PD-01 Jet Engine

[madpatty]

Hi Racket,
Altitude is 219 meters above sea level.

Cheers,
Patty

[racket]

Hi Patty

So we can work on using a Standard Atmosphere , 3:1 PR will be ~ 29.4 psi gauge .

Tach might be reading "reflections", also if under AC lighting sometimes this can interfere with readings .

Maybe "blacken" everything in the inducer area with a matt black spray paint other than the one "reflective" surface you use for providing your tach readout from, this will remove any chance of reflections, and do the reading in natural sunlight , no artificial lighting .

If this still produces aberrant rpm vs P2 figures then I have no idea what the problem could be other than engine components not working at their optimum .

There is one other thing you could try to lower your temperatures a tad more and that is fitting a conical diffuser exhaust pipe , ~10 degrees included angle ~150mm long with inlet ID of 58mm to match your turb exducer , this will convert any dynamic pressure exiting your exducer into static pressure and increase the pressure ratio across the turb stage from less exhaust wastage .

Cheers
John .

[finiteparts]

Apr 3, 2015 16:03:46 GMT -5 racket said:

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

John,

That comment didn't help your case.  I expected better from you!  Ha!  

Diffusing wake flows makes them worse.  Putting struts in, even to "straighten" the flow makes more wakes, which makes things worse....diffusing flows in general makes things worse...none of those things would have helped reduce the variation in the static pressure field in the exhaust flow. So I repeat my statement..."What you found out was that measuring flow speed in a "jet pipe" was problematic...this is different. "

I agree with you though...I think the inlet should be bellmouthed to reduce the entrance loss...or the other option is to make it's length at least 10 time the ID of the tube to mix out the vena contracta.  We should work this out, because everyone on this site could benefit from a good mass flow measurement at the inlet.

So, I did some digging in a few books I have " Principles and Practice of Flow Meter Engineering" and "Aerodynamic Measurements" and found that my idea of measuring in a low speed flow region is wrong...well maybe not wrong, but definitely not a good idea.  

Mass flows of gases are ideally measured with choked nozzles.  Choked nozzles give the most accurate conversion of pressure measurements to velocity and thus mass flow.  But, since many flows can't take a choked nozzle upstream of the device, there is a class of flow meters that operate in subsonic flows, which include plate orifice meters, venturi meters, etc.  The trick here is to measure the flow at higher speeds because the static pressure to total pressure difference is larger.  So, because you will always have some amount of error in any pressure gauge, the effect of cascading that error through the calculations is larger at low velocities were the measurement error is a larger percentage of the differential pressure measurement.  

So, instead of measuring in a large, expanded passage like I did on my home-brewed mass flow meter, I would want to measure in a converging section or in a throat of a venturi.  This has the benefit of getting the flow velocity up to have the larger static/total pressure difference, plus the accelerating section upstream will suppress boundary layer thickening.

By taking a single measurement of the gas velocity and assuming that the entire cross-sectional area is flowing at that velocity, we are inducing an error due to that assumption.  Measuring the gas velocity is very achievable and should be fairly accurate, but using that to calculate a mass flow introduces more errors.

There are devices that measure at multiple points and average the gas velocity to reduce the error that should be relatively easy to make or buy.  

www.dwyer-inst.com/Product/Flow/FlowSensors/In-Line-AveragingPitotTube-/SeriesDS

Other devices have a known calibration factor (K factor) which reduces the velocity from the center measured velocity to a more accurate average velocity. Finally, some reduce the flow area with a coefficient that corrects for the blockage of the boundary layer and this corrects the mass flow measurements, since the flow in the turbulent range has a very straight bulk flow velocity profile, this works well in that regime.  Getting a close K-factor should be achievable especially if someone works it out on a standard turbocharger setup with a good compressor map....worst case scenario, base it on a similar known device K-Factor.
 
So, the page you provided is an excellent resource on a modified version of the standard ASME type nozzle designs, which measure the static pressure of the inlet on the wall and the total pressure is just ambient pressure, since no work has been done on the fluid prior to measuring the static pressure.  Using static pressure measurements at the wall or in the flow is equally common in various styles of flowmeter design, so that shouldn't be any real issue.  The wall mounted static pressure taps require very accurate hole creation (no burrs, sharp edges without chamfers, exactly 90 degress, etc) so as to not skew the measurement, since they are measuring at the base of the boundary layer and there are guidelines out there showing this.  The key thing here, for any of the flowmeter designs, is to create a flat velocity profile so that one measurement can accurately represent the bulk flow speed and know how your static pressure behaves compared to the free stream flow.  

The pitot is easy to come by and fairly cheap to get ASME types....and you have the benefit of having confidence in the velocity measurement at the point your probe is at, thus making a velocity profile is relatively easy.  Measuring static pressure from the wall, you have to infer a velocity profile.

technodepot.com/product/dwyer-instruments-160-8-series-160-stainless-steel-pitot-tube-8-5-8-insertion-length?format=visible&source=ads&gclid=Cj0KEQjwl_6oBRDHxNGz6ueJufMBEiQAvm_k_lYpi97eOKPhN0PDz1l2Bks9t5ffACO2H-veftOsV5caAk6K8P8HAQ

A homemade one will have errors, but there is no reason that it can't be +/- 10% if done well...and it beats people guessing where their engines are running. I doubt everyone needs the level of accuracy of flow measurement that is provided by the one in the Scaled Jets link that you provided.  I mean look at the thermocouples and pressure gauges people are using...just as far from calibrated gauges with proper design installations.  

~ Chris

[racket]

Hi Chris

LOL, ...............we are talking 14 years ago in 2001 when I did the measurements with the "long" pipe , I'd only been at this DIY turbine game for a bit over a decade at that time, there wasn't much info around , no Internet where everything is now freely available, it really was "trial and error" development as I never ever had a comp map for the TV84 , comp maps were extremely rare , only authorised Dealers had access to them .

I made up the "long pipe" to get more accurate total pressure and temperature readings so that I could have more faith in my calculations for what I felt was the rather pathetic thrust reading of "only" 50 kgs/110lbs that the TV84 turbo was producing , the static vs total readings were a sideline by that time due to the wildly "inaccurate??" sets of numbers I'd previously been producing .

On the 19 the December 1999 I recorded static at 5.5psi and total at 15 psi whilst thrust was in excess of 50 kgs/110 lbs , this was when using a "small" pipe, ID of pipe matching exducer diameter, with flow straighteners as well as a central tube and "bullet" in the jetnozzle centre , this pipe gave the best thrust of all my experimental pipes, it was only ~6 inches long .

Eventually I realised that I'd been getting maximum thrust for the previous few years , just didn't know it at the time :-(

Yeh , a nice inlet flow meter is the easiest way to go , and I agree with you that we can't be too pedantic about "accuracy" , I stick to a plus or minus 5% as as good as we can do with the instrumentation we have .

I was very particular about thrust measurements though, as taken in conjunction with other data readings we can workout just how well the engine is performing.

Cheers
John

[madpatty]

Hi Racket,
Yet another test run...this time with all the blades colored matte black except one painted white.

And results:

P2(Psi)            RPM

7                    71000
10                   81300
12                   88000
14.5                95600
15                  97300
15.5                98100
16.5                A2000
17                  A3800
18                  A5500

Readings are similar to the previous ones so i can say i was reading correctly. 

Same rpm same everything still temperatures higher this time as i can see the turbine wheel glowing reddish color.
That's really annoying if we can't interpret the behavior of the engine now. Because P2 vs RPM are telling a different story.

Cheers,
Patty

[racket]

Hi Patty

OK, lets look elsewhere .

You are running an 83mm comp exducer similar to the Garrett GT35 www.turbobygarrett.com/turbobygarrett/turbocharger#GT3582R (1)

At your 7 psi P2 - 1.48 PR you have 71,000 rpm , the Garrett is at ~60,000 , 18% more rpm for you

10 psi - 1.68 PR at 81,300 rpm the Garrett ~70,000, ~16% diff.

14.5 ~2 PR at ~96,000 , the Garrett ~85,000 , ~13% diff

16.5psi 2.12 PR at 98,100, the Garrett ~87,000 rpm ~13% diff.

18 psi 2.22 PR at 105,500 , the Garrett ~92,000 ~15% diff

An average of ~15-16% difference across the range...............some relative consistency .

But really theres not much I can say about why you are down on pressure .

You could try measuring your T2 temperatures to see if you can find an aberration there as well , at a 14.7psi P2 - 2:1 PR you should have a "stabilized" T2 of ~107 deg C on a 20 C day , if it is then your P2 gauge is working OK , but if its a lot hotter at >130 C then I'd say your gauge isn't correct and you are producing a higher PR .

Without a complete inspection of your engine to determine potential areas that could be causing you problems its very difficult to make a decision , it could be anything .

Have you measure your comp wheel to shroud axial clearance between the 65mm to 83mm diameters , the last bit of the "tip" ??

If so , what measurement did you find ??

Cheers
John

[madpatty]

Hi Racket,
During the making of the compressor shroud i used the method of repeatedly putting on the part on the engine and checking if it is touching the compressor wheel by pushing the wheel on all sides as well as pushing the shaft from the turbine towards compressor side....
When i started to touch then a last final POLishing cut to complete the part.

Today i will fitting the T2 pickup at a point near the diffuser outlet....

Cheers,
Patty

[madpatty]

Hi Guys,
Another successful spoolup of the engine and Temperatures also back to Normal.
There was a leaking gas injector that must have causing the TOTs to rise due to fuel burning at unwanted places.

Also did a few tests with manometer again, this time with a bellmouth at the manometer inlet.
As finiteparts(Chris) wanted to see the effect of the bellmouth on the manometer readings so i went for a slide fit type bellmouth which i can add and remove to the running engine.

Some pics of the bellmouth setup:-








Unfortunately gas pressure didn allow me to go further 15 Psi P2 but still i manged to get a few readings. Also fitted a T2 Guage at a point near the diffuser outlet.

Here are the few readings that i collected:-


P2         T2           TOT           Manometer(without bellmouth)          Manometer(with bellmouth)(mm of water)


5.5        77           556            58

7           101         537                                                                               78

10         112         538            130                                                            120

12         123         552                                                                               152

13         126         553

14         134         557          

14.5      136         561            186                                                            156

15         136         560


Ambient temperature:- 25 degrees, pressure:- 1013 mB


Here is the video showing the behaviour of the manometer water level with addition and removal of the bellmouth.





Cheers,
Patty

[racket]

Hi Patty

Those T2 readings are "interesting" :-((

At 7 psi P2 it appears that the comp was running at 67.5% effic , at 12 psi its dropped to 56% and only marginally better at 14.5 psi at 58% , either your P2 gauge is under reading by a large amount or your compressor stage isn't working very well .

What is positively interesting is your T O T s staying in the mid 500s , I'd have expected them to be higher if the comp was running at 56% .

OK lets do the maths for 12 psi - 1.8 PR , 98 C degree rise in the comp would need a ~85 deg C drop in the turb , if so, your T I T was 552 + 85 = 637 C - 910 K , turb effic would be no more than 80% , you'd be needing a 1.64 PR to drive the comp, 1.8 PR X 0.95 = 1.71 PR going into the turb stage , leaving only a 1.04 PR for your exhaust gas dynamic energy , thats only 0.6 psi , not enough energy to produce the required speed to exhaust your gases.

Conclusion , your PR gauge isn't reading correctly based on the other data you provided

As for your drop in manometer readings with the bellmouth in place , your airflow into the straight section of tubing is across the entire cross section and at a lower speed that the "throat" speed of the vena contracta on just the bare straight section .

I would have expected to see a greater "height" in your manometer .

Cheers
John

[madpatty]

Hi Racket,
That was the same pressure guage which i checked at 5,10,15,20,25 & 30 Psi with a reference calibrated unit....

What TOT were you expecting??can you specify me a range?

Cheers,
Patty

[madpatty]

The video of the test run:-




Cheers,
Patty

[racket]

Hi Patty

If your P2 gauge is indicating correctly then you have a compressor stage problem and/or your T2 pickup is registering an incorrect temperature , and/or you are running in semi surge situation , which is a compressor stage problem .

At a 2:1 PR you should be having your best compression , and it should be at at least 74% efficiency , in which case you would have a T2 of 113 deg C with a 25 C ambient temp .

I would have been expecting a TOT of >600C with such poor compressor efficiency , my use of an 80% turb effic was being optimistic considering the NGV construction, a lower turb efficiency would have required a higher pressure ratio across the turbine stage, effectively using up all of the available pressure leaving none for the dynamic exhaust flow , but with a higher gas temperature the "balance" would have been "acceptable".

Its absolutely critical that any measuring is done to the highest standard possible, taking every effort to minimise aberrant readings , your TOT readings should be taken within a jetpipe some distance downstream of the turbine wheel to allow mixing of the flows producing a more accurate reading of "average" TOT .

From what I understand of your inlet manometer readings there was only a difference of ~0.25 psi ( ~6 inches of water) between static and total pressures , 0.25 psi represents a very low airspeed which also means not a lot of air going into the engine possibly indicating "surge" , or at the very least a compressor stage thats not processing air as it should .

Cheers
John

[madpatty]

Hi Racket,
But 6 inches of water column in my case signifies almost 0.31 kg/s of airflow, which is my design airflow at 3.0 PR.

cheers,
Patty

[racket]

Hi Patty

But you are assuming that the airflow rate is even across the entire cross section of the inlet tube that your pressure pickups are mounted in .

The fact that they are mounted within a tube >2 times greater in area than the inducer as well as relatively close to the inducer would indicate to me a large potential for inaccuracies .

If those pressure pickups were several inducer diameters away from the inducer then those readings would more accurately indicate flow , but there would still be a velocity gradient across the tube.

One more thing to try, weigh your fuel flow rate at a steady power setting , sit your gas cylinder on a set of scales and note the weight of fuel flow per minute at say 12 psi P2 .

You have your ambient temp , your T2 temp and your TOT temp so its easy enough to calculate the temperature rise in the combustor .

The temperature rise will require a certain amount of BTUs/minute/kg of air flow rate.

Currently you have a T2 of 123 C and lets assume my calculated T I T of 637 C , so a 514 C temp rise , with Sp Ht of 0.24 and 925 F deg rise you'd need 222 BTU's /pound of air per second . or 13,332 BTU's/min /lb, assuming you are flowing ~0.5 lbs/sec -0.22 kg/sec , then you'll only be needing ~6,500 BTUs/min .

Now I'm not certain of your fuels BTU rating , but you should be able to find that out and do the calculations, just be aware that your fuel burn efficiency may not be 100% so there could be an indication of a greater airflow rate simply to consume the fuel flow rate .

This will only produce a rough outcome as we haven't factored in everything but it will give you an idea of whats happening .

Cheers
John

[madpatty]

Hi Guys,

I had a question regarding running of the engine on liquid fuel(diesel).
What i am doing till now is testing the engine on gaseous fuel only so as to eliminate any liquid fuel related problems and i will continue to do that until i get her fully sorted out(as far as Max. P2 and RPM are concerned).

But as i have tried twice to run her on liquid fuel and both the times The result was melted NGVs. So i thought maybe the method i was adopting is wrong. 
What i currently do is i have two separate fuel manifolds, 16 fuel injectors(small copper capillaries) each as i have got 16 evaporator tubes in my combustion chamber.

fuel injector on diesel manifold has 0.5 mm ID and on gaseous fuel have 1 mm ID.

i am feeding gas to all the evaporators rather than to 2 or 3 evaporators for the preheat purpose.Its not actually preheating but complete start of the engine on gaseous fuel and once it is idling at say 4-5 Psi P2, i start feeding the liquid fuel with simultaneously reducing the amount of gaseous fuel till the point when the engine is fully running on liquid fuel.

Is this method OK of feeding the liquid fuel once after the engine has started(self sustained) or this way the liquid fuel is being blown away too fast from the evaps without being fully vaporized due to high velocity air??
 what just came to my mind was Maybe when we start feeding liquid fuel from the very beginning, just after preheating(when the air speed inside the engine is low yet) then the flame gets some time to find a spot to pivot itself in a region of slow speed air and then engine can be throttled up from there.  


Cheers,
Patty