A simulation based on the Monte Carlo method allows to introduce uncertainty into a mathematical model. This is
achieved by randomly varying the input parameters and conducting tens or hundreds of deterministic analyses. The
aim of this work is to show that such a stochastic simulation provides a deeper insight into the behaviour of space
structures than purely deterministic analysis. This has been proven for the structural model of a space telescope. The
simulation reflects the main sensitivity of the real hardware. For a comparable telescope the simulation indicates were
a redesign could improve its performance. The finite element solver Nastran and the stochastic software MSC.Robust
Design were applied. Practical advice is given on how to prepare a structural model for a stochastic simulation. - A
simple model was tested to reveal the characteristics of a method called "stochastic design improvement" (SDI). Other
than optimisation an SDI takes uncertainty into account. Thus, for every iteration a cloud of samples is generated. This
cloud is moved towards a target, i.e.\ the desired performance. SDI does not aim at creating an optimum but a cluster
of acceptable solutions. The solution which is robust has to be identified afterwards. Where an optimal design exhausts
the resources a robust design can absorb uncertainty. - The greatest advantage of the Monte Carlo method is its
simple idea of repeating the same task several times. This allows to spend more time in studying the results of the
simulation than in tuning complex mathematical algorithms. Since the method is based on randomness, it showed to
be very robust. Another advantage is that it can be applied to various engineering tasks. The computational effort is
acceptable if simulations are run outside office hours. Utilising its intrinsic parallelism is not limited by the hardware
but the number of solver licences available. Finally, a stochastic simulation allows to fully exploit the potential of an
already existing model.