Abstract. Experiments in a laboratory or a company, either physical or virtual, generate a large amount of data and information (raw data, consolidated results, experimental conditions, etc.) that has to be managed in order to preserve and enhance the knowledge and the expertise of the organization, especially when turnover happens. If simulation data management has been especially explored and standardized (STEP AP 209 for instance), experimental data did not create the same interests and only standard data management tools are proposed. In this paper, we propose an ontology to capitalize such information and knowledge, with an efficient reuse objective. This ontology, for which specific taxonomies have been defined according to the relevant literature of the domain, is illustrated on a set of data on previous experimental campaigns, corresponding to two kinds of mechanical experiments: tribology tests and material characterization tests.