Aerodynamics and aeroelasticity

[drgondog]

Originally Posted by Soren

I knew you would chicken out.

And as for this being your thread and me not being on topic, well do you see now how irritating that is Bill ? You did this in countless threads.

Go set up another thread and we'll dicuss Gene's math.

Then we will work the math at three different altitudes using the manufacturer's charts for Hp versus altitude at Military and WEP for those altitudes and we will look at drag (I noticed you used best case 109K-4 performance against a P-51D, which by the way has far less drag and i noticed you didn't show it above)) and we will drop 1000 pounds and compare P-51B against K as well as against 109G5, 6, 10 and 14's also with dramatically reduced ata, and you will bring out reports discussing the complete Cl profile for ships.

Then I would like to to comment on both charts from Gene (not you) relative to the Fw 190 and ask yourself

At which point in the nice smooth high G versus TAS does the BOTTOM DROP OUT as we have been discussing for the last week?

Take it to another thread.

Answer the questions I posed you on this thread.

Thank you

[drgondog]

Originally Posted by kool kitty89

Okay, well on the topic I'll again pose the question on elliptical wings:

How do varying elliptical planforms effect the lift distribution, ie the Spit's wing has the ellipse stretched toward the leading edge, compared to a pure elliptical wing as seen on the He 70, or He 112, or the straight LE with elliptical trailing edge of the P-47/P-35/P-43 (and the Re.2000 series fighters) or He 280. Or elliptcal with clipped tips like the CW spitfire, Tempest, or P-47N.

Or adding rounded wingtips to a trapezoidal planform. (ie Bf 109F)

First and most important. All wing section data is maximum, or best, that can be expected because it represents a two dimensional wing - in other words one stretching to infinity with no downwash distribution and no tip vortex.

If you presented a lift distribution it would be constant from wing tip to wing tip. You would visualize a rectangle. This could never exist in real life as there would be a discontinuity at the tip because the lifting line vortices distributed along the span have to 'go somewhwere' and they do.. and it is manifested as two tip vortexes that 'theoretically in a perfect non viscous air) would extend all the way back to original point of lift.

All three dimensional wings will exhibit the above characteristics regarding tip vortex.

All three dimensional wings with constant or tapering chords (either rectangular, trapezoidal, delta or elliptical) will have maximum Lift (Circulation in wing theory) in the center of the wing (when looking at Lift Distribution versus semi span plots), and then all will exhibit 'elliptical like' Lift distribution and curve 'downward' from Maximum at Center to zero at the tip.

A TRUE elliptical wing from a Planview, the spanwise downwash from center to tip due to the distibuted Circulation is a Constant and will exhibit the least induced drag in comparison with all the other planform types with same aspect ratio..

Note: in the last thread where you posed this discussion I pointed out that ALL real world (and specifically the Mustang, Fw 190 and Spitfire) exhibited 'ellipital like' distributions when Soren seemed obsessed with a.) Spitfire didn't have an elliptical lift distribution but the Fw 190 designers did that just to 'improve high G turn capability'..

I pointed out that while all exhibited the 'elliptical like' Lift distribution, the Spit was better than the Fw 190 and the P-51.

In a trapezoidal wing for example, varying the tip/chord ratio 'higher' to say .4 will mitigate the differences in efficiency (for subsonic - actually for supersonic the delta is essentially more efficient).

But real world says you do a lot of things to a wing to a.) perform best in as wide a spectrum as possible, and b.) be safe to fly in a wide range.

Wing twist is primarily to enhance the low speed landing safety by lowering or redcucing the chord angle of attack relative to freestream velocity as you move outboard from fuselage. When you do that you are trying to reduce the effective CL going outboard. When you are successful you still get the high lift in the inboard section of your wing (where you want it) but when it reaches CL max, the inboard section will stall first (by design) and the local angle of attack outboard keeps that section from reaching CLmax at the same time..

So you get buffeting over elevators feeding that to your stick, but you still have lift over the ailerons and a chance to correct to a lower angle of attack (push stick slightly forward) to regain control while not losing roll control.

(which the 190 did because of the unanticipated structural deformation in the tip area causing it to stall at the same time - or earlier as the inboard.)

Now we're back.

The rounded tip does slightly diffuse the tip vortex (in theory) from the last point of true airfoil to the tip. In many cases it will increase the wing efficiency slightly in a similar way to increasing the aspect ratio (slightly) - but more costly.

The Trapezoidal wing planform is simpler that an elliptical wing planform - and it is easier to design placing the main spars (for both structural efficiency and weight) for trapezoidal wings than elliptical wings like a Spit and a Jug.

I'll check to see if I answered your questions

Net Elliptical Wings are more efficient with respect to induced drag given same airfoil, same twist, same aspect ratio..

ALL have 'elliptical like' Lift distribution with max at center, zero at tip.

[drgondog]

Originally Posted by Soren

Christ

Bill when will you sieze with the pissing matches?? Don't you think I know what this is all about ?? Could you answer every single of those questions above Bill ?? No. Also when did I ever become an a/c designer Bill ? Have I ever claimed to be one ?? All you want is a fight, you have no intentions of keeping this cordial.

Yes, I can answer the questions. It is clear you can not. But you acclaim your knowledge which gets us into these disputes. Simply, do you 'know aeroelasticity' or do you not? the answer could be yes, no, or it depends - we can take it from there.

Second point - you have variously claimed that I did not, and do not know what I am talking about - with respect to aerodynamicsa, structures and aeroelasticity. Is the answer 'yes I do", or 'No", or "it depends" - we can take it from there

It's also very convenient that you avoided all other of my questions, and the reason is clear: You can't support your claims, your bold claims that the P-51 & 109 are close in terms of turn performance and that the field of aeroelasticity was considered witchcraft by aerodynamicists during WW2 being perfect examples.

I am not avoiding them - get out your math.. and the reason I say this is that I have the math for simple turn limits, zero altitude loss, as a function of G and CLmax.. the question I asked Gene in the past was to what degree did his plots take into consideration the respective manufacturer specs for max power as function of altitude, and thrust as a function of the prop efficiency. We got sidetracked about the time he quit posting - so maybe you can answer that question? But take it to another thread.

Set up a thread for the 109 vs 51, bring your math. absent bringing the math, explain how you derived Thrust, and how you calculated drag, and bring your powerplant performance as a function of altitude... but another thread - or you can do it here

Now I can tell you what aeroelasticity is and what its effects are on an a/c (Although wiki covers allot of it), I can also tell you that it was in no way witchcraft during WW2 which you claimed it was and that even the Soviets had Scientists specializing in this field, namely M.V. Keldysh, in the early 1930's. That having been said we get better at each field within science as time goes by, and ofcourse aeroelasticity is better understood today, and also A LOT easier to guard against because of the ability construct and test an airframe in sophisticated computer simulations before ever deciding to actually build it. During WW2 the methods were crude by comparison and the most reliable results were achieved by conducting test flights. One method used was carefully examining the wing profile under heavy loads while at the same time establishing the maximum load factor of the wing itself.

Finally I asked you to wait until Crumpp came on the scene, why did you ignore this Bill??

Simply because I respect Gene a great deal but I am fairly secure in my own opinions and don't feel a requirement for Gene in my debate with you. If Gene feels I am wrong he can say so and state his reasons - if I disagree we can have a civilized debate between two people who know a lot about what they are talking about - and if his points are dead on I have zero problem admitting it. BTW, it was email exchanges on the 109 wing that brought us together and forced me to look at some of my old textbooks. I HAD forgotten some stuff. Does that answer your question?

Anyway following your next reply I'll consider wether it is at all worth participating in this thread..

Participation is an option

All in all I consider myself friendly not to just ignore this thread..

Soren, you choose to ignore facts and opinions that are at variance with your own. That could be one reason to ignore this thread.

You could choose to ignore because you may realize you may be out of your depth. We haven't fully tested mine so my 'expertise' may be subject to question also. This thread is one way to focus on that without getting personal.

You could choose to ignore the thread, but 'peek' every once in awhile because it is possible I have a body of knowledge that you don't and you want to learn but are too proud to ask... particularly after all the bi-directional insults (both of us) - but very specifically because you have completely dismissed my knowledge in these fields in a very contemptuous manner and it would be embarrasing to ask.

I hope you stick around - maybe both of us can learn

[drgondog]

I am going to do this in multiple parts because it is necessarily complicated and will bore the crap out of Marcel.

We will start at the point the (Fw 190, P-51 etc) preliminary design teams had settled on the airfoils and wing design criteria, sized the tail (Vertical stabilzer, rudder, Horizontal stabilizer and elevator, trim tabs, etc) based on stability and control requirements for all the flight profiles, fuselage, fuel, engine, landing gear, length of take off and landing length req's, etc., etc.

The aerodynamicists have done all their calculations and are working on deriving wind tunnel results, in parallel, and will be working constantly with the airframe design guys to discuss changes as issues are discovered relative to stall, fixed and free static margins relative to unanticipated aerodynamic effects or cg changes, etc, discovered over time.

But they feed to loads to the Airframe team and the detail structural analysis begins with Free Body diagrams of the airframe under external loads such as engine torque, thrust, lift, lift distribution, drag, landing G's,etc. Very important in this beginning are also asymmetrical loads due to rolling and diving turns.

A first place to start might be looking at all the load combination on the tail - first to look at the Vertical (dive pull out) , Horizontal (Rudder loads in roll) and Moments about the longitudinal axis of the airframe (largely due to the asymmetrical loads- both rudder in roll as well as torque from prop/engine applied to fuselage)

BTW - this is HUGE in a typical helicopter tail boom and rudder design due to the rotor being required to counter the huge torque of the main rotor systems.

They look at the external loads on the wing because they are interested in a.) the wing not failing, and b.) the transmitted loads to the fuselage, and c.) what compromises they make have to make in wing structure due to placement of landing gear, flaps and ailerons, armament, fuel, etc.

They TRY to anticipate structure to resist 'too much' deflection and torsion due to aileron and flap loads, particularly in turns, to maintain the anticipated aerodynamic efficiency to achieve their design targets.

They look at the powerplant and how the loads tranmitted by that big ass prop (thrust and torque) must be carried to the airfarme - and usually anticipate that both hp and torque will increase over time and the structure must be a.) capable without redesign or b.) redesign of key components are not difficult or long lead time. Example if the primary structur is a forging instead of tubes - you are hosed as that woul be a long lead time item to change.

More later - we'll get to 'aeroelasticity as it existed' in WWII in next post or the one after it

[syscom3]

And it was all done with slide rules.

In my experience in the aerospace industry, talking to the engineers , they told me something about design.

Some of the smarter engineers know intuitively whats going to happen, and the lesser engineers end up proving them right.

[Glider]

Originally Posted by syscom3

And it was all done with slide rules.

In my experience in the aerospace industry, talking to the engineers , they told me something about design.

Some of the smarter engineers know intuitively whats going to happen, and the lesser engineers end up proving them right.

That I like

[DerAdlerIstGelandet]

Originally Posted by Soren

It doesn't if you read his posts in which he AGAIN insults me to try and pour fuel onto the already enormous fire.

Also keeping in mind all the threads he has sidetracked I have every reason to believe this is just another attempt at creating a fight.

At no point in this thread so far (atleast until this point) has Bill been anything but cordial. He is asking you questions, you are skirting around them.

This has the making of a good thread, that a lot of people could learn from. Just answer the fricken questions man!
Frankly Soren he is only talking to you in the way you talk to other people.

Seriously Soren, he has not flamed once in this thread.

[parsifal]

I know virtually nothing about this topic, but a question keeps floating into my head and was wondering if someone would want to answer it. My rudimentary understanding (if you could call it that) of the flutter phenomena is the wing bending or vibrating because of the forces caused by the airflow past the wings. Yes? Now, if there is more airflow further from the wing root, than there is closer to it, isnt there going to be a greater effect because (for want of a better expression) there is a greater leverage for the forces at work to exert on the wing and airframe structure?

If that is correct, wouldnt an elliptical wing form be inherently less likley to suffer than a trapazoidal wing form, where inherntly a greater percentage of the wing are is further away from the wing root.

Or is this completely wrong......

[drgondog]

Originally Posted by parsifal

I know virtually nothing about this topic, but a question keeps floating into my head and was wondering if someone would want to answer it. My rudimentary understanding (if you could call it that) of the flutter phenomena is the wing bending or vibrating because of the forces caused by the airflow past the wings. Yes? Now, if there is more airflow further from the wing root, than there is closer to it, isnt there going to be a greater effect because (for want of a better expression) there is a greater leverage for the forces at work to exert on the wing and airframe structure?

If that is correct, wouldnt an elliptical wing form be inherently less likley to suffer than a trapazoidal wing form, where inherntly a greater percentage of the wing are is further away from the wing root.

Or is this completely wrong......

Flutter is usually experienced primarily with control surfaces and usually because they have a rotational degree of freedom (i.e. pivotal as in elevator or aileron.)

The manifestations of 'flutter' is an oscillating deflection normal to the freestream. It is caused by one of two primary reasons.

1.) it is immersed in turbulent or unsteady flow (say near inboard wing stall or ,compressibility, or indicial gusting) and that turbulent flow is shedding vortices on the elevator in a way to approach resonant frequency
2.) it is resonating with an input frequency of some other type input - say the beat frequency of the engine.. and the structural frequency of the elevator/aileron is close to the input frequency

If the wing, for example starts 'resonating' you are in deep and serious trouble as resonance by definition amplifies the deflections - usually in the case of an airframe - to failure.

[drgondog]

Next phase - you have all the loads and work on individual primary assemblies, for example the tail. With the assumed worst case rudder loads normal to the rudder and the axix of symmetry for the airframe, and the assumed worst case loads on the elevator you design each sub component within the "lines drawings" established during Preliminary Design. You have two dominant boundary conditions - stress and weight.

As the rudder and elevator are movable surfaces, Flight Controls group, is engaged to help design control horns, tab linkages, etc - which are largely determined by the forces that have to be applied over a moment arm, and the moment arm length or 'swing' are largely determined by the space available.. So you look at the design, the hinge points and control forces required, the amount of deflection permitted by the aero and stability and control guys for rudder and elevator, the spars, torque box, etc to take out bending cause by the aero forces applied to them.. If they (load carrying structure) are built up beams or torque boxes then compression is carried in one cap, tension in the opposite cap, and the shear transfer in the web between the caps - normally riveted.

So visualize an "I" beam with the very top surface being the skin of the elevator, then an "L" or a "T" extruded aluminum Cap section, then a Web made of thin aluminum connected to both the top and bottom "Cap" section, with rivets. It will be sized based on buckling stress threshold in carrying shear from the Compression Cap to the Tension Cap. When Buckling is calculated it usually means going up one more dimension in the sheet metal web - say from .040 to .050.

I am going somewhere with this relative to aeroelastic analysis, but later

Usually a spar built up like this will be stiffened with vertical Caps joining the top Cap, Web and bottom Cap.. this results in a closed solution for shear analysis and enables stiffening the web in some cases without going to a thicker web - or simply a desirable loacation to attach a rib.. or located there (think three D) to attach another web perpendicular to this beam we are building to yet another Spar built up the same way - But Now we are building a Torque Box rather than a simple spar.

So now assume we have built the top and bottom skins, we have our beam or torque box analyzed, we have all the cut outs required to position control linkages, that it successfully meets the 8 G limit load to Yield at its WEAKEST point, and meets 12G ULTIMATE for its weakest point also.

Maybe the elevator is not so large as to require torque boxes and simple tube truss works, but we have used the same approach for the Horizontal and Vertical Stabilizer..

For the moment we have finished the elevator, and similarly the rudder and say for the moment we know how to build the vertical spar of the Vertical stabilizer, and we have finished looking at tail wheels and are at the bulkhead that will be the 'Tail" to "Fuselage" interface.

At that point we know we have to safely and efficiently carry the loads from the tail in all the worst case scenarios up to 12 Ultimate (at least we have done the best we can - and fought with the airframe design guys in the eternal Strong versus Light battle.

I know I can declare victory with the analysis because safety usually trumps Performance in US doctrine - but I do care because I want MY ship to be the best of class and weight is HUGE.

So Now I'm at the interface aft and I know I have to take out Forces normalized to three Planes and Bending in at least two axes - but I'm gonna check with all three.

At this point you will be looking at perhaps 4 main beams of the airframe (two top, one on each side, and two bottom, one on each side... they will transmit the Compression and Tension Loads, a Bulkhead will take some of the Torsion Load, pass it to the shear panels connecting the Four Main beams, and in turn distribute the load successfuly going forward... i.e "the Load path" for the artists in the audience - because this analysis combines art with the know physics and math.

Syscom had it right.

If some one wants me to go into more boring detail about the wing, what to do about cut outs for inspection panels, ammo doors, cowlings, etc I will.

The Net Net.

The airframe must take all the applied external loads from High turns, to dive pullout, to rolling manuevers in dive, to the propeller torgue, account for all the holes and inefficient structure (like a cockpit) or a mid wing versus a low wing design - and it's gotta fly.

The airframe is not a "tube" or a "box" of homogeneous material properties that are simple to calculate in many cases for a.) natural frequency, b.) Stiffness, and c.) harmonic motion to different forces.

No, it is a complex mess of torque boxes made up of stiffeners and sheet metal and rivets, connected to castings and forgings through shear or tension devices, with different properties at different locations or different properties along a constant line span wise of fuselage station wise. It is not even simple to calculate DEFLECTION or TORSION Displacements due to Precise Loads - much less complex aerodynamic loads.

Now, (an edit to this post) I have seen attempts at striffness calculations performed after the fact to attempt to solve a problem - the spitfire wing tip comes to mind when the control reversal issues came to play.. and the equations they used from my perspective were technically correct for that small region as an approximation for the stiffness)

Now, Soren tell me how the 'Aeroelastic' properties, some of which I have named
were not a 'well known science' even AFTER the Korean War... much less during the day of Kurt Tank?

And once you have given me the background I don't have on that subject, tell me how the Germans analyzed theri design for a.) natural frequency, b.) deflections under all the theoretical loads, and c.) changed their designs before the a/c was built?

[drgondog]

Chris - BTW an arcane branch of structures was first applied to your beloved bird, the Blackhawk. I can't remember his name but the guy Bell hired (a very smart Brit) to do 'survivavbilty' on the Bell competition for the AH-60 was whicked away by Hughes but not before I learned a little bit.

The spec called for crew survivability in an autorotate crash of 2G's (IIRC) and the problem was to UNWIND classic structures of desing to a higher load limit. Simply we had to design the bottom of the ship, (and by definition the bottom of the AH-60) to elastically deform (fail) 'outside-in' and absorb the energy of the crash - protecting the crew section while screwing up the bottom of the bird.

A sloppy 'cushion' of collapsing beer cans.

I say the 'first' only because it was totally unknown as an applied structures science at Hughes, Bell, Boeing, Kaman and Sikhorsky because we had to search hard to find John ?? (CRS). Obviously the art was known in UK but not applied to helicopers.. Maybe used elsewhere but first applied in states in airframe biz at Bell and Hughes in early 1970's.. I suspect even the auto business should have been studying it

[DerAdlerIstGelandet]

Interesting. You worked on that project?

[drgondog]

Originally Posted by DerAdlerIstGelandet

Interesting. You worked on that project?

I did - on the losing side. The ship we proposed to the Army was a much bigger Huey like Utility Tactical bird - but in my opinion we lost because our ship was still 'wedded' to the two blaed rotor and we could not meet the 'reduced internal g load' below the threshold - it still bounced your ass around at cruise to high cruise.

I was part of a three man team with the chief Aerodynamicist and head of the transmission group to solve 'another way'.. slipping back to mating structures with aeroelasticity - I designed a 'flat plate' interface between the Transmission and the fuselage attach points.

If you can visualize holding a long metal ruler or scale in your hands and bouncing (like a yo yo) it, you will notice a point on either side of the force (your oscillating hand motion) there is a location where the vertical deflection in both directions is zero. so when there is an 'up' force the ruler is convex up, and with down force, convex down. the points at the end are widely deflected at max range of motion up or down - but the 'nodes' don't move.

The challenge was to build a 'plate' analogous to the ruler and replace your oscillating hand with the rotor/transmission system.

So I designed a semi plate strong enough to take and distribute the rotot/transmission loads, bounce in the middle - but attach to fuselage at the four 'neutral nodes'. result was greatly improved ride because the fusleage was attached at the calm point while the rotor/transmission was bouncing around.

We actually built one for a Jet Ranger and it was called the 'Noda Magic Ride'

Big issues were pilot control 'feel' differences and had to do funky things on the drive shaft from engine to Transmission to be able to accomodate the deflections from a fixed engine mount to a bouncing transmission.

Great solution but not too practical.

[DerAdlerIstGelandet]

Basically you were working on a type of vibration absorber?

[drgondog]

Originally Posted by DerAdlerIstGelandet

Basically you were working on a type of vibration absorber?

How about 'isolating the vibration to the rotor pylon/platform' and based on harmonics of motion, letting that system flail away, but at the 4 nodes which attached fuselage to the Rotor/Pylon/Transmission system - it was at very low vibration.

think of a Pogo stick attached to, and bouncing the tuned plate. Where the plate had a natural zero deflection point ('Node') from the pogo stick bouncing up and down on it (the thin ruler analogy) and located far from the center of the plate (or thin ruler), it was smooth as glass.

If I could draw a picture from the side - of the plate - you would see one mode of the 'ruler' as convex up, and the other mode as convex down. Superimpose those two curves to the point where they 'cross' at two points - one at each outer 'third' of the two ruler pictures?

Where they cross is the zero deflection point of the two alternate deflections from a two motion vertical force (up then down) in the middle of the ruler. In the middle of the ruler is the maximum deflection up from the 'up force', then the maximum deflection down is from the 'down force'

The analogy from the helicopter is that lift from rotors - particularly two rotors is really not a constant 'single value.. the blades tend to unload a little in a two per rev beat..so its a vertical force, then a slightly lower vertical force, then a slightly higher vertical force, etc etc (the pogo stick analogy).. and your butt in the seat is going up and down to the same beat and you feel the G force variation as 'vibration'..

And equally the reason that fatigue is such a bitch in the helicopter Biz (segue back to Aeroelasticity).. IIRC we used about only 20,000 psi for 2024-t4 as the 'yield'/allowable stress for Limit Loads.



and that is what beats the shi# out of you in a high speed run in a Huey.. and also creates the 'wop wop wop' sound that is Sooo distinctive for Bell Ships