TV94 Turboshaft Project

[ashpowers]

Hi John,

Yes, I made a small mandrel that fits into the inner race bore with a small shoulder so it only slips into the i.d. about 10mm. I put this mandrel into the vice with the bearing about 1/4" above the top of the vice clamps. I then took a small propane torch to the outer race and spun the bearing with the other hand. Continued heating this way until the outer race "dropped" onto the top of the vice... then allowed everything to air cool on its own.

Reassembly was a bit trickier - I ended up making another mandrel to insert into the inner raceway that ran the full length of the inner race - with a nice fit - not interference fit, but a snug fit you could insert it by hand. I then put the inner race with this shaft into the freezer, and once it was nice and cold I packed all of the balls between the races. I have an electric stovetop that has a pyrex glass surface above the heating coils so I put the bearing one and turned up the heat. Only the outer raceway touches the glass heading surface. Holding the mandrel as the outer raceway heated up I gently rotated the inner race back and forth until the outer race expanded and the balls and inner raceway dropped into position. I slid the bearing over off of the heating zone and let it air cool.

Despite all of this and the high quality RR bearings I have obtained, I dont think I will end up using rolling element bearings in this motor, LOL.

I spent a good deal of time this weekend researching air foil bearings, put together a nice informative email about the project, and sent out several emails to companies that manufacture foil bearings. They look a lot more promising than rolling element bearings.

Fingers are crossed - hopefully I'll have some more luck and find a company that will "hook me up".

[racket]

Hi Ash

Thanks for the info on the bearings , I might just have to give that a try with some angular contact bearings , I've got a lot of these used SiN balls from the failed hybrid deep groove ones I made up, they're still in good shape despite "eating" the steel raceways .

Air foil bearings ...............now you're getting technical ....................theres more info coming out about their use in turbo machinery , but I gather the thrust bearings can be a bit "problematic" .

Looking forward to hearing more :-)

Cheers
John

[ashpowers]

The angular contact hybrids I assembled use a 40-degree contact angle. From what I recall, in the model RC engines they tend to typically use 15 degree contact angles. Since the contact angle affects the relationship between the bearing's axial load carrying capacity vs. its radial capacity and seeing that the RC guys have settled on lower contact angles, I'm not so sure these bearings I've made would be properly suited. It is all dependent on the shaft dynamics, which is dependent on geometry, balance accuracy, and bearing damping systems.... and analyzed across the RPM range... and dependent on axial loading and bearing preloading.. The process is just one enormous mess for us as we have to make some assumptions along the way.

Using Air foil bearings has already been done in both a Holset turbocharger as well as a gas turbine engine very similar in design to your FM.
Turbo:
www.miti.cc/newsletters/06_oilfree_turbocharger_gas_engine_applications.pdf

Turbojet:
www.youtube.com/watch?v=BQzh2lh2zPs

In that turbocharger you can clearly see the use of an axial air bearing between the radial support bearings. Integrating a system like that into my engine would be possible - got lots of room around the bearing housing and can easily split the housing into two parts to accommodate the thrust bearing. This system would require no lubrication so that eliminates a lot of hardware and carefully refined delivery system components. I would also be inclined to believe that the need for the labyrinth seal behind the comp wheel would be unnecessary. A properly sized air thrust bearing would be capable of countering the axial loads.. or perhaps leave the laby seal in place and allow the leaked air to pass through the bearing tube and dump behind the turbine wheel into that lower P3.5 area... Keep the foil bearings cooler as well as dump some cooling air onto the turbine... These air bearings will operate at ridiculous temperatures and theoretically, infinite speeds. Additionally, there are dry film lubricants available for the contact surfaces which will withstand temps to 1500F.

AND, with this arrangement it will allow for a shaft "body" of larger diameter. This component would have both radial bearing races as well as the thrust bearing rotor. It would be hollow with a bore diameter matching the turbine shaft - and would run the full span of distance between the compressor's rear seat and the turbine's shaft base. This would significantly improve rotordynamics - raising the critical speeds likely moving at least one band beyond the RPM scope of the engine as well as reducing mode amplitudes. Additionally, the larger diameter "center shaft" would provide more surface area for the bearings as well as provide greater tangential velocities - producing shaft liftoff at lower speeds and offering greater hydrodynamic support.

I got a reply back from DynaTech corp regarding my request for the trial version of their iSTRDYN rotordynamic software app - that email I sent to them also laid out some details of the project.. They offered to do a "pro-bono" pricing to help out with the project and do the analysis for me, but what they do and what disposable financial resources I have are a universe apart from each other. I sent back a rather length reply going into a good bit of detail about myself, the turbines I've built, what this engine is for,... hope they will understand that this hobby serves no interest for profit and consider my offer of a 3rd party review and some advertising real estate on the side of the VTOL aircraft for when it makes cover of Popular Mechanics. =)

[ashpowers]

Radial NGV completed. =) Outer transfer ring and topring w/axial pre-swirlers to come shortly. =)

[ashpowers]

And..... finished the new CAD system this afternoon:
Intel Core i7 2700K 3.5GHz Native, Quad-Core Processor
Asus P8P67 Socket 1155 Motherboard
8GB (2X4GB) 2200MHz DDR3 memory
PCIEX NVidia Quadro 2000D dual head video w/1GB memory
OCZ Vertex3 120GB SATA3 Solid-State Drive
Cooler Master 750W Power Supply
Windows 7 Pro.
Shown here with wireless MS KB, MS USB Mouse, Hewlett-Packard W19e 19" TFT monitor.
Custom Built Enclosure - all aluminum, w/0.093" polycoarbonate panels. A couple of brushed stainless handles, a momentary nitrous oxide power turn-on switch, and we're rocking the CAD modeling in RealView now =).
Windows Experience Rating (avg): 7.4
SSD Bandwidth: 535MB/s Read, 500MB/s Write.

Will be purchasing a large widescreen 3D monitor to complement this powerhouse. =) REAL 3D CAD modeling coming soon!

[Johansson]

That NGV is a work of art!

[ashpowers]

Got a few more pics of the NGV on Monday when I received the part. It certainly is one of the coolest looking pieces in the motor. =)

Only problem is, when they machined the mounting holes they got real sloppy with them and I am seeing as much as 0.030" variation between hole centers, which unfortunately is unacceptable. Obviously I'm a bit upset about this - hopefully this company that made the part for me wont jerk me around - I've sent pictures of the measurements and the screwups and have requested that a new part be machined. =(







A local shop is procuring the 8.25" round 316SS billet to produce the outer ring and top ring from. I have pulled the manufacture of these parts from the company that made the NGV ring, for obvious reasons. Actually ended up getting this billet for $140USD, which is being laser cut at the 3.34" radius so I can produce both of these parts from a single piece of stock.

A bit of a disappointment but I'll work through it....

[dennis]

Hi Ash,

did they CAD-CAM manufacture it without drawing or did you send them a drawing?

Did you give them specific tolerances for the holes?
If not they might be in the clear unfortunately, according to norm-tolerances (or whatever it's called in english)
I hope you will get a new one for free or at least with a substantial discount.

Keep up the good work, I like you pictures

Cheers,
Dennis

[ashpowers]

Hi Dennis,

I sent them the CAD drawing of the part. I stated that I would like the part to within a thousandth but that 3-5 thou would still work. I'm still waiting a response back to my email. Not looking good. =(

[racket]

Hi Ash

What a bummer :-( ...................such a beautiful bit of equipment ....................it always pisses me off when I get a "professional??" to do a job for me and they stuff it up ,hope you get it resolved.

Cheers
John

[dennis]

Oh 0.001" to 0.003" - 0.005" are very tight...
Anyway if they agreed they should deliver.

It looks as if there was a miscommunication on their part and the machinist just used ANSI B4.1 Standard Tolerances as guideline...





Almost the same thing happened to one of my co-workers and in the end he was screwed because only instructions written on the drawing are legally binding here

[ashpowers]

If it comes to it, I can make the part work - I will just have to drill the mounting holes in the baseplate and topring using this part as the template.

What I dont understand is how they can machine the vanes and basering so accurately and then completely botch the mounting holes. The vanes are all within a couple of thousandths face to face, edge to edge, height, etc etc..

Even if they want to dig in somehow that the specs weren't on the drawing itself, one of the vanes it is clear that they breached the sidewall when drilling the hole and then they welded it in and ground it back to shape by hand. You can also see where a mounting hole was begun and then the bore was interrupted and the bit re-positioned. This was the same mounting hole that breached the sidewall. That is not CNC machine work, that is botchy repair work for something that should have been scrapped and a new part made. So I do have them on that part of the work - that has nothing to do with tolerances either.

Additionally, they chamfered every edge on the part - some more than others, and the original drawing does not call for chamfering. I filleted the base of the vanes to the ring for strength and all other edges were not to be chamfered as there are other parts mating to this piece and I wanted crisp edges for those mating faces.

[racket]

Hi Ash

Thats real " backyard" tolerances :-((......... sounds like they gave it to the first year apprentice to hand drill the holes using a centre punch and pedestal drill, not a CNC machine .

Cheers
John

[dennis]

Hi Ash,
I hope you didn't get the feeling that I wanted to defend the machine shop and blame you...
I just know how difficult it is to machine a part like this, especially the holes.
Why they used caution to machine the shape but then completely screwed up making the holes is beyond me though. I hate it when someone delivers more than sub-standard parts, let alone does a totally crooked job

cheers,
Dennis

[turbochris]

That computer is sweet! Unlocked cpu! Get yourself 2 big hard drives and put them in a mirrored raid array, keep the important stuff on it. Had a Vertex go bad on me after 2 yrs. It hurts when gigabytes are lost.....