ESE's works engine tuner

[aljaxon]

Yes, hopefully. Otherwise the waves couldn't do much good, could they?

You've got that right, Al.

In the old days thinking was also that the earth was flat. Every change in cross flow area causes a reflection, that is to say: over the entire length of the cone. The fact that in those old days people assumed the wave to start from the center of the cone, or, even more daft, from the middle of the total length of the imaginary end cone extended to zero diameter, was just an aid to form a mental picture (pun intended).

the earth is flat.
especially where i live. in fact its concave. and i assume you live/lived just across the north sea from me in the land of spetters? (very memorable (for more than one reason) dutch biking film) cant get much flatter than that.

i thought having the strongest stuffing wave returning just before ex port closure was best. sort of like halfway between transfer closure and ex port closure.
but the waves that occur at the start of the rear cone return far too early or at the desired time but at a much higher than peak rpm. those waves arent wanted if you dont want over rev. so im thinking why not have a much longer belly? purely in the case of my situation wanting little overev and more front side power so as to be able to cruise without tap dancing on the gear lever.

[skako]

Skako, the sim is calculating the pipe bulk temp in real time as the rpm points progress.
It cannot suddenly change the accumulated results at one specific rpm , unless you only run that one rpm individually and overlay the two results.
Thus what was happening directly before 9000 is affecting that result - just as would happen in reality.
Neels may have a better logic than mine here.

Two things to note - its absolutely impossible to run 95 Unleaded at 13:1 , I use 11.75 A/F ratio and that's pushing it hard.
Even 110 Octane would struggle to not detonate at 13:1 when making serious power.

Also you have not run the sim in Turbulent mode , then created a new Combustion Model from that run , with the real Delay/Duration and Vbe numbers that are affected by Compression/MSV/ timing and A/F etc
specific to your setup.
You enter the new saved Combustion file parameters , and then continue in Prescribed - its also way faster.
Just running the generic inputs can lead you well down a blind rabbit hole of non reality - again.

EDIT - running 22* of advance at peak power the thing will absolutely destroy itself , so something else is seriously wrong for the screen to not be screaming DETONATION at you during a sim run.

ok wobbly,
now I see, the simulation should go in several steps and not as I did in just one step.
I am new to using this simulator, still learning how to use it properly.
Thanks wobbly for pointing out my mistake.

[wobbly]

Wos , absolutely correct about the hysteresis in the temp system.
That sort of answers the next point for Sako - most tuners use inertial or Eddy dynos with constant load for testing of 2T engines specifically as it emulates
the very dynamics we see on track.
That is , more power = more acceleration.

We never , ever see 9000 rpm for more than a split second after getting off the gas and down changing into a corner.
The pipe temp has dropped in that period and then what happens on the dyno , happens on the track - we give it full throttle.
Thus a good sim on petrol will reflect this with a wall temp of near 325*C at the bottom of the powerband , then if the dyno run time is correct , we will see the wall temp the
pipe sees on the track at full noise - around 425.

This is not the egt as seen in the header , that is affected by the A/F ratio used and the combustion parameters ( Compression/MSV/Timing ) and this influences the pipe bulk gas temp
that ultimately affects the wave speed just as much as the pipe wall temp does.
This pipe wall temp on petrol has been proven very accurate in many sims from 50cc to 300cc - but only if a realistic A/F ratio is used for the fuel.

I can see no useable benefit at all to doing step and hold testing of a 2T - doing correctly timed dyno runs simulates reality - just as we are trying to do with a sim.

And Al , you have a serious Catch 22 with belly length - make it longer and that generates a steeper rear cone , sure that limits overev , but even more so it affects negatively the front side power.
Once again my favorite saying , sorry - no free lunch.
And as I have already said , that R1 pipe tells you nothing apart from the 3 cone rear setup, as its designed to generate max power spread from 10,000 to 14800 from an engine with no PV , no PJ
and is limited by a flat line ignition.

[aljaxon]

Wos , absolutely correct about the hysteresis in the temp system.
That sort of answers the next point for Sako - most tuners use inertial or Eddy dynos with constant load for testing of 2T engines specifically as it emulates
the very dynamics we see on track.
That is , more power = more acceleration.

We never , ever see 9000 rpm for more than a split second after getting off the gas and down changing into a corner.
The pipe temp has dropped in that period and then what happens on the dyno , happens on the track - we give it full throttle.
Thus a good sim on petrol will reflect this with a wall temp of near 325*C at the bottom of the powerband , then if the dyno run time is correct , we will see the wall temp the
pipe sees on the track at full noise - around 425.

This is not the egt as seen in the header , that is affected by the A/F ratio used and the combustion parameters ( Compression/MSV/Timing ) and this influences the pipe bulk gas temp
that ultimately affects the wave speed just as much as the pipe wall temp does.
This pipe wall temp on petrol has been proven very accurate in many sims from 50cc to 300cc - but only if a realistic A/F ratio is used for the fuel.

I can see no useable benefit at all to doing step and hold testing of a 2T - doing correctly timed dyno runs simulates reality - just as we are trying to do with a sim.

And Al , you have a serious Catch 22 with belly length - make it longer and that generates a steeper rear cone , sure that limits overev , but even more so it affects negatively the front side power.
Once again my favorite saying , sorry - no free lunch.
And as I have already said , that R1 pipe tells you nothing apart from the 3 cone rear setup, as its designed to generate max power spread from 10,000 to 14800 from an engine with no PV , no PJ
and is limited by a flat line ignition.

cheers wobbly
i started thinking i can fiddle with multi different angled rear cone sections to get the one that matters shallower. like an initial sharp angled cone followed by a shallower cone. but then i suppose the initial sharper cone would rob too much of the available energy?
only way i can think to do anything is is to have a longer belly but with a narrow diameter to reduce that the rear cone angle. so all my diffuser angles need reducing. also maybe a slowly tapering belly?

im getting a bit too trigger happy now cos i need to get the sections ordered from the laser cutters in time for me to play over christmas.

[wobbly]

Rear cones with a steep initial angle that then flattens our are useless , been there last century - no breakfast or free lunch.
Saw that as a recommended example in an iterative sequence recommended by the Mota simulation package - they obviously havnt actually built one as it for sure doesnt work.
Yes , tapering the belly to reduce the rear cone works - the R1 has that feature as does the R2

[Wos]

Hello wobbly,

With programmable ignition we try to produce cold egt with high retard to force pipe resonance begin as soon as possible...


For me i recogniced that its alwasys a gamble were to stop this high retard and lowering it as this is necessary to avoid Detonation risk
Every engine Konfiguration has diffrent behaviour...

So far found non rule...for a first curve it has been always a shoot from the hipps...if i had a pipe with resonance peek 8500 took 5000 ...from 26 degree to 19 in 500 reve... better falling earlier and deeper to minimize risk for first try

Is there any rule of thumb how much lowering... and were to start?



Thank you very much once again!

Wolfgang

[JanBros]

@ Wobbly : you once said that for a CVT +/- 33-67% lengths for header and diffusor are too long, better 30-64%. what do you do with the lost length ? just make the pipe shorter (which seems wrong to me) or add to the belly ?

[wobbly]

Wos - the most useable midrange advance is 30 to 32* and this can be a straight line from around 2000 rpm to the point that the pipe starts to be effective and power is rising sharply
per rpm step . This much static can only be used ( on petrol ) as long as the engine is not run for any extended period in that area - even warming up under load at part throttle.
The next hard and fast rule is that if the compression is optimized for the fuel then its all but guaranteed you will need 15* advance at peak power.
Thus you can draw a straight line from the 30* turnover point to that rpm.
How much retard after peak power depends upon how much overev is needed. Then a flat line prevents excess egt rise that would need richer jetting to cool/ not make power.
The straight line retard can be bent upward into a convex form by looking at the TubMax graph shape and pushing your luck with detonation.

But the logic of using alot of advance to " get the pipe into resonance quicker " seems counterintuitive to me.
The less advance , the hotter the pipe becomes , quicker .
And if you have a port timing/pipe TL pairing that creates superposition within the powerband - the hotter the pipe , the quicker you get into that area.
Sure , pushing the advance gives great throttle response , but Hp wise you are fighting excess compression losses BTDC as the piston is pushing against static compression as well as the quickly rising
combustion pressure , thus slowing down the piston speed driven by only by crank inertia.

JanBros - yes a shorter header pulls down the port negative pressure ratio earlier in the depression cycle , favouring power around peak.
And the shorter total diffuser length gives steeper diffuser angles that favor power around peak. Front side , after the clutch has locked is set by the length and steepness of the
final diffuser , then the rear cone angle(s) will set the amount of useable power directly after peak.

If you use the small duct exit approach with no steps - and a steeply angled slip joint , the peak and overev will be dramatically enhanced , so you the have the choice of either steepening the rear cones to pump up the peak and reduce the overev potential or just lengthening the pipe to get more front side.
In all of this you are manipulating the powerband shape with the cones - all the mid does is to connect the effects together at a specific TL.
Sure it has an effect on pipe volume but imho that is the least useable aspect of pipe design variables - its a by product , not an input.

[Wos]

Wos - the most useable midrange advance is 30 to 32* and this can be a straight line from around 2000 rpm to the point that the pipe starts to be effective and power is rising sharply
per rpm step . This much static can only be used ( on petrol ) as long as the engine is not run for any extended period in that area - even warming up under load at part throttle.
The next hard and fast rule is that if the compression is optimized for the fuel then its all but guaranteed you will need 15* advance at peak power.
Thus you can draw a straight line from the 30* turnover point to that rpm.
How much retard after peak power depends upon how much overev is needed. Then a flat line prevents excess egt rise that would need richer jetting to cool/ not make power.
The straight line retard can be bent upward into a convex form by looking at the TubMax graph shape and pushing your luck with detonation.

But the logic of using alot of advance to " get the pipe into resonance quicker " seems counterintuitive to me.
The less advance , the hotter the pipe becomes , quicker .
And if you have a port timing/pipe TL pairing that creates superposition within the powerband - the hotter the pipe , the quicker you get into that area.
Sure , pushing the advance gives great throttle response , but Hp wise you are fighting excess compression losses BTDC as the piston is pushing against static compression as well as the quickly rising
combustion pressure , thus slowing down the piston speed driven by only by crank inertia.

JanBros - yes a shorter header pulls down the port negative pressure ratio earlier in the depression cycle , favouring power around peak.
And the shorter total diffuser length gives steeper diffuser angles that favor power around peak. Front side , after the clutch has locked is set by the length and steepness of the
final diffuser , then the rear cone angle(s) will set the amount of useable power directly after peak.

If you use the small duct exit approach with no steps - and a steeply angled slip joint , the peak and overev will be dramatically enhanced , so you the have the choice of either steepening the rear cones to pump up the peak and reduce the overev potential or just lengthening the pipe to get more front side.
In all of this you are manipulating the powerband shape with the cones - all the mid does is to connect the effects together at a specific TL.
Sure it has an effect on pipe volume but imho that is the least useable aspect of pipe design variables - its a by product , not an input.

Thank you wob

Spoke in case of low reving oldi classic enduros but also from enduro in general

There we need driveability... torque bevore beinn on pipe...and this is the reason why i stay far under 30 degree...never had needs over 26 degree as drivability was lost ...and we are happy to use the area of reso when its begins to start with "cold" egt... as our pipe is not not long enought ...to short

...after this have to keep too long thing on pipe by hotter egt.. and having some overev by more hot egt

Maybe very contraintuitiv ...as we are limited in exhaust port on oldi 1980 too...

I confirm 15 degree to powerpeak


so my conclusions is... its more difficult to find good timing on 1980 enduro than on some roadracers

On our contests conditions vary...tracks unknown...no Chance to learn shifting points...

But...have another Projekt where your advice will help us for sure in future ..a roadracer

Thanks!!!! One more!

Wolfgang

[lodgernz]

We never , ever see 9000 rpm for more than a split second after getting off the gas and down changing into a corner.
The pipe temp has dropped in that period and then what happens on the dyno , happens on the track - we give it full throttle.
Thus a good sim on petrol will reflect this with a wall temp of near 325*C at the bottom of the powerband , then if the dyno run time is correct , we will see the wall temp the
pipe sees on the track at full noise - around 425.

This is not the egt as seen in the header , that is affected by the A/F ratio used and the combustion parameters ( Compression/MSV/Timing ) and this influences the pipe bulk gas temp
that ultimately affects the wave speed just as much as the pipe wall temp does.
This pipe wall temp on petrol has been proven very accurate in many sims from 50cc to 300cc - but only if a realistic A/F ratio is used for the fuel.

Wobbly, this is the first time I have seen any mention of pipe wall temperature.

Would you mind giving us a quick rundown of where to place the sensor, what best sensor, how to interpret results?

[Condyn]

What is the reason why so many championship winning engines seem to end up square, or nearly square? I realize I could spend ages simulating different bore/stroke, port configurations, but the curiosity just struck.

Most of the CVT snowmobile engines I work on are over-square ( short stroke ) so I am asking mainly to get opinions, or more likely facts, on what short comings I am or could be up against with them.

[wobbly]

Lodger , when I asked Neels to add in the option of varying pipe wall temp ( to better simulate reality ) he asked for some actual numbers.
I used a K surface probe clamped onto the outer wall in the mid header , mid body and rear cone end.
These figures from back to back multiple runs re created the egt as seen on track data from the lowest to the highest rpm used , and gave 3 sets of surface numbers.
These then averaged gave 325*C at the bottom of the power band , rising to 425*C after peak power at full throttle.
And when then inputted as actual temps in the sim I got really good correlation for the KZ engine as dynoed and simulated.

I have put these into many , many sims and they have proven to give good baselines for many different applications.
The code uses this input as well as the gas temp generated by all the combustion factors such as A/F , advance , MSV , Compression etc as they all affect actual bulk temp inside
the pipe.

A/F is very important with 11.5 showing good results for unleaded pump and up to 13 for 110 leaded race gas - using compression to suit.
This translates into around actual 620 EGT and 680 EGT for those fuels @ 3X bore from the piston.
If you use the pipe programs from FOS or Neels version in the sim , this also seems to correlate well to between 500*C and 600*C of bulk gas temp throughout the pipe.

[lodgernz]

Wonderful info. Thank you Wobbly.

[wobbly]

A square engine is the best compromise between rpm capability and achievable port time area.
Sure a short stroke can theoretically rev higher , but the piston is heavier so not alot is usually gained.
For the same port timing the short stroke is severely limited in port height despite its larger bore circumferences useable area.

Best example was Yamaha doggedly sticking to 56/50.6 in the TZ all the way up thru the 4DP models - they were completely useless in 125 and 250 GP.
They didnt rev any harder than a square Honda or Aprilia and got their arses kicked Hp wise.
All the way up to 2000 when they finally swallowed corporate pride , bought a Czech anemometric flow visualizer , dumped shit that never worked like twin boost ports , went square
and won first and second in 260GP easily with Jacques.

Sure the Aprilia boys stole points off each other that year , but I was at Philip Island for the final round and the two Yamaha's scrapped over the title right to the finish line -
the whole length of the straight ahead.

[Frits Overmars]

Condyn, Here's a text I happened to have lying around, that might clarify things a bit.

Oversquare is good for four-strokes, where the diameter of the valves that fit into the head is determined by the cylinder bore.
For two-strokes it's the other way around.
I'll give you an example, with a simple rectangular exhaust port, with a port height of 50% of the stroke and a port width of 70% of the bore.
For an engine with 100 mm bore and 100 mm stroke (that's 785,4 cc) this would mean a port height of 50 mm, a port width of 70 mm and a port area of 50 x 70 = 3500 mm².

Let us see what happens when we make that engine really oversquare. We double its bore: 200 mm instead of 100 mm.
But in order to maintain the original cylinder capacity, the stroke must shrink from 100 mm to 25 mm.

Now let's look at that exhaust port again. Its height becomes 50% of 25 mm = 12,5 mm, its width becomes 70% of 200 mm = 140 mm, and its port area becomes 12,5 x 140 = 1750 mm². We've lost 50% of our original port area!

This means that our oversquare engine can rev only half as high as our original engine before running out of breath.
It can produce only half the power of the original engine. Oversquare is not good for two-strokes.

PS Wob, those short-stroke Yamahas were bad alright, but you are doing them an injustice.
From the RD56 onward (my favourite Yam, a simple aircooled twin that beat the mighty six cylinder Honda) the stroke has always been 50.7 mm, not 50.6 mm.