中華民國內政部建築研究所中華民國內政部建築研究所

一、研究緣起
鋼構造由於強度高、韌性佳、耐震性能優良且施工期短，近年來已逐漸成為國內建築業界興建大型集合式住宅或商業辦公大樓的主流，然而，高溫下鋼材有強度折減與軟化的現象，使得火害對鋼構造建築的危害性較為嚴重，民國90年的汐止東方科學園區鋼構大樓火災，其延燒範圍廣且延燒時間長，造成該鋼構大樓內部之重要構件(如：鋼柱、鋼梁、樓板…等)受火害高溫而損壞嚴重，為歷年所僅見，顯示有必要針對鋼構造建築的火害影響進行研究。鋼構建築物的柱構件受火害影響可分為內柱（四面受火）、邊柱（三面受火）及角柱（二面受火），由於角柱端點僅與兩支鋼梁連接，其鋼梁連接方式較內柱(四支鋼梁連接)與邊柱(三支鋼梁連接)不對稱，因而造成角柱端點彎矩束制較弱，使得柱的有效長度較長，致挫屈強度較小，且因鋼梁連接方式的不對稱，造成角柱受到較大的彎矩，加以角柱在火場中曝火面為二側，使得溫度分佈不對稱，角柱在火害下的結構行為更加複雜，以往研究大多針對內柱火害行為，鮮少對於角柱的火害行為進行研究，由於鋼柱屬於鋼構建築的重要構件，若受到火害高溫侵襲，嚴重情況將造成鋼柱挫屈破壞，進而造成鋼構建築之嚴重倒塌。目前我國現行對於建築物之鋼柱構件的耐火能力檢驗，主要依據CNS 12514-1「建築物構造構件耐火試驗法－第1部:一般要求事項」及CNS 12514-7「建築物構造構件耐火試驗法－第7部：柱特定要求」，以單一構件的鋼柱進行耐火試驗，然而，此種方法未考慮真實鋼構造建築受到火害時，其鋼柱受到其他構件之束制影響。此外，角柱在火害中受到鋼梁膨脹外推變形之效應，所衍生的二次彎矩之影響，亦無法在標準試驗中呈現，因此透過實尺寸鋼構屋的火害實驗，始能夠探討角柱於火害下的真實行為。
二、研究方法及過程
本計畫的主題是針對貴所在台南歸仁戶外實驗場中既有的實尺寸鋼構實驗屋內的「H型鋼角柱」來進行火害相關的研究，因此，本研究團隊將在實尺寸鋼構實驗屋內包含「H型鋼角柱」的部分區劃空間，進行火災實驗，藉以研究此區劃空間內「H型鋼角柱」於真實的結構束制情況下受真實火害之行為，此外，本計畫亦將針對此次所進行的實尺寸鋼構屋角柱之現地火害實驗進行數值模擬與分析，藉以瞭解實驗量測所無法獲得的角柱在火害下的結構行為，以及整體鋼構屋在角柱受到火害高溫的結構行為和影響，本計畫亦將進行實驗研究資料庫之建置。
三、重要發現 
(1)本次鋼構實驗屋的火害實驗在總火載量達43,488 MJ的18組木框架燃燒下(火載量密度為604 MJ/m2)，在火害實驗區內的鋼柱、鋼梁與樓板之變位皆未達我國CNS 12514-1第10.2.1節承重能力之(a)項及(b)項所規定之破壞判定值。
(2)經由初步量測所得之火害後西北角柱0.9H處之斷面位置圖推測，西北角柱受到柱頂兩個方向相接之兩支大梁火害後之變形與收縮，造成柱頂有側移和扭轉(由上俯視為逆時針扭轉)之現象，使其偏離原設計位置；火害後角柱頂端若偏離原設計位置，此角柱將會因P-效應而在火害後受到多餘的二次彎矩作用。
(3)由數值模擬結果發現，模型(一)之西北角柱在降溫後(200C)之軸力載重比為9.69%，為常溫時軸力載重比之7.13倍，降溫後(200C)之軸力載重比增加原因為火害實驗中所有鋼柱隨著溫度變化而發生力量重新分配現象，造成降溫後(200C)西北角柱之軸力增加，但其軸力載重比仍不大，故未發生柱體挫屈現象。
四、主要建議事項
建議一 
火害後鋼柱構件之耐震性能研究：立即可行建議
主辦機關：內政部建築研究所
協辦機關：國立成功大學、國家地震工程研究中心、中國鋼鐵結構股份有限公司
鋼柱是鋼構建築中最重要之構件之一，其對整體鋼構建築之耐震性能影響甚鉅，當鋼柱受到高溫火害的侵襲，其材質或銲道可能受到高溫影響，造成其強度與延展性發生變化，使得鋼柱的耐震性能下降，吾人可透過未受火害與受火害鋼柱構件的側向反覆載重實驗，對火害後鋼柱構件之耐震性能進行評估。
建議二
火害後鋼構屋之結構耐震性能評估：中長期建議
主辦機關：內政部建築研究所
協辦機關：國立成功大學、國家地震工程研究中心、中國鋼鐵結構股份有限公司
鋼構建築受到火災侵襲後，其構件與梁柱接頭之銲道、螺栓，常因火場的高溫與冷卻方式造成材質變化，造成構件與接頭之強度下降，使得鋼構建築整體耐震能力下降，吾人可透過未受火害與受火害之鋼構屋的大型振動台實驗，藉以研究火害後鋼構屋之結構耐震性能。

Due to high strength, good ductility, superior seismic resistance and short construction duration, steel structures have become the main stream of residential complexes and commercial office buildings in Taiwan. However, strength reduction and softening make fire more hazardous for steel buildings. In 2001, the wide and long fire spreading and high temperatures in the steel building of East Science Park in Si-Jhih City of Taipei County caused serious damage in the important structural members, such as steel columns, steel beams, floors, etc. This rarely-seen fire shows that it is necessary to proceed studies on steel framed buildings in fire. The columns subjected to fire in steel buildings can be divided to interior columns (fire on four sides), edge columns (fire on three sides) and corner columns (fire on two adjacent sides). A corner column end connects only two beams, and this kind of connecting detail is more asymmetric than interior columns (connecting four beams in one end) and edge columns (connecting three beams in one end). As a result, the rotational restraint of corner column is weaker, which causes the longer effective length of the corner column, so the buckling strength of the corner column is smaller. Due to the asymmetric beam connecting detail, corner columns are subjected to larger moments. In addition, corner columns have two sides exposed to high temperatures in a fire compartment, and this asymmertric temperature distribution makes the more complicated structural behaviors of corner columns in fire. The past studies mostly focused on the structural behaviors of interior columns in fire, and very few studies were done for the structural behaviors of corner columns in fire. Steel columns are the important components in steel buildings. If the steel columns are fire damaged and failed by buckling, the steel building might collapse seriously. At present in Taiwan, the fire-resistance test for a steel column mainly follows CNS 12514-1 and CNS 12514-7, which requires to test a single steel column in a standard furnace. However, this kind of test method does not consider the restraint forces between steel columns and other components in a real steel building in fire. Besides, corner columns might be pushed by the expanded beams in fire and this might cause second-order moment. This kind of effect cannot be presented in a standard fire test. Therefore, through the real fire test of a steel experimental house, we can study the real structural behavior of steel H-shaped corner column.
The important findings of this research project are as follows:
(1)Two zones (B-zone & D-zone) in the experimental building were burned by the fire load of 43,488 MJ using eighteen wood cribs (i.e. fire load density = 604 MJ/m2). The downward displacements of the steel columns, beams and slabs in the two zones did not exceed the failure displacement limits and displacement rates determined by the our country’s CNS 12514-1.
(2)Based on the cross-sectional location of the northwest corner column at 0.9H after the preliminary measurement of the fire damage, it is inferred that the northwest corner column was influenced by deformation and contraction of the two beams that were connected in two directions at the top of the column. This influence caused the column top to move sideways and to twist counterclockwise from the top view. As a result, the corner column top deviated from the original design position. If the top of the corner column deviated from the original design position after fire damage, this corner column would be subjected to the redundant secondary bending moment after fire damage due to the P- effect.
(3)From the numerical simulation results, it was found that the axial-force load ratio of the northwest corner column of Model (1) after cooling (i.e. at 200C) was 9.69%, which was 7.13 times the axial-force load ratio at room temperature. The reason for the increase in the axial-force load ratio after cooling (i.e. at 200C) was due to the force redistribution of all steel columns with temperature variations. This caused the axial force of the northwest corner column to increase after cooling, but its axial-force load ratio was still not large. Therefore, no column buckling occurred in Model (1).
This project proposes two immediate strategies.
For the first immediate strategy:
Steel columns are one of the most important components in steel structures, and they have a great impact on the seismic performance of overall steel structures. The change in the strength and ductility would reduce the seismic performance of steel columns. We can evaluate the seismic performance of post-fire steel column members through lateral cyclic loading tests of unfired and fired steel column members
For the second immediate strategy:
After a steel structure is damaged by a fire, the weld passes and bolts of its components and beam-to-column connections often cause material changes due to the high-temperature and cooling method of the fire site, which causes the strength of the components and connections to decrease. As a result, it reduces the overall seismic resistance of the steel structure. We can study the seismic performance of a post-fire steel structure through the large shaking table experiments of the steel structures with and without fire damage.