2 shaft turbine kart build

[racket]

Further "lobsterbacks" were added between diffuser cone and a heavier section ring at the compressor housing end of the delivery duct , there being an ~1/4" gap between delivery duct and comp housing to allow for any alignment difficulties with the flexible coupling ( 3 inches of radiator hose )

Attachments:

[racket]

The finished flametube , it has an overly long "snout" at the outlet end so as to provide a plenum between it and the outer can for the delivery air to feed into and disperse prior to it flowing evenly up the sides of the flametube to the wall holes and evaporators , the "snout only protrudes into the turb scroll by ~half an inch

Attachments:

[racket]

Flametube and outer can ready for assembly onto the turbo

Attachments:

[racket]

Combustor mounted , it has a slight "downhill" slope to the primary end where the combustor drain is located .

The flexible joint between delivery tube and compressor outlet has 3 hose clamps , one on the delivery tube and one on the comp housing , with a third clamp around a band of metal between the other clamps , this is to prevent "clocking" of the housings at high P2 and TITs , a problem I encountered with my turbine bike when using this turbo .

A "Hobbs" oil pressure switch can be seen in the oil line to the turbo , this was interlocked with the Bosch EFI fuel pump to cut power to the fuel pump at oil pressures below 40psi , it proved invaluable when I accidentally deactivated the electric oil pump on one occasion , the fuel supply was cut and the turbo was saved

Attachments:

[racket]

Fuel inlet end of the combustor .

The flametube locating nut can be seen screwed into place on the locating tube with sufficient thread left for the fuel injector mounting fittings to screw on, as well as triangulation struts to support the combustor and turbo at the other end , producing a rigid assembly .

The small chromed cock is for preheat LPG/propane entry , it flows into the securing tube and coaxially around the fuel injector into the evaporators and then to the flametube , the cock is mounted into the original but modified fire extinguisher handle/valve fitting .

Spark plug is position just downstream of the evaporators and provided positive ignition of the preheat gas .

Attachments:

[ernie wrenn]

racket

Nice job.. you answered a lot of questions that were running around my mind. Beautiful work.

Question.. Why are you running the heavier oils? I have very limited knowledge of the turbo jets but with the J-34's and R/R 201>522 we use Automatic transmission oil (ATF). The j-34 holds 5 quarts with the R/R @ 4 quarts.

Is the turbine producing that much heat to thin the oils down? Any adea of the oil temps?

Ernie

[racket]

Hi Ernie

Brass bush bearings in turbos need "normal" engine oil , turbine oil/ATF is for rolling contact bearings and unsuitable , I use 10W-40 synthetic in my 10/98 engine with its brass bush bearings to provide the correct "damping" to control rotor dynamics , they suffer badly from very complicated "modes" and whirl instability unless dampened , one of the reason ball bearing turbos are "crap" according to a very experienced overhaul specialist I've spoken to, it got so bad he wrote to Garrett telling them he wasn't going to stock them anymore , ball bearing turbos only have one good feature, faster spoolup , after that , if you want reliability, go "brass", all the large industrial turbos uses "brass" for their >30,000 hour between overhaul units.

With the kart I had to change from 10W-30 to 15W-50 to keep oil pressures up when it got hot (>60 deg C -140 F ) as I needed the high oil pressures to keep the "fragile" thrust bearing alive .

Oil temperature rise of ~30 deg C going thru turbo , the oil temp gauge picked up oil temps in turbo drain pipe of 80deg C whilst tank temp of 50 deg C , its normal for there to be an 80 deg F rise in oil temps across a turbo .

With my 10/98 engine , it has a "heavy duty" thrust bearing , some 2.5 times greater thrust bearing surface area to comp rear wall area (source of axial thrust from static air pressure loads) than the karts TV84 turbos thrust bearing , so the 10/98 happily copes with running 40psi oil pressure at max rpm/power settings even with oil temps of ~100 deg -210 F

Cheers
John

[ernie wrenn]

Thanks for the info...

I will keep up with your build.

Nice to have solid knowledge on the oils. 80 deg rise across the turbo is a lot of friction. Would a water jacket on the turbo bearing area help reduce temp build up with causing distorsion?

Ernie

Ernie

[racket]

Hi Ernie

A lot of the temp rise is from absorbing heat from the turbine shaft , the turbos with water jackets have then to minimise that heat soakback and reduce the chances of lube coking

Actual differences in frictional losses between ball bearings and brass bushes are minimal at higher rpm , the big difference is at those lower spoolup rpm , ........ when I was playing with hybid bearing in my FM-1 engine , the ceramic balls were so much freer running than steel balls , steel balls have quite "rough" surfaces compared to SiN balls .

I feel that its easier to get rid of the lube heat than try to "cool" the turbo , an oil cooler in the feed line to the turbo is the easiest way , you'll notice I used an automatic transmission cooler on the kart , the turbos with brass bush bearings are designed to use the high lube flow as a cooling agent , there are some large turbos with water cooling passageways in the centre housing , but its generally smaller turbos on autos that have them , probably because of the shorter heat transfer pathways and the higher gas temperatures encountered, big diesel engine turbos generally run substantially cooler gases , so not quite the same requirements , the ones being being used for continuous high power settings would benefit from a water jacket if they were frequently shut down from those high settings without a decent cool down interval at idling conditions before shutting off the engine .

Cheers
John

[racket]

Hi Ernie

Just thought of something else that might be of interest , .....................with my 10/98 engine , due to the initial problems I had with the airflow within the engine , when I pulled the engine apart there was considerable evidence of heat soakback along the shaft , the metal had "purpled" , whereas when I pulled the engine apart after the modifications were made and airflow behaved as per design , there was no sign of heat soakback on the replacement shaft despite a long run time at high power/temp settings , ................I originally incorperated bleed air holes thru the NGV wall to cool the turbine hub , but due to the initial airflow problems they weren't operating, allowing the hub to heat up to the gas temps going thru the turb wheel with heat soakback the result .

The bleed air holes ( 6 X 2mm dia ) did such a good job of cooling the hub that the balance grind marks in the hub were still "silvery" .

Cheers
John

[Johansson]

Time to get busy John, we are waiting!

[racket]

Hi Jamie

Here it is :-)

Cheers
John

[jamiep]

Very cool. I had thought a clutch after the freepower turbine would help performance...I noticed about midway through this thread you thought the same.

I am still on the hunt for an appropriate freepower turbine...probably use the largest hot section off of a turbocharger I can find.

Any reason you didn't just use the scroll that came on your free power wheel? weight?

Tell us a little more about your bypass air just before your free turbine...is it automatic or manual?

Jamie

[Johansson]

Perhaps a turbine wheel from one of the first stages in an Allison C20 would work? They come up now and then on Ebay and it is most often the 4th stage that ends up expensive.

[racket]

Hi Anders

There was a nice second stage wheel on Ebay last week , $150 ................they're good for ~90 hp in a DIY situation, but because they have only a stub shaft its not as easy to mount as a third or fourth stage wheel with the thru bore .

But they'd be almost indestructible for us as they'd normally see T I T s of ~850 C and 56,000 rpm , way beyond what we'd feed them , the blades would be lightly stressed due to a max flow of ~1.1 lbs -0.5 kgs/sec compared with the 3.6 lbs/sec of the C20

Cheers
John