Wednesday, 23 April 2008

I got it! I got it! - - - I don't got it! (groan)

So what's happenin' Dude? I hear you ask.

Well, I almost had it running - but you've heard that story before? no - really!

My careful measurements of the power required to drive the engine from an external motor as I brought the fuel flow up have convinced me I'm on the right track.
The plot of power versus temperature exactly parallels the theoretical calculation.

Except for two effects.

The predicted relation is a straight line but I found I was getting a curve that showed power being lost in two ways:-
One was at startup. The effort required to spin the engine cold is much greater than expected. I put this down to the effect of oil viscosity.

As an aside, it's amazing how fine the operating point of the turbine is - If losses exceed gains by even the smallest part of a watt the engine won't run.
If gains exceed losses by just as much, the turbine can accelerate out of control in seconds.
Despite the massive power potential, the friction losses resulting from a few grammes of resistance observed at the turbine diameter make the difference between life and death.

The second effect is a similar frictional loss that appears at high temperature - I have no idea what is causing it. It's tiny - I can barely feel it when turning the shaft by hand - but it kills the engine stone dead.

I said I almost had it running?
Yes: With the engine being driven flat-out by my blower and pushing the combustion temperature right up to the limit, I reached a region where the engine definitely tried to run.

No doubt in my mind at all. It suddenly accelerated, becoming very sensitive to the amount of fuel being fed in - for a moment I thought I had lost control from the way it accelerated. But then effect #2 kicked in (I think) and the engine slowed again -then picked up - and slowed again.
After a few such spurts, I decided to stop the test because the turbine wheel and NGV were glowing brightly and I know where that leads! (See last post)

Anyway - Got to find the cause of that friction.


Anonymous said...

Hi, just found your blog and was wondering whether you were still trying to build a model gas turbine? Reading the posts from 2007 through to this last one, I had a few comments which i thought you may be interested in ...

Anthony said...

What with one thing and another, this project got put on the back-burner.
I came to the conclusion that the only way to work out what was wrong was to get some experience of running a car turbo type jet. But limited space and budget make that problematical.
While I decide what to do, I've cleared the workshop space and am using it for other purposes.

Anonymous said...

Hi, thats too bad. I had a look at your methodology / tests and thought it was all pretty good. I was having a think about what the issue was, I did a few calculations and would be pretty confident that your pressure ratio is too low (1.1) to allow for enough expansion in the turbine to produce enough energy to run the compressor when the compressor and turbine losses are taken into account i.e. because the efficiencies in these small components will be quite low (~50 - 60%) compared with full scale machines (~80 - 90%), with a such a small pressure ratio, you dont have enough expansion ratio to compensate for the losses as well as run the compressor.

Seeing as you got evidence of surge / some turbine work / other turbo machinery related effects, i would be encouraged the components are working, albeit probably quite far of optimum efficiencies.

Boosting compressor diameter + engine RPM as much as possible would be a good basic start. Adding an inducer to the compressor inlet would also help reduce flow losses / help prevent blade stall and boost overall compressor efficiency. Would estimate target PR of 1.3 - 1.4 would be absolute minimum to cover the inefficiencies / bearing losses.

Diffuser efficiency - I havent looked closely, but as with any diffuser, always better to be slightly conservative with your vane angles (smaller rather than bigger, longer rather than shorter) to prevent separation and major flow losses. 7deg is the reported theoretical optimum, but ~15deg is fairly common for model gas turbines designs i have looked at. I dont think they are too caught up on efficiency.

You may need to experiment, although ultimately the inlet / outlet area ratio is what will govern the degree of pressure increase in the diffuser.

anyway, hope this helps a bit ...

Anthony said...

I agree.
The reason it's on the back-burner is that I think I need a complete rebuild.
The small size of the components make them difficult to work accurately and on the whole I think I would be better doubling up on all sizes.
Increasing rpm might help get me 'over the hump' but I was hoping to keep that down as much as possible for safety.
It's interesting that you focussed on the compressor/diffuser area. That's where I feel myself that the design falls down.

Anonymous said...

hi. i have also looked into jet turbines. pretty much since i saw fusionman fly across the english channel i began to think....
"boy that's incredibly stupid.... I WANT ONE!"

and so my research began.

I think you're right about having a poor compressor diffuser area. it looks like it's trying to make up it's mind on wether it's a centrifugal or axial compressor.

If i were you i would stick to an axial compressor (it looks simpler) you need 3 things:
1. inlet guid vane
2. compressor
3. Stator (this is what you don't have.)

I have taken a look at a few engines. (google yay!)

wikipedia has a few nice images.

as well what type of bearings are you using? It doesn't matter to me, but even bearings can cause friction. i suggest magnetic bearings (unless that's what you're using) because they have less friction and it's more predictable.

Anthony said...

Axial flow compressors are more efficient at larger scale, as in full-size engines. However at small-scale, the engineering tolerances are too severe.
That's why car turbos use radial compressors, as do all model jet engines.
Your point about bearing friction is absolutely valid but magnetic bearings cannot function at elevated temperatures.
Neodymium for example has a curie point around 300 degrees C.
Even steel or ceramic bearings need significant cooling to survive in the jet flame and need to be replaced almost every run. It took several tries before I was able to prevent the bearings seizing, by ducting a cooling mix of air and lubricant vapour down the drive shaft tunnel using a pressure bleed from the compressor.

warman45 said...

i wish to thank you for replying to my previous questions. i am the last anonymous person to post.

i'm incredibly interested in you're jet turbine. i'm not gonna pretend i know something. as a matter of fact i only understand the basic theory of jet engines. but i want to learn.

do you have more photos of the engine. from the photo's you presented i still have no clue how you managed to provide adaquet cooling to the bearings. as well i would really like to get a closer look at you're compressor.

btw you mentioned that axial compressors could not withstand the stress in smaller sizes. what about them causes them to fail (ets say i wanted to make one with a diameter of 15 cm.)

what type of aluminum did you use?

how much did the aluminum blanks cost and where did you get them?

did you try making the turbine slightly larger to compensate for friction?

did you try using a more flammable feul?

did you try bleeding off some of the air just after the compressor to increase efficiency? (turbofan)

how much of the original vacume is still in use?

is this model intended to simulate a power generation turbine or an aircraft engine?

my dad actually has a few books on aircraft maintenence. probably useless beyond basic theory but still pretty cool.

do you draw out you're desighns before building them?

why does you're NVG IVG only extend partway into the engine?

what material did you use for the engine housing?

about surging. did you try implementing a bleep valve that would stop the pressure from exceding operational tolerances? (and breaking something)

can i see more photos of you're engine?

are you still working on you're engine?

where can i learn more?

Anthony said...

That's quite a list...

I'm sorry, I've got no more photos.

I said the engineering 'tolerances are too severe' - That means the accuracy of the parts has to be impossibly fine. When you make things smaller, you have to make the gaps between parts smaller too.
Axial compressors are very complicated with hundreds of tiny blades spinning close to each other.

Nothing special about the aluminium - standard eBay stuff.

I've documented all of the work so far in this Blog - If it isn't mentioned here - I didn't try it.

None of the vacuum cleaner parts were used in the end.

The engine isn't for any specific purpose - it's just for the fun of building it.

I don't make full engineering drawings - sometimes I might sketch out critical parts roughly to clarify my thoughts.

If you really want to learn about model jet engines, get a copy of Kurt Schreckling's book (Google his name)
All of your questions are answered very fully in that book.

warman45 said...

By the way. My dad has a General Electric engine part. it is either a stator, compressor, or turbine. if you want i can take some pictures and send them to you via E-mail

warman45 said...

By the way. My dad has a General Electric engine part. it is either a stator, compressor, or turbine. if you want i can take some pictures and send them to you via E-mail

cory3101 said...

Anthony, it seems your having trouble with getting your jet engine to work. I've started to attempt to build my own, the same way you have done.
For that friction that you cant find, I ran tests on my own and i realized that the bearings weren't exactly centered, close, but multiply that close by 50,000-150,000 and it becomes major friction as im sure you know.
And with about getting your jet engine to run. T having a water injection in between the turbine and the combustion chamber. Water spraying into the combustion chamber can increase the efficiency of jet engines by 50%. Because of the law of Conservation of Energy, the 2 forms of energy in a jet engine are heat and pressure, corresponding to each other. So a decrease in heat means a decrease in pressure. Steam replaces the pressure lost and can allow you to run higher temperatures before the turbine.
Just some hopeful ideas about how to help from my own experimentation and some other groups help. Hope this helps, and thank you for all of your amazing information and documentation.

Anthony said...

Hi Cory.
I'm glad to hear someone else is going to have a go at this.
Hopefully you will have more patience and skill than I.
My conclusion was that it needed more time, money and ability than I possess.
It's obviously possible because it's been done and these engines are available commercially. Sadly, the cost of developing a homebrew version is likely to be more than the purchase price of an off-the-shelf version. That kinda kills the fun for me.
So many things are like that these days - it's no wonder kids spend their time playing with ready-made toys and games.

folu said...

Hi Anthony. I was browsing the internet for home made turbine engines when I came across your site. I must say you're a genius. At some point after reading from the first post, I skipped to the last page to see the outcome and was shell shocked to find out that you dumped the project.
After reflecting on this, I came to one conclusion- you involved too many calculations. Even though I actually learnt a lot, I believe the calculations are not as important as having faith that your engine will work even when it doesn't obey physical laws. I'd be very excited to see your engine running soon.

chennai said...
This comment has been removed by the author.
suresh said...

Hi Anthony...I found this site luckily when i googled about self-sustaining RPM. This site has tremondous insite into the challenges in the putting a gas turbine together. I m involved in arriving at a calculation/methodology to compute self-sustaining RPM. Is it possible for you to throw some insight in calculation of self-sustain RPM for a gas turbine? or else any books/references, which i can go through. Thanks for this brilliant blog

novelinflow said...

Technology Submission - State of the Art - Novel InFlow Tech - Featured Project Development;|/ ·1; Rotary-Turbo-InFlow Tech / - GEARTURBINE PROJECT Have the similar basic system of the Aeolipilie Heron Steam Turbine device from Alexandria 10-70 AD * With Retrodynamic = DextroRPM VS LevoInFlow + Ying Yang Way Power Type - Non Waste Looses *8X/Y Thermodynamic CYCLE Way Steps. 4 Turbos, Higher efficient percent. No blade erosion by sand & very low heat target signature Pat:197187IMPI MX Dic1991 Atypical Motor Engine Type. |/·2; Imploturbocompressor; One Moving Part System Excellence Design - The InFlow Interaction comes from Macro-Flow and goes to Micro-Flow by Implossion - Only One Compression Step; Inflow, Compression and outflow at one simple circular dynamic motion / New Concept. To see a Imploturbocompressor animation, is possible on a simple way, just to check an Hurricane Satellite view, and is the same implo inflow way nature.