Dated: 11 November 1965
Addressed to: N.V Vliegtuigbouw, Teuge, Netherlands.
Dears Sirs,
Sagitta II, ZK-GDO, serial no. 010
Following receipt of information from Australia about the fatal accident to a Sagitta glider serial No. 004 at Singleton, N .S.W., on 24. July 1965, this Department called for a precautionary inspection of the wing root area on all Sagitta gliders in New Zealand and imposed a temporary ban on cloud flying and. aerobatics pending the results of the Australian investigations.
In the case of ZK-GDO, serial No. 010, the precautionary inspection resulted in reports of some disturbing defects in the visible main spar construction at the root. The glider was then further examined by officers of this Department and a timber engineer from the New Zealand Forest Service whose report is given in N.Z. Forest Service letter dated 4 November 1965, a copy of which is enclosed for your information.
Also enclosed is a set of 5 photographs Nos. 1 to 4, showing the defects in the root attachment area of the starboard main spar, while No. 5 shows the delamination visible between the laminated blocks of the port main spar root.
Where visible, these defects may be of no great importance in the structural integrity of the glider, but unfortunately the method of construction is such that critical areas cannot be inspected without some destructive opening up. No such opening up has yet been attempted. On the port spar, inspection of the boom ends has of course not been possible due to the closed metal end fittings.
The defects are apparently all due to excessive shrinkage of the timber. Note also the clearance that has developed between the spar and the metal fittings (see Photograph No.4), a total clearance of 0.074 inch (1.9 mm.) having been measured between the starboard upper fitting and the spar. Inquiries indicate that this particular glider has always been adequately housed and protected from the weather when not in use.
The general finish and coarse hand trimming of the spar root ends to accommodate the metal fittings are not considered to be of the standards normally associated with aircraft construction and coupled with lack of any effective spar end protection, such as adequate painting, have probably aggravated the cracking and delamination now evident.
The owners have agreed to ground ZK-GDO pending further advice which they are naturally anxious to receive.
I will be pleased to receive your urgent advice with your recommendations regarding repair or replacement of the defective wings. If repair is proposed, please supply full details of the necessary rework including such opening up as may be necessary to check on the integrity of the structure.
Recent advice from Australia also indicates that you are preparing strengthening modifications for the wing root area. I will be pleased to receive details of these modifications at your earliest convenience.
In addition to the above requested advice, I will be most grateful if you can provide the following further information:
1. A copy of the applicable specification for the timber used in the spar construction.
2. Moisture contents limits for the timber at the time of manufacture.
3. Identity of the glue used in manufacture.
4. Copies of the drawings covering wing spar construction and the root attachment fittings. probably consisting of drawing numbers
50-00
50-01
50-02
50-30
50-31 (3 sheets)
5. Details of any static tests that have been performed on the wing structure.
6. Details of any similar or related defects in other Sagitta gliders.
Yours faithfully,
Director of Operations & Technical Services,
E.F. Carpenter
Attention: Mr. R. Heald
Date: 4 November 1965
Sagitta II Glider
Referring to our joint visit to the above aircraft at Hastings on Monday, November 3rd. Examination of the visible portion of the wing spar roots shows the following damage.
(a) A shake at least 3” in depth in a laminate of the end block at about mid height of the cross section.
(b) Small checks in some of the chord laminates.
(c) Local delamination of the glue lines of the end blocks.
(d) Local delamination of plywood web from the chord at one corner.
(e) Some breaking out of end grain timber on one end block.
There is no apparent evidence that the damage is due to overloading during flight. The most probable cause of a, b and c is shrinkage due to drying out of the timber.
Unless special measures were taken to control the climate of the factory the winter equilibrium moisture content of the timber in Holland would be considerably higher than the equilibrium moisture content occurring in a dry Hawkes Bay summer. The dimensional variation of timber with moisture content depends upon grain orientation. The dimensional change is the greatest in the direction tangential to growth rings, least along the grain and somewhere in between for the direction radial to grain.
The plywood webs would have considerably less dimensional change than the laminated timber because the orientation of grain in adjacent veneers of plywood helps to give dimensional stability. The restraint due to the plywood would cause tension to develop across the laminates of the chord and end block at right angles to the glue-lines. The tension would relieve itself by shakes (a) or by glueline failure (c).
A well made glueline will be stronger than the timber but small local areas of low strength glueline do sometimes occur due to variation in the glueing properties of the timber.
The radial checks in the chord laminates (b) can be explained similarly as due to the restraint against shrinkage provided by two relatively dry laminates upon a laminate of higher moisture content sandwiched between them. Differential moisture content in the order of 5% have been known to cause such checking.
The local delamination of one corner of one ply web from the chord (d) cannot be accounted for by drying shrinkage. The most probable cause is damage resulting from a blow normal to the plywood placing the glueline in tension. This is possible because the ply is slightly proud of the chord at this point. It is noted that the failure is in the glue rather than the timber indicating less than the optimum glue … (text missing, ed.)
The breaking out of the end grain (e) could be caused by poor sawing technique. Some fragments of wood taken from this area are being examined by the Forest Research Institute for evidence of fungal attack. The result of the investigation will be reported later.
Conclusion
(1) The dry checks and shakes may progress further if a spell of hot dry weather occurs to cause further drying to below the moisture content level achieved during last summer.
(2) It would be advisable to check the horizontal shear capacity of the member, when drawings are available with the defects a and c subtracted from the section. The allowable bearing stress on the bolts might also need to be checked.
(3) If restrained construction of this type is to be used the only way to avoid drying checks is to fabricate the member at the timber moisture content approximately the lowest likely to be encountered in New Zealand. I would suggest 10%.
(4) In order to slow down the rate of loss of moisture under drying conditions the exposed end grain should be carefully sealed with a paint of low permeability.
A.L. Poole
Director General
Per: T.M. Wardis