Facet Opel

This demostrates that all bodies will develop a vortex flow at one point or another, although of the ones shown, the angle of attack will differ significantly between them as to when. To do something like the Rasberry, you need to design it in such a way that normal flight and that which generates the vortex of the forebody are well correlated in order to make sure that the airplane does not experience detrimental pitch trim changes. This will most likely require that the airplane use fairly rounded leading edges in the forebody strake.

I think that if one had the tools and testing facilites that it would be possible to reconfigure this with a sharp strake, but I think that would be quite a bit of work.

Originally posted by orion
The document you reference deals mostly with wings with sharp leading edges.

The J35 draken has a very nasty stall mode that the Saab engineers called superstall. At certain combinations of AoA, pitch rate, speed and altitude it would suddenly pitch up to 30+ degrees and stabilize. In this stalled condition the plane could be rolled with the ailerons and the elevator could control pitch between 30 and 80 degrees. There wasn't enough elevator authority to push the nose down below ~30 degrees but it could be recovered by pulling the nose up to maximum and then pushing down fast. Hopefully the plane's momentum would carry it through the unstable region. I haven't been able to find any information about the airflow leading up to and during superstall. Have you?

Draken has a blunt leading edge all over. My reading on deltas leads me to suspect that the forward part probably didn't form a LEV until the aft part was close to it's stall and that under certain conditions the forward part started vortex lift just as the aft part stalled. This would result in the fast rotation observed. Then as the plane went past 35 or 40 degrees the vortices would burst and the plane would be a bluff body with some control. When the pilot pushed the nose down the LEV would reestablish on the forward delta at ~30 degrees and the positive moment would reappear.

A sharp leading edge at the nose reduces the influence of the variables mentioned above and makes the low speed behavior of deltas much more predictable.

Orion,

what kind of design effort would be needed to make Raspberry work?

Indeed, because of the pitch behaviour concerns that I had, I shifted my interest to double delta's. Could Raspberry be somewhat easier to analyse when the wing planform would be turned into a double delta?

What kind of effert & commitment would be realistically required to create Raspberry for real?

It doesnt have a rudder for one thing...

Off the top of my head? I don't know. When I was looking at it some time back I guessed that some of the preliminary ideas could have been tried with small hand launch gliders, building up in size scale until you had something predictable at about 1/3 or 1/2 scale, but that would only have gotten the configuration to a rough and very preliminary shape.

Following that, the work would have needed to continue to a higher level of analysis. Given the unique nature of the airframe, nothing short of CFD would come close to predicting the shiftng patterns of the pressure distribution. I normally don't like using CFD for up front work but in this particular case I don't think there's much choice short of building a large model and sticking it into a wind tunnel. At that poit though, the CFD is cheaper.

The problem though is that the "affordable" CFD (vortex-lattice and panel code analysis) packages don't tend to do a good job of dealing with separated flow (which is what the vortex is) - that requires a full 3D Navier-Stokes solution, which can be as pricey as a wind tunnel ($15k to $30k per model).

If I was to take a rough stab at a number for a ground-up development, including complete analysis through flying prototype (composite), and if we did all of it here and swallowed some of the costs ourselves, I think the development would still come in at $200k, maybe a bit more.

And going to a double delta would not make it simpler since you now have two surfaces that generate this vortex flow, probably at two different aoa's since the deltas have two sweep angles. The only way that would be simpler is if you had the design database for the Lil' Draken.

Originally posted by orion

The problem though is that the "affordable" CFD (vortex-lattice and panel code analysis) packages don't tend to do a good job of dealing with separated flow (which is what the vortex is) - that requires a full 3D Navier-Stokes solution, which can be as pricey as a wind tunnel ($15k to $30k per model).

thanks for your (gu)estimates

I've seen two software efforts, that try to combine spanwise wing analysis with using XFoil for local section information at discrete chord positions. Call it 2.5D analysis if you will, instead of full 3D analysis.
One is written in C, the other one (written by Matthieu Schrerer) is written in Java. Could it be that this adds significantly to using bare VLM methods, as these 2.5D methods add the local section behaviour to the total analysis as well?

My hunch is that you're going to answer: "It depends, for straight wings, yes, likely; but since the underlying spanwise analysis doesn't support highly swept wings too well...., etc... "

Hans

VLM allows the input of camber lines so using a 2.5D analysis werhe you can input a number of spanwise sections should actually be pretty good. However, not having seen the software I can't even guess at its applicability of accuracy. It does seem like a good intermediate (low cost?) approach. How does it deal with tip issues (or does it)?

Orion,

on my IBIS project site, I keep a page with links to interesting software.
You can check it out here

The two products I referred to above are called XFLR5 and MIAReX. Both are free and/or open source programs.

Thank you - interesting stuff. I'll take a closer look when I have a bit more time. Looks promising for first cut trade-offs and even a bit of optimization.

Thank you indeed, Hans. I downloaded XFLR5 and was able to run through my trade study on elevon span in a fraction of the time I was expecting. While I wasn't looking for strictly accurate numerical results, I was able to make comparisons of one version versus another in short order.

Nice program.

Originally posted by Topaz
Thank you indeed, Hans. I downloaded XFLR5 and was able to run through my trade study on elevon span in a fraction of the time I was expecting. While I wasn't looking for strictly accurate numerical results, I was able to make comparisons of one version versus another in short order.

Nice program.

Hi Topaz, I'm glad you liked it! Perhaps you can use this tool to underpin your 'fear/respect' for swept-back configurations?


On another note, I still don't really see/understand where the underlying techniques of these two programs could support the analysis of a delta or double delta configuration. Even if knowing the local section behaviour in such a wing configuration is cool, I reckon that outside the linear aerodynamics regime these programs can't really help. How would they correctly analyse the high AOA (vorticity) case?

In Hoerner's "Fluid Dynamic Lift" there is a section about delta's, that contains formulae that tell you how much lift due to vorticity and how much 'normal' lift is produced at any AOA. These formulae are approximations, perhaps good ones, but approximations nonetheless. What is left out is what happens with moments when sweeping through an AOA range... Also, these formulae apply to single delta's only...

Vorticity and separated flows are rather difficult to pin down accurately short of putting a model into a wind tunnel or running the expensive CFD codes. When I still worked in the mainstream aerospace sector I remember running accross Cm - alpha plots in connection with strake and swept wing planform combinations: Some of the results looked like abstract string art.

Given that most of us do not have the ready cash to shell out for this level of analysis, the only way to proceed is to go as far as you can using the basic mathematical tools, then transition to models. True, it will be difficult (if not impossible) to gain any numerical data, but building larger and larger models will allow you to tweak the configuration and along the way, to come closer to familiarizing yourself with the behaviors prior to building something big enough to actually sit in. Hopefully, by that time the models allowed you to arrive at a configuration and mass distribution that demostrates stable and predictable behavior.

Originally posted by h_zwakenberg
Hi Topaz, I'm glad you liked it! Perhaps you can use this tool to underpin your 'fear/respect' for swept-back configurations?

Oh, I think we still have a way to go on that.

Originally posted by Norman much earlier in this thread
... Your observation about wetted area is basically what I've thought for some time. i. e. You're just trading all that non working skin on the fuselage and empennage for working skin on the wing. Although the wing is substantially bigger on a tailless than on a conventional of the same weight the total surface isn't realy much greater. ...

Yeah, that's nice. And as mentioned above, the major contributor to the wetted area is a structure that tends to have a low parasite drag in the first place, and is relatively easy to build to 'laminar' standards of smoothness and (lack of) waviness. Fuselages, by and large, are not.

In fact, as I understand it, wings (as a structure) are relatively easier to build than fuselages, so by going to a tailless aircraft with a small pod, we've put the majority of the aircraft (by volume, weight, or however you want to cut it) into a type of structure that's relatively easy and quick to build. And as Hans mentions too, tends to have a relatively low parts count.

This, along with the ability to store such an aircraft in a very small space, is what drew me to such designs in the first place. Performance gains, if any, would just be 'frosting' to me.

We're neglecting induced drag, of course, but that's one of the places where you 'pay for your supper' in designing tailless aircraft. It's much harder to get that optimized than with conventional aircraft and still have good handling characteristics.

Finally I'm finding the sort of info I've been looking for to help design something I wanted to build about 10 years ago!
I tried to contact Wainfarn years back to ask if he was creating vortex lift on the Facetmobile?
some may know I have a BD-5 kit (well most of one) but it wont fit into our ultralight catagory here in Australia.
It has to have more area, I tried increasing the span, but the wing got fairly heavy (structurally) and defeated the purpose somewhat.
So then I thought of making it a delta, lots of area and with only an 8% section at the root, was nearly a foot thick which meant I could make a real light spar (IF ANY!)
Then I started hearing about the problems the Velocity's were having with the lifting strakes while I was reading some of the stuff dealing with the Lippisch deltas in the american wind tunnels.
It all got a bit confusing, so I went and built something else.
Too many projects
Arthur.