powerband that came on like a bolt of lightning at 3-4k. It wasn't enough to be confused with a 500cc dirt bike but when it transitioned from weed-eater to howling demon it would leave modified Banshees in its wake.

I've thought about doing something similar to tinker with designs as purely a thought exercise when thinking of how cool it would be to be like John Britten.

My thoughts about building a dyno (on a beer budget) would be to find a water pump of modest size, maybe scrounging some old ranching operation or a trip to a farm supply place. Now mount the pump so the body can pivot around the centerline of the input shaft, with this pivot held rigidly to the engine/dnyo frame. Now since the body is free to rotate an arm could be attached to it to stop the rotation. With a spring between the pump arm and the frame the torque applied to the pump case would be directly reflected by the spring deflection (and different springs would provide different ranges). Mount a dial indicator to measure the deflection. Then to properly load the engine a simple valve on the water outlet would vary the load and it'd be simple to adjust with one hand while getting the throttle open with the other.

It'd only likely be precise enough to provide relative readings and also only at steady RPM measurements, unless maybe you could setup a video camera to monitor the dial indicator and RPM gauge to plot the points afterwards. Oh and to monitor RPM a couple white strips on the motor shaft and a $30 RC prop RPM gauge would likely be precise enough as well. Most expensive part would likely be a coupler between the engine and the pump, but maybe something like a rag joint from a steering shaft could hold up to the power and again would keep the cost super low instead of a lovejoy or other coupling that can handle radial and axial offsets.

ghetto-fabulous-tastic! for what could be under $200 if you have the tubing and welder already, add a couple safety shields and it would be easy to start testing.

 

[RG500]

Cliff J.

Thanks for the web link and for the water pump idea, I've passed this information along to my design guy (I am just an idiot machinist) for his approval, I'm thinking I'd have to gear the pump down to a reasonable speed, my engine will run 12,000 RPM all day long and I doubt a water pump would really like those sort of speeds too much, safety shields indeed :)

Thanks,
RG500

 

[Cliff_J]

Hmm, maybe a belt drive to achieve the gearing would allow for slight misalignment and provide a slip mechanism should said water pump protest. A chain seems a little scary at those speeds but would only add a few more percent loss and positive transfer (and its likely the final drive anyways...)

An impeller/centrifigal pump would likely hold up at speed much better than a roller or other pump design, I hadn't considered the high RPM, I was thinking of 4-cycle revs. :) I'm just uncertain how reliable they would be at transferring the torque from the impeller to the case and not just dissipate it into the water versus a postive displacement pump that would simply need to increase its pressure to push the water past the restriction.

 

[RG500]

The water pump idea is sort of growing on me, at first my design friend suggested a flywheel of around 400 pounds or so connected to my poor little engine, my concern about this was that it would certainly grind my engine into a frothy pulp before everything came to rest sould I have a mechanical failure at a high speed.
If a chain drive sounds a little scary I could always use a gear set to lower the pump speed and perhaps keep from launching a stray chain through the garage and into the living room, the misses would not be pleased. I also thought of placing the engine on a rotating mount but I guess it really wouldn't make a difference if I took my readings off the engine or the pump.
Still thinking.

Thanks again,
RG500

 

[Cliff_J]

Well my thoughts on a rotating flywheel revolve around a few negative aspects. First you'd only be able to measure rate of rpm increase/moment of inertia to compute power. While very applicable to how the power is used real world, you'd need some sort of fast data capture to pull it off in segments that could be meaningful, to know that before change X it took 12 seconds to go 1000-5000RPM and after change X it took 11.5 seconds would indicate that the average power increased. Instead you'd likely want to know by 500RPM or so increments and this could be a very short time window unless you have a heavy flywheel.

A one-way clutch would solve the problem of the flywheel driving the engine, and yes it would hurt a lot of stuff should a failure occur. Which is the second negative, safety. At 5000RPM and say a 18inch diameter it doesn't take math to figure out that even 100lbs is dangerous. It would need to be very well balanced and at some point the centrifical force would cause the outside weights to fly off (better be made of good steel with good welds) and should a bearing fail - Yikes! At least with a water pump and maybe 200psi the worst is a big bath as long as the shields could catch the pump shrapnel.

Third you have the fact that the energy in the flywheel is the square of the velocity, meaning you'd need to factor that into power calcs. Now that isn't super difficult math or anything but while NASA can cite Apollo 13 as a case of their guys getting some fast accurate calcs done quickly they are also unable to have the public forget they had a contractor mistake metric for english units and run a probe into the surface of mars. If you're making 25HP at 5000RPM but mistaken think you're at 40HP at 8000RPM and make a change that reads 25.3HP and 40.3HP that's seemingly a small change. But if instead its 25HP and 60HP that changes to 25.3HP and 63HP then its a more meaningful change.

Its not like the water pump is perfect either, it may be difficult to source one of the correct size to be linear enough to make worthwhile readings. A known good source of data like if you were find a CBR600RR in a salvage yard and use a cycle world dyno graph to calibrate your dyno to match its HP curve then you'd have a pretty nice setup. But this could be a budget buster and now you got to find a use for the CBR engine like hooking it up to the lawn mower or something... ;-)

BTW, have you checked into commercial options? Some of the dynos out there are actually cheaper then I thought they would be. Kinda kills the American DIY attitude when you can find a roughly similar cost item and can then move onto bigger and better things like a nice smokey burnout...

These guys have the biggest selection but there are plenty of them out there, you may even just want to read around their website for ideas.
http://www.land-and-sea.com/dyno.htm