The project has first focused on the definition of the Multi -scale and Multi-Disciplinary Model. The MSMDM is a set of interconnected models necessary to achieve the project goals. It integrates the various physics and the various scales as well as the relevant interdependencies to enable an improved prediction of the mechanical behaviour of a precooler heat exchanger (Hx) in operation. Thus, a proper architecture of models of the MSMDM had be defined and applied in order to derive an efficient, reliable and evolutive MSMDM. This architecture is the definition of all the required necessary models, including scale and physics, as well as the workflows which need to be addressed and therefore the required data transfers between software/tools.
The architecture of models of the MSMDM finally obtained is synthetized in Figure 1. It describes the process which leads to a relevant modelling of the thermomechanical phenomena occurring in an Hx. It shows the matrix nature of this architecture in the physics dimension (Fluid mechanics, Thermal & Mechanical) as well as in the scale dimension (0 to 3).
All the required level and physics required have been defined as well as the interconnection (data transfer)
Besides, in order to perform mechanical simulation of local areas with relevant boundary condition a hybrid modelling methodology has been identified, which relies on the derivation of a macroscale model of the Heat exchanger obtained through an homogenisation process.
Also, the statistical tools to perform the sensitivity analyses have been defined with a specific focus on the uncertainties within the heat exchanger, should they be the uncertainties arising during manufacturing or during usage, and their impact on the structural and/or thermodynamic performance, and the reliability of the system. To model the potential defects, the choice was made to induce geometrically correlated errors, using stochastic processes.
The developments are in progress for now and first tests on basic problems have been initiated, and the results are for now in line with the results presented in the literature.
The mechanical testing on material samples has also been defined, from the nature of required tests to the number required and nature of required specimens. Then, the test apparatus has been tested to ensure that all tests could be performed with a good level of confidence, at room temperature and at high temperature for both tensile tests and fatigue tests.
Some testing has been performed on the thinnest samples (50 microns) with dummy samples so that the feasibility of testing such a narrow sample is ensured. The main goal with these tests was achieved, proving feasibility of testing very thin samples. This is illustrated in Figure 3, where different samples were tested with different gauge lengths but all of them with 50 microns width. The graph bellow shows the strain-stress curves being noticeable that some differences or uncertainties exist from test to test , but these are due to the manufacturing process of the dummy samples.
In the meantime, some non-destructive testing methodologies have been studied and evaluated to identify the best candidate to include crack/failure detection in in-service compact heat exchangers
