ESE's works engine tuner

[Hoebra]

@Frits Overmars, how is it possible to get some of the Piston Pin Plugs you showed here many times in 12mm version?

[ceci]

[OopsClunkThud]

results: The flat top piston with 1mm radius at the edge flows less exhaust than the 4° conical. When the port is open only 5° the radius flows 10% less, and just before the transfers open it's only 2% less. Looking at the images it's clear that relative to the conical piston the radius is shrouding the flow and reducing the area. The radius also does not add much to the ability of the gas to turn the corner at the edge of the piston. At sonic speeds this is driven by the heat and pressure of the gas. If there is enough heat and pressure (as when the port first opens) the gas can turn ~20° around the edge of the piston, and if there is not enough (at the end of blowdown) then the flow separates from the radius.



I did some quick tests with a larger radius (keeping the .5mm flat height above the timing edge) and it flowed as well as the conical at the end of blowdown. I'm building out a larger test plan to zero in on an optimal value.

[wobbly]

So I wonder if up the front side the transfer flow Cd effect is greater on power output than the downside of the Exhuast flow Cd.
But in the overev area , the effects reverse efficacy.
Interesting that Jan had to radius the port timing edge to regain lost power in the Aprilia , but a radius on the piston has a negative effect as shown in CFD.

[Frits Overmars]

@Frits Overmars, how is it possible to get some of the Piston Pin Plugs you showed here many times in 12mm version?

It was never my intention to start a business with the plugs, which is why I posted both the drawings and the choice of material here, so that anyone who is interested can make them himself or have them made. They are quite simple really.

Interesting that Jan had to radius the port timing edge to regain lost power in the Aprilia , but a radius on the piston has a negative effect as shown in CFD.

In the above pictures the piston edge is flush with the exhaust floor. I would like to see the effect of a radiused edge at the onset of blowdown.
I'd also like to see the effect of the radius on the flow from the transfer ducts and the exhaust duct into the cylinder, and the effect it has on piston cooling.

[OopsClunkThud]

In the above pictures the piston edge is flush with the exhaust floor. I'd rather like to see the effect of a radiused edge at the onset of blowdown.
I'd also like to see the effect of the radius on the flow from the transfer ducts and the exhaust duct into the cylinder, and the effect it has on piston cooling.

Frits,

both pictures are at the same same level of zoom, but the exhaust floor is well below the frame of the view. as this is a simple 2D profile, the piston and top of the port/cylinder are all in gray. The subsonic flow in blue and supersonic flow in red.

the first picture is 5° into blowdown, and the second is right at the end of blowdown. Reversing the exhaust is easy enough, I'll pull the pressure values from engmod, the rest of the model is the same (input flow vector is odd, but not hard).

As for the transfers, I'll get to it.

I'm thinking snapshots at 5° increments should be enough, with maybe an extra at 1-2° after EPO.

[Flettner]

Thankyou, thats worthwhile and intetesting work.

[Vannik]

The subsonic flow in blue and supersonic flow in red.

This makes me doubt the software as there is nothing there that can cause supersonic flow, and sonic flow will form a shock and jump back to subsonic. The pressure ratio and duct shape required for supersonic flow does not exist in an engine.

[Muhr]

Sometimes you can be fooled by which wall condition you use, maybe so that you use a slip?
I have been there myself a few times

I had thought to see something similar to below. but I do not have access to so much data power on my simple variant

A little more interesting reading

https://www.e3s-conferences.org/arti...2020_06005.pdf

[OopsClunkThud]

I'll be the first to admit that with regards to CFD (and many other topics) I'm no expert. I've been trying to square the "no supersonic" bits of Blair's book with what I was seeing in the 3D CFD on simscale. That's part of what drove me to try some simplified 2D running in SU2. I also very much appreciate the input and help, even (or especially) when I have it wrong.

I was thinking that since the flow is choked, the only way to increase the mass flow rate was to change Cd or A. To decompose the problem, I was looking at the side slice of the 2D flow to see what could be done to improve Cd*A would inform how to shape the port roof and piston, while a top view 2D slice of the side wall angle would inform, well the side wall. Then I could put those back into a 3D and see how it works out.

In both 3D and 2D the walls were modeled as no-slip, but for 2D I made the walls adiabatic. This was all steady state, so that's a known difference from reality.

[Vannik]

Supersonic flow requires a different solving methodology as the waves move in one direction only, as the medium (gas) moves faster than the speed of sound the "rearwards" wave also moves forward because its speed rearwards is less than the gas flow speed. If the numerical method does not do this transition you can get wrong results.

There has been a number of PhD theses in the last few years on the Cd value during sonic flow, it is no longer assumed to be constant. This is very important to 2T engines as the blowdown phase is mostly sonic flow. This is a good area to focus on in my opinion.

[jamathi]

So I wonder if up the front side the transfer flow Cd effect is greater on power output than the downside of the Exhuast flow Cd.
But in the overev area , the effects reverse efficacy.
Interesting that Jan had to radius the port timing edge to regain lost power in the Aprilia , but a radius on the piston has a negative effect as shown in CFD.

Our boss made a new pipe.
It had MUCH better power at intermediate revs.
But the engine did not want to rev anymore.
So I tried to radius the exhaust port:
Max revs got back, same as before
And the power at intermediate revs was the same as with the old pipe......
Nothing gained and nothing lost....
But he could not sell the new pipes to the various teams!!!
Al new cylinders were radiused after this.
So he could sell a new pipe with each new cylinder.
Selling pipes and pistons was a very profitable business.
There were 'light' and 'heavy' pistons.
The light ones had to be changed after each practice, or you would lose a lot of power.
With the heavy ones you could do all practices and the race....
So using light pistons was highly recommended!!!!

[Vannik]

Selling pipes and pistons was a very profitable business.
There were 'light' and 'heavy' pistons.
The light ones had to be changed after each practice, or you would lose a lot of power.
With the heavy ones you could do all practices and the race....
So using light pistons was highly recommended!!!!

We once shared a pit with an Aprillia factory supported team and it was the first time in my life where I saw them removing pistons after each session and checking the dome height in a special jig. Depending on the amount of collapse of the dome they would either bin it or use it with a different head in the next session. Factory support meant they could buy the special stuff, not getting them free.

[wobbly]

Amazing really - if you had not found the solution Jan then the Great Leaders new pipe would have been exposed as being what it was , a flop.
But in the end he came out smelling of Roses as a ton of cylinders and pipes and pistons got sold , and certainly not because he was a clever tuner.

[jamathi]

We once shared a pit with an Aprillia factory supported team and it was the first time in my life where I saw them removing pistons after each session and checking the dome height in a special jig. Depending on the amount of collapse of the dome they would either bin it or use it with a different head in the next session. Factory support meant they could buy the special stuff, not getting them free.

No, nothing for free
On the pistons they had a gain margin of 80%...