gtbph
Veteran Member
Joined: August 2013
Posts: 101
|
Post by gtbph on Jun 12, 2015 12:26:45 GMT -5
Hi Steve, Wow, nice model, and very impressive machining! The surface looks perfectly smooth! Did you use a cylindrical mill to cut this? I think I will have to use a ball nose cutter because some surfaces are slightly concave, I wonder if the finish will be good enough. I've seen your mill on photobucket, very nice work! Mine is about the same size I think. I envy you still having the hand wheels on it, I removed mine, and miss them to quickly make simple parts. The airfoil generator is awesome! That's exactly what I need, it's nice that the coordinates can be downloaded. The book looks very interesting and detailed, I'll take a deep look at it. Thank you very much for your help, Alain
|
|
|
Post by racket on Jun 12, 2015 16:28:03 GMT -5
Hi Alain
The wheel feels like its made of high tensile steel with a "tinned" coating as its worn off in places and theres some "rust" on the bare spots , the blading is very thin at the tips.
Cheers John
|
|
rythmnbls
Veteran Member
Joined: August 2011
Posts: 145
|
Post by rythmnbls on Jun 13, 2015 9:25:27 GMT -5
Hi Alain, Thanks ! I used 3 passes to machine the profile, a rough pass with a cylindrical endmill to rough out the shape, then two passes with a ball endmill to finish. There's a clip on U-tube showing part of the roughing cut here. Its not very clear, typical amateur cell phone photography Thanks again , I call it an X2 franken mill , Its a lot more rigid with the new frame. I thought I would still need the hand wheels as well, but after I got a pendant for it I've not used them. Yep, I especially liked that they published the equations at the bottom the page. Made it much easier to add to the spreadsheet. Regards, Steve.
|
|
|
Post by britishrocket on Jun 13, 2015 21:06:32 GMT -5
Hello Alain, Your work so far is really impressive. I'm particularly interested in the swirler you made and the tooling you produced to do this. Swirl injectors are a subject particularly close to my heart, and something I have done a lot of work on for my rocket engine project. I decided to cheat a little bit by using metric trapezoidal threading tooling to make my swirlers. A smaller angle of the slot compared to the normal axis increases the swirl velocity in a shorter space and contributes to shorter break up length and smaller droplets. Here is a photograph of a triple start swirler I machined using a metric trapezoidal threading insert:- I have a few posts on swirl injection on my rocket engine build thread here, as well as on my blog at www.britishreactionresearch.blogspot.co.uk , should this be of interest to you. I'll keep watching your progress as I am learning a lot from you. Regards, Carl.
|
|
gtbph
Veteran Member
Joined: August 2013
Posts: 101
|
Post by gtbph on Jun 14, 2015 10:58:34 GMT -5
Hi John, Oh that's interesting, thanks. When thinking about it, steel could be a good alternative material to aluminium or titanium. Aluminium has the fatigue problems that Chris mentioned, and I heard titanium is extremely difficult to machine. Steel does not have these problems, I read the strenght/weight ratio of steel and aluminium are about the same. Titanium would be the most capable material I think, but difficult to obtain and machine, and expensive.
What would be very interesting for me, is the profile of the baldes. You said it is very thin at the tips, does it have a sharp leading edge? Where is the maximum thickness?
Cheers, Alain
|
|
gtbph
Veteran Member
Joined: August 2013
Posts: 101
|
Post by gtbph on Jun 14, 2015 11:00:15 GMT -5
Hi Steve, "Franken mill" This pendant thing is great! I'd love to have one, but it would be quite difficult to interface it with my CNC system, because the program is self-made and running on an old laptop with DOS. But maybe I'll upgrade to LinuxCNC too, one day. Thanks for the video link. I was thinking that maybe you had used a cylindrical endmill parallel to the long side of the piece, to get such a good finish. Regards, Alain
|
|
gtbph
Veteran Member
Joined: August 2013
Posts: 101
|
Post by gtbph on Jun 14, 2015 11:04:22 GMT -5
Hi Carl, Thanks! I just enjoyed reading your blog posts from spring '13, I really liked the thorough scientific investigation on the droplet sizes and spray patterns! Very nice work, I'll have to re-read it again, the amount of informations is overwhelming. My nozzle is far from finished, I'll use a smaller slot angle in the next try, thanks for pointing it out. I'll continue with the nozzle tests when I have a high pressure pump. The slot cutter tool works ok like it is, but it produces burrs. I think it needs additional sharpening to cut cleanly. The red area will be removed before it is used the next time, maybe it will cut a bit better: On the other hand, for smaller-angled slots the lathe works perfectly well as I can see on the photo, so maybe I'll also use the lathe next time. Regards, Alain
|
|
|
Post by britishrocket on Jun 14, 2015 14:42:46 GMT -5
Hello Alain,
Thanks for looking at my blog, I am glad you found something of use. Initially I'd thought about using commercially available atomising nozzles in my design, like those used in oil burners. Like you, I thought that these were just too big and not at all easy to integrate into my design. Having then decided to make my own, it became important to try to gain a better understanding of the major controlling factors governing swirl atomisation. I started further back with shear atomisation, which led me to swirl.
In a lot of my early swirl trials I used coarse threaded wood screws as the inducers, to quickly get a feel for the problem. I knew that to get more control I'd have to make my own, and that is where the trapezoidal thread tooling came in. I was also inspired to do this by seeing some swirl inducers in a Russian paper on the subject.
The book "Atomisation and Sprays", by Arthur Lefebvre, is a definite must for understanding these devices. It is quite expensive, but there is a more sanely priced Kindle version.
The cutter you made for generating the swirler slots is a fantastic thing. I enjoy reading about your turbine exploits. I lurk around on the turbine pages here as someday I will want to build a turbine driven centrifugal pump.
|
|
|
Post by racket on Jun 14, 2015 21:43:23 GMT -5
Hi Alain I dug the wheel out and took some pics Leading and trailing edges are "sharp" , max thickness at tip is 1.2mm , this increases to only ~2-2.5mm at root ( bit hard to measure) , theres very little fillet at root , maybe a 2mm radius at most I can provide more measurements/angles if you need them :-) Cheers John
|
|
rythmnbls
Veteran Member
Joined: August 2011
Posts: 145
|
Post by rythmnbls on Jun 15, 2015 8:57:04 GMT -5
That's a beautiful looking comp wheel Racket, the blades remind me of those big bypass fan blades on the RR Trent engines.
Any ideas from looking at it on how it was made?
Regards,
Steve.
|
|
gtbph
Veteran Member
Joined: August 2013
Posts: 101
|
Post by gtbph on Jun 15, 2015 17:29:23 GMT -5
Hello, Thanks for the detailed pictures John! Beautiful indeed. Amazing how thin the blades are, I wonder if it's forged maybe? Because the forces must be huge at the root... And I'm baffled by the sharp turning of the air at the blade root. The tip profile is also very interesting, it looks like there is just a tiny bit of turning at about 3/4 of the blade. I found some interesting papers: This one is about optimizing an existing design by different stacking of the blades, with nice pictures and interesting references: Three-dimensional multi objective design optimization of a transonic compressor rotorAnd this one is especially interesting, it's the description of the original rotor being optimized in the previous paper, and it includes the blade manufacturing coordinates for four different rotors and their stators: Design and Overall Performance of Four Highly Loaded, High-Speed Inlet Stages for an Advanced High-Pressure-Ratio Core CompressorI'll try to use these coordinates for the profile of my wheel. This is a new model with 1.5mm fillet approximations added, to get an idea. The rest is exactly the same as before. It's an approximation because the fillets were added only in the direction perpendicular to the chord arc, and it assumes the blade walls are perpendicular to the hub, which is not exactly the case. At the moment, there is no room for larger fillets, if this is not enough, the wheel must be made bigger or with fewer blades. I didn't have the time to read more on this yet. Alain
|
|
|
Post by racket on Jun 16, 2015 1:03:10 GMT -5
Hi Alain
Good observation on the tip shape, I always thought it was straight until you mentioned it , so I laid a straight edge along it, and yes, theres an ~5 degree angle change from about that 3/4 distance moving the trailing end towards the preceding vane .
I still can't workout how the wheel has been constructed , its very complicated with both internal and external threads , external splines , under race lube with what looks like the raceway ground directly on the shaft stub and then those very thin blades which would be difficult to machine from solid .
The wheel is magnetic , so lotsa ferrous material ...............maybe chrome molly.
The wheels has the potential for a flow of ~10 lbs/sec , I was going to run it using a 4th stage Allison C20 freepower turbine wheel , direct coupled , but with only ~150 hp to drive the T700 comp wheel, rpm would have been reduced a bit along with the pressure ratio , I was planning on a 1.3 : 1 PR to maximise flow for the available horsepower.
Cheers John
|
|
gtbph
Veteran Member
Joined: August 2013
Posts: 101
|
Post by gtbph on Jun 16, 2015 14:26:10 GMT -5
Hi John, Emm, I know very little about DIY turbine history, but I would bet this would be the most powerful DIY turbine ever made! Something like 1200 N / 270 lbf ? Cheers, Alain
|
|
gtbph
Veteran Member
Joined: August 2013
Posts: 101
|
Post by gtbph on Jun 16, 2015 14:32:43 GMT -5
Hi Carl, Yes it would be nice to have small nozzles, I will need at least six of them for the annular combustion chamber. This is a sketch of the nozzle as it is now, "A." means area, "D." diameter (Millimeter): The next version will have a smaller slot angle, as you suggested. What do you think about the area ratios, would it for example be good to reduce the slot area, or maybe the bypass hole area? Any advice or comment would be highly welcome. Regards, Alain
|
|
|
Post by racket on Jun 16, 2015 17:00:17 GMT -5
Hi Alain
I was planning on using an afterburner as well , so we can roughly double that thrust , I was looking at ~520 lbs - 2300N :-)
Too many projects , not enough time or energy to do them :-(
Have you looked at "Fogging Nozzles" , there are some very small ones available
Cheers John
|
|