axial flow compressor home build

[michaelgolden]

Hey guys,

I'm currently living in Alberta, Canada. I'm working on the design of a home built axial flow jet engine. The first step is to successfully build the first stage rotor of the compressor. Before I go ahead with any other part of the engine I'd like to have it balanced at design speed. I'll most likely spin it up past design speed to do some destructive testing. After that the rest of the compressor will follow.

I've done most of the calculations through excel. Moved onto the design through solidworks and have begun making the blades for the first stage.

I was hoping to have the first stage tested and balanced by the end of the year. Although I have been unlucky with the machinists in town. Its mostly oil rig stuff out here. No time for the little guys. I dont have machines of my own yet, and I'm heading to school in a couple months. So the project might be put on hold until next year.

I'll attach some photo's and part of the assembly.

[michaelgolden]

[racket]

Hi

Before you go too far with construction have you thought about your combustion , the Bladon micro axial engine wasn't a success because of the undersized flametube , axial compressors flow a lot of air and unless you use transonic blading at high tip speeds you won't be getting a very high pressure ratio from your 5 stages , ...............with lowish air densities going into the flametube you need a fairly large crossectional area to keep the flame alight , so unless your flametube is considerably larger in diameter than your first stage comp wheel you'll have combustion problems , we need roughly 3 times the crossectional area of the compressor inducer for a flametube when using a radial compressor wheel which will produce a healthy pressure ratio.

Cheers
John

[michaelgolden]

Hey John,
I have given some thought to combustion. The numbers can be adjusted in the calc anytime. I've done some full cycle calculations. The compressor has been given very modest numbers without going transonic. Has a compression ratio of just about 2.2 with a mass flow of 0.8kg/s, under standard atmospheric conditions of 1.013 bar. On a day of 37C. The density entering the combustion chamber is around 1.9kg/m3 with an engine pressure ratio of 1.5.
The compressor tips diameter are 4" with an exit blade height of 13mm and the resulting combustion chamber diameter is just over 5" to give you an idea in the difference in area of my current model.

Its too bad the bladon brothers had issues with their flame, it was a beautiful engine.
The photo up top was pulled from the excel numbers. Length was just an educated guess. But I'm still in the process of figuring out how the flow moves through the holes in the kj-66 and tk-50. But not too concerned about it right now until I get a working compressor.
It cant be accurate though with the current model assuming no loss's. Which is why I see the only real way to build it is to start from the compressor, match my values to the calculations and asses it from there.

Off to work. Cheers!
Mike

[racket]

Hi Mike

With a 4 inch tip and ~3 inch root diameter for the first stage , that'll give an inlet flow area of ~5.5 sq ins , multiply by 3 gives us ~16.5 sq ins for the flametube crossection, assuming a 3 inch ID for the flametube , that'll mean a ~5.5" OD for the FT , allow a gap of say 0.5 inch between FT and outer can and you'll be needing a ~6.5 inch OD can to get air velocities down to reasonable levels to burn kero in .

I hope you're using fairly modest compression efficiency values with the design , small axial wheels need very good blading to work .

Is there any particular reason why you want to go axial rather than radial , the OD of the engine won't be much different, if at all, because of the need for a large diameter flametube for the mass flow and pressure ratios you'll be using

Cheers
John

[michaelgolden]

Hey John,

I've done a few combustion calculations. Its interesting to know how you guys have been doing it. Multiply by 3 seems like an easy rule of thumb and would have saved me a lot of playing around with the numbers. I've been calculating the static values of the flow to yield the chamber cross sectional area. But it requires knowing the velocity of the air through a working combustion chamber.

The efficiencies are typically low. And the actual pressure ratio for each stage only ends up being just a little over 1.1 while keeping everything very modest. And the diffusion factor for the blades have been kept very close too 0.4. But its hard to factor in any drag or other loss's without spooling up a completed compressor.

I find them fascinating. Especially high powered transonic turbofan's. I'm a fan of watching AgentJayZ's in depth videos on axial flow overhauls. Definitely a design challenge. If it does work. There's a ton of potential to be made. The current mathematical design yields a force of 365 newtons/s on the ground during a hot day, burning around 38-40oz/m.

I hope everything goes according to plan. So far, so good. It'll be exciting to see it spool up for the first time. And then break it haha.

Cheers
Mike

[michaelgolden]

Hey guys, little update. Full build progress is being posted on GTBA website. In case anyone is interested, here are some pictures of the project so far.

[racket]

Hi Mike

Thats some beautiful work there , yep , been checking out your build on GTBA :-)

One question , ....... looking at the pics of the inlet, there appears to be an alignment of front supports and comp blades, is this so ??

The reason I ask is if this is so then you might experience "harmonics?" which can cause problems , its probably a good idea to have uneven rather than a synchronized phasing.

That first stage comp wheel will be shifting some air , its kinda large when held in your hand .

You've been busy :-)

Cheers
John

[michaelgolden]

Hey John,

Thanks very much, and regarding your question. I concur its a known problem to have two stages with the same amount of blades one after the other. And having more or less blades with each succeeding stage is one way to avoid the harmful harmonics. But I have never heard or read of aligning or staggering the struts to cause any issues. I would be interested in reading more about it. Unfortunately its too late for this engine so we'll just have to cross our fingers!

I'm blocking some of the total massflow available. But there is about 1/4" gap between the struts and the blades so any undue effects should be negligible.

Its my first attempt and its a huge learning curve. Although everything has been kept very moderate in terms of a 4" inlet. So if it ever does run one day. There is much room for improvement.
Total massflow should be around 0.8 kg/s on a hot day. For a total of 365 newtons of force from the nozzle. Some un-calculated loss has been ignored.

Cheers

[sanokanta]

Can someone post the link for the GTBA Building Progress one...

[racket]

Hi Mike

There have been turbochargers that needed their inlets changed where the struts holding the inducer shroud and map width enhancing slot caused failure of the blading due to phasing, a simple change to the spacing of the struts cured the problems .

You should be OK in the short term.

You will have wake turbulence from your struts as they're not a true aero shape , the trailing edge is a bit rounded and fairly close to the blading so you might have a fair bit of "strange" inlet noise other than the usual sucking sound .

For a first attempt you're doing a magnificent job of it :-)

Cheers
John

[ernie wrenn]

365 Newtons = 82.055285 Pounds force
Newtons Pounds force
1 N = 0.224809 lb
1 lb = 4.44822 N

[michaelgolden]

Thanks John, its a fun project. If I was to start over again I would probably dispense with the struts and go with IGV's. Its really more of a proof of concept for me right now. Once I have most of the design dialed in, it will be easy to make modifications for a mark II if I get that far. Can be lighter and I could really amp up the design speed. Like you say, should be ok for now.

www.gtba.co.uk/

Should be right around 400 newtons or so at 10*C.

Thanks for the interest guys. More posts to come.

[racket]

Hi Mike

LOL...........Mk 11 , 111 and 1V all seem to be a bit better , every time I finish a build I look at it and think ....mmmmmmm, that could be better , then incorporate it in the next build :-)

Yes please , keep those updates coming .

Cheers
John

[michaelgolden]

http://instagram.com/p/zgppzggZ9u

http://instagr.am/p/zoHk5wAZ5s

Here's a link to some pictures and short videos. Posts on jets have been dwindling because I've been spending my time on the 5 axis machine. Its really convenient to post photo's on instagram off the phone. So here you go!