ESE's works engine tuner

[koba]

But I have to agree that cooling the exhaust piston would require a brain wave.

It's not at all a brain wave but it would be much easier if using a wet sump on the exhaust sides, then the undersides if the pistons could be spayed with cooling oil; like in a lot of turbo engines.

Would one have to supercharge it then, or could one have a small 'starter charger' and completely forsake the crankcase pumping effect, relying entriely on the pipe action for running.
Feasible for a race engine?

This probably makes no sense at all.

Just thinking with keyboard in hand.

[Flettner]

My Kawasaki has a similar " chamber bleed " welded into it, accept mine is not a bleed, it's a pressure take off for the EFI system to help it make up it's mind on fueling. Seems to help because it fell off the other day at a trail ride. Straight away I could tell something had changed.

[dinamik2t]

Hey, Wob, could you give a little more info on how you do the ignition unit mapping with your ignition dyno?

[wobbly]

In the pic you see the rotor mounting.
On the other end of the shaft I fit my large ( 400mm ) alloy protractor, with a stiff pointer set to 0/360 degrees, on the timing marks.
The vacuum cleaner motor has a light dimmer rheostat in the field winding, and with that and a small digital rev counter triggered off the plug lead,I video record the unit spinning up
slowly thru the rpm range.
I used to use a VHS video camera years ago, but the player died, so I had to buy digital camera that records onto a CD.
With a strobe shining on the degree marks as it spins, I can play back the video back on a big screen in slomo, watch the digital rpm gauge, and note down the change in timing every 100 rpm or whatever.
You can pause the recording as well to see the occasional wavering result.
At the end I go back around to the rotor end and strobe the timing marks to find out at what rpm the timing marks are aligned, and double check that the protractor is seeing zero at that rpm.
From the rpm/degree chart I can then add the calculated static timing to the curve.
If for example you run 1.4 mm static on a 125 this would add say 15* to the zero reading on the protractor,at the timing mark rpm seen on the strobed rotor.
This gives the actual timing the engine sees at every noted rpm.
Its also then easy to change the static timing, and draw the resulting new curve.
I have a standard calculation spread sheet that I simply fill in the angles,the static timing / rpm, stroke and rod length , and Mr Gates draws the graph for me.
Can you have the spreadsheet - no.
Im dumb, so i got one of the dumber PhD students I supervised at Uni to do it for me.

[dinamik2t]

Thanks for the morning class Wob

To be honest I know almost nothing on this topic. I am looking around the web in order to find info, but it seems I 'm not using the correct keywords.
I was thinking on calculating the RPM from the crank position sensor (reluctance) via software (I have a daq device) and find a way of calculating the advance from the cdi ignition signal.
Static advance is the easiest part I think, as it can be calculated from the rotor-pickup directly. The RPM signal must be easy too if you convert reluctance to digital. Main problem to me is how to measure the advance over RPM -strobe can't be the only way.

The objective is to map different CDI versions from 2 bikes-rotor specs (static advance is stated in the manual). So, I thought of cutting thick iron disks corresponding to the rotor diameters/lobes, in order to avoid mounting issues and to use a less powerful motor. Like the ones sold from Ignitech. (this way I can also keep a 'record' of those engines' rotors.) Would you think of any potential problem with this?

There is a guy here doing something that looks like mapping via oscilloscope & software: http://cbr250.com/forum/thread-11008-page-7.html ; but the topic is a little messy and I am a little dumb.

Anyway, thanks again Wob


p.s. doesn't your camera have a card slot, like SD or something? I hate CDs...

[wobbly]

I bought the trick CD camera for a fortune when we still had VHS video cassettes, phones nowdays have better resolution.
As per the video all this is easy now with digital storage scopes and you can buy digital output rotary position sensors that have 360 pulses per rev.
With Matlab you can easily use the position sensor and a pulse from the coil wire, calculate the time delay from the rpm, and it will draw the graph for you, with an offset for the static setting.

[Forgi]

Hi Friends!

Wobbly! Do you have any logical explanation that in case of the simple exhaust port why the narrow exhaust nozzle did not work? If Engmod is right the 75% rule has to work, according to the simulations. What's more, I think that in case of a simple exhaust port a smaller exhaust nozzle can also work!

A simple trapezoid exhaust port with a chordal of 70-75% is almost the same like a not too wide T port, so from this point of view it also has to be work! I have just wanted to do some tests, if you are interested I'd share my results.

[TZ350]

A simple trapezoid exhaust port with a chordal of 70-75% is almost the same like a not too wide T port, so from this point of view it also has to be work! I have just wanted to do some tests, if you are interested I'd share my results.

I would be very interested in the results.

[wobbly]

The exhaust nozzle I believe when used with a big T port or a 3 port with plenty of blowdown area is doing two things.
Firstly it is reducing the volume of the duct, and secondly it is matching the flow capability ( or increasing velocity, depends on how you look at it ) of the duct to the actual
requirement of the power being made.
If you take the area of the main port and add the two eye ports, this dramatically increases the blowdown STA and thus the power potential.
But in reality this total area isnt needed for the power potential, its the blowdown that is important.
From doing hundreds of simulations and building more engines than I care to count, the best power is made by matching the header inlet area to the total Ex effective area, but reducing
the duct to around the 75% as mentioned.

But when you have a single Ex port, out at 72% or whatever, and lift it to the max at around 200+ duration, the power is ALWAYs limited by the blowdown STA available.
A simple trapezoid will never be able to match a T port for ultimate blowdown STA capability.
Thus having a limited power capability, we already have a matching small port and duct.
I have found in many cases that power will go up if the duct exit is made around 90% for things like a TZ350, that has a big port, but the bmep ,thus power, is fairly low.

As Frits has alluded to we should be matching the duct to the blowdown available, and this would give a much better approximation to the ideal duct exit needed.

[SwePatrick]

I sure do. I'm just wondering now: should I address you as Flettner, as Uniflow or simply as Neil? In any case, I will take this opportunity to compliment you on your projects. The variable rotary valve housing is clever, I like your EFI, and your uniflow engine is impressive.

You don't happen to have Italian roots, do you?

Fast / slow is the direction to go, but for really turbulent-free flow you would need 'very fast / stop', with the channel through the ball in line with the stationary parts of the exhaust duct. But then the acceleration forces would become huge.
Maybe you could take another look at Alex Hofmanns four-stroke head; for each 'valve' he uses a hollow sphere that intersects the stationary duct; the part of the duct inside the sphere is also stationary; good for flow and for keeping heat away from the sphere.
Still, for uniflow I would prefer piston ports for both transfer and exhaust. But I have to agree that cooling the exhaust piston would require a brain wave.

Look at this video,, i think this is the way so solve some issues
http://www.youtube.com/watch?v=Bch5B...ature=youtu.be

[TZ350]

Deto in the Over Rev Zone.



The mid chamber bleed.



Tried it open and closed, Red open, Blue closed. The big surprise was how quiet it was when open, hardly any more noise at all, certainly wouldn't need any sort of mufler.

Actually I tried it lots to see if I could join the two curves and if it would kill the high rpm deto. But it was difficult to coordinate by hand and the runs were all over the place. Although the results were good enough to make me very interested in developing a proper ATAC Valve.



Kel gave me a hand to try and figure out where this top end deto is coming from.

I tried to take pictures when the Red light flashed, but all I got was a lot of unfocused blurry photos.

Although I could see the Red light flashing, in all of the shots the green light obscured the Red deto one.



At the end of the runs the engine went into auto ignition and ran on, in the clip you can just see Kel pull in the clutch and thumb the kill button but the engine still ran on for a bit.



Kel has suggested that the dings in the copper squish area left from an earlier blowup may be the problem.

Thanks to Kels help I now know that the deto is likely a part of the auto ignition problem. Changing the timing and/or jetting did not help.

[koba]

Tried it open and closed, Red open, Blue closed. The big surprise was how quiet it was when open, hardly any more noise at all, certainly wouldn't need any sort of mufler.

Interesting curves.

Maybe there is more to it but 'linking the curves' looks like a potential direction for progression.

[koba]

Kel has suggested that the dings in the copper squish area left from an earlier blowup may be the problem.

I'd certainly suspect that.


The video on the 'free valve system' was interesting.

I'd wondered why such a technology wasn't already in production, perhaps it isn't so far away.

[Bert]

Deto in the Over Rev Zone.



The mid chamber bleed.



Kel has suggested that the dings in the copper squish area left from an earlier blowup may be the problem.

[/URL]

Really interesting development Rob (should be easy to control the bleed with the ignitech).

I've been following this Deto thread with interest and often wandered if the copper squish area may be an area for concern.
two things that come to mind:
1. with likely expansion, is the squish (being copper sheet) expanding into the main combustion chamber; thus creating potential heat pockets (or sharp edges) which start the combustion process early (with the burning gasses not escaping cleanly).
2. is the copper sheet being lifted under compression (the photo eludes to it being out of shape); is there a possible limited life to the use of a thin copper sheet before it anerls (sp) itself to a point that it deforms under high temperature/pressure? again generating areas of heat pockets...
the reason for this thought is the only deto rings I've ever seen personally, (and other Fits & Wobbly will; I'm sure put me right) are quite solid (3-5mm thick); and machined into either the head or the top of the barrel.

two cents, probably only worth one.

[Frits Overmars]

Look at this video, i think this is the way so solve some issues http://www.youtube.com/watch?v=Bch5B...ature=youtu.be

A very interesting video, Patrick; I enjoyed watching it. No bla bla sales talk, just facts. But I doubt that these oil pressure actuated valves will be useful for us.
I have seen similar valve actuation in a Lotus prototype engine and there I noticed the same points of attention that Christian von Koenigsegg mentions in the video.

First, the 2 ms delay in opening the valve. For an engine revving at 10,000 rpm that means a delay of 120 crank degrees... And no matter how fast the valve can be accelerated, it has to start moving from standstill; its initial velocity is zero. In my book that is a very slow rate of port opening, compared to a piston that is already at its maximum velocity when it starts opening an exhaust port.

Then there is the available port area. You can cram 2, 3, 4 or even more valves in a cylinder head, but the more valves you fit, the more they are in each other's way, especially when they all open at the same time.
If you leave valve acceleration forces aside for as moment and just concentrate on maximum flow area, one central exhaust valve in the head is optimal (its diameter should be cylinder bore / SQR(2) ). But even this optimal area, combined with the slow rate of opening, cannot match the angle.area of piston-controlled exhaust ports.

Then there is the rev limit. In the video a maximum of 20,000 rpm is mentioned, no doubt for the tiny valves of a high-revving four-valve four-stroke motorcycle engine. That means each of these small valves can open and close up to 10,000 times per minute.
But we are not dealing with four-strokes (at least I am not). In a two-stroke each valve has to open during each revolution. That means a maximum rpm of only 10,000 with these small valves. And if we go for the maximum flow area with a large central valve, that 10,000 rpm maximum will shrink to 8,000 (more mass and more lift).

Next: cooling. In a four-stroke the exhaust valves sit on their seats three-quarters of the time, dissipating heat. No such luxury in a two-stroke....

Summary: not enough angle.area, not enough revs and not enough cooling. If you really want to use valves, poppet valves won't do; look at rotating valves.
Or look a bit further and use piston ports .