FJK3 - 88mm DIY Gas turbine

[nersut]

Hi again Jet people!

I have finally started on my new build using my 88 mm Garrett TV8513 turbocharger as the core, named: FJK3 

As usual every input from you guys is always welcome! 

Step 1.
Compressor scroll was removed & the compressor wheel cleaned & inspected, luckily with no sign of wear or damage.

 flic.kr/p/VNsq9x

Step 2.
I made a turbo flange using a piece of 6 mm stainless plate, I used the gasket as template.
Is 6 mm too thin?

 

Step 3.
I calculated the x-sectional area of the compressor inducer to find the recommended combustor liner diameter, using this formula: Area = (radius² x π)
Compressor wheel inducer area: 6.082 mm² = (44 mm² x π)

The recommended x-sectional area for the combustor liner is 3 times the x-sectional area of the compressor inducer, so: 3 x 6.082 mm²  = 18.246mm²
To find the recommended diameter I used: Diameter = ((√((inducer area x 3) / π)) x 2)
Recommended combustor liner diameter: ((√((6.082mm² x 3) / π)) x 2) = 152 mm

Using a (free 21 day trial) easy to use cone layout program (www.conelayout.com) I made some paper cone templates for the combustor liner to outlet transition area.
Using the dimensions of the single inlet turbine scroll : (59 mm x 88 mm) = 5.192 mm² I calculated the circular outlet diameter, using:  Diameter = (√(turbine scroll inlet area / π) x 2)
Combustor outlet diameter: (√(5.192 mm² / π) x 2) = 81 mm

I canibalized the FJK2 & cut the stainless parts using the calculated transition cone templates.

  

I found a stainless tube which was a bit smaller diameter than the recommended 152 mm, so I cut it & welded a 47 mm strip to get it to the required diameter.
The cone was MIG-welded & a short piece of 80 mm tube (close enough to the calculated 81 mm, I think)

    

The combustor outer casing will be made using an old stainless water fire extinguisher, I could use the end cap to the combustor liner top piece.



Next step will probably be drilling holes for the primary, secondary & tertiary air zones.
I will use the awesome JetSpecs calculator program (www.nuclearprojects.com/jetspecs/jetspecs.exe)

Any recommendations for the hole sizes & locations?

Example:

• 8 mm for the primary zone, 36 holes
• 14 mm for the secondary zone, 8 holes
• 22 mm for the tertiary zone, 8 holes

I think I will lay the combustor down with a 90 degree bend at the turbine inlet to make the overall size of the gas turbine engine more compact.


Best regards
Erik

[racket]

Hi Erik

Nice to see the project getting started :-)

A few comments .............

6mm is a tad thin for large flanges unless lotsa securing bolts used , 10 mm would be a minimum for the scroll foot flange with its 4 bolts , I've used up to 15 mm thick if a large endplate for the combustor so as to take pressure loads as well as remaining flat enough so that the gasket on the scroll doesn't leak, the thickness of the scroll housing rectangular flange at the bolts is a fair indication of the thickness required.

90 degree bends at the combustor discharge aren't the best idea , a straight in approach gives best results , a bend creates turbulence in the relatively high speed gases that can "alter" the available scroll throat area making it "smaller" than its potential flow .

Your turbo will flow roughly the same as my TV84 did ,I used an ~140 mm ID flametube..................are you going to used a spray nozzle for the fuel or evaporators ??

If a spray nozzle , you may need to take into account its spray angle in determining the position of your primary holes , I'm assuming you aren't going to have any other primary "holes" other than the wall holes , if this is the case then I'd be having maybe 8 larger holes of say 12 mm dia to get air deeper into the fuel spray to promote turbulence , some 8 mm holes and plenty of 2 mm wall cooling holes

Cheers
John

[nersut]

Hi again John

Thanks for guiding me, again

I will have to redo the flange with a thicker one then.

I am aware of 90 degree bent is not optimal, but I favor a more compact overall size of the engine (height/width)
I will try not to narrow more than the x-sectional flow area in the bent.
How much negative effect can the 90 degree bent cause?

I am opted to use high pressure furnace fuel injector as with my previous builds with a 80 degree spray angle.

Probably also a secondary injector with lower fuel flow (good atomization) for easier ignition, startup/idle & preheat before the main injector opens.
The secondary injector is almost necessary with a single high flow main fuel injector, because the main injector does not creates a fine spray at low/startup fuel pressure (streams of fuel/big droplets) & causes wet/hot startups and is very difficult to ignite diesel fuel with spark ignition. It can also very easily flood the combustor with pools of fuel. Done that... Very scary & causes uncontrolled fueling with big flames exiting the nozzle & crazy reving of the turbo! 

Here is the sketch of the combustor for the FJK3, now with correctly located air delivery tube & cone shaped. 



I am willing to copy the top piece of the combustor liner in the Allison/RR 250 engine with the holes & flaps to create good mixing of the fuel/air (swirl & vortex).
The Allison/RR 250 have a "similar" setup with tangential air delivery tubes, dome & a single high pressure fuel injector on top.
Pros & cons?




Best regards
Erik

[racket]

Hi Erik

A thicker flange is a good idea , any leakage from a thin flange that warps is just a waste of power .

If you can keep the 90 degree bend with a decent sized radius it'll help minimise any "problems" ............it won't eliminate them , but if compactness is important then thats a compromise that has to be accepted.

Some air entry "above" the fuel spray will help keep the flametubes domed cap a bit cooler , I had a multitude of small holes ( 36 X 3mm dia) in my TV84 FT cap as well as the swirl vanes ( 12 of at 9 X 5 mm flow area) .

Cheers
John

[nersut]

Okey, thanks.

With some help from my work I now have two flanges for the combustor outer casing & bottom piece, 6 mm in thickness.





For clamping the two flanges together I will use bolts & nuts.
What should the spacing between the holes be for good sealing? Is 1½" (~38 mm) too far between?


Best regards
Erik

[racket]

Hi Erik

How does the 38 mm spacing look on paper ?

LOL, I'd probably have "over engineered" and gone for ~25 mm -1" using 6mm screws/nuts to give it the "aerospace" look ;-)

Can you fit a rebate/step arrangement so that the parts are easily aligned concentrically??

Cheers
John

[nersut]

Hi John

Here are two models of the flange, the first one with about 38 mm spacing & 15 holes. The second model is with 25 mm spacing & 23 holes, surely looks more "aerospace" grade



I will go with 25 mm spacing, less likely to cause air leakage.

I didn't really get your question about rebate/step arrangement? Sorry for my bad english.

To minimize troubles caused by the 90 degree bend, I searched online for solutions.





I will go with box type 90 degree elbow with rounded shape. Maybe adding 1-3 guide vane(s) could reduce turbulence & "choking" effects?


Best regards
Erik

[Johansson]

I don't think that uncooled metal flaps inside the combustor outlet is a very good idea, unless they are made out of inconel or MA253 they will most likely burn out over time and possibly damage the turbine wheel.

[racket]

Hi Erik

Take Anders advice ...........no turning vanes please .

By "rebate" I mean machine the "fat" flange face in to a width of the "skinny" flange ID , it only needs to be a millimetre deep cut , the "step" produced also helps with sealing , and if done before drilling the securing holes , both flanges can be drilled at the same time whilst maintaining concentricity .

Cheers
John

[nersut]

Hi Anders & John thank you for the advice/warning, no guide vane(s) then.

Here is the new turbo flange, now in 10 mm thickness & fits nicely.



John, I understand now about the recess cut, I will ask the guy if he could make the lathe work for me again.

Will a v-clamp work well instead of "tons" of bolts & nuts?



I also made a paper template for the square-to-circle elbow. I noticed that the compressor discharge port/neck is relative massive, about 14 mm wall thickness & about 33 mm long.



No, it's not holes on the side of the combutor outer casing, just reflections from the top hole.   



Maybe I should trim off the neck or just some off it in length & thickness (dotted lines). This way i won't need extra 90 degree bends & extra tubing.
Just a strait funnel/cone shaped transition piece instead (extra compact)




Best regards
Erik

[nersut]

A little off topic.

I have another quick question about this exhaust duct for the Pratt & Whitney PT6T engine used in the Bell 212.
I found one at the scrapyard, what materiel is it, stainless steel & more exotic?

It looks exactly like one in the pictures below.
 


Best regards
Erik

[Johansson]

Hi Erik,

A V-band clamp for the lid works just fine, I try to use them wherever I can to avoid the nerve wrecking job of drilling dozens of small holes in stainless.

Cheers!
/Anders

[racket]

Yep , V-bands are great .............work best when having a rebate so that both flanges stay concentric during tightening .

Exhaust duct , probably only need stainless as temps are pretty low by that stage and not requiring "exotica" .

Your comp discharge straight into the combustor looks very neat and tidy .

[nersut]

Hi Anders & John

Sounds nice with v-band clamp solution, easy access to the internals for inspection & possible further modifications.  

The bigger flange for the end cap is now recessed, about 1 mm deep cut.
The smaller flange sits nicely in the groove & there is a a little play for heat expansion, the bigger flange will have direct contact with the combustor liners exhaust pipe.





I also made another flange to use the original v-band clamp for the jet nozzle (turbo outlet).



Step by step, it's slowly going forward with the build.


Best regards
Erik

[nersut]

I have been puzzling with cardboard & masking tape to make templates for the exhaust elbow.





I also copied the hole arrangement from the Rolls Royce/Allison 250 engine for the combustor liner dome, total 54 holes maybe all with 3 mm diameter, this is about 6 % of the inducer area.
I will also put flaps to make mixing/swirling effect & lots of small holes for film cooling near the edge of the dome.

Before I start drilling the holes, should I go with a different diameter?

• 3,0 mm  = 6,3 %
• 3,5 mm = 8,5 %
• 4,0 mm = 11,2 %
• 4,5 mm = 14,1 %
• 5,0 mm = 17,4 %

The three inner rows of holes will produce counter-clockwise swirling effect (viewing from the top) due to the internal flaps & the three outer rows will produce clockwise swirling effect.
The two opposite swirling effects is also copied from the RR/Allison 250 engine, this will produce good mixing of air & fuel.


Best regards
Erik