Improving work efficiency and thermal comfort by negative vertical temperature gradient in heating model of impinging jet ventilation

In heating model, impinging jet ventilation (IJV) supplies warm air jet with high momentum which impinges and spread over floor. Thermal stratification in opposite direction (temperature of foot area is higher than head area.) is found in the area near supply opening, because high residual momentum of the jet after impinging can resist thermal buoyancy in a considerable distance on floor surface. Negative vertical temperature gradient was seldom explored. To study how negative vertical temperature gradient affects local thermal comfort and work performance, 16 subjects (8 males and 8 females) were asked to work in a room with three negative vertical temperature gradients (1, 3, 5k/m) between head (1.1 m) and ankle levels (0.1 m). In this study, overall thermal sensation, local thermal sensation of body segments and work performance were collected. The results show that work performance and thermal comfort could be improved by higher negative vertical temperature gradients. * Corresponding author: yangbin@xauat.edu.cn E3S Web of Conferences 356, 03017 (2022) ROOMVENT 2022 https://doi.org/10.1051/e3sconf/202235603017 © The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0 (http://creativecommons.org/licenses/by/4.0/).


Introduction
Impinging jet ventilation (IJV) proposed by Karimipanah and Awbi is a ventilation method that can be applied to winter heating [1].In an IJV room, a jet of air with high momentum is discharged from the supply diffuser at low room level [2], impinges onto the floor and spreads over it [3], and the resulting flow from the IJV is a thin layer of air over the floor because the high supply velocity of IJV can prevent the warm air from rising too quickly [4].The special air supply mode makes the working area within the air supply distance with a temperature distribution of cold head and hot feet, which is different from the common vertical temperature gradient (fig. 1) [5].
Vertical temperature gradient is an important factor affecting indoor thermal comfort.Avoiding local thermal discomfort, whether caused by a vertical air temperature difference between the feet and the head, by an asymmetric radiant field, by local convective cooling (draft), or by contact with a hot or cold floor, is essential to providing acceptable thermal comfort [6].In the study upon which ASHRAE 55 and ISO 7730 [7] specifications are based, Olesen et al [8].exposed 16 seated subjects (8males, 8 females) with light clothing in a tiny chamber at four vertical temperature gradients (0.4, 2.5, 5.0, and 7.5 °C/m) for 3 h in summer.As a result ,only one of the sixteen participants (6.3%) reported discomfort at 2.5 °C/m and two (12.5%) reported discomfort at 5 °C/m;Experiments conducted by Yu et al. and Cheong et al [9] .Had 60 college subjects seated in an environmental chamber [10].A vertical temperature gradient up to 5 °C/m was acceptable.Shi Chao Liu et al. [11] mentioned in the article that the reasons for different researchers' different maximum limits on vertical temperature gradient in the room may be the different evaluation criteria for thermal discomfort caused by vertical temperature gradient in different studies, and that the whole body thermal sensation was not taken as the input parameter of the prediction model.
All of the above researches on vertical temperature gradient are conducted to study the limit of vertical temperature gradient of head and foot in the room with displacement ventilation or floor air supply in summer, but there are very few researches on the vertical temperature gradient of head and foot in IJV heating mode.This experiment will study the effect of adverse vertical temperature difference on thermal comfort in the convective heat exchange room of IJV in winter.

Laboratory layout
Experiment was carried out in an artificial climate experimental chamber inside a large laboratory, which was 5.4 meters long, 5 meters wide and 2.6 meters high.The west and south walls of the artificial laboratory were external walls, the east and north walls were internal walls, the outer laboratory was adjacent to the east wall, and the corridor was adjacent to the north wall.Two stations are arranged in the experimental cabin, each 1.7 meters away from the air-supplying oppening.
Impact jet ventilation adopts a circular air supply duct with a diameter of 0.5 meters and is placed in the middle of the east wall.The air-supplying oppening is 0.4 meters away from the ground, and the size of the return air outlet is 0.4×0.4meters, which is set in the ceiling.

Measuring points and instruments
Ankle position (apart from the ground 0.1 meters), torso position (apart from the ground 0.6 meters) and sitting head position (apart from the ground 1.1 meters) of the dry bulb temperature and air velocity on working station were collected.Randomly selected subjects' skin temperature of head, chest, arm, hand, thigh, calf and foot were also collected.

Conditions and procedures
In the experiment, IJV was used to create a reverse vertical temperature gradient (the air temperature of the feet was greater than the temperature of the head), and the nominal vertical temperature gradient of 1K /m, 3k/m and 5K /m was created by adjusting the air supply temperature.There were three working conditions (Table1).IJV was used to generate negative vertical temperature gradients in the experiment chamber.The temperature at 0.6m was controlled at 19-21 ℃ , and the subjects could adjust the opening and closing of the zipper of the coat by themselves in the first 30 minutes to make the subjects' whole body thermal sensation at a neutral level (the overall thermal sensation vote was at -0.5~0.5)[12].Subjects will come to the lab 20 minutes in advance to paste I-button and understand experiment process.After entering the chamber, subjects can read and browse the web and other activities.During this period, subjects will have enough time to adjust their clothes and adapt to the environment.During this period, data statistics will not be carried out in the questionnaire, so that subjects can get familiar with the content of the questionnaire in advance.The questionnaires were filled in every 10 minutes for 90min.

Questionnaire content
Subjects were asked to conduct individual evaluations of thermal sensation and thermal comfort on both whole and local body parts which included the head, hands, hands and feet to indicate how hot or cold they felt and how comfortable they were.Thermal sensation was evaluated by the 7-point scale suggested by the ASHRAE 55 standard (−3 cold, −2 cool, −1 slightly cool, 0 neutral, 1 slightly warm, 2 warm, 3 hot) and the thermal comfort was evaluated by a 7 point scale (−3 very uncomfortable, −2 uncomfortable, −1 slightly uncomfortable, 0 no feeling, 1 slightly comfortable, 2 comfortable, 3 very comfortable).Subjects used a 5-point scale to evaluate how acceptable they felt about the indoor thermal environment (−2unacceptable, −1 slightly unacceptable, 0 neutral, 1 slightly acceptable, 2 acceptable) and also, their preference for thermal environment was collected i.e., cooler, no change or warmer.

Objective
The figure.2 shows the air temperature and airflow velocity of the station 1.7 meters away from the air supply outlet and 0.1, 0.6 and 1.1 meters away from the ground in case1 (1k/m), case2 (3k/m) and case3 (5k/m).It can be seen from the figure that the air velocity near the station in the experimental chamber is basically the same from 0.1m to 1.1m above the ground.The air velocity is 0.50m/s at 0.1m and 0.15m/s at 1.1m above the ground.The vertical temperature gradients of 1K /m, 3K /m and 5K /m were created in the experiment conditions from 0.1 meters to 1.1 meters above the ground.Even if the air temperature at 0.6 meters was not controlled, the subjects could adjust the opening and closing of the jacket zipper to make the thermal sensation of the whole body reach the neutral level.

Subjective
The statistical analysis and plotting were conducted with origin2018.Our initial data analysis using the Shapiro-Wilk normality test found that the dataset were non-normally distributed [13].We therefore assessed the difference of the median with the paired Wilcoxon signed-rank test [14].Also, the effect size of the difference was calculated in terms of Cohen's d.The thresholds of the Cohen's d were |d| < 0.147 "negligible", |d| < 0.33 "small", |d| < 0.474 "medium", otherwise "large".The statistical significance was based on p < 0.05 (*), p < 0.01(**), and p < 0.001 (***) [11].

Statistics of thermal sensation
Since general sensation can lead to local discomfort, we keep it at the same level of all three temperature gradients.The experiment in this study was designed to maintain a thermal neutral environment for most subjects during the test with the temperature gradient being the only variable.The median (Q1, Q3) thermal sensation of the whole body obtained in each condition was -0.2 (-0.4,0.1), 0.3 (-0.1, 0.6), 0.4 (-0.1, 0.7) (figure .3).

Fig. 3 Thermal sensation varying with vertical temperature gradient
We use Wilcoxon signed -rank test method to test the each part of the body and whole body between the three conditions, every two conditions which have significant difference were calculated the Cohen 's d("medium" and "large" | d | we marked with red * in each group).As shown in the figure .3, the thermal sensation of the whole body and hands did not change significantly with the change of the negative temperature gradient.The thermal sensation of head changed strongly from 1K /m to 3k/m, and the thermal sensation of the feet changed greatly from 1K /m to 3K /m or 5K /m.However, both head and feet responded slightly to the change of the vertical temperature gradient from 3k/m to 5K /m.It may imply that increasing the negative vertical temperature gradient can improve local thermal sensation, but the effect is not obvious after increasing to a certain limit.

Statistics of thermal comfort
The Wilcoxon signed-rank test and calculating Cohen's d method were also used to statistical analysis which was the same as the thermal sensation processing method (fig.4).The thermal comfort of the whole body was significantly changed from 1K /m to 3k/m or 5K /m, and the thermal sensation of the head and feet was significantly improved with the increase of the negative vertical temperature gradient from 1K /m to 3K /m or 5K /m.As with thermal sensation, there was no significant improvement in thermal comfort of the head and feet when increasing negative vertical temperature gradient from 3k/m to 5K /m.It may suggest that a negative vertical temperature gradient can improve the thermal comfort of the room in winter, but the gain effect is not very detailed when the negative vertical temperature gradient is too large.

Statistics of thermal acceptability
The percentage of votes for unacceptable on the whole body, head, and feet in Fig. 5 shows that the unacceptable rating for a certain value of vertical temperature gradient is highest at the feet and the lowest at the head.With little effect on discomfort rating of the head, increasing the vertical temperature gradient from 1 k/m to 3 k/m causes larger changes on the feet and the whole body.The statistics of thermal sensation and comfort suggests that the whole body and head are less likely influenced by negative vertical temperature gradient than feet when temperature gradient increased from low levels (1k/m) to high level (3or5 k/m).And increasing the negative vertical temperature gradient from 1K /m to 3k/m decreased the unacceptable rates of the whole body and the foot significantly, while increasing the temperature gradient from 3k/m to 5K /m did not significantly decrease the unacceptable rate, even increased unacceptability rate of the head.

Fig. 4
Fig. 4 Thermal comfort varying with vertical temperature gradient

Fig. 5
Fig. 5 Thermal acceptability varying with vertical temperature gradient