ESE's works engine tuner

[wobbly]

Muhr , theoretically the hook angle passing thru bore center would seem logical but CFD would sort that in no time.

[Muhr]

I guess this statement was intended for Patrick. I think exactly as you say that the center bore will not be far from the truth, then the question is how long in the cycle this will be a supplement. I guess as soon as you get an established negative pressure on the exhaust side, these "hooks" will probably be a dead space.
So simulating this you probably needs a transient simulation with the exhaust side and piston movement

[andreas]

To maintain the arched cross area of the port (which becomes an issue only when the port is almost fully open), why put the divergence (hooks) in the side walls, and not in the floor preferably, as it will eventually be a circular path?

[Frits Overmars]

Patrick , one concept I have always wanted to try on a single port ( very rare , work wise nowdays ) is to add concave hooks on the sides of the duct above TPO. The shape as is common on T port outer edges - that increases considerably that ports Cd value.

concave as in adding a side wall angle to the port as pictured below? If so, I've been trying to model this in engmod as a bridged exhaust with a 0 width bridge. Actually started down this path trying to figure out how much side wall angle to add. Clearly the effective port area would be between 0° side wall as the min, and a radial angle with the area of the curved port face (adjusted for down angle of 25%) for the max. I think that will be the last and largest part of this study.

....Theoretically the hook angle passing thru bore center would seem logical but CFD would sort that in no time.

.... I think exactly as you say that the center bore will not be far from the truth, then the question is how long in the cycle this will be a supplement. I guess as soon as you get an established negative pressure on the exhaust side, these "hooks" will probably be a dead space. So simulating this you probably needs a transient simulation with the exhaust side and piston movement

To maintain the arched cross area of the port (which becomes an issue only when the port is almost fully open), why put the divergence (hooks) in the side walls, and not in the floor preferably, as it will eventually be a circular path?

I started playing with CFD when I was still young and handsome (I had to write my own software, it's that long ago) and it did, and still does prove Wobbly, Patrick, Muhr and Andreas right.
The initial hook angle passes through the bore center and the sidewalls will start as a circular path, but this path will almost immediately tighten, as the difference between the exhaust duct roof angle and its floor angle will take care of maintaining a rather constant cross flow area ('rather' because of some CFD finesses that may be outside the scope of this post).
In English: as the floor drops, the sidewalls should close in.
Maintaining the circular path would lead to an undesirable increase of the duct volume. I have been distributing a couple of sketches which show the 20° difference angle that also appeared in Patricks drawing, instead of the initial radial angle, because that hopefully reduces the number of cylinders ruined by overzealous hookers tuners.


Below: an exhaust duct solid, modelled from the points cloud generated by my CFD, and a pair of cut-away pics showing the dropping floor and the closing-in sidewalls of the twin exhaust ducts in my cylinder.

[Muhr]

It's a bit hard core to code your own CFD. But what do you do if the only ones who have access to it are NASA .I have tried to find info on how it went with that project, I think if I remember correctly I saw a picture of a cylinder. Also high interesting to hear more from a perspective on the raised exhaust port floor.

[OopsClunkThud]

Frits, in your CFD results did the radial alignment of the hook flow better than the 20°? I had assumed that the 20° came about as a balance of the initial radial outward expansion vs the developing flow velocity down the exhaust.

To make sure the port is maintaining a constant taper rate (from port area down to the ~90% or whatever value is settled on in engmod), I've placed sketches at the .25, and .5 bore distance along the path. These have a target area based on position and with the top and bottom shape of the port fixed, the needed width can then be found. From those widths a smooth path for the rails that form the hooks can be found and then the final lofted shape can be formed.



I'll try and write these steps up like Lohring did with the "3D Engine Modeling.pdf"

Thanks for all the guidance from everyone!

[Frits Overmars]

Frits, in your CFD results did the radial alignment of the hook flow better than the 20°? I had assumed that the 20° came about as a balance of the initial radial outward expansion vs the developing flow velocity down the exhaust.

It was fractionally better at outflow and definitely worse at reverse flow. Makes you think twice about transfer duct hooks....The 20° also helped to keep the exhaust duct volume small.

To make sure the port is maintaining a constant taper rate (from port area down to the ~90% or whatever value is settled on in engmod), I've placed sketches at the .25, and .5 bore distance along the path. These have a target area based on position and with the top and bottom shape of the port fixed, the needed width can then be found. From those widths a smooth path for the rails that form the hooks can be found and then the final lofted shape can be formed.

The only 'taper' I used came from the radiused exhaust port top edge and the radiused piston timing edge; between them they gave a rather nice impersonation of a De Laval nozzle.
Downstream I used a 3° divergence. The tapering comes from Wobbly's experience in making the best use of a conventional exhaust duct. It certainly works in those circumstances, but mine were different, wich a much higher than conventional exhaust floor.

[wobbly]

The only caveat I have seen about lifting the port floor ( thus the reason I havnt been down that path as yet ) is that Jan mentioned to me lifting the floor only worked
as long as the Aux ports were fully optimised ( around to bore center ) to achieve the Blowdown needed for the power produced.
As a single port is always going to be inherently Blowdown limited , I would be carefull .

The other point I have kept in mind ( as the KZ engines I work mostly on are certainly not yet Blowdown optimised - maybe soon with my new Pankle pin copies ) is that Jans
actual documented floor lifting experiments ceased at + 3mm . So as yet there is no documented dyno proof I have seen , that going higher is beneficial.

[Frits Overmars]

The only caveat I have seen about lifting the port floor is that Jan mentioned to me lifting the floor only worked as long as the Aux ports were fully optimised ( around to bore center ) to achieve the Blowdown needed for the power produced.

Wob, as you may have noticed in the above pics, my symmetrical scavenging cylinder has abundant total exhaust port width. At 190° exhaust timing and 130° transfer timing it has clearly more blowdown STA than the corresponding transfer STA, so the transfers set the optimum rpm, with a save blowdown reserve for overrev. The Aprilia RSA does not have such a reserve, and raising the exhaust floor to above the transfers, as I have done, would create serious problems. But I think that for the RSA an exhaust floor 10° below the top of the transfers should leave sufficient blowdown reserve for overrev.

The other point I have kept in mind is that Jans actual documented floor lifting experiments ceased at + 3mm . So as yet there is no documented dyno proof I have seen , that going higher is beneficial.

That documented proof is not available because Jan went into retirement. He had planned to carry on experimenting with higher floors and the cylinders for those experiments had already been cast when Jan realised that he would be gone before he could test them, so he cut those floors down in order to prevent certain people from claiming future improvements as their own.

[TZ350]

.
More from 2 Stroke Stuffing.
.

[Muhr]

The only caveat I have seen about lifting the port floor ( thus the reason I havnt been down that path as yet ) is that Jan mentioned to me lifting the floor only worked
as long as the Aux ports were fully optimised ( around to bore center ) to achieve the Blowdown needed for the power produced.
As a single port is always going to be inherently Blowdown limited , I would be carefull .

The other point I have kept in mind ( as the KZ engines I work mostly on are certainly not yet Blowdown optimised - maybe soon with my new Pankle pin copies ) is that Jans
actual documented floor lifting experiments ceased at + 3mm . So as yet there is no documented dyno proof I have seen , that going higher is beneficial.

This with raised exhaust port floor keeps me awake in the evenings sometimes, I put exhaust 2.2mm over transfer floor on my little 50cc hobby project, I can only assume that you lose a lot of top if this would be a failed attempt?
Have also had some trouble with the weight of the crank assembly, have now ordered materials that I will pick up tomorrow to make the balance shaft and rotor shaft in titanium. Do not know if it was a particularly good idea but I guess I will get answers. have not dared to start macining the cast parts before I got control of all the small parts

[jamathi]

Wob, as you may have noticed in the above pics, my symmetrical scavenging cylinder has abundant total exhaust port width. At 190° exhaust timing and 130° transfer timing it has clearly more blowdown STA than the corresponding transfer STA, so the transfers set the optimum rpm, with a save blowdown reserve for overrev. The Aprilia RSA does not have such a reserve, and raising the exhaust floor to above the transfers, as I have done, would create serious problems. But I think that for the RSA an exhaust floor 10° below the top of the transfers should leave sufficient blowdown reserve for overrev.

That documented proof is not available because Jan went into retirement. He had planned to carry on experimenting with higher floors and the cylinders for those experiments had already been cast when Jan realised that he would be gone before he could test them, so he cut those floors down in order to prevent certain people from claiming future improvements as their own.

Now I regret very much not to have left those cylinders with raised exhaust port floor to my successors, at least the result would have been known now, 12 years later....
But I don't understand why nobody else had the same idea!
It just takes 1 day of testing...
If successful there would have been more room for the transfers..

By the way Frits, how did your cylinder run? Any test results?
Someone else had the same idea: it would not rev at all.....
Was yours any better?
Blowdown was certainly enough!
I liked the design very much when I first saw it!
But some people tried it without result...

[Frits Overmars]

By the way Frits, how did your cylinder run? Any test results?
Someone else had the same idea: it would not rev at all. Was yours any better?
Blowdown was certainly enough! I liked the design very much when I first saw it! But some people tried it without result...

Wasn't that the same 'someone else' who claimed he was the only person who knew everything about the Ryger and who publicly wrote: "Jan Thiel doesn't understand how a two-stroke works" ?

[dutchpower]

Why your cil.is stil on the shelf by Harm Frits

[jamathi]

Wasn't that the same 'someone else' who claimed he was the only person who knew everything about the Ryger and who publicly wrote: "Jan Thiel doesn't understand how a two-stroke works" ?

Yes, the same person, now the RYGER has died......
He sent me a video to show his 'system' worked.
Later it did not rev.
Was yours any better, or was it never tried at all....?
And what are your thoughts about the RYGER now, now that you can safely talk about it? You once tried it, on a kart, you were very enthusiastic.....
But it never achieved anything at all....!!!