A relaxation-based probabilistic approach for PDE-constrained optimization under uncertainty with pointwise state constraints