PD-01 Jet Engine

[finiteparts]

Also, a quick way to "move" the operating line over towards the surge line is to put an exhaust nozzle on...it would be much more simple to do that than my previous suggestion of plugging a NGV passage.

If you make up an adjustable nozzle, it would allow you to "tune" the mass flow and pressure. You could watch the manometer to see how the mass flow was responding to the engine backpressure.

Also, if you experience surging, the water on the static pressure side will response pretty well, even with the fluid inertial effects.

Finally, John, if Patty was going for super accurate measurements, then the bellmouth like the one provided in the link would be justified, but for simple testing, this one is accurate enough.

The values I calculated were using the geometric area of the pipe, but in actuality, there is a blockage effect of the boundary layers, the flow is not going to be a fixed velocity across the face, etc...but this is a good ball park number.  

The numbers fit, high RPMs, low PR's...he is on the choke side of the map.

~ Chris

[madpatty]

Adjustable nozzle?
I need some design suggestions....

But its a bit of relief because didn want to plug the NGVs...

Cheers,
Patty

[finiteparts]

I'll have to get back to you on that. We made a 2-D nozzle for our Senior Design project engine that was super simple. A hinged plate in a square tube. The hinged plate had a long bar that we could pull while it was running and "set" the nozzle area.

I'll try to think up some other ideas.

~ Chris

[racket]

Hi Patty

A few numbers for you to think about ............

Assuming a T I T of 900 C -1173 K at a PR of 2.18 ( 19psi P2 = 2.3PR X 0.95 = 2.18 into turb NGV ) entering your NGV stator which has a throat area of 1.39 sq inches - 0.00965 sq ft

A choked NGV will have a ~1.9 PR drop , therefore 2.18 divide by 1.9 = 1.15 PR at the throat , there being an ~155 deg C drop across the "nozzles" so a static temperature of 1018 K -745 C .

Density of the gases at 1.15 PR at 1018 K is ~40 cu ft/lb - 0.025 lbs/cu ft .

Speed will be ~2,000 ft/sec

Mass low = speed X density X area = 2000 X 0.025 X 0.00965 = 0.48 lbs/sec = 0.219 Kg/sec

Your calculation of 0.39 Kg/sec is 77.8% greater .

Your turbine wheel has a tip height of 12mm , with a 70mm dia the inlet area is ~4.090 sq ins - 0.0284 sq ft , to get 0.39 kg/sec - 0.858 lbs/sec with density of 0.025 lbs/cu ft through that area the radial speed would need to be 1208 ft/sec , and with a 2,000 ft/sec gas speed the NGV discharge angle would need to be ~37 degrees , and you'd need a tip speed of ~1600 ft/sec .

Now we come to getting the gases out of the exducer , if exducer is 58mm and hub 18mm then annulus is ~3.7 sq ins - 0.0257 sq ft , because the exducer area is smaller than the inlet area there needs to be a pressure drop , but we don't have much to play with as we used up most of the pressure at the NGV throat

Lets assume we use up the remaining 1.15 PR -2.2 psi and the gases are at ambient pressure, and temp has dropped another 30 C degrees , so the gases exit the turbine wheel at 988 K with a density of 44.8 cu ft/lb - 0.0223 lbs/cu ft , your 0.858 lbs/sec will need to be travelling at 1432 ft/sec axially out of the exducer annulus .

Unfortunately you'd need a gas speed within the wheel nearly double that to produce that exhaust speed, and we can't get that from a 1.15 PR .

A second "unfortunate " for you to consider .............1432 ft/sec at static pressure of 14.7psi represents a total pressure of probably close to 20 psit , which you don't have .

As I told you a long time ago , do the sums .................you cannot get the gases from your BIG compressor wheel through the SMALL turbine at the high temperatures you are using .

The addition of a jet nozzle will only exacerbate the situation , any restriction to the flow will raise backpressure , reduce the pressure drop across your turbine wheel and your temperatures will skyrocket .

Please get your calculator out and prove me wrong

Cheers
John

[racket]

Hi Chris

I found out may years ago that trying to calculate flow from the difference in static vs total pressure was "troublesome" , I was doing it within the jet pipe , it being purposely enlarged to reduce the velocity .

I feel we need a "correct" bellmouth and duct on the inlet to do it even half well using DIY methods .

Cheers
John

[madpatty]

Hi Racket,

There are certain parameters which appear conflicting.
As the mass flow you a have calculated puts me right in the best efficiency of 74% and rpm also should be near 89000 rpm which is clearly not the case as can see.

This only point towards one error in your calculation which is the TIT of 900 degrees celsius you have assumed.

Yesterday when i ran the engine i got a TOT reading of 603 degrees celsius. I rejected that thinking that the thermocouple probe is faulty which i bought new a day ago. But starting the calculations taking the TOT of 603 degrees at 2.3 PR the mass flow calculated begins to fall into the ballpark.

From the map 2.3 PR at the rpm measured puts me in a efficiency island of 72%.

Therefore temp rise in compressor 107 deg.C.
Required temperature drop in turbine ~ 94 degrees.

TOT= 603 degrees celsius
ITT(total)= 603+94= 697 degrees clesius or 970 kelvin.

Therefore static temperature at Turbine inlet(flow choked condition)~ 833 Kelvin
Density(static) of air at throat ~ 0.50 Kg/m3

Throat area~ 1010 sq.mm (1.39 sq. mm was the theoretical area, i made it a bit larger)
Air velocity~ 580 m/s

Therfore mass flow= 0.001010 x 0.49 x 580 = 0.3 Kg/s.


Though it is not near 0.39 kg/s but still on the choke side.

Cheers,
Patty

[racket]

Hi Patty

OK , lets rework for the lower temperature .

As you are adamant that you are flowing in the choke region of the map your compression efficiency will be down at say 65% or below , but lets use 65% for calcs .

2.3 PR at 65% gives us a 119 C degree rise in compression , this will require a 104 deg drop through the turbine stage, therefore using your 603 C TOT the TIT would need to be 707 C - 980K

Considering your turbine stage construction I'll assume a generous 75% efficiency , this would then require a 1.84 PR across the turbine stage to power the compressor .

Assuming our 2.18PR going into the stage , you'd have an ~ 1.2 PR coming out of the stage.

At this point it doesn't look like you need to run a choked NGV , but lets do it anyway .

A 980 deg K T I T , and a 1.9 PR , this will produce a ~130 deg C drop at 90% effic., and a gas velocity of ~1800 ft/sec , gas density at 850 K and a PR of 1.15 is ~34 cu ft/lb or 0.0294 lbs/cu ft , we'll use our previous 1.39 sq in - 0.00965 sq ft throat area .

Mass flow = 1800 X 0.0294 X 0.00965 = 0.51 lbs/sec or 0.233 kg/sec .............we've squeezed another 6.4% extra through the NGV at the lower temperature , but your calculated mass flow through the comp is still 67% greater than what can be processed by the NGV .

But you still haven't addressed the issue of getting the flow through the turbine wheel .

As a side issue , assuming 0.39 kgs - 0.858 lbs/sec mass flow into the comp , your 54 mm inducer has an area of 3.55 sq ins , lets knock off 0.25 sq ins for the hub leaving an inducer annulus area of 3.3 sq ins or 0.0229 sq ft , assuming an air velocity of 600 ft/sec going into the comp , this will require a 16.6 C degree drop in temperature , assuming your ambient is 20 C , then the actual air temp at the comp face is 3.4 deg C , static pressure would be 11.98 psia , density at 273.5K is 0.065 lbs/cu ft , mass flow would be 0.89 lbs/sec if there weren't any blades in the way , so 600 ft/sec would be a reasonable assumption to make for airflow into the passageways, this is an extremely high figure and is usually only achieved at reasonable compression efficiencies at extremely high pressure ratios for a turbo running billet wheels , at lower pressure ratios it implies very poor compressor efficiencies .

LOL.......the ball is now in your court............hit it back :-)

Cheers
John

[madpatty]

Hi Racket,
LOL....here comes the ball....;-)

There is no fun doing and reading the same calculations...
Talking about the flow you are calculating(0.23 Kg/s) and which i have calculated(0.3 Kg/s, accroding to present geometric NGV throat area and compressor map), the thing that is very sure is my compressor is certainly not into surge or anywhere near it, if not choke.

The flow which you have calculated is in the best efficiency island of 74%.....how you justify the use of smaller compressor then?

Cheers,
Patty

[finiteparts]

Apr 3, 2015 0:01:11 GMT -5 racket said:

Hi Chris

I found out may years ago that trying to calculate flow from the difference in static vs total pressure was "troublesome" , I was doing it within the jet pipe , it being purposely enlarged to reduce the velocity .

I feel we need a "correct" bellmouth and duct on the inlet to do it even half well using DIY methods .

Cheers
John

 John, 

What you found out was that measuring flow speed in a "jet pipe" was problematic...this is different.  In a jet pipe you have highly distorted flow due to the blade passage wakes, residual swirl, etc...the static pressure field is highly non-uniform and time varying...thus it would be difficult to accurately measure local flow velocities with home-made means.  For the industry, they do it all the time by using fast response pressure measurement instruments, like Kulites.  

In the inlet duct, the flow is stable, there should be no wakes to produce a time varying static pressure field and so this method is very valid for inlet flow measurements.   The use of an ASME flowmeter design is more so that you can have confidence in your numbers because it has been very well documented.  I would agree that if someone was wanting to use the data for more critical reasons, a more accurately designed flowmeter might be called for.

~Chris

[racket]

Hi Chris

I was using a long ( 2 foot ) shallow conical diffusing jetpipe with flow straighteners, with measuring taken from the far end where velocities were down at below comp inlet speeds .

The Link I provided Patty with should give him an idea of whats required to produce accurate measurements .

Cheers
John

[racket]

Hi Patty

If redoing calculations isn't any fun, then you are doomed to perpetuating the same mistakes .

You are assuming that your vaned diffuser is producing the same compression efficiency as the vaneless turbo housing despite your "calculated flow" of 0.39 kg/sec being completely "off design" for the vaned diffuser system in place.

My calculated flow of 0.23 kgs/sec is purely for a choked NGV, it isn't for an entire turbine stage , this would require measuring the flow passageways of the exducer along with the exducer angles so as to draw up the velocity triangles and calculate the energy "lost" in the exhaust , this "energy" or dynamic pressure is part of the overall pressure ratio of the stage, you currently have a 2.18 PR going into the stage , an "average" 800 ft/sec exhaust velocity is worth an ~1.12 PR , this leaves a 1.94 PR for the rest of the stage , ie, the NGV and the turb wheel , if we subtract the turb wheel that leaves a PR unable to choke the NGV

Currently you won't be able to choke your NGV because it doesn't leave enough energy to get the gases through the rest of the stage , without the NGV being choked its difficult to do the calculations for its actual flow .

But lets assume we have an equal pressure ratio across both the NGV and turb wheel , to get the PR across the NGV we square root that 1.94 PR which gives us a ~1.4 PR for the NGV .

Using our 980 K T I T and a 1.4 PR from a 2.18 PR going in the throat PR is 1.55 .

Assuming an optimistic 90% nozzling effic , there will be a ~90 deg C temp drop to 890 K , gas velocity ~1500 ft/sec , area 0.00965 sq ft , gas density 26 cu ft/lb - 0.0384 lbs/cu ft

Mass flow 0.55 lbs/sec - 0.25 kgs/sec , we've picked up another 10% of flow over the previous choked nozzle , but your 0.39 kgs/sec estimate is still some 56% greater .

But we still haven't done the velocity triangles for the inlet to the turb wheel , with a 20 degree NGV and 1500 ft/sec gas velocity the blade tip shouldn't exceed ~1400 ft/sec which equates to ~116,000 rpm for a purely radial entry into the wheel , BUT , to produce horsepower to drive the wheel we generally need there to still be some swirl component "impulsing" onto the tip, this then reduces the optimum tip speed by some 10% or so , this would reduce the rpm back to ~104,000 in your case.

Then theres the restriction at the exducer , the "nozzling" through the wheel would only produce another ~1500 ft/sec plus any "left over" entry velocity , lets say 1,600 ft/sec at the exducer , with a 30 degree tip angle the axial exit velocity will be ~800 ft/sec .

Now 800 ft/sec axially out though an annulus of 3.7 sq ins - 0.0257 sq ft , gases at ~830 K say , so density at ~37 cu ft/lb - 0.027 lbs/cu ft .

Mass flow 800 X 0.0257 X 0.027 = 0.555 lbs/sec - 0.252 Kgs/sec

If you wish to "stick your head in the sand" and hope that miraculously the numbers will add up differently, you're mistaken .

Without the benefit of meticulously measuring every part of your engine I can only provide rough calculations , but theres no way you can choke your compressor with the size turbine wheel you are using unless you are bleeding off massive quantities of air , but if you were doing this your temperatures would be much higher due to the imbalance of mass flow rates through the wheels .

Ball returned :-)

Cheers
John

[madpatty]

Hi Racket,

You were against the bigger impeller from the day 1...reason was:-
It will cause surging.

I did tach vs P2.....it pointed towards NO surging.
You said check tach....i did that.
You said check P2 guage...i did that and added a second guage near the turbine end.

I did the Compressor end string test....pointed towards No surging.
And many more tests with the various impellers without any suitable result with this impeller shwoing the best data.

You want to say compressor cannot choke in this setup...i agree, Even when the whole data i have collected(same way any DIY enthusiast will do) points towards something else.

But after all these ball passing calculations.....you know what the best part is we have reached the consensus that my engine is flowing very well in the centre of the map....what else i need if you are agreeing with 0.22-0.25 kg/s flow....
The problem is when other people with similar setups are making same P2 at sub 89K rpm, i am making at well above 100K rpm.

Who needs choking or surging...i need a working engine.

Cheers,
Patty

[racket]

Hi Patty

I was against using the big compressor inducer because at the temperatures ( >1,000 C TIT) you were running at that time, you would end up in surge, the smaller comp wheel with the mods I suggested would have effectively "increased" your turbine wheels power relative to the compressor allowing lower temperatures to be used, but you insisted on using the big comp . .

For whatever reason you are now able to run at more normal temperatures ( 750 C TIT ) there is less chance of surge unless you add a jet nozzle to restrict the pressure drop across the turbine stage at which point temperatures will rise and surge will again be a consideration .

Your compressor wheel is a very sophisticated bit of engineering , and with its swept back blading it copes well with "surge conditions" to the point where it may not exhibit classic surge even when in a state of partial surge , modern compressor wheels don't suddenly surge like older radial bladed wheels .

You are assuming your tachometer is reading correctly , but that is only because of "remote testing" on a high speed motor , it is not an in situ test .

The Holset HX35 A77 turbo comp map I have indicates for a 2.3 PR a rpm of ~97,000 rpm , a 3:1 PR at ~114,000 .

Now we need to be 100% certain of ambient conditions at the time of the test ............

Ambient Temperature

Ambient Pressure

Are you at altitude ??

If so, what altitude ??

Cheers
John

[madpatty]

Ambient temperature- 20 degrees
Ambient pressure- 1003mBar, 14.54 Psi.

Cheers,
Patty

[racket]

Hi Patty

You didn't provide me with your altitude

Cheers
John