Loughborough University
Thesis-2011-Kelly.pdf (13.85 MB)

Thermal comfort on train journeys

Download (13.85 MB)
posted on 2011-06-07, 07:48 authored by Lisa K. Kelly
This thesis presents a body of work conducted to determine thermal comfort on train journeys. Relatively little research has been conducted on trains in comparison with the vast body of work conducted within building environments. This thesis aimed to expand our knowledge of rail passenger thermal comfort throughout the journey; platform to destination. The train journey was separated into its component parts and analysed by conducting both laboratory and field experiments that either simulated or measured aspects of a train journey. Laboratory experiment 1 examined appropriate methods of data collection during train journeys. Participants (9 males and 9 females) were exposed to a simulated train environment three times and used a different data collection method on each occasion; a paper-based method, a voice recorder or a Personal Digital Assistant (PDA). Results concluded that the three methods can be used interchangeably when recording thermal comfort data. Participants preferred the PDA over the other two methods because they felt it afforded them a level of privacy in addition to blending in with other rail passengers using similar technologies. The second laboratory experiment measured thermal comfort following a change of environment. Participants (12 males and 12 females) were exposed to three environmental conditions (warm, neutral and slightly cool) in a thermal chamber on three separate occasions. The exposure lasted 30 minutes, after which, participants entered a new environment that was the same on each occasion (slightly cool). Results showed that overshoots in sensation (beyond those predicted by the Predicted Mean Vote thermal comfort index PMV) are observed following downward steps (warmer to cooler) in environmental conditions. No overshoots were observed following the upward step (cooler to warmer) in environment, with sensations immediately reflecting the predicted steady-state values. Laboratory experiment 3 (22 males and 26 females) expanded the research conducted in laboratory experiment 2 by exposing participants to greater magnitudes of environmental change. In addition, sensation was measured after this change until steady-state was reached. Participants were exposed to four environmental conditions (cool to warm to neutral to cool or cool to cold to warm to cool) consecutively over a 2 hour period with 30 minutes spent in each location. Results demonstrated similar effects to those observed during laboratory experiment 2 with overshoots observed following downward steps in environmental conditions and none observed in the opposite direction. Sensations demonstrating overshoots gradually increased until steady-state was achieved after approximately 25 minutes. Field experiment 1 (12 males and 32 females) measured thermal comfort while boarding trains. Participants were taken on a short train journey and recorded sensations whilst on the platform and during boarding. Results showed that overshoots may also be observed following step up and step down in environments. It is hypothesised that change in air velocity is influential in this effect. Thermal comfort throughout a train journey was measured in field experiment 2. Participants (16 males and 16 females) reported on thermal comfort on the platform, during boarding and throughout a return train journey from Loughborough to London St Pancras. Results also demonstrated overshoots following upward transients indicating that there are factors in the field that do not occur in laboratory conditions. Subjective parameters reach steady-state after approximately 20 minutes and PMV accurately predicted sensations during the journey. Again, air velocities may have interacted with other variables resulting in the overshoots following upward steps in environmental conditions. Laboratory experiments 2 and 3 resulted in the creation of a model predicting sensation following a change of environment, PMVTRANS. When the model was compared with the field data, it could not accurately predict sensations observed during transients. It also could not predict the sensation overshoots observed following upward transients. A new model is now proposed, NEW PMVTRANS. This model shows greater correlation with actual sensation than PMV; however it does require further validation from field data. Research has shown that PMV is an accurate estimator of sensation within a train carriage and should be used by train designers to optimise the environmental conditions for passengers.



  • Design and Creative Arts


  • Design


Loughborough University

Rights holder

© Lisa K. Kelly

Publication date



A Doctoral Thesis. Submitted in partial fulfillment of the requirements for the award of Doctor of Philosophy of Loughborough University.

EThOS Persistent ID



  • en


Ken Parsons

Qualification name

  • PhD

Qualification level

  • Doctoral

This submission includes a signed certificate in addition to the thesis file(s)

  • I have submitted a signed certificate

Usage metrics

    Design Theses


    Ref. manager