Very well done Frits, thinking rationally like that obviously comes naturally
Sadly my immediate reaction only accounts for the obvious proven 54 bore at a 70% chordal, where 2.5mm is commonplace - so yes 2.6mm is too big.
Also in my opinion anyone saying a 65% chordal is happy with 2mm ( even if it is a 40mm bore ) simply means its closer to a lawnmower than a race engine.
Some might opine that Im so narrow minded it could fit between the blocks at Giza - and no a credit card wont go, been there , tried that.
Frits! Are the winner bikes in liedolsheim lawn mowers
Fact is...bidalot has only main cordal 65...max 66...and 2mm bridges to the boosts...
Not the optimum as nearly everybody knows ! ��
But these zylinders are strongest in this competition...untouched Ducts... 21 -22 hp/ crank
Very fast lawnmowers.
Frits! Once you were involved in liedolsheim...supported jan schäffer with a simson...and he did second place inthe 8h race!!
Wob, you're underestimating me. On the job you learn words from your colleagues that you won't find in a dictionary. I think back fondly on Miguel Sauca, the Spanish draughtsman who came with Jan Thiel from Bultaco to Garelli. So now I can also swear perfectly in Italian and Spanish. And in Dutch of course, and a little in French, and even in Finnish.
Come to think of it, I cannot swear in Japanese. I wonder it that's a flaw in my upbringing or a testimony of Japanese civilization.
OK, back to the harsh reality.
I try to answer this type of questions with a universal answer that is independent of bore, stroke, etc.
In this case my answer would be that the width of a vertical bar between two ports must be at least 5° of center angle.
Example: a cylinder with a 54 mm bore has a circumference of Pi x 54 mm = 169,65 mm.
That circumference comprises a total center angle of 360°. One degree center ⁶angle has a cordal width of 1/360 x 169.65 mm = 0.47 mm. And a bar with a width of 5° center angle is therefore 5 x 0.47 = 2.36 mm wide.
Now would be the time to say that you can forget most of what I wrote above. I can simplify it to:
The minimum bar width must be 0.044 x the cylinder bore diameter.
Thanks Frits for the Formula!
Is the Formula independent from width of main ex or made for the known sturdy and effektiv 70 percent cordal width?
Think the narrower the main...the higher is life span of every involved part there.
Conclusion could be that it is possible to use narrower bridge bars, if main is under 70 percent?? Right or wrong??
Engines for eight-hour races are prepared for reliability rather than power. They do not nearly reach the piston speeds that Wobbly is use to and in this respect I can understand Wobbly's remark.
Fact is...bidalot has only main cordal 65...max 66...and 2mm bridges to the boosts. Not the optimum as nearly everybody knows !
A port width of 70% of the cylinder bore in combination with the optimum port shape will give the optimum blowdown angle.area. But making the main exhaust port narrower yields more place for auxiliary exhaust ports, in which case the 70% main port width is not sacred and keeping it narrower may give a better total blowdown angle.area.
Originally Posted by Wos
Frits! Once you were involved in Liedolsheim...supported Jan Schäffer with a Simson...and he won the 8h race! Can you give us a idea, which power the engine made?
Sorry Wolfgang, I don't even remember what year it was; I would have to look it up in the Langtuning dyno-computer but that computer is not online for security reasons and at the moment I am not even in the same country.
Thanks Frits for the Formula! Is the Formula independent from width of main ex or made for the known sturdy and effective 70 percent cordal width?
Conclusion could be that it is possible to use narrower bridge bars, if main is under 70 percent?
That formula is only dependent on the factors mentioned in it, as should be the case with any formula. Do you see any port width mentioned?
The bridges between the ports in a cylinder have more than one function. They must guide the flow but they must also keep the cylinder together. And they do not always succeed, as you can see below. So do not make them too narrow.
Viking - if you want more upper front side then that is achieved by increasing the length and angle of the last part of the diffuser ,leading up to the parallel section.
What is the geometry of your current rear cone - maybe I can suggest an alternative.
As I said your design has good superposition, but even with that the pressure ratio rise after EPO is not spectacular - many " good " designs I have are " off the chart " past 3X.
Wobbly,
I´m in the middle of a over one year long moving project, so my calculation and simulation computer is lost in some of the 94 moving boxes i have. Most likely its in another place 450km from my current location.
I have not done any EngMod2T simulation for over two years now, so my abstinence is profound.
Searched for some information on my 3D CAD computer, but the only expansion chamber models to this engine, that I found, had a rear cone of 13°/22°.
Very kind of you to offer some help with my simulations of this engine, but I think we can spend our
time on a more interesting engine project.
Attached you see pictures of a 4-cylinder engine for another of my snowmobile project.
Vikings like brutal force, and so do I.
The engine is based on dual Rotax 670 rotary valve inlet snowmobile engines, with my own designed cylinder head and flexible crankshaft coupling device.
I have not yet decide if the cylinder volume will be 1340cc, 1480cc or 1000cc (to be able to participate in dragracing competitions).
Attached also a simulation in EngMod2T of one cylinder of a 2-cyl 740cc (or 4-cyl 1480cc).
Do you see any room for improvements?
Yes, Vannik need to extend the Y-axis pressure ratios to at least past 4X.
Aha, so do I start calling you Ragnar, your Viking name from the TV series, I will keep an eye out for it on the EngMod Group.
As in the previous screenshot, the only real issue is not turning that huge EPO pressure ratio energy into a negative ,soon enough around BDC.
In the best sims I have the wave is negative before the 60 marker - this shows that the diffuser is not efficient enough, early in the cycle.
I assume this engine will be CVT, so you can also use a rear cone geometry I have been working with lately to seriously pump up power around peak.
For a CVT, a two angle 50/50 with the steepest first, for a roadrace engine the same but with 3 decreasing angles.
In the roadrace version the shallow last rear cone gives a huge boost in overev, not needed in your case.
But this geometry pumps up peak and or overev power, at the expense of front side power - no free lunch as usual.
So - I have only used it in cylinders with 3 ports that are all activated with a servo PV.
With those cheap arse Rotax diaphram things it is impossible to lift the valves slow enough over a wide enough span.
Using an Ignitech running this, you can lift the PV at any rate up to say only 800 rpm before peak, and have more power everywhere.
Are you using the so called " Wobbly Duct " - if not, PM me your email and I will send you the paper I wrote on this.
In this screenshot the return amplitude is restricted, enabling much more overev power.
Ive got a thing thats unique and new.To prove it I'll have the last laugh on you.Cause instead of one head I got two.And you know two heads are better than one.
Hi Koba. Rick Ford was running an air cooled TF/TS125 with 22rwhp and that proved really reliable with little heat fade. We could get much more hp out of our aircooled GP125's but the heat fade made the extra power self defeating. We tried all sorts of tricks, like ceramic coatings and extra copper fins. The copper went all the way in to form the squish band. But air cooling was just not up to handling the waste heat from the higher HP.
I've found the same thing with my MB100 based stuff.
The old aircooled nail worked great at 13.5:1, for 3 laps.
I think it ended up at about 11.8:1, running on any old fuel.
This one was built to keep going, be easy to ride and still use standard road internals.
EDIT: Yay, saved Frits from looking at his thumb for too long. My comments may not be better!
Last edited by koba; 13th August 2025 at 04:18.
Reason: Additional comment
Neels, fully agree. I was pointing out how the sim wave front could be improved, but should have addressed as well the consequences lying in wait for sub optimal transfer duct geometry.
I am not familiar with that Rotax engine - does it have lift shafts ?
Ive got a thing thats unique and new.To prove it I'll have the last laugh on you.Cause instead of one head I got two.And you know two heads are better than one.
Neels, whilst working late last night it occurred to me that assuming the Rotax depression wave acts along the same lines as the R3 I am testing, then we have a
situation where at 1000 rpm below peak power, the pipe - now too long for the tuned rpm, acts to create an even sooner, greater depression at the port.
Thus it makes me think that at BDC, there will be a large, unavoidable span of rpm where even with an inefficient diffuser, we will have excessive short circuiting, simply due to the pipe geometry and
the port acting in concert, to pull the suction early and low.
I dont have any " inefficient " diffuser designs to test in a big cylinder, to see the wave front shape created in that Rotax for example at lower rpm.
I would be really interested if Viking could post the screen shot, or send me the .pack to understand what happens in this scenario.
Here is the KZ at 1000 rpm below peak to show the effect in that engine.
I assume we would only see the down side of this effect on lift shafts, using your excessive short circuit, scavenging model.
Ive got a thing thats unique and new.To prove it I'll have the last laugh on you.Cause instead of one head I got two.And you know two heads are better than one.
Wayne, that is huge suction, and it comes from the pipe still resonating (non zero positive pressure at export opening to allow buildup of pressure and suction pulses) although the plugging pulse is too early but still strong. My thinking is that it will cause short circuiting with lift shaft ports.
The effect of short circuiting on pipe bulk temperature is quite severe, the following is some guidelines I am working on (The names in brackets are the particular scavenging model names):
When using the different scavenging models in EngMod2T the following pipe wall temperatures are a rough guide:
• Normal Scavenging, almost no short circuiting - ~350°C
• Small amount of short circuiting (Yam12) - ~300°C
• Medium amount of short circuiting (Small) - ~250°C
• Severe short circuiting (Severe Short Circuiting) - ~200°C
I do wish there was an elegant way to get EngMod2T to predict the amount of short circuiting and to calculate its effect on the bulk pipe temperature!
... we have a situation where at 1000 rpm below peak power, the pipe - now too long ?? for the tuned rpm, acts to create an even sooner, greater depression at the port.
Thus it makes me think that at BDC, there will be a large, unavoidable span of rpm where even with an inefficient diffuser, we will have excessive short circuiting, simply due to the pipe geometry....
Wob, I noticed that you tend to use a somewhat smaller header length percentage than what I am used to. The difference is maybe just 1 or 2 %, but it could have a noticeable effect.
Short-circuiting due to too short a pipe is somewhat self-regulating due to the much lower exhaust gas temperature resulting from the escaping cold mixture, as Neels points out.
I think this was a reason for provoking deliberate short-circuiting and/or high-speed loop-scavenging in yesterday's direct-drive kart engines.
Originally Posted by wobbly
Reply With Quote
..........
Bookmarks