1

Adhesive Coupled Aluminum Aircraft Joints By R. W. Bradley (EAA 12788) 44 Beach Street Marblehead, Massachusetts

W. Tabroff

and (EAA 52103)

47 Raymond Circle

Peabody, Massachusetts

A

SIGNIFICANT advance in the quality of amateur-built aluminum aircraft can be realized through the use of reliable adhesive bonds between the structural parts. We have all witnessed the successful use of aluminum adhesive-bonded joints in commercial and military aircraft for over the past 20 years. The integrity of the adhesive joints is based on carefully controlled processing which in some form should be approached by the amateur. With the proposed metal processing minimal

labor and cost are achieved along with ease of fabrication. In comparison with riveted joints, the adhesive structures show an increased lap-width efficiency of from two to three times. The aerodynamics are excellent, eliminating dimpling, 'riveting, and filling. The adhesive structure shows improved life and reduced joint corrosion. Part of this improvement in life comes from the adhesive's reduction of joint vibration. We have all noted that in riveted aircraft structures, the rivets vibrate loose to lower the strength to a point where entire cowlings must be periodically removed for re-riveting. Surprisingly, this is similarly

true of both standard and blind rivets.1 Additionally, the use of adhe34

SEPTEMBER

1970

SUGGESTED SMOOTH CONTOUR STRIP JOINT DESIGN: An improved strip structure for simple assembly to be used in conjunction with Si lane 6050 treated aluminum and epoxy adhesive. The rivet is a 7/64 in. plated monel rivet with 300 Ibs. shear and 350 Ibs. tensile strength. No. MK321 BS3. Note how the epoxy dip has followed the mandrel head up up into the shell to hold the head and seal the rivet. Epoxy top-filled, this should make a beautiful high-strength vibration-free assembly.

sivcs reduces the number of potential hole sites where strain initiation or crack propagation can occur. The greatest deterrent to progress of adhesive-metal joints has been the optimal surface treatment that is necessary; it is not as easy as indicated in much of the sales literature. It could be disastrous for a 3000-4000 Ib./sq. in. of shear epoxy cement to be only buttered onto aluminum for joint use and fail through the lack of adhesion to either surface. The least trace of grease or corrosion can ruin the adhesive joint strength. The accepted surface treatment — MIL (FPL Etch) consisting of vapor

degrease, alkaline degrease, and sulphuric acid etch — is practically impossible for the homebuilder to use. The process must be isolated and controlled, as even the fumes can corrode shop tools. Because of these problems, another method of metal surface preparation more practical for the homebuilder is needed. In our investigation of various products it was found that metal-to-plastic coupling agents such

as the Dow2 silanes offer an answer. For example, a diluted solution of a suitable silane when painted onto aluminum provides surface adhesion and reactive sites for epoxy poly-

merization and adherence. The coupling agent also require moisture penetration into the epoxy joint. A controlled test program was conducted, the results of which indicate great possibilities. Some questions have been raised as to how good the Dow silanc-trcated joints are relative to the accepted MIL-FPL surface treatment. The data given here arc the result of preliminary standard pull shear tests following treatment with cither of two silane coupling agents. It appears that even very simple surface preparation consisting of acetone wash and silane solution coating can result in a highly reliable joint.

These tests were made with critically thin cement films of approximately .001 in. thickness in which serious variations in joint strength should show up immediately. It should be cautioned here that critically thin or starved cement film joints, such as used in these tests, must be avoided in normal aircraft

manufacture. A .003-.007 in. adhesive thickness is considered in the practical optimum range. At this range, aircraft vibrations on one side of the joint have a chance to damp out before reaching the other side of the joint to prevent adhesive shear. It is

surprising how practically easy it is to put joint parts together resulting

in an adhesive film thickness of .005 to .010 in. It is a matter of experiments, adhesive squeeze-out, and slight joint pressure. Typically now, the homebuilder would sand the areas to be treated

Shown here is a letter from iwo local FAA inspectors who have been contacted relative to this type of construction.

20

«9 70

with No. 120 to No. 240 grit cloth,

acetone spray wash the areas twice, and brush on a diluted solution of Dow's 6050 silane (0.5 in .methanol 99 percent, water 1 percent). Immediately after the silane solution is airdried, the epoxy adhesive is applied to both parts. Leave a fillet of cement around the edge of each joint to retard further moisture penetration into the joint. Do not use MEK for cleaning, as many commercial MEK samples contain undesirable

residues. Happily, this system now allows the easy use of "closed-end"3 solidaluminum "Pop" rivets to prevent aluminum to adhesive peel and eliminate the damage inflicted by riveting hammers and back-up irons. A stronger, vibration-free, long-life structure has been accomplished. As authors, we fully agree with the local FAA inspectors' opinion reported in the accompanying letter, and sincerely appreciate their interest and effort. It becomes obvious that meaningful tools must be put into the amateur's hands for top quality adhesive bonding, for example: • Measured amounts of silane 6050 for a one percent solution (twice the strength used in tests) should be available to mix in given quantities of methanol, and should be kept in a dessicated chamber • which will be provided;4 • To eliminate s t a r v e d cement joints, rounded inert fragments (mostly silica) graded from .0035 in. to .010 in. should be made available. About a half dozen particles sprinkled on the spread cement surface before assembly has given excellent control;4 • An improved smooth contour strip joint design which can be assembled without pounding and which has sufficient strength without adhesive should be advanced. See the suggested strip construction and note that the epoxy of the dipped rivet has

followed the mandrel head into the sleeve and hold the head as a wedge and make the rivet watertight.3 • A small, low-cost, reliable shop adhesive test strip strength measurement system to guarantee the

(Continued on next page)

.!^^£^2. " ~"~i',';-

.•ssi's,--— «„

^^•^si?*^^ 'j