ESE's works engine tuner

[Niels Abildgaard]

I'm trying to understand why the volume of the transfer ports is important.

In order for air to flow OUT of the transfer ports, doesn't an equal volume of air have to flow IN to the transfer ports?

So that the volume of the ports is not so important compared to the resistance to flowing in and out? (and through, and the angles and shapes and such).

Let us talk of transfer port areas and tranfer channels volume and debate will surely fuck up in some other entertaining way.

[wobbly]

Can admin please block the fuckwit posting nothing to do with the subject at hand.

[wobbly]

I think that the relationship between port area and transfer volume isnt something we can discuss with any hope of accuracy as the examples to hand
vary so hugely , but the subject would be a great PhD dissertation paper with a huge amount of comparative testing needed.
But of course that isnt going to happen in todays anti 2T environment.

In my RSA sim the transfer total volume is quite a bit less than the the swept 125cc , and this makes sense as the B port entry is tiny in comparison to even its own A port entry area.
So in this case we have a " small " transfer volume/swept volume relationship , but the highest Delivery Ratio ever seen of 1.4.
The B port area is much greater than that of the A due to the chordal width as well as the higher timing - so we cant draw any definitive conclusions from Jans work.

Many RSA clones used in 250 Superkart racing have the B port entry much larger , as the rear case studs have been moved back - away from the original Rotax pattern Jan had to use.
If this is " better" I do not know in any absolute terms, as the newer generation pipes on those engines probably account for the greater power produced , not the duct configuration in my opinion.

Another example I am working on is the 3XV/4DP on a TZ250. This is making well over 50 Hp in the sim , but it has a short cylinder height and thus short ducts , but the much larger entry areas give almost a 1:1
duct to swept volume ratio , with a DR of close to 1:1.
So in short ,I cant draw any definitive guides from any of this.

Edit - one point of interest was that when I visited Bud Asklands workshop and dyno facility he was working on Raineys new YZR cylinders , they looked almost identical from the bottom as a 3XV
but he had epoxied the A, B , C outer walls with Devcon that was around 5mm thick , when I asked he said the smaller volume made a couple more Hp but that Wayne had remarked that the " throttle connection " feeling
was way better when tested back to back. Being a smartarse I asked was the difference due to case volume - he handed me a case 1/2 with quite large pockets machined above the mains directly next
to the flywheel faces , the answer was " no , it took forever to get the CCR correct so it wasnt changed at all".
Smart man.

[Condyn]

Speaking of devcon Wobbly,
A few months back you were talking about cutting a transfer duct vinamold to fix the rearward leading edge angle by injecting liquid Devcon. I have a cylinder with this same problem and was wondering if you have any results on this test?
And as I am typing this, I remember the strange 2 piece, steep to shallow baffle cone pipe, perhaps on the same project. I am also curious if this has been built and tested on the dyno to verify the sim results. A side note is that I do not doubt the sim, but the oddity still has me wondering. If the project is stalled, I think we all can relate to that. Thank you.

[wobbly]

Making the 3XV B port front wall perpendicular to the bore CL proved to be way easier than I had imagined. The Devcon Liquid B , is very liquid.
So I took the cylinder I cut up to check everything , and simply used a bubble level to get the cylinder 1/2 sitting dead horizontally square , and put some race tape across
the transfer entry at the level I wanted , and using a big plastic syringe I basically kept pouring it in until I could see the liquid creep up to the port edge in the bore.
I left it ovenight to cure , ripped off the tape , and voila ,a perfectly angled and flat duct fill in.
I am going to do the same with the roof of the A port , simply angle the cylinder to 22* and pour in the liquid until it reaches the port edge.
But this will be done after the cylinders are plated.

The " reverse " angled rear cone has been dyno tested in another project , but as a 3 step ,all equal length decreasing contraction.
In the sim the two angle and 3 angle make virtually the same power , but to make 3 shorter rear cones is easier to do in stainless.
I am currently awaiting confirmation that the pipe centerlines actually work on the bike , if the guy doing this fucks around much longer I
have arranged for a 4DP to be delivered to my shop in November , so I can confirm that the laser pattern shapes actually fit.
Its an issue as the pipes are fatter than anything Yamaha put on that series of bikes , and the 3XV has a much steeper duct exit angle than the 3YL/4DP.

EDIT - note that the reverse cone idea only works in this particular situation where everything was done to pump up peak and overev power , the front side being
taken care of due to the fact the 3XV Power Valve closes off all 3 Exhaust ports.
The diffuser geometry was the same - if alot of mid power was needed the last diffuser would be long and steep.

[wobbly]

Here is the reverse rear cone idea used in an overbored TZM150 for short course road racing in Vietnam , again this closes off all 3 Exhaust ports so that easily takes care of the front side power.

[Rudex]

Here is the reverse rear cone idea used in an overbored TZM150 for short course road racing in Vietnam , again this closes off all 3 Exhaust ports so that easily takes care of the front side power.

Wobbly,

The first segment, is the flange?. Why this huge angle before to header?. BTW, I thought header greater than 3.6º it was very high, it is very angled and short.


Another thing. Why the new generation of cylinders in karts or little engines like 2fast, 8.1, malossi, take the parallel wall with the front Aand bottom Bports and the RSA take the B very straight?. If I remember correctly, you posted RSA is for top end power and the parallel separation is for more mid power.

[Condyn]

Thank you for the reply wobbly. I like the liquid epoxy method. It makes it quick and repeatable.

[wobbly]

Mr Rude , the first question re the flange internal angle. This is all covered in the paper I wrote that Neels makes available to EngMod users.
The Exhaust duct exit is sized to create 0.8 Mach velocity ( 0.75 x the port effective chordal area ) at this point , and the header inlet is back to 100% ( or slightly greater ).
This steep angle and the exit velocity starts the diffuser action very early in the the cycle, and gives a huge increase in power around peak and in the overev.
And this can then allow the use of a longer pipe if needed, to generate more front side power.

The header angle and length again dictates the amount of overev generated ( in concert with the 1st diffuser length and angle ) where shorter and steeper again gives peak and overev power
always at the cost of front side.
But in both cases shown all 3 Exhaust ports are closed off by the old ( but very effective ) Yamaha drum PV, that can be servo driven to be fully up only about 1000 rpm from peak
and this gives a huge boost to the front side mechanically , not using pipe the geometry.

Header angle evolved from a very common 3.2* included to then being split 50/50 - 3.5/5.5*, then to a constant 4.5* with the 1st diffuser then being shorter and steeper.
I first used the 4.5* angle on Dr Henise's TZ350 LSR breaking engine - where peak and overev were all important.
But in the latest R2 KZ design we are forced to compromise , by the need to generate more front side power , and use much shallower header angles , but keeping it very short
to then give plenty of overev as well.
All this specific tailoring is super simple to nail down , reliably , using Neels genius code.

The A and B port centerline crossing points are , again , a compromise to generate a wide powerband or simply extra power around peak.
The RSA has a powervalve , so every element of the design could be directed toward creating top end.
Jan did , exactly what I have just done with the 3XV , filled in the B port front wall with epoxy , to make it perpendicular to the bore centerline , and this instantly gave a boost in power around peak.

But the scavenging regime ( including the port stagger geometry ) in say the TM KZ engine ( no PV ) needs a naturally much wider spread of power - again no free lunch, and the peak is suppressed.
Thus we now commonly see differing sets of port heights and wall angles , that are tailored specifically to the end use and the engines mechanical layout dictated by the presence, or lack of a PV.

EDit - the A port front wall in your A ( LH ) version should intersect at the same point as the A and B rear wall - the RSA scavenging pattern.

[jamathi]

Mr Rude , the first question re the flange internal angle. This is all covered in the paper I wrote that Neels makes available to EngMod users.
The Exhaust duct exit is sized to create 0.8 Mach velocity ( 0.75 x the port effective chordal area ) at this point , and the header inlet is back to 100% ( or slightly greater ).
This steep angle and the exit velocity starts the diffuser action very early in the the cycle, and gives a huge increase in power around peak and in the overev.
And this can then allow the use of a longer pipe if needed, to generate more front side power.

The header angle and length again dictates the amount of overev generated ( in concert with the 1st diffuser length and angle ) where shorter and steeper again gives peak and overev power
always at the cost of front side.
But in both cases shown all 3 Exhaust ports are closed off by the old ( but very effective ) Yamaha drum PV, that can be servo driven to be fully up only about 1000 rpm from peak
and this gives a huge boost to the front side mechanically , not using pipe the geometry.

Header angle evolved from a very common 3.2* included to then being split 50/50 - 3.5/5.5*, then to a constant 4.5* with the 1st diffuser then being shorter and steeper.
I first used the 4.5* angle on Dr Henise's TZ350 LSR breaking engine - where peak and overev were all important.
But in the latest R2 KZ design we are forced to compromise , by the need to generate more front side power , and use much shallower header angles , but keeping it very short
to then give plenty of overev as well.
All this specific tailoring is super simple to nail down , reliably , using Neels genius code.

The A and B port centerline crossing points are , again , a compromise to generate a wide powerband or simply extra power around peak.
The RSA has a powervalve , so every element of the design could be directed toward creating top end.
Jan did , exactly what I have just done with the 3XV , filled in the B port front wall with epoxy , to make it perpendicular to the bore centerline , and this instantly gave a boost in power around peak.

But the scavenging regime ( including the port stagger geometry ) in say the TM KZ engine ( no PV ) needs a naturally much wider spread of power - again no free lunch, and the peak is suppressed.
Thus we now commonly see differing sets of port heights and wall angles , that are tailored specifically to the end use and the engines mechanical layout dictated by the presence, or lack of a PV.

EDit - the A port front wall in your A ( LH ) version should intersect at the same point as the A and B rear wall - the RSA scavenging pattern.

The first time I epoxied the B-duct was in 1997, with a very good result, after that all my cylinders were cast that way.
I made my first triple-port exhaust in 1971, may have been the very first to do this...
Far better than a bridged port: More power and a LOT less problems!!!
And you can make a wider A-port...

[Rudex]

Mr Rude , the first question re the flange internal angle. This is all covered in the paper I wrote that Neels makes available to EngMod users.
The Exhaust duct exit is sized to create 0.8 Mach velocity ( 0.75 x the port effective chordal area ) at this point , and the header inlet is back to 100% ( or slightly greater ).
This steep angle and the exit velocity starts the diffuser action very early in the the cycle, and gives a huge increase in power around peak and in the overev.
And this can then allow the use of a longer pipe if needed, to generate more front side power.

The header angle and length again dictates the amount of overev generated ( in concert with the 1st diffuser length and angle ) where shorter and steeper again gives peak and overev power
always at the cost of front side.
But in both cases shown all 3 Exhaust ports are closed off by the old ( but very effective ) Yamaha drum PV, that can be servo driven to be fully up only about 1000 rpm from peak
and this gives a huge boost to the front side mechanically , not using pipe the geometry.

Header angle evolved from a very common 3.2* included to then being split 50/50 - 3.5/5.5*, then to a constant 4.5* with the 1st diffuser then being shorter and steeper.
I first used the 4.5* angle on Dr Henise's TZ350 LSR breaking engine - where peak and overev were all important.
But in the latest R2 KZ design we are forced to compromise , by the need to generate more front side power , and use much shallower header angles , but keeping it very short
to then give plenty of overev as well.
All this specific tailoring is super simple to nail down , reliably , using Neels genius code.

The A and B port centerline crossing points are , again , a compromise to generate a wide powerband or simply extra power around peak.
The RSA has a powervalve , so every element of the design could be directed toward creating top end.
Jan did , exactly what I have just done with the 3XV , filled in the B port front wall with epoxy , to make it perpendicular to the bore centerline , and this instantly gave a boost in power around peak.

But the scavenging regime ( including the port stagger geometry ) in say the TM KZ engine ( no PV ) needs a naturally much wider spread of power - again no free lunch, and the peak is suppressed.
Thus we now commonly see differing sets of port heights and wall angles , that are tailored specifically to the end use and the engines mechanical layout dictated by the presence, or lack of a PV.

EDit - the A port front wall in your A ( LH ) version should intersect at the same point as the A and B rear wall - the RSA scavenging pattern.

Ok, Thanks Wobbly!.

About the layout port. I got it, but what is the current layout on kz-r2 for example?. I mod the picture

[wobbly]

I cant give you the exact figures of the R2 as its current homologation and under NDA , but the R1 is very similar with slightly differing timings.
The boost is wider at 32% and both the B walls and A rear cross near to 75% , the A front @ 78%.
I think that the wider boost naturally forces the A & B crossing point further forward , such that their rearward bias does not mask the area directly in front.
Your LH drawing of the RSA the crossing point looks too far forward , its only 18% of bore = 9.7mm , thus the TM is closer to the centerline.
The boost has been run narrower along with the RSA reverse stagger , and it made large power gains above peak , but lost way too much down at 10,000
where it runs off several 2nd gear hairpins during a lap.
The R2 also has a new concept of the smallest CSA of the main Exhaust duct is at around 1/3 length up from the flange face , keeping the original area at the exit.

[diesel pig]

I made my first triple-port exhaust in 1971, may have been the very first to do this...
Far better than a bridged port: More power and a LOT less problems!!!

Jan puts why I hate bridged exhaust ports and love triple-port exhaust, so well, I need say no more on the subject.

[Rudex]

I cant give you the exact figures of the R2 as its current homologation and under NDA , but the R1 is very similar with slightly differing timings.
The boost is wider at 32% and both the B walls and A rear cross near to 75% , the A front @ 78%.
I think that the wider boost naturally forces the A & B crossing point further forward , such that their rearward bias does not mask the area directly in front.
Your LH drawing of the RSA the crossing point looks too far forward , its only 18% of bore = 9.7mm , thus the TM is closer to the centerline.
The boost has been run narrower along with the RSA reverse stagger , and it made large power gains above peak , but lost way too much down at 10,000
where it runs off several 2nd gear hairpins during a lap.
The R2 also has a new concept of the smallest CSA of the main Exhaust duct is at around 1/3 length up from the flange face , keeping the original area at the exit.

Thank you Wobbly,

The picture is from forum i got it to show you grafiically the separation of ports A and B in RSA and style TM cylinders. But I checked it putting in scale 100 and it is ok 82% and 68% rear B, are ok.
I modified the picture to simulate the tm layout. 75% for both A and B rear walls and 78% for front wall A, front B wall is not clear.

With the picture I see visually the diference.

Thank you for your help.

[wobbly]

If you widen the boost port on the TM side , and shift the B rear wall over as well , the exit angle would be I believe almost identical to the RSA
but crossing closer to the bore center.
The A rear wall crosses at the same 75% point as the B front and rear. But the A front should be closer to the boost at 78%.
Having each side marked as A and B is confusing things as well, should be RSA and TM.