HC5A Kart build

[racket]

Hi Keith

I'd double the number to 16 at 11mm dia., the greater number will provide more even coverage of the cross section as the holes will be closer together , the fewer large dia holes for the Primary are to get air right into the centre and create the turbulence and recirculation required for the main combustion , the Secondary holes are only for "cleaning up" any remaining fuel before the gases are chilled by the main dilution holes which would make combustion difficult.

Cheers
John

[ausjet]

Made a bit of progress on the flame tube design today


Read through the recent post on swirlers by one of the fellas and had a go at a pattern. I can understand that it's optimal to have the flame anchored in place by a swirler creating an area of turbulence and lower pressure (I think) in the centre, but I don't fully get what effects the different designs will have. Eg if I have the swirler ring close to the injector vs further out, greater number of thinner fins or fewer wider ones. I'm guessing it's similar to the primary zone holes..(fewer large diameter holes penetrate more)
I started off with straight blades to the centre point but changed them to 20degrees offset, thinking that might give more swirl being pointed away from the wall.

[racket]

Hi Keith

You need more of those 2mm wall cooling holes either side of the main Primary holes , probably one every 12 mm in a ring around the wall , you've got a lot of wall to cool .

As for the swirl vanes , you can have the "radial slots" radially , what we are trying to achieve is a sort of doughnut swirl within the primary zone , the fuel spray if a hollow cone starts the process, the main primary holes produce jets of air that collide in the centre of the flametube with air then travelling back up towards the spray nozzle taking flame with it to impact the newly discharged fuel and air entering through your cap swirlers to initiate combustion as early as possible , the flames are closer to the wall than the centre of the flametube , hence the need for the wall cooling as well as a greater number of secondary holes that don't need to discharge as deeply into the fametube .

Hope this helps

Cheers
John

[ausjet]

Cheers John, have put some more cooling holes in, 11.75mm apart, smack on 300mm2
Time to give the drill press a workout

[racket]

Happy drilling :-)

[finiteparts]

Hi Keith,

"Read through the recent post on swirlers by one of the fellas and had a go at a pattern. I can understand that it's optimal to have the flame anchored in place by a swirler creating an area of turbulence and lower pressure (I think) in the centre, but I don't fully get what effects the different designs will have. Eg if I have the swirler ring close to the injector vs further out, greater number of thinner fins or fewer wider ones. I'm guessing it's similar to the primary zone holes..(fewer large diameter holes penetrate more)
I started off with straight blades to the centre point but changed them to 20degrees offset, thinking that might give more swirl being pointed away from the wall. "

Yes, the presence of a swirling flow leads to a region of low pressure in the center. Maybe I can make that more clear to everyone. When a flow turns, it does that because there is a pressure difference. So lets say we have a hockey puck shooting across the ice in a straight line. If we wanted it to turn, we would have to push it with some sideways force, so we smack it with the stick and it veers off at an angle. So if we wanted the path to be a curve, we could imaging giving it a bunch of small taps as it progresses. This is what is happening to the fluid as it zips along...except instead of the stick providing the force to turn it's path, we have pressure fields. So we know that fluids flow from high pressure to low pressure because there is a net positive force, we can see that in order for the flow turn in (and create a rotating path) the high pressure has to be on the outside acting inward on a lower pressure center region. When the rotational momentum of the fluid equals the axial momentum of the fluid, the vortex breaks down and the flow reverses creating the recirculation bubble.

How does the design of the swirler impact this? Well, from above you can see that we are interested in the amount of swirl momentum (which is referred to as the swirl number) but we are also concerned with how much air we are putting in the primary zone. So the larger the diameter, the smaller the slots will be to meet a target airflow. The angle of the swirler will also change because to make the recirculation bubble, you are "trapping" a lot more volume in the bubble for the larger swirler and thus need more swirl energy from the swirler to balance the swirl and axial momentum to form the vortex breakdown.

Usually, a smaller swirler is used, but there are instances when the cooling holes are used to form a vortex in the dome. The primary jets can be used to pump some momentum into the swirling flow to form the torodial vortex. It is just harder to predict.

Just by looking at your pictures, I am curious how much air you are putting in through those slots? I think they would work better closer to the injector, but that is part of the fun in experimenting.

Good luck!

Chris

[ausjet]

Thanks for detailed explanation Chris, makes a bit more sense 😊 the swirlers turned out to be a bit undersized at 600sq mm (6x 3.7mm x 27mm)
Was aiming for 750sq mm.

[CH3NO2]

Aug 12, 2017 13:41:34 GMT -5 finiteparts said:

...

How does the design of the swirler impact this? Well, from above you can see that we are interested in the amount of swirl momentum (which is referred to as the swirl number) but we are also concerned with how much air we are putting in the primary zone. So the larger the diameter, the smaller the slots will be to meet a target airflow. The angle of the swirler will also change because to make the recirculation bubble, you are "trapping" a lot more volume in the bubble for the larger swirler and thus need more swirl energy from the swirler to balance the swirl and axial momentum to form the vortex breakdown.

Usually, a smaller swirler is used, but there are instances when the cooling holes are used to form a vortex in the dome. The primary jets can be used to pump some momentum into the swirling flow to form the torodial vortex. It is just harder to predict.
....

Hi Chris,

Of the numerous methods to create a recirculation bubble in Zone 1, what if the air were injected outwardly-radially from the top of the dome through an annular orifice as shown in (1) of the drawing below?  

Then supplemented by perpendicular air injection at (2) near the circumference of the dome.  Which is then charged once again perpendicularly at inlet area (3).  The swirl energy is supplied from three directions. 

The drawing shows heavy black lines of the flame can walls with blue lines indicating air flow direction.

The fuel nozzle also shown in blue, at the top, is only a hypothetical location.  It can can be moved axially such that the fuel can be sprayed above or even below the plate creating the radial flow orifice.

Tony

[finiteparts]

Hi Tony,

Sure that would probably work well for simple combustor. There were a few papers back in the 1970s or '80s that used fuel impingement on a surface as a means to breakup the liquid into smaller droplets and then having the sharp edge on the plate allowed for further flow shearing of the sheet, basically a simple airblast atomizer. Rolls explored several impingement atomization schemes for use in the RB162, which was a lift jet engine (when they were exploring other avenues to make supersonic VTOL aircraft propulsion systems). They had several schemes that used single jets impinging on scrolls and angled plates in the airflow to produce small droplet clouds. My guess is that they were not that successful, since I haven't heard about them since. I will see if I can find some of those papers.

The things that come to mind to look out for are:

1. The disc has to be attached to the dome somehow and the struts or mounting features might collect liquid fuel causing locally heavy fuel concentrations. You could probably get around this by shaping the struts to be thin and sharp in the radial direction, offering the least amount of wetted surface for the fuel to "stick" to. If the struts are too "big" the droplets will collect and coalesce into larger droplet that might cause combustion issues when they are shed off the strut into the passing flow.

2. The flow through the gap between the disc and the dome should be sized to get as high a flow speed as practical, while making sure that you are getting sufficient primary zone air. You want to impart a high level of kinetic energy to the airflow because that is the mechanism for breaking up the fluid into smaller droplets. It takes energy for a large droplet to break up into smaller droplets, because of surface tension. When you break a large droplet into a bunch of smaller droplets, you increase the surface area between the liquid and the air...this means that for a given amount of liquid, there is a larger overall amount of surface tension. You can provide the energy needed to for the additional surface tension by imparting kinetic or thermal energy to the fluid. Vaporizers do this by providing thermal energy, but to mechanically atomize fluids, you can push the fluid with a pump to high speeds or you can use the energy in high speed airflows to increase the fluids kinetic energy. The higher energy fluid can then be perturbed, which will allow it to break apart into smaller droplets. Keep the edge as sharp as possible to help "break" the liquid sheet and get the air flows on either side introduced to the liquid sheet as rapidly as possible.

3. There will be a stagnation point in the recirculating flow ON the hot side of that disc. That means you will likely hold the flame on the back side of that disc. You might want to bring your injector fairly close to the cold side of that disc so you make sure the fuel cools the disc as effectively as possible. You might even want to use a solid cone injector to make sure the center of the disc receives fuel droplets to cool the metal.

Good luck!

Chris

[ausjet]

Have been searching for a few days on how to pie cut a cone accurately.. couldn't find any useful info. Plenty of stuff on pie cutting straight pipes. Anywhoo had a crack and eyeballed it. Pretty sharp bend but hoping at the same time it will lose its air velocity quickly like this.

[CH3NO2]

A diffuser and elbow all in one. Very cool.

[racket]

Nothing wrong with "lobster backs" , one of my favourite way to produce tight bends ............... a nice thick ring next to the comp housing and a short length of silicone turbo hose will connect things up nicely :-)

[ausjet]

Slowly getting there! This is the new intake. I had to make another one as the old one was too short once fishmouthed to fit the combustor.







Today I finished off the flame tube and slip joint.







Started working on the mounting the igniter (ceramic glow style). Thought to myself, better not drop this. Well... I now have another one on order

[racket]

Looking good :-)

What are those extra holes for at the mounting flange ??

Cheers
John

[Deleted]

Nov 8, 2017 16:07:11 GMT -5 racket said:

Looking good :-)

What are those extra holes for at the mounting flange ??

Cheers
John

Hi, was just think the same thing, love the build :-)

All the best

Andy