Nowadays, experimental biology gathers a large number of molecular and genetic data to understand the processes in living systems. Many of these data are evaluated on the level of cells, resulting in a changed phenotype of cells. Tools are required to translate the information on the cellular scale to the whole tissue, where multiple interacting cell types are involved. A recently developed multiscale three-dimensional 3-D simulation framework is designed for tissues in a flow equilibrium of highspeed migrating cells, as found frequently in immunological tissue. Cells can enter and leave the tissue according to some dynamics imposed by the considered biological system. The core component is an agent-based simulation of single cells that is build on top of a Delaunay triangulation. Reminiscent of molecular dynamics, the simulation tool is called Delaunay object dynamics (DOD). The DOD framework is applied to immunological tissue. In particular, the compartmental homeostasis of B lymphocytes and T lymphocytes in secondary lymphoid tissue is investigated on the basis of DOD simulations. This paper clearly demonstrates that the DOD method has the potential to 1) link different length scales between the molecular and the tissue level, 2) be a quantitative method in the sense that each parameter of the model corresponds to a measurable quantity, and 3) exhibit predictive power for novel conclusive experiments. The method can help to understand how global properties of a given biological system emerge from (complex) local interactions. This can help to design and guide experiments to test mechanisms that are identified by theoretical considerations of biological systems. We claim that the DOD architecture is suitable to investigate complex interactions of biological systems with large numbers of highly motile cells or more generally of discrete objects in space.