Been a while - new project, T04 Gas Turbine.

[Jason]

Thanks a lot sir. I have made several GTs with old turbo chargers from junkyards, so now i want to try some real stuff. :D

The major problem with my GTs is this that even with huge amounts of fuel they do not move on as they already have enough weight. :cry:

Now as the axial flow engines are much lighter and easier to construct, so i chose to make a 5 compressor stage axial flow jet.

I know that this sounds crazy but it all would be fun.

Hope that all this doesn't bother you


I want to ask one more question that wheather should i use regulator on my LPG cylinder while using it on a GT???

[saugumas]

There where discuasion about quickly wearing out bearings.
Maybe this info be useful:
turbines need good lubricatin systems(except for foil brearings), if turbine is hot, and you want to shutdown engine, you must wait some time because if don't wait bearings could be damaged and durability greatly degrade.
in some cars there is device called Turbo Timer, read http://en.wikipedia.org/wiki/Turbo_timer


I am not proffesional, so critics are welcome :)


Sorry for bad english.

[Ash Powers]

Although I haven't posted much on the GT, there has been a lot of work going on "in the background". The new compressor diffuser appears to be working quite well but I have been sorting out issues with the turbine section. I finally got the rotating group balanced which made a healthy difference but I'm finding that I cannot get the engine to run up to much more than around 11-12psi without the TOT's getting out of hand.

The previous post with the airflow meter info and comp map had a bug in it - when looking at the compressor flow map and knowing the massflow and pressure, you must divide the measured airflow by the pressure to know the x-axis value. I previously didn't do that and my "measured" 36lbs/min was actually represented by the x-position of 22.5lbs/min, which puts the compressor operating out in its lower efficiency to the left of center. I just figured this out this afternoon and it clears a lot of things up - I had blocked off three of the ducts in the NGV to reduce the flowrate, but the engine ran even worse. Well, now I know why. :) Interestingly enough, when I did this on the original NGV, the compressor went HARD into surge. With the new diffuser in the engine now and blocking off three ports, there was no surging from the comp so I get the impression that the new diffuser is working better.

I was also concerned with the previous revamp of the NGV as I had cut the vanes back quite a ways from the turbine's periphery. I built a new NGV and this time the vanes are curved which will allow for some tuneability. I also didn't put the recess into the NGV plate like previously and now have a NGV duct height of 0.485" vs. 0.440 of previously. The diameter of the NGV plate has also been reduced by about 0.200" as the entrance throat area was a little too small previously.

ON TOP of all of this, I also clipped the turbine wheel back about 10 degrees to allow it to flow more air.

Unfortunately I ran out of 3/8" steel plate and still need to machine the rear cone fitting on the NGV for the combustion chamber to mate to, so I'm going to shelve the engine for the weekend and get back to it next week.

Hopefully these adjustments will improve airflow through the engine, allow the turbine to soak up energy a little better, and give the compressor suitable conditions for optimum efficiency.

Here's some pics:

[Jim Berquist]

How in the world do you all balance these things?

Jim

[Ash Powers]

I used the dry-bearing shaft support with a vibratory source. I put the turbine and shaft extension together and cleaned out the bearings using brake parts cleaner and installed this into the bearing tube. These ceramic bearings, when dry, have exceptionally low friction and the heavy side of the wheel will find its way to the bottom. I actually use my air compressor as the vibrating source - I just put the bearing tube on the compressor's handle. I then used a dremel with a grinding stone to remove material from the heavy side until the wheel would stay in any position I put it into. It made all the difference in the world - before I did this, as the engine came up in speed, all the little items on the test bench would start dancing around from all the vibration.

[Jim Berquist]

Smart, Simple sort of and gets you most close yep! Like a electric motor you just dimple the outer portion until you can no longer find changes? Do you have to dimple the front and rear faces? If you could dimple the leading edge would that be better?

Jim

[Ash Powers]

Been making good progress on the scooter. :) I got the engine mounted, built the fuel tank with the in-tank pump, built a case for the electronics, mounted the battery, mounted the PDA, wired everything up.

Unfortunately when I fab'd the fuel tank bulkhead with the electrical fittings, I somehow created a short on the pump's power lines which destroyed the power transistors for the fuel pump. I was hoping to have run the engine again but now I'm waiting for the replacement transistors to arrive.

In the mean time, I discovered something which has got me pretty excited. Apaprently somewhere along the lines, the calibration for the thermocouple inputs went awry. The thermocouples are reading about 100C higher around the 600C mark than what the temperature actually is. A lot of the modifications I have been making to the turbine have been directly a result of the turbine outlet temps I've been seeing and just about each time I've made a change to the engine, I have improved on this condition. With this new info, I have actually only been coming up to around 550-585C TOT's. I'm really excited about this discovery as I'm inclined to believe that once I get the engine running again, it will likely perform as it should, or at least a LOT closer to what I've been expecting.

So, until I get the transistors, here's some pics to check out. :)

Engine mounted:


Tank built: 2.75 gallon capacity, 1/16" aluminum.


This image shows the drop-box in the bottom of the fuel tank where the pump's pickup sock resides. The front wall of this drop-box is retro-angled so that when the brakes are applied, the box will hold fuel to prevent air from entering the pump. This drop-box is 2" deep.


PDA and switch panel mounted: (PDA is held in place by velcro to make removing it simple)


Engine set back a few more inches to accomodate the starter (not shown) and provide enough room for a seat:


I also started building the starter. This is using a 540-sized radio controlled car motor: 55,000RPM @ 12V with lots of torque to spool the engine - I've just machined the housing and mounted the motor inside but I will get to the tri-pod mounts and bendix assembly this weekend.




Custom tag: :) :)


And this is about-how I will sit on the scooter (note: my butt will not be on the engine, LOL)

[Johansson]

Ha ha, wonderful!

I am really impressed by the quality of your work, have you any idea how many hours you have spent on the engine so far?

//Anders

[Ash Powers]

Anders, Thanks for the feedback - I'm pretty particular when it comes to building things and a vast majority of the design takes place in the grey matter between my ears before I go with "tools in hand". Although the development of the engine hasn't been as such, I prefer to build it right the first time around rather than hassle with the troubles that will come up by doing a half-rate job. The engine obviously being something where dimensions, geometries, etc, play a critical role to the actual performance and it is expected to have to rebuild items to get them running as expected, but even still, all the parts I've built for the engine have been built to the best quality I can muster.

I dont have an exact figure on the amount of hours I've put into the engine build - if I were to guess, over a 6-month period of time now since I began, I think it would be fair to say I've burned 10 hours per week at least - Roughly estimating, perhaps 250-350 hours total in the engine and engine electronics. A sharp contrast to the ~8 hours in the scooter itself and its just about done - building the seat and installing some foot pegs can be done in about two more hours.

[Johansson]

Every hour on such a project is well spent, I know people who spend most of their spare time in front of the TV and still think that it is a total waste of time and money to build things without a practical use...

How did you make the tachometer work by the way? The IR-diode and reciever sits next to each other so the nut must be reflecting the signal somehow, is it painted black and white perhaps?

//Anders

[Ash Powers]

The tach uses the IRLED and phototransistor with half the compressor nut colored black using a felt-tip marker. I machined the compressor nut from a piece of 6061 aluminum - it doesn't have "points" like a real nut - it is smooth all the way around. That part of the tach needs some work though as it is pretty sensitive to small notches in the comp nut and causes extra pulses that throw off the tach reading. I think using some optical cable in a tube with a laser light source will work much better and be less susceptible to ambient light conditions and small notches in the nut (from tightening/loosening it with a pair of pliers).

[Johansson]

What kind of threads are you using, clockwise or counter-clockwise?

With the standard threads on a turbo shaft is is necessary to tighten the nut with enough torque that it won´t come loose while running the engine, the counter-clockwise threads on most micro turbines makes the nut tighten itself while running so it can be made of softer materials and tightend by hand.

I think that I am going to try a SS domed nut on my engine since it can be tightened with the right tools and still offers a smooth surface for the tacho to reflect on.

//Anders

[Ash Powers]

Anders,

The rotating group turns clockwise so I threaded the end of the turbine and shaft extension with left-hand threads - it is all left-hand threaded to keep things from coming loose. :)

I might turn a new one though and use SS as well - I will be able to cross-drill it and make a pin wrench so I dont have to worry about marring up the sides of it..

I haven't been using the tach that much anymore as it is a bit tedious to setup each time I pull the engine apart and make a change. I've been using the mass airflow sensor and the pressure gauge to see where I'm falling on the compressor map, of which the RPM of the group is also indicated in.... gives a much better idea of where you are in the map when using the MAS.

[Ash Powers]

Things are coming along nicely. The scooter part of the setup is more/less complete at this point. I built the seat for it and installed some foot pegs - feels like sitting on one of those custom built WCC motorcycles; arms and legs straight forward. :)

I also machined the mounting legs for the starter motor and attached it all to the compressor inlet duct. I had to custom order some metric drill bits as the shaft on the electric motor is 3mm dia and I dont have that here so I'll have to wait until wed/thurs for the bit set to arrive so I can machine the starter shaft.

Instead of using a bendix assembly on the end of the shaft, the electric motor is able to slide back/forth in the housing. Where I originally had the two fixing screws in place has been replaced by two 0.25" dia pins I cut out on the lathe and opened up the holes in the starter casing. These pins slide into the bores and will prevent the electric motor from twisting under its own torque when it spins up the rotating group. A return spring will be inserted to push the electric motor back away from the compressor nut. I'll end up machining a button to put into the front cover of the casing that can be depressed to engage the starter shaft to the comp nut as well. There is about 0.25" worth of forward/backward room for the motor to move which will be plenty. The end of the starter shaft will be cut from a 5/8" dia rod stock and the compressor end will have a bore cut in it to house a short piece of 7/16" dia rubber hose. This will provide the friction pad to engage the comp nut.

The only main thing left on the scooter is to mount the propane starter tank. This will be mounted in front of the battery. The tank is one of the tall skinny cylinders used for the brazing setups you can purcahse at Home Depot/Lowes, etc and will be good for ~20 starts. I may make a more elaborate seat later on - I'm sure that aluminum plate will get a bit uncomfortable in short time.. Something with ~2" foam padding and some vinyl would work just fine.... and I'll get to use my sewing skillz again, LOL. The seat assembly is detachable as well - you can see the mounting pins welded to the engine mounts. The seat assembly simply slips onto these pins and comes right off for easy access to the electronics and the front of the engine.

Here's some pics:

[Jim Berquist]

Man you do some good work. It looks so clean, neat and even funtional.

Jim