Contemplating an HX82 Turbojet

[monty]

Speaking of compressor wheels...

I've been working on the detail design of the engine components. I've also been playing around with different compressor wheels and seeing how it changes the turbine/ngv design. I have two compressors on hand I am looking at. One is the X856, and the other is X861. At first I ruled out the X861, because it's so much taller than the X856. Fitting it in the engine is more difficult. However, it has a 35 degree inlet angle and 102mm inducer. The X856 has a 106.5mm inducer with a 30 degree inlet angle. Believe it or not the X861 should flow about 8% more than the X856. It also has a better inlet/outlet diameter ratio and should produce higher pressure ratios. The lower tip ht of the X861 makes the diffuser design more of a challenge. There isn't any practical way to keep the diffuser throat aspect ratio of 1. It would require around 30 channels, and the vanes are too thin to allow for any pass throughs for fuel and oil. Everything would need to come through the engine casing. The other option is accept a lower throat aspect ratio of .6 and less length to width....so the diffuser efficiency would suffer a bit. Theoretically, the X861 should give about 10% more thrust....which is tempting to chase. I can just barely fit the taller wheel in the engine. The X861 has a slightly smaller inducer throat...and I'm not totally convinced it won't choke before the higher flow can be realized.

As far as the turbine NGV goes: Both wheels are pushing the limits of the turbine. I've run multiple "what ifs". Regarding clipped vs. not, I don't think either of these comp wheels is going to cause the turbine exducer to choke. The mach #s I get exiting are in the .4-.5 range. Clipping the wheel helps reduce exit swirl, but it forces the inducer/NGV to work harder. Balancing the NGV area and angle to match is where I think you have been struggling John. I've been iterating around the same issue with the spreadsheet. Clipping the turbine changes both the NGV angle and area requirement. The inlet incedence angles are pushing the limits for both comp wheels with the turbine clipped or not. A slight change can push things over the edge. Operating this close to the edge means small changes can be the difference between working and not.

For my calculations I have neglected the deviation angles of both the NGV and exducer. I don't think I have enough data to even guess at these numbers, and the literature doesn't give me any confidence they do either. So all of this is a big fat SWAG. However, the trends are instructive. How the variables play off one another is useful to know.

Clear as mud.

Monty

[ripp]

Hi,

an interim question, what effect does an NGV have a larger inflow area?







Thanks in advance

Ralph

[monty]

Ralph,

If I understand correctly: The larger NGV inlet allows the flow to transition smoothly into the nozzle. It will be more important for engines like you show, because the combustion can has been pushed to the minimum size. Missile engine designers are trying for the absolute smallest engine possible. I'm not that constrained. I can accept a slightly larger, slightly heavier engine.

Monty

[racket]

Hi Ralph

I'd love to be able to have a nice curved inlet like that , but I don't have any way of making it, or fitting it , in the space available , my vapouriser tubes are already very close to my NGV outer diameter, ..........just another compromise :-(

Cheers
John

[ripp]

Hi Monty, John, Chris and others,

How can I imagine that,

Does a larger inlet into the NGV allow the hot gases to flow out of the combustion chamber more quickly?

or just more even?

or a little more mass flow?, .....

Kurt Schreckling also used an NGV with a larger inflow area on his FD3/64



and with an existing system like yours 11/127 engine, what could change?

Thanks in advance

Ralph

[monty]

Ralph,

You might see a few percentage points gain in nozzle efficiency, but everything else will be the same. Instead of a 95% efficient nozzle, you will have a 96-97% adiabatic eff.

Not a huge gain in the overall scheme of things.

Monty

[racket]

Hi Ralph

As Monty said , theres always better flow if there isn't sudden changes in flow areas , the smoother the better , but sometimes its not possible :-(

Cheers
John

[monty]

Thought I'd post a quick update. Given my original goals, I have failed miserably.....

I finally got my spreadsheet working correctly. Confirmed now by plugging in data from a Solar T62 (thanks John!). That was highly instructive. Previous to that, I could not seem to make heads or tails of anything. I could never get the work, flow and other coefficients to line up with what is recommended in all the texts and papers I have been reading. I could come up with a solution, but it was horribly off all the charts and was never going to produce good results. When I plugged in the data from the T62, everything magically clicked into place! all the coefficients right down the line were "perfect". Obviously the guys at Solar knew a thing or three.. My brain also clicked finally. The whole time I was causing the problem! I was trying to push too much mass through the turbine. DUH!! I also discovered where the limits are and what variables change what. It's just as well I struggled so long, because I have been through the spreadsheet so many times I caught a few minor errors. But now I'm confident in the numbers it's spitting out. This allowed me to do a bunch of "what ifs??" The first big one was to revisit the turbofan I've been working on. The news was not great. I designed everything for 150shp. But with the Holset turbine.....I couldn't get more than 50hp until John metaphorically told me I was full of beans...Which prompted the T62 data, which led to my understanding a few things. Armed with that I could get 100-110 Shp with the Holset. Not really enough. I could do better if I upped the rpm until the turbine tip speed is 1700ft/s. That's the secret. Doing that allows for a choked NGV, while keeping the inlet triangles optimized. Unfortunately with the Holset this is way beyond the pitch line velocity my gear set design would survive....it's on the bleeding edge as it is. I need a TF15 turbine to make the turbofan work.....that could take a while it seems.

Ah well, the turbo-jet:

No matter how hard I tried, I just couldn't get too excited about the turbojet. I was getting some good numbers. Around 200-230 lbs of thrust. The diffuser looked good on paper......But the FUEL BURN!!! 35 gal/hr.!!! Yikes..The engine....she was a big girl with quite the appetite.....Not really my type. Kinda homely too. I just wasn't feeling it. 

So I ran another what if....based on that missile engine posted earlier (thanks Ralph!) and the numbers were promising. I did a few calcs on flow areas. Looked for suitable comp wheels, and stuck some parts in cad to see the general layout of things. THIS I LIKE!!

No gearbox. Fan is direct drive. The comp nut is a Hirth coupling. The starter/generator motor is correctly sized and also uses a Hirth coupling. (same one I was planning to use for the turbo-jet.) I have incorporated inlet swirl calcs, so I can match the inducer of whatever shows up from KTS. WHY CAN'T THEY PUT INDUCER ANGLES IN THE CATALOG!!??

The components are all small and easy to make compared to the parts for the big fan. I can print the casting forms and use ceramic shell. MUCH easier, SOOOO much smaller. I'll need to switch from a .8mm to .4mm nozzle on my printer!

The fan is going to be the hardest part to make. But I can come up with something, even if I have to pay to have it machined.

Rough numbers:

Fan: PR 1.22 Mdot 6.5 lb/s Diameter ~ 6in. 65krpm. ~75 shp.

Core: PR 3.2 Mdot 2.1 lb/s Diameter = too small...but, 2 lbs is easier to pass than 4+ and the modest pressure ratio means the diffuser can pretty well suck and not hurt things too much.

Engine Bypass ratio ~3. Overall PR ~3.9 It burns at least half the fuel of the fat girl...and should make around 200lb thrust.

It's an interesting design. The Fan limits rpm, so you give up some turbine performance. This sets how much power the fan can take. I've come to think of these turbines as having an expansion budget. You can decide what to spend it on. Fan, Compressor, or Nozzle. The fan was fixed because of RPM, so the game was get the best combination of PR and mass flow for the size, and spend the rest on the nozzle of the core.

To get this to work, I've clipped the turbine tip ht from 19 to 13. The exducer is also clipped to 45 deg. using a curved clip. Not because it's choked. It's to set the flow and work coefficient and limit exit swirl. It finally occurred to me that flow coeff. for a jet is going to be higher than a turbo-shaft or turbocharger. As long as all the other numbers are good, this doesn't matter too much. The meriodonal velocity ratio is all screwed up though. Makes it hard to get good inflow conditions. This combo put the numbers in the sweet spot with the flow just barely pushing the high side.

Design build thread soon.

Monty

[monty]

I thought I might explain the reasoning for going to the fan vs turbojet. My calculations show that the radial inflow turbine is not ideal for a turbojet. When you start trying to optimize for thrust the flow coefficient becomes too high, and the work coefficient too low. The specific speed gets out of hand and starts pointing towards a mixed flow or axial flow turbine. The option is then to increase the pressure ratio to get things into the happy zone. This results in a PR~7....too high for a single stage radial comp wheel. An additional axial stage is needed. This creates a high PR low mass flow engine...and low thrust. This configuration is fairly efficient...but not of much practical interest to me. An efficient 150lb thrust engine might be OK for a missile or some other application, but nothing I'm interested in. This is why axial turbines predominate turbojet designs, and radial inflow turbines predominate the turboshaft designs. The solution is a fan of some sort with bypass. This gets the work and flow coefficients in line for the radial turbine. Unfortunately, the direct drive fan limits the rpm of the turbine, so you wind up with a compromise. It would be better to have a geared fan, but the gears become an issue at the rpm this turbine wants to operate. So I have maximized the fan power and flow for the rpm, and let the rest of the power provide core PR and flow. The turbine will be a bit high on the flow side of things, and low on rpm and tip speed....but this is the best compromise I can find for this off the shelf turbine. A larger diameter turbine is needed to make the geared fan a reality.

Monty

[racket]

Hi Monty

I've sent you a PM about TF15M turb wheels :-)

There are titanium comp wheels for the HX83 ;-)

Yep , for thrust we need lotsa mass flow ..............bigger hole in the front .

Cheers
John

[monty]

What do you guys think? Any way this noodle of a rotating assembly survives?

I switched to all brass because the ball bearings were looking marginal with the thrust from the fan, plus I needed to minimize the hub diameter of the fan. By the time you put a squeeze film damper and jump through all those hoops....might as well just use bushings. Small cheap and bullet proof with built in damping.

I looked into calculating mode shapes and critical speeds....But I don't think I have enough data. Too many unknowns.

[racket]

Hi Monty

Thats gunna have a pretty high Mach number at the tip of the fan , ~ M 1.6 ??

Brass is good :-)

Cheers
John

[monty]

John,

Inlet guide vanes deflect the air and reduce the relative mach to ~1.2. That's a pretty common number for the first stage of axial compressors. I have an example of a transonic fan design in Hill and Peterson with M1R of 1.4. The copyright date on that book is 1992. I'm not sure what the state of the art is for transonic stages these days, but considering some of the faster biz jets are doing M.9+ at 50Kft......it must be impressive. I don't think that's something I can replicate in my garage though! I'd like to get the fan mdot up around 7 lbm/s but that starts pushing the M1R up to 1.3+. Thus my focus on reducing the hub dia. to maximize mass flow.

I have decided to go ahead and exploit the max mass flow for the core turbine if possible. I'm not going to trim the tip ht. unless I have to. As long as the inflow angles, specific speed, and work coefficient are in the acceptable range, I'm going to let the flow coefficient drift up as far as possible. That unlocks quite a bit of performance that would otherwise be on the table. Letting the core be the best jet it can, given the limits of the turbine...while still driving the fan. This lets the fan power go up to around 110 hp. Unfortunately M1R limits the mass flow, so the extra power only goes into pressure ratio. But the overall engine pressure ratio goes up and it helps with fuel efficiency. Plus can help offset pressure loss in the jet pipe. I could squeeze a bit more out of this if the fan and core nozzle are expanded to ambient. Right now I have them discharging to a mixer and going out a common nozzle. It's hard to do separate nozzles unless the engine is in an external nacelle.

The radial comp wheel will be the determining factor. The inducer inlet flow angles needed to do this are in the 35-40 degree (from tangential) range (55-50 in axial aerospace parlance). Especially given that I'd like to limit stator turning in front of the inducer. I'd like to allow around 7-10 deg of swirl to simplify the blade shapes.

I'll see what turns up in my mail box this week....

Monty