|
Post by madpatty on Oct 26, 2020 21:37:37 GMT -5
Hi Patty, I think you may have a few issues with the turbine bearing region. First, it looks to me (as best as I can estimated from the few pictures) that your bolting pocket that you have machined in there is exposed to the hot gas flow. This is not good since any "cooling" jets that you try to introduce will be totally insufficient to fill the pocket. Usually, you try to have the jets introduced behind the backface of the turbine wheel, with the idea that the cooling gas will try to purge the space behind the turbine wheel and not allow hot gas to get ingested behind the wheel. Here is my attempt to see how the bottom of the scallop between turbine blades matches up with the bolting pocket...I know they are not in the same direction and there is a bit of perspective error, but I think it scales pretty close. If this is correct, then your hot gas flow has direct line of sight to the pocket and hot gas will just get churned up in there. The cooling jets that you have in there will actually act to pull hot gas into the pocket by driving local vorticity. If this is an accurate representation of the geometry that you have in your engine, you need to do something to fill this pocket. And here is an attempt to draw in what I am thinking. The second major issue that I can see is that you have a giant, easy conduction path straight to the outer race of the bearing. The flange on the center tube acts like a giant fin to pull heat from the very solid and thick NGV plate. The bolts themselves are bathed in hot gas on the one side and then they conduct that heat straight to the center tube flange, right above the bearing. Turbochargers have heatshields as a means to control the ability of the hot gas path thermal energy from heating up the aft bearing, which would coke up without it. Ball bearings can coke up too and this varnish begins to act as an insulative layer, impeding the bearings ability to shed the built up thermal energy into the oil. I hope this makes clear why I was asking about the scallop and the turbine picture. If this is the case, then we can start to think about ways to thermally isolate the turbine bearing. Have you seen any oil staining on the NGV plate or any other signs to lead you to believe that the total loss oil system is discharging properly out the aft end? Also, be careful reading into the heat stains on the front diffuser structure...you melted a bearing there and the thermal staining could be due to that only. Good luck, Chris Hi Chris. The outer diameter of the bolt pocket is 85mm and approximate OD of turbine scallop is 65mm so there's approximately 10mm radial width for gases to get into that packet. We have to start looking for ways to thermally insulate the turbine bearing. Any/All ideas are welcome. I have a few pictures of NGV plate which I think show oil stains MAYBE- Regards. Patty
|
|
|
Post by madpatty on Oct 26, 2020 22:04:51 GMT -5
Hi racket and finiteparts . Another gut feeling at this point is about the fact that bearings fail too rapidly in my case for any of the above mentioned factors to take effect. I have had bearings failing almost within seconds every time and shaft came out without any coloration at all. See pics- Well at this point I can't be sure if it's heat at turbine bearing ALONE thats causing bearing to fail every-time, BUT i will expect to run at-least to something above idle before any substantial heat effects start deteriorating the bearing. To me it appears its more of a misalignment or bent shaft thing that maybe causing it because every-time before failing the vibrations/gyrations have risen and then a sudden lockup. I am not aware the amount of load the bent shaft(as from dial gauge readings above that I posted yesterday) will pose on the bearing. Like Racket got his engine running to idle and had most of bearing failures near 50k rpm due to either preload, heat or skidding problems. Regards. Patty
|
|
|
Post by racket on Oct 26, 2020 22:27:23 GMT -5
Hi Chris My 12/118 NGV imgbb.com/Cnc2m6X has a 75 mm dia pocket with the ~60 mm turb wheel hub , the bleed air holes are angled at probably 45 degrees radially so that 6 of them impact the screw head and forces the air to hopefully flow a bit circumferentially ?? around the pocket and cool it , whilst the other 6 holes are directed at the wheel boss/hub to cool it . Despite there being an ~7 mm radial "gap" between hub and pocket the bleed air appears to be doing its job of keeping things cool , there virtually no heat colouration on the shaft. I once tried to machine a 65 mm dia "lip" on the 75mm pocket to act as a "heat shield" but I feel good bleed air supply is a better option as it'll also act as a bit of a "blanket" over the bottom of the wheels passageways insulating the hub a tad and cooling the blade roots. My current thinking is that with the new diffuser design there is swirl exiting the diffuser as there was with my similarly diffusered design 10/98 engine jetandturbineowners.proboards.com/attachment/download/9 , the swirl exiting the diffuser prohibits radial flow down the back of the diffuser without a pressure drop and a velocity increase . That potential static pressure drop at the bearing bleed air holes means no entry of air as the pressure at the hot end is probably greater , also the lack of air travelling along the OD of the shaft tunnel limits any cooling of it or the NGV wall at the other end , theres also no excess static pressure to force air through the NGV to cool the turbine pocket and bearing The simple addition of the deflector and radial straightening vanes meant the difference between this jetandturbineowners.proboards.com/attachment/download/25 and this jetandturbineowners.proboards.com/attachment/download/44 Cheers John
|
|
|
Post by racket on Oct 26, 2020 22:35:03 GMT -5
Hi Patty
Agreed , a few seconds running even with no lube shouldn't cause failure at such low rpm , thats why I suggested in my email of the 24th that you check the shaft tunnel for trueness .
Also checking "endfloat" and doing the various maths for axial expansion of the shaft tunnel .
Cheers John
|
|
|
Post by turboron on Oct 27, 2020 7:49:09 GMT -5
Patty, I still think you have rotor dynamics/natural frequency/critical speed problem. You said "because every-time before failing the vibrations/gyration have risen and then a sudden lockup.". As I said in an earlier post this is exactly what happen in the locked rotor test as we tried to go through the critical.
Also, I ran an engine test where we turned the oil off during a power run. The engine ran for three or four minutes before the vibration started to climb and we shut it down. The thrust bearing balls were about 25% of their original size.
Thanks, Ron
|
|
|
Post by madpatty on Oct 27, 2020 8:11:49 GMT -5
Patty, I still think you have rotor dynamics/natural frequency/critical speed problem. You said "because every-time before failing the vibrations/gyration have risen and then a sudden lockup.". As I said in an earlier post this is exactly what happen in the locked rotor test as we tried to go through the critical. Also, I ran an engine test where we turned the oil off during a power run. The engine ran for three or four minutes before the vibration started to climb and we shut it down. The thrust bearing balls were about 25% of their original size. Thanks, Ron Hi Ron. I am also more inclined towards vibrations causing such a rapid premature failure every-time. BUT regarding the critical speed I am not sure because of the fact that I have used the same exact turbine shaft in the original design engine where it successfully ran for many minutes. I am inclining towards a bad turbine shaft with a bent turbine wheel(the inconel portion only has been bent me thinks). Regards. Patty
|
|
|
Post by turboron on Oct 27, 2020 9:04:58 GMT -5
Patty, the spring rates of the housings have significant effect on the rotor dynamics. If for some reason, you changed the spring rates of the housings it would change the rotor dynamics significantly.
Thanks, Ron
|
|
jetric
Veteran Member
Joined: December 2014
Posts: 132
|
Post by jetric on Oct 27, 2020 10:14:02 GMT -5
Hi Patty, That bearing is way too crowded, you need to remove one ball from each of your bearings so there are 16 balls in each bearing then retest your engine, as you have stated there will be a high level of inter-ball rubbing causing them to lock up against each other, also there is not much gap between the balls in your current 17 ball count to allow for expansion before the balls are hard up against each other. With a 16 ball count there will be enough gap between the balls. I think it was mentioned on the GTBA site that at high rotational speeds full complement bearings have a natural tendency to equalize the gaps between the balls so they never end up touching each other in use (provided you have got enough gaps/clearance between the balls unlike in your 17 ball count bearings), Its worth a try anyway before you get too deep into modifying the engine, Richard S. Hi Guys. Another question that came to mind after I made my own full complement bearings is- How many bearing balls should be used for making a normal bearing to full complement? Should it be the maximum number of balls you can fit? In my normal 6003 bearing, they by default come with a 10 x 3/16” balls and you van fit maximum of 17 balls. But 17 balls ‘feel’ sorta tight. Though bearing rotates freely but it seems there will be lot of inter-ball rubbing. Here is 1 bearing that I test made with full complement of ceramic balls- Thanks. Patty
|
|
ripp
Veteran Member
I'm sorry, I don't speak english, so I torment you (and myself) with a translation program,Sorry
Joined: January 2013
Posts: 231
|
Post by ripp on Oct 28, 2020 2:30:51 GMT -5
Hi Patty, In various forums I have read the following points. first, the distance between the outer diameter of the combustion chamber and the wall should not be too small, second, the outer diameter of the combustion chamber should be larger than the lower edge of the diffuse Patty, how about your turbine? Please post photos of your combustion chamber. I really like the shape of the combustion chamber of the anders ju-02 Cheers Ralph translate.google.at
|
|
|
Post by madpatty on Oct 28, 2020 3:23:26 GMT -5
Hi Patty, In various forums I have read the following points. first, the distance between the outer diameter of the combustion chamber and the wall should not be too small, second, the outer diameter of the combustion chamber should be larger than the lower edge of the diffuse Patty, how about your turbine? Please post photos of your combustion chamber. I really like the shape of the combustion chamber of the anders ju-02 Cheers Ralph translate.google.at Hi Ralph. The diameter of lower edge of the diffuser in my case is 240mm and the OD of the combustion chamber is 250mm, so slightly larger. The outer can is 280mm in diameter. Here are some of the pics for reference- Regards. Patty
|
|
ripp
Veteran Member
I'm sorry, I don't speak english, so I torment you (and myself) with a translation program,Sorry
Joined: January 2013
Posts: 231
|
Post by ripp on Oct 28, 2020 6:46:46 GMT -5
Hi Patty, Your cc-design is strange for me, can you please explain your numbers of data.your cc ID look like too small for me. Please sam Fotos. cheers Ralph
|
|
|
Post by madpatty on Oct 28, 2020 7:50:15 GMT -5
Hi Ralph.
The cc ID is 105mm Id. The shaft tunnel OD is just 43mm. Why do you think it’s too small?
Thanks Patty
|
|
|
Post by racket on Oct 28, 2020 15:58:55 GMT -5
Hi Ralph
I agree with you , from looking at the pics the flow area at the inner "corner" appears to be a bit restrictive, especially as there'll be expansion forward when its heated during operation .
Hi Patty
How does the flow area at the corner compare to the flametube hole areas its feeding , it needs to be larger to compensate for the turning "losses" .
Cheers John
|
|
ripp
Veteran Member
I'm sorry, I don't speak english, so I torment you (and myself) with a translation program,Sorry
Joined: January 2013
Posts: 231
|
Post by ripp on Oct 28, 2020 16:02:36 GMT -5
Hi Ralph I agree with you , from looking at the pics the flow area at the inner "corner" appears to be a bit restrictive, especially as there'll be expansion forward when its heated during operation . Hi Patty How does the flow area at the corner compare to the flametube hole areas its feeding , it needs to be larger to compensate for the turning "losses" . Cheers John Cheers Ralph translate.google.at
|
|
|
Post by madpatty on Oct 28, 2020 19:52:53 GMT -5
Hi Ralph I agree with you , from looking at the pics the flow area at the inner "corner" appears to be a bit restrictive, especially as there'll be expansion forward when its heated during operation . Hi Patty How does the flow area at the corner compare to the flametube hole areas its feeding , it needs to be larger to compensate for the turning "losses" . Cheers John Hi Racket and Ralph. That thing was in mind when I designed the combustion chamber. If I remember correctly the area at inner corner was 3-4 times the total area of holes in the inner flametube. BUT i agree. There are really sharp turns for air otherwise wanting to go straight from the diffuser exit. It definitely needs an air deflection system it seems. Regards Patty
|
|