Gurus: Im a newbie owner of Designfoil (Dreesecode) which uses the snack program to generate lift & drag coefficients etc for airfoils. Im wondering if this program or similar airfoil simulators can be used to input a fuselage profile coordinates as though it were an airfoil & determine lift, drag & moment parameters. I realize that meaningful information is probably asking too much because of many limitations, the equivalent 'span' is so short (fuse width) & it varies (width taper), but can this method be qualitatively used to compare profile A vs B all other things being equal sort of thing?

Specifically, Im wondering about design variations of a high speed top wing mounted configuration where 1) the typical fuse outline meets the wing at the airfoil chord line & therefore the top surface of the wing section presents a 'bump' to the airflow 2) the wing is now placed lower in the fuse with semi blended fillets & the bump is reduced 3) lower wing position yet, the fuse top is an uninterrupted flowing line. Are the 'relative' calculated lift & drag & pitching moment signatures between options 1,2,3 reliable?

I'm not familiar with the particular airfoil program you refer to. However, airfoils generally are designed on the basis of two dimensional flow. Their parameters are defined and calculated on the basis of two dimensional flow. The flow around wing tips, very low aspect ratio wings and fuselages is generally three dimensional and the usual mathematical models of airfoil parameters do not apply directly in three dimensional flow cases. The third dimension of flow has to be removed for the two dimensional mathematical modelling to apply. The question of what the third dimensional component is, has to be resolved. This is easier for a swept wing or, wing tip than it is for a wing-fuselage interface. The lift distribution along the span is usually reduced some at the fuselage. The high wing and low wing cases usually involve more three dimensional flow than the mid wing case. In the mid wing case, the fuselage acts more like an end plate, reducing the spanwise flow relative to the other two cases. For minimum drag, the fuselage center line has to be curved to align with the up wash in front of the wing and the down wash behind the wing. At high speeds, which generally correspond to low lift coefficient, the up wash and down wash angles are very small compared to the high lift coefficient case. Therefore, the fuselage shape in the vicinity of the wing is a compromise for a variety of up wash and down wash conditions associated with different speeds.

ptxman,

Unfortunately, your new software will not help you out with fuselage aerodynamics. Fortunately, there are some fairly simple design rules you can apply. The drag associated with wing-fuselage junctions is often called interference drag. The main things to avoid are unnecessary junctions and acute angles. This is why fillets are good. The conventional wisdom is that a mid-wing configuration is best, but this probably is assuming that the side of the fuselage is convex ( bulging out ), so that the mid-wing layout gives two obtuse angles, where a high wing or low wing would give you an acute angle at the junction. Since our planes often have flat sides, you might be better off with a high wing, which only gives one junction ( which should be filleted ). There is no fundamental problem with the 'hump' of the upper surface of the wing being on top of the fuselage, although a perfectly smooth top surface would have a little less drag. If you are able to make the top nice and smooth, and not add an additional junction, you probably have a shot at the best of both worlds.

Good luck,

banktoturn

Here is a link
http://www.dreesecode.com/

Re your 2d vs 3d flow discussion, thats kind of what I figured. But let me try & pose the question a different way. Assume the pgm is a good airfoil simulator. It allows entry of any airfoil shape expressed in 2d. So I input the x,y values of airfoil A & B, it calculates the polars & coefficients at assumed fluid & velocity parameters. I dont know actual lift & drag of the 'wing' until the 3d info (span, planform geometry etc) is defined, but, if the wings will be identical in that regard, doesnt the 2d results allow me to qualitatively compare airfoil shapes A & B?

So now if I input the coordinates for 2 'airfoils' which just so happen to look like a fuse side view & again assume the span (fuse width) is the same in both cases, isnt there some information to be learned?

I seriously doubt it. The fuselage is basically acting as two very short wingtips with no actual wing between them. In other words there is just no two dimensional flow here to analyse. Sure, you'll get numbers but they won't mean anything that you can actually use.

From reports I've read in the past about sailplanes there IS some benifit to using airfoil profiles for the nose back to the mid point or so of the wing. After that it just becomes too turbulent to really matter. You want to pick an airfoil that offers a high point at the leading edge of the wing and then flow the rear curves into the boom. If you can determine the upwash and down wash angles AT THE SPEED OF MAXIMUM IMPORTANCE and align the nose and tail boom with these then you'll gain a slight benifit at that one speed. This is why F3B models don't really have swoopy fuselage designs, the speed runs are all important so they set them up for that region of flight. The same report also said that there would only be laminar flow drag reduction benifits for plug on nose cone style arrangements. As soon as the laminar flow hits the end joint of the cone it's back turbulent again. And any nose hatch types are turbulent as of the leading edge of the hatch. And needless to say if there's an electric motor propellor on the nose then all bets are off.

The documentation for my old David Fraser sailplane analysis program models fuselage drag using a simple cross section model that assumes a regular streamlined shape and calculates the profile drag from basic calculations of cross section area and skin area.

For a fuselage whose width is a small fraction of the fuselage length, the induced angle of attack will be many times greater than the airfoil's characteristic angle of attack. The induced drag will completely swamp out the profile drag and very little lift can be generated because the tip vortices will be so close together that very, very little downwash will be produced. In effect you will always be looking at the small difference between two large numbers, trying to deside what the difference means. The significance of the small difference will be lost in the noise of the uncertainty in the large numbers.

Instead of paying $295 for Design Foil, I would down load X-Foil free from:
http://raphael.mit.edu/xfoil/
In the description of Design Foil there is no mention of laminar seperation bubbles. It may be that it doesn't deal with this important aspect of low reynolds number performance prediction. I don't know of any program that deals with low reynolds number airfoils better than X-Foil.

So xfoil will do a good job? I was thinking of trying it. Does it account for the separation bubble when computing lift/drag coefficients, or just let you know they are likely?

How about Eppler's Profile software, is that public domain, I've read about it but haven't been able to find it anywhere. Supposedly it identifies when separation bubbles are likely, hence identifying when its own result is suspect.

X-Foil takes seperation bubbles into account when calculating the polars.

I don't know if the Eppler code is public domain or not.

ptxman,

You are asking the right question; it would be really useful, in principle, to have a tool to tell you which fuselage profile is best. Fortunately, a fuselage is a little "easier" to design than a wing, since we don't care how good it is at generating lift, only how bad it is at generating drag. Make it look as much like a stretched-out teardrop as you can. You are right to want to get rid of the hump on the top. Get rid of all the bumps and acute junctions that you can, and you will be doing everything that a computer program could tell you to do.

Xfoil, the Eppler code, and all the similar programs do the best they can with separation bubbles, which is not great. The more prevalent the separted region, the more suspect the results. The part of those programs that predicts the onset of separation is probably pretty good. Predicting whether the flow will reattach, forming a bubble, is much harder for the kind of methods that are used. A program that could do that reliably could not be run on your computer in a reasonable time.

banktoturn

Actually, I only paid 1/3 that price on a promo they had going which put it below the price of other pseudo-cad plotting packages & had no simulator component hatsoever. The file handling capabilities looked like 1-stop shopping: import & conversion (Selig .arc .dat, Drela .mse, Compufoil .cor). It exports into autocad, rhino & solidworks which was important to me. The ability to manipulate, modify, stretch airfoils in a simple windows interface is convenient. The aerodynamics stuff was almost a secondary issue although I am interested.

Re your low Re lam sep comment, Im not an expert in this area. FWIW the manual mentions usage of SNACK airfoil analysis engine, boundary layer calcs based on von Karman, transition model based on Schlicting or Michel, turbulent flow based on Buri, drag coeef based on Squire-Young. Color coded indicator circles do appear on the airfoil chosen indicating stagnation pt, laminar bubble or laminar separation resulting from the air, velocity, AOA etc inputs.

Whether these methods are accurate, to what degree, better/worse than x-foil, I havent got a clue. It would be great to hear from more qualified guys like you who work with this stuff routinely. I think its available as a free download for a period, why dont you give it a whirl & enlighten us?

Ptxman,

I'm glad you got your moneys worth. I have no doubt that the program works well at full scale reynolds numbers and the CAD aspects are very useful but I haven't seen anything that leads me to believe that it will be a good predictor of airfoil performance at model reynolds numbers. Why don't you just send them an E-mail and ask them the minimum reynolds number range that the program is intended to be used for? They should know better than any of us. I feel sure that being a satisfied customer, they will give you an honest answer. If they don't answer, I would be cautious about applying the program to model airfoil performance questions.