ESE's works engine tuner

[Muhr]

No..
I googled it.
It was just that i was writing about.
If you have this concern, how does car manufacturers solve this problen in the heads with far more heat in exhaust than a twostroke?

Think about it for a while

I will try to clarify what I'm aiming for.
If we think that the water rises from the crankcase quite close to the cylinder wall, then the flow of water will take the easiest way to the cylinder head, which means you will get a slightly reduced flow agents the exhaust flange. If I add cooling fins or do some small cavities between the bolt patterns of the flange then I wondered if the even more reduced flow along with the increased specific surface could cause it to start locally film boiling.
which would lead to a deterioration of cooling effect in that area.
And would mean less heat transported from the heat source and then a theoretically lower water temperature but less cooling at load peaks.

I am not saying that I am right but that was my question.

If you have this concern, how does car manufacturers solve this problen in the heads with far more heat in exhaust than a twostroke?

They think of such things I suppose

[Muhr]

Muhr, are you drawing this cylinder in Fusion 360? It looks like it from the pictures. I use Fusion 360 and have been trying to figure out how one goes about drawing a cylinder with all the passages the way you want them in that program. I'm not proficient (obviously) in CAD drawing, but its neat to see what you have done.

yes it's 360! However, my experience is very limited when I only used the program for 2 weeks. I have used other programs, so I'm not a total beginner.
I started myself with sketchup to get an understanding of the methodology. It's not as overwhelming at first and has many good exercise videos you can do. (and it free)

[Michael Moore]

I can do pretty easy stuff like, draw a shape as an entry to the port and then the same for the exit, and then create a loft to follow a spline curve, but that doesn't give me enough detail. would you just draw the cross section of the port ever few mm lets say, and then just loft that?

You are probably about where I am. If beginning and ending shapes lofted following a guide curve doesn't get you where you want to go then I guess you'd have to have some shapes in-between to loft to. And much of that will probably depend on the capabilities of your CAD software, whether you can loft with more than one guide curve etc.

Can you do a prismatic extrusion of the basic shape of the port in one direction and then use the outline of how you'd like things to look in a different axis to trim that, and then take that shape and fillet edges?

My guess is that whatever you can model is probably not exactly what you want, so you get close, cast and port it until you like it, pull a Vinamold or similar flexible casting from the port and then 3D scan the molding and then clean up the scan and work from that for the next iteration of the model.

cheers,
Michael

[Jannem]

So, revoking my foil hat project of trying a supersonic nozzle in the exaust.

It's been proven over and over by guys way beyond my level that a properly developed engine will make max. power when velocities are at .8mach, so why bother?

I don't have an access to engmod, but I recall from posts of others that sim may show benefits beyond .8, but that is where the reality sets in. I'm speculating this would be a product of non uniform velocity distribution in the nozzle, which is evident in a cfd picture posted here before. .8 mach may well be locally choked already.
Whether going above mach1 would give anything is also questionable from multitude of viewpoints, but #1 question is the inevitable mass flow limitation at nozzle choke. As it seems from the sims presented here, a highly tuned engine at full song may still have approximately 30% of it's blowdown port time area subsonic, so maybe there could be something.

To get an idea if this is an immediate failure or if there really could be something to be gained, I designed a nozzle, which should surely make the test engine visibly blowdown limited by the nozzle, but also go supersonic before that. Comparing that against the non-limited duct back-to-back ought to show if its just a brick wall situation, with power diving already at lower revs. The brick wall ought to be there when the port is sonic choked, but the interesting bit is before that.

As I don't own a lathe, I DIY'd a simple way to get area changes I wanted. The side benefit is to be able to test for the impact of uneven velocity distribution by 90deg rotation of the nozzle. For subsonic part of the cycle the horizontal orientation ought to be better, for supersonic, I have no idea.
As there are so many areas where possible claims of gains could be shot down by scratching the surface of my testing implementation, I wont be making any. And of course, if results show loss of power everywhere, then things are more straight forward.

So, here's the nozzle in the fitting phase with sliced pipe and the, not so beautiful, finished form.

[SwePatrick]

I will try to clarify what I'm aiming for.
If we think that the water rises from the crankcase quite close to the cylinder wall, then the flow of water will take the easiest way to the cylinder head, which means you will get a slightly reduced flow agents the exhaust flange. If I add cooling fins or do some small cavities between the bolt patterns of the flange then I wondered if the even more reduced flow along with the increased specific surface could cause it to start locally film boiling.
which would lead to a deterioration of cooling effect in that area.
And would mean less heat transported from the heat source and then a theoretically lower water temperature but less cooling at load peaks.

I am not saying that I am right but that was my question.


They think of such things I suppose

I actually think you are overthinking or overdeveloping, just make sure no air can be trapped and you´re good to go.
That´s what i have done in several carengines that i have tuned to about 1000hp, they have never shown any heatproblems even though materialthickness in heads were very thin after my portjob.
And they run closer to boiling and a lot hotter exhaustgasses, sometimes above 1000c, never been any problems.
But if air is trapped somewhere i have seen engineblocks crack due to steam shockwaves in those airpockets.

You also need to design waterpassages so the water flows evenly around every path.
This is often done with the headgasket.

If i look into my kawasaki cylinder, water inlet and outlet are on the same side of the barrel and head, rear against the carb.
Water goes into cylinder above the reedcase, then flows to the front, turns upwards through headgasket and flows rearwards in head to the outlet.
In gasket one can see it´s the same tech as in an car engine gasket.
It restricts flow at rear to control flow so it heads forwards in cylinder.
but,, it has holes in the rear to vent out airbubbles it there are any, just as in an carengine

So, lead flow to every path with some finesse, and you´re good.

No need to think of boundarylayer etc etc boiling.
To minimize that if you are afraid of it you can add some additive in your coolant to lower surfacetension, the water becomes 'wetter'.
It actually works fine with soap
But just a couple of drops if you don´t like suds

[lohring]

I've not used Fusion 360 (I have Rhino and Alibre) but you might try modeling each passage individually then doing a Boolean subtraction from a large solid with each one. Remember to look for axes of symmetry that will let you draw a feature and then mirror it across the axis, reducing the amount you have to create.

cheers,
Michael

I'm also learning Fusion 360 to draw 2 stroke passages. What I've learned so far:

Don't use surface models (Patch) unless a solid model won't work.
If you use surfaces for passages you will need to use the split body command and then remove the piece you split off. Do that as the very last step.

Model the passages with solid lofts for both the inside and outside parts.
Use the combine feature as a cut to hollow out the part. Remember to turn off the parts you aren't working on.

Lofts like center lines. I can only suggest trial and error to see what lofts will work. Sharp bends will need more sketches to get through the curve, but the best passages I've designed only use a beginning and end sketch connected by a center line with the right curve.

Lohring Miller

[Muhr]

I actually think you are overthinking or overdeveloping, just make sure no air can be trapped and you´re good to go.
That´s what i have done in several carengines that i have tuned to about 1000hp, they have never shown any heatproblems even though materialthickness in heads were very thin after my portjob.
And they run closer to boiling and a lot hotter exhaustgasses, sometimes above 1000c, never been any problems.
But if air is trapped somewhere i have seen engineblocks crack due to steam shockwaves in those airpockets.

You also need to design waterpassages so the water flows evenly around every path.
This is often done with the headgasket.

If i look into my kawasaki cylinder, water inlet and outlet are on the same side of the barrel and head, rear against the carb.
Water goes into cylinder above the reedcase, then flows to the front, turns upwards through headgasket and flows rearwards in head to the outlet.
In gasket one can see it´s the same tech as in an car engine gasket.
It restricts flow at rear to control flow so it heads forwards in cylinder.
but,, it has holes in the rear to vent out airbubbles it there are any, just as in an carengine

So, lead flow to every path with some finesse, and you´re good.

No need to think of boundarylayer etc etc boiling.
To minimize that if you are afraid of it you can add some additive in your coolant to lower surfacetension, the water becomes 'wetter'.
It actually works fine with soap
But just a couple of drops if you don´t like suds

I took the opportunity to ventilate the issue to a close friend who is one of the people in charge of engine block at Volvo and he estimated that about 70% of the total development time, is flow optimising on cooling to prevent film boiling.

I think you have a lot to learn the automotive industry!

http://publications.lib.chalmers.se/...401/207401.pdf

[TZ350]

cooling to prevent film boiling.

When I was doing a lot of engine reconditioning work we used to get engines with wet liners that had pin holes right through them.

The culprit seemed to be localized cavitation, a bubble would form on the liner surface then collapse. This happened often enough in the same place that the resulting water hammer against the liner would eventually eroded a hole through it.

I thought it was localized boiling but it is apparently a result of cylinder wall flexing.

https://www.cumminsfiltration.com/si...s/3300963A.pdf

http://www.emestlab.com/lab/Portals/...rs/KOKA001.pdf

A blurb that might be interesting about Evans water-less coolants, no local boiling.

http://www.evanscoolants-uae.com/tec...no-problems/72

[Haufen]

Does anyone have any tips or experience regarding the design such as wall thickness or the like?

I guess so. A bit more information on your cylinder would be helpful. Displacement? Bore? Performance / BMEP you aim at? From the pictures, it looks about right.
Personally, when designing a new cylinder, I'd never start with the final exhaust port timing right away as it can easily be too much already to start with. Adjusting in small steps on the dyno makes it much easier to hit the sweet spot. And you can learn a lot from it.

I will try to clarify what I'm aiming for.
If we think that the water rises from the crankcase quite close to the cylinder wall, then the flow of water will take the easiest way to the cylinder head, which means you will get a slightly reduced flow agents the exhaust flange. If I add cooling fins or do some small cavities between the bolt patterns of the flange then I wondered if the even more reduced flow along with the increased specific surface could cause it to start locally film boiling.
which would lead to a deterioration of cooling effect in that area.
And would mean less heat transported from the heat source and then a theoretically lower water temperature but less cooling at load peaks.

I am not saying that I am right but that was my question.


They think of such things I suppose

Nowadays, gasoline car engines often come with a cooled exhaust gas passage. This not only speeds up engine warm up, it also allows them to keep exhaust gas temperatures low enough to run lambda 1 within a wider area in the engine map. Of course, this needs some serious simulation work before it can hit the market.



Nevertheless, comparing car engines with two-stroke engines in regards of boiling risk is comparing apples with oranges. On car engines, the coolant enters the engine with 90°C and exits with an average temperature of 110°C or so, whereas on a two-stroke you enter with 40°C and if you exit with 60°C that would already be quite much. So the delta temperature between average temperature of the coolant and the temperature needed for local film boiling is much much larger on the two-stroke, hence this problem is much less likely to occur. If you just take care that there are no areas of dead flow and install a cooling system which allows you to run at two-stroke ideal (low) coolant temperarures, you should be fine.

On the other hand, as you are designing your cylinder yourself, you can already implement a nice coolant flow path through your cylinder which keeps every area nicely cooled and has no dead areas. On the RSA, part of this is achieved through coolant openings of varying sizes in the cylinder head in order to direct the bulk flow where you want it to go. And there are tiny holes to allow for venting of bubbles in the direct (mostly blocked) part (as described by SwePatrick).

[Muhr]

I guess so. A bit more information on your cylinder would be helpful. Displacement? Bore? Performance / BMEP you aim at? From the pictures, it looks about right.
Personally, when designing a new cylinder, I'd never start with the final exhaust port timing right away as it can easily be too much already to start with. Adjusting in small steps on the dyno makes it much easier to hit the sweet spot. And you can learn a lot from it.

Thank you! Who does not want a hand on his shoulder
it is a 40 borr/stroke
50,265cc
BMEP 205.7 psi

i am 15% under calculated area on exhaust .

[Muhr]

On the other hand, as you are designing your cylinder yourself, you can already implement a nice coolant flow path through your cylinder which keeps every area nicely cooled and has no dead areas. On the RSA, part of this is achieved through coolant openings of varying sizes in the cylinder head in order to direct the bulk flow where you want it to go

Yes, that was a bit so I thought, why not do well when you can.
I will not be able to make full sim on the flow before the summer just static

[Muhr]

I'm also learning Fusion 360 to draw 2 stroke passages. What I've learned so far:

Don't use surface models (Patch) unless a solid model won't work.
If you use surfaces for passages you will need to use the split body command and then remove the piece you split off. Do that as the very last step.

Model the passages with solid lofts for both the inside and outside parts.
Use the combine feature as a cut to hollow out the part. Remember to turn off the parts you aren't working on.

Lofts like center lines. I can only suggest trial and error to see what lofts will work. Sharp bends will need more sketches to get through the curve, but the best passages I've designed only use a beginning and end sketch connected by a center line with the right curve.

Lohring Miller

Loft i great for exhaust design! you may need to do some 3D sketching witch can be a bit tricky. For transfer ports i recommend using revolving! get a straight over and underside with the correct angles whit loft can be a nightmare.
expect a lot of error 40

[Frits Overmars]

engmod.. may show benefits beyond .8 .... I'm speculating this would be a product of non uniform velocity distribution in the nozzle.... .8 mach may well be locally choked already.

Correct.

#1 question is the inevitable mass flow limitation at nozzle choke.

It's my pleasure to get that # 1 question out of the way then: there is no mass flow limitation. When the ratio between upstream pressure and downstream pressure reaches a value of about 2, flow velocity will reach Mach 1. Lowering the downstream pressure any further will not raise the flow velocity any further.
Raising the upstream pressure won't raise the flow velocity any further either, but it will increase the density of the flowing medium, so even when the flow velocity remains the same, increasing the upstream pressure will increase the mass flow.
The picture with the nice colours below shows how at the beginning of exhaust port opening, the flow tries to cling to the exhaust duct ceiling, using it as a nozzle. The exhaust duct floor is too far away to help create a full nozzle, but that temporary half-nozzle at the ceiling is worth thinking about. How can we encourage that?
That's why I'm in favour of a radiused top edge on the exhaust port and a radiused piston top edge. It may not prevent a whole lot of energy loss, but every bit helps.

[190mech]

" That's why I'm in favour of a radiused top edge on the exhaust port and a radiused piston top edge. It may not prevent a whole lot of energy loss, but every bit helps."

Frits,Would not a radius on the piston edge create a void in the squish area promoting detonation?Would a custom combustion chamber need to be made to match the piston edge at the exhaust side?I'm sure we'll need to maintain a sharp edge on the piston dome at the transfer ducts to promote piston cooling also..

[katinas]

" That's why I'm in favour of a radiused top edge on the exhaust port and a radiused piston top edge. It may not prevent a whole lot of energy loss, but every bit helps."

I'm sure we'll need to maintain a sharp edge on the piston dome at the transfer ducts to promote piston cooling also..

and maybe sharp edges on the piston dome and on transfers windows help to resist little, when exhaust gas enter to transfers.