Main Spar Structural Defect
By accident we discovered a substantial manufacturing flaw at the DG-300/303 main spar, triggering extensive discussions about next steps and ways to publish this to our customers.
As it is the long-standing custom of this web site, I will follow what I have always done in difficult situations: I will tell you exactly what happened and what the resulting consequences are. And as always, I promise not to hide anything:
A repair shop received a DG-300 with severe wing root damage following a landing accident. During the repair process they exposed both spar caps for inspection. At this point they discovered a manufacturing flaw of the spar caps: Instead of being absolutely straight, the rovings of one spar cap showed a slightly wave-like pattern. Alarmed by this discovery we investigated the spar caps of several other DG-300 wings, which were in for repair in our factory or other approved maintenance shops. We found a similar, but less severe pattern at some of these wings too.
The most important question right to the beginning:
“Why may DG-300 spars exhibit this wave-like pattern, and is this flaw also possible at spars of other DG airplanes?”
The answer:
This specific flaw is limited to the DG-300/303 series!
For all DG wings, except for the DG-300 and DG-600, the spar caps are manufactured in separate a mould, which allows maximum precision for the roving placement. While the rovings for the DG-600 are placed directly into the wing mould, they are inserted into a prefabricated channel which becomes an integral part of the wing structure. This allows the same precision for the roving placement. Only the DG-300/303 uses a different manufacturing method (which is also used by other manufacturers), therefore only the DG-300/303 is prone to the described manufacturing flaw.
For the DG-300 wing spar caps, the glass-rovings are placed directly on the inside of the cured composite skin of the wing. The shape of the spar cap is defined by auxiliary moulds on each side of the spar cap, which are removed later. For the center spar stubs, the rovings continue into a separate mould, bolted to the wing mould.
And here is the problem:
In the root area, the composite structure of the wing skin is thinner; the foam core of the composite is gradually shaped down to make room for the root rib caps. This creates a triangular-shaped opening between the spar stub moulds, the spar cap and the shaped down area of the composite skin. According to the manufacturing procedure for the spar cap, this area shall be filled with a mixture of microballoon and polyester resin. After curing it shall be sanded to provide an exactly even transition from the wing skin of full thickness into the spar stub mold. Wilhelm Dirks intentionally selected polyester resin due to its fast curing, providing stable support for the spar cap rovings without slowing down the production process. The wing used for the stress test had been produced following this procedure. And this was the procedure the production team at ELAN in Slovenia had been trained for. All DG-300 had been produced by ELAN for the previous Glaser-Dirks GmbH.
At some point in time, nobody remembers exactly when, someone at ELAN had the “genius” idea of “improving” the process:
Instead of building and post-processing the polyester transition in the triangular void it would be much easier to fill the space directly with epoxy, wait one or two hours until the epoxy gels, and then lay the rovings for the spar cap. A process modification like this needed certification from the original manufacturer, but was unfortunately never communicated back to Bruchsal. Wilhelm Dirks had no chance to warn about the high potential for manufacturing flaws introduced by this process change.
When the epoxy in the supporting triangular foundation was not sufficiently cured before the rovings were laid, som rovings could lose their straight orientation before the spar cap had cured. This caused the wave-shaped pattern, which we call “ondulation”.
Ondulated spar caps do not have the same strength as those with correctly oriented rovings.
Unfortunately this critical area is covered by the root rib and the spar shear web, so that the inspectors had no chance to discover the ondulation during their intermediate and final inspections. Therefore nobody discovered this manufacturing flaw until now!
The airplane quoted above, on which we discovered the problem, was 20 years old and had 1,500 trouble-free hours until the (unrelated) landing accident.
About 500 DG-300 and DG-303 are currently in use, with an average age of 15 years and most likely more than a million hours total. And never ever had a DG-300/303 wing failed due to structural problems.
This statement tastes sweet like sugar, but what does it mean, what can we do with that knowledge?
The inspector who discovered the problem followed the procedure and made a report to the EASA. This way it became “official”.
All involved engineers, technicians, and also the EASA agree that neither the 1,500 hours of the particular aircraft, nor the about one million total hours accumulated by all DG-300/303 are sufficient evidence that there is no problem. The ondulation looks just too scary to the specialists. And nobody knows exactly how much the structure has been weakened. No scientific data is available for this case.
But one thing is obvious:
We have a potentially serious problem at a point of very high safety impact.
Nobody is willing to bear the risk that eventually a wing will fail.
And this includes us too.
But on the other hand we believe that there is no need to ground all 500 planes immediately, since:
-
There had been no problem during 15 years
-
The original stress test of the wing showed a substantial higher structural strenght than the required j=1,725.
As a mater of fact, the stress test was terminated at j=2.1 without failure of the wing.
This means the wing has fortunately much higher strength than required by the regulations. -
We don’t know how widespread the problem is. Of eight so far inspected spar caps only three were ondulated. But this truly not a representative sampling.
The inspection is unfortunately very complex, the wing needs to be cut open.
So, what shall we do?
There are three options:
-
All DG-300/303 shall be grounded permanently – But that is not a viable option!
-
All DG-300/303 need to be inspected and if necessary repaired. We estimate the inspection cost to €6,000 for each aircraft. The wing has to be opened at four places since the critical areas are not reachable with an endoscope. A repair of the spar is very complex and would possibly cost €5,000 for each spar cap. In addition, all aircraft need to be shipped to Bruchsal, since only we can make the decision what is tolerable and what needs repair. We are considering to train our representatives in US or other oversea areas, so they also can do that job.
Given the enormous cost, this option is also not viable for our customers. Also we do gladly without this service revenue. -
We try to prove by testing ondulated structures that the structural reserves of the DG-300/303 are sufficient to continue safe operation without the need for expensive repair, if necessary with performance restrictions.
We decided to explore the third option and have already concluded the tests with heavily ondulated spar caps. The ondulated spar caps demonstrated surprisingly high strength on the test harness, but unfortunately not enough to fulfill the design regulations without the need for performance restrictions.
The problem solution:
The following solution was approved by the EASA and allows the continuing operation of all DG-300/303 without further cost for the owners, conditional to the below listed performance limitations.
The operational limits of all DG-300 Series will be constrained to the following values:
-
New Revised Limit Original Limit Maximum speed 250 km/h (135 kt) - old: 275 km/h (148 kt)
-
Maximum maneuvering speed 175 km/h (94 kt) - old: 200 km/h (108 kt)
-
Maximum take-off and landing mass with water ballast 450 kg (992 lb) - old: 525 kg (1157 lb)
-
Maximum mass of non-lifting parts 240 kg (529 lb) - old: 246 kg (542 lb)
-
Arobatics (including DG-303 Acro) prohibited
Owners finding these limitations unacceptable and prefer to utilize their DG-300/303 according to the original performance envelope have to return their aircraft to the DG factory in Bruchsal, Germany for inspection and possible repair. Charges will apply.
Cost and Responsibilities
When deliberating on the possible solutions, we investigated also the associated cost and responsibilities.
The majority of affected aircraft was purchased at and delivered by the later bankrupt company Glaser-Dirks GmbH. DG Flugzeugbau acquired the type certificate and maintenance rights for the DG-300/303, but not the manufacturing and product responsibilities. We have no advantage from supporting the DG-300/303 fleet. They don’t really contribute to our service and spare parts business since these sailplanes almost never break. However, our own cost to support the pilots is increasing from year to year.
And in this particular case, ELAN is solely responsible for the manufacturing flaw, and not even the former Glaser-Dirks GmbH. ELAN is admitting the mistake, but refusing any financial responsibilities. They just don’t respond to any inquiries of this kind and are totally uncooperative.
A few airplanes (about 10) have been manufactured by ELAN or their successor AMS, and delivered through us. But all these sailplanes are out of warranty. AMS had built another batch of about 25 airplanes and sold directly. AMS told us that they returned to the original manufacturing procedure at some point of time, but they couldn’t say when or at which serial number. Most likely are the latest DG-303, which are still under warranty, technically without problem, but only an inspection could tell for sure.
DG Flugzeugbau has no financial or technical responsibility for the majority of the DG-300/303 fleet, at least up to serial number 475. It would be not justified to demand extensive tests and corrective actions to be executed by DG Flugzeugbau on its own cost. We have already invested more than €10,000 Euro in material and labor to conduct stress tests with the honest interest to keep the 500 DG-300/303 flying.
But with at everything we do to support these pilots, we need to keep the cost in mind.
What happens now?
We created an airworthiness directive containing the above listed restrictions and sent it to the EASA for approval.
Even though the approval process will take some time, we urge all pilots to honor these restrictions already now in the interest of their own safety.
Based on our knowledge – and confirmed by EASA – the operation of these sailplanes is safe within the revised operational envelope. No further actions are required.
This means, “we”, the DG-300/303 owners got off still reasonable.
- friedel weber / wilhelm dirks - 
Discussion:
That such an incident stirs up a lot of discussion in the relevant forums was to be expected and is ok.
That ELAN did a very bad job in those days (negligent, but not intentional!) is no question.
But many other postings have had lowest tabloid niveau!
Some claim the inspector had intentionally signed-off bad airplanes as “good” to make the profit margins. And they cry even for prosecution. What kind of people are these? I know this inspector very well. He was as shocked as we were after the discovery. And he was instrumental in the investigation which finally discovered the unauthorized change of the production method!
The real responsible management of ELAN remains in "noble silence". They are absolutely uncooperative; they even refuse to reimburse us for all our testing efforts. Unfortunately the time for claims has lapsed, therefore suing is not a viable option.
And then there are pictures… Of course we have pictures from the bad spar, but we intentionally abstained from publishing. Untrained viewers could conclude nothing, but be scared. But somebody out of the community has nothing better to do than to publish a picture. What you can see is a totally ondulated spar cap. But of course you see only the surface, the first layer. How is the 5th, the 10th or the 20th layer? Nobody knows!
We removed exactly this very spar cap of the photo and exposed it to tension test. It held J=1.4. Everything ok? Do you understand?
No, of course not! So what does it mean?
The calculated maximum stress for a wing is not resulting from a too tight looping, as you might believe. It results from passing a vertical gust with 15 m/sec (2,945 ft/min) in otherwise calm air. This is a rule in the design regulations, stating that the sailplane has to endure this load at maximum speed.
This sets "J=1" and is equivalent to 6.4 g. At 6.4 g I would weigh (unfortunately) half a ton, and would be unconscious within seconds.
But the bad, terribly ondulated spar held J=1.4, that is about 9 g!
Yes, the DG-300 wing is that strong!
It would even be ok to fly aerobatics. But we prohibited it since an emergency pull-up from a busted figure might exceed the limits.
But J=1.4 is just not enough to satisfy the regulations.
The regulations require J=1.725 for a new airplane and provide for the influence of aging. Under these conditions, j=1.5 would be sufficient for the DG-300. But the tested spar cap didn’t reach that - "noly" J=1,4. The very moderate restrictions reduce the maximum load sufficiently so that even this ondulated spar goes beyond J=1.5, which makes it compliant with the regulations.
But what makes me really angry is the accusation that we attempted to “sweep the problem under the rug”! And only because the photo is dated September 2006! Of course, the investigation took time, but all involved engineers, including the EASA, concluded that after 15 trouble-free years and more than a million hours there is no imminent danger,
I’m very sensitive to such kind of accusations… What would you propose we should have done?
Ground all 500 sailplanes until the investigations were complete because of a suspicion only???
On the other hand: Nobody knows when Elan had changed the production process and how long they did it.
So all DG-300 gliders are affected without any exception.
In contrast to a case with another sailplane in 2003, our problem caused no accident, and our spar had no gaping bonding areas, but was besides the ondulation perfectly manufactured.
I refrain from imagine the discussion such a mass-grounding would have triggered!
Finally the question:
Who pays for this?
So far only DG Flugzeugbau has paid, or precisely, my own bank account; approximately €10,000. But this saved all owners from the need to paying anything! They can continue to enjoy their sailplanes with just obeying those mild operation restrictions.
But to suggest we should be responsible for bad workmanship at a company to which we had almost no business relationship. For a mistake that was made long before my company started to exist… That is gross!
I believe we succeeded with our hard work to keep the effect on all pilots world-wide to a minimum, without incurring extra cost for our pilots.
I think this is the maximum we could do in this situation.
End of discussion…
- friedel weber -
Translation: Manfred Koethe, Boston 
Technical Note 359-24 - final version
Some Months later
The season 2007 is coming to its end.
During the winter we like to offer the inspection of all gliders which are under warranty of AMS, Slovenia as well as those of other customers. The cost of the inspection of a complete glider (opening the wing, inspection, closing the wing and finish) will be about 7.600 Euro + V.A.T.
In the moment the work only can be done in Bruchsal.
The company AMS is rather sure not to have made the same mistake as ELAN. That would mean that these gliders will be free of any defect, but they must be checked to fly within the old limits including Acro-Flights. If a defect should be found, of course it must be repaired to fly without the new restrictions. Otherwise if both sides of one wing part should be defect, it is more economically to exchange the complete half wing. For that purpose AMS issued a special price so that half a new wing will cost about 11.000 Euros + tax.
So we ask all owners of a DG-300:
If you like that we inspect your glider, please call our Mr. Goetze and arrange an appointment.
The work will be done probably beginning 2008.First Results of Inspection
The first inspections of the DG-300 main spar have been made in Bruchsal. Six gliders with a number of 24 potential defects have been examined - all were okay! No problems anywhere.
Now we can issue a certification that the owners are allowed to continue flying within the old limits of safety.
We think that it is a good idea, also to bring your DG-300 to an inspection. It will give you a feeling of safety and you will get a higher price, when you should sell your glider so you can compensate a greater part or the total cost of the insepction.
Here you can see one of the main spars. You have to look very carefully to recognize that the rovings are straight without a wave-like pattern.
