ZODIAC UNSAFE/ GROUNDED?

[Dan Thomas]

I cound never get past the thickness of it's airfoil. I see the CH650XL has a decent looking wing with a lower profile airfoil that undoubtely will give better numbers.

I once flew a 601 with that fat wing, and was astounded at the performance. It had good STOL characteristics as well as an outstanding cruise speed. All its numbers made the Cessna 150s (that were using in training at the time) look sick. And it did all of it on about the same power-to-weight ratio as the 150's.

Chris Heinz, the designer, is a very clever fellow. His designs are good. But the 601's aileron hinge was the top wing skin, and I was never too sure about that. It also contributed to some stiffness in aileron control, while the elevator and rudder moved easily and had lots of authority.

Dan

 

[Ivorlink]

Hello All,
I am a new member and couldn't resist adding my bit to this thread. I had already started on a Zodiac 601XL when the issues started last year. I put it on hold as I am also in the process of immigration to New Zealand and so have been monitoring all the investigations since.
The 601XL and the new 650 are essentially the same a/c with a thinner wing profile than the first 601 HD series which had the fat wing. The XL uses flaps and joins the fuselage and not to a stub wing as the first HD series.
Now as to the issues- There are recommendations from Zenith to keep the tension of the aileron cables to spec, but when a European XL pranged in Holland, the Dutch authorites clamped down followed by other Countries - Flutter was considered, and so a German company was hired to test the flutter. ( I have the details if any one wants it) This was extensive and the results are available on the web- the outcome exonerated Zenith and flutter was ruled out. Obviously Zenith was not about to assume responsibility as the US legal system would have a field day in liability suits. The net result - zip. The wing fault still is not known for sure and there is a lot of conjecture. Chris Heinz also issued a change to the elevator movement limiting the down angle - the thought is that the elevator sensitivity is very high and light forces mean a sudden full down deflection would create excessive G loads. Some of the accidents were also in moutainous areas with known heavy down drafts. - But that does not excuse what happened on those aircraft that appeared straight and level. What Zenith has done in the interim pending their decision on how to deal with the issue, is recomend a reduction in MAUW and slowing the flight cruise speed. This has not been very nice as the advertised capability has taken the ac to that of a slow single seater! With many 100's of XL out there, the litigation would put the company out of business if they don't come up with some answers soon. The UK has issued some Modifications as mentioned above by others, but my take on whether I continue with my XL or not is similar to what wsimpso1 mentioned in an earlier post + to wait and see what else unfolds from the Zenith skunk works.
A South American group have the option of using 2024T4 instead of 6061T6 in key areas related to the centre spars and the fuselage spar support angles, as well as mass balancing the ailerons - personally think they may have the right track. As we are experimental I have also toyed with a spar redesign and using pushpull aileron and elevator controls. Obviously the math has to be done and probably would have to fall outside of the sport pilot category if the weight increase mandated it - then again this is a whole new plane - isn't it?
Ivor

 

[Ivorlink]

The profile of the ZODIAC XL wing is 15% thick (Riblett GA 35-A-415 airfoil).

 

[wsimpso1]

Pilots have reported and witnesses have seen wings oscillating up and down rapidly. And yet some expert can not find flutter? Either somebody was not trying hard enough or somebody is not telling the truth. And "the cables were not tight enough" just does not inspire confidence. I am willing to bet that lots of airplanes are flying around with cable tensions a little low and they don't fold up in flight. I am inclined to buy the eyewitness accounts.

My thoughts from looking at the planes were that the wing was not terribly thick, and the 15% Riblett airfoils are fine foils. The net drag difference between Harry's 12% and 15% foils is positively tiny. Is there an even thicker foil that was used on some of the birds?

As to the airframe stiffness, foil thickness is the small contributor. If you are looking to drive resistance to twisting modes up in frequency, the figure of merit is G*J. G is a material characteristic, and is a constant in aluminum structures. J is the second torsional moment of the section, and can be gotten too by adding Ixx and Iyy. For the sheet metal structure other than the spar, they basically go with the thickness of the sheetmetal and the square of the dimension in the direction measured. The contribution of the skin from a 15% thickness compared to 75% chordwise is a ratio of 5^2 = 25 - yeah, the thickness adds only 1 part for every 25 parts that the chordwise direction adds. You also have the spars in there, but again, they are not terribly stiff in bending compared to what the seperation chordwise does to fore and aft contribution. Add them all together to get the torsional resistence. Not much comes from thickness compared to the chordwise distribution of sheet metal.

Next up on structural stiffness is if the structure had enough elastic stability. There is a tendency in aviation to design using sem-tension fields and the like. The designers are reducing weight by tiptoeing up to the edge of buckling failures, and that is inherently low torsional stiffness terrain... If the spar or the skins had just a little too much focus on light and not quite enough focus on stiff, then local elastic instability could result low torsional stiffness and give a tendency towards flutter.

The users are not usually going to be that different from the users of other airplanes. And the build quality will not usually be that different either. If an airplane has a tendency to fold up in flight, look at the design. The history of airplanes is full of such occurences...

Billski

 

[vortilon]

The users are not usually going to be that different from the users of other airplanes. And the build quality will not usually be that different either. If an airplane has a tendency to fold up in flight, look at the design. The history of airplanes is full of such occurences...

Billski

The graveyards are full of better pilots and builders than I. I tend to lean toward what our fore fathers have learned the hard way. The hingeless aileron was a pretty bold move and may or may not be the problem. I hope for all concerned they find the problem and it is not detramental to the company and all the builders and owner operators of these aircraft.

 

[sadams]

There is a lot of misinformation being posted here. I have no interest in the 601, but don't like to see the spread of misinformation. First off, the FAA has reviewed the accidents, engineering data atc. and feels that the ntsb recomendation to ground the fleet lacks any justification. Secondly, the flutter testing was perfromed by an independant lab with extensive experience, and found no tendancy to flutter, even with control cables set way below recomended tension (test results - http://www.zenithair.com/zodiac/xl/data/GVT-Report-CH601xl.pdf ). Accident analysis in many of the cases demonstrated initial failure in a manner consistent with overstress in negative G's. Most of the accidents involved pilots with only a few hours in the xl. These findings, along with a lot of elevator authority with very light stick forces required for elevator movement have led most to believe that the accidents likely due to the pilot inadvertently overstressing the aircraft with abrupt down elevator movement. In other words, it most likely represents a lack of training or understanding of one of the quirks of this particular aircraft. Again, I don't own a 601 and likely never will, but I don't understand why armchair engineers have dogedly held on to their theories of flutter even though there is no supporting evidence, in fact, all of the evidence suggests otherwise.

 

[MadRocketScientist]

There is a lot of misinformation being posted here.

Ohhh the irony.:gig:

Shannon.

 

[wsimpso1]

Be careful. Three engineering degrees, one of them focused on structures, analysis, and failures, 27 years experience accurately diagnosing and fixing product with failures, including time reconstructing accidents, instrument rated private pilot flying all over the USA, designed and building a composite airplane... NOT an armchair engineer.

Witnesses - Rapid wing oscillations came from eyewitnesses. While eyewitness accounts are notorious for inaccuracy, if wing oscillations are reported in several cases, you gotta investigate it seriously. This does indicate that flutter occurred at some point along the accident sequence.

Overstress - Accident investigators found overstressed parts. Whoopee. This IS the expected case in an accident. And whether the pilot over loaded parts of the airframe with a strong push, or if flutter grew to the point of overloading parts of the airplane, or flutter plus pilot inputs overloaded parts of the airframe, you are going to have evidence of overload. If nothing got way out of shape, there would not have been a crash...

Now one path that could be taken is that flutter was secondary - it occurred after the airframe was damaged by excess negative-g or excess negative g rate. In this theory, the wing system was damaged by sudden nose down inputs, with some parts weakened or detached, and then the airframe was too soft and fluttered, resulting in the final failure.

How a failure analyst can tell the difference between broken parts from flutter and broken parts from overload is beyond me. I have investigated a bunch of parts damaged by two-way loading in resonance, but in many, the damage was one-way only. If the structure shows damage in both directions, it is an indicator of possible flutter, but if it the structure has somewhat more resistence to damage in one direction than in the other, it will only show one-way damage. Lack of two-way damage does not conclusively prove that flutter did not occur.

Human Factors - I shall put to you that if a pilot can easily over-g the airframe, EITHER the structure is short on strength somewhere OR the control system has an inadequate feedback to the pilot OR both.

Fixes involve:

Raising the control gradient;

Beef up the limiting parts of the structure.

Each distance you from failure. The former gives more feedback to the pilot, the latter has the simultaneous advantages of putting failure at greater g-loadings and reducing susceptibility to flutter. A combination of the two would sure drive this failure much farther from normal operations.

Other approaches could certainly be taken, but I am still left with the feeling the this airplane is short on something significant. Blaming the pilots when almost all other airplanes of similar mission are not folding up and falling from the sky is neglecting the bigger issue.

Billski

 

[orion]

And of course the other aspect of this is asking whether that structure had sufficient margins in the first place. As this subject started to see more public light, a group of owners approached me in order to see whether i would be willing to undertake a structural review of the aircraft's wings and recommend possible alternatives and/or solutions to what is now perceived as a definite problem. This group was not wholly sold on the idea of flutter since their information indicated that several of the crashes occurred at times of seemingly normal, level flight, or during relatively gentle maneuvers, with no indications of structural instability (flutter).

I was unable to do the work due to prior commitments so I cannot comment too much as to cause of the failures but, I can make one comment: Our company investigated the limitations of the LSA category, especially from the standpoint of structural margins and based on that investigation we determined that we would take on no LSA type projects as we didn't feel we could provide sufficient margins and still keep functional payload capabilities. If the category remained as originally formulated (registration of fat ultralights) then maybe however, although the limitations stayed the same, most companies are marketing their aircraft as much more than that. Given the durability requirements of aircraft used for training or for any form of utility use, we felt that the LSA products are marginal at best for the more advanced and/or rigorous missions.

Now, I don't mean to imply that sufficient margins could not be incorporated but these capabilities will of course come at a cost, and most of these LSAs are already way to expensive, especially considering the originally stated price goals. It's too bad though - the goals were good and it could've been a leg up to general aviation but the unreasonably low weight limit and the current costs of many of these still make flying unattainable to many.

 

[sadams]

By armchair engineer I meant engineers sitting in an armchair with little knowledge of the actual facts of these accidents, or the actual structural properties of the aircraft. I was not questioning anybodies credentials. I have heard of 2 pilot reports of wing ascillations, one while flying over a power station that went away by diving to increase speed, and a second from a pilot who's cable tensions were way out of spec. I don't remember seeing any witness accounts of one of the failed aircraft having oscillations prior to the wing failure. Which accident was that?

My point is that I think you are barking up the wrong tree. You are saying there is no proof that flutter didn't occur and I am saying there is no proof that flutter did occur. I think my point of view is backed by the data available and yours is speculation. The problem with this aircraft is excess elevator control authority combined with very light feedback to the pilot with increasing elevator deflection. The excess control authority has been addressed, but as of yet I have not heard of changes to address the light stick forces.

 

[Dieselfume]

I've just posted in a separate thread the recently issued FAA Special Airworthiness Information Bulletin regarding the Zodiac.

Having reviewed some of the data (linked in my above mentioned thread) from the static test to limit load and ultimate load, as witnessed by a Transport Canada DE (equivalent to an FAA DER), it appears that the wing with structural mod is capable of handling loads past ~4.0g before failure. Boy I'd sure be curious what load the upper spar cap was buckling at before this mod. Hmmm, maybe I don't want to know... (full body shiver)

Also, as Orion alludes to above, the ASTM consensus standards do not match up with 14 CFR part 23 standards. Working backward from the static test data, for a 600Kg (1320lb) GW airplane, limit load is 1518Kg (3347lb), and Ultimate load is 2280Kg (5027lb) with a demonstration past ultimate to 2446Kg (5393lb). This works out to limit load being 2.53g, ultimate of 3.8g, and demonstrated 4.08g.

Lacking a copy of ASTM F 2245-06 (since you have to pay for it), it would appear that the ASTM standard only requires limit load of +2.5g, substantially less than the normal category Part 23 standard of +3.8g we are all used to seeing.

Personally, after recently talking with Orion, I don't have a warm fuzzy feeling with what I know about these ASTM standards, and therefore, don't have a complete warm fuzzy with the certified light sport aircraft. I like very much having the transparency of the current Part 23/25 regulations and generous amounts of guidance material.

 

[mcjon77]

Quick question.

When you have your load numbers, don't you usually divide those numbers by the gross weight of the plane minus the wings, sense the wings carry themselves?

 

[vortilon]

I work directly with an engineering company fabricating assemblies for R&D aircraft. I am not an engineer and don't claim to be. That said I have been around airplanes for fifty plus years and have a pretty good sense of what works and what does not. I have twice now seen areas of structure recently that made me scratch my head and say that needs more strength but kept my mouth shut not wanting to upset the slide rule folks. I am now reworking what I already knew was too weak and was found later by a computer model.

Sometimes a little horse sense trumps all the magical formulas.

 

[autoreply]

Quick question.

When you have your load numbers, don't you usually divide those numbers by the gross weight of the plane minus the wings, sense the wings carry themselves?

No.

You're assuming that the wing has a perfect lift distribution, but it doesn't. Lift is only back to normal about a foot from the fuselage. That means your wings are loaded more towards the tips than the theoretical results. Including the weight of your wings less or more balances this effect out

 

[PTAirco]

Indeed, mcjon77 is correct.

I just read the document describing the load test. A total of 5381.2lbs was applied to the wings. To determine the load factor we first have to subtract the wing weight itself and that figure is not given anywhere. Let's make a guess and call it 200 lbs : 5381.2/(1320 -220) = 4.8 G.

The thing to remember is that although they called this the "Exceeding Ultimate Load Test" there was no permanent deformation or failure, except for some elongation in the spar bolt holes. If at 4.8 G nothing has buckled or broken, one could assume the ultimate load to be close to 7.2 G, being a metal airframe. The point is you should be able to fly this airplane to utility class limits without structural problems.

 

[PTAirco]

No.

You're assuming that the wing has a perfect lift distribution, but it doesn't. Lift is only back to normal about a foot from the fuselage. That means your wings are loaded more towards the tips than the theoretical results. Including the weight of your wings less or more balances this effect out

If you look at the load distribution on the document, you can see they took this into account. You still need to subtract the wing weight for your calculations.

Remember, these test were done with people present who know their stuff (just read their credentials).

 

[bmcj]

OK, subtract the weight of the wings, but to compare apples to apples, isn't the same true for 14 CFR part 23? In other words, wouldn't the 14 CFR part 43 plane still be 50% stronger than the ASTM plane because one uses 4g for the limit load and the other uses 4g for the ultimate load?

 

[mcjon77]

I always thought (or should I say ASSUMED) that the zodiac was tested to 6g ultimate load and 4g limit load. Is this not the case?

 

[bmcj]

I always thought (or should I say ASSUMED) that the zodiac was tested to 6g ultimate load and 4g limit load. Is this not the case?

If you are asking me, then I don't know. From Dieselfume's post, it sounds like it was only required to achieve 4g ultimate per the ASTM standard. The other thing I took from the post is that the Zodiac meets the 4g ultimate load requirement with the wing mod, but may not have before the wing mod.

 

[mcjon77]

I just checked Zenith's website. They state that the ultimate load is +/- 6gs. On seperate pages, they show load testing but do not give specific numbers on how much weight is on the wings.