JP2005043197A - Concrete inspection method by electromagnetic wave, and concrete inspection device by electromagnetic wave - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform more detailed determination than hitherto with respect to the kind, the depth, the thickness or the like of a defect part in a concrete structure. <P>SOLUTION: An electromagnetic wave is transmitted to the concrete structure and a reflected wave is received, and A<SB>0</SB>×r is determined by the relation between A/G on an initial motion peak of the measured internal defect reflected wave and a round-trip propagation time t and the following expression, and the kind of the defect part is discriminated based on A<SB>0</SB>×r. The expression is as follows: ln(A/G)=-(α×V×t)+ln(A<SB>0</SB>×r), where, A is the amplitude recorded in a measuring device, G is an amplification factor, α is an attenuation factor of the amplitude in concrete, V is an electromagnetic wave velocity, A<SB>0</SB>is an initial amplitude on the concrete surface of the electromagnetic wave transmitted from a transmission antenna, and r is a reflectivity of the electromagnetic wave on the internal defect. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

【００３１】
測定には、米国ＧＳＳＩ社（Ｇｅｏｐｈｙｓｉｃａｌ Ｓｕｒｖｅｙ Ｓｙｓｔｅｍｓ， Ｉｎｃ．）の商品名ＳＩＲ−２なる測定装置を利用した。アンテナにはダイポールアンテナを利用し、中心周波数によって透過深度や分解能が異なることから、１５００ＭＨｚ、９００ＭＨｚ、４００ＭＨｚのものを使用した。測定は欠陥上部のコンクリート打設後、１、２、３、４、６、８、１０週経過時に行った。その際、測線はアンテナが内部欠陥の中央を通るように設定した。主な測定パラメータを表２に示す。
【００３２】
【表２】

【００３３】
取得する波形データはＳＩＲ−２のハードディスクに収録されるが、データ長を１６ビットにしたことから、記録上の振幅は、−３２７６８〜３２７６７の範囲になる。これらは増幅後の値であるが、後にヘッダファイルから読み取った増幅率で除して、増幅前の振幅を求めることができる。
【００３４】
送信アンテナＴから放射された電磁波が内部欠陥で反射し、受信アンテナＲに捉えられるまでの各振幅を図８に示すように、送信アンテナＴから送信されるコンクリートへの入射波の振幅（初期振幅）をＡ_０、内部欠陥への入射波振幅をＡ_１、内部欠陥からの反射波の振幅をＡ_２、受信アンテナＴでの受信振幅をＡ_３とすると、これらの関係は次式で表される。
【００３５】
【数１】

【００３６】
ここで、Ａは測定器に記録される振幅で（Ａ_３・Ｇ）で与えられる。Ｇは増幅率、（Ａ_２／Ａ_１）＝ｒは内部欠陥の反射率、αはコンクリート中での減衰率、Ｚ（＝Ｖ・ｔ、Ｖ：電磁波速度、ｔ：往復伝播時間）は伝播距離で、Ｚ^ｎは幾何学的な発散による減衰を表す。Ｚ^ｎについて、コンクリート中に異常反射波がない道路トンネルにおける支保工反射波の初動に関する振幅／増幅率〜往復伝播時間関係を図９に示す。これからほぼ直線関係にあり、ｎ≒０が成り立つことが判る。これらを［数１］式に代入し、対数を取ると式のようになる。
【００３７】
【数２】

【００３８】
Ｖをワイドアングル法などにより測定すれば、複数深度にわたるｌｎ（Ａ／Ｇ）〜ｔ関係の勾配からαを、ｙ切片からＡ_０・ｒを求めることができる。Ａ_０は表面の状態が変わらない限りアンテナ毎にほぼ一定であるため、反射率の評価が可能である。ここで、ワイドアングル法とは、送受信アンテナの中心位置を固定し送受信アンテナ間隔を変化させる方法である。例えば、図８において、送信アンテナＴと受信アンテナＲとの間隔を暫時離隔させる。なお、本方法では識別のし易さと、背面反射波のような他の反射波の影響を最も受け難いことから、内部欠陥反射波の初動のピークに関するＡ／Ｇ、ｔを取り扱うことにしている。
【００３９】
アンテナ毎の内部欠陥の検出結果について、材齢１週における内部欠陥の検出例を図１０〜１３に示す。これらの左側は、測定記録から無欠陥個所の平均波形を差し引く背景除去処理を行った後のラインスキャン表示記録である。右側には左側の↓で示すトレースにおける欠陥波形の初動に関する振幅Ａ、往復伝播時間ｔの読み取り位置を示した。
【００４０】
一般に、コンクリート中における電磁波の減衰は取り扱うアンテナの中心周波数や主にコンクリートの比抵抗によって変化し、周波数が高いほど、比抵抗が低いほど減衰が大きくなる。このため、使用するアンテナとコンクリートの材齢によって、検出可能な欠陥の最大かぶりが変化する。コンクリートでは、材齢が経過するほど比抵抗が大きくなることが知られており、材齢が浅いほど比抵抗が小さく、減衰が大きいことから、深いかぶりの内部欠陥検出が難しい。
【００４１】
材齢１週から１０週に渡り測定を行った。以下、アンテナ毎の検出結果について説明する。
【００４２】
１）１５００ＭＨｚアンテナ
空洞は材齢１週ではかぶり２０ｃｍまで、２週以降は３０ｃｍまで検出可能であった。ジャンカはかぶり２０ｃｍまでは材齢１週から検出可能であったが、かぶり３０ｃｍは１０週経過後も検出できなかった。低強度はかぶり１０ｃｍのみ検出できているようであったが、骨材によると見られる不規則な散乱波が重なっており、反射面が不鮮明であった。
【００４３】
２）９００ＭＨｚアンテナ
空洞とジャンカについては、材齢１週からかぶり３０ｃｍまで検出可能であった。低強度はかぶり１０ｃｍのみであるが材齢１週から検出できた。
【００４４】
３）４００ＭＨｚアンテナ
９００ＭＨｚアンテナとほぼ同様の検出状況であった。
【００４５】
このようにかぶり３０ｃｍまでの範囲であれば、複数のアンテナを使用することにより、空洞とジャンカは材齢１週から検出可能である。一方低強度はかぶりが浅いもののみ材齢１週から検出可能で、かぶりが２０ｃｍ以上では材齢によらず検出困難であることになった。これは、減衰率というよりはむしろ、空洞とジャンカではコンクリートとの比誘電率の差あるいは電磁波速度の差が大きく、入射波に対する反射波の振幅比を表す反射率が大きいのに対して、低強度では小さく、反射波が微弱であることに起因している。
【００４６】
次に、内部欠陥の厚さの影響に関する検討を加える。本方法では内部欠陥からの反射波の初動に関する振幅／増幅率と往復伝播時間の関係から減衰率、初期振幅×反射率を求めることにしている。この場合、図１４のように、反射波が欠陥の前面反射波と背面反射波の合成波になることから、欠陥の厚さと使用する電磁波の波長との関係で反射波が乱れ、初動が変化することが考えられる。そこで、測定記録からアンテナ毎に初動反射波が変化しない欠陥厚を検討した。
【００４７】
図１５〜１７は順に、１５００ＭＨｚ、９００ＭＨｚ、４００ＭＨｚアンテナによりかぶり１０ｃｍの空洞模型の検査を行った際に得られた、欠陥中央部の６０トレース分の往復伝播時間、反射波の初動振幅／増幅率をプロットしたものである。これらから、往復伝播時間は欠陥の厚さによらずアンテナ毎にほぼ一定であるが、初動振幅は欠陥の厚さの影響を受けることが判る。すなわち、受信波形の半波長が４ｃｍ程度である１５００ＭＨｚでは、欠陥の厚さが５ｃｍ、１０ｃｍ、２０ｃｍのいずれでも振幅はあまり変わらない。
【００４８】
しかし、半波長が８ｃｍ程度の９００ＭＨｚでは１０ｃｍ、２０ｃｍはほぼ同じであるが、５ｃｍでは５０％近く小さくなる。４００ＭＨｚは半波長が１３ｃｍ程度であるため、いずれの厚さにおいても振幅が異なる。従って、予想されるように、初動振幅が欠陥の厚さの影響を受けないのは半波長程度以上を有する場合に限られることになった。ただし、ジャンカでは電磁波速度が空洞よりも遅く、波長が短くなることから、より薄いものまで適用可能である。
【００４９】
次に、減衰率、初期振幅×反射率の値と種別判定について考察する。１５００ＭＨｚでは欠陥厚５、１０、２０ｃｍの全てを用い、９００ＭＨｚでは１０、２０ｃｍの記録を用い、４００ＭＨｚでは２０ｃｍのみを用いた材齢１週におけるｌｎ（Ａ／Ｇ）〜ｔ関係を図１８〜２０に示す。これから、欠陥の種別によってＡ／Ｇが異なり、低強度、ジャンカ、空洞の順に大きくなっていること及び種別毎に傾きがほぼ同じでｙ切片が異なるグループに分かれることが判る。傾きはコンクリート中の電磁波振幅の減衰率であり、内部欠陥の種別とは無関係になることに整合している。一方、ｙ切片は初期振幅×反射率を表しており、反射率は次式の関係がある。ただし、ε_１、Ｖ_１は通常強度コンクリートの比誘電率及び電磁波速度、ε_２、Ｖ_２は内部欠陥の比誘電率及び電磁波速度である。
【００５０】
【数３】

【００５１】
ここで、Ｖ_１＝９ｃｍ／ｎｓとし、空洞でＶ_２＝２７ｃｍ／ｎｓ、ジャンカでＶ_２＝１５ｃｍ／ｎｓとおくと、反射率は空洞とジャンカで０．５１、０．２５になるはずであるが、求めたｙ切片の値も９００ＭＨｚ、４００ＭＨｚではそれに近い比率になっている。次に、９００ＭＨｚアンテナから求めた減衰率と初期振幅×反射率の材齢による変化を図２１、２２に示す。このように、減衰率は空洞とジャンカでほぼ同じで、６週までは減少し、その後一定値に落ち着く傾向にある。一方、初期振幅×反射率は、測定日における表面の含水条件によって初期振幅が変化するため、見かけ上、材齢によって大きく変化する。しかし、空洞とジャンカでの比率は材齢によらず１：０．５に近い値で推移する。従って、減衰率の変化や初期振幅の変化によって振幅が変化することがあったとしても、空洞とジャンカではｌｎ（Ａ／Ｇ）〜ｔ図上で倍近い差があり識別可能である。このように、内部欠陥として空洞と作成したジャンカに近い配合のものがあり、半波長以上の厚さであれば、材齢１週から種別判定が可能である。しかし、低強度は、電磁波速度が通常強度コンクリートと差がないようであり、現状では検出すること自体が難しく、初期振幅×反射率ｒによる種別の判定は困難である。
【００５２】
それ以外の適用条件について、半波長未満では、数値的に未検討であるが、厚さが薄くなるほど種別判定が困難になる。しかし、ジャンカの配合に関しては、次のように、ある程度変化したとしても判定可能である。
【００５３】
予備実験で作成した４種類のジャンカの配合と１５００ＭＨｚアンテナを使用して求めた電磁波速度を表３に、電磁波速度と水を除く全重量に対する粗骨材の重量比の関係を図２３に示す。模型に採用したのは配合２であるが、粗骨材の割合によって電磁波速度が変化することが判る。通常強度コンクリ―トの電磁波速度を９．５ｃｍ／ｎｓとし、上記数３の式から、配合２の反射率を１とした場合の比率を求めると、配合１、配合２、配合３、配合４、空洞では０．６５、１、０．３１、１．４１、２．２２になる。種別判定の面からは空洞との差があるほど容易であることから、配合１、配合３は配合２よりも判定し易い。また、粗骨材１００％の配合４は空洞の反射率に近くなるが、それでも６０％程度の反射率であることから判定できる可能性がある。
【００５４】
【表３】

【００５５】
次に、本発明の第二の実施形態について説明する。本実施形態では、主信号と背面信号との比較により、欠陥部の種類が判定可能である。
【００５６】
まず、散乱源及び反射源解析について考える。電磁波はコンクリート内部の欠陥で散乱されたり、境界面で反射されることにより、散乱波及び反射波として観測される。この散乱波及び反射波の発生状態は、欠陥の種類によって異なると考えられる。
【００５７】
元データでは、コンクリート表面の反射波や背景的なノイズがあり、これらが欠陥波形に影響を与えることが考えられる。したがって、欠陥のない部分のデータを用いて、背景除去を行った。これより、欠陥波形のみの抽出ができ不要な信号の影響が低下すると考えられる。また、レーダー計測では、深い位置の信号を確認できるようにするため、深さが深いほどゲインを高めて、データ採取を行っている。これは、レーダー波形を歪めることになるので、解析時にデータ採取時のゲインの増分を補正した（以後、元データを増幅率で除したものをゲイン補正と呼ぶことにする）。図２４，２５に１５００ ＭＨｚのアンテナで、材齢８週におけるかぶり１０ｃｍの空洞及びジャンカの元データと背景除去処理＋ゲイン補正したレーダーのイメージ像を示す。
【００５８】
この散乱波及び反射波の信号強度ならびに形状から、欠陥の特徴を検討するために、欠陥波形を抽出しヒルベルト変換により包絡線信号を求めて比較した。
【００５９】
このヒルベルト変換は次式により定義される。
【００６０】
ｙ（ｔ）＝ｘ（ｔ）＊ｈ（ｔ）
ここで、＊はたたみ込み演算であり、ｈ（ｔ）＝１／（πｔ）である。
【００６１】
図２６にゲイン補正後、背景除去処理により欠陥波形を抽出し、包絡線信号に変換した結果例を示す。（ａ）は空洞部を有する試験体（かぶり１０ｃｍ，欠陥厚５ｃｍ）を用いた場合に得られる欠陥波形を、（ｂ）は欠陥のない健全部波形を示す。前半部（１ｎｓ付近）の波形は両者ともほぼ同様な形状であることから、背景除去処理には、健全部の波形を利用することにより欠陥信号のＳ／Ｎを良くした。背景除去処理した結果は（ｃ）に示す。（ｄ）は、背景除去した波形を絶対値に変換した整流信号、（ｅ）はその包絡線信号を示す。包絡線信号は、ヒルベルト変換により計算した。この欠陥信号が出現している範囲で、散乱及び反射強度を求めるため包絡線信号を平均した。
【００６２】
次いで、電磁波速度解析を行う。コンクリート内部に存在するジャンカ、低強度および空洞の誘電率はそれぞれ異なるため、これらの欠陥部分の電磁波速度も異なる。したがって電磁波が伝搬する経路中の電磁波速度が知れれば、その経路内に存在する欠陥種別を判定することが可能である。本研究に用いたコンクリートモデルには、鋼製支保工が表面から４０ｃｍの深さに挿入されている。欠陥部および健全部の支保工信号の出現位置から、各部の電磁波速度を測定した。その結果、空気中の電磁波速度をｃとして、健全部の電磁波速度は約０．３４ｃと見積もられた。また低強度部の速度は健全部と同じ０．３４ｃで、ジャンカ部は健全部と低強度部より速く約０．６５ｃであった。これらの値は材齢が８週までで大きな変化がないこと、並びに用いた周波数の範囲で周波数に依存しないことも確認できた。
【００６３】
図２７にかぶり１０ｃｍのジャンカを８週目に探査して得られた支保工信号の例を示す。支保工の深さが既知であれば、支保工信号が最も表面に近付いたときの伝搬時間から経路中の電磁波速度を計算することが可能であるが、実際のトンネル覆工では支保工の深さは不明である。このため本研究では、支保工信号の軌跡を最小二乗近似を用いてカーブフィットすることで電磁波速度を求めた。なお、空洞部では欠陥の反射率が大きいため、空洞の下に存在する支保工信号を検出できず、したがって電磁波速度を検出することもできなかった。但しこのような欠陥部分における支保工信号の消失が確認できれば、存在する欠陥は空洞であると判断できる。このため本研究では、ジャンカと低強度の判定を目的として検討を行った。また深い位置にある支保工を扱うため、４００ＭＨｚの信号で検討を行った。
【００６４】
ここで、散乱源及び反射源解析結果について検討を加える。図２８に１５００ＭＨｚアンテナで、材齢２，４，８週におけるかぶり１０ｃｍの包絡線信号を示す。１５００ＭＨｚの結果では、低強度の反射波は確認できなかったため空洞とジャンカのみ示す。図２８では、縦軸は欠陥の前面信号のピークで規格化し、横軸は欠陥前面信号のピーク付近の時間が０になるように調整している。
【００６５】
図２８から空洞での包絡線信号は、ほとんど欠陥の前面信号のみであるのに対して、図２９ではジャンカでの包絡線信号は厚さ１０，２０ｃｍでは欠陥の背面信号が明瞭に認められた。これは、前述する電磁波速度解析の結果から得られたジャンカの電磁波速度０．６５ｃ（光速ｃ＝３×１０^８ｍ／ｓ）から確認した。空洞においても、２，４週目ではピークが確認できなかったが、８週目の厚さ２０ｃｍの信号で欠陥の背面信号と推定されるピークが認められた。８週目データの空洞、ジャンカの厚さ２０ｃｍの包絡線信号を比較すると、空洞でも欠陥の背面信号が認められているものの、ジャンカにおける欠陥の背面信号の方が明瞭に確認できることが分かった。以上の結果は、空洞はジャンカに比べて電磁波速度が速いため、前面信号と背面信号が、本研究で製作した厚さ程度であると分離できないことや、空洞は欠陥の前面信号の強度がジャンカに比べて大きく、欠陥表面で電磁波が大きく反射され透過する量が減少したことにより、欠陥の背面信号が小さくなったことが考えられる。また、初期の材齢において、空洞では欠陥の背面信号が認められないこと等を含めて判断すると、欠陥の背面信号が明瞭に確認できればジャンカであると推定できることが考えられる。
【００６６】
さらに、空洞の包絡線信号の２，４週目のデータでは、２〜３ｎｓの間で信号強度が増加している部分が認められた。これは、電磁波速度解析から求めたコンクリート中の電磁波速度０．３４ｃから計算すると、かぶり深さ１０ｃｍの約２倍の間隔に相当し、欠陥の前面信号の繰り返し信号と推定される。このような繰り返し信号は、ジャンカには認められず、空洞の特徴と判断することができる。
【００６７】
かぶりの深い部分のデータを確認するために、９００，４００ＭＨｚアンテナで、材齢８週における結果を図２９，３０に示す。
【００６８】
９００ＭＨｚにおいては、ジャンカにおける欠陥の背面信号は厚さ２０ｃｍで明瞭に確認できた。厚さ１０ｃｍでは、欠陥の背面信号と欠陥の前面信号との重なりで、信号の幅は大きくなっているものの明瞭に確認はできなかった。４００ＭＨｚの結果においても、ジャンカにおける欠陥の背面信号は厚さ２０ｃｍについてのみ確認できた。
【００６９】
９００ＭＨｚのかぶり１０ｃｍ、厚さ５ｃｍのデータを比較すると、空洞とジャンカともに約２ｎｓｅｃの時点で信号強度が増加している部分が認められた。１５００ＭＨｚの結果と比較すると９００ＭＨｚは、Ｓ／Ｎが良く、ジャンカにおいても認められたものと考えられる。その他の９００ＭＨｚと４００ＭＨｚの結果は、パルス幅が大きいことや鋼製支保工の影響により、欠陥の前面信号の繰り返し信号を確認することは困難であった。
【００７０】
以上の結果から、空洞とジャンカを識別するためには、欠陥の厚さや使用するアンテナやかぶり深さ等の条件が限られるが、１）欠陥の背面信号の有無を確認すること、また、２）補足的に欠陥の前面信号の繰り返し信号を確認することにより、種別判定が可能であると考えられる。
【００７１】
また、包絡線信号から欠陥の背面信号が検出できた空洞とジャンカについて、欠陥の前面信号と背面信号のピークの時間差を求めて、電磁波速度解析から求めた電磁波速度より欠陥の厚さを求めた。１５００ＭＨｚの結果を表４に示す。表４に示すように若干誤差があるもののほぼ妥当な値を示している。この結果から、速度既知で計算することにより欠陥の厚さが推定できることが分かる。
【００７２】
【表４】

【００７３】
このように、包絡線信号からの散乱源及び反射源解析から欠陥種別判定の可能性を検討した結果、欠陥の背面信号が明瞭に確認できれば、ジャンカを識別できる可能性が得られた。また、補足的に欠陥の前面信号の繰り返し信号を確認することにより欠陥種別判定できる可能性も得られた。欠陥の背面信号が得られれば、欠陥の前面信号と背面信号のピークの時間差により厚さ推定可能であることが分かった。
【００７４】
次に、本発明の第三実施形態について説明する。本実施形態では測定される平均電磁波速度の値により、欠陥の種類を判定する。Ｈ型鋼である支保工のウェブ部分は図３１に示すように平坦であるが、これが透過撮影レーダー探査では、図２７に示すように湾曲した形状となる。したがって、図３１と図３２との幾何学的関係から最小二乗法を用いれば、平均電磁波速度を求めることができる。
【００７５】
図３１において、ｄ；支保工の深さ、ｘ；支保工直上から横方向へのアンテナの移動距離、ｖ；コンクリート中の電磁波速度と定義する。図３１でアンテナをコンクリート表面に沿って移動させ探査を行うと、支保工信号の軌跡は図３２の如くになる。ここで、ｉ；測定位値の値のサフィックスで、ｎ点存在、ｘｉ；ｉ番目の横方向のアンテナ位置、ｙｉ；ｘｉでの支保工までの電磁波の往復伝搬時間とそれぞれ定義する。
【００７６】
ｙｉ＝２・（ｘｉ^２＋ｄ^２）^０．５／ｖ
但し、ｄとｖは未知。
【００７７】
ｘｉとｙｉは測定値で、これからｄとｖを求める。
【００７８】
上式から
（ａ・ｘｉ^２＋ｂ）^０．５＝ｙｉ／２
ただしａ＝１／ｖ^２ ｂ＝（ｄ／ｖ）^２
ａ・ｘｉ^２＋ｂ＝ｙｉ^２／４＝ｚｉ
ｒ＝Σ（ａ・ｘｉ^２＋ｂ−ｚｉ）^２ （Σは１〜ｎ）
ｒは残差でこれを最小にする
【００７９】
∂ｒ／∂ａ＝０から
ａΣｘｉ^４＋ｂΣｘｉ^２＝Σｘｉ^２・ｚｉ
∂ｒ／∂ｂ＝０から
ａΣｘｉ^２＋ｂΣ＝Σｚｉ
よって，
ａΣｘｉ^２＋ｎｂ＝Σｚｉ
これを解いて
【００８０】
【数４】
ａ＝｛ｎΣｘｉ^２・ｚｉ−Σｘｉ^２・Σｚｉ｝／｛ｎΣｘｉ^４−（Σｘｉ^２）^２｝
ｂ＝｛Σｘｉ^２・Σｚｉ−ａ（Σｘｉ^２）^２｝／ｎΣｘｉ^２
【００８１】
したがって、
平均電磁波速度：ｖ＝（１／ａ）^０．５、支保工の深さ：ｄ＝ｖ・ｂ^０．５ となる。
【００８２】
図３３に材齢２週から８週の範囲で、このようにして求めた健全部の電磁波速度を測定した結果を示す。図中の２点は、ジャンカと低強度のコンクリートモデルでの測定結果を表す。支保工の深さと支保工信号の出現位置から別途求めた健全部の電磁波速度である０．３４ｃを、図に横線で合わせて示している。図のようにカーブフィットで求めた電磁波速度は、材齢及び用いたコンクリートモデルによらず、ほぼ横線で示した値に一致する。
【００８３】
次に図３４に同じく材齢２週から８週で測定した低強度部の電磁波速度を示す。図では、各材齢をかぶり深さ１０ｃｍ、２０ｃｍ及び３０ｃｍに分けて示している。さらに各かぶり深さ中の３点は、欠陥厚さ５ｃｍ、１０ｃｍ及び２０ｃｍを表す。かぶりと欠陥の厚さが大きくなると、支保工信号と欠陥反射波が重なり、支保工の軌跡を正確にトレースできなかった。このような場合は計測不可とした。図に見られるように、ばらつきは認められるものの低強度部の電磁波速度は、図３３の健全部とほぼ同じ値になった。同じくカーブフィットによりジャンカ部で測定した電磁波速度を図３５に示す。低強度部の測定結果よりもばらつきは大きいが、０．５ｃ前後の値となった。先に述べたようにジャンカ部の電磁波速度は約０．６５ｃであるが、図３５はこのような速度の速いジャンカ部を伝播経路中の一部に含む場合の測定結果である。このため得られた結果は、概ね妥当な値であると判断される。この図３５のジャンカ部の結果と、図３４の低強度部及び図３３の健全部の結果とあわせて考えると、この方法により欠陥種別を判定することが可能である。
【００８４】
このように、電磁波速度解析から検討した結果、健全部と低強度部の電磁波速度は約０．３４ｃであるのに対し、ジャンカ部のそれは約０．６５ｃであった。これらの電磁波速度は材齢８週までの範囲で大きく変化せず、また周波数依存性も認められなかった。支保工信号の軌跡から電磁波速度を求めることが可能で、この結果からジャンカと低強度の判定を行える可能性が得られた。また、空洞部分で支保工信号が消失することから空洞の判定の可能性も得られた。さらに、上述の速度に関する考察結果より、図３６に示す如き欠陥からの反射波が得られたにも拘わらず、図３６の波線円で示す如き信号があらわれたときは、低強度部との判断になる。
【００８５】
電磁波平均速度を用いて図３７（ｃ）に示す欠陥部ｘの厚みを求める方法について説明する。図３７（ａ）、（ｂ）に示すように、コンクリート製の覆工トンネル軸方向に対して間欠的に配置された支保工により補強支持されている。ここで、支保工の深さ：ｄ、健全部の電磁波速度：Ｖ_１、欠陥の電磁波速度：Ｖ_２は既知であるとする。また、検査機のレーダー記録から、健全部における表面から支保工までの時間差：ｔ_１、欠陥部における表面から支保工までの時間差：ｔ_２を読み取る。そして、ｔ，ｖの関係よりｘを求めることができる。
【００８６】
【数５】

【００８７】
【数６】
ｘ＝（（ｔ_２−ｔ_１）／２）・（（ｖ_１・ｖ_２）／（ｖ_１−ｖ_２））
【００８８】
次に、反射波のスペクトル解析に関する第四実施形態について説明する。
内部欠陥で反射された電磁波の波形は、減衰や反射等の周波数依存性の影響により生じる波形分散等の欠陥の反射特性を反映しており、波形解析することで欠陥種類を識別できる可能性がある。これらの影響は反射波の周波数スペクトルの差異として検出できることが期待される。これを明らかにするために、３種類のアンテナのうち最もＳ／Ｎの良かった４００ＭＨｚのデータを用いて、スペクトル解析を実施し、検討した。
【００８９】
図３８に材齢８週目におけるかぶり１０ｃｍの空洞の元データ，背景除去処理，背景除去処理＋ゲイン補正したレーダーのイメージ像を示す。欠陥のない健全部分のデータを用いた背景除去処理とゲイン補正をすることで、計測条件によらず歪みのない欠陥波形の抽出を行った。
【００９０】
図３９に背景除去処理により欠陥波形を抽出した波形例（空洞：材齢８週，かぶり１０ｃｍ，厚さ５ｃｍ）を示す。４００ ＭＨｚアンテナの場合、かぶりが浅いとき特にコンクリートの表面から受信される反射波との重なりの影響を受けていることが分かった。欠陥波形を欠陥からの反射波以外の信号と分離することにより、欠陥のみの情報を抽出することができる。
【００９１】
スペクトル解析を行う際、各々の欠陥反射波が出現している範囲で空間平均し、反射波形のＳ／Ｎを向上させた。解析ゲートは、図３９（ｃ）に示すような欠陥反射波の立ち上がり近傍からゲートを設定した。スペクトル解析は、欠陥反射波のフーリエ変換後にパワースペクトルを求めた。
【００９２】
ここに、パワースペクトル計算結果を示す。図４０〜４２にかぶり１０，２０及び３０ｃｍで８週目に計測した厚さ５，１０及び２０ｃｍの欠陥反射波のパワースペクトル計算結果を示す。また、図４３〜４５は図４０に示すパワースペクトルを欠陥厚さ５ｃｍ、１０ｃｍ、２０ｃｍの各々の場合につき、空中直接波のデータ、空洞模型にて得られたデータ、ジャンカ模型にて得られたデータ毎に分離させて表示したものである。図４０〜４２には、計測時に別途採取した空中直接波のスペクトルも示す。この空中直接波のデータは、送受信アンテナを直接伝わった信号であり、アンテナ個体の特性と考えることができる。第一実施形態に示されたように、空洞、ジャンカの反射波は明瞭に捕らえられていたが、低強度の反射波はかぶり１０ｃｍ、厚さ５ｃｍの条件以外はＳ／Ｎが低く検出されなかった。したがって、本スペクトル解析では、空洞とジャンカの種別判定について検討した。
【００９３】
図４０〜４２から、空洞反射波のスペクトルは、かぶり１０ｃｍにおいて、空中直接波のスペクトルにほぼ類似していることが分かった。かぶりが２０，３０ｃｍと深くなるにつれて、空中直接波のスペクトルからの歪み度合いが大きくなり、かぶり３０ｃｍになると低周波数側にスペクトルが移動していることが分かった。これは、コンクリート中の減衰の周波数依存性による波形分散等の影響が考えられる。欠陥厚さの差異は、顕著に認められなかった。このように、かぶりが大きくなることにより、空洞反射波のスペクトル形状が若干歪むものの、空中直接波のスペクトルの形状に類似していると思われる。
【００９４】
一方、ジャンカ反射波のスペクトルは、空中直接波のスペクトルに対して歪み度合いが大きいことが分かった。空中直接波のスペクトルを基準とした場合、高周波数側に分布しているものもあれば、低周波数側に分布しているものもあり、一貫した傾向が認められなかった。この原因を調べるために、スペクトル解析前の欠陥波形（かぶり１０ｃｍ）を見直した。図４６，４７に欠陥波形を示す。
【００９５】
空洞は、厚さが変わっても振幅が異なるのみでほぼ同様な形状を示していた。ジャンカは、図４５に示すように、厚さが大きくなるにつれて波形が広がった形状になることが分かった。これは、空洞に比べて、ジャンカの電磁波速度が小さいため、ジャンカの背面信号が前面信号の後方に出現するため前面信号と背面信号が重なった影響と考えられる。すなわち、空洞反射波のスペクトルは背面信号の影響は少ないが、ジャンカ反射波のスペクトルは、電磁波速度が小さいため、背面信号の影響が大きくなったため歪み度合いが大きくなったと推察される。また、かぶりが大きくなることによるコンクリート中の減衰の影響，空洞反射波に比べてジャンカ反射波のＳ／Ｎの低下によるコンクリート中の散乱ノイズの影響等がジャンカ反射波のスペクトルを大きく歪ませたものと考えられる
【００９６】
以上の結果から、空中直接波のスペクトルを基準とすると、空洞反射波のスペクトルは、ジャンカ反射波のスペクトルに対して、歪み度合いが少なく、空中直接波のスペクトルに類似していることが分かった。
【００９７】
パワースペクトル計算結果から、空洞反射波のスペクトルに比べてジャンカ反射波のスペクトルは、空中直接波のスペクトルに対して歪むことが分かった。空洞反射波とジャンカ反射波のスペクトルの差異は分かったが、これを判別するには定量化を行う必要がある。推定したパワースペクトル形状を定量化するため、式（Ａ５），（Ａ６）から式（Ａ１）〜（Ａ４）に示す４つの周波数特徴量（形状パラメータ）を計算した。式（Ａ５），（Ａ６）は、それぞれ原点周りのｎ次積率（モーメント）、算術平均周りのｎ次積率（モーメント）である。
【００９８】
４つの特徴量は、パワースペクトルのような分布の形状（特性）を表すパラメータであり、このように数値化することにより、定量化が可能となる。
【００９９】
【数７】

【０１００】
パワースペクトルの結果から、空中直接波のスペクトルからの歪み度合いが、空洞とジャンカで異なることが分かったので、各欠陥について求めたこれら４つの特徴量は、空中直接波のスペクトルから求めた特徴量との絶対値差とした。
【０１０１】
このかぶり１０ｃｍのデータを図４８に示す。このように特徴量を絶対値差に変換することで、かぶり１０ｃｍにおいては、ジャンカの方が空洞に比べて大きな値となり、識別の可能性が得られた。かぶり２０及び３０ｃｍにおいても、かぶり１０ｃｍほど分離性は良くなかったがほぼ同様な傾向を示した。
【０１０２】
以上の結果から、特徴量を求めることにより、定量的な種別判定の可能性を見出せた。
【０１０３】
特徴量を求めて種別判定の可能性が得られた。これを種別判定するためには、客観的な判断が必要になる。したがって、４つの特徴量と材齢を用いた５次元の多変量解析を検討した。空洞とジャンカを識別するために、それぞれ本解析で求めた特徴量のマハラノビス距離を求めて、判別分析を行った。マハラノビス距離とは、各欠陥特徴量空間の重心までの距離（ユークリッド距離）を各欠陥特徴量のばらつきを考慮した距離である。マハラノビス距離は１次元（１変数）で考えると、ユークリッド距離を標準偏差で割った値の２乗と定義することができる。これを多次元（多変量）で考えるとｉ番目サンプルの両群の重心までのマハラノビス距離は以下の式で求められる。
【０１０４】
【数８】

【０１０５】
表５〜７にかぶり１０，２０及び３０ｃｍの判別分析した結果を示す。判別的中率は、どのかぶりにおいても約８０％以上の結果を示しており良好であると思われる。今回の解析からは、ジャンカを空洞として誤判別することが多かった。すなわち、本解析結果からは、ジャンカの判定は正答率が高いものと考えられる。また、今後、特徴量の見直しや実構造物のデータ等によりデータベースを増やすことにより、評価に対する信頼度が向上することが期待される。
【０１０６】
【表５】

【０１０７】
【表６】

【０１０８】
【表７】

【０１０９】
ここで、上記各実施形態に係る手法を総合して相互補完的に用いる形態について説明する。本研究において、コンクリート構造物の非破壊的な地下レーダー探査を用いて、主に欠陥種別判定を対象に欠陥評価について検討した。これらをまとめたものを表８に示す。
【０１１０】
【表８】

【０１１１】
上記各実施形態で行った解析手法は欠陥の種々の特性を反映しており、解析手法ごとに影響される因子が異なった。また各解析アプローチの仕方が異なっており、こうした多方面の解析結果から欠陥の評価を行うことにより評価精度が向上することが考えられる。さらに、適用範囲が異なることも考慮すると、実際のコンクリート構造物に適応するには、適応範囲に対応した解析方法を数種類組み合わせて、総合的に判定するのが望ましいと考えられる。
【図面の簡単な説明】
【図１】本発明の実験のために作成したコンクリートモデルの縦断面図である。
【図２】試験体に用いる二種類のコンクリートに対する一軸圧縮強度試験の結果を示すグラフである。
【図３】試験体に用いる二種類のコンクリートの有効間隙率の測定結果を示すグラフである。
【図４】試験体に用いる二種類のコンクリートの超音波速度（Ｐ波速度：Ｖｐ、Ｓ波速度：Ｖｓ）の測定結果を示すグラフである。
【図５】試験体に用いる二種類のコンクリートの含水率の測定結果を示すグラフである。
【図６】検査を行ったコンクリートモデルを示す写真である。
【図７】図６に示すコンクリートモデルの内部欠陥の埋設状況を示す写真である。
【図８】送信アンテナから放射された電磁波が内部欠陥で反射し、受信アンテナに捉えられ、測定器に記録されるまでの経路における振幅変化を示す図である。
【図９】道路トンネルにおいて計測した無欠陥箇所の支保工反射波に関する振幅／増幅率と、往復伝播時間の関係をプロットしたグラフである。
【図１０】１５００ＭＨｚアンテナによりかぶり２０ｃｍ、厚さ５〜２０ｃｍの空洞模型の検査を行った際の階調表示によるレーダーイメージ及び矢印位置における波形を示すグラフである。
【図１１】１５００ＭＨｚアンテナによりかぶり２０ｃｍ、厚さ５〜２０ｃｍのジャンカ模型の検査を行った際の階調表示によるレーダーイメージ及び矢印位置における波形を示すグラフである。
【図１２】９００ＭＨｚアンテナによりかぶり１０ｃｍ、厚さ５〜２０ｃｍの低強度模型の検査を行った際の階調表示によるレーダーイメージ及び矢印位置における波形を示すグラフである。
【図１３】４００ＭＨｚアンテナによりかぶり３０ｃｍ、厚さ５〜２０ｃｍの空洞模型の検査を行った際の階調表示によるレーダーイメージ及び矢印位置における波形を示すグラフである。
【図１４】試験体内の欠陥厚さと反射波波形との関係を示す図である。
【図１５】１５００ＭＨｚアンテナによりかぶり１０ｃｍの空洞模型の検査を行った際の各サンプルトレースにおける測定結果を欠陥厚毎に示すグラフであり、（ａ）は往復伝播時間を、（ｂ）は初動振幅／増幅率をプロットしたものである。
【図１６】９００ＭＨｚアンテナによりかぶり１０ｃｍの空洞模型の検査を行った際の各サンプルトレースにおける測定結果を欠陥厚毎に示すグラフであり、（ａ）は往復伝播時間を、（ｂ）は初動振幅／増幅率をプロットしたものである。
【図１７】４００ＭＨｚアンテナによりかぶり１０ｃｍの空洞模型の検査を行った際の各サンプルトレースにおける測定結果を欠陥厚毎に示すグラフであり、（ａ）は往復伝播時間を、（ｂ）は初動振幅／増幅率をプロットしたものである。
【図１８】１５００ＭＨｚアンテナにより材齢１週の空洞模型及びジャンカ模型の検査により得られた、初動振幅／増幅率と往復伝播時間の関係を示すグラフである。
【図１９】９００ＭＨｚアンテナにより材齢１週の空洞模型、ジャンカ模型及び低強度模型の検査により得られた、初動振幅／増幅率と往復伝播時間の関係を示すグラフである。
【図２０】４００ＭＨｚアンテナにより材齢１週の空洞模型、ジャンカ模型及び低強度模型の検査により得られた、初動振幅／増幅率と往復伝播時間の関係を示すグラフである。
【図２１】９００ＭＨｚアンテナの記録より求めた減衰率と試験体の材齢との関係を示すグラフである。
【図２２】９００ＭＨｚアンテナの記録より求めた初期振幅×反射率の値と、試験体の材齢との関係を示すグラフである。
【図２３】粗骨材／（粗骨材＋細骨材＋セメント）と電磁波速度関係の関係を示すグラフである。
【図２４】空洞を有するコンクリート試験体に対する検査により得られた階調表示によるレーダーイメージであって、（ａ）は元データを、（ｂ）は背景除去とゲイン補正をしたデータをそれぞれ表示したものである。
【図２５】ジャンカを有するコンクリート試験体に対する検査により得られた階調表示によるレーダーイメージであって、（ａ）は元データを、（ｂ）は背景除去とゲイン補正をしたデータをそれぞれ表示したものである。
【図２６】本発明の第二実施形態における受信波形及び波形処理の結果を示すグラフであって、（ａ）は空洞部を有する試験体を用いた場合の受信波形、（ｂ）は健全な試験体を用いた場合の受信波形、（ｃ）は（ａ）の受信波形からバックグラウンドを除去した波形、（ｄ）は（ｃ）の波形を整流して得られた波形、（ｅ）は（ｄ）の波形をヒルベルト変換して得られる包絡線波形（信号）にそれぞれ相当する。
【図２７】支保工を有する試験体に対する検査により得られた階調表示によるレーダーイメージを示す図である。
【図２８】１５００ＭＨｚのアンテナを用いてかぶり１０ｃｍの試験体に欠陥厚５ｃｍ、１０ｃｍ、２０ｃｍの欠陥を設けた場合にそれぞれ得られた包絡線信号であり、（ａ）材齢２週目の空洞模型、（ｂ）は材齢４週目の空洞模型、（ｃ）材齢８週目の空洞模型、（ｄ）は材齢２週目のジャンカ模型、（ｅ）は材齢４週目のジャンカ模型、（ｆ）は材齢８週目のジャンカ模型を用いた場合にそれぞれ相当する。
【図２９】９００ＭＨｚのアンテナを用いて材齢８週目の試験体に欠陥厚５ｃｍ、１０ｃｍ、２０ｃｍの欠陥を設けた場合にそれぞれ得られた包絡線信号であり、（ａ）はかぶり１０ｃｍの空洞模型、（ｂ）はかぶり２０ｃｍの空洞模型、（ｃ）はかぶり３０ｃｍの空洞模型、（ｄ）はかぶり１０ｃｍのジャンカ模型、（ｅ）はかぶり２０ｃｍのジャンカ模型、（ｆ）はかぶり３０ｃｍのジャンカ模型を用いた場合にそれぞれ相当する。
【図３０】４００ＭＨｚのアンテナを用いて材齢８週目の試験体に欠陥厚５ｃｍ、１０ｃｍ、２０ｃｍの欠陥を設けた場合にそれぞれ得られた包絡線信号であり、（ａ）はかぶり１０ｃｍの空洞模型、（ｂ）はかぶり２０ｃｍの空洞模型、（ｃ）はかぶり３０ｃｍの空洞模型、（ｄ）はかぶり１０ｃｍのジャンカ模型、（ｅ）はかぶり２０ｃｍのジャンカ模型、（ｆ）はかぶり３０ｃｍのジャンカ模型を用いた場合にそれぞれ相当する。
【図３１】アンテナと支保工の位置関係の概略図である。
【図３２】支保工信号の軌跡を示す概略図である。
【図３３】本発明の第三実施形態にかかる手法（カーブフィット）により、健全なコンクリート試験体中の電磁波速度と空気中の電磁波速度の比を材齢毎にプロットした図である。
【図３４】本発明の第三実施形態にかかる手法（カーブフィット）により、低強度部を有するコンクリート試験体中の電磁波速度と空気中の電磁波速度の比を材齢毎にプロットした図である。
【図３５】本発明の第三実施形態にかかる手法（カーブフィット）により、ジャンカを有するコンクリート試験体中の電磁波速度と空気中の電磁波速度の比を材齢毎にプロットした図である。
【図３６】低強度部の欠陥信号を示す階調表示によるレーダーイメージである。
【図３７】（ａ）トンネルの長手軸方向からみた断面図、（ｂ）は（ａ）のＡ−Ａ断面図、（ｃ）は（ａ）における欠陥部の介在状況を示す図である。
【図３８】欠陥信号を示す階調表示によるレーダーイメージであり、（ａ）は元データ、（ｂ）は背景除去を行ったイメージ、（ｃ）はゲイン補正を行ったイメージである。
【図３９】（ａ）は健全な試験体の受信波形、（ｂ）は空洞部を有する試験体の受信波形、（ｃ）は（ｂ）の受信波形に対して背景除去処理及びゲイン補正を施して得られた波形である。
【図４０】かぶり１０ｃｍの試験体を用いて測定した厚さ５ｃｍ，１０ｃｍ，２０ｃｍの欠陥のパワースペクトルを示し、ａｉｒは空中直接波、ｖｏｉｄは空洞、ｊａｎｋａはジャンカ、ｌｏｗは低強度部をそれぞれ示す。
【図４１】かぶり２０ｃｍの試験体を用いて測定した厚さ５ｃｍ，１０ｃｍ，２０ｃｍの欠陥のパワースペクトルを示し、ａｉｒは空中直接波、ｖｏｉｄは空洞、ｊａｎｋａはジャンカ、ｌｏｗは低強度部をそれぞれ示す。
【図４２】かぶり３０ｃｍの試験体を用いて測定した厚さ５ｃｍ，１０ｃｍ，２０ｃｍの欠陥のパワースペクトルを示し、ａｉｒは空中直接波、ｖｏｉｄは空洞、ｊａｎｋａはジャンカ、ｌｏｗは低強度部をそれぞれ示す。
【図４３】かぶり１０ｃｍ、欠陥厚さ５ｃｍの試験体を用いて測定した空中直接波、空洞、ジャンカのパワースペクトルを示す。
【図４４】かぶり１０ｃｍ、欠陥厚さ１０ｃｍの試験体を用いて測定した空中直接波、空洞、ジャンカのパワースペクトルを示す。
【図４５】かぶり１０ｃｍ、欠陥厚さ２０ｃｍの試験体を用いて測定した空中直接波、空洞、ジャンカのパワースペクトルを示す。
【図４６】空洞を有する試験体の検査により得られた波形を示す図であって、（ａ）は空洞の厚さが５ｃｍの場合、（ｂ）は空洞の厚さが１０ｃｍの場合、（ｃ）は空洞の厚さが２０ｃｍの場合にそれぞれ相当する。
【図４７】ジャンカを有する試験体の検査により得られた波形を示す図であって、（ａ）はジャンカの厚さが５ｃｍの場合、（ｂ）はジャンカの厚さが１０ｃｍの場合、（ｃ）はジャンカの厚さが２０ｃｍの場合にそれぞれ相当する。
【図４８】４つの周波数特徴量に対する試験体種別毎の材齢に対する分布を示すグラフであり、（ａ）は基準化平均周波数、（ｂ）は基準化標準偏差周波数、（ｃ）は基準化歪度、（ｄ）は基準化尖度の分布を示す。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a concrete inspection method using electromagnetic waves and a concrete inspection device using electromagnetic waves. More specifically, the present invention relates to a concrete inspection method using electromagnetic waves for inspecting the inside of the concrete structure by transmitting electromagnetic waves to a concrete structure and receiving reflected waves, and a concrete inspection apparatus using electromagnetic waves used therefor.
[0002]
[Prior art]
In recent years, the demand for safety of concrete structures such as tunnels has increased, and various nondestructive inspection methods have been studied. In evaluating the soundness of concrete structures, it is important to detect not only defects such as cracks and jumpers appearing on the surface, but also defects such as cavities hidden inside the surface, jumpers and low strength. it is conceivable that.
[0003]
Examples of methods for detecting internal defects include sounding methods, infrared methods, ultrasonic methods, and underground radar exploration, which is an electromagnetic wave inspection. Among these, the electromagnetic wave inspection has a number of advantages because it can perform continuous measurement along the survey line quickly and quantitatively, and can cover parts with a depth of cover as compared to other methods. Inspection method. As this type of electromagnetic wave inspection, the one described in the following Patent Document 1 is known.
[0004]
[Patent Document 1]
JP 2002-257744 A
[0005]
According to this document, an image device and an infrared detector are used together with an electromagnetic wave radar to detect the soundness of concrete such as a tunnel without contact.
[0006]
However, this document only detects the presence of a back cavity in tunnel concrete, and attempts to improve the reliability of inspection by forming a complementary relationship with other inspection items. Therefore, it has not been possible to make a more clear qualitative or quantitative judgment on the type of defect part, the position and scale of the defect part, and the like.
[0007]
[Problems to be solved by the invention]
In view of such a conventional situation, an object of the present invention is to make a more detailed determination of the type, depth, thickness, and the like of a defective portion in a concrete structure than before.
[0008]
[Means for Solving the Problems]
In order to achieve the above object, the feature of the concrete inspection method using electromagnetic waves according to the present invention is that a concrete inspection method using electromagnetic waves inspects the inside of the concrete structure by transmitting electromagnetic waves to the concrete structure and receiving reflected waves. The A / G and the round-trip propagation time t regarding the initial peak of the internal defect reflected wave and the relationship between the following equation (1) and A₀・ R is obtained and this A₀It is to determine the type of the defective part based on r.
[0009]
ln (A / G) = − (α · V · t) + ln (A₀・ R) (1)
Where A is the amplitude recorded in the measuring instrument, G is the amplification factor, α is the attenuation factor of the amplitude in the concrete, V is the electromagnetic wave velocity, A₀  Is the initial amplitude of the electromagnetic wave transmitted from the transmitting antenna on the concrete surface, and r is the reflectance of the electromagnetic wave at the internal defect. A₀  As long as the condition of the concrete surface does not change, it is almost constant for each type of antenna used.
[0010]
Specifically, by measuring the A / G and the round-trip propagation time t at a plurality of locations having different depths such as defects, A can be obtained from the above equation (1) by the least square method.₀Find r and α.
[0011]
In addition to the above characteristics, the depth of the target portion may be obtained from the measured round-trip propagation time of the initial peak of the reflected wave and the electromagnetic wave velocity inside the concrete structure. That is, according to the feature, it is possible to specifically obtain a quantitative value such as the depth of a target portion such as a support work or a defective portion. In addition, what is necessary is just to obtain | require the electromagnetic wave velocity inside a concrete structure using a wide angle method etc.
[0012]
Moreover, it is desirable to set the frequency of the electromagnetic wave so that the half wavelength of the electromagnetic wave is smaller than the expected thickness of the defect portion.
[0013]
On the other hand, in the electromagnetic wave method of inspecting the inside of the concrete structure by transmitting the electromagnetic wave to the concrete structure and receiving the reflected wave, in order to more specifically determine the type of the defect portion, the measured defect It is possible to determine whether or not the defective portion is a jumper portion based on the presence or absence of a back signal having another peak with respect to the main signal having a peak in the processed signal that has been subjected to envelope processing while removing the healthy portion signal from the signal desirable.
[0014]
And the thickness of the said jumper part can be calculated | required by the difference of the peak of the said main signal, and the peak of a back signal.
[0015]
Further, empirically, it is possible to determine whether or not the defective portion is a hollow portion by a slight front repeating signal generated after the main signal.
[0016]
More specifically, the processing signal is rectified, and the envelope processing is preferably performed by analytical signal processing. For the envelope processing, various processing can be used as appropriate, including Hilbert transform. These processes are not limited to software processes, and can be realized by hardware processes using a rectifier circuit or a low-pass filter.
[0017]
Another configuration of the present invention is a method for inspecting the inside of a concrete structure by transmitting electromagnetic waves to the concrete structure and receiving reflected waves, and the curvature of the flat reference body photographed in a curved manner Using a plurality of coordinates of each point in the image, the depth of the flat reference body and the average electromagnetic wave velocity in the concrete can be obtained, and it can be determined from this average electromagnetic wave velocity whether the defective part is a junker part or not. .
[0018]
And the thickness of the said jumper part is computable from the variation | change_quantity of the said average propagation speed. In addition, the presence of the cavity can be determined based on the disappearance of the curved image.
[0019]
When the average electromagnetic wave velocity is substantially the same as that of the healthy part and a reflected wave appears, it can be determined that the low intensity part exists.
[0020]
Specifically, the flat reference body is a tunnel support. Further, the depth of the flat reference body and the average electromagnetic wave velocity are obtained by a least square method.
[0021]
Still another configuration of the present invention is a method for inspecting the inside of a concrete structure by transmitting electromagnetic waves to the concrete structure and receiving reflected waves, and removes a healthy part signal from the measured defect signal. At the same time, the power spectrum of the signal is obtained, and the type of the defect portion is determined by approximating the obtained power spectrum shape and the cavity power spectrum shape and the jumper power spectrum shape obtained in advance. The approximation determination is performed by obtaining the Mahalanobis distance of the measured reflected wave for a plurality of parameters obtained from the power spectrum. In this case, it is desirable that the parameter has a concrete age. The parameter is the average frequency ν of the power spectrum.₁, Standard deviation frequency (μ₂)^1/2, Skewness μ₃/ (Μ₂)^3/2And / or kurtosis μ₄/ (Μ₂)²-3 may be included.
[0022]
The above methods can also be combined with each other. For example, you may make it implement both the methods in any one of Claims 1-8, or in any one of Claims 9-18. This is because the former has a correlation with the thickness of the defect portion in the frequency, and the latter has no restriction due to the correlation.
[0023]
In addition, a method selected from at least two of the four groups according to any one of claims 1 to 4, claim 5 to 8, claim 9 to 14, or claim 15 to 18 is performed. May be. As a result, the type determination of the defective portion is more reliable.
[0024]
And in the two or more inspection methods implemented, you may vary the frequency of the electromagnetic waves used. This is because by making the frequency different, it is possible to use different characteristics for each examination.
[0025]
On the other hand, the concrete inspection apparatus according to the present invention is characterized by including a transmission antenna for transmitting electromagnetic waves, a reception antenna for receiving reflected electromagnetic waves from a concrete structure, and a measuring instrument for measuring the received waves. It is in.
[0026]
【The invention's effect】
As described above, according to the characteristics of the concrete inspection method using electromagnetic waves according to the present invention, the concrete structure can be obtained by skillfully using the attenuation characteristics of the electromagnetic waves, the speed difference of the defect portion, the flat reference body, and the shape characteristics of the power spectrum. It is possible to make a more detailed determination as to the type, depth, thickness, and the like of the defective portion in the inner portion than before. In addition, by combining multiple methods in a mutually complementary manner, it has become possible to further increase the judgment system.
[0027]
Other objects, configurations and effects of the present invention will become apparent in the section of the embodiment of the present invention shown below.
[0028]
DETAILED DESCRIPTION OF THE INVENTION
Next, the present invention will be described in more detail with reference to the accompanying drawings.
First, the concrete model created for the experiment is a model of tunnel lining, as shown in Fig. 1. In normal concrete (N-18, JIS standard) with a thickness of 40 cm on H-shaped steel, Three types of internal defects, ie, cavities, mortar-deficient junkers, and low strength due to excessive moisture, were arranged with three different thicknesses of 5 cm, 10 cm, and 20 cm, and three different foggings of 10 cm, 20 cm, and 30 cm.
[0029]
The composition of normal strength concrete, low strength concrete and junker is shown in Table 1, and the results of laboratory tests conducted by preparing specimens with φ100 mm and h = 200 mm for normal strength and low strength are shown in FIGS. Thus, it has been confirmed that there is a significant difference between the normal strength and the low strength in terms of uniaxial compressive strength, ultrasonic velocity, and the like. For the junkers, four types of specimens with different formulations were prepared and measured with an antenna having a center frequency of 1500 MHz, and the final formulation was determined from the appearance of reflected waves. The electromagnetic wave velocity at that time was 15 cm / ns. Further, foamed polystyrene having an electromagnetic wave velocity of 27 cm / ns was used for the cavity. FIG. 6 shows the created concrete model, and FIG. 7 shows the situation of the defect model embedded inside.
[0030]
[Table 1]

[0031]
For the measurement, a measuring device called trade name SIR-2 manufactured by GSSI (Geophysical Survey Systems, Inc.) was used. Dipole antennas are used as antennas, and those having 1500 MHz, 900 MHz, and 400 MHz are used because the penetration depth and resolution differ depending on the center frequency. The measurement was performed when 1, 2, 3, 4, 6, 8, and 10 weeks passed after placing the concrete above the defect. At that time, the survey line was set so that the antenna passed through the center of the internal defect. Table 2 shows the main measurement parameters.
[0032]
[Table 2]

[0033]
The waveform data to be acquired is recorded on the SIR-2 hard disk, but since the data length is 16 bits, the recording amplitude is in the range of -32768 to 32767. Although these are values after amplification, the amplitude before amplification can be obtained by dividing by the amplification factor read from the header file later.
[0034]
As shown in FIG. 8, the amplitude of the incident wave to the concrete transmitted from the transmitting antenna T (initial amplitude) until the electromagnetic waves radiated from the transmitting antenna T are reflected by internal defects and captured by the receiving antenna R is shown. ) A₀, The incident wave amplitude to the internal defect is A₁, The amplitude of the reflected wave from the internal defect A₂, The received amplitude at the receiving antenna T is A₃Then, these relations are expressed by the following equations.
[0035]
[Expression 1]

[0036]
Where A is the amplitude recorded in the measuring instrument (A₃• Given in G). G is the amplification factor, (A₂/ A₁) = R is the reflectivity of internal defects, α is the decay rate in concrete, Z (= V · t, V: electromagnetic wave velocity, t: round-trip propagation time) is the propagation distance, ZⁿRepresents attenuation due to geometrical divergence. ZⁿFig. 9 shows the relationship between the amplitude / amplification factor and the round-trip propagation time for the initial motion of the support reflected wave in the road tunnel where there is no abnormal reflected wave in the concrete. From this, it can be seen that the relationship is almost linear, and n≈0 holds. Substituting these into the [Formula 1] formula and taking the logarithm gives the formula:
[0037]
[Expression 2]

[0038]
If V is measured by a wide angle method or the like, α is obtained from the gradient of ln (A / G) to t over a plurality of depths, and A is obtained from the y intercept.₀R can be obtained. A₀Is substantially constant for each antenna as long as the surface state does not change, so that the reflectance can be evaluated. Here, the wide angle method is a method in which the center position of the transmission / reception antenna is fixed and the transmission / reception antenna interval is changed. For example, in FIG. 8, the interval between the transmission antenna T and the reception antenna R is temporarily separated. In this method, A / G and t relating to the initial peak of the internal defect reflected wave are handled because it is easy to identify and hardly affected by other reflected waves such as the back reflected wave. .
[0039]
About the detection result of the internal defect for every antenna, the detection example of the internal defect in material age 1 week is shown in FIGS. The left side of these is a line scan display record after performing a background removal process of subtracting the average waveform of the defect-free portion from the measurement record. On the right side, the reading position of the amplitude A and the round-trip propagation time t for the initial movement of the defect waveform in the trace indicated by ↓ on the left side is shown.
[0040]
In general, the attenuation of electromagnetic waves in concrete varies depending on the center frequency of the antenna to be handled and mainly the specific resistance of concrete. The higher the frequency and the lower the specific resistance, the greater the attenuation. For this reason, the maximum fog of a detectable defect changes with the antenna to be used and the age of concrete. In concrete, it is known that the specific resistance increases as the age of the material elapses, and as the age of the material decreases, the specific resistance decreases and the attenuation increases. Therefore, it is difficult to detect internal defects in deep fog.
[0041]
Measurements were taken over the age of 1 to 10 weeks. Hereinafter, the detection result for each antenna will be described.
[0042]
1) 1500MHz antenna
The cavities were detectable up to a cover of 20 cm at the age of 1 week and up to 30 cm after the 2nd week. Junka was detectable up to 20 cm in cover from 1 week of age, but 30 cm in cover was not detected even after 10 weeks. The low intensity seemed to be able to detect only 10 cm of fog, but the irregular scattered wave seen by the aggregate overlapped and the reflecting surface was unclear.
[0043]
2) 900MHz antenna
About the cavity and the jumper, it was detectable from the material age 1 week to the cover 30 cm. Low intensity was detected only from 10 weeks of age, although the cover was only 10 cm.
[0044]
3) 400MHz antenna
The detection situation was almost the same as that of the 900 MHz antenna.
[0045]
Thus, if it is the range to cover 30 cm, a cavity and a jumper can be detected from the age of 1 week by using a some antenna. On the other hand, the low intensity can be detected only from a material age of 1 week when the cover is shallow, and when the cover is 20 cm or more, it is difficult to detect regardless of the material age. This is because the difference between the relative permittivity or the electromagnetic wave velocity between the cavity and the jumper is large, rather than the attenuation rate, and the reflectivity representing the amplitude ratio of the reflected wave with respect to the incident wave is large. The intensity is small and the reflected wave is weak.
[0046]
Next, we will consider the effect of the thickness of internal defects. In this method, the attenuation rate and the initial amplitude × reflectance are obtained from the relationship between the amplitude / amplification factor and the round-trip propagation time for the initial motion of the reflected wave from the internal defect. In this case, as shown in FIG. 14, since the reflected wave is a composite wave of the front reflected wave and the back reflected wave of the defect, the reflected wave is disturbed due to the relationship between the thickness of the defect and the wavelength of the electromagnetic wave used, and the initial motion changes. It is possible to do. Therefore, the defect thickness from which the initial reflected wave does not change for each antenna was examined from the measurement record.
[0047]
15 to 17 show the round trip propagation time of 60 traces in the center of the defect and the initial amplitude / amplification factor of the reflected wave obtained when the cavity model having a cover of 10 cm is inspected with 1500 MHz, 900 MHz, and 400 MHz antennas in order. Are plotted. From these, it can be seen that the round-trip propagation time is almost constant for each antenna regardless of the thickness of the defect, but the initial motion amplitude is affected by the thickness of the defect. That is, at 1500 MHz where the half-wavelength of the received waveform is about 4 cm, the amplitude does not change much regardless of whether the defect thickness is 5 cm, 10 cm, or 20 cm.
[0048]
However, at 900 MHz with a half wavelength of about 8 cm, 10 cm and 20 cm are almost the same, but at 5 cm, the frequency is nearly 50% smaller. Since 400 MHz has a half wavelength of about 13 cm, the amplitude is different at any thickness. Therefore, as expected, the initial motion amplitude is not affected by the thickness of the defect only when it has about half a wavelength or more. However, with a jumper, the electromagnetic wave velocity is slower than that of the cavity and the wavelength becomes shorter, so that even thinner ones can be applied.
[0049]
Next, the attenuation rate, initial amplitude × reflectance value and type determination will be considered. The relationship of ln (A / G) to t at 1 week of age using only 20 cm at 400 MHz, using only 20 cm at 400 MHz, using all defect thicknesses 5, 10, and 20 cm at 1500 MHz is shown in FIGS. Shown in From this, it can be seen that A / G differs depending on the type of defect, and increases in the order of low strength, jumper, and cavity, and is divided into groups with the same slope and different y-intercept for each type. The slope is the decay rate of the electromagnetic wave amplitude in the concrete and is consistent with being independent of the type of internal defect. On the other hand, the y-intercept represents initial amplitude × reflectance, and the reflectivity has the following relationship. Where ε₁, V₁Is the relative permittivity and electromagnetic wave velocity of normal strength concrete, ε₂, V₂Is the relative permittivity and electromagnetic wave velocity of internal defects.
[0050]
[Equation 3]

[0051]
Where V₁= 9 cm / ns, V in the cavity₂= 27cm / ns, V in Junka₂= 15 cm / ns, the reflectivity should be 0.51 and 0.25 for the cavity and the jumper, but the calculated y-intercept values are close to those at 900 MHz and 400 MHz. Next, FIGS. 21 and 22 show changes in attenuation rate and initial amplitude × reflectance obtained from the 900 MHz antenna depending on the material age. Thus, the attenuation rate is almost the same between the cavity and the jumper, and tends to decrease until 6 weeks and then settle to a constant value. On the other hand, since the initial amplitude changes depending on the moisture condition of the surface on the measurement date, the initial amplitude × reflectance apparently changes greatly depending on the material age. However, the ratio between the cavity and the jumper changes at a value close to 1: 0.5 regardless of the age. Therefore, even if the amplitude may change due to a change in the attenuation factor or a change in the initial amplitude, there is a nearly double difference between the ln (A / G) and t diagrams in the cavity and the jumper, and they can be identified. As described above, there is an internal defect having a composition similar to a hollow and created jumper, and the type can be determined from the age of 1 week if the thickness is a half wavelength or more. However, it seems that low strength does not seem to have a difference in electromagnetic wave velocity from that of normal strength concrete, and it is difficult to detect it at present, and it is difficult to determine the type based on initial amplitude × reflectance r.
[0052]
Regarding other application conditions, if the wavelength is less than half a wavelength, it has not been numerically studied. However, as the thickness is reduced, the type determination becomes difficult. However, regarding the blending of janka, it can be determined even if it changes to some extent as follows.
[0053]
Table 3 shows the electromagnetic wave velocities obtained by using the mixture of the four types of jumpers prepared in the preliminary experiment and the 1500 MHz antenna, and FIG. 23 shows the relationship between the electromagnetic wave velocity and the weight ratio of the coarse aggregate to the total weight excluding water. Formula 2 was adopted for the model, but it can be seen that the electromagnetic wave velocity changes depending on the proportion of coarse aggregate. The ratio when the electromagnetic wave velocity of the normal strength concrete is 9.5 cm / ns and the reflectance of Formulation 2 is 1 is obtained from Formula 3 above, and Formulation 1, Formulation 2, Formulation 3, Formulation 4 In the cavity, it becomes 0.65, 1, 0.31, 1.41, and 2.22. Since there is a difference from the cavity in terms of type determination, it is easier to determine the combination 1 and the combination 3 than the combination 2. Moreover, although the composition 4 of 100% of coarse aggregate is close to the reflectance of the cavity, it can still be judged from the reflectance of about 60%.
[0054]
[Table 3]

[0055]
Next, a second embodiment of the present invention will be described. In the present embodiment, the type of the defective portion can be determined by comparing the main signal and the back signal.
[0056]
First, consider scattering source and reflection source analysis. An electromagnetic wave is observed as a scattered wave and a reflected wave by being scattered by a defect inside the concrete or reflected by a boundary surface. It is considered that the generation state of the scattered wave and the reflected wave varies depending on the type of defect.
[0057]
In the original data, there are reflected waves on the concrete surface and background noise, which may affect the defect waveform. Therefore, background removal was performed using data of a portion having no defect. As a result, it is considered that only the defect waveform can be extracted and the influence of unnecessary signals is reduced. In radar measurement, in order to be able to confirm signals at deep positions, data is collected by increasing the gain as the depth increases. This distorts the radar waveform, so the gain increment at the time of data collection was corrected during analysis (hereinafter, the original data divided by the amplification factor will be referred to as gain correction). FIGS. 24 and 25 show the radar image image obtained by using the 1500 MHz antenna and the original data of the cavity and the jumper with a cover of 10 cm and the background removal processing and gain correction at the age of 8 weeks.
[0058]
In order to investigate the characteristics of the defect from the signal intensity and shape of the scattered wave and the reflected wave, the defect waveform was extracted and the envelope signal was obtained by Hilbert transform and compared.
[0059]
This Hilbert transform is defined by the following equation.
[0060]
y (t) = x (t) * h (t)
Here, * is a convolution operation, and h (t) = 1 / (πt).
[0061]
FIG. 26 shows an example of a result obtained by extracting a defect waveform by background removal processing after gain correction and converting it into an envelope signal. (A) shows a defect waveform obtained when a test body having a cavity (a cover of 10 cm and a defect thickness of 5 cm) is used, and (b) shows a healthy part waveform without defects. Since both of the waveforms in the first half (near 1 ns) have substantially the same shape, the S / N of the defect signal is improved by using the waveform of the sound part in the background removal process. The result of the background removal process is shown in (c). (D) shows the rectified signal obtained by converting the waveform with the background removed into an absolute value, and (e) shows the envelope signal. The envelope signal was calculated by the Hilbert transform. The envelope signal was averaged in order to obtain the scattering and reflection intensities in the range where the defect signal appeared.
[0062]
Next, electromagnetic wave velocity analysis is performed. Since the dielectric constants of the junkers, low strength, and cavities existing inside the concrete are different, the electromagnetic wave velocity of these defective parts is also different. Therefore, if the electromagnetic wave velocity in the path through which the electromagnetic wave propagates is known, it is possible to determine the type of defect existing in the path. In the concrete model used in this study, a steel support is inserted at a depth of 40 cm from the surface. The electromagnetic wave velocity of each part was measured from the appearance position of the supporting part signal of the defective part and the healthy part. As a result, the electromagnetic wave velocity of the healthy part was estimated to be about 0.34c, where c is the electromagnetic wave velocity in the air. The speed of the low strength portion was 0.34c, the same as that of the healthy portion, and the jumper portion was about 0.65c faster than the healthy portion and the low strength portion. It was also confirmed that these values did not change greatly until the age of 8 weeks, and that they did not depend on the frequency in the frequency range used.
[0063]
FIG. 27 shows an example of a support signal obtained by exploring a 10 cm-covered jumper in the eighth week. If the depth of the support is known, it is possible to calculate the velocity of the electromagnetic wave in the path from the propagation time when the support signal is closest to the surface. That is unknown. Therefore, in this study, the electromagnetic wave velocity was obtained by curve fitting the support signal trajectory using the least square approximation. In addition, since the reflectance of the defect was large in the cavity, the support work signal existing under the cavity could not be detected, and therefore the electromagnetic wave velocity could not be detected. However, if the disappearance of the support signal at such a defective portion can be confirmed, it can be determined that the existing defect is a cavity. For this reason, in this research, it examined for the purpose of the judgment of junker and low intensity. In addition, in order to deal with support work in a deep position, examination was performed with a 400 MHz signal.
[0064]
Here, the scattering source and reflection source analysis results are examined. FIG. 28 shows an envelope signal of a cover of 10 cm with a 1500 MHz antenna at ages 2, 4, and 8 weeks. In the result of 1500 MHz, only a cavity and a jumper are shown because a low-intensity reflected wave could not be confirmed. In FIG. 28, the vertical axis is normalized by the peak of the defect front signal, and the horizontal axis is adjusted so that the time near the peak of the defect front signal is zero.
[0065]
From FIG. 28, the envelope signal in the cavity is almost only the front signal of the defect, whereas in FIG. 29, the envelope signal of the jumper clearly shows the rear signal of the defect at a thickness of 10, 20 cm. . This is because the electromagnetic wave velocity 0.65c of the jumper obtained from the result of the electromagnetic wave velocity analysis described above (light velocity c = 3 × 10⁸m / s). Even in the cavity, no peak could be confirmed at the 2nd and 4th weeks, but a peak estimated to be the back signal of the defect was recognized by a signal having a thickness of 20 cm at the 8th week. Comparing the envelope signal of the 8th week data cavity and the jumper with a thickness of 20 cm, it was found that although the back signal of the defect was recognized even in the cavity, the back signal of the defect in the jumper can be clearly confirmed. The above results indicate that the cavity has a higher electromagnetic wave velocity than the jumper, so that the front signal and the rear signal cannot be separated if the thickness is the same as that produced in this study. It is conceivable that the back signal of the defect is reduced by reducing the amount of electromagnetic waves that are greatly reflected and transmitted on the defect surface. In addition, if it is determined that the defect back signal is not recognized in the cavity at the initial age, it can be estimated that the defect can be assumed to be a jumper if the defect back signal can be clearly confirmed.
[0066]
Furthermore, in the data of the second and fourth weeks of the envelope signal of the cavity, a portion where the signal intensity increased between 2 and 3 ns was recognized. When this is calculated from the electromagnetic wave velocity 0.34c in the concrete obtained from the electromagnetic wave velocity analysis, it corresponds to an interval of about twice the cover depth of 10 cm, and is assumed to be a repetitive signal of the front signal of the defect. Such a repetitive signal is not recognized by the jumper and can be determined as a feature of the cavity.
[0067]
In order to confirm the data of the part where the fog is deep, the results at the age of 8 weeks with the 900, 400 MHz antenna are shown in FIGS.
[0068]
At 900 MHz, the back signal of the defect in the jumper was clearly confirmed at a thickness of 20 cm. At a thickness of 10 cm, the width of the signal was increased due to the overlap of the back signal of the defect and the front signal of the defect, but could not be clearly confirmed. Even in the result of 400 MHz, the back signal of the defect in the jumper was confirmed only for the thickness of 20 cm.
[0069]
Comparing data of 900 MHz fog 10 cm and thickness 5 cm, it was found that the signal intensity increased at about 2 nsec for both the cavity and the jumper. Compared with the result of 1500 MHz, 900 MHz has a good S / N, and it is considered that it was recognized also in the jumper. From the other 900 MHz and 400 MHz results, it was difficult to confirm the repeated signal of the front signal of the defect due to the large pulse width and the influence of the steel support.
[0070]
From the above results, in order to identify the cavity and the jumper, conditions such as the thickness of the defect, the antenna to be used and the depth of cover are limited. 1) Confirm the presence or absence of a back signal of the defect, and 2 ) It is considered that the type can be determined by confirming the repetitive signal of the defect front signal.
[0071]
In addition, for the cavity and jumper where the back signal of the defect could be detected from the envelope signal, the time difference between the peak of the front signal and the back signal of the defect was obtained, and the thickness of the defect was obtained from the electromagnetic wave velocity obtained from the electromagnetic wave velocity analysis. . The results at 1500 MHz are shown in Table 4. As shown in Table 4, although there are some errors, the values are almost reasonable. From this result, it can be seen that the thickness of the defect can be estimated by calculating at a known speed.
[0072]
[Table 4]

[0073]
Thus, as a result of examining the possibility of determining the defect type from the analysis of the scattering source and the reflection source from the envelope signal, the possibility of identifying the jumper is obtained if the back signal of the defect can be clearly confirmed. In addition, there is a possibility that the defect type can be determined by confirming the repetitive signal of the front signal of the defect. It was found that if the back signal of the defect is obtained, the thickness can be estimated from the time difference between the front signal of the defect and the peak of the back signal.
[0074]
Next, a third embodiment of the present invention will be described. In the present embodiment, the type of defect is determined based on the value of the average electromagnetic wave velocity measured. The web portion of the support that is H-shaped steel is flat as shown in FIG. 31, but this is a curved shape as shown in FIG. Therefore, if the least square method is used from the geometric relationship between FIG. 31 and FIG. 32, the average electromagnetic wave velocity can be obtained.
[0075]
In FIG. 31, d is defined as the depth of the support work, x is the moving distance of the antenna from directly above the support work, and v is the electromagnetic wave velocity in the concrete. When the antenna is moved along the concrete surface in FIG. 31 and the survey is performed, the locus of the support signal becomes as shown in FIG. Here, i is the suffix of the measured position value, and is defined as the presence of n points, xi; i-th lateral antenna position, and yi;
[0076]
yi = 2 · (xi²+ D²)^0.5/ V
However, d and v are unknown.
[0077]
xi and yi are measured values, and d and v are obtained from the measured values.
[0078]
From the above formula
(A ・ xi²+ B)^0.5= Yi / 2
However, a = 1 / v²    b = (d / v)²
a ・ xi²+ B = yi²/ 4 = zi
r = Σ (a · xi²+ B-zi)²    (Σ is 1 to n)
r is the residual to minimize this
[0079]
From ∂r / ∂a = 0
aΣxi⁴+ BΣxi²= Σxi²・ Zi
From ∂r / ∂b = 0
aΣxi²+ BΣ = Σzi
Therefore,
aΣxi²+ Nb = Σzi
Solve this
[0080]
[Expression 4]
a = {nΣxi²・ Zi-Σxi²・ Σzi} / {nΣxi⁴-(Σxi²)²}
b = {Σxi²・ Σzi-a (Σxi²)²} / NΣxi²
[0081]
Therefore,
Average electromagnetic wave velocity: v = (1 / a)^0.5, Depth of support work: d = v · b^0.5  It becomes.
[0082]
FIG. 33 shows the results of measuring the electromagnetic wave velocity of the healthy part thus obtained in the range of 2 weeks to 8 weeks of age. The two points in the figure represent the measurement results with a jumper and a low-strength concrete model. 0.34c which is the electromagnetic wave velocity of the healthy part separately obtained from the depth of the support work and the appearance position of the support work signal is shown by a horizontal line in the figure. As shown in the figure, the electromagnetic wave velocity obtained by curve fitting almost coincides with the value indicated by the horizontal line regardless of the age and the concrete model used.
[0083]
Next, FIG. 34 shows the electromagnetic wave velocity of the low-intensity part similarly measured from the age of 2 weeks to 8 weeks. In the figure, each material age is divided into cover depths of 10 cm, 20 cm and 30 cm. Further, three points in each cover depth represent a defect thickness of 5 cm, 10 cm, and 20 cm. When the thickness of the cover and the defect increased, the support signal overlapped with the reflected wave of the defect, and the support track could not be traced accurately. In such a case, measurement was impossible. As can be seen in the figure, the electromagnetic wave velocity in the low intensity part was almost the same as that in the healthy part in FIG. Similarly, FIG. 35 shows the electromagnetic wave velocity measured at the jumper portion by curve fitting. Although the variation was larger than the measurement result of the low-strength part, the value was around 0.5c. As described above, the electromagnetic wave velocity of the junker portion is about 0.65c. FIG. 35 shows the measurement result when such a fast junker portion is included in a part of the propagation path. For this reason, it is judged that the obtained result is a reasonable value. When considering the result of the jumper part in FIG. 35 together with the result of the low-intensity part in FIG. 34 and the healthy part in FIG. 33, the defect type can be determined by this method.
[0084]
Thus, as a result of examination from the electromagnetic wave velocity analysis, the electromagnetic wave velocity of the healthy part and the low-intensity part was about 0.34c, whereas that of the jumper part was about 0.65c. These electromagnetic wave velocities did not change greatly in the range up to the age of 8 weeks, and no frequency dependence was observed. It was possible to determine the velocity of electromagnetic waves from the trajectory of the support signal, and from this result, there was a possibility of making a judgment of junkers and low intensity. In addition, since the support signal disappeared in the cavity, the possibility of determining the cavity was also obtained. Further, based on the results of the above-described consideration regarding the speed, when a reflected wave from a defect as shown in FIG. 36 is obtained and a signal as shown by a wavy circle in FIG. become.
[0085]
A method of obtaining the thickness of the defect portion x shown in FIG. 37C using the electromagnetic wave average speed will be described. As shown in FIGS. 37 (a) and 37 (b), it is reinforced and supported by a support that is intermittently arranged in the axial direction of the concrete lining tunnel. Here, depth of support work: d, electromagnetic wave velocity of healthy part: V₁, Electromagnetic velocity of defects: V₂Is known. Also, the time difference from the radar record of the inspection machine to the support in the healthy part: t₁The time difference from the surface to the support work at the defective part: t₂Read. Then, x can be obtained from the relationship between t and v.
[0086]
[Equation 5]

[0087]
[Formula 6]
x = ((t₂-T₁) / 2) ・ ((v₁・ V₂) / (V₁-V₂))
[0088]
Next, a fourth embodiment relating to spectrum analysis of reflected waves will be described.
The waveform of electromagnetic waves reflected by internal defects reflects the reflection characteristics of defects such as waveform dispersion caused by frequency-dependent effects such as attenuation and reflection, and the possibility of identifying the type of defect by analyzing the waveform is there. It is expected that these effects can be detected as a difference in the frequency spectrum of the reflected wave. In order to clarify this, a spectrum analysis was performed using 400 MHz data having the best S / N among the three types of antennas.
[0089]
FIG. 38 shows the original data of the cavity of 10 cm in the cover at the age of 8 weeks, the background removal process, the background removal process and the gain-corrected radar image. By performing background removal processing and gain correction using data of healthy parts without defects, defect waveforms without distortion were extracted regardless of measurement conditions.
[0090]
FIG. 39 shows a waveform example (cavity: material age 8 weeks, cover 10 cm, thickness 5 cm) in which a defect waveform is extracted by background removal processing. In the case of a 400 MHz antenna, it was found that when the cover was shallow, it was influenced by the overlap with the reflected wave received from the concrete surface. By separating the defect waveform from a signal other than the reflected wave from the defect, information on only the defect can be extracted.
[0091]
When performing the spectrum analysis, spatial averaging was performed in the range where each defect reflected wave appeared, and the S / N of the reflected waveform was improved. The analysis gate was set from the vicinity of the rising edge of the defect reflected wave as shown in FIG. In the spectrum analysis, the power spectrum was obtained after Fourier transform of the defect reflected wave.
[0092]
Here, the power spectrum calculation result is shown. 40 to 42 show the power spectrum calculation results of defect reflected waves with thicknesses of 10, 10 and 20 cm measured at the eighth week with fogs of 10, 20 and 30 cm. In addition, FIGS. 43 to 45 show the power spectrum shown in FIG. 40 for each of the defect thicknesses of 5 cm, 10 cm, and 20 cm, obtained from the air direct wave data, the data obtained from the cavity model, and the jumper model. It is displayed separately for each data. 40 to 42 also show airborne direct wave spectra separately collected during measurement. This airborne direct wave data is a signal directly transmitted through the transmitting / receiving antenna, and can be considered as a characteristic of the individual antenna. As shown in the first embodiment, the reflected wave of the cavity and the jumper was clearly captured, but the low-intensity reflected wave was not detected with a low S / N except for the conditions of the cover of 10 cm and the thickness of 5 cm. It was. Therefore, in this spectrum analysis, the type determination of the cavity and the jumper was examined.
[0093]
From FIGS. 40 to 42, it was found that the spectrum of the cavity reflected wave is almost similar to the spectrum of the direct air wave at a cover of 10 cm. It was found that the degree of distortion from the aerial direct wave spectrum increased as the fog depth increased to 20, 30 cm, and the spectrum shifted to the lower frequency side when the fog depth was 30 cm. This may be due to the influence of waveform dispersion due to the frequency dependence of attenuation in concrete. The difference in defect thickness was not noticeable. Thus, although the fogging becomes large, the spectral shape of the cavity reflected wave is slightly distorted, but it seems to be similar to the spectral shape of the direct air wave.
[0094]
On the other hand, it was found that the spectrum of the jumper reflected wave has a larger degree of distortion than the spectrum of the direct air wave. When the aerial direct wave spectrum was used as a reference, some of them were distributed on the high frequency side and others were distributed on the low frequency side, and no consistent trend was observed. In order to investigate this cause, the defect waveform (covering 10 cm) before the spectrum analysis was reviewed. 46 and 47 show defect waveforms.
[0095]
The cavities had almost the same shape with different amplitudes even when the thickness changed. As shown in FIG. 45, it was found that the jumper has a shape in which the waveform increases as the thickness increases. This is considered to be due to the fact that the front signal and the back signal overlap each other because the back signal of the jumper appears behind the front signal because the electromagnetic wave velocity of the jumper is lower than that of the cavity. In other words, the spectrum of the cavity reflected wave is less influenced by the back signal, but the spectrum of the junker reflected wave is presumed that the degree of distortion is increased because the influence of the back signal is increased because the electromagnetic wave velocity is small. In addition, the effect of attenuation in concrete due to the increase of fogging, and the effect of scattered noise in concrete due to the decrease in S / N of the junker reflected wave compared to the cavity reflected wave greatly distorted the spectrum of the junker reflected wave. Conceivable
[0096]
From the above results, it was found that the spectrum of the cavity reflected wave is similar to the spectrum of the direct air wave with less distortion than the spectrum of the junker reflected wave, based on the spectrum of the direct air wave. .
[0097]
From the power spectrum calculation results, it was found that the spectrum of the junker reflected wave was distorted with respect to the spectrum of the direct air wave compared to the spectrum of the cavity reflected wave. Although the difference between the spectrum of the cavity reflected wave and the jumper reflected wave is known, it is necessary to quantify to distinguish this. In order to quantify the estimated power spectrum shape, four frequency feature amounts (shape parameters) shown in equations (A1) to (A4) were calculated from equations (A5) and (A6). Expressions (A5) and (A6) are the n-th moment (moment) around the origin and the n-th moment (moment) around the arithmetic mean, respectively.
[0098]
The four feature amounts are parameters representing the shape (characteristic) of the distribution such as the power spectrum, and can be quantified by quantifying in this way.
[0099]
[Expression 7]

[0100]
From the results of the power spectrum, it was found that the degree of distortion from the airborne direct wave spectrum differs between the cavity and the jumper. Therefore, these four feature values obtained for each defect are the feature values obtained from the airborne direct wave spectrum. And the absolute value difference.
[0101]
FIG. 48 shows the data for the 10 cm fog. By converting the feature value into the absolute value difference in this way, the jumper has a larger value than the cavity at the cover of 10 cm, and the possibility of identification is obtained. Even at the cover of 20 and 30 cm, the separability was not as good as that of the cover of 10 cm, but the same tendency was shown.
[0102]
From the above results, the possibility of quantitative type determination could be found by obtaining the feature amount.
[0103]
The possibility of the type determination was obtained by obtaining the feature amount. In order to determine this type, an objective determination is required. Therefore, a five-dimensional multivariate analysis using four features and age was examined. To discriminate between cavities and junkers, we determined the Mahalanobis distance of the features obtained in this analysis, and performed discriminant analysis. The Mahalanobis distance is a distance in consideration of variation of each defect feature amount with respect to the distance (Euclidean distance) to the center of gravity of each defect feature amount space. The Mahalanobis distance can be defined as the square of the value obtained by dividing the Euclidean distance by the standard deviation when considered in one dimension (one variable). When this is considered in multi-dimension (multivariate), the Mahalanobis distance to the center of gravity of both groups of the i-th sample can be obtained by the following equation.
[0104]
[Equation 8]

[0105]
Tables 5 to 7 show the results of discriminant analysis of the covers 10, 20 and 30 cm. The discriminant predictive value shows a result of about 80% or more in any fogging and seems to be good. From this analysis, there were many cases where misidentification was made with junka as a hollow. That is, from this analysis result, it is considered that the correct answer rate is high in the determination of the jumper. In the future, it is expected that the reliability of the evaluation will be improved by increasing the number of databases by reviewing the feature values and data of actual structures.
[0106]
[Table 5]

[0107]
[Table 6]

[0108]
[Table 7]

[0109]
Here, an embodiment in which the methods according to the above embodiments are used in a comprehensive manner will be described. In this study, we examined defect evaluation mainly for defect type determination using nondestructive underground radar exploration of concrete structures. These are summarized in Table 8.
[0110]
[Table 8]

[0111]
The analysis methods performed in the above embodiments reflect various characteristics of defects, and the factors that are affected differ for each analysis method. In addition, each analysis approach is different, and it is conceivable that evaluation accuracy can be improved by evaluating defects based on the analysis results from various fields. Furthermore, considering that the application range is different, it is desirable to make a comprehensive judgment by combining several analysis methods corresponding to the application range in order to adapt to an actual concrete structure.
[Brief description of the drawings]
FIG. 1 is a longitudinal sectional view of a concrete model created for an experiment of the present invention.
FIG. 2 is a graph showing the results of a uniaxial compressive strength test for two types of concrete used for the test body.
FIG. 3 is a graph showing the measurement results of the effective porosity of two types of concrete used for the specimen.
FIG. 4 is a graph showing measurement results of ultrasonic velocities (P wave velocity: Vp, S wave velocity: Vs) of two types of concrete used for a test specimen.
FIG. 5 is a graph showing the measurement results of the moisture content of two types of concrete used in the test specimen.
FIG. 6 is a photograph showing a concrete model subjected to inspection.
7 is a photograph showing the state of burying internal defects in the concrete model shown in FIG. 6;
FIG. 8 is a diagram showing a change in amplitude in a path from when an electromagnetic wave radiated from a transmitting antenna is reflected by an internal defect, captured by a receiving antenna, and recorded on a measuring instrument;
FIG. 9 is a graph plotting the relationship between the amplitude / amplification factor related to a support reflected wave at a defect-free portion measured in a road tunnel and the round-trip propagation time.
FIG. 10 is a graph showing a radar image by gradation display and a waveform at an arrow position when a cavity model having a cover of 20 cm and a thickness of 5 to 20 cm is inspected with a 1500 MHz antenna.
FIG. 11 is a graph showing a radar image by gradation display and a waveform at an arrow position when a jumper model having a cover of 20 cm and a thickness of 5 to 20 cm is inspected with a 1500 MHz antenna.
FIG. 12 is a graph showing a radar image by gradation display and a waveform at an arrow position when a low-intensity model having a cover of 10 cm and a thickness of 5 to 20 cm is inspected with a 900 MHz antenna.
FIG. 13 is a graph showing a radar image by gradation display and a waveform at an arrow position when a cavity model having a cover of 30 cm and a thickness of 5 to 20 cm is inspected with a 400 MHz antenna.
FIG. 14 is a diagram showing a relationship between a defect thickness in a test body and a reflected wave waveform.
FIGS. 15A and 15B are graphs showing measurement results of each sample trace for each defect thickness when a cavity model having a cover of 10 cm is inspected with a 1500 MHz antenna; FIG. 15A shows a round-trip propagation time, and FIG. / Amplification rate is plotted.
FIGS. 16A and 16B are graphs showing the measurement results of each sample trace for each defect thickness when a cavity model having a cover of 10 cm is inspected with a 900 MHz antenna, wherein FIG. 16A shows the round-trip propagation time, and FIG. 16B shows the initial amplitude. / Amplification rate is plotted.
FIGS. 17A and 17B are graphs showing measurement results of each sample trace for each defect thickness when a cavity model having a cover of 10 cm is inspected with a 400 MHz antenna, wherein FIG. 17A shows the round-trip propagation time, and FIG. 17B shows the initial amplitude. / Amplification rate is plotted.
FIG. 18 is a graph showing the relationship between initial motion amplitude / amplification factor and round-trip propagation time obtained by inspection of a cavity model and a jumper model with a 1-week-old material using a 1500 MHz antenna.
FIG. 19 is a graph showing the relationship between initial motion amplitude / amplification factor and round-trip propagation time obtained by inspection of a cavity model, a jumper model, and a low-strength model with a 1-week-old material using a 900 MHz antenna.
FIG. 20 is a graph showing the relationship between initial motion amplitude / amplification factor and round-trip propagation time obtained by inspection of a cavity model, a jumper model, and a low-strength model with a 1-week-old material using a 400 MHz antenna.
FIG. 21 is a graph showing the relationship between the attenuation rate obtained from the 900 MHz antenna recording and the age of the specimen.
FIG. 22 is a graph showing the relationship between the value of initial amplitude × reflectance obtained from 900 MHz antenna recording and the age of the specimen.
FIG. 23 is a graph showing the relationship between the coarse aggregate / (coarse aggregate + fine aggregate + cement) and the electromagnetic wave velocity relationship.
FIG. 24 is a radar image by gradation display obtained by inspection of a concrete specimen having a cavity, where (a) shows the original data and (b) shows the data after background removal and gain correction. Is.
FIG. 25 is a radar image by gradation display obtained by inspection of a concrete specimen having a jumper, in which (a) shows original data and (b) shows data after background removal and gain correction. Is.
FIG. 26 is a graph showing a reception waveform and a result of waveform processing in the second embodiment of the present invention, where (a) is a reception waveform when a test body having a cavity is used, and (b) is a sound waveform. (C) is a waveform obtained by removing the background from the received waveform of (a), (d) is a waveform obtained by rectifying the waveform of (c), and (e) is a waveform obtained by rectifying the waveform of (c). It corresponds to an envelope waveform (signal) obtained by Hilbert transform of the waveform of (d).
FIG. 27 is a diagram showing a radar image by gradation display obtained by inspection of a test body having a support work.
FIG. 28 is an envelope signal obtained when a defect having a defect thickness of 5 cm, 10 cm, and 20 cm is provided on a specimen having a cover of 10 cm using an antenna of 1500 MHz, and (a) a cavity at the age of 2 weeks. Model, (b) Cavity model at 4 weeks of age, (c) Cavity model at 8 weeks of age, (d) Junker model at 2 weeks of age, (e) 4th week of material The jumper model (f) corresponds to the case where the jumper model at the age of 8 weeks is used.
FIG. 29 is an envelope signal obtained when a defect having a defect thickness of 5 cm, 10 cm, and 20 cm is provided on a specimen having an age of 8 weeks using a 900 MHz antenna, and (a) shows a cover with a cover of 10 cm. Cavity model, (b) Cavity model with a cover of 20 cm, (c) Cavity model with a cover of 30 cm, (d) Crasher model with a cover of 10 cm, (e) Crasher model with a cover of 20 cm, (f) Cushion model with a cover of 30 cm This corresponds to the case of using the Junka model.
FIG. 30 is an envelope signal obtained when a defect having a defect thickness of 5 cm, 10 cm, and 20 cm is provided on an 8 week old specimen using a 400 MHz antenna, and (a) shows a cover with a cover of 10 cm. Cavity model, (b) 20 cm cover model, (c) 30 cm cover model, (d) 10 cm cover model, (e) 20 cm cover model, (f) 30 cm cover model. This corresponds to the case of using the Junka model.
FIG. 31 is a schematic view of a positional relationship between an antenna and a support work.
FIG. 32 is a schematic diagram showing a locus of a support work signal.
FIG. 33 is a diagram in which the ratio of the electromagnetic wave velocity in a sound concrete specimen and the electromagnetic wave velocity in the air is plotted for each material age by the technique (curve fit) according to the third embodiment of the present invention.
FIG. 34 is a diagram in which the ratio of the electromagnetic wave velocity in a concrete specimen having a low-strength portion and the electromagnetic wave velocity in the air is plotted for each material age by the method (curve fit) according to the third embodiment of the present invention. .
FIG. 35 is a diagram in which the ratio of the electromagnetic wave velocity in the concrete specimen having a jumper to the electromagnetic wave velocity in the air is plotted for each material age by the technique (curve fit) according to the third embodiment of the present invention.
FIG. 36 is a radar image by gradation display showing a defect signal in a low intensity portion.
37A is a cross-sectional view of the tunnel as viewed from the longitudinal axis direction, FIG. 37B is a cross-sectional view taken along the line AA of FIG. 37A, and FIG.
38A and 38B are radar images by gradation display showing defect signals, where FIG. 38A is the original data, FIG. 38B is the image after background removal, and FIG. 38C is the image after gain correction.
39 (a) shows a received waveform of a healthy specimen, (b) shows a received waveform of a specimen having a hollow portion, and (c) shows background removal processing and gain correction for the received waveform of (b). It is the waveform obtained by giving.
FIG. 40 shows the power spectrum of a defect having a thickness of 5 cm, 10 cm, and 20 cm measured using a specimen with a cover of 10 cm, air is a direct wave in the air, void is a cavity, junka is a junker, and low is a low-strength part. Show.
FIG. 41 shows the power spectrum of a defect having a thickness of 5 cm, 10 cm, and 20 cm measured using a specimen with a cover of 20 cm, where air is a direct wave in the air, void is a cavity, junka is a junker, and low is a low-strength part. Show.
FIG. 42 shows the power spectrum of a defect having a thickness of 5 cm, 10 cm, and 20 cm measured using a specimen with a cover of 30 cm, air is a direct wave in the air, void is a cavity, junka is a junker, and low is a low-strength part Show.
FIG. 43 shows the power spectrum of airborne direct waves, cavities, and jumpers measured using a test specimen with a cover of 10 cm and a defect thickness of 5 cm.
FIG. 44 shows the power spectrum of an airborne direct wave, cavity, and jumper measured using a test specimen with a cover of 10 cm and a defect thickness of 10 cm.
FIG. 45 shows power spectra of airborne direct waves, cavities, and jumpers measured using a test specimen with a cover of 10 cm and a defect thickness of 20 cm.
46A and 46B are diagrams showing waveforms obtained by inspection of a test specimen having a cavity, where FIG. 46A shows a case where the thickness of the cavity is 5 cm, FIG. 46B shows a case where the thickness of the cavity is 10 cm; c) corresponds to the case where the thickness of the cavity is 20 cm.
47A and 47B are diagrams showing waveforms obtained by inspection of a specimen having a jumper, where FIG. 47A shows a case where the thickness of the jumper is 5 cm, FIG. 47B shows a case where the thickness of the jumper is 10 cm; c) corresponds to the case where the thickness of the jumper is 20 cm.
FIG. 48 is a graph showing a distribution with respect to age for each specimen type with respect to four frequency feature amounts, (a) is a normalized average frequency, (b) is a normalized standard deviation frequency, and (c) is normalized. Skewness, (d) shows the distribution of normalized kurtosis.

Claims (22)

A method for inspecting a concrete by electromagnetic waves in which the inside of the concrete structure is inspected by transmitting an electromagnetic wave to the concrete structure and receiving a reflected wave. A concrete inspection method using electromagnetic waves, characterized in that A ₀ · r is obtained from the relationship between the propagation time t and the following equation (1), and the type of the defective portion is determined based on the A ₀ · r.
ln (A / G) =-(α · V · t) + ln (A ₀ · r) (1)
Where A is the amplitude recorded in the measuring instrument, G is the amplification factor, α is the attenuation factor of the amplitude in the concrete, V is the electromagnetic wave velocity, A ₀ is the initial amplitude of the electromagnetic wave transmitted from the transmitting antenna on the concrete surface, r Is the reflectivity of electromagnetic waves at internal defects. The measurement of the A / G and the round-trip propagation time t is performed at a plurality of locations having different depths such as defects, and A ₀ · r and α are obtained from the above equation (1) by the least square method. Item 2. A concrete inspection method using electromagnetic waves according to Item 1. 3. The electromagnetic wave according to claim 1 or 2, wherein the depth of the target part is obtained from the measured round-trip propagation time t of the initial peak (P) of the reflected wave and the electromagnetic wave velocity inside the concrete structure. Concrete inspection method. The concrete inspection method using electromagnetic waves according to any one of claims 1 to 3, wherein the frequency is set so that a half wavelength of the electromagnetic wave is smaller than an expected thickness of the defect portion. A method for inspecting a concrete by using electromagnetic waves to inspect the inside of the concrete structure by transmitting an electromagnetic wave to the concrete structure and receiving a reflected wave, and removes a sound part signal from the measured defect signal and performs envelope processing A method for inspecting concrete using electromagnetic waves, comprising: determining whether a defective portion is a junker portion based on the presence or absence of a back signal having another peak with respect to a main signal having a peak in the processed signal. 6. The method for inspecting concrete using electromagnetic waves according to claim 5, wherein the thickness of the junker portion is obtained from the difference between the peak of the main signal and the peak of the back signal. The method for inspecting concrete using electromagnetic waves according to claim 5 or 6, wherein it is determined whether or not the defective portion is a hollow portion by a slight front repeating signal generated after the main signal. The concrete inspection method using electromagnetic waves according to claim 5, wherein the processing signal is rectified, and the envelope processing is performed by analytical signal processing. A method for inspecting a concrete by an electromagnetic wave inspecting the inside of the concrete structure by transmitting an electromagnetic wave to the concrete structure and receiving a reflected wave. By using a plurality of coordinates, the depth of the flat reference body and the average electromagnetic wave velocity in the concrete are obtained, and it is determined whether or not the defect portion is a junker portion by this average electromagnetic wave velocity. Concrete inspection method. The concrete inspection method using electromagnetic waves according to claim 9, wherein the thickness of the jumper portion is calculated from the amount of change in the average propagation speed. The concrete inspection method using electromagnetic waves according to claim 9 or 10, wherein the presence of a cavity is determined based on disappearance of the curved image. The method for inspecting concrete using electromagnetic waves according to any one of claims 9 to 11, wherein when the average radio wave velocity is substantially the same as that of a healthy part and a reflected wave appears, it is determined that a low intensity part exists. The concrete inspection method using electromagnetic waves according to claim 9, wherein the flat reference body is a tunnel support. The method for inspecting concrete using electromagnetic waves according to any one of claims 9 to 13, wherein the depth of the flat reference body and the average electromagnetic wave velocity are obtained by a least square method. A method for inspecting concrete with an electromagnetic wave in which the inside of the concrete structure is inspected by transmitting an electromagnetic wave to the concrete structure and receiving a reflected wave, and removes the sound part signal from the measured defect signal and A method for inspecting concrete using electromagnetic waves, characterized in that a power spectrum is obtained, and the type of the defect portion is determined by an approximate determination of the obtained power spectrum shape and the cavity power spectrum shape and the junker power spectrum shape obtained in advance. The concrete inspection method using electromagnetic waves according to claim 15, wherein the approximation determination is performed by obtaining a Mahalanobis distance of a measured reflected wave for a plurality of parameters obtained from a power spectrum. The method for inspecting concrete using electromagnetic waves according to claim 15 or 16, wherein the parameter has age of concrete. The parameters are the average frequency ν _{1 of} the power spectrum, the standard deviation frequency (μ ₂ ) ^1/2 , the skewness μ ₃ / (μ ₂ ) ^3/2 and / or the kurtosis μ ₄ / (μ ₂ ) ² -3. The concrete inspection method using electromagnetic waves according to any one of claims 15 to 17, characterized by comprising: A concrete inspection method using electromagnetic waves, wherein both the methods according to any one of claims 1 to 8 or any one of claims 9 to 18 are carried out. Implementing a method selected from at least two of the four groups according to any one of claims 1 to 4, claim 5 to 8, claim 9 to 14, or claim 15 to 18. A concrete inspection method using electromagnetic waves. 21. The method for inspecting concrete using electromagnetic waves according to claim 19 or 20, wherein the frequency of electromagnetic waves used in the two or more inspection methods is different. The concrete by the electromagnetic wave in any one of Claims 1-21 provided with the transmitting antenna which transmits electromagnetic waves, the receiving antenna for receiving the reflected electromagnetic waves from a concrete structure, and the measuring device which measures a received wave Concrete inspection equipment using electromagnetic waves used for inspection methods.

Images