Abstract
The design of adhesive joints is discussed. The simplest distribution of applied stress occurs when two thin-walled circular tubes are joined coaxially by an annular adhesive layer which is normal to the axis and of constant thickness. When this assembly is subjected to torsion, the resultant pure shear is evenly distributed throughout the joint, in distinction from any other design. This is therefore a potential technique for measuring the fundamental adhesive strength. The published work on this system is reviewed, with particular reference to the effect of varying the joint dimensions.
Experimental work with one adhesive (M.S.9160) is described, in which joint strengths independent of the dimensions of the steel adherends are achieved, although the scatter is large. Thus it is confirmed that this joint design offers considerable potential advantages over others in yielding fundamental data.
In contrast, with another adhesive (R.D.1286) the joint strength is found to depend on the joint area. This effect is perhaps due to the partial retention of water formed during setting.
The joint thickness has been controlled by including wire spacers in the adhesive. This technique has been shown not to affect the joint strength.
A reduction in strength is found with increasing thickness, confirming general experience with many other designs of joint.