Design for Disassembly (DfD) is a novel paradigm where parts can disassemble autonomously by being exposed to an external trigger. Using this approach, the production of parts that are more environmentally friendly and easier to recycle becomes possible. The DfD process is started with a short heat pulse, this heat pulse triggers a substantial volume expansion of the utilized material, leading to cracking and final separation of defined sub-assemblies. To design such DfD-objects, a software that is capable of handling at least two materials is required. The current predominant vector-based approach is not suitable for modelling structural interactions at the level of individual voxels, which is a requirement for designing parts for DfD. Therefore, the voxel-based approach is chosen for this application. One of the main challenges with this approach is the compression of data, since the uncompressed voxel-based data would require too much memory. Various compression methods and different setups of these methods are investigated. The goal of these tests is to find a data-structure that can handle the required data for DLP-based stereolithography. For our machine that means a resolution of up to 1920*1080 pixel and two materials (the base material and the DfD specific material). The main advantage of a voxel-based approach must still be considered when compressing data: For each smallest printable point (voxel) the material can be defined. Our work shows challenges that arise with compression of data: The maximum achievable compression rate, the access times and the required time for compressing the data. The outcome of this research effort is a data structure designed for the creation of components that follow the DfD principles. This advancement represents a crucial step towards a complete voxel-based modelling environment, where parts for DfD can be designed and efficiently processed in the following manufacturing chain.