Built engine, having problems getting it to start

[hosedup]

We had it running for about 2 or 3 seconds and that is it.  Except for when it ran, it has a big blue flame that comes out the exducer.  Im thinking the flame isnt getting seated in the tube.  We have tried small amounts of fuel all the way up to wide open, all we get is lots of fire.  We are using a leaf blower to start but also a shop vac, it spins it faster.  My thoughts are there need to be more primary holes and the tertiary holes need to be bigger.  One thing I noticed inside the can are burn marks that appear to match the secondary holes.  I do not know how long it burned like this, I suspect short because the outer surface of the can does not have heat scaring. 

Here are pics of the build.

Flame tube is 4.5" OD, holes are:
30x 0.25in, 2 rows, 15 each, 1st row offset from top by 1.2in and 1.7 for the second.
13x 0.38in, 1 row, offset by 6 in from the top.
7x 0.5in, 1 row, offset by 11in from the top.  
Flame tube is 15 inches to reducer, then 2 more inches to flange.

Injector has 6x 0.1in holes radially, fed from one 0.25 hole down the middle.  There is not a hole in the face of the injector.

Spark plug is standard automotive plug being continuously fed high voltage from an old laser power supply.

Fuel system is throttle valve attached to propane tank, no regulator.

Outer can has a 1in gap all around from the flame tube.  Charge air does not enter at tangent.  

There is a small air leak at the joint from the can and lower flange.  I have not sealed this up thinking that it might be good to have a small air leak so the compressor doesn't stall as easily.

[racket]

Hi

In an earlier email you mentioned the comp inducer was 50 mm and that....." Ive played with jetspecs a bit, inducer dia as 50mm, it suggests primary holes 6mm x 21, secondary 10mm x 5 and tertiary 16mm x 5"

Why didn't you use the Jetspecs recommendations ??

Currently your 30 X 0.25" Primary holes represent 1.47 sq inches of hole area , your 50mm inducer represents ~3.04 sq ins , your Primary holes equate to ~48% of the inducer , far too much , should be no more than 30% .

Your 13 X 0.38" Secondary holes again represent 1.47 sq ins , so ~48% of inducer , again too much, should be ~20%

The Tertiary holes at 7 X 0.5" represent only 1.37 sq ins or just ~45% of inducer , not enough , should be ~50%

The flametube holes are "out of tune" , you have ~40% too much hole area , which is probably being exacerbated by the missmatch between comp and turb stage sizes/flows which will most likely be reducing flow and increasing that excessive hole area to closer to 60% .

May I suggest a new/modified flametube with 15 X 7 mm Primary holes at your 1.7" displacement distance ( blank off top current row and slightly open the diameters of second row ) and have your fuel injection radially towards those holes.

The Secondary holes need to be greatly reduced in total area , either blank off half of them ( every second one) , or blank off the entire row and drill 10 X 7 mm new holes .

The Tertiary holes can be left as is or slightly opened to ~13 mm .

Cheers
John

[hosedup]

John,
Thank you your advice. In the process of converting metric in jetspecs along with bad math, I screwed it up. Going to fill in holes and fix it.

[mrcoffee]

Hello everyone. I've just joined the forums. Thank you all for providing such a resource!


I'm the other guy in the "we" that Hosedup mentioned. We go back quite a ways, and have a hearty love for projects that make our other friends ask "Why? -Oh, Nevermind, it's you guys."


Having read this thread, I feel it necessary to mention that Hosed and I are in the land of oversized fast food portions and strange fractional math (the U.S.)


John,
First off, thank you for all of your advice. For the ~4 seconds that our turbojet made "jet" noises, we were very excited. While we're definitely champing at the bit to make it happen again, we both would like to do it more properly this next time around.

You've mentioned 15x7 mm holes for the modified flame tube. Our drill bits here are sized by 64ths of an inch. These bits land us at 7.14mm or 6.74mm as the closest drill bit size.

Are these differences enough to make a palpable difference, or should we just pick the closest match and go with that? (leaning towards "Just go with the 9/32" (7.14mm) bit).

There are similar discrepancies with the other holes, but I'm guessing they'd be in the same situation - get it as close as we can and proceed from there.


On another front, how do you all feel about oil? One source I read said they were using 5w-50. Given how hot the oil can get I can see why that could be beneficial. We ran our first test run with 5w-20, Definitely had to increase pump speed as the oil warmed up and pressure dropped (aiming to maintain 30 psi).

We're going to change the oil before the next session. Any thoughts as to putting more 5w-20 in, or should we step up to 5w-50?

Regardless, we're having a blast working this problem. Thank you all again for your help.

-Rhys

[racket]

Hi Rhys

LOL...........I have a set of Metric and a set of Imperial drills and use either/both , whichever is sharp ;-)

Go the 9/32" , we don't have to be too pedantic about size as theres generally far bigger issues with our engine that will change overall flow rates .

The general advice with oils is to use a full synthetic , and the oil specs for the engine the turbo came off , but if you'll be starting the engine with very cold oil then you'll need the 5W ... , as for the "hot end" of the specs , generally we don't run long enough to need "50" , though I have used it on occassion,.................we don't want too thin a hot oil as the rotor dynamics can be compromised along with difficulty in holding pressure .

Pressure , measured at the turbo oil inlet port , is critical to survival of the engine , so aim to have at least as much oil presure as max comp discharge air pressure ( P2) , preferably at least another 10 psi or so as we run our "turbos" harder than when on the IC engine ..........I start my engine on 70 psi oil pressure , my old Garrett TV84 often saw 100+ psi using 15W-50 .

All the best with the next testrun :-)

Cheers
John

[hosedup]

John, when you spoke about the secondary holes, either blank half or blank all and redrill at 7 mm, which would be better? I can do both with nearly the same effort. Since there are an odd number of holes, its impossible to fill every other one. I believe this would cause one area to either have too much or not enough air. For this reason, I'm assuming it would be better to weld up all holes and drill again.

My plan is to do this tonight as well as open the tertiary to 13 mm.

Looks like I have an excuse to buy a metric drill index.

[slittlewing]

Nov 6, 2019 1:01:11 GMT -5 mrcoffee said:

On another front, how do you all feel about oil? One source I read said they were using 5w-50. Given how hot the oil can get I can see why that could be beneficial. We ran our first test run with 5w-20, Definitely had to increase pump speed as the oil warmed up and pressure dropped (aiming to maintain 30 psi).

We're going to change the oil before the next session. Any thoughts as to putting more 5w-20 in, or should we step up to 5w-50?

I have been using "thin" oil (Aeroshell) up until now, and struggled to make high pressure when hot (it gets hot very fast). Based on advice from this forum I will be moving to 10W50 next time - extra pressure cant hurt!

[racket]

Yeh , preferable to weld up all holes and drill new Secondary ones

[hosedup]

Just filled the holes and redrilled.  Hoping to try a test firing tomorrow. 

[hosedup]

It lives!!  It was surprisingly easy to start with a reversed shop vac.  Once it lit and was starting to spool up, it would pull a vacuum on the hose supplying start air, so much so it would hold itself to the inducer.

Some observations:
We were running the propane tank nearly wide open.  In the video below, about half open.  It seems like there is room to grow, need more fuel.
There is smoke (I am assuming combustion products) coming out of the oil tank.  We were running ~30 PSI oil
There is an air leak on the combustor can that I need to seal up.  Then again could this leak may be beneficial because its a hybrid T3T4 turbo?
The flame tube seems to have even heat marks.

Short video of it running:
www.youtube.com/watch?v=Dk-PQFxEmc8

[racket]

That looks better ...........good ol' Jetspecs comes to the rescue :-)

Yep , oil tank is a bit fumey ...................is your P2 pressure gauge needle steady of bouncing around a bit ??

Cheers
John

[mrcoffee]

Hello again everyone.

Just figured I'd drop a couple of links in here to wrap up the story of our first Turbojet.

The good: We got it running! Twice!

The bad: We destroyed the turbo! Twice!

However, the turbos were $35 each. And we learned a lot about what NOT to do. So I'm ok with destroying the cheap turbos before we move on to more expensive models.

Definitely need more granularity and data points over the oil system. Pressure gauges also probably not a bad idea.

Also, I learned not to spool a turbo too quickly. I think that and oil starvation killed turbo #2 (the vanes crashed into the housing, I'm wondering if they didn't heat up too quickly and expand faster than the exducer housing.)

Naturally the gopro crashed during test fire #2, so there's just this 15 second clip of it running:

www.youtube.com/watch?v=Dk-PQFxEmc8


The next day we got it running again, but as I mentioned, I feel I ran it up too quickly. Also forgot to keep an eye on the oil pressure:

www.youtube.com/watch?v=UgS1LSFkRt0


So, yes. Much learning to do. Also for the next build we need to slow down and build out the supporting systems more fully.

Regardless, Thank you all for the help on our first Turbojet! I figured the least I could do was to post a conclusion on it.

On to the next one,
-Rhys

[racket]

Hi Rhys

A couple of points worth mentioning .

Firstly gas turbines have reatively long start/spoolup times compared to IC engines , we need to keep the "starter" engaged until they are up to idle .

Secondly it would appear you were running into surge, the flames from the turbine exducer were "pulsing"

Thirdly , oil pressure is necesary to not only lubricate the bearings but cool them and to provide stabilisation of the rotor , unless the rotor has that stabilisation the shaft can gyrate causing contact between wheels and housings , this can be exacerbated if there is surge present from a mismatch of comp to turb flow capabilities.

We must have gauges , the more the better , oil and P2 a must , then turbine outlet temp thermocouple to help keep the turb wheel from "glowing" and destroying itself.

And to finish off , its not a good idea to check the turbine wheel by looking at it whilst its throwing off red hot bits of metal

All the best with the next build, I'll be looking forward to progress pics :-)

Cheers
John

[hosedup]

Yeah, looking at the business end of a self distrusting homebrew jet engine wasn't a great idea. The safety glasses don't do much for the face. Next time either a camera or mirror will be employed.

[mrcoffee]

Wait, Who looked at it? The video was from a gopro tied to a table. I'm not going downstream of that thing while it's running.