The increasing competition and the urge to reduce costs and shorten product-life-cycles in the automotive industries imply the application of virtual manufacturing mock-ups. Computer aided engineering methods including simulations of the structural behavior during manufacturing and assembly contribute to eliminate trialand-error along the production chains of vehicle structures. Such digital mock-ups also provide an input for additional tolerance analyses and allow the identification of the pending quality of structure components in advance. In this way large numbers of experiments and adjustments on the existent production lines can be avoided. In this paper an approach is introduced, which allows the replication of the physical effects along manufacturing chains based on thermo-mechanical models. The rendered results after considering successive operational steps provide the basis for initial estimations regarding the shape accuracy of structure assemblies. For a specific example of the automotive body-in-white a first run of a manufacturing process chain is investigated in order to predict the shape quality of a probable production cycle by applying a tolerance chain analysis. The results enable the scrap identification due to non-satisfied quality of assemblies. Finally, based on the achieved results a re-engineering of the manufacturing chain is suggested in order to fulfill quality requirements.Keywords Computer aided engineering Á Body-in-white Á Tolerance chain analysis 1 Background and motivation A major aim of production plants is to continuously preserve high efficiency and product quality in order to retain competitiveness and achieve maximum profitability. A central sector in automotive industries and particularly where high reproducibility is feasible and concurrently high product quality is required, is the body-in-white [1].In order to identify errors and predict inaccuracy of the final product, different methods have been developed to observe product quality along the manufacturing process chain. For this purposes an online acquisition of key quality attributes in the body-in-white with the aid of sensor systems concerning i.e. apparatus setups, material properties or semi-product accuracy is recommended [2]. Moreover such quality characteristics can be estimated in advance by applying computer aided methods, initially for single manufacturing operations and afterwards for complete manufacturing chains of major assemblies of modern automotive light-weight structures [3].Virtual manufacturing mock-ups are nowadays widely used to support production lines. They may span optimization algorithms to define successive steps within a manufacturing operation [4] or virtual reality programming in order to enable tele-presence [5]. Furthermore, simulations with the aid of the finite element analysis allow the replication of the mechanical and heat effects of manufacturing and assembly tasks [6,7] up to the crash properties of frame structures [8]. Such digital mock-ups can additionally contribute to the qua...