APS/Ω Gas turbine bike build

[Seldom Seen Slim]

I disassembled the engine after the smoke had cleared and found that the compressor nut had come lose.






OUCH!!

[manta22]

Anders;

Loctite won't provide much retention when its temperature goes high. You might try a prevailing torque locknut such as a "jet nut".

Regards, Neil  Tucson, AZ

[Mobacken Racing]

R&D; it never gets boring.

You are right Neil, the heat from the shaft will soften the Loctite and make it mostly useless. What I need to do is to thoroughly clean the rotor parts from oil before assembly and torque the nut even more, the billet compressor won´t give like a cast one so more torque than the specs state is needed.

A friend has calculated that the rotor at 70.000rpm is trying to loosen the nut with only 30 ft lbs of torque, so it shouldn´t have come loose in the first place if everything was done properly. We´ll have to think about this a bit so it won´t happen again.

[tauruck]

The R&D on this project is off the charts. I'm reading but don't understand yet I enjoy the processes.
If you and Bo ever got together, Lord help us!!!!.

Awesome work and I really hope you get what you're after?. Thanks for making this so interesting.

Mike.

[Interested Observer]

Anders,  probably venturing beyond the scope of my familiarity, but...
Could you in effect key the wheel to the shaft by sinking a longitudinal dowel into the shoulder at the shaft enlargement that would interface with a hole in the wheel?  Or vice-versa. 
Also, the wheel looks like aluminum?  I don’t know what kind of temperatures it gets to but thermal expansion from a previous run may have resulted in a loss of preload.

[Mobacken Racing]

Thanks Mike!

The thing is that the torque acting on the nut trying to unscrew it is only like 30 ft lbs, so a 130° angle torquing of the nut when assembling the rotor should be plenty enough to keep it in place. One theory is that the compressor wheel has touched the alloy cover and unscrewed itself, possibly from some vibration mode plus a too tight blade clearance.

I´ll digest this while waiting for the new compressor wheel and the 24v converter.

Cheers!
/Anders

[maj]

Should the thread be in the other direction ?

[wobblywalrus]

Sometimes a thick titanium washer under a flange nut will make a bolted connection less susceptible to loosening from vibration.  Ti is a springy material.

[Mobacken Racing]

Maj: It is more common that a clockwise turning turbine shaft has a left hand thread, that way the turbine torque only tightens the nut further.

Wobblywalrus: I am discussing this with some guys at the jet community, the tractorpulling guys make keyways to lock the compressors on the shafts and some suggests double nuts that lock against each other.

I am sort of hesitant to adding anything to the rotor since it is designed by Garrett to work with a single nut and a proper torquing down, adding stuff to a rotor that spins up to 75.000rpm feels like begging for more problems.

Cheers!
/Anders

[Mobacken Racing]

I disassembled the rest of the engine tonight, and luckily I found no big surprises.



A bit of soot from the post run flames were expected, but look at this mark on the shaft tunnel. It looks to me like there has been oil leaking past the o-ring seal on the oil inlet. A later check of the seal confirmed that it was damaged.



The bearings look just fine to me, a bit of shine to the pressure pads but they don´t look worn. The same goes for the journal bearings, no problems there.



I wonder why the silicone seal around the shaft tunnel/diffusor plate joint has worn like this?



Some of the welds on the NGV vanes had cracks in them, I will add more filler to the welds to make them sturdier.



Here is the compressor cover, no use trying to interpret the wear marks since there is no way of knowing if any of them are from a rub that loosened the comp nut.



I got the suction oil filter in the mail today, now I just have to wait for the AN10 hose fittings before I can fit it to the suction line.



Cheers!
/Anders

[wobblywalrus]

Anders, close-up pictures of the rotor, top and bottom, where the shaft goes in and out, might be a big help.

[Interested Observer]

Damaged O-ring looks like it got pinched during assembly.  Especially since it doesn’t appear to be containing any appreciable pressure.  Any sign of the missing material?

[Mobacken Racing]

Wobblywalrus: What do you mean?

Interested Observer: Nope, I see the seal when I assemble the core so no risk of pinching it. My theory is that the pump that sucks oil back to the tank from the engine is pulling hot air from the engine past the seal and into the oil gallery, this explains both the damaged seal and why the returning oil is unusually hot.

I don´t know how to solve this though, a spring loaded bleed valve on the drain line is probably the easiest approach. This way I can control how much underpressure the scavenge pump can produce.

Cheers!
/Anders

[charlie101]

Maybe a harder rubber quality would solve it. Or/and make a thin narrow sealing crush ring by turn the sealing surface off a few 100th mm leave a ring at the outer perimeter?

[Mobacken Racing]

I think the problem is that I didn´t mill the seal groove all the way around because of the hole for the axial bearing oil, air is leaking at that point and damage the seal edge that is just next to it.

I can use a Dremel to extend the groove until the ends meet just outside the oil hole, that in combination to a more controlled under pressure should fix the problem.