A review of human thermal comfort experiments in controlled and semi-controlled environments

There are three main methods to improve thermal comfort in existing buildings: modeling, experiments and measurements. Regarding experiments, no standardized procedure exists. This article provides an answer to the question: "What is the most common practice for human thermal comfort experiments in (semi-) controlled environments?". A total of 166 articles presenting results on 206 experiments were collected and analyzed to extrapolate the most common practice. The results are arranged in five main themes: subjects (e.g. number and age), climate chamber (e.g. surface area), thermal environment, experimental procedure (e.g. phases and duration), and questionnaire. A typical experiment was found to employ 25 subjects and to take place in a permanent climate chamber with a floor area of 24 m(2) During the experiment, 3 air temperature variations are used. The test itself takes 115 min, but is preceded by a preconditioning and conditioning phase. The subject is given a questionnaire at regular intervals of 15 min, with questions highly dependent on topic, but including thermal sensation and comfort vote rated on a bipolar 7-level scale. Number of subjects, gender distribution, type and floor area of the climate chamber and utilization rate of the scale for rating thermal comfort and sensation are all linked to topic, as well as number of different air temperatures, whether conditioning is employed and questions in the questionnaire. Several links between experiment characteristics reciprocally are also identified.