本研究以探討利用一個可在現場測量煙洩漏量的試驗儀器，其具備控制門內壓力，並可測量風速、壓力、溫度，在不破壞門組件及不妨礙門扇正常開關閉的情況下，進行測試了解既有建築用門的煙洩漏情形。本研究所得結果如下：
既有建築用門關閉狀態下洩漏量情形
在台灣科技大學研揚大樓現場門(以下簡稱TR)及實驗室用門(以下簡稱LAB)測試下，結果顯示在50 Pa的艙體壓差時，LAB洩漏量為592m^3/h，TR洩漏量為455m^3/h，而在25Pa的艙體壓差時，LAB洩漏量為417.7m^3/h，TR洩漏量為300m^3/h，而兩者在關閉時洩漏量遠大於CNS15038規定的25m^3/h。
二、既有建築用門在不同開關度下洩漏量情形
試驗以每5mm為一級距，進行建築用門在不同開關度洩漏量測試，結果顯示，受限於軸流風機功率能力，TR及LAB用門開啟後均無法達到50 Pa的艙體壓差，而開口寬度在10mm 以上TR與LAB均無法維持25 Pa的艙體壓差。維持艙體壓差在10 Pa狀態下，開口寬度在5mm至40mm之間洩漏量結果維：
TR部分：關閉裝態下時洩漏量為268.8m^3/h，5mm開口洩漏量為320m^3/h至40mm開口洩漏量為592m^3/h。
LAB部分；關閉裝態下時洩漏量為278m^3/h，5mm開口洩漏量為348m^3/h至40mm開口洩漏量為522m^3/h。
三、既有門各縫隙洩漏量分析
LAB用門為一般木門，門框與門之間無任何氣密條，門下方縫隙佔整體門洩漏量80.93%，而TR用門下方縫隙洩漏量佔整體門洩漏量16.23%，表示安裝氣密條在門四邊的洩漏量上能達到較良好止洩漏效果。

This study aims at getting to have a proper idea of how well two subject structures for existing buildings doors, when kept intact and left in its normal position, perform their function of smoke-resistance. Various tests are carried out using on-site instruments to pick up data on smoke leakage as well as related data on wind velocities and temperatures. The following findings are what all these tests present:

A. Amounts of leakage picked up when the built-in doors are kept tightly closed:
Based on the results of tests on the door of AAEON Building, National Taiwan University of Science and Technology (TR for abbreviation) and that of Laboratory (LAB for short), when the in-building pressure is controlled at 50 Pa, the leakage for TR door is 455m3/h and the leakage for LAB door is 592m3/h. when the in-building pressure is controlled at 25 Pa, the leakage for TR door is 300m3/h and the leakage for LAB door is 417.7m3/h. Both far exceed the maximum leakage allowance of 25m3/h required by CNS15038.

B. Amounts of leakage picked up when the built-in doors are kept partially closed:
Varied results come up when tests are done with the built-in door closed at different gaping widths (at an adding-up rate of 5mm for each new test). As confined to the working capacity of the axial flow fan employed in the tests, neither the door of TR nor that of LAB is able to reach the required 50 Pa. in terms of differential cabin pressure when both are kept totally agape. Neither is able to reach the requirement of 25 Pa in terms of differential cabin pressure when the gaping width is controlled at above 10mm, nor do they maintain 10 Pa in terms of differential cabin pressure when left agape at a width of 40mm. The related data collected at various gaping widths are listed as follows:

In the case of TR door:
Leakage readings ranging from 320m3/h at the width of 5mm up to 592m3/h at the width of 40mm are obtained. Doors are kept tightly closed the leakage is
268.8m3/h

In the case of LAB door:
Leakage readings ranging from 348m3/h at the width of 5mm up to 552m3/h at the width of 40mm are picked up. Doors are kept tightly closed the leakage is
278.8m3/h

C. Analysis of varied leakage amounts from cracks between the door and its frame:
The material used for the LAB door is wood and there is no smoke-resistant strip positioned between the door and its frame. Leakage measured at the bottom crack accounts for 80.93% of the total amount of escaped smoke. By contrast, leakage measured at the bottom crack of TR door accounts for only 16.23%, which is proof that smoke-resistant strips positioned between the door and its frame can be effective stoppage to smoke leakage.

[1] 趙鋼，林金宏，「火場求生新觀念」，商訊文化，2011，p.25
[2] 林金宏，「消防月刊」，2003，p.4
[3] F.B. Clarke, (1997). Physiological effects of smoke: Managing escape, ASHRAE Journal, 39(3), pp.47-56.
[4] CNS15038建築用門遮煙性試驗法，民國99年11月。
[5] 欒中丕，「煙洩漏量試驗儀器之研發與應用」，國立台灣科技大學建築系博士論文，民國98年1月。
[6] 鄭孟昌，「建築物開口部構件常溫遮煙性實驗方法之研究」，國立台灣科技大學建築系碩士論文，民國95年6月。
[7]U. Kazuki, (1997). Fire testing of window glass for integrity and smoke leakage, Research Report, Architecture Institute of Japan (AIJ), Kanto branch, Paper No. 8.
[8] J.H. Klote, (1995). SMOKE CONTROL, SFPE Handbook of Fire Protection
Engineering, Chapter 12 Section 4, pp.230~245.
[9]J. Rakic, (2000). The Performance of Unit Entry Doors When Exposed to Simulated Sprinkler Controlled Fires,pp.1-13.
[10] D. M. Brani and W. Z. Black, (1992). Two-Zone Model for a Single-Room Fire, Fire Safety, pp.189-216.
[11] EE. Zukoski, (1995). Properties of fire plumes, In: Cox, G. (ed.),
Combustion Fundamentals of Fire, Academic Press, London, Chapter 3, pp. 101-219.
[12]G.K. Yuill , K.H. Haddad. Effect of Opening Stairwell Doors on Performance of a Stairshaft Pressurization System. ASHRAE Transaction (1994).100(1):990-999
[13] 田中哮義，「建築火災安全工學入門」，平成14年改訂版。
[14] 邱聖幃，「建築物開口部設備遮煙試驗裝置之開發」，國立台灣科技大學建築系碩士論文，民國96年1月。
[15] 莊英宏，「建築鋼製門之遮煙性能研究」，國立台灣科技大學建築系碩士論文，民國97年6月。
[16]Fang Liu, Lan Gong and Zhi-Jun Yan. (2000). Opening Flow Coefficient and Its Effect on Fire Smoke Flow, Journal of Chongqing Jianzhu University, pp.86-92.
[17] D. Etheridge and M. Sandberg (1996). Building Ventilation:
Theory and Measurement: 53-54, John Wiley and Sons, U.K.
[18]D. Gross, (1991). Estimating Air Leakage Through Doors for Smoke Control, Fire Safety Journal NO.17, pp.171-171