Conventionally, simulation of product behaviour is employed as a pre-realization type of assessment at the end of the design process, making only late feedback for improvement possible. Enabling the start of optimization in the conceptualization is expected to have significant influence on design efficiency. However, the available information at that stage is uncertain, incomplete, multifold and imprecise, which calls for new simulation techniques. This paper proposes nucleus-based modelling and simulation as a solution. A nucleus is a modelling entity to capture the relationships between the lowest level metric elements of the product and to represent the physical effects governing the behaviour of the product. Tolerating uncertainty, incompleteness, modality and imprecision, a nucleus-based model is able to provide an integral model of the actors of the use process. Simulations are controlled by so-called scenarios that arrange a logical structure of feasible situations for the integral model. The paper describes the content of the nucleus-based integral model and presents an application case study to illustrate the potentials of this new approach.