本計畫的第一年，成功的建立的數值模擬平台，節省了大量的實 驗材料成本。而經由實驗平台得到的測試結果與數值模擬之評估結果 比對，更確立了模擬平台的準確性，故此完整實驗與數值結合之平台 可作為垂直軸風力發電機之測試基準依據。
依照測試的結果對所設計的風力發電機的風車部分做了最佳化 的改良，進而找出了適合往後與光能結合的最佳化效能實驗平台。
112
計畫編號：NSC 99－2221－E－011－077 學門領域：流體力學
技術/創作名稱垂直軸風力發電機的數值模擬與性能測試之整合分析平台
For the aspect of establishing a reliable numerical means, incorporated with LES scheme. Also, its aerodynamic characteristics are close to the experimental result after the supporting arms of blades is taken into account in calculating the VAWT performance. In addition, a parametric study, including blade number, radius, and solidity, is carried out with the aids of this new simulation model for realizing their corresponding influences on the VAWT performance.
可利用之產業及
113
推廣及運用的價值
透過本技術，得以更快速的完成風力發電機之設計，有助於提升台灣在
能源開發技術。
114
參考文獻
[1] Paraschivoiu, Ion, “Wind Turbine Design with Emphasis on Darrieus Concept,” Ecole Polytechnique de Motreal, 2002.
[2] Sandia National Laboratories Staff, ”Vertical Axis Wind Turbines-The History of the DOE Program.”
[3] Strickland, J. H. “The Darrieus Turbine: A Performance Prediction odel Using Multiple Streamtubes,” SAND75-0431, 1975.
[4] Blackwell, B. F., Sullivan, W. N., Reuter, R. C., and Banas, J.F.
“Engineering Development Status of the Darrieus Wind Turbine,” J.
ENERGY, Vol. 1, No. 1, pp, 50-64, Jan. 1977.
[5] Blackwell, B. F., and Sheldahl, R. E., ”Selected Wind Tunnel Test Result for the Darrieus Wind Turbine,” J. ENERGY, Vol. 1, No. 6, pp, 382-386, Nov. -Dec. 1977.
[6] Sheldahl, R. E. and Bleckwell, B. F., “Free-Air Performance Test of a 5-Meter-Diameter Darrieus Turbine,” SAND77-1063, 1977.
[7] Klimas, P. C. and Sheldahl, R. E., “Four Aerodynamic Prediction Schemes for Vertical-Axis Wind Turbines: A Compendium,”
SAND78-0014, 1978.
[8] Sheldahl, R. E., Klimas, P. C., and Feltz, L. V., “Aerodynamic Performance of a 5-Meter-Diameter Darrieus Turbine with
115
Extruded Aluminum NACA-0015 Blades,” SADN80-0179, 1980.
[9] Johnson, D. A. and King, L. S., “A Methematically Simple Turbulence Closure Model for Attached and Separated Turbulent Boundary Layers,” AIAA, Vol. 23, No. 11, pp, 1682-1692, Nov. 1985.
[10] Wolfe, W. P. and Ochs, S. S., “CFD Calculations of S809 Aerodynamic Characteristics,” AIAA-97-0973, 1997.
[11] Brian, K. K., “Evaluation of Self-Staring Vertical Axis Wind Turbines for Stand-Alone Applications,” Grifftth Univerisity, Apr.
1998.
[12] 嚴坤政，“小型風力發電系統設置與葉片氣動力分析”，南台科技 大學機械工程研究所碩士論文，2003 年。
[13] 孫明忠，“風力發電機葉片快速設計程序與軟體設計”，長庚大學 機械工程研究所碩士論文，2005 年。
[14] Chua, K.L., http://windturbine-analysis.netfirms.com
[15] Fujisawa, N. and Shibuya, S., “Observations of Dynamic Stallom Darrieus Wind Turbine Blades,” Journal of Wind Engineering and Industrial Aerodynamics, Vol. 89, pp, 201-214, 2001.
[16] Kirke, B. K., “Evaluation of Self-Staring Vertical Axis Wind Turbines for Stand-Alone Applications,” Griffith University, Apr.
1998.
[17] Young, D. F., Munson, B. R., and Okiishi, T. H., “A Brief Introduction to Fluid Mechanics”, 2nd Edition, John Wiley & Sons, Inc., 2000.
116
1
(英文) The 4th WSEAS International Conference on Fluid Mechanics and Heat & Mass Transfer
發表論文 題目
(1) Real-time thermal model of chemical mechanical polishing.
(Paper ID 303-251 )
(2) Influence of elevator moving pattern and velocity on the airflow uniformity for an LCD panel deliver.
(Paper ID 303-234 )
2
國科會研究計畫出席國際研討會報告
研討會名稱
The 4th WSEAS International Conference on Fluid Mechanics and Heat & Mass Transfer
July 14- 17, 2011 Corfu Island, Greece
所發表之兩篇論文:
Real-time Thermal Model of Chemical Mechanical Polishing.
(Paper ID 303-251 )
Influence of Elevator Moving Pattern and Velocity on the Airflow Uniformity for an LCD Panel Deliver.
(Paper ID 303-234 )
報 告 人: 林 顯 群 教授
國立台灣科技大學機械工程系
3
此次參加第四屆 WSEAS 流體力學與熱質傳聯合研討會是由 WSEAS 學會所舉辦，共有四百七十八篇論文投稿發表，所有論文皆 經過全文審查通過才被接受，論文皆經過適當之審查。此次研討會屬 於大型之多項主題聯合國際性會議，研討會期間有來自三十幾個國家 之相關學者，齊聚在希臘科芙島之假日皇家飯店切磋研究成果，雖以 歐洲國家論文數較多，但整體之國際性參與度佳。並發表之三篇論文:
(1) Real-time Thermal Model of Chemical Mechanical Polishing. (Paper ID 303-251 )
(2) Influence of Elevator Moving Pattern and Velocity on the Airflow Uniformity for an LCD Panel Deliver Facility.
(Paper ID 303-234 )
此次大型國際聯合研討會論文總共有四百七十八篇論文投稿 發表，分成六十八個 Section，分別在六個會場及四天內發表，研討 會期間可於會場加入有興趣之論文發表會場次。由此次研討會期間除 了解到流體力學與熱質傳之國際發展現況，亦可參與了解國際性相關 領域之研究與應用，以及在各先進與開發中國家之發展與現況，此次 國內有多篇論文在此研討會發表，但是今年歐洲來回旅費激漲，我們 皆是一半自費前往參加，很高興國內有多位學者參與類似領域之國際 研討會，提高國際能見度與建立更好之國際溝通討論基礎。
7 月 13 日下午由台北出發，經過二十幾小時之飛行及轉機等待，
於 7 月 14 日下午三點左右才到達此次第十五屆 WSEAS 國際研討會
4
之地點，希臘科芙島之假日皇家飯店，先行安頓好住宿問題。到達會 場 即 快 速辦 好註冊 手 續 ，趕 上第一 篇 論 文發 表之場 次 ， 題目 為 Influence of Elevator Moving Pattern and Velocity on the Airflow Uniformity for an LCD Panel Deliver Facility. (Paper ID 303-234 )是在 節約能源方面的研究。每篇論文發表時間 15 分鐘，在論文發表後下 午則穿梭於 Thermal/Fluid 之 Session 聆聽有興趣之論文發表。此飯店 為郊區度假型 RESORT 形式，離市區有點距離，還好當初透過旅行 社安排，直接訂住此四五星級飯店，雖然價錢昂貴到非國科會補助之 日支費總額足以進駐，但是可直接參與相關研討會之所有活動，省去 來往研討會旅館之交通情況，基本上是個較為合適之安排。
第 二 天 主 要 參 與 Heat Exchanger 與 Thermal management in electric devices 兩個領域之論文發表會場，第二篇論文發表之題目為 Real-time Thermal Model of Chemical Mechanical Polishing. (Paper ID 303-251 ) ，是本人近年來產業與學界之合作研究重點的研究成 果。
大會除了 Keynote Lectures 與 Plenary Lectures 之邀請演講大廳外，
另有其他五個研討會會場進行論文之同步發表，四天研討會期間，總
共有 68 個研討會 Sections，從早上八、九點開始， 每天有五個 Sections
連著，一直到晚上八點鐘皆有論文發表，以消化四百七十八篇投稿論
文。第二天到第四天 (7 月 15 日到 7 月 17 日) 個人則選擇性之參與
5
各會場之相關研究主題之論文研討會活動，此研討會 Sections 每天皆 排定六個會場，每場約六到十篇論文表，並沒有詳細將論文領域分得 詳細，是以要聽特定領域之研究成果發表，倒是因此可稍微了解各領 域之應用情況。上下午各有一個 Coffee Break 時間以供大家休息交 誼，此次會議參與人員泛歐洲國家佔多數，是以英文口音皆不太純 正，需要稍微仔細聆聽，溝通沒問題 。
本次大會科芙島之假日皇家飯店因位於海邊之坡地上，有一 個淡水游泳池與一個海水浴場可供大家休閒活動筋骨，旅館建有小型 電纜車可來往上方游泳池與海邊，第二天晚上之大會晚宴就在游泳池 旁進行自助式晚餐，是在當天最後一個 Sections 完成後之晚上九點才 開始，食物有希臘風味，但是亦不是很豐富，也安排希臘傳統舞蹈表 演。此項國際研討會之特色為，其投稿論文涵蓋機械方面之所有相關 研究領域，可藉機參與聆聽各個研究主題之考量，稍為了解相關問題。
Influence of elevator moving pattern and velocity on the airflow uniformity for an LCD panel
delivery facility
Sheam-Chyun Lin, Bor-Jang Tsai, Cheng-Ju Chang
Abstract—Owing to the increasing LCD panel size, the difficulty on delivering the glass substrate has been enhanced dramatically and become a critical problem in the LCD manufacturing industry. Nowadays, most of panel fabrication factory utilize the fully-automated delivering technology instead of the traditional labor delivery for diminishing the possibility of polluted particles on the LCD board. Thus, this study intends to investigate on maintaining the air quality inside the delivering facility with a moving elevator. Also, special emphasis is focused on the influence on the moving pattern and velocity of the elevator via numerical technique. Firstly, CFD code Fluent is used to execute the transient flow simulation and evaluate the flow patterns inside this delivery equipment. From analyzing the calculated results, it is found that the inferior air is generated mainly by the increasing vortex inside the delivery equipment for an upward-moving elevator. On the contrary, the flow field becomes very smooth without obvious vortex phenomenon, and thus induces a better air quality when the elevator moves downward. However, a better uniform flow field occurred when the elevator is moving upward. In addition, the airflow uniformity is not effectively improved by reducing the elevator velocity and increasing the FFU airflow velocity. It is concluded that the moving pattern of elevator has an essential impact and can be utilized to improve the air quality inside the LCD delivery facility.
Keywords—Delivery Facility, LCD, Transient Simulation, Vortex
Sheam-Chyun Lin is with the Department of Mechanical Engineering National Taiwan University of Science and Technology Taipei, Taiwan.
(Phone: 886-2-27333141#6453; fax:886-2-27376460; e-mail:
Bor-Jang Tsai is with Department of Mechanical Engineering Chung Hua University Taipei, Taiwan. (Phone: 886-3-5186478;
fax:886-3-5186478; e-mail: [email protected])
Cheng-Ju Chang is with the Department of Mechanical Engineering National Taiwan University of Science and Technology Taipei, Taiwan.
(e-mail: [email protected])
I. INTRODUCTION
Recently, low-pollution process and superior quality product have drawn significant attention due to the increasing demands in semiconductor, biotechnology, and pharmaceutical industry. Due to the extra small size and rigorous purity requirement on the high-performance products, it could result in a vital damage if there is a 0.1 micro-meter pollute particle adhere to its surface during the process. Therefore, many alternatives are proposed to solve those contaminating challenges for reducing the individual damage and increasing product’s yield rate. For example, dealing with multiple glass substrates in the same process is used frequently and proven effectively for the situations adopted the NEMI (National Environmental Methods Index) or JEDEC (Joint Electron Device Engineering Council) standard. Nevertheless, installing those substrates requires extra step and thus increases the damage probability.
This scratch problem becomes even worse in manufacturing the LCD panel due to the fragile TFT-LCD glass, which can be scraped because of a slight speed difference between two delivering motors in the transmission belt. Moreover, effectively controlling the cleanness level in the delivering facility turns into a crucial technology for ensuring a satisfactory yield rate. Therefore, maintaining the air quality inside the transport elevator becomes the topic of this research.
In 1996, Hu et al. [1] used an ultrasonic anemometer to measure the velocity distribution in a full-scale clean room equipped with the FFU (fan filter unit) system. The results showed that the non-uniformity and deflection angle Recent Researches in Mechanics
ISBN: 978-1-61804-020-6 81
of airstream inside a clean bench are 13% and 7.9 degrees, respectively. In 1999, Cheng et al. [2] used a finite-volume CFD code STAR-CD to simulate the flow field of clean room by solving the Reynolds-averaged Navier-Stokes equations. The numerical results demonstrated that the airflow uniformity increases along with the proper allocation of floor porosity or by controlling the distribution of inlet velocity profile.
In 2010, Shih et al. [3] investigated the particles wafer moving velocity. Later, Giannoulis [4] examined the airflow around a raising panel via experimental and numerical efforts. Numerical and experimental outcomes correlate well and illustrate a significant difference between the airflows around the panel covered by impermeable and permeable materials. Clearly, the airflow around the elevating panel becomes smoother when the plastic film is replaced by permeable nets.
In 2010, Lambert et al. [5] utilized CFD code to estimate the time-elapsed decay of contaminants within a chamber experiencing a high-Reynolds-number flow. They found that despite different flow rates, the measured contaminant washout took 12~13% longer than the numerically predicted value. Furthermore, deviation between computational and experimental data is as low as 5.32%, which implies that CFD is a useful tool for studying ventilation phenomena.
Kobayashi et al. [5] used experimental and CFD tools to analyze cross-ventilated flow through a single room.
Their emphases are (a) to clarify and understand the airflow characteristics on the windward vortex and the leeward wake for various openings; (b) to explore the flow pattern above the ground from pressure measurements; (c) to identify the accuracy of CFD. The results showed that the
pressure and vortex dropping decreased with an increased opening size. Recently, Saidi et al. [6] numerically evaluated the effectiveness of ventilation system in a full-scale clean room. The results show that the contaminant source motion and its path have a great influence on the contaminant dispersion through the room.
Based on the literature reviews, it is summarized that the polluted particles have a severe impact on the yield rate and CFD code employed dynamic mesh can provide accurate flow visualization for analyzing the physical phenomena in a cleanroom. Hence, this study utilizes CFD code Fluent [7] together with the dynamic-mesh technique to investigate the transient flow field inside the LCD delivering facility with a moving elevator. The main goals include maintaining the air quality of the facility and understanding the influence on the elevator moving pattern, elevator speed, and FFU discharge air velocity. Also, special emphases are focused on identifying the vortex location and airflow uniformity in the delivering equipment via the numerical flow visualization.
Finally, an appropriate parameter setting can be found through the above-mentioned parametric study to ensure the yield rate of LED substrates.
II. PHYSICAL MODEL DESCRIPTION
To save delivering time and prevent panel damage during moving the glass substrate, this study employs a vertical laminar-flow facility to transport the LCD glass substrate into storage cabinets or next station. This vertical laminar-flow delivering facility is built based on standard of mini-environment, and is equipped with six FFUs installed on the top ceiling (see Fig.1) to serve as air source for driving the internal flow. The incoming airstream firstly hits the moving elevator and LCD panel; then partial flow is reflected and expelled to ambient atmosphere through the top exhauster while the rest of airflow passes through the channel between delivering facility and elevator. This bypass airstream divides into several small streams, flowing between glass substrates and generating vortex inside the moving elevator, and finally exhausts from the Recent Researches in Mechanics
ISBN: 978-1-61804-020-6 82
lower and bottom ventilating holes.
It is worthy to note that transporting glass substrate is accomplished by moving elevator upward and downward, which is the dominant factor for casting the internal flow pattern. The other driving force is the incoming air current which is generated by FFU and expelled through the ventilation openings at the top, lower, and bottom locations. The interaction between the aforementioned phenomena results in an unsteady and complicate flow field, which needs sophisticate CFD software to simulate and analyze. The proper assumptions and boundary conditions is set according to the practical conditions and described in the following subsections.
A. Assumptions
To capture the actual physical phenomena, several appropriate assumptions and boundary conditions are made to simulate the flow field. The following assumptions are enforced in order to simplify the flow complexity
(1) Incompressible flow is assumed because the fluid velocity is quite low for an internal flow;
(2) The flow is treated as Newtonian fluid, the density is constant, and viscosity is isotropic;
(3) Dynamic mesh is applied to simulate the moving patterns of elevator;
(4) the standard k- ε turbulent model is adopted for turbulence calculation;
(5) The influences of radiation heat and floatation terms are neglected;
(6) Spherical particles are assumed.
B. Boundary Conditions
Several boundary conditions are used in this work and are described as follows:
(1) Velocity inlet boundary condition
The vertical laminar-flow delivering facility is a typical mini-environment, which is usually driven out the polluting particles by the airflow. So as to control airflow velocity easily, FFU system is installed on the ceiling to ensure a uniform airflow direction.
(2) Pressure outlet boundary condition
In order to maintain the positive pressure in the clean room, the pressure is set to be 25Pa. Positive
pressure represents to control the pressure difference in two spaces.
(3) Dynamic Mesh method
Instead of using relative velocity, this study utilizes dynamic mesh method to simulate moving elevator for an accurate calculation.
III. NUMERICAL SIMULATION
As stated in last section, the physical phenomena caused by the interaction of moving elevator and FFU discharge airflow is very complex and unsteady. Thus, the numerical tool is adopted to simulate and investigate this complicate flow field. This section describes several important numerical models and judging terminologies used in this work.
A. Numerical method and turbulent model
In the numerical algorithms, the second upwind turbulence dissipation rate (ε).
B. Dynamic mesh and moving patter of elevator
Dynamic mesh is known for its capability that could automatically change the grid number and shape for fitting with the moving rigid body. Therefore, this investigation employs the dynamic-mesh technology to examine the flow field inside the LCD delivering facility. Also, the moving patterns of elevator are planned as：
Pattern-1:
The elevator moves downward from top to bottom exports at a constant speed, and stays 5 seconds after reaching the lowest position.
Pattern-2:
The elevator moves upward at a constant speed, and stays 5 seconds after reaching the top position.
There are two moving patterns together with three moving speeds 0.08, 0.12, and 0.16 m/s considered in this Recent Researches in Mechanics
ISBN: 978-1-61804-020-6 83
investigation. In accordance with the different of moving patterns and speeds observed airflow patterns to obtain the best moving pattern and speed.
C. Index of airflow uniformity (λ)
It is essential to maintain high uniformity airflow in the delivering facili1ty with class-1 level. Also, the indication of the uniformity of airflow is defined as
(1 ) 100%
v
λ
= −σ
× (1) where σ is the standard deviation of the velocity distribution, and V is the average velocity of the cross-section area. Clearly, a higher λ value represents that airflow uniformity is better and flow pattern becomes smoother in the delivering facility.A comprehensive simulation program is arranged to execute the systematic parametric investigation on the airflow uniformity and ability to exclude particles in the delivering facility. These parameters include FFU air velocity, moving pattern and velocity of elevator. Three FFU outlet velocity (0.54, 0.63, 0.71 m/s) and elevator speed (0.08, 0.12, 0.16 m/s) are designated for consideration in this work. Thereafter, the outcome can provide an important reference for building a panel plant.
Also, an appropriate combination of the above parameters can be reached for keeping uniformity flow in the delivering facility.
IV. NUMERICAL RESULTS AND DISCUSSIONS The cleanliness level of this clean room is set to be class 1 and the ceiling-installed FFU system has an outlet area of 6.5*12.5 m2. CFD simulation indicates that, as illustrated in Fig. 2, FFU discharge air flows in the delivering facility and splits into two streams after directly striking the top of elevator. One stream is reflected and expelled via the top exhausting holes, and the remaining airflow successfully enters the enclosing channel between wall and elevator. This passing-through airflow is the driving force to eject the pollutant form the thin layer between LCD boards, and also is the reason to generate unfavorable vortex and recirculation. The distinctions on
flow phenomena induced via different parameter settings are described and discussed in details under various elevator moving patterns.
Firstly, the FFU air velocity and the elevator moving speed are set at 0.63 m/s and 0.16 m/s for comparison under the same criterion. For moving pattern-1, as indicated in Fig. 3, the airflow above the elevator is induced to shift downward with a descending elevator at the beginning (t=1 sec). Also, since most of airflows are blocked by elevator, thus the flow beneath the elevator is quiet smooth except slight recirculation occurs near the bottom exhausting openings. At t=3s, the airflow around the elevator flows smoothly when the elevator locates at the middle point of stroke. Finally, the elevator reaches the lowest position (t=6s), the flow field near the elevator becomes more complex due to the strong interaction between the downward elevator and ventilating holes at the bottom.
However, as shown in Table 1, the airflow uniformity recovers to 84.79% quickly after a short stop (5 seconds), which is needed for loading/unloading LCD panels.
As regards the upward pattern-2, the rising elevator compresses the air above it, therefore a low-velocity zone is formed immediately on the top of elevator (see Fig. 4). compressed and induced airflows above and beneath elevator speed up to maximum strength and quickly expel out through the top ventilating openings. Similar to the patter-1, after a 5-second stop, the airflow uniformity restores to 97.43% (see Table I).
Consequently, after carefully comparing above results, it is obvious that the airflow uniformity and flow field in the moving pattern-2 is superior to that of pattern-1;
nevertheless, recirculation phenomenon is identified around the corner of delivering facility in both patterns. This conclusion can serve as an essential reference for arranging the working procedures and layouts of LCD panel factory.
Recent Researches in Mechanics
ISBN: 978-1-61804-020-6 84
V. RESULTS AND DISCUSSIONS
In this work, a comprehensive CFD investigation incorporated with dynamic mesh is carried out for the parametric study on the airflow uniformity and ability to exclude particles in the delivering facility. As a result, moving pattern of the elevator has an obvious affect on the flow field and airflow uniformity. Also, occurrences of vortex and recirculation influence the general trend of flow field and downgrade the airflow uniformity. Based on the above results and discussions, the following conclusions are obtained.
(1) Regardless of any moving speed of elevator, better airflow uniformity is presented in the delivering facility when the elevator is moving with pattern-2.
(2) Enhancing the discharge velocity of FFU and reducing the speed of moving elevator illustrate insignificant impact on airflow uniformity in the delivering facility.
(3) Since decreasing the elevator speed requires extra time and cost to transport LCD panel, thus it is helpful to increase the elevator speed and FFU air velocity as long as no serious airflow uniformity appears.
In conclusions, an accurate numerical model together with dynamic mesh is established and used for understanding the flow field and air quality inside the delivering facility at different elevator speeds and FFU airflows. Consequently, an appropriate parameter combination is obtained for keeping uniformity flow in the delivering facility. Also, this outcome can serve as an important reference for transporting the LCD panel.
REFERENCE
[1] S. C. Hu, Y. Y. Wu, and C. J. Liu, “Measurements of Air Flow Characteristics in a Full-Scale Clean Room”, Building and Environment, Vol. 31, No. 2, pp. 119-128, 1996.
[2] M. Cheng, G. R. Liu, K. Y. Lam, W. J. Cai, and E. L.
Lee, ”Approaches for Improving Airflow Uniformity in Unidirectional Flow Cleanrooms”, Building and Environment, Vol.
34, pp. 275-284, 1999.
[3] A. Mistriotis Giannoulis and D. Briassoulis, “Experimental and Numerical Investigation of the Airflow around a Raised Permeable Panel”, Journal of Wind Engineering and Industrial Aerodynamics, pp. 808-817, 2010.
[4] R. Lambert, C. L. Lin, E. Mardorf, and P. O’Shaughnessy, ”CFD Simulation of Contaminant Decay for High Reynolds Flow in a Controlled Environment”, Behalf of the British Occupational Hygiene Society, Vol. 54, No. 1, pp. 88-99, 2010.
[5] T. Kobayashi, M. Sandberg, H. Kotani, and L. Claesson,
“Experimental Investigation and CFD Analysis of Cross-Ventilated Flow through Single Room Detached House Model”, Building and Environment, Vol. 45, pp. 2723-2734, 2010.
[6] M. H. Saidi, B. Sajadi, and G. R. Molaeimanesh, ”The Effect of Source Motion on Contaminant Distribution in the Cleanrooms”, Energy and Buildings, Vol. 43, pp. 966-970, 2011.
[7] Fluent 6.3 documentation. FLUENT INC.
[8] B. E. Launder and D. B. Spalding, Lectures in Mathematical Models of Turbulence. Academic Press, London, England, 1972.
[8] B. E. Launder and D. B. Spalding, Lectures in Mathematical Models of Turbulence. Academic Press, London, England, 1972.