Abstract

Acceptable approaches that can be utilized for establishing safe operational loads for adhesively-bonded composite joints invariably adopt an integrated methodology. The methodology selected typically combines theoretical analysis, both numerical and closedform, supported by a carefully-planned experimental program. Speciflc details of the approach may range widely with regard to the balance of analysis versus experimental content as dictated by the environmental conditions and the nature of the processing and fabrication methods that are to be adopted. For the orbiting space instrument platforms of interest in this paper the range of temperature extremes are progressively increasing and consequently major concerns with respect to thermal stress and distortion of composite and metallic assemblies have become particularly challenging. In most instances a combination of analysis techniques are utilized for space hardware designs, e.g. a macroscopic flnite element analysis to deflne mechanical and thermal loadings developed in the various structural joints and attachments followed by use of either detailed closed-form analyses or more reflned flnite element analyses of the individual joint conflgurations. As an example the complex arrangement of structural joints on the Mars Exploration Rover (MER) Lander Structure was analyzed by using a global flnite element model plus a series of theoretical analyses of adhesive bondline stresses based on a combination of Volkersen, shear lag and Beam-on-Elastic-Foundation models. These closed form expressions were used to evaluate the distributions of shear and peel stresses developed in the bonded joint details. In this paper the acceptability of the above methodology is evaluated using, as background, results of previous UCSB research, directed at general aviation composite aircraft structures for which the mechanical loading conditions are the dominant concern. Acknowledging that most geometrically practicable designs of bonded lap joints exhibit high shear and peel stresses in the end regions of the lap length conflned to less than 5% of the total lap length. To illustrate the localized nature of the bondline stress distribution the more critical region at the termination of the outer adherend of a double lap joint where general tensile loading is considered. Two flnite element modeling techniques were also used to predict both the shear and the peel stress distributions also along the adhesive central plane. A simplifled flnite element model (FEM) utilizes two noded interface elements to represent the adhesive layer, and combines both shear and peel efiects where appropriate. The results for this model are illustrated and indicate slightly lower maximum shear stresses but again violates the free edge condition. Finally a more conventional 8-node plane strain quadrilateral FEM is used for both adhesive and adherend regions and is shown to yield the lowest peak shear stresses. It is noted here, however, that by using the relatively high fldelity mesh subdivision indicated that the vanishing free edge shear stress condition is satisfled with this model. Recommendations will be presented along with expressions for predicting peel stress states developed under mechanical and thermal loadings. The conclusions are intended to serve the design/analyst who is chartered with the task of developing reliable criteria, based on selective experimental evidence, for bonded composite joints.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call