Techniques

For those that weren't there, Matt was doing an outside square and, when making the corner from the vertical dive to inverted flight, the inboard wing seperated cleanly at the fuselage side resulting in an immediate crash and pretty much total destruction of the airplane. The obvious cause was the failure of the left wing right at the natural stress riser where it enters the fuselage.

Some speculation occurred at the accident site that the spars might not have been properly spliced in the center section. While that might be true, in my estimation the failure would never have gotten to the spars if a building/covering technique I strongly advocate had been adhered to.

As are many stunters, the Stuka was assembled prior to covering the wing and tail structures. As a result, when the covering was applied it ended right at the fuse side. Paper covering adds tremendous stiffness to any soft structure on which it is placed. (Remember, an I-beam wing gets almost all of its torsional rigidity from the covering. A C/D-tube wing is stiffer to start with but anywhere the covering is attached becomes much more rigid as a result.)

When the covering ends right at the fuselage side the result is an unconscious multiplier to the natural level of stress concentrated at the fuse. The fuse is an aerodynamic cleaver placed right at the middle of the wing attempting to bisect it with each high "G" flexing of the wings. When we cover the wings right up to that "cleaver" we have made the cleaver's job that much easier since the wing is nice and rigid right up to the fuse but soft and supple...to a point...within the fuse. That such a structure will fail under repeated duress is almost a given.

It is true that better spar splicing might well have prevented the failure as well but the fact is that had the following covering technique been utilized it wouldn't have made a bit of difference.

It is *extremely* important that the rigidity of a spanwise loaded surface be as consistent as possible so that any stresses applied to it will be evenly distributed. The best way to assure this is to cover the wing *before* installing in the fuse insuring that the covering for the left and right wings overlap an inch or so at the centerline. The result will be that aforementioned consistently load resistant structure with no natural or built in stress risers.

I always (almost) cover the flying surfaces completely before installation and dope them as well. A compromise must be made in doping to insure that there is enough "nap" of the cloth/paper left to insure proper glue adhesion when the surface is installed. I dope until I start to see occassional shiney areas start to appear then stop. When gluing the wing into the fuse I rough up the area where the glue will be applied with medium grit sandpaper being careful to never sand through the paper itself.

From time to time I've had some dispute that paper can be so important to structural integrity. In my opinion they are seriously underestimating both the loads they are putting on their wings and the wisdom of insuring consistency of flex along the entire apan rather than concentrating it mechanically at the natural stress riser produced by the fuse.

Matt's experience is just the latest in a growing number of proofs of the matter.

Steve

Sorry to hear about the loss of the plane.

Ted,

It's good to see you're posting here. I, for one, can use your advice.

Denny Malott

Now, on Ted's comments. First off, it wasn't the spars that failed. It was the sheeting. The spar was a single piece with carbon fiber top and bottom that broke in the middle. The carbon fiber peeled on one side, but only due to the crash. There is a trememdous amount of stress in the middle of that wing. This was the largest wing, and the heaviest serious stunt plane (because it was larger) that Matt had ever built. And it was dialed in to turn sharp corners. And Matt was turning them sharp. This all adds to the stress. So, where an earlier plane using this construction may not have failed, this one did.

Now, if just one plane fails, that is one too many, so changes need to be made. Ted's advice is well taken and just that alone might have been enough to prevent the failure. We have determined that there was one other cause in addition, however. This was Matt's first pipe ship. The entire under belly of the fuselage was open and the bottom of the wing formed the bottom of the pipe tunnel. So, not only were there stress risers at the fuselage joint as Ted mentions, but there was no fuselage bottom to assist in compression loads. We feel that the wing actually weakened in the very center because of the ability to flex (just slightly)in the absence of a fuselage bottom. It was not the spars, it was the sheeting that let go and that took the spars with it. Part of the wing broke inside the fuselage, and part of it broke at the fuselage joint. But the real point is it should not have broken in the first place. On the next one it will be corrected.

(Whenever you have an accident, do an autopsy. If something broke in mid air, it was built too light. If something did not break on impact, it was built too heavy.)

The bottom line is this: I think Ted's idea of covering the entire wing, including the center section, with one continuous layer of silkspan is good. That, alone, could have prevented this failure. Perhaps some stiffening material in the fuselage section under the wing would have helped. But the present plan on the next one (the wing is already built, but not covered) is to cover the center section, both top and bottom, in several layers of light fiberglass. This will be brought outside the fuselage, with the first layer extending, in part, to the edges of the sheeting with an emphasis towards the front. The second layer will be a smaller one, but, also, extending beyond the fuselage. Both of these layers will be either eliptical or tapered in such a way that they do not leave a nice, clean shear edge which would then be the next breaking point. And, by bringing these out beyond the fuselage, it will stiffen the center section sufficiently that there should be no fear of failure in that area again. And then, even with the open belly of the pipe tunnel, the absence of a fuselage bottom should not be a problem.

Ted thought this might be overkill. Perhaps. But if the previous wing was built to 95% structural integrity (it did survive more than 1000 flights) and Ted's suggestion would take it up to 101% (Maybe more, I don't know) I would rather have a little overkill at minimal weight expense to insure it doesn't happen again. But, yes, even doing this, one needs to make sure there is not a clean edge beyond the fiberglass that would lead to the next structural failure.

The key here is to learn what happened, why it happened, and try not repeat any mistakes that were made.

Leonard Neumann

I understand the process of overlapping the tissue on the wing before gluing it into the fuselage, but I'm am uneasy about the strength of the fuselage/wing glue joint when made this way. It seems like *balsa / glue/ balsa* would always be the preferred sequence. If you do this type of construction on a regular basis, than it obviously works, and my doubts are unfounded. (I've never seen your fuselage fall off the wing <g>).

The profile of the glue joint would be *balsa fuse / glue / tissue / DOPE / balsa wing*. If seems like the 'weakest link in the chain' is the DOPE layer. I certainly wouldn't try to "glue" a wing in using dope, but dope is one of the layers in this 'joint sandwich'. (I'll make up a sample in the shop and do a little distructive testing).

Could you give me some thoughts that would explain why the joint is strong enough with dope being one of the adhesion layers.

As always, it was a pleasure talking to you at the Nats. I always appreciate your time.

Cheers,

The question re plastic coverings is the one exception which I briefly referred to but failed to follow up on.

The plastic coverings add essentially zero (OK, maybe a couple percent) rigidity to the wing on which they are installed. Thus, the effects they have on spanwise rigidity are negligible. Properly constructed, sheeted and sparred wings for plastic coverings will remain essentially consistent in spanwise rigidity and thus don't fall under the same sort of problem area.

My original "IMITATION" had a monokote covered wing and tail and had literally thousands of flights with no apparent deterioration. I think it is still flyable in another person's hands now and its been flying regularly since 1978!

Warren's comment re the relative merits of gluing the fuse to paper vice wood to wood is well taken. In my estimation there are adequate means available to make this joint more than strong enough for our purposes.

The problem with the wing's spanwise rigidity re paper covering is another matter. We are actually reducing the basic wing's ability to absorb flex when we cover "to the fuse". The wing is actually more resistant to failure before covering than after. In my opinion this is a red flag no-no to purposely increase this risk in this fashion when simple and expedient means to eliminate it are available.

Re the number of flights for which we should aim our structures. There is really no reason not to expect a well built wing to last more or less indefinitely. The means to do so are there without significant weight penalty and to propose a design limit to failure seems counterproductive.

Good discussion.

Hope that you enjoyed it.

>last more or less indefinitely.
> The means to do
>so are there without significant
>weight penalty and to propose
>a design limit to failure
>seems counterproductive.
>
>Good discussion.

All this said, I agree with most points. Silkspan (or silk) certainly helps reduce flexing and spreading the load over a structure.

I am not sure, though, that silkspan really helps that much at a shear point such as where a wing runs into a fuselage. Silkspan's ability to resist a tear at a shear point is questionable. I use a straightedge to "cut" my silkspan. That straightedge gives me a shear point or line very similar to a fuselage - and it tears easily. I can't do the same with silk or fiberglass. Just how much strength silkspan adds at a stress riser is unknown. It has to add some, but because of silkspan's tendency to tear along a shear line - well ....

Regardless, your point on insuring that the area of a stress riser is reinforced is a good reminder. One of the things that really impressed me when I got back into this hobby after only sporadic flying over a 20 year break was how much harder planes are flown now than they were back in the early 1970s. The planes of yore were smaller, the engines less powerful, and the patterns were, for the most part, somewhat softer. Perhaps it is the effect, but the the bigger ships of today with the much more powerful mills in the front seem to be turning harder and sharper than what I remember. I've noticed that wing construction hasn't changed all that much, either, albeit it seems that wings are a little thicker than before. I suspect the stresses we put on wings at the root are much higher than they used to be - again this is only opinion.

One thing is for sure. The two ships I have on the board right now WILL get some glass reinforcement across the center section before final assembly...

I believe the above is essentially correct--that silkspan offers little shear strength of its own. There are a couple of observations I'de make however. First, a composite material's strength is not simply the sum of its components. Silkspan alone is not very strong but when doped properly to a surface, the combination is stronger than the sum of the individual contributions. Also, the typical contribution that a wing skin provides is as a tension member of the wing structure I believe. While resisting a spanwise bending of the wing (this is the primary acceleration-induced loading) the skin is in tension on the outside of the bend and compression on the inside. In order for the silkspan/dope to properly develop its resistance to these forces, it must remain attached to the balsa substrate. On both the tension side and the compression side, the dope interface to the wood surface is in shear. If the interface fails, the material no longer acts as a composite sandwich. It may continue to function on the tension side but would be catastrophic on the compression side since the skin would buckle and peel off the substrate.

Second thought is that this failure, from the descriptions above, and from a typical wing bending load perspective, most likely was a result of a material tensile or compressive failure brought on by the "hard hinge point" that Ted describes. Repetitive cyclic loads applied to our wings will degrade the silkspan/dope/wood interface to the point of separation without any significant outward signs of failure. When this occurs, then the surface acts as two individual pieces, neither of which or in combination can resist the applied loads.

Heavily overlapping the covering and then ending it at some point outside the fuse is equally as bad, as Leonard has indicated also. The trick is to remove any discontinuity in the load/stress application when the wing bends--and this goes for the internal structure as well. Tapering, rounding, thinning, etc. when stopping these reinforcements is mandatory for maximum survivability.

All of the above is the reason I have been very active on these forums in discussions concerning application of reinforcement (carbon, glass, silkspan, etc) with dope or white glue, or whatever household goop someone is willing to try. The bonding of the material fiber to its neibor fibers and to the substrate surface is absolutely critical to the strength of the finished piece. It really doent make any sense to reinforce with carbon fibre applied with a resin that wont develop the proper bonds with the materials for instance. On the other hand, try a sample of thin epoxy and silkspan--it will suprise you.

Final thought--it is possible that the glass repair that Matt performed on the original LE crack in some way contributed to this failure? Unless done carefully, and tapered properly, adding reinforcements to a broken area will very often guarantee a stress riser "hard point".

Anyway...Didn't mean to ramble or monopolize the soapbox, but hope to add a bit to the understanding of the mechanics of the composite structures we all employ.

FWIW

>the typical contribution that a
>wing skin provides is as
>a tension member of the
>wing structure I believe.
>While resisting a spanwise bending
>of the wing (this is
>the primary acceleration-induced loading) the
>skin is in tension on
>the outside of the bend
>and compression on the inside.
>In order for the silkspan/dope
>to properly develop its resistance
>to these forces, it must

Thanks for the summary. Nicely done. Very good points on tensive versus compressive loads, laminates, and laminate deteriorating.

Ted has literally tens of thousands of flights and is a world class competitor - and the experience that comes with both. Normally I don't question what he writes. He is a generous source of information and is willing to share with the occasional duffer like myself. So, if my post seemed like a flame, please disregard. Nothing of the sort was inteneded. I just still have to wonder on this issue, though, about the ability of the silkspan to handle loads at a line where there is a shear (sp?). With that in mind and with what happened to Matt Neumann, might there be some other methods to keep this from happening to others? It seems that this discussion brought out a wealth of techniques. That is a good thing.

More important to me was the gist of Ted's message, e.g. that the loss of a good ship could have been avoided by reinforcing the center sheeting of a wing to mitigate the stress riser at the line between the wing and fuse.

One thing is for sure - all other things being equal - a wing silkspanned completely through the center web prior to installation will be less likely to fold at the fuse line than one that wasn't. Perhaps others methods are better, perhaps they are overkill.

Thanks, Ted, for broaching the subject objectively and getting a great discussion going.

I'm with you in listening VERY intently when Ted and others speak here.

Hi Bob,

Certainly some valid observations. I would opine re the relative strentgh of the paper in the following manner. I, too, use a straight edge to trim sheets to size. I do so, however, in the same manner the "strongman" tears a phone book in half...by starting at one "edge" and working into each subsequent edge separately. In other words, you are separating the sheet of silkspan from one edge to the other, thus literally tearing each fibre individually. This is quite different from trying to pull apart all the fibres on a straight line simultaneously as in the case of the fuse acting on the wing.

In "my" final analysis it seems that adequate resistance to the type of failure encountered by Matt (or many others over the years of my involvement) can be achieved with the addition of only a couple of grams of extra paper properly distributed. This will result in a wing virtually impervious to this type of failure at a virtually zero cost in wing loading. Good insurance for a cheap premium.

Another thought I have here is that sometimes we may be striving to build things too light. In our search for lightness we use 4 pound wood in the wing when maybe we should be sheeting the center section with something like 8 pound instead (or maybe even 10 pound). It would have double the strength at the same thickness, much better crush resistance at the fuselage joint, etc. Just using a little heavier (stronger) wood across that center section plus covering all the way with silkspan might make the difference.

I know some have said to use the heaviest ribs at the outboard tip AND the center section. The same goes for sheeting, too.

Leonard Neumann

Very good point.

There is one thing that still stick in my mind, though. Matt's plane passed the pull test. Then the wing failed. This made me think of other failures that I have seen over the years (including a couple of my own). All of these passed numerous pull tests as well.

All failures I'ved seen started at one point, then spread. Whether its fuel soaking through, stress riser, pinch point, etc. From the initial fault it spreads in a manner in which each section, or, the case of silkspan, fiber has to take an inordinate amount of strain until it too fails. Not too dissimilar to your excellant strongman/telephone book analogy. I freely admit that I do not have 1/100th the experience of you, Paul, Brett, Leonard, et al. That in mind, I have never seen a "clean" failure, foam or wood, where the wing just parted and pulled either towards or away from the pilot. Rather, it started as a small fault somewhere and spread along a shear line (for lack of a better term), and did so in a manner similar to tearing silkspan with a straightedge.

Hence my concern with silkspan. To be honest, I don't know what could withstand that once a fault starts.

In my limited experience, it all comes down to the old weight versus strength question. We know that wings will flex - as you, the man who makes his living sitting between two wings "X" hours per month knows all too well. The thing to know is which way they will flex, and which we will try to dampen out. I doubt we will ever find full consensus on that.

Although the stresses are different, I think we could all benefit from some of the theory that goes into helicopter blades. During my time in the service I had the chance to pick up a couple of "chunks" of rotor blade from semi-rigid systems after a crash. (UH1-X and AH1s). I was always amazed at the lightness and the torsional rigidity of modern blades when I held a chunk, and how much they flexed when I was a passenger in one. My uncle was a design engineer for Kaman and worked on the AH1 Cobra blades (along with the fully articulated Husky and Seasprite). His comment was you can't make a good blade that won't flex. Its just knowing which flex is good and which isn't, stopping the latter and controlling the former. That over simplification came from the man who was hypertensive, had ulcers, and almost jumped every time he heard a chopper with one of his blades went down. He was normally on the site within 24 - 48 hours. Now in his retirement, he can't stand the boredom, and consults doing the same thing.

The answer? Who knows.

But it won't give me ulcers... .

Regards - Bob

JAK

I have been overlapping my silkspan across the center section recently with good results. It doesn't hurt that the center wing section is also reinforced al-la Bob Hunt's center sections with the added benefit of giving a suspended bellcrank mount. I have never had a hint of stress wear on the fuselage/wing joint using this method. Instead of the basswood that Bob uses, I use a hard balsa/plywood sandwich connected to both spars, center ribs and the lower/upper center section sheeting. Then after this, the overlaping of the silkspan or polyspan to about 3/4" outside the fuselage sides. One other thing, both spars are tied to each other with a form of shear webbing X'ed front and back of the spar out to the sixth wing bay. This might be slight overkill, but it is still relatively light and the structure is extremely strong.

Ted,

It was the first outside corner of the reverse wingover....I was standing about 15 ft from the impact.

Jim;

I scanned through my modest mag collection (sporadic back to the 1970s), but did not see the method you mentioned. Can you recommend a source or elaborate further?

It is in Stunt News, in the first Lincoln Log wing article by Tom Morris.

Jim;

Ahhhh, January/February 2000 issue, no?

Got it. I was focused on looking up all of BH's old articles...

Thanks.

I think the original Lincoln Log Article in SN was around MAR/APR or MAY/JUN of 1997.

I'm just completing a Prowler kit (a profile) and applied the glass tape in 1" and 3" laminations over the center joint as per instructions. I then used .2 oz carbon veil in lieu of silkspan. The veil is one single piece on the top and another on the bottom.

On built up wings, wouldn't it make sense to have a single 36" spar in the middle of the wing and have the spar joints, say, 5" from the wingtips?

I've also heard about overlapping the silkspan at the center to get a double layer of paper (but wider than the fuselage).

>Jim Pollock

Jim, your comment of "never a hint of stress wear on the fuselage/wing joint" is interesting. Up to the moment when we launched Matt's plane for that final flight there wasn't a hint of stress wear on the fuselage/wing joint of his plane, either. It had earlier developed a crack in the center of the leading edge on the bottom under the pipe. But Matt noticed that crack and had glassed over it. The crack had not extended around to the top as the center section there is still intact, as is the area that he glassed over on the bottom. But the center section farther back did give way and then the wing sheared at the top, right next to the fuselage. Or did it shear right at the top next to the fuselage and then the center section of the bottom gave way?

Whichever way it happened there was no visable evidence that it was about to fail. In fact, if one could have removed the wing right before that last flight I don't think one could have noticed anything. It passed the pull test fine (done before each official flight). The controls were solid (and the belcrank mount was attached to that spar in the center as part of the reinforcement.) There was simply no evidence of anything until it popped.

But the things you describe may, indeed, be sufficient additional reinforcement to keep it from ever happening to you. We hope so. Just don't be lulled into a false sense of security into thinking that since you have never seen any evidence of stress failure, that it can't or won't happen.

Leonard Neumann

On covering the wing center section with silkspan, just don't dope right where the side/fillets go. Epoxy will go right through the silkspan and the joint will be bullet proof; well, as bullet proof as it gets.

Doing some glassing on the center section is a good idea. If you are going to carry excess weight, the center of the wing and the bellcrank mount are the best places for it.

One needs to be carefull when sanding the fillets. We have seen this type of failure occur when one sands into the sheeting when shaping the fillets. The same failure can happen at the Stab/Fuse fillets.

I have heard of this, too, and it is a good warning, Curt. But in examining Matt's plane the wood is full thickness at the points of failure. This wasn't a case of sanding through and leaving a weakened area. But I do believe that it was weakened, as Ted says, by bringing the silkspan right up to the fuselage side and stopping there.

Leonard Neumann

Great thread! The cleaver "way of thinking" is probably even more important in the profile world than in the competition stunt world, and regarless of any specific solution, the answers are there just by picturing the problem correctly.

By the way, the Nats video played well at the Bridgewater, VA Rotary Club, and your handle, plane, are now well known there. Thanks for your co-operation!

Some people use sheer webbing or cross braces between the top and bottom spars. I use cross bracing all the way on every bay. Some people only go out a certain number of bays. Question; is it possible that the wing is to stiff and instead of flexing when needed it just broke? Leonard said there was a crack they has glassed over. Well that made a very strong SPOT in the wing joint and the rest stayed the same strength. If the crack continued out further on the inside where it could not be seen then this could have been the culprit. The Strong spot became less flexable then the rest of it and something had to give. It starts flexing at the joint and POW it's gone.

I wonder if the wings are to stiff? Will the carbon wing cause fuselages to just blow up?

I am getting ready to cover my next plane. It is filled and ready. Should I cover all the way to the fuse or should I cover just the bays?

Doug Moon

My situation was a little different in that I had pancaked it the previous week in too low of a pull out, more of a hard touch and go. I gave it what I thought was full inspection and straightend the lgear, and flew it the following week. Boom! Folded the wing at the top of the outside square. (Saw the same thing happen to Gene Martine at a KOI on his Tempest, with the same pancacke story he did on his Tempest the previous week!)

Not fun, bent the crank on a good ST60, spinner - gone- CF prop - history - tank - schumtzed too, ... The nose pod was 25ft from the crash. The lead outs cut two nice slots in the wing and part of the fuse. Toothpicks everywhere. No salvage here, I'd like to see the "crash repairs" column try this one!

My wing was film covered, and I had glassed the center. I've heard Windy's method is to use football shaped cutouts of glass cloth accross the joint that extend beyond the fuse/wing joint. Does anyone know more about this method, or has it already been written up somewhere???
EricV

It was already decided to do this before the "incident", but what happened to the previous ship simply reinforces the decision (good word to use?).

Would simply bringing the silkspan all the way across the wing as Ted suggests be sufficient? In light of the fact that the plane survived more than a thousand flights without it--probably. At this point we are not looking at probably.

When Matt lost a Still Stuka that had used .025 solid leadouts, all the planes were switched to .031s. This change would "probably" be sufficient, but with the larger ships (the pipe ship included) they are now .039s. A little overkill in some of these areas is not a bad deal if you expect the plane to last. As Ted said, one should be able to expect the wing on the plane to last virtually forever (as long as it doesn't get oil soaked or meet the ground in some unfortunate attitude).

Leonard Neumann

I assume that you will also be using Bob's center section reinforcement that is show on his Lost Foam Video. What I use is similar, except that it is made from hard 3/8" blasa with 1/16 or 3/32 plywood laminated on the inside surface. The Balsa part is then contoured to the shape of the ribs and the center section sheeting is glued directly over the top of it. If you want, you can then add a small plywood cap over the bellcrank mount protruding through each side. This seems to make the center section fairly bullet proof. At least it has in mine, Tom Morris's and Bob Hunt's planes to date at least. Of course, there's always room for improvement, so if you find a better way, please let the rest of us know!

I think a good summation is that the sparring/sheeting through the center section of the wing must be made at least as strong as the contiuous sparring/monocoque structure of the the wing beyond the center section. This means either continuous spars (a good suggestion) or proper spar splicing and center section sheeting which is continous and extends beyond the fuse side stress riser.

Once this is achieved the covering attached to the surface of the structure must be so arranged as to make the resistance to spanwise flex either as contiuously consistant tip to tip or (less necessary but certainly arguable in this case) more rigid in the center with GRADUAL reduction of rigidity as you proceed toward the tips.

I vote for continuously adequate flex resistance since "adequate" satisfies the strength requirements and the method I employ...continous paper covering tip to tip...results in the least additional weight to accomplish that objective. To the extent that a builder is able to utilize the second method while retaining a reasonable final weight I've no objection to doing so.

The center section sheeting was one continuous piece and that did fail underneath. The top portion of the wing sheared basically at the fuselage. The bottom portion sheared at the leading edge (it was glassed in the center), but sheared more in the center from that point back.

I think a combination of what Ted has been proclaiming and the use of a little heavier sheeting (already used by Ted, so I understand) in the center section would have saved the day.

The next one will still be glassed. It may add a half an ounce in weight, but will definitely take the stress.

Leonard Neumann

The wing for the new plane is already built, but, yes, Matt said he followed Bob Hunt's methods on this one (that was before the crash) so it should be stronger. It is too late to incorporate heavier balsa sheeting in the center section, but adding the sections of football shaped 1/2 ounce glass cloth in the center, as per Windy's recommendations (and he has folded a couple, so he ought to know) ought to insure the structural integrity of this one.

Leonard Neumann

With an adequate spar (i.e. full depth shear web that has proper stiffness), this wrinkle will not happen and this "type: of failure will not happen. This "type" of failure has been witnessed on foam wings with no full depth spar in them. I have always used a full depth light ply spar in foam wings even without wing mounted gear. I have had no failures of these.

Once again, my background as a structural engineer suggests that you look further into the spar and make sure it is adequately stiff.

On Matt's plane, he did have shear webbing between the spars. The spars each had carbon fiber top and bottom (something more than what the plans called for) but no shear webbing in the middle because of the belcrank -- and right where it is needed the most. The wing sheared at the top--pretty clean right at the fuselage. But it broke inside at the bottom pretty much in the middle. Here it was wood separation, as there is still a thin layer of wood covered with silkspan in the tunnel area. There was no evidence of oil soaking at all. vibration, I am sure. Repetitive stress from maneuvers, absolutely!

I am not an engineer and could not tell if it sheared at the airplane top of the wing (plane inverted) which caused the center of the bottom to fail, or if it collapsed in the center of the bottom which than allowed the top to shear.

My thought on the fiberglass, however, is this: The carbon fiber wings now being produced (first by Kaz and now by Windy) do not use any internal structure or spars at all (except maybe at the very center), but are supported mainly by the outside structure of the carbon fiber shell. If the center section of a wooden wing were fiberglassed in a similar fashion, in theory, that could be made strong enough to hold without any internal structure. And if this were done in a lighter fashion to a "normal" wing, it would give additional structure to the outside, that, added to the internal structure, should make the wing adequately strong.

As I recall Matt's comments, the new wing does have a plywood spar. By chance it was built differently. By choice it will now be made even stronger. And, no, we don't want to build the same wing with the simple addition of covering the silkspan all the way to the center to see how many flights it could then take before failure (if indeed there would be failure, or if that, alone, would make it adequate). I would rather be 200% sure the next time.

Leonard Neumann

In-efficient: spreading mass quantities of glass cloth on the sheeting to keep the wing skin from buckling.

Efficient: Beefing up the wing spar to carry most of the wing vertical shear, along with proper joints of same to the wing, so that the loads do not have to be transfered through the weak direction of the sheeting.

The Impact that won the W.C. had a carbon fiber wing spar. It was a box section, tapered unidirectional tape on the caps and +/- 45 degree tape on both sides of the "box" section (this webbing supports the vertical shear loads). This is quite rigid and weighed only 2 ounces, and this spar supported 100% of the wing shear and bending loads. There was extra reinforcement around the cutout for the bellcrank, as this is a critical area for the wing shear and bending loads. This spar was crashed twice in other planes prior to being used in the W.C's plane. (That's another story for another day) Both were pancakes, one into asphalt and one in the grass. The first was with a broken cable in my handle and the second was a mystery loss of air flying in the dead calm. The point is that the spar was so strong it survived two crashes before going on to its final glory, at only 2 ounces! This included the entire spar and the mounts for the bellcrank.

The proper use of these materials makes for structures that will last almost a lifetime. My own experience with these materials tells me that CA is not a proper adhesive to use on carbon. It is just not a strong enough bond to the carbon fiber. A good epoxy is much better, and cleanliness is next to godliness!

Please excuse me for making this sound like a lecture.

p.s. I would be glad to review your plans for the next version of your wing prior to closing up the assembly forever.

The new one is essentially Bob Hunt engineered (same area, same airfoil, but lost foam process). It is all done and buttoned up already, but Matt told me what he did. There is a plywood spar in it, he said. He has been working on the plans for this and a succeeding ship and I know he has the pencils, but I don't know if he has them for the new wing construction. I will need to check with him. Eventually they will be done in CAD, but not yet.

Leonard Neumann

Far be it for me to debate Paul when it comes to structures. I'm certain all he says is true. Whether or not access to such sophisticated structures is (or even needs to be)available to the group at large is another question.

I certainly agree that layers of cloth and epoxy/resin are probably not efficient in terms of strength to weight. I suggested as much to Leonard when we were discussing the subjects at the team trials "post Matt trauma".

I still believe for the average modeler (and I are one since I've no access to sophisticated manufacturing techniques) the traditional sparring arrangements (stressing continuity of spar viability through the center-section via proper splicing),coupled with close attention to a covering/finsihing technique that emphasizes continuity of stress and flex along the span is adequate for essentially trouble free wing structures.

Having said all the foregoing, a little CYA is in order regarding my Trivial Pursuit/Great Expectations/Final Edition, just in case...

When originally built the wing was covered tip to tip with silkspan as I've advocated. The wing performed just fine for several years up to and including a Walker Cup at the 1995 Nats.

When the plane was stripped of its Red, White and Blue Great Expectation plumage ALL OF THE COVERING was removed right down to bare wood for refinishing in the current Final Edition purple pond scum version. The base on the wing for the refinish was 0.2 oz CF applied with dope over the sheet surfaces and poly-span over only the open bays. I used the CF because it would both provide more strength than the silkspan and requires less dope to fill, thus saving weight (although the concept of Ted Fancher "saving" weight might seem more than a bit oxy-moronic to his close associates).

Those following this thread closely will immediately note that I have grossly disregarded my own advice in that the CF was applied only UP TO THE FUSELAGE SIDE since I didn't disassemble the ship while refinishing. This is not only as bad but much worse structurally than simply papering to the fuse since I've got paper inside the fuse joint, a seam, and then CF outside. Given the even greater rigidity of the CF this is like asking for a failure! The ship as currently configured is flying with the operator's fingers firmly crossed that an event duplicating Matt's doesn't happen.

Now, I'm not totally stupid (again, my close associates may be rolling their eyes), the ONLY reason I felt I could do so successfully was because the 1/2" X 1/8" spars within the wing were highly re-inforced with a sandwich of dual 0.5" X 0.007" unidirectional epoxy impregnated carbon fibre strips, vertically oriented full span at both the main spar and rear spar just forward of the trailing edge sheeting. Combined with shear webbing and spar splices with which I was fully confident I was willing to take the chance that the sparring (as Paul advocates) was adequate to the task notwithstanding the dimwitted approach to the covering.

It's worked just fine. So far....

HI Paul

I can tell you for certain that there is NOT a design Flaw in this or any other wings that I designed
There are over 3000 of these planes flying with a large portion of them built up, and many have had over 2000 flights on them
I have used this design for almost 20 years with ZERO failures
I do not know for a fact but I was told the center was left RAW, the LE was cut away to clear a tank and the sheeting did not wrap around in the front center
This would all be reasons for it to shear at the sheeting. cutting away the LE will let the sheeting flex more than it should
This wing is every bit as strong as any lost foam wing built also
The open bottom is also NO problem, these are very well tested and have also 1000s of flights on them
Proper procedures would be over lap the covering in the center....never cut into the LE.... never butt glue carbon in the center...Glue the LE to the rear former, always use epoxy, never CA, ...I wrap the entire center with light cloth ... and I put fillets of epoxylite both inside and outside of the wing joining area at the fuse sides this gives a very wide strong joint with little weight.
to date I have never had a failure on any of these..
there are dozens maybe hundreds of reasons why a wing would fail but with this one...design isn't it

Regards

I guess we have rumors flying all over the place, so I have to dispel this one. First off:

There is nothing wrong with the designs of Randy Smith. I will say this again on another post, but if his designs are followed to the letter and building practices followed as he outlines above, and proper materials are used and glue joints done properly, there should be no problems whatsoever with his designs. (We can add our own flaws. But it is not the designer, then, who is at fault.) Structurally Randy's designs are more than sufficient. Aerodynamically, they are excellent. However, flawed materials can slip past us, our glue joints could fail ( our fault ), anything could happen. And if a catastrophic failure like this does happen, we need to look for the cause (if we ever do determine it for sure). I in no way have ever or would ever place any of the blame on the structural design of Randy Smith's planes.

That said, Matt's '97 ship--totally different from this one--did have the center section of the wing cut away to clear the tank. This had additional structure added to the center section, however, and has experienced no sign of failure.

Matt's White Stuka was built using the aerodynamics and general building practices of one of Randy Smith's designs. (Thanks, Randy.) He couldn't help himself, however, and he had to make it look like a Stuka. (Sorry, Randy.) The ship flew beautifully, and the next ship is being built on the same aerodynamics. As Bob Hunt advised at the FAI trials (before the "incident") "I wouldn't change a thing. There is nothing wrong with the plane." (There will be a few minor "tweaks", but these have to do with Matt's modifications, and have nothing to do with Randy's design.)

Now, on the ship that did fail, Matt just e-mailed me and I found out something I did not know before. He did have a cut out in the leading edge of the wing similar to the .97 ship but not going as deep. This being his first pipe ship, and not knowing what he was going to need, he was trying to prepare for all contingencies and made room for a 7 ounce tank that was never needed. This cut out will not be placed in the new plane, nor will it be in the new wing that will be fitted to the failed plane when it is eventually rebuilt. However, there was carbon fiber mat placed around the leading edge to fuselage joint to reinforce this area and prevent the "stress cracks" that so often occur. There was no sign of stressing in this area prior to the final flight, and this carbon fiber was pulled from the fuselage when the wing failed pulling strands of wood with it. I don't believe that this, of itself, was the cause of the failure.

I will add that, yes, there was "raw" wood was on the top sheeting. Matt covered the wing after it was attached to the fuselage and did not bring the covering all the way to and overlapping at the center, nor did he glass the center as Randy says is his practice. The intent on the next wing(s) is to glass the center out beyond the fuselage with light fiberglass prior to covering it. Little things, but each will add a little additional strength.

Yes, Matt did use CA on some of the joints, and this, as I will mention in a moment, could have contributed to the failure. Furthermore, there was no fuel soaking contributing to the failure. There was no thinning of the sheeting due to excess sanding. There was a 1/8 x 3/8 main spar, tapering to 1/8 at the tip that was covered top and bottom with .007 carbon fiber. There was a 6 inch long sub spar 1 to 2 inches behind the main spar that was 6 inches long which, therefore, extended out beyond the fuselage. This was, also, covered top and bottom with carbon fiber.
There was some separation of the carbon fiber from the spar due to the crash, but it has wood strands attached to it, which means the glue joints held.

The main spar was spliced with a 1/8 x 3/8 inch balsa splice in the center section and also reinforced by the belcrank mount. The plane passed the pull test and the controls were solid through the turn to inverted in the wingover, which leads me to believe there was no internal failure of the belcrank mount or the controls up to the point that the wing folded.

It was about a week or so before the Nats that Matt heard a "pop" during one of his practice flights. He was not sure whether it came from the engine or from the plane. (There were a few backfire pops in the pipe at the FAI trials as those who were there may remember.) Matt noticed a crack in the silkspan under the pipe in the center section by the leading edge of the bottom of the wing shortly after this. At the time, he thought this may have been the result of the pop that the heard. He peeled the silkspan away and covered that area with fiberglass. Neither that area, nor any portion of the center sheeting above it failed as could be observed after the crash. The silkspan and very top of the center sheeting in the pipe tunnel is also intact (leaving the entire pipe tunnel intact) after the crash. So there was no crack here that led to failure. Also, a thought that we both had that the open fuselage bottom might have contributed to the failure has now been discounted. That area is solid.

In reexamining the plane, it does appear now that the glue joint on the splice on the top spar may have failed and that this, then, contributed to, or caused the catastrophic failure in its final flight. It is possible that this was the "pop" that Matt heard prior to the Nats. Maybe not. But in reexamining the plane, the top spar remains intact with the splice having separated fairly cleanly. The sheeting on the top of the wing, also, sheared fairly cleanly at the fuselage. The bottom spar is crushed in several places and the sheeting buckled internally. All of this is consistent with the way the wing folded and leads to this possible conclusion: The total construction methods used were adequate and could possibly have permitted several thousands of hours of flying except for the failure of the glue joint on the top spar. If the top spar splice joint failed, and it appears now that this could be the case, the wing was weakened to the point that the repeated stress of violent outside corners finally led to the ultimate catastrophic failure. If the separation of the spar joint occurred a week prior to the Nats, which, also, appears possible, then the plane survived several hundred more flights before experiencing the catastrophic failure that prematurely ended its life. This leads to my conclusion that the structure was adequately built had the glue joint not failed (if, indeed, that was the cause.).

So, maybe Ted's suggestion of covering the wing prior to installing it in the fuselage would have helped. Either this or fiberglassing the center section of the fuselage would appear to be good practice and will be followed in the future. The use of epoxy in structural areas where there is either a lot of vibration or flex is probably, also, a good idea. We already use epoxy to glue in motor mounts. Using epoxy or even white or yellow glue in spar joints and braces is probably a better idea than CA as these glues are probably better able to flex under loads which could lead to CA's failure.

We must remember that it is all a total package, and the plane is never stronger than its weakest link. We, also, need to remember, that Matt's plane had over 1000 flights on it over a two year span. He put over 25 gallons of fuel through it this year alone. The failure did not happen in one flight or one month or one year . But there was no way of telling that this was going to be its final flight. It didn't show its age. There were no external stress cracks. There were no "funnies" going on.

Wouldn't it be nice if we could just go back and replace the wood in the center section and see if that does the trick? Or maybe replace the glue. Or cover it differently. Or...

But we can't. So the point is, we just need to review our construction procedures, double check our wood and glue selections, and see where we can make improvements. If we, ourselves, introduce a design flaw, we need to correct it. We can learn. We can share ideas. And we can hope that we never witness this happening again.

(Let's see, how many flyers has this happened to? Shall we line up and raise our hands?)

Leonard Neumann

With the input from "Capt" Walker I've had some grey matter flashes that have me thinking. See if you can follow my scatter brained musings.

Firstly: Taking a page out of Boeings book of 1/1 scale airliners, they stress test the wings before the aircraft is commissioned for service, ..this is part of the design/certification process. We've all perhaps seen the dramatic footage of the 777 wing hitting the failure point. Anyway, ..point is, no fuselage is involved in these tests. This brings into focus(for me) Paul's comments about the soundness of the/a wing design, ..or perhaps the correct execution of a sound wing design construction, ..or perhaps construction material quality control, (or lack there of). Anyone of these could bring a wing design to it's knees if not of a high standard, right?

Boeing do not test the complete spans, only half of it, and what we're discussing is a total span situation. Is there a difference to consider?

If the structure that the fuselage adds to the wing is sufficient to stop/delay structural failure of the wing, ..how is it that some of us build take apart models that do not succumb to wing failure? I'll give it to you that "take-a-part" technology may be managed in different ways, but do you see where I'm going? How is it that the T-A-P models do not have wing failure? Perhaps some do, if so why?

Or is it that the fuselage does not add to wing structure, but is the root cause of the wing stress loads?

I believe that if the wing structure in unto itself is of: A) sound design (Randy Smiths fine work) B) is of sound construction (Matt and Leonard's building practices are at the top of there game), and C) the construction material quality controls are in place (this one may be suspect by degree as mother nature has made bad balsa before I should think), ..the end result will be a successful wing, ..that's wing unto itself. If all the above are well and good, and you ask Boeing to put it on the "Stress Rig" It should fail at a point near to/at the/ some point past, the designed in limits. That said, we do not test to find what these limits are, ..we go by what has worked in the past for a given flight envelope. We do not test to failure with our models prior to first flights, ..the testing is done in the field, in the real world of sport or competition flying. The limits of a design are based on experience and good construction practices, ...design and construction practices that are proven from past experiences and updated as experience dictates.

If Matt's wing could have been flown as a flying wing, or tested to destruction as per Boeing, would it have failed at the same load as it did, or would it have failed at more or less then the actual load failure point?, ..this is without the fuselage in place. What part does the fuselage play in the overall wing load bearing ability? Can a sound design unto itself be compromised once the fuselage is added to the mix?

If the "T-A-P" model in effect allows the wing to act more as a self contained structural unit, in which the fuselage is loosly attached (for lack of better words), does this "loose fit" allow the wing to disapate loads over the total wing structure better then the unitbody construction of a glued fuselage/wing combination? It's got me wondering.

Regards.

Now, we could have built in a stronger spar and it probably would have held. (How about an iron I-beam?)

We could have used stronger sheeting and it probably would have held. (How about 3/4 inch Maple Plywood?)

We could have glassed the center section and it probably would have held. (How about three layers of 8 ounce cloth with carbon fiber in between?)

We could have maybe even simply covered the wing with silkspan all the way across the center before attaching the wing and it MIGHT have held.

We have probably milked this thing to death, but most people have not seen the wing and most people are just guessing. A little stronger next time, yes. Failure is not an option. But sometimes striving for a balance between strength and weight is like the rubber powered free flight plane. Question: "How many turns do you give it?" Answer: "Wind it up until it breaks, and then back it off two turns." In this case, since it broke after 1000 (or so) flights, a little stronger should do it. Next time it will be.

Next topic please.

Leonard Neumann

Tony.

>
>My situation was a little different
>in that I had pancaked
>it the previous week in
>too low of a pull
>out, more of a hard
>touch and go.
>EricV

Same thing happened to me, with a scratch-built Flite Streak. In my case, the "pancake" ended the flight - it was like a hard bounce-landing at the bottom of an attempted loop, and killed the engine. There was a small tear in one wingtip, and the music-wire landing gear (1/8") were bent back pretty good, but there didn't seem to be any other damage, so I brushed it off, did an inspection, straightened things out as much as I could, and put it back in the air. After a few level laps, the plane suddenly nose-dived into the ground and the inboard wing disappeared in a cloud of balsa fragments. What had happened was that when the plane hit at the bottom of the loop, the bellcrank mount was damaged, but this couldn't be seen or detected without taking the plane apart. It seemed perfectly fine. When it was next in the air, the mount gave way in level flight, and the bellcrank proceeded to crash its way through each successive rib in the wing almost all the way to the tip. In the process, of course, it yanked hard on the control rod, which pulled the elevator, and since the control horn was on the bottom of the ele, this resulted in a "full down" condition which nose-dived the plane uncontrollably.

Fortunately, the only damage was to the inboard wing and the elevator, and I've almost got it put back together - had to cut 7 new ribs. Luckily, I don't expect the plane to last 1,000 flights - the above happened on flight #13 for this airframe...

Tim

Tim