11/127 manned flight engine.

[ripp]

Wow!!!

I'm really looking forward to your analysis

Ralph

[Sweetenough]

Hi John,
Nice to see that she is up and running again.
Intreseting fenomena, do you thing the "ratteling" comes from the fuel supply or from the engine?
Also looking forward to your analysis

Kind Regards
Thomas

[racket]

Hi Guys,

She went into severe surge at ~5minute mark in the video ,it was rather loud , so I removed the sealing band I had around the surge slot exhaust , this allowed me to get up to 3.5 Bar before surge started again at ~7min 10 sec in the video .

Something is restricting flow , most likely the turb exducer as I only clipped back the exducer by a limited amount , changing the original exducer tip angle from a "tight" 25 degrees to a more normal 30 degrees , the old turb wheel was clipped back to achieve a 40 degree angle with extra throat area .

The old setup was at one extreme , the new setup is at the other , I need something in the middle .

Temperatures today were very low , down in the 400 deg C range at idle , indicating lotsa gas deflection , too much actually , I'd prefer temps ~500C at idle .

Otherwise , the engine worked OK , rather nice to get her up to a 4.5 :1 Pressure Ratio , thats the highest I've ever taken one of my engines :-)

The T2 minus T1 temps were also indicating poor comp efficiency which can happen if flowing too close to the surge line , another 10% flow might fix things .

Lotsa "numbers" to digest and calcs to make before I pull her apart to make any mods , the easiest modification will be a bit more ground off the exducer , maybe another 5 degree angle change .

Cheers
John

[pitciblackscotland]

Hi John,
Nice Now that start was nice and i did not see any glowing turbine wheel Yeb a bit more modifications and you will have a nice strong engine there me thinks

Cheers,
Mark.

[racket]

Hi Mark

Yep , its provided me with another "boundary" , I now know the "surge line" of the comp , so as long as I can move the flow a bit more to the right side , things should be OK .

That surge scared the crap outa me as I thought it was the engine getting into difficulties, something failing , at least I know my surge slot is working , but theres a limit to how far I can abuse it .

Hopefully the comp efficiency will improve with a tad more flow , which should help temps even more , lots to consider :-)

Cheers
John

[finiteparts]

John,

Interesting results. I had a few thoughts on the surge.

The first thing to consider is that the use of a vaned diffuser drastically reduces the map width and thus the surge margin, especially as you move toward the upper portions of the map. If you make a change in the turbine exducer I would suggest making a small trim only on the outer portion of the blade similar to how OEMs due....an angled trim on the maybe less than 30% of the exducer blade span. The largest impact on area is done on the higher radius portions, so that is why they target that area first. I am not sure that I would do this though, because I don't believe that the change in flow rate is worth the loss in turbine power and efficiency.

Another thought is that your impeller may be choking, especially since the surge slot made such a drastic impact and you are at the upper portion of the compressor map. Since diffuser choke usually occurs at lower PRs and the inducer chokes at higher PRs. If this is the case, trimming the turbine exducer will likely not help, if anything it may make things worse. Do you have any idea what the peak pressure ratio of the original impeller application was? It may be that you have reached the capability of the inducer design. If you run the numbers on the inducer vectors, what is the estimate of the inducer Mach number if you assume a zero incidence? That might help you understand how much more "room" you have to go.

It was good to see such a change in the performance due to your changes and it will be interesting to see how more changes impact performance. I think the another thought is that there are several confounding variables that were changed, so it may be harder to identify a single cause/effect. You changed the NGV throat area and you changed the turbine exducer area. It may be hard to say that the exducer area impacted the performance because it may be just as likely to be due to the NGV throat area change...which is actually more likely to control the engines mass flow.

Do you have an exhaust nozzle or is that just a straight pipe? If it is a nozzle, you might try a run without it to see if more mass flow helps, without the need to modify the turbine.

Anyway, great work and it is a real inspiration to us all that you've had such a success, even if there are still some learnings to be had. Keep up the great work!

Good luck!

Chris

[racket]

Hi Chris

Thank you for your thoughts , much appreciated :-)

I'm currently experiencing low comp efficiency , the T2 minus T1 is indicating effic. in the mid to high 60% , whereas earlier in the year it was ~10% greater.

Fuel pressures are much lower now than in the previous test , some of the decrease is from lower combustion temps but some from lower mass flows .

I did a set of calcs for the 4:1 PR numbers where I had 0.57 Bar in the jetpipe pitot at an "average" 660 deg C and 145 lbs of thrust , the jet nozzle is 118 mm dia , very oversized , the resulting velocity was probably ~1540 ft/sec , which would work out at a mass flow of only ~3 lbs/sec , rather than the ~4 lbs expected .

The comp temp rise was ~215 C degrees which would require ~180 C deg drop through the turb stage requiring a ~2.45 PR , our 4:1 PR becomes say 3.8 going into the turb stage , 3.8/2.45 = 1.53 coming out , my measured jetpipe total was ~1.57 , pretty close.

Unfortunately I have no idea what the comps heritage is , it was just an aftermarket wheel , but looking at it, it appears to be a "reasonable" normal comp wheel capable of 4:1 PR on a big diesel engine .

Its a 53 Trim wheel with inducer:exducer flow area ratio of ~1.7 : 1 , inducer throat area of 6390 sq mms , diffuser throats ~40% of inducer throats .

The previous NGV had total throat area of 6430 sq mms , somewhat "oversized" , new NGV ~5060 sq mms ,

The new turb exducer was clipped on a 7 degree slope which removed ~4 mm axially at the tip and changed tip angle from a tight 25 degrees to 30 degrees with throat increased from ~4800 sq mms to 5700 sq mms , the previous wheel was clipped back further to produce a 40 degree tip and a throat area of ~6590 sq mms.

Interestingly , my TV84 turb wheel was clipped to ~47 degrees ( 43 deg from axial) yet performed OK with a choked turb scroll at 4:1 PR , it was relying more on impact energy than reaction energy , I'm hoping that I might be able to replicate that scenario with even better results due to the higher comp efficiencies I was achieving earlier in the year .

My turbine wheel is "undersized" for the comp I'm using , a "marriage of convenience" that needs to be massaged gently to get it to work :-)

Thank you again for your thoughts

Cheers
John

[racket]

Hi Guys

Turb wheel has been back in the "grinder" to have its exducer clipped further , now back to where it was with the last wheel , the exducer throat is now 37% bigger than the original which had a comp flowing 3 lbs/sec , so hopefully big enough for 4 lbs/sec , .............I just need to find what needs to be done to get it work , the G Trim Garrett wheel used in the 12/118 engine had an exducer throat area of ~5080 sq mms and flowed ~3.3 lbs/sec , with 6590 sq mms , some ~30% larger , 4lbs/sec should be easy.

Currently my poor comp efficiency from running on the surge line is requiring more pressure drop through the turb stage than optimum , producing lower gas pressures/densities that require more flow area ,........I'm "on the wrong side of the curve" :-(

With flow a bit further away from surge I hope to pickup an extra 10% comp effic improvement which will mean less pressure drop required thru the turb stage and higher pressures and densities adding to the extra mass flow..........maybe :-)

If I'm still able to run relatively cool temps I might consider a tad more off the exducer .

Rotor off to the balancer tomorrow morning , she's boxed up and ready to go .

The engine looks much "happier" inside , less heat stress :-)

Cheers
John

[finiteparts]

John,

I take back my previous concern about you being near choke.  That was based on radial bladed compressors with vaned diffusers that really loose map width as you go up in PR.  I did some digging and found a very illustrative image in a paper by D Flaxington and E Swain (PCA Engineers, Lincoln, UK), called "Turbocharger Aerodynamic Design", (Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science 1999 213: 43)

It gives an image of a radial vaned compressor verses a compressor with backsweep, both using the same vaned diffuser.  No other details are given about the diffuser design of sizing, so, this will just be a data point for us to "understand" how the compressor map might change due to a vaned diffuser with a more modern compressor design.  Unfortunately, they didn't have it with a vaneless diffuser....that would have been an even better data point.  But I thought everyone might learn as much as I did from this comparison.



It is very illustrative that due to the impeller backsweep reducing the diffuser inlet Mach number and diffuser loading, the entire system can more stably operate with more variation in the diffuser inlet conditions (incidence angle, flow velocity, etc.).  Notice too, the speed lines shapes are less severe, which are typical indicators of when the system goes unstable on the surge side (the lines go flat or change to a positive slope).

Enjoy!

Chris

[racket]

Hi Chris

Yep , my comp doesn't have a lot of backsweep compared to some and might account for why the full blown surge came in rather suddenly rather than being preceeded by "rough air" .

LOL......maybe the poor old comp was doing its best and finally said , ENOUGH :-)

It'll be interesting to see what difference the extra exducer flow area makes as it'll be the only change made this time , it was unfortunate that I had to make two changes together with the last test , but it allowed me to test a "tight" system , so at least I know where that lies now .

Cheers
John

[dieselguy86]

So how EXACTLY does backsweep change things so much? If to get x pressure the air needs to be sped up to y amount, then how does changing the backsweep change that? If the air has to hit 1,000ft/sec to get 2.0pr (just pulled random numbers here) then how does changing the blade angle change that?

Can't remember which book I read it in, but I remember reading about slip factor, and I can see in the above picture that the wheel with backswept blades has to spin faster to hit the same pr as the radially tipped wheel. I'm just wondering how it not only widens the map, but also picks up efficiency.

-Joe

[finiteparts]

May 4, 2022 21:19:31 GMT -5 dieselguy86 said:

So how EXACTLY does backsweep change things so much? If to get x pressure the air needs to be sped up to y amount, then how does changing the backsweep change that? If the air has to hit 1,000ft/sec to get 2.0pr (just pulled random numbers here) then how does changing the blade angle change that?

Can't remember which book I read it in, but I remember reading about slip factor, and I can see in the above picture that the wheel with backswept blades has to spin faster to hit the same pr as the radially tipped wheel. I'm just wondering how it not only widens the map, but also picks up efficiency.

-Joe

Joe,

When you have a backswept impeller, for a given rpm, the absolute discharge velocity is reduced.  If you look at the vector diagrams, the relative discharge has a small aft velocity component that reduces the absolute discharge velocity, and thus also the total change in absolute tangential velocity across the impeller.  The amount of power that the impeller imparts to the air is determined by how large the change in tangential velocity is, so if there is a reduction of the total change in tangential velocity, it will produce less power transferred into the airflow and thus a lower PR.  It is very similar to the reduction of impeller work due to slip. 

Yes, with the reduced impeller work of a backswept impeller, that impeller needs to be rotated at a higher speed to make the same stage PR.  A pure radial bladed impeller produces more imparted work per stage, but does it at a reduced map width due to the lower stability of the speed lines at lower flows.

Due to the reduction of the impeller discharge velocity the diffuser is less challenged, which helps increase the system efficiency.  More diffusion is done in the impeller with backsweep, thus the amount of diffusion that is needed in the diffuser is reduced.

If I had more time, I would add a few sketches...but, time is very limited right now...so if it is not clear, I will try to add something later.

- Chris 

[racket]

Hi Guys

Its a bit quiet on the Site at present , isn't anybody doing anything ?? :-(

Picked up my rotor from the balancing guy this morning , hopefully have the engine back on the test stand in a couple of days .

Meanwhile I've been getting the afterburner electrics sorted in case they're required soon .

Hotstreak plumbing made up , with a 12V solenoid allowing combustor fuel bleed to be used .

Cheers
John

[wannabebuilderuk]

May 11, 2022 22:53:30 GMT -5 racket said:

Hi Guys

Its a bit quiet on the Site at present , isn't anybody doing anything ?? :-(

I've been pretending to be a jedi with my plasma cutter 😁

[racket]

Hi Ben

Nice to see progress is being made :-)

I've never used the plasma function on my machine, only the TIG , my nice MIG machine hasn't been used for a decade :-(

I like my oxy torch for a lot of jobs , it use to be expensive hiring cylinders , but thankful we can now purchase and swap which is much better for this DIY'er , I must have wasted enough money on cylinder hire over 40 years to buy a decent car .

Cheers
John