Thrust Vectoring Design Project - Unleash Your Ideas!

[Aero Stud]

Hey people,

I've started a group undergrad project that involves designing, and probably building I hope, an addition to a small jet engine, for thrust vectoring. I got a few ideas already, but of course a few more from all the jet enthusiasts out there won't hurt. More specificly we prefer to smoothly divert the flow with some type of sleeve rather than just stick some fin in there. And, maybe also use some composite materials both for innovation and as a possible feature of heat signature reduction.

Don't be shy to shoot crazy ideas, I like those. Thanx.

 

[Danger]

Don't be shy to shoot crazy ideas, I like those. Thanx.

Glad to hear it; that's my specialty.
Okay. To start with, I only know of three ways that vectoring is currently achieved: 1) gimbal the whole engine (or the wing that it's on), 2) gimbal the nozzle only, or 3) introduce steerable vanes into the exhaust flow.
I'm sure that you're more than familiar with all of those. I have an idea for a possible approach, but I have no idea whether or not it's possible, let alone practical. Here goes, anyhow:
I'm thinking of bleeding maybe 5-10% of the combustion gas into a manifold plenum before it gets to the nozzle. (Or, perhaps more practically, compressor efflux before the combustion chamber.) In any event, four runners lead from the plenum to nozzles aimed inward from the edges of the main exhaust nozzle. Each has its own valving tied to the control system. As the valve opens, it introduces the high-pressure gas into the exhaust flow at a tangent, thus diverting it in the opposite direction.
Any chance of that working?

 

[FredGarvin]

Danger,
You have a pretty good idea in using compressor discharge (CD) for a source of motive power. However simply ejecting it into the flow won't do terribly much. In order for something like that to work you would need a source with higher flow and pressure than the turbine exit. CD is only a small percentage of the total engine flow.

You have to look around and see what is currently being done. There are reasons why things are as they are. Look at the F-22 and the JSF. Those are the two leaders in North America and have two different approaches. Honestly, I would expect that something like what is done on the F-22 would be the easiest to adapt to your situation. If you get an operational vectoring device, you will come to find out that erosion of the surfaces and required coatings of the inner surfaces could provide you research work for the next decade.A word of warning: Depending on the engine you select, many engines are sensitive to back pressure effects. You need to be aware that diverting exhaust flow may increase the back pressure on the engine which in turn will, at the least, cause an increase in TIT and problems lighting off. Worst case you could also induce a surge. You need to be very aware of your engine's mechanical limits. Proceed slowly and deliberately.

 

[Danger]

you would need a source with higher flow and pressure than the turbine exit.

I suspected as much, but was rather hoping that I was wrong.

 

[Aero Stud]

Thank you Danger and FredGarvin.

But Danger, like I said, we're using an existing small jet engine. So we're not really planning to take it apart, rather just make an addition to it.

A word of warning: Depending on the engine you select, many engines are sensitive to back pressure effects. You need to be aware that diverting exhaust flow may increase the back pressure on the engine which in turn will, at the least, cause an increase in TIT and problems lighting off. Worst case you could also induce a surge. You need to be very aware of your engine's mechanical limits. Proceed slowly and deliberately.

Of course, that's one of the things that are important to us. And part of the reason we don't want any fins or something that will directly interfere with the flow. Either way we'll plan and study the changes in operation (the flow, thrust and so on) due to the thrust vectoring. Our engine has a bypass pressure ratio of just 4 and nowhere near to be choked though, so it's not a major problem, just a matter of efficiency.

If you get an operational vectoring device, you will come to find out that erosion of the surfaces and required coatings of the inner surfaces could provide you research work for the next decade.

Speaking of which, we'll probably need an elastic material and like a said it better be composite. So I searched the web a bit and found this: http://www.epp.goodrich.com/fyreroc/curr_products.shtml (I was first looking for the different materials that make up the protective tiles of the space shuttles because they encounter conditions which could be considered similar in general, and realized they probably won't work here - we need something both elastic and as thin as possible, so I found that this company makes those as well as the material in the link.) So, maybe you heard about this material or could guess whether it's something that can be put in the exhaust flow without eroding (too fast...).

 

[Danger]

How much deflection are we talking about here, in terms of degrees? While it won't do anything for your IR footprint, I suspect that gimballing the engine might be your most practical option. Maybe something like my original suggestion, but just using scoops on the exterior of the plane, would suffice to cool the exhaust a few hundred degrees.

 

[Aero Stud]

The whole IR thing is just a possible secondary objective. Anyway, if we divert the flow at relatively large angles inside some geometry then we can always consider adding some reflective material on the outside. Regarding gimballing the engine, we have to work on the engine itself actually, so it's a good idea but probably won't fit us.

About the angle, we have no restrictions for now. Our proffesor was generally talking about a 10-15 degrees range but did not limmit us, I just find that boring and not ambitous.. because I'd love to install it on a model aircraft later and have it take off vertically or do some impressive maneuvering.

 

[FredGarvin]

Unless you have a crack team in controls system, I'd forget about that last part.

 

[Danger]

I'm going on the assumption that this thing is non-afterburning. Can you describe or post a picture of the arse end so we can see what you're attaching to?

 

[Aero Stud]

http://www.amtjets.com/
It won't let me put the link directly, click "Specfications" on the left. Out of the 3 engines, it's the "largest" one on the right - Olympus HP.

Unless you have a crack team in controls system, I'd forget about that last part.

Nope we don't really - those guys are doing a different project, and I really don't like control theory myself either, but when it's something real instead of just some weird math, it's a different issue. And I think it depends also on how you divert the thrust and how good is the person flying it. We'll see about all this later.

 

[Danger]

Thanks for the link, Aero. I didn't realize how small the engine is. Since the only good photos of the rear section were all Pegasus models, I'll just assume that the Olympus is similar and start thinking on it some more.

 

[Aero Stud]

Thanx man.

 

[Danger]

Okay, I have something. Unfortunately, I can't get ImageShack to put it up properly.
What I'm thinking of is using braided stainless steel (as used to sheathe hoses) to form a flexible sheath attached to the exhaust nozzle. Tri-axial activators, such as pulling cables or pulling/pushing cylinders can deflect the sheath by the same sort of method that remote control car mirrors work. I'll keep trying to get the diagrams up.

 

[Aero Stud]

NOW you're talking !

I just went with the composite materials because that's what I prefer and also on the way neglected the thought that using metal doesn't neccessarily mean it has to be solid. I had an idea at first about using a construction made of small shutters from steel (partially like on windows, just that those are made of plastic), but the circular geometry ruins it.

This is good. Because it's elastic and I can fold it - which is great both to save space and primarily do what I wanted - divert the flow smoothly and with possibility of large angles. About how to open it, I think I know that already from an idea I had so far, using two different servos. But of course, I'll be happy to see yours as well. I have a different follow-up question though - do you know how flat can you make the surface of braided stainless steel ? Because obviously to divert a jet requires it to be as smooth as possible. I'm going to do a search.

I wonder if anybody heard of steel cloth, that's elastic anough and thin anough, and fit at least up to 1400F. I looked for it before and couldn't find something like that.

 

[Danger]

The surface texture was a concern for me as well. It's suprisingly smooth to the touch (similar to fibreglass cloth), but I don't know enough about fluid dynamics to even guess at what it will do to the exhaust flow.
The reason that I suggested 3 actuators is that those will allow you to flex the sleeve in any direction, not just up and down or sideways. If you only need one axis of movement, 1 or 2 will suffice.

 

[Aero Stud]

ok.. you mean 3 axial actuators - along the main axis on the sides of the nozzle ? My idea is to save some material and leave anough to open a sleeve of equal diameter to the solid exit nozzle, up to 90 degrees or less, by turning a circular frame that the sleeve section is attached to and at one side is connected with the solid nozzle exit the whole time, or accordingly move it back and roll or push the sleeve back around the solid nozzle on the outside (depending on the material). Now using another servo you can turn this whole cap with the first servo around the outside of the solid nozzle so you can deploy it at any angle you choose.

Do you happen to know the particular type of braided stainless steel we're talking about, the firm, a hose example or something like that. :)

And of course, thank you Danger. :)

 

[Danger]

ok.. you mean 3 axial actuators - along the main axis on the sides of the nozzle ?

Right; the sleeve would be welded, bolted or clamped to the original nozzle as a tailpipe extension. The movement would be similar to an elephant trunk or snake. I was assuming that you would want instantaneous response, to be tied in with control surface movements. If I understand your idea correctly, it seems to be quite complicated mechanically and slow to get into service.
I have to get ready for work now. I'll log back in later.

 

[Aero Stud]

I see what you mean. However I see a problem here, same as with most ideas of this type including mine - in that when you turn this sleeve the axial length has to vary radialy. So then something has to be done with the excess material and the axial crossection changes and all that unwanted stuff. But... this gives me another idea. Say, if I connect another solid section at the end of the nozzle.. I have to think about this some more.

 

[Danger]

Well, my first thought, before the braided cloth, was to weld a rounded collar to the outer surface of the existing nozzle. A two-piece pipe with a matching circular depression on the inner surface of the end could then be welded together around it. The function would be similar to a spherical rod end, with the same sort of actuators as the braided design.
Here's a link to a braided stainless supplier: http://www.n-seisen.co.jp/english/products/fabric.html"
I'm afraid that I don't understand what you mean by the axial cross-section changing radially. If it means what I think it does, I can see no way to eliminate it with any design.
I'm still working on the drawings anyhow. Maybe if I transfer them to this computer I can get them accepted by ImageShack. Still a few details to add first, though.

 

[Danger]

As usual, I'm probably way too late to be of any use even if you like the design. I just finally managed to get ImageShack working by using tif files instead of jpg. Here goes:
"[IMG[/URL]

 

[tospok]

hi

Hi guys, its seems you have some problems about materials. First of all I do not know the temperature range at which the nozzle operate but guess its 500 - 900 C, than creep will be a major problem and materials strengths at elevated temperatures decreases dramatically. Thus using a thin flexible sheet is a dream for me. one way seems using some layered structure and using some internal air streams to prevent direct contact of material with hot gases. The layered structure can be some ceramic coating on top of a super alloy but thermal expansion coefficients must be similar. If the temperature is lower your design may work, but I have never seen an elastic thrust vectoring (but i don't know much about thrust vectoring too). I hope I could help a bit.

 

[Danger]

Welcome to PF, Tospok.
Thanks for pitching in. Any input whatsoever is valuable throughout the forums, as long as it's based upon physical possibility rather than some bone-headed personal theory. I must admit that matching thermal expansion characteristics never even occurred to me. Good point. It seems to me that you are going to be a very helpful addition to our community.
I fear, however, that both of us are too late for this particular project. I'm glad that you pulled this thread out of mothballs, though, because I'd really like to know what ended up being done and how well it worked.
How about it, Aero Stud? Update, please.

 

[kach22i]

This UAV/Model has some simple thrust vectoring you may want to look at (and it's a cool site).

http://www.canosoarus.com/05UMAAV/UMAAV01.htm

 

[FredGarvin]

It's a completely different thing to direct the thrust of a ducted fan. Add a fluid static temperature of 800° or so to the equation and things get difficult.

 

[kach22i]

It's a completely different thing to direct the thrust of a ducted fan. Add a fluid static temperature of 800° or so to the equation and things get difficult.

Correct, which might be why a few of the designs I've seen published actually bleed in outside cooling air (secondary mass flow) prior to any attempt to vector the thrust.

This one (below) is using a secondary flow for other than cooling effects.

http://www.geocities.com/m_mason007/home.html

Follow-up:
http://flux.aps.org/meetings/YR97/BAPSDFD97/abs/G4600006.html

 

[FredGarvin]

Yes. Bleed air and secondary flow has a well established use for cooling. It has been used in burners for quite some time.

 

[minger]

May be a little late, but I know of a couple other ways to thrust vector.

The first involves inserting small tabs in the nozzle itself near the throat I believe (I didn't do the research, just went to the defense). You can basically have 3 or 4 tabs, and vector by inserting the right combination to achieve the desired side force.

The other that I seen was pretty cool. The exit nozzle had a series of holes. During normal operation, these holes were closed. One could then vector the thurst by opening these holes in the direction you wanted to vector.

I'm not real sure about the mechanics of how it worked, but a projectile simulation was shown with no fins, only the holes, and it was able to completely steer itself.