JP4187064B2 - Method for calculating replacement load in underpinning and leveling method for existing structures - Google Patents

Description

ここに、ｙ _Ｉｉｊ ：建物各点の変位
ｂ _ｉ ^Ｉｉｊ ：重回帰式の偏回帰係数
Ｘ _Ｉｉｊ ：建物各点の受替荷重の予測値
Ｔ p ：雰囲気温度
ｂ _０ ^Ｉｉｊ ：重回帰式の定数項
【００１５】
アンダーピニングでは、構造物本体の変形特性、周辺地盤の変形特性、構造物の真の荷重などの不明確な要因の影響が顕著であるため、その挙動を精度良く把握することが困難であったが、上記請求項１記載の本発明においては、既設構造物の各点の変位、受替荷重、および雰囲気温度の実測値を変数として重回帰式を求める統計的手法により任意の変位を与える受替荷重の条件を算定するようにした。従って、前記重回帰式は建物の変形特性や周辺地盤の変形特性などの影響を既に含んでいるため、それらの個々のメカニズムは明らかにできないが、高い精度で受替荷重を算定することが可能となる。
【００１６】
請求項２に係る本発明として、杭基礎を備える既設構造物の一部又は全部を、地下に新たに設けた杭基礎及び受替版からなる受替構造体によって受替えるアンダーピニングにおいて前記受替構造体による受替を完了し埋戻し復旧を行った状態にて前記既設構造物をレベリング状態とするためのレベリング方法であって、
前記埋戻しに先だって、少なくとも埋戻しによる荷重増分を考慮して受替版の各点沈下量を求め、この受替版の各点沈下量の逆符号値を埋戻し前における既設構造物の変位として設定するとともに、この変位を与条件として請求項１記載のアンダーピニングにおける受替荷重の算定方法により受替荷重を算定し、この算定された受替荷重に基づいて既設構造物の受替荷重を調整した後、埋戻しを行うことを特徴とするアンダーピニングにおける既設構造物のレベリング方法が提供される。
【００１７】
上記請求項２記載の本発明は、具体的に埋戻し後に建物をレベリング状態とするための手順に係るものである。すなわち、建物の将来変位（埋戻し以降に生じる変位）に影響する要因としては、例えば(1)流動化処理土の埋戻しによる荷重増分、(2)埋戻しにより生じる受替版下および建物下の地盤反力、(3)シールドの掘進による周辺地山の緩みなどが挙げられる。また、施工条件によっては上載荷重の影響や近い将来に想定される建設物の影響などについても考慮する方が望ましいと思われる。これらの要因の中で少なくとも埋戻しによる荷重増分を考慮して受替版の各点沈下量を求め、埋戻し後の建物変位が０になるように、前記受替版の各点沈下量の逆符号値を埋戻し前の既設構造物の変位として設定し、この変位を与条件として前記重回帰式を用いた受替荷重の算定方法により受替荷重を算定し、この算定された受替荷重に基づいて既設構造物の受替荷重を調整した後、埋戻しを行うようにする。
【００１８】
請求項３に係る本発明として、前記受替版の各点沈下量の算定は、受替版をはり要素でモデル化するとともに、新設杭の杭頭ばねで支持した構造モデルとし、荷重は載荷点から各はりの軸線までの距離に応じた受替荷重を作用させることに行う請求項２記載のアンダーピニングにおける既設構造物のレベリング方法が提供される。
【００１９】
請求項４に係る本発明として、前記受替版を支持する新設杭の杭頭ばねは、受替版の沈下量の実測値と、前記構造モデルによる解析値が適合するように逆解析を行った結果から決定する請求項３記載のアンダーピニングにおける既設構造物のレベリング方法が提供される。
【００２０】
請求項５に係る本発明として、算定された受替荷重に基づいて既設構造物の受替荷重を調整するに当たり、既設構造物の受替荷重の変動により受替版各点の沈下量が変化することを考慮する場合は、前記構造モデルにより受替版の各点変位が所定範囲に収束するまで繰り返し計算を行うようにする請求項３、４いずれかに記載のアンダーピニングにおける既設構造物のレベリング方法が提供される。
【００２１】
【発明の実施の形態】
以下、本発明の実施の形態について図面を参照しながら詳述する。
【００２２】
〔計測データに基づく重回帰式の誘導〕
建物の不等変位を除去できる最適な受替荷重の予測には、実測された建物の変位、受替荷重、および雰囲気温度の３つの変数から求めた重回帰式を用いる。以下では、計測データに基づいた重回帰式を算定する具体的な手順を述べる．
(1)工事概要と計測概要
対象とした工事は、臨海副都心線二期工事の第１広町トンネル工区におけるアンダーピニングである。大井町駅と大崎駅との間に位置する第１広町トンネルは、シールド工法により建設される外径7,100mmの鉄道単線断面トンネルである。この路線上には、鉄筋コンクリート造１２階建てのJR広町社宅２棟と、鉄筋コンクリート造８階建ての品川区防災センターの計３棟の建物が位置し、それらの建物の基礎杭の一部がトンネル掘削の支障となるため下受け形式によるアンダーピニングが実施された。解析の対象は、アンダーピニングされる３棟の建物の中で最も多い点数で支持されたJR社宅６号棟とした（図１及び図２参照）。JR社宅６号棟は建物総重量167MNで、このうち62.3MNを受替版を介した１３本の新設杭により受替える工事である。
【００２３】
受替荷重の算定には、実測された建物の変位、受替荷重、および雰囲気温度の３種類の変数から求めた重回帰式を用いた。受替制御は応力導入から既設杭の撤去が完了するまでの比較的長期にわたる。このことから、建物各点の変位挙動は周辺部の温度変化の影響を受けると考え、説明変数に雰囲気温度を選定することにした。この重回帰式は、建物本体の変形特性や周辺地盤の変形特性などの不確かな要因の影響をすでに含んでいるため、それらのメカニズムは明らかにできないが、高い精度で受替荷重を算定することが可能となる。
【００２４】
図３は、重回帰式の算定に用いた受替荷重、雰囲気温度、および建物の変位の実測値を示したものである。これらの計測データは、応力導入直後から既設杭の杭頭の切断が完了するまでの期間に計測したものであり、その総数は約5,000である。これらの図からも同一箇所の受替荷重と建物の変位とが、単純な相関関係にないことが判明している。
【００２５】
〔重回帰式の誘導〕
目的変数ｙ_ｉを建物ｉ点の変位、説明変数ｘ_ｉ、Tをそれぞれ建物ｉ点の受替荷重と雰囲気温度とすれば、建物各点の変位は次式により表される。なお、受替荷重の算定が目的であるが、制御するのは受替荷重であるため建物各点の変位を目的変数とすることにした。
【００２６】
【数１】

ここに、ｙ_ｉ：建物ｉ点の変位の実測値
ｘ_ｉ：建物ｉ点の受替荷重の実測値
ｂ_１、ｂ_２、…ｂ_i、ｂ_ｔ：重回帰式の偏回帰係数（ここではi=14とした）
Ｔ：雰囲気温度の実測値
ｂ_０：重回帰式の定数項
【００２７】
上式(1)を満たす偏回帰係数と定数項を決定すれば、目標とする建物各点の変位を代入しｘ_ｉについて解くことで受替荷重を求めることができる。偏回帰係数は次式により算定した。
【００２８】
【数２】

ここに、Ｓ^２ _ｘｉ：変数ｘ_ｉの分散
Ｓ_ｘｉｙｉ：変数ｘ_ｉと変数ｙ_ｉの共分散
【００２９】
また、重回帰式の定数項は次式(3)で表される。
【００３０】
【数３】

ここに、[ｙ_ｉ]：変数ｙ_ｉの平均
[ｘ_ｉ]：変数ｘ_ｉの平均
なお、上記[]表記は便宜上、平均値を意味し、文字上側の上線表示の代替えである（以下、同様）。
【００３１】
重回帰式はそれが示す意味の明確化などの観点から、極力少ない変数で誤差の小さい目的変数を与えるのがよく、精度の高い重回帰式が得られても、むやみに変数が多くなると計算が繁雑になり実用的でない。このため、重回帰式の算定においては、次に述べる方法によりその説明変数を選定することにした。
【００３２】
一般に、説明変数間が高い相関を有するときは、偏回帰係数が求められず、偏回帰係数の符号と目的変数との単相関係数の符号が一致しないなどの多重共線性が生じることが知られている。このため、下式(4)により説明変数間の相関性を判定し、相関が認められる時は一方の変数を除外して重回帰式を作成することにした。ここでは、その判定の基準を0.9以上とした。
【００３３】
【数４】

ここに、Ｒ_ＸＸ：説明変数間の単相関係数
Ｓ_ＸｉＸｊ：変数ｘ_ｉと変数ｘ_ｊの共分散
Ｓ_Ｘｉ：変数ｘ_ｉの標準偏差
Ｓ_Ｘｊ：変数ｘ_ｊの標準偏差
次に、下式(5)により目的変数と説明変数との相関性を判定し、目的変数に与える影響が極端に小さいものを除外する。ここでは判定の基準を0.1未満とした。
【００３４】
【数５】

ここに、Ｒ_Ｘｙ：説明変数と目的変数の単相関係数
Ｓ_Ｘｉｙｉ：説明変数ｘ_ｉと目的変数ｙ_ｉの共分散
Ｓ_Ｘｉ：変数ｘ_ｉの標準偏差
Ｓ_ｙｉ：変数ｙ_ｉの標準偏差
これらの手順から、本工事の場合は、受替荷重に関する１４元の連立方程式が得られる。連立方程式を行列式で表せば下式(6)となる。
【００３５】
【数６】

ここに、ｙ _Ｉｉｊ：目標とする建物各点の変位
ｂ _ｉ ^Ｉｉｊ：重回帰式の偏回帰係数
Ｘ_Ｉｉｊ：建物各点の受替荷重の予測値
Ｔp：雰囲気温度の予測値
ｂ _０ ^Ｉｉｊ：重回帰式の定数項
上式(6)において、右辺第１項の偏回帰係数を表す行列は14行15列である。雰囲気温度Ｔは受替荷重から独立した変数であるため、偏回帰係数を表す行列は14行14列の正方行列に変換することができる。したがって、上式(6)を下式(7)のように変換すれば、目標とする建物各点の変位を代入することで受替荷重が算定できる重回帰式が得られる。
【００３６】
【数７】

ここに、ｙ _Ｉｉｊ ：建物各点の変位
ｂ _ｉ ^Ｉｉｊ ：重回帰式の偏回帰係数
Ｘ _Ｉｉｊ ：建物各点の受替荷重の予測値
Ｔ p ：雰囲気温度
ｂ _０ ^Ｉｉｊ ：重回帰式の定数項
【００３７】
〔重回帰式を用いた変位予測手法の妥当性の検証〕
本節では現場計測結果と重回帰式から算定した受替荷重の予測値とを比較し、重回帰式を用いる施工段階の変位予測手法の妥当性を検証する。以下では、実測値と予測値との比較検討に先立ち、求めた重回帰式の信頼性について考察を加える。
【００３８】
(1)重回帰式の検定
【００３９】
【表１】

【００４０】
上表１は、前節で述べた方法により求めた重回帰式の自由度を調整した決定係数と、予測値と残差との分散比を示したものである。なお、求められた重回帰式は、次式により求めた分散比Ｆが自由度q、n-q-lのＦ分布（ここに、n：標本数、q：説明変数の数）に従うものとして右片側検定を行い、その信頼性を確認した。
【００４１】
【数８】

【００４２】
【数９】

【００４３】
【数１０】

ここに、Ｆ：分散比
Ｖ_Ｒ：予測値の不偏分散
Ｖ_Ｅ：残差の不偏分散
ｑ：説明変数の数
ｎ：サンプルの数
すなわち、Ｆが有意水準より大きいならば、重回帰式が予測に役立たないという仮説が棄却され、重回帰式は予測に役立つとする考えである。
【００４４】
表１をみると、建物各点の変位の実測値は、重回帰式から求めたその予測値との相関が高く、また分散比はＦ_{ｑ、ｎ−ｑ−ｌ}(５％)（本事例の場合は最大で1.940）と比較して大きいことがわかる。したがって、これらの重回帰式は十分な信頼性を有するものと判断され、目標とする建物各点の変位を与えれば、それを満たす建物各点の受替荷重を十分な精度で求められることを示している。
【００４５】
(2)偏回帰係数による受替荷重と構造物の変位の特性
説明変数の受替荷重と雰囲気温度では単位が異なる。求められた重回帰式においてどの変数の影響がより大きいかを考察するために、標準化した場合の偏回帰係数を求めることにした。計測データの標準化は次式により行う。
【００４６】
【数１１】

ここに、ｚ_ｉ：変数ｘ_ｉの標準化されたデータ
[ｘ_ｉ]：変数ｘ_ｉの平均
Ｓ_Ｘｉ：変数ｘ_ｉの標準偏差
【００４７】
【数１２】

【００４８】
上式(12)は計測データを標準化した場合の偏回帰係数を示したものである。本来、偏回帰係数は近似値であるため、建物各点の受替荷重がそれらの変位に与える影響を厳密に判断するのは困難である。しかしながら本事例の場合、説明変数間の相関係数が最大でも0.5程度であることから、説明変数はほぼ独立しているものと見なして、偏回帰係数の観点から荷重制御と建物の挙動特性に関する考察を試みる。
【００４９】
図４は、標準偏回帰係数の絶対値の平均|[ｂ]|、平均[ｂ]、および分散Ｓ^２ _ｂを各々の受替位置について表したものである。この図をみると、B-10点、C-10点、C-9点、C-7点、E-6点、およびTの各変数は、標準偏回帰係数の絶対値の平均|[ｂ]|が比較的大きく、一方、その分散Ｓ^２ _ｂが比較的小さいことがわかる。このことから、これらの各変数は建物全体の変位に与える影響が大きく、本工事における受替制御では、これらの受替荷重を先行して決定するのがよいと判断される。また、C-11点の標準偏回帰係数は絶対値の平均|[ｂ]|、分散Ｓ^２ _ｂがともに比較的大きい。このことは、C-11点の受替荷重がある特定の箇所の建物の変位に大きく影響を与えることを示している。式(4.12)からそれらがA-9点、B-11点、C-11点の建物の変位であることがわかる。A-9点とB-11点、C-11点は建物の対角に位置することから、建物は図５に示すように対角線a-aを軸にねじれるような挙動をしていることが推察できる。
【００５０】
また、標準偏回帰係数の平均[ｂ]の分布を見ると、建物の外側に位置するA列，C列およびE列のそれは全体的にマイナスの傾向を、建物の中央部に位置するB列のそれは全体的にプラスの傾向を示していることがわかる。これは、建物外側のジャッキが建物を相対的に沈下させる方向へ働くこと、建物中央部のジャッキが建物全体を相対的に隆起させる方向へ働くことを示していると考えられる。すなわち、A列をジャッキアップした場合には、B列が支点となりそれと反対側のC列やE列が沈下する。この時、C列やE列のジャッキ荷重は増加を示しそれらは設計荷重値以上であることから、ジャッキ荷重を再調整する必要はなく、C列やE列の沈下が保持されたためと思われる。
【００５１】
一方、建物中央部のB列をジャッキアップした場合には、外側に位置するA列やC列も若干の隆起を示す、あるいは隆起を示さない場合でもそれらのジャッキ荷重は減少するため、技術者はジャッキ荷重を増加させ設計荷重の保持に努める。この結果としてA列やC列の建物変位は相対的に隆起を示すものと考えられる。
【００５２】
(3)実測値と予測値との比較およびその考察
図６、図７は、建物各点の変位と受替荷重の実測値、および重回帰式により算定したそれらの予測値を示したものである。図６のCASE１は、建物の変位と受替荷重の実測値が重回帰式の算定に用いた計測データの最大値と最小値との間に位置する場合である。ここでは、便宜上このようなデータを標本の範囲R内と呼ぶ。
【００５３】
図７のCASE２は、建物の変位と受替荷重の実測値の一部が、重回帰式の算定に用いた計測データの最大値と最小値との間以外に位置する場合である。ここでは、便宜上このようなデータを標本の範囲Ｒ外のデータと呼ぶ。
【００５４】
建物の変位の予測値は上式(6)に受替荷重の実測値を代入して求めたものであり、受替荷重の予測値は上式(7)に同時期の建物の変位の実測値を代入して求めたものである。なお、これらの実測値は重回帰式の算定に用いた標本以外の計測データの中から任意に抽出したものである。これらの図から、重回帰式による建物の変位と受替荷重の予測値は、それらの実測値と全体的によい一致を示していることがわかる。下式(13)に示すように誤差率ｅを設定すれば、CASE１は建物の変位で6.0％、受替荷重で1.6％であり、CASE２は建物の変位で6.9%、受替荷重で7.7%である。このことから、本手法による建物の変位および受替荷重の予測値はそれらの実測値とよい適合を示していると判断できる。このことは、建物のレベリングに重回帰式による統計的手法を用いれば、その工程の労力が大きく低減できることを示している。
【００５５】
【数１３】

ここに、ｘ_ｉ：建物ｉ点の受替荷重の実測値
Ｘ_ｉ：建物ｉ点の受替荷重の予測値
ｎ：受替箇所の数
次に、CASE２の誤差率がCASE１のそれと比較して若干大きい原因について考察を加える。CASE２では受替荷重の実測値が重回帰式の標本の範囲Rを越えている箇所があるため、その予想精度が低下したものと考えられる。一般に、重回帰式用いて予測を行う場合は、標本の範囲R内で説明変数を設定するのが望ましく、それが標本の範囲Rを大きく越える場合は誤差が大きくなる傾向がある。このような場合は、上式(7)に理想的な変位の条件（目標とする建物各点の変位）を与えて受替荷重の目安を算定し、上式(6)を用いて建物の変位が設定した許容範囲内（建物の健全性を確保する管理値とは異なる）に収まるように受替荷重を調整するなどの手段が必要になる。逆に、受替荷重を少ない制約で精度よく算定するためには、建物のレベリングまでの工程中に許容される範囲内（この場合は建物の健全性を確保できる管理値以内）でジャッキ荷重を意図的に増減させるなどの手段を講じ、重回帰式の算定に用いる標本がより多くの情報を含むようにすることが重要である。
【００５６】
〔建物のレベリングについて〕
下受け形式のアンダーピニングでは，図２に示すように建物の直下を掘削してそれを受替える。このため、受替が終了した後には、その部分を流動化処理土などにより埋戻して復旧する必要がある。埋戻しに先立ち、油圧ジャッキの機械的なロック、あるいはサポートジャッキへの盛替えを行うとともに、沈下計などの計測器を撤去するのが通例である。したがって、工事の最終時点における受替位置の建物変位の計測は不可能であり、建物外周部の計測結果などからその健全性を判断せざるを得ない。このように、工事の最終時点における建物各点の変位の実測値と予測値との対比やその結果に基づく補正も不可能なことから、埋戻し前に最適な受替荷重を算定し建物の各点に不等変位を残さないことが重要になる。
【００５７】
これらのことから、ここではCASE1を建物のレベリング前の状態とした場合の最適な受替荷重の試計算を行い、その過程についてまとめることとする．
建物の最終変位に影響を与える要因は、(1)流動化処理土の埋戻しによる荷重増分、(2)埋戻しにより生じる受替版下および建物下の地盤反力、(3)シールドの掘進による周辺地山のゆるみなどが挙げられる。施工条件によっては上載荷重の影響や近い将来に想定される建設物の影響などについても考慮する方が望ましいと思われる。本例では、(2)の地盤反力については、荷重増分により生じる受替版や建物の沈下（数mm程度）は流動化処理土の表面に生じるブリージングの領域である可能性が高いこと、(3)のシールド掘進の影響については、周辺地山が洪積粘性土層であること、応力導入時の新設杭の軸力計測結果から杭の周面摩擦力が相当に期待できることなどの理由から考慮しないことにした。このため建物の最終変位量（埋戻し以降の将来変位）は、埋戻しの荷重増分による受替版の沈下量から判断した。
【００５８】
受替版の沈下量は、平面あるいは立体の要素でモデル化し算定するのがよいと思われるが、ここでは計算の省力化の観点から受替版を図８に示すように、ばね２，２…で支持されたはり要素３，３…でモデル化した構造モデル１とし、荷重は載荷点から各はりの軸線までの距離に応じた受替荷重を作用させることで受替版の沈下量を算定することにした。はりの剛性はその分担幅と等価なものとし、新設杭の杭頭のばねのばね定数は受替版の沈下量の実測値と図８に示す構造モデルによる解析値が適合するように逆解析を行った結果から決定している。
【００５９】
図９はCASE１の荷重条件下における受替版各点の沈下量の算出結果を示したものである。この結果から埋戻し前の建物各点の変位が図１０(a)に示すような状態であるならば、埋戻した後の建物各点の変位がすべて0.00mmになり不等変位が残留しないと判断される。このときの建物各点の受替荷重は、本発明で示した統計的手法を用いて算出すれば図１０(b)に示すような分布となる。
【００６０】
ここで注意すべきことは、埋戻しによる荷重増分量と受替荷重の変動量との割合である。本工事の場合、前者が97.62MN、後者が11.63MNであるため、埋戻しによる荷重増分量の方が受替荷重の変動量よりも十分に大きく、受替版の沈下量はほぼ埋戻しによる荷重増分により決定されるが、埋戻し量が比較的少ない施工条件の場合には、建物各点の受替荷重の変動により受替版各点の沈下量が変化する割合が大きくなるため、受替版各点の変位がある程度の範囲に収束するまで繰り返し計算を行う必要がある。このことを考えれば、受替版の沈下量を算定する構造モデルは、図８に示したような簡易なはり-ばね系の構造モデルを用い、異なる荷重系における繰り返し計算を容易にする方が合理的と思われる。
【００６１】
以上、アンダーピニング工事における最終的な受替荷重の算定に重回帰式を用いる統計的手法の適用を提案するとともに、現場実測値と統計的手法による予測値とを比較することにより本手法の妥当性を検証した。これらから得られた知見を列挙すれば次のとおりである。
【００６２】
(1)建物各点の変位、受替荷重、および雰囲気温度の実測値を変数として重回帰式を求めることにより、任意の建物の変位を与える受替荷重の条件を算定することができる。すなわち、従来は多くの労力を要した建物のレベリングが容易になり、また残さざるを得なかった建物の不等変位を小さくすることができるため、工費の縮減やより高い品質の施工が可能になる。
【００６３】
(2)重回帰式による予測値の精度を向上するためには、多くの情報を含む標本を抽出することが重要である。このためには建物のレベリング時点までの工程中に、建物の変状が許す範囲で受替荷重の増減を行うなどの手段を講じ、重回帰式の標本の範囲を広げることが望ましい。また、受替荷重は一様に増減させるのではなく、それぞれ独立した標本が得られるように配慮した操作をする必要がある。
【００６４】
(3)個々の説明変数がほぼ独立していると見なせる場合には、標準偏回帰係数から、建物の変位の制御に重要になるジャッキの位置が判断できる。また建物の挙動は周辺の温度変化の影響大きく受けることがわかった。
【００６５】
(4)受替荷重の算定に重回帰式を用いる手法の最大の利点は、この重回帰式がアンダーピニング工事中の計測データから誘導されることにある。すなわち、建物本体の変形特性や周辺地盤の変形特性など、アンダーピニングの挙動予測に際して不明確となる要因を重回帰式がすでに含んでいるため、高い精度で受替荷重を算定することが可能である。しかし、この方法ではこれらの要因が建物各点の受替荷重にどの程度の影響を与えるかは明確にできない。
【００６６】
【発明の効果】
以上詳説のとおり本発明によれば、アンダーピニングにおいて、建物各点の変位、受替荷重、および雰囲気温度の実測値を変数として重回帰式を求めることにより、任意の建物の変位を与える受替荷重を精度良く算定することができる。
【００６７】
従って、埋戻し後の建物がレベリング状態となるように、埋戻し前の建物変位を求め、この変位を与える受替荷重を前記重回帰式から求め、この算定された受替荷重に基づいて既設構造物の受替荷重を調整した後、埋戻しを行うことにより、アンダーピニング工事完了後の建物に不等変位が残留しないようにできる。
【図面の簡単な説明】
【図１】 アンダーピニングによる受替状況を示す平面図である。
【図２】 アンダーピニングによる受替状況を示す断面図である。
【図３】 建物各点の受替荷重と変位および雰囲気温度の経時変化図である。
【図４】 標準偏回帰係数の特性を示す、(a)は絶対値の平均の分布図、(b)は分散の分布図、(c)は平均の分布図である。
【図５】 建物のねじれる挙動を示す説明図である。
【図６】 建物各点の変位と受替荷重の分布図（ＣＡＳＥ１）である。
【図７】 建物各点の変位と受替荷重の分布図（ＣＡＳＥ２）である。
【図８】 受替版の沈下量を求める構造モデルのイメージ図である。
【図９】 受替版各点の沈下量の予測値である。
【図１０】 建物のレベリングに係る、(a)は建物各点の変位の予測値、(b)は目標変位を与える受替荷重の予測値である。
【符号の説明】
１…構造モデル、２…ばね、３…はり要素[0001]
BACKGROUND OF THE INVENTION
  The present invention provides a method for accurately calculating a replacement load that gives an arbitrary displacement to an existing structure in underpinning work, and prevents the existing structure from remaining unequally displaced from the original leveling state when underpinning is completed. It is related with the leveling method for doing.
[0002]
[Prior art]
  In urban areas such as Tokyo and Osaka, overcrowding and agglomeration accompanying rapid economic growth are intense. Underground is no exception, and facilities such as railways, sewers, electric power, and communications are congested under the road, and the underground areas up to the middle and shallow are already full. Under these circumstances, the construction of new underground facilities such as railways is increasingly being carried out in the remaining deep areas and private underground areas.
[0003]
  In tunnel construction, the shield construction method is often used because it has little impact on the above-ground facilities and underground facilities. However, high-rise buildings are located on private land, and bridge piers such as elevated railways and elevated roads are located on the road, and the foundation piles become an obstacle when shields dig under them. For this reason, underpinning is often required to replace a part or all of the existing structure with a new foundation.
[0004]
  In particular, in the case of underpinning that replaces a part of an existing structure, the displacement behavior of the building is affected by the rigidity of the structure body such as underground beams, foundation footings, columns, walls, and upper beams, and foundations of non-replaced parts. It is quite complicated because it is affected by the peripheral friction of the pile, the tip support force, and the ground reaction force under the bottom of the building.
[0005]
  On the other hand, the soundness of the building is ensured by controlling the replacement load using a hydraulic jack. In this case, the replacement load is often controlled to hold the design load of the building, but the design load is often different from the true building load, resulting in unequal displacement at each point of the building. Many are seen. Even if this unequal displacement does not exceed the control value determined by the vertical displacement of the existing piles or the relative displacement between the existing piles etc., if the feeling of the user and further quality improvement are taken into account, each point of the building It is best to completely remove the unequal displacement and restore it to the state before construction. Hereinafter, the process of removing the unequal displacement generated at each point of the building in the final stage of the underpinning work is referred to as “leveling”.
[0006]
  In building leveling, unloading all jack loads at the same time may cause unexpected building deformation. For this reason, leveling of buildings is usually performed by unloading one or several jack loads. However, because the building is supported at multiple points as described above, the unloaded load is distributed to other locations, other jacks share the non-replacement part ground, etc. Although it is assumed, the mechanism is not clear and there are many cases where the behavior intended by engineers is not exhibited.
[0007]
  On the other hand, there are several studies on the prediction of behavior of existing structures in underpinning work. These prediction methods include calculating the vertical displacement of the pile head of an existing pile, calculating the displacement of an arbitrary point by modeling an existing structure with a frame structure of a spring support, and the existing structure from the subsidence amount of the underlying structure. It can be roughly divided into three that predict the behavior of things.
[0008]
  For example, the following Non-Patent Document 1 proposes a method of calculating the vertical displacement of a pile head using a displacement transfer function that takes into account the interaction between an existing pile and a new pile. Non-Patent Document 2 below proposes a method of calculating the vertical displacement of the pile head using a beam-spring structural model that takes into account the interaction between the existing pile and the new pile.
[0009]
  Further, the following Non-Patent Documents 3 and 4 propose a method of modeling an existing structure with a frame structure and a pile or ground with a spring. In this method, the rigidity of an aged structure, the selection of structural members that should be considered for rigidity, and the ground characteristics of the non-replacement part (the ground characteristics of the replacement part are the There are many unclear factors, such as (can be estimated to some extent from the subsidence characteristics), especially when the replacement structure is asymmetrical, and these effects become significant, making it difficult to properly model existing structures. There are many cases. For this reason, as shown in the following Non-Patent Document 5, etc., the amount of settlement of the underlying structure in the final state is calculated, and the final amount is calculated from the sum of the amount of elevation of the existing structure at the time when the removal of the existing pile is completed The method of predicting the state of a typical existing structure is often used in practice.
[0010]
[Non-Patent Document 1]
          Jiro Hirose, Joji Konishi, Yuji Nakamura: On the behavior of buildings supported by pile foundations during underpinning, Proceedings of the Japan Society of Civil Engineers, No.435 / VI-15, pp.43-50, 1991.9.
[Non-Patent Document 2]
          Ichino Michiaki, Shimizu Yukinori, Koizumi Satoshi: On the method of predicting displacement of pile foundation structures in underpinning works, Journal of Japan Society of Civil Engineers, No.700 / VI-54, pp.159-175, 2002.3.
[Non-Patent Document 3]
          Jiro Hayashi, Naotake Nishimura, Seto Matsuo, Hiroshi Koyama: Design and practice of subpin underpinning work in the construction of the Minamimorimachi, Katafuku Line, pp.86-94, 1996.3.
[Non-Patent Document 4]
          Hikaru Kato, Fumiyoshi Takatsuki, Honor Kondo, Shoichi Furuyama, Yoshio Kajiuchi: Impact prediction and measurement management in Shinkansen viaduct underpass, Tunnel Engineering Research Papers, Vol. 6, pp.309-314, 1996.11.
[Non-Patent Document 5]
          Seibu Nishibayashi, Shuichi Yahata: Underpinning of high-rise buildings, tunnel and underground, Vol. 21, No. 3, pp.7-16, 1990.3.
[0011]
[Problems to be solved by the invention]
  However, the methods of Non-Patent Documents 1 and 2 are compared with other methods because the interaction between the pile and the pile through the surrounding ground is taken into consideration for the period until the existing pile is removed. It shows a good agreement with the measured value, but it cannot be applied at the time of leveling the building after the existing pile is removed.
[0012]
  Moreover, in the case of the said nonpatent literature 5, although the health of a building can be checked and the displacement of each point of a building can be kept within a management value, the leveling of a building has to rely on the control according to the field etc. There was a problem.
[0013]
  Therefore, the main problem of the present invention is to propose a method for accurately calculating a replacement load that gives an arbitrary displacement in underpinning, and then to propose a method for leveling an existing structure using this calculation method. .
[0014]
[Means for Solving the Problems]
  In order to solve the above-mentioned problem, as the present invention according to claim 1, a part or all of an existing structure including a pile foundation is received by a replacement structure comprising a pile foundation newly provided in the basement and a replacement plate. In underpinning to change,
  Using the measured values of displacement, replacement load, and ambient temperature at each point of the existing structure as variablesThe following formula (7) ofMultiple regression equationByThere is provided a method of calculating a replacement load in underpinning, wherein a replacement load that gives an arbitrary displacement to the existing structure is calculated.
[Expression 7]

Where y _Iij : Displacement of each point in the building
       b _i ^Iij : Partial regression coefficient of multiple regression equation
        X _Iij : Predicted value of replacement load at each point of the building
          T p : Atmosphere temperature
       b ₀ ^Iij : Constant term of multiple regression equation
[0015]
  In underpinning, the effects of unclear factors such as the deformation characteristics of the structure body, the deformation characteristics of the surrounding ground, and the true load of the structure are significant, making it difficult to accurately grasp the behavior. However, in the present invention described in claim 1 above, a receiving device which applies an arbitrary displacement by a statistical method for obtaining a multiple regression equation using the displacement of each point of the existing structure, the replacement load, and the measured value of the ambient temperature as variables. The condition of replacement load was calculated. Therefore, since the multiple regression equation already includes the influence of the deformation characteristics of the building and the deformation characteristics of the surrounding ground, their individual mechanisms cannot be clarified, but the replacement load can be calculated with high accuracy. It becomes.
[0016]
  As the present invention according to claim 2, the replacement is performed in underpinning in which part or all of an existing structure including a pile foundation is replaced by a replacement structure including a pile foundation newly provided in the basement and a replacement version. A leveling method for setting the existing structure in a leveling state in a state in which replacement by a structure is completed and backfilling is performed,
  Prior to the backfill, at least the load increase due to backfillConsiderIn consideration, the amount of settlement of each point of the replacement version is obtained, and the opposite sign value of the amount of settlement of each point of the replacement version is set as the displacement of the existing structure before the backfilling, and this displacement is given as a condition The replacement load is calculated by the calculation method of the replacement load in the underpinning described in 1, and after the replacement load of the existing structure is adjusted based on the calculated replacement load, backfilling is performed. An existing structure leveling method in underpinning is provided.
[0017]
  The present invention according to claim 2 specifically relates to a procedure for bringing a building into a leveling state after backfilling. In other words, factors affecting future displacement of buildings (displacement after backfilling) include, for example, (1) load increment due to backfilling of fluidized soil, (2) under replacement plates and underbuildings caused by backfilling The ground reaction force of (3), the loosening of the surrounding ground due to the excavation of the shield. In addition, depending on the construction conditions, it may be desirable to consider the effects of the loading load and the effects of structures expected in the near future. Among these factors, at least the load increment due to backfilling is taken into account to determine the amount of settlement of each point of the replacement version, and the amount of settlement of each point of the replacement version is set so that the building displacement after backfilling becomes zero. The reverse sign value is set as the displacement of the existing structure before backfilling, the replacement load is calculated by the replacement load calculation method using the multiple regression equation with this displacement as a condition, and the calculated replacement After adjusting the replacement load of the existing structure based on the load, backfilling is performed.
[0018]
  As for the present invention according to claim 3, the calculation of each subsidence amount of the replacement version is made by modeling the replacement version with a beam element and using a structural model supported by a pile head spring of a new pile, and the load is loaded The leveling method of the existing structure in the underpinning of Claim 2 performed by applying the replacement load according to the distance from the point to the axis of each beam.
[0019]
  As the present invention according to claim 4, the pile head spring of the new pile supporting the replacement plate is subjected to an inverse analysis so that the measured value of the subsidence amount of the replacement plate matches the analysis value of the structural model. The leveling method of the existing structure in the underpinning according to claim 3 determined from the results.
[0020]
  In the present invention according to claim 5, in adjusting the replacement load of the existing structure based on the calculated replacement load, the amount of settlement at each point of the replacement plate changes due to the change in the replacement load of the existing structure. In consideration of what to do, the calculation is repeated until each point displacement of the replacement version converges to a predetermined range by the structural model. A leveling method is provided.
[0021]
DETAILED DESCRIPTION OF THE INVENTION
  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
[0022]
[Induction of multiple regression equation based on measurement data]
  For the prediction of the optimum replacement load that can remove the unequal displacement of the building, a multiple regression equation obtained from three variables of the actually measured displacement of the building, the replacement load, and the ambient temperature is used. The specific procedure for calculating the multiple regression equation based on the measurement data is described below.
(1) Construction overview and measurement overview
  The target construction was underpinning in the No. 1 Hiromachi tunnel construction area for the second stage construction of the Rinkai Fukutoshin Line. Daiichi Hiromachi Tunnel, located between Oimachi Station and Osaki Station, is a single-track railway tunnel with an outer diameter of 7,100mm constructed by the shield method. On this route, there are two buildings, 12 reinforced concrete JR Hiromachi offices and 8 reinforced concrete Shinagawa Ward Disaster Prevention Center. A part of the foundation piles of these buildings are located. Underpinning was carried out in the form of an undercarriage to hinder tunnel excavation. The subject of the analysis was JR Company's No. 6 building, which was supported by the most points among the three buildings underpinned (see Figs. 1 and 2). JR Company's House Building No. 6 has a total weight of 167 MN, of which 62.3 MN is replaced by 13 new piles via a replacement version.
[0023]
  For the calculation of the replacement load, a multiple regression equation obtained from three types of variables of the measured displacement of the building, the replacement load, and the ambient temperature was used. Replacement control takes a relatively long time from the introduction of stress to the removal of the existing piles. Based on this, we assumed that the displacement behavior of each point of the building was affected by the temperature change in the surrounding area, and decided to select the ambient temperature as the explanatory variable. This multiple regression equation already includes the influence of uncertain factors such as the deformation characteristics of the building body and the deformation characteristics of the surrounding ground, so the mechanism cannot be clarified, but the replacement load can be calculated with high accuracy. Is possible.
[0024]
  FIG. 3 shows the actual values of replacement load, ambient temperature, and building displacement used in the calculation of the multiple regression equation. These measurement data were measured from the time immediately after the introduction of stress until the completion of the cutting of the pile head of the existing pile, and the total number is about 5,000. Also from these figures, it has been found that the replacement load at the same location and the displacement of the building do not have a simple correlation.
[0025]
[Induction of multiple regression equation]
  Objective variable y_iIs the displacement of building i point, explanatory variable x_i,If T is the replacement load and the ambient temperature of the building i point, the displacement of each point of the building is expressed by the following equation. Although the purpose is to calculate the replacement load, the displacement is controlled by the displacement of each point in the building.
[0026]
[Expression 1]

Where y_i: Measured value of displacement at building i
        x_i: Measured value of replacement load at building i
        b₁, B₂... b_i, B_t: Partial regression coefficient of multiple regression equation (here, i = 14)
        T: Measured value of ambient temperature
        b₀: Constant term of multiple regression equation
[0027]
  Once the partial regression coefficient and constant term satisfying the above equation (1) are determined, the displacement of each target building point is substituted and x_iThe replacement load can be obtained by solving for. The partial regression coefficient was calculated by the following formula.
[0028]
[Expression 2]

Where S² _xi: Variable x_iDistribution of
       S_xiii: Variable x_iAnd variable y_iCovariance
[0029]
  The constant term of the multiple regression equation is expressed by the following equation (3).
[0030]
[Equation 3]

Where [y_i]: Variable y_iAverage of
        [x_i]: Variable x_iAverage of
  In addition, the above [] notation means an average value for convenience, and is an alternative to the overline display on the upper side of the character (the same applies hereinafter).
[0031]
  In order to clarify the meaning of multiple regression equations, it is better to give objective variables with small errors with as few variables as possible, and even if a highly accurate multiple regression equation is obtained, it will be calculated if there are too many variables Becomes complicated and impractical. For this reason, in the calculation of the multiple regression equation, the explanatory variables were selected by the method described below.
[0032]
  In general, when there is a high correlation between explanatory variables, the partial regression coefficient cannot be obtained, and it is known that multicollinearity occurs, for example, the sign of the partial regression coefficient does not match the sign of the single correlation coefficient of the objective variable. It has been. For this reason, the correlation between explanatory variables was determined by the following equation (4), and when a correlation was observed, one variable was excluded and a multiple regression equation was created. Here, the criterion for the determination is 0.9 or more.
[0033]
[Expression 4]

Where R_XX: Single correlation coefficient between explanatory variables
      S_XiXj: Variable x_iAnd variable x_jCovariance
        S_Xi: Variable x_iStandard deviation of
        S_Xj: Variable x_jStandard deviation of
  Next, the correlation between the objective variable and the explanatory variable is determined by the following equation (5), and those having an extremely small influence on the objective variable are excluded. Here, the criterion of determination was set to less than 0.1.
[0034]
[Equation 5]

Where R_Xy: Single correlation coefficient between explanatory variable and objective variable
      S_Xiyi: Explanatory variable x_iAnd the objective variable y_iCovariance
        S_Xi: Variable x_iStandard deviation of
        S_yi: Variable y_iStandard deviation of
  From these procedures, in the case of this construction, a 14-element simultaneous equation relating to the replacement load is obtained. If the simultaneous equations are represented by a determinant, the following equation (6) is obtained.
[0035]
[Formula 6]

here,y _Iij: Displacement of each point of the target building
       b _i ^Iij: Partial regression coefficient of multiple regression equation
        X_Iij: Predicted value of replacement load at each point of the building
          Tp: Predicted ambient temperature
       b ₀ ^Iij: Constant term of multiple regression equation
  In the above equation (6), the matrix representing the partial regression coefficient of the first term on the right side is 14 rows and 15 columns. Since the ambient temperature T is a variable independent of the replacement load, the matrix representing the partial regression coefficient can be converted into a 14 × 14 square matrix. Therefore, if the above equation (6) is converted into the following equation (7), a multiple regression equation that can calculate the replacement load can be obtained by substituting the displacement of each target building point.
[0036]
[Expression 7]

Where y _Iij : Displacement of each point in the building
       b _i ^Iij : Partial regression coefficient of multiple regression equation
        X _Iij : Predicted value of replacement load at each point of the building
          T p : Atmosphere temperature
       b ₀ ^Iij : Constant term of multiple regression equation
[0037]
[Verification of validity of displacement prediction method using multiple regression equation]
  In this section, we compare the field measurement results with the predicted value of the replacement load calculated from the multiple regression equation to verify the validity of the displacement prediction method at the construction stage using the multiple regression equation. In the following, the reliability of the obtained multiple regression equation will be considered prior to the comparative study between the actual measurement value and the predicted value.
[0038]
(1) Multiple regression equation test
[0039]
[Table 1]

[0040]
  Table 1 above shows the coefficient of determination adjusted for the degree of freedom of the multiple regression equation obtained by the method described in the previous section, and the variance ratio between the predicted value and the residual. Note that the multiple regression equation obtained follows a right-sided test assuming that the variance ratio F obtained by the following equation follows an F distribution with q and nql degrees of freedom (where n is the number of samples and q is the number of explanatory variables). And confirmed its reliability.
[0041]
[Equation 8]

[0042]
[Equation 9]

[0043]
[Expression 10]

Where F: dispersion ratio
       V_R: Unbiased variance of predicted values
       V_E: Unbiased variance of residuals
        q: Number of explanatory variables
        n: number of samples
  That is, if F is greater than the significance level, the hypothesis that the multiple regression equation is not useful for prediction is rejected, and the multiple regression equation is useful for prediction.
[0044]
  Looking at Table 1, the measured displacement values at each point of the building are highly correlated with the predicted values obtained from the multiple regression equation, and the variance ratio is F_{q, n-q-l}It can be seen that it is larger than (5%) (maximum 1.940 in this case). Therefore, it is judged that these multiple regression equations have sufficient reliability, and if the displacement of each target building point is given, the replacement load of each building point satisfying it can be obtained with sufficient accuracy. Show.
[0045]
(2) Characteristics of replacement load and displacement of structure by partial regression coefficient
  The units differ depending on the explanatory variable replacement load and the ambient temperature. In order to consider which variable has a greater influence in the obtained multiple regression equation, it was decided to obtain a partial regression coefficient when standardized. Standardization of measurement data is performed using the following formula.
[0046]
[Expression 11]

Where z_i: Variable x_iStandardized data for
      [x_i]: Variable x_iAverage of
       S_Xi: Variable x_iStandard deviation of
[0047]
[Expression 12]

[0048]
  The above equation (12) shows the partial regression coefficient when the measurement data is standardized. Originally, since the partial regression coefficient is an approximate value, it is difficult to accurately determine the influence of the replacement load at each point on the building on the displacement. However, in this case, since the correlation coefficient between explanatory variables is about 0.5 at the maximum, it is considered that the explanatory variables are almost independent, and it is related to load control and building behavior characteristics from the viewpoint of partial regression coefficient. Try to consider.
[0049]
  FIG. 4 shows the average of the standard partial regression coefficients | [b] |, the average [b], and the variance S² _bIs shown for each replacement position. In this figure, the B-10 point, C-10 point, C-9 point, C-7 point, E-6 point, and T variable are the average of the absolute values of the standard partial regression coefficients | [b ] | Is relatively large, while its variance S² _bIs relatively small. From this, each of these variables has a large influence on the displacement of the entire building, and it is determined that it is better to determine these replacement loads in advance in the replacement control in this construction. Also, the standard partial regression coefficient at the C-11 point is the average of absolute values | [b] |² _bAre relatively large. This indicates that the displacement of the building at a specific location with a C-11 replacement load has a significant effect. Equation (4.12) shows that these are the displacements of the building at points A-9, B-11, and C-11. Since points A-9, B-11, and C-11 are located on the diagonal of the building, it can be inferred that the building behaves like a twist around the diagonal aa as shown in Fig. 5. .
[0050]
  In addition, looking at the distribution of the mean [b] of the standard partial regression coefficient, the A column, C column and E column located outside the building show a generally negative trend, the B column located in the center of the building It can be seen that it shows a positive trend overall. This is considered to indicate that the jack on the outside of the building works in a direction that causes the building to sink relatively, and that the jack in the center of the building works in a direction that raises the entire building relatively. That is, when row A is jacked up, row B serves as a fulcrum, and rows C and E on the opposite side sink. At this time, the jack load of the C row and the E row showed an increase and they were higher than the design load value, so it was not necessary to readjust the jack load, and it seems that the subsidence of the C row and the E row was maintained. .
[0051]
  On the other hand, when the row B in the center of the building is jacked up, even if the rows A and C located on the outside also show some bulges, or even if they do not show any bulges, their jack load will decrease. Tries to maintain the design load by increasing the jack load. As a result, it is considered that the building displacements in row A and row C are relatively raised.
[0052]
(3) Comparison between the measured value and the predicted value
  FIG. 6 and FIG. 7 show the displacement of each point of the building, the actual measurement value of the replacement load, and the predicted values calculated by the multiple regression equation. Case 1 in FIG. 6 is a case where the measured values of the displacement of the building and the replacement load are located between the maximum value and the minimum value of the measurement data used for the calculation of the multiple regression equation. Here, for convenience, such data is referred to as the sample range R.
[0053]
  Case 2 in FIG. 7 is a case where some of the measured values of the displacement of the building and the replacement load are located other than between the maximum value and the minimum value of the measurement data used in the calculation of the multiple regression equation. Here, for convenience, such data is referred to as data outside the sample range R.
[0054]
  The predicted value of the displacement of the building is obtained by substituting the measured value of the replacement load into the above formula (6), and the predicted value of the replacement load is the measured value of the displacement of the building at the same time in the above formula (7). It is obtained by assigning a value. Note that these actual measurement values are arbitrarily extracted from measurement data other than the sample used for the calculation of the multiple regression equation. From these figures, it can be seen that the predicted values of the displacement of the building and the replacement load according to the multiple regression equation are in good agreement with the actual measured values. If the error rate e is set as shown in the following equation (13), CASE1 is 6.0% for the displacement of the building and 1.6% for the replacement load, CASE2 is 6.9% for the displacement of the building and 7.7% for the replacement load It is. From this, it can be determined that the predicted values of the displacement of the building and the replacement load according to the present method are in good agreement with those actually measured values. This indicates that if a statistical method based on multiple regression equations is used for building leveling, the labor of the process can be greatly reduced.
[0055]
[Formula 13]

Where x_i: Measured value of replacement load at building i
        X_i: Predicted value of replacement load at building i
         n: Number of replacement locations
  Next, the reason why the error rate of CASE2 is slightly larger than that of CASE1 will be considered. In CASE2, the actual value of the replacement load exceeds the sample range R of the multiple regression equation. In general, when prediction is performed using a multiple regression equation, it is desirable to set an explanatory variable within the sample range R, and when it greatly exceeds the sample range R, the error tends to increase. In such a case, the ideal displacement condition (target displacement of each building point) is given to the above equation (7) to calculate a guideline for the replacement load, and the above equation (6) is used to calculate the building Means such as adjusting the replacement load so that the displacement is within the set allowable range (different from the management value for ensuring the soundness of the building) are required. On the other hand, in order to accurately calculate the replacement load with few restrictions, the jack load must be within the allowable range during the process until the leveling of the building (in this case, within the control value that can ensure the soundness of the building). It is important to take measures such as intentionally increasing / decreasing so that the sample used for calculating the multiple regression equation contains more information.
[0056]
[About building leveling]
  In the underpinning type underpinning, as shown in Fig. 2, excavation is performed directly under the building to replace it. For this reason, after the replacement is completed, it is necessary to recover the portion by backfilling it with fluidized soil. Prior to backfilling, it is customary to mechanically lock the hydraulic jack or replace it with a support jack, and remove the measuring instrument such as a settlement meter. Therefore, it is impossible to measure the displacement of the building at the replacement position at the end of the construction, and it is necessary to judge the soundness from the measurement result of the outer periphery of the building. In this way, it is impossible to compare the actual displacement and predicted value of the displacement at each point of the building at the final point of construction and to correct it based on the result, so the optimal replacement load is calculated before backfilling. It is important not to leave unequal displacement at each point.
[0057]
Based on these facts, the trial calculation of the optimum replacement load when CASE1 is in the state before the leveling of the building is performed, and the process is summarized.
  Factors affecting the final displacement of the building are: (1) Increase in load due to backfilling of fluidized soil, (2) Ground reaction force under replacement plate and under building caused by backfill, (3) Shield advancement For example, loosening of the surrounding ground. Depending on the construction conditions, it may be desirable to take into account the effects of the loading load and the effects of structures expected in the near future. In this example, with regard to the ground reaction force of (2), it is highly possible that the replacement plate or subsidence of the building (about several mm) caused by the load increment is the region of breathing that occurs on the surface of the fluidized soil, Regarding the influence of shield excavation in (3), the surrounding ground is a sedimentary clay layer, and the axial frictional force measurement results of the new pile at the time of stress introduction can reasonably expect the peripheral friction force of the pile. I decided not to consider it. For this reason, the final displacement of the building (future displacement after backfilling) was judged from the sinking amount of the replacement version due to the incremental load of backfilling.
[0058]
  The sinking amount of the replacement version should be modeled and calculated with a plane or solid element, but here, from the viewpoint of labor saving in calculation, the replacement version is shown in FIG. The structural model 1 is modeled by the beam elements 3, 3 ... supported by…, and the load is subsidized by applying a replacement load according to the distance from the loading point to the axis of each beam. I decided to calculate. The rigidity of the beam is assumed to be equivalent to the shared width, and the spring constant of the pile head spring of the new pile is inversely analyzed so that the measured value of the subsidence amount of the replacement version matches the analysis value by the structural model shown in FIG. Is determined from the results of
[0059]
  FIG. 9 shows the calculation result of the amount of settlement at each point of the replacement version under the load condition of CASE1. From this result, if the displacement of each point of the building before backfilling is in the state as shown in FIG. 10 (a), the displacement of each point of the building after backfilling is all 0.00mm and no unequal displacement remains. It is judged. The replacement load at each point of the building at this time has a distribution as shown in FIG. 10B if it is calculated using the statistical method shown in the present invention.
[0060]
  What should be noted here is the ratio between the load increment due to backfilling and the amount of change in the replacement load. In the case of this construction, since the former is 97.62MN and the latter is 11.63MN, the load increment due to backfilling is sufficiently larger than the fluctuation amount of the replacement load, and the subsidence amount of the replacement version is almost due to backfilling. Although it is determined by the load increment, under construction conditions where the amount of backfill is relatively small, the rate of change in the amount of settlement at each point of the replacement version increases due to the change in the replacement load at each point of the building. It is necessary to repeat the calculation until the displacement of each point on the replacement version converges to a certain range. Considering this, the structural model for calculating the sinking amount of the replacement version should be a simple beam-spring system structural model as shown in Fig. 8, which facilitates repeated calculations in different load systems. It seems reasonable.
[0061]
  As described above, we propose the application of a statistical method using multiple regression equations to calculate the final replacement load in underpinning work, and compare the actual measured values with the predicted values obtained by the statistical method. The sex was verified. The findings obtained from these are enumerated as follows.
[0062]
(1) By obtaining a multiple regression equation using the measured values of the displacement of each point of the building, the replacement load, and the ambient temperature as variables, the condition of the replacement load that gives the displacement of any building can be calculated. In other words, it has become easier to level buildings that require a lot of labor in the past, and the unequal displacement of buildings that had to be left behind can be reduced, so that construction costs can be reduced and higher quality construction can be achieved. Become.
[0063]
(2) In order to improve the accuracy of the predicted value by the multiple regression equation, it is important to extract a sample containing a lot of information. For this purpose, it is desirable to increase the range of the multiple regression equation samples by taking measures such as increasing or decreasing the replacement load within the range permitted by the deformation of the building during the process up to the leveling of the building. Moreover, it is necessary not to increase or decrease the replacement load uniformly but to perform an operation in consideration of obtaining independent samples.
[0064]
(3) If each explanatory variable can be regarded as almost independent, the position of the jack that is important for controlling the displacement of the building can be judged from the standard partial regression coefficient. It was also found that the behavior of the building is greatly affected by the surrounding temperature change.
[0065]
(4) The greatest advantage of the method using the multiple regression equation for the calculation of replacement load is that this multiple regression equation is derived from measurement data during underpinning work. In other words, the multiple regression equation already contains factors that are unclear in predicting the behavior of underpinning, such as the deformation characteristics of the building body and the surrounding ground, so it is possible to calculate the replacement load with high accuracy. is there. However, this method cannot clearly determine how much these factors affect the replacement load at each point of the building.
[0066]
【The invention's effect】
  As described above in detail, according to the present invention, in underpinning, a replacement that gives a displacement of an arbitrary building by obtaining a multiple regression equation using the measured values of the displacement of each point of the building, the replacement load, and the ambient temperature as variables. The load can be calculated with high accuracy.
[0067]
  Therefore, the building displacement before backfilling is calculated so that the building after backfilling is in a leveled state, the replacement load that gives this displacement is determined from the multiple regression equation, and the existing displacement is calculated based on the calculated replacement load. By performing backfill after adjusting the replacement load of the structure, it is possible to prevent unequal displacement from remaining in the building after completion of the underpinning work.
[Brief description of the drawings]
FIG. 1 is a plan view showing a replacement situation due to underpinning.
FIG. 2 is a cross-sectional view showing a replacement situation due to underpinning.
FIG. 3 is a time-dependent change diagram of replacement load and displacement at each point of the building and ambient temperature.
4A and 4B show characteristics of a standard partial regression coefficient, where FIG. 4A is a distribution chart of the average of absolute values, FIG. 4B is a distribution chart of dispersion, and FIG.
FIG. 5 is an explanatory diagram showing the twisting behavior of a building.
FIG. 6 is a distribution diagram (CASE 1) of displacement of each point of the building and replacement load.
FIG. 7 is a distribution diagram (CASE 2) of displacement of each point of the building and replacement load.
FIG. 8 is an image diagram of a structural model for obtaining a sinking amount of a replacement version.
FIG. 9 is a predicted value of the amount of settlement at each point of the replacement version.
10A is a predicted value of the displacement of each point of the building, and FIG. 10B is a predicted value of a replacement load that gives a target displacement.
[Explanation of symbols]
  1 ... Structural model, 2 ... Spring, 3 ... Beam element

Claims (5)

In the underpinning that replaces part or all of the existing structure with a pile foundation by a replacement structure consisting of a pile foundation newly installed in the basement and a replacement version,
The replacement load that gives an arbitrary displacement to the existing structure is calculated by the multiple regression equation of the following equation (7) obtained by using the displacement of each point of the existing structure, the replacement load, and the measured values of the ambient temperature as variables. A method for calculating a replacement load in underpinning.

Where y _{Iij is} the displacement of each point in the building
b _i ^Iij : Partial regression coefficient of multiple regression equation
X _Iij : Predicted value of replacement load at each point of the building
T p : Atmospheric temperature
b ₀ ^Iij : Constant term of multiple regression equation In the underpinning in which part or all of the existing structure with pile foundation is replaced by a replacement structure consisting of a pile foundation newly installed in the basement and a replacement version, the replacement by the replacement structure is completed and buried. A leveling method for bringing the existing structure into a leveling state in a state where restoration is performed,
Prior to returning the filled, obtains each point subsidence of受替plate by taking into account the load increment by at least backfilling, the existing structures before backfilling the opposite sign values of each point subsidence of the受替edition The displacement is set as a displacement, and the displacement is calculated by the displacement load calculation method in the underpinning according to claim 1 with the displacement as a condition, and the replacement of the existing structure is performed based on the calculated displacement load. A method for leveling an existing structure in underpinning, characterized by performing backfilling after adjusting the load. The amount of settlement of each point of the replacement version is calculated by modeling the replacement version with beam elements and using a structural model supported by the pile head spring of the new pile, and the load is the distance from the loading point to the axis of each beam. The leveling method of the existing structure in the underpinning according to claim 2, which is performed by applying a replacement load according to the above. The pile head spring of the new pile that supports the replacement plate is determined from a result of performing an inverse analysis so that an actual measurement value of the subsidence amount of the replacement plate matches an analysis value of the structural model. Leveling method for existing structures in underpinning. When adjusting the replacement load of the existing structure based on the calculated replacement load, when taking into account that the amount of settlement at each point of the replacement version changes due to the change in the replacement load of the existing structure, The leveling method for an existing structure in underpinning according to any one of claims 3 and 4, wherein the calculation is repeated until each point displacement of the replacement version converges to a predetermined range by the structural model.

Images