JP3707473B2 - Method and apparatus for ultrasonic inspection of steel pipes - Google Patents

Description

但し、ｆ_C は中心周波数、Ｂは周波数掃引帯域幅、Ｔはパルス幅である。
【００８５】
送信用メモリ１９に記憶されたチャープ波は、同期信号発生器１８からの送信周期Ｔ₀ を有する送信タイミング信号ａに同期して読出されて、Ｄ／Ａ変換器２１でアナログ信号に変換され、さらに送信用増幅器２２で増幅されて、周期Ｔ₀ を有するチャープパルス信号ｂとして、ダイプレクサ２３を経由して斜角探触子５へ印加される。
【００８６】
斜角探触子５から水などの音響接触媒質を介して被検査体としての鋼管１の内部に入射された超音波６は、溶接部３内の面状欠陥４等で反射され再び斜角探触子５で受信される。斜角探触子５から出力されたエコー信号ｃは、ダイプレクサ２３を経由して受信用増幅器２４で増幅され、Ａ／Ｄ変換器２５でデジタル信号に変換され、デジタルフィルタとしてのＦＩＲフィルタ（有限インパルス応答フィルタ）２６を通り、同期加算平均化回路２７に入力される。
【００８７】
ＦＩＲフィルタ２６は、図６に示すように、１２８個の乗算器２６ａと、１２８個の加算器２６ｂと、１２８個の遅延器２６ｃとで構成されている。ＦＩＲフィルタ２６の各乗算器２６ａに印加される各係数Ｃ_o 〜Ｃ₁₂₇ には、波形演算器２０で発生させたチャープ波形の信号が参照信データとして設定されている。また、各遅延器２６ｃは入力信号に対して送信周期Ｔ_o に相当する時間の遅延を行い出力する。
【００８８】
このような構成のＦＩＲフィルタ２６においては、デジタル信号に離散化されたエコー信号ｃの波形ｘ（τ）と相関演算を行うための参照波形は、ある一定のサンプリング周波数でサンプリングされ（離散化され）る。この例では各離散化データ値は、１２８個の係数Ｃ_o 〜Ｃ₁₂₇ として、それぞれ乗算器２６ａ一方に入力される。
【００８９】
一方、入力端から送信周期Ｔ₀ 毎に入力される離散化受信データｘ（τ）は、各乗算器２６ａの他方に入力に直接供給され、前記参照データ（係数）Ｃ_o 〜Ｃ₁₂₇とそれぞれ個別に乗算される。先頭の係数Ｃ₁₂₇との乗算結果を除く各乗算結果はそれぞれ１２７個の遅延器２６ｃと加算器２６ｂとが交互に直列接続された該当加算器の入力の一方に供給される。
【００９０】
そして、先頭の係数Ｃ₁₂₇との乗算結果のみが前記交互に直列接続された先頭の遅延器２６ｃに直接供給され、この遅延器２６ｃの後段に直列接続される加算器２６ｂの入力の他方に係数Ｃ₁₂₆ との乗算結果が供給されている。そして、前記直列結合における最後の加算器２６ｂの出力が相関演算出力ｙ（τ）となる。
【００９１】
いま入力信号をｘ_i(j)、係数（参照データ）をＣ_(k)、出力信号をｙ_i(j)、タップ数をｖ_c ，探傷信号の繰り返しをｉ、一周期の探傷信号データ点数をｎとすると、出力信号ｙ_i(j)は(10)式の畳み込み演算を行うことで得られる。
【００９２】
【数１】

【００９３】
(10)式において参照信号を逆順にすると次の(11)式となる。
【００９４】
【数２】

【００９５】
この結果、ＦＩＲフィルタ２６を用いてエコー信号（受信信号）ｃに対する相関演算を行うことができる。この相関の動作は、参照信号と受信信号をｊだけずらしながら相互相関を計算していくものである。
【００９６】
このＦＩＲフィルタ２６の具体的動作を図７示す波形図を用いて説明する。
【００９７】
図７において、時間τ₁ の点は式でｊ＝０の位置に相当する。まず、τ_i の位置で、受信信号（エコー信号）と参照信号との相関演算を０〜Ｎ_c-1 のデータ分だけ行う。ここでＮ_c は、ＦＩＲフィルタ２６のタップ長であり、すなわち係数Ｃ_o 〜Ｃ₁₂₇ の数である。相関の結果は、図中一番下の信号、すなわち新たなエコー信号ｄとして出力される。このτ_i の時点では参照信号と受信信号（エコー信号）は類似していないため、出力はほとんど零である。
【００９８】
次に、ｊを一つずつ増やし、図中τ₂ ，τ₃ ，…のように順次演算を行っていく。この結果、受信信号（エコー信号）中のエコーと参照信号の位相が一致した点（図中の時間軸のほぼ中央）で、最大のピークの相関信号が得られる。この結果、チャープ波のパルス幅は圧縮され、参照信号と相関のない電気的なノイズ信号は、大幅に低減される。
【００９９】
ＦＩＲフィルタ２６における相関処理によりパルス圧縮されたエコー信号ｄは次の同期加算平均化回路２７で同期加算平均処理が実施される。
【０１００】
図８は同期加算平均化回路２７の詳細構成図である。
【０１０１】
この同期加算平均化回路２７は、大きく分けて、第一のメモリ２８、加算器２９、第一のメモリ２８の書込・読出アドレス発生器３０ａ、３０ｂ、減算器３１、第二のメモリ３２、第二のメモリ３２の書込・読出アドレス発生器３３ａ、３３ｂ、減算器３４とで構成されている。
【０１０２】
なお、第一のメモリ２８は、同期信号発生器１８からの送信タイミング信号ａの送信周期Ｔ₀ 毎にＦＩＲフィルタ２６から受信した受信信号（エコー信号ｄ）のＮ＋１周期分以上のデータを格納できる容量を有し、この例では各周期毎の受信信号（エコー信号ｄ）はＭ個のデータを含み、各データは８ビットで構成されているとしている。
【０１０３】
また第２のメモリ３２は、各受信周期Ｔ₀ におけるＭ個の各データをそれぞれ個別にＮ周期分加算したＭ個の和データを格納できる容量を有し、この例ではそれぞれ１６ビットのデータをＭ個分格納するものとする。
【０１０４】
次に各部の動作を説明する。最初、ＦＩＲフィルタ２６から受信信号（エコー信号ｄ）が全く入力されてない初期状態においては、第一のメモリ２８及び第二のメモリ３２には何も記憶されてない。１回目の周期Ｔ₀ の受信信号（エコー信号ｄ）が入力されると、この受信信号は第一のメモリ２８の先頭の領域に記憶されると共に、加算器２９の入力の一方に加えられる。この１回目の周期では、第一のメモリ２８の信号読出アドレスＲＡ₁ が指定する領域及び第二のメモリ３２には何も記憶されていないので、加算器２９の入力の一方に加えられた受信信号（エコー信号ｄ）は、そのまま加算器２９を通り、さらに減算器３１を素通りし、第二のメモリ３２に格納される。
【０１０５】
２回目の周期Ｔ₀ の受信信号（エコー信号ｄ）が入力されると、この２回目の周期の受信信号と第二のメモリ３２に記憶されている前回の周期の受信信号（エコー信号ｄ）とが加算器２９で加算される。この加算結果は、減算器３１を素通りし、第二のメモリ３２に格納される。なおこの２回目の受信信号（エコー信号ｄ）は、第一のメモリ２８における第二の領域にも記憶される。
【０１０６】
以後平均回数Ｎ周期分の受信信号（エコー信号ｄ）が入力されるまでは、このように第２のメモリ３２に受信信号（エコー信号ｄ）が順次加算されるのと、第一のメモリ２８に該当領域に順次受信信号が格納される動作が続く。
【０１０７】
平均回数Ｎ周期以上の受信信号（エコー信号ｄ）が入力されると、新たに入力された受信信号（エコー信号ｄ）のＮ周期前の受信信号（エコー信号ｄ）が第一のメモリ２８の信号読出アドレスＲＡ_i が指定する領域から読み出され、減算器３１に減算信号として供給される。
【０１０８】
このため加算器２９により、第二のメモリ３２に記憶されている信号と新たに入力された受信信号（エコー信号ｄ）が加算され、その加算結果が出力されると、この加算結果より第一のメモリ２８から読出されたＮ周期前の受信信号が減算器３１によって減算されるので、減算器３１の出力信号は常にＮ周期分の受信信号を加算したものとなる。
【０１０９】
このＮ周期分の受信信号の加算結果が、第二のメモリ３２に格納されると共に、除算器３４に入力されてＮで除算されるので、除算器３４の出力は、超音波パルスの送信周期Ｔ₀ 毎に、Ｎ周期分の受信信号（エコー信号ｄ）の同期加算平均結果が新たなエコー信号ｅとして、次の表示部３５へ出力される。
【０１１０】
ここで受信信号（エコー信号ｄ）を記憶し、また読出すための書込アドレス発生器３０ａ、３３ａ、読出アドレス発生器３０ｂ、３３ｂに供給するタイミング信号は、同期信号発生器１８から供給される送信タイミング信号ａを用いる。
【０１１１】
このように構成された同期加算平均化回路２７を採用することにより、エコー信号をチャープパルス信号の出力周期に同期して複数周期に亘って平均化することによって、ランダムな周期性を有しない材料ノイズや電気的なノイズの信号レベルが低下し、周期性を有する欠陥エコーの信号レベルは上昇する。
【０１１２】
より具体的の説明すると、この実施形態の超音波探傷装置においては、チャープ波のパルス圧縮を用いているため広い周波数帯の超音波６の送受信を可能にする広帯域の斜角探触子５を用いている。広帯域の斜角探触子５は一般に感度が悪いが、チャープ波のパルス圧縮を用いているため送信パルス幅は短いため、表面性状に起因して反射・干渉しているノイズエコーの１パルス毎の変化をより大きくすることができる。
【０１１３】
その結果、この信号を同期加算平均することによりイズエコーを狭帯域探触子の場合よりも低減することができる。よって、エコー信号に含まれる欠陥エコーのＳ／Ｎ比を大幅に向上できる。
【０１１４】
図９は、表面性状に起因するノイズエコーの波形を拡大したものであり、図９の（ａ）〜（ｄ）は、探傷位置を１〜４と変えたときに、狭帯域探触子と広帯域探触子によるそれぞれの波形例を示している。また図９の（ｃ）は、（ａ）〜（ｄ）の４つの波形を平均化処理した波形である。このように、エコー信号ｄを平均化することにより、Ｓ／Ｎ比を向上させることができる。
【０１１５】
【発明の効果】
以上説明したように、本発明の鋼管の超音波探傷方法及び鋼管の超音波探傷装置においては、鋼管の溶接部内における面状欠陥に対し超音波が垂直に入射するように斜角探触子を配設している。
【０１１６】
したがって、欠陥からのエコーを強く検出でき、かつ溶接部の余盛からのエコーを発生しづらくし、また、鋼管内を伝搬する超音波の路程が比較的長いことより、溶接金属からの組織エコーも少なく、斜角探触子から出力されるエコー信号に含まれる欠陥エコーのＳ／Ｎ比を向上できる。
【０１１７】
さらに、鋼管の肉厚対外径比（ｔ／Ｄ）が小さいと大屈折角の探触子を用いる必要があるが、その際の超音波の往復透過率の低下もデジタルフィルタで実現されるチャープパルス圧縮技術を用いて斜角探触子から出力されるエコー信号に対する信号処理を実施している。
【０１１８】
よって、電気的ノイズを増加させることなくエコー信号の信号レベルを増幅回路で回復できる。
【０１１９】
さらに、検査対象の鋼管の材料に音響異方性が認められる場合でも、音速分布を求めて必要な探傷屈折角を満たす斜角探触子を選定し探傷するために正常に探傷できる。
【図面の簡単な説明】
【図１】本発明の一実施形態に係る鋼管の超音波探傷方法が適用される鋼管の超音波探傷装置の機器構成図
【図２】音響異方性材料の超音波の速度分布特性を示す図
【図３】屈折角と斜角探触子における必要な振動子幅との関係を示す図
【図４】超音波の斜め入射時の往復透過率を示す図
【図５】斜角探触子に印加されるチャープパルス信号を示す波形図
【図６】ＦＩＲフィルタの概略構成を示すブロック図
【図７】同ＦＩＲフィルタの動作を説明するための図
【図８】同期加算平均化回路の構成構成を示すブロック図
【図９】狭帯域と広帯域の探触子によるノイズエコー波形とその平均化波形の例を示す図
【図１０】斜角探触子を用いた鋼管の一般的欠陥探傷方法を示す図
【符号の説明】
１…鋼管
３…溶接部
４…面状欠陥
５…斜角探触子
６…超音波
１１…探触子選定演算装置
１２，３５…表示部
１３…上位コンピューター
１４…ＰＷＤ演算装置
１５…マニュピレータ
１６…溶接シーム検出器
１７…ホルダー
１８…同期信号発生器
１９…送信用メモリ
２０…波形演算器
２１…Ｄ／Ａ変換器
２２…送信用増幅器
２３…ダイプレクサ
２４…受信用増幅器
２５…Ａ／Ｄ変換器
２６…ＦＩＲフィルタ
２７…同期加算平均化回路
１６マニピュレーター
２８…第一のメモリ
２９…加算器
３０ａ…第一のメモリの書込アドレス発生器
３０ｂ…第一のメモリの読出アドレス発生器
３１…減算器
３２…第二のメモリ
３３ａ…第二のメモリの書込アドレス発生器
３３ｂ…第二のメモリの読出アドレス発生器
３４…除算器[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a defect detection method for detecting a defect existing inside a steel pipe, and in particular, a planar shape existing in a plane including a radial direction (thickness direction) generated in a steel pipe welded portion and an axial direction. The present invention relates to an ultrasonic flaw detection method for steel pipes and an ultrasonic flaw detection apparatus for steel pipes, which detect ultrasonic defects using an oblique probe attached to an outer peripheral surface of the steel pipe.
[0002]
[Prior art]
For example, a steel pipe having an outer diameter exceeding 1 m is manufactured by bending a strip-shaped steel plate into a cylindrical shape and then welding the side surfaces of the steel plates. In addition, the steel pipe manufactured by such a method may be called a welded steel pipe. Therefore, as shown to Fig.10 (a), in the outer peripheral surface 2 of the manufactured steel pipe 1, the linear welding part 3 extended in an axial direction exists.
[0003]
In such a steel pipe 1, the probability that the planar defect 4 resulting from the welding defect at the time of manufacture will generate | occur | produce inside the welding part 3 becomes high. Naturally, this planar defect 4 occurs in the bonding surface including the radial direction (thickness direction) and the axial direction.
[0004]
As a method for flaw detection of the planar defect 4 generated in the welded portion 3 of the steel pipe 1, an oblique ultrasonic flaw detection method as shown in FIGS. 10B and 10B has been put into practical use. .
[0005]
FIG. 10B shows a flaw detection in which ultrasonic waves 6 generated from the oblique probe 5 attached to the outer peripheral surface 2 of the steel pipe 1 are incident on the planar defect 4 inside the welded portion 3 with 0.25 skip. FIG. 10C shows a flaw detection method in which the ultrasonic wave 6 is similarly incident on the planar defect 4 with 0.75 skip.
[0006]
The angle formed by the propagation direction of the ultrasonic wave 6 incident into the steel pipe 1 from the oblique probe 5 and the perpendicular to the incident surface is referred to as the refraction angle θ of the flaw detection. The refraction angle θ of the touch element 5 is nominally 70 °.
[0007]
However, in the case of the steel pipe 1 having the cross-sectional shape 4 in which the ratio of the wall thickness t to the outer diameter D, that is, t / D is greater than 3%, the ultrasonic wave 6 incident on the steel pipe 1 has a refraction angle θ = 70 at the bottom surface. Since reflection does not occur at °, flaw detection at a refraction angle smaller than the nominal refraction angle of 70 ° (for example, 65 °, 60 °, and 55 °) is performed in the 0.75 skip shown in FIG. 10C.
[0008]
As a result, the t / D of the steel pipe 1 to be inspected is about 6% and 0.75 shown in FIG. 10C except for the case of performing flaw detection with a 0.25 skip / refractive angle of 70 ° shown in FIG. 10B. In skip flaw detection, the ultrasonic wave 6 does not enter the planar defect 4 to be detected vertically. Therefore, most of the reflection of the ultrasonic wave 6 by the planar defect 4 is not in the direction of the oblique probe 5, and the oblique probe 5 cannot receive the echo of the planar defect 4, so that the defect detection capability is reduced. .
[0009]
Further, in recent years, since a steel plate that has been controlled and rolled, such as a TMCP steel plate, is used as the base material of the welded steel pipe, the case in which the object to be inspected has acoustic anisotropy is the majority. The material having such acoustic anisotropy does not necessarily have a transverse wave sound velocity of the ultrasonic wave 6 of 3230 m / s, and the sound velocity varies depending on the propagation direction.
[0010]
For this reason, in the oblique angle probe 5 manufactured assuming that the shear wave velocity of the object to be inspected is 3230 m / s which is the reference velocity, the nominal refraction angle described in the specification is different from the actual refraction angle θ, and this nominal refraction angle The value of the refraction angle is no longer reliable. Depending on the degree of acoustic anisotropy of the object to be inspected, the actual flaw detection refraction angle may be about 63 ° even when a nominal refraction angle of 70 ° is used.
[0011]
Therefore, in the standard JIS Z 3060 of the ultrasonic flaw detection test method for steel welds, the flaw detection of a material having acoustic anisotropy is an oblique probe so that the deviation between the nominal refraction angle and the actual flaw detection refraction angle is small. 5 is selected to be smaller.
[0012]
[Problems to be solved by the invention]
However, in the above-mentioned JIS standard, the joint welded part in the steel pipe manufacturing process is not applicable, and the automatic flaw detection for the surface defect 4 existing in the welded part 3 in the welded steel pipe is not a measure against acoustic anisotropy. Little has been done.
[0013]
For example, when an arc welded steel pipe is taken as an example, the t / D of a generally manufactured steel pipe 1 is 0.7 to 7%, and the conventional flaw detection method uses ultrasonic waves with respect to the planar defect 4 except for special cases. Since 6 is incident obliquely, the ultrasonic wave 6 reflected in the direction of the oblique probe 5 (the incident direction of the ultrasonic wave) is weak and has a low defect detection capability.
[0014]
Furthermore, in the conventional method, echoes reflected from the vicinity of the welded portion 3 and returning to the oblique probe 5 include not only those from the surface defect 4 but also many echoes from the overfill of the welded portion 3. Therefore, if an amplifier or the like is used to improve the defect detection capability, the echo from the planar defect 4 is amplified, but the echo from the surplus of the welded portion 3 and the tissue echo from the weld metal are also amplified. Since electrical noise also increases, only defective echoes cannot be detected with a high S / N ratio.
[0015]
On the other hand, in the flaw detection with 0.25 skip shown in FIG. 10B, the position of the planar defect 4 and the oblique probe 5 is closer than that of the 0.75 skip flaw detection. 5 can detect a defect signal relatively strongly. However, since the strong ultrasonic wave 6 is incident on the weld metal portion in the weld portion 3, the echo reflected in the weld metal is observed as a forest echo and the echo output from the oblique probe 5. The S / N ratio of the defect echo included in the signal is hardly improved.
[0016]
In addition, in the thin steel pipe 1, as shown in FIG. 10 (b), there is a concern that the holder that supports the bevel probe 5 may come into contact with the welded portion 3. This makes it impossible to detect flaws in the welded portion 3.
[0017]
In addition, when the material of the steel pipe 1 that is the object to be inspected has acoustic anisotropy, in the conventional manual flaw detection method, the deviation between the nominal refraction angle and the actual flaw detection refraction angle θ is relatively small, and the nominal refraction angle is 70. The flaw detection was performed using an oblique probe 5 smaller than °. However, the incident angle of the ultrasonic wave 6 with respect to the planar defect 4 is further away from the vertical as compared with the flaw detection with a 70 ° refraction angle. Less.
[0018]
On the other hand, in the case of automatic flaw detection of a welded portion of a steel pipe, as described above, there is almost no countermeasure against acoustic anisotropy, and an actual flaw detection refraction angle θ using the oblique probe 5 having a nominal refraction angle of 70 °. When the angle reaches 63 °, the flaw detection is performed at 63 °, and when the actual flaw detection refraction angle θ is 76 °, the flaw detection at 63 ° is performed using the oblique probe 5 having a nominal refraction angle of 70 °. It is carried out.
[0019]
For this reason, when the oblique probe 5 manufactured to perform the 70 ° flaw detection is to perform the 63 ° flaw detection, there is a problem that the directivity of the beam of the ultrasonic wave 6 is reduced and the flaw detection range is narrowed. is there. Further, when flaw detection is performed at 76 °, the directivity of the beam of the ultrasonic wave 6 is increased and the flaw detection range is widened. Along with this, the sensitivity also changed, and there was a problem that the analysis of the flaw detection result became very complicated.
[0020]
The present invention has been made in view of such circumstances, and by detecting the flaw detection refraction angle of the ultrasonic wave incident on the steel pipe by a predetermined procedure instead of the nominal refraction angle, a probe from a defect is obtained. It is possible to increase the reflection intensity of the ultrasonic wave in the direction, and it is caused by materials such as echoes from the weld overlay included in the echo signal output from the oblique probe and forest echoes from the weld metal Noise can be reduced, and flaw detection can be realized at a position that is not subject to mechanical limitations such as the approach limit distance, and even materials with acoustic anisotropy can be flawed at a predetermined refraction angle, and apparent vibration It is an object of the present invention to propose an ultrasonic flaw detection method for steel pipes and an ultrasonic flaw detection apparatus for steel pipes that can perform flaw detection with a constant beam spread in ultrasonic waves without fluctuation in the width of the element.
[0021]
[Means for Solving the Problems]
The present invention relates to a planar defect existing in a plane including a radial direction (thickness direction) and an axial direction in a welded portion existing in an axial direction of a steel pipe manufactured by bending a strip-shaped steel plate into a cylindrical shape. The present invention is applied to an ultrasonic flaw detection method for a steel pipe that performs ultrasonic flaw detection using an oblique probe attached to the outer peripheral surface of the steel pipe.
[0022]
And in order to eliminate the above-mentioned subject, in the ultrasonic flaw detection method of the steel pipe of the present invention,Based on the ratio of the wall thickness to the outer diameter of the steel pipe, the ultrasonic wave incident from the oblique probe into the steel pipe enters the steel pipe perpendicularly to the surface defect. The ultrasonic inspection flaw refraction angle is set, and the incident angle of the ultrasonic wave to the steel pipe in the oblique probe is set based on the velocity characteristic in the ultrasonic propagation direction and the flaw detection refraction angle. Then, the width of the vibrator in the oblique probe is determined based on the reduction ratio of the vibrator obtained from the set incident angle and flaw detection refraction angle.
Further, the incident angle of the ultrasonic wave with respect to the steel pipe in the oblique probe may be obtained based on the velocity characteristic in the direction corresponding to the ultrasonic wave propagation direction in the acoustic anisotropy of the steel pipe and the flaw detection refraction angle.
[0023]
Specifically, the flaw detection refraction angle θ of the ultrasonic wave is calculated by the equation (1) using the wall thickness t and the outer diameter D of the steel pipe.
[0024]
θ ＝ Sin^-1[1- (t / D)] (1)
In the ultrasonic flaw detection method for steel pipes configured in this way, the ultrasonic waves that are incident on the steel pipe by the oblique probe are incident on the surface defects of the welded portion perpendicularly, and are therefore reflected from the surface defects. The propagation direction of the wave is also equal to the incident direction, and the defect echo can be detected most efficiently by the oblique probe. Further, flaw detection from this direction reduces the reflection of ultrasonic waves due to extra welding at the welded portion, and has the effect of reducing noise.
[0027]
The reason for setting the incident angle of the ultrasonic wave to the steel pipe and the width of the vibrator by such a method will be described.
[0028]
As explained earlier,Since the flaw detection refraction angle θ of the ultrasonic wave incident on the steel pipe is determined by the ratio of the thickness t of the steel pipe to be inspected to the outer diameter D, that is, t / D, the ultrasonic wave is refracted at that angle. An oblique probe having an incident angle with respect to the surface of a steel pipe, and an oblique probe having a vibrator whose size of the ultrasonic beam (apparent vibrator width) after being refracted corresponds to the size of the surface defect to be flawed. It is necessary to use an angle probe.
[0029]
Generally, since the transverse wave velocity of ultrasonic waves propagating in steel is 3230 m / s, it is possible to easily select an oblique probe having an incident angle to be used for flaw detection by using Snell's law. Since the transducer reduction rate F in oblique flaw detection can be calculated from the relationship between the incident angle i and the refraction angle θ, the actual transducer width can also be calculated from the apparent transducer width required for flaw detection.
[0030]
However, since the controlled rolled steel sheet having toughness at low temperature has acoustic anisotropy, the speed of sound is not always 3230 m / s uniformly in all propagation directions of ultrasonic waves. In the inspected body having acoustic anisotropy, the sound velocity distribution of the ultrasonic wave (transverse wave) propagating in the plane perpendicular to the rolling direction (C direction: in the plane perpendicular to the axial direction in the steel pipe) is shown in FIG. The sound speed distribution is as shown. That is, the speed varies depending on the value of the refraction angle θ indicating the propagation direction of the ultrasonic wave.
[0031]
The ultrasonic velocity V (θ) can be approximated by equation (2).
[0032]
V (θ) = [(V₀+ V₄₅) / 2] + [(V₀―V₄₅) / 2] cos (4θ) ... (2)
However, V₀, V₄₅Are the sound speeds when the refraction angles are 0 ° and 45 °, respectively.
[0033]
Equation (2) can be determined by measuring the speed of sound at two different angles of refraction because there are two unknown variables. Here, if the sound speed distribution V (θ) can be determined, the flaw detection refraction angle θ is determined from the t / D of the steel pipe, so the sound speed V (θ) of the ultrasonic wave propagating in the flaw detection direction can be determined. As a result, the incident angle i of the ultrasonic wave incident on the steel pipe surface from the vibrator can be calculated according to Snell's law.
[0034]
Therefore, since the transducer reduction rate F by oblique flaw detection can be calculated from the values of the incident angle i and the refraction angle θ as described above, the vibrator width necessary for flaw detection can be determined, and the oblique angle under the optimum conditions for flaw detection. A probe can be selected.
[0035]
As described above, even if the steel pipe that is the object of flaw detection is made of a material having acoustic anisotropy, the surface defects present in the welded portion can be detected with high accuracy.
[0036]
Further, the present invention relates to surface defects existing in a plane including a radial direction and an axial direction in a welded portion existing in the axial direction of a steel pipe manufactured by bending a strip-shaped steel plate into a cylindrical shape. The present invention is applied to an ultrasonic flaw detection apparatus for a steel pipe that performs ultrasonic flaw detection using an oblique probe attached to an outer peripheral surface.
[0037]
And in order to eliminate the above-mentioned subject, in the ultrasonic flaw detector of the steel pipe of the present invention,Based on the ratio of the wall thickness to the outer diameter of the steel pipe, the ultrasonic wave that entered the steel pipe from the oblique probe enters the steel pipe perpendicularly to the planar defect. A flaw detection refraction angle calculating means for calculating a flaw detection refraction angle of the incident ultrasonic wave, a velocity characteristic in the propagation direction of the ultrasonic wave, and the calculated flaw detection refraction angle, the ultrasonic wave in the oblique probe with respect to the steel pipe The vibration for calculating the width of the vibrator in the oblique probe based on the incident angle calculating means for calculating the incident angle and the reduction ratio of the vibrator obtained from the relationship between the calculated incident angle and the flaw detection refraction angle. Child width calculation means andIt has.
The incident angle calculation means calculates the incident angle of the ultrasonic wave on the steel pipe in the oblique probe based on the velocity characteristic in the direction corresponding to the ultrasonic wave propagation direction and the flaw detection refraction angle in the acoustic anisotropy of the steel pipe. You may make it do.
[0038]
In the ultrasonic inspection apparatus for steel pipes configured as described above, the ultrasonic inspection refraction angle θ is obtained by the equation (1), as in the above-described ultrasonic inspection method for steel pipes.
[0039]
θ ＝ Sin^-1[1- (t / D)] (1)
Further, since the ratio t / D between the thickness t and the outer diameter D of the steel pipe is as small as about 3%, for example, the distance PWD between the welded portion probes between the surface defect in the welded portion and the oblique angle probe. Can be obtained by Eq. (3) by simple geometrical considerations.
[0040]
PWD = [(90−θ (degree)) / 360] Dπ (3)
By positioning the oblique angle probe so as to satisfy the calculated flaw detection refraction angle θ and the weld-part-probe distance PWD, the ultrasonic wave incident into the steel pipe is applied to the planar defect. Therefore, the S / N of the defect echo included in the echo signal output from the oblique angle probe can be improved.
[0042]
Also,The incident angle i is obtained by the above-described method, and further, the reduction rate F of the vibrator by the oblique flaw detection is obtained from the values of the incident angle i and the refraction angle θ. From this reduction rate F, the width W of the vibrator is obtained. Therefore, the transducer width W necessary for flaw detection can be determined, and an oblique angle probe having the optimum conditions for flaw detection can be selected.
[0043]
Yet another inventionInIn addition to the ultrasonic flaw detector for a steel pipe according to the above-described invention, probe driving means for applying a chirp pulse signal whose frequency is swept within a predetermined pulse width to the oblique angle probe at a constant period; An A / D converter that performs A / D conversion on the echo signal output from the angle probe, and a correlation operation is performed on the A / D converted echo signal using a chirp pulse signal. And a digital filter for outputting as a new echo signal.
[0046]
Here, digital using chirp pulse signalPerform filteringExplain the technical reason.
[0047]
In the oblique angle probe, a wedge made of, for example, acrylic is inserted between the vibrator and the inspection target surface in order to make the ultrasonic wave enter the inspection target surface while being inclined. As shown in FIG. 4, the reciprocal transmittance T of ultrasonic waves between acrylic and steel._SVaries greatly depending on the refraction angle θ or the incident angle i.
[0048]
Therefore, when oblique angle flaw detection is performed using an acrylic wedge, the flaw detection sensitivity changes according to the refraction angle θ of the ultrasonic wave. Then, as the refraction angle θ increases, the flaw detection sensitivity decreases, so that it becomes necessary to electrically amplify the echo signal output from the oblique angle probe.
[0049]
However, generally, when the detected echo signal is passed through an electrical amplification circuit, electrical noise increases, so the intensity of the echo signal from the object to be inspected is naturally amplified. The above-described electrical noise is added to the material noise such as the forest echo of the signal to reduce the signal-to-noise ratio (S / N).
[0050]
For this reason, in the past, flaw detection with a large refraction angle θ of 70 ° or more has been considered disadvantageous from the viewpoint of the S / N ratio even if the defect echo becomes strong.
[0051]
Therefore, for example, an autocorrelation operation is performed with a digital filter on the echo signal synchronized with the chirp pulse signal output from the oblique angle probe, which is output at a fixed period, that is, an echo signal where a defect echo exists at a fixed period. By doing so, material noise and electrical noise that do not have periodicity other than defect echoes can be removed.
[0052]
Therefore, since the echo signal can be amplified by an electric amplifier circuit with almost no electrical noise amplification, a sufficiently high S / N ratio is maintained even in flaw detection at a large refraction angle θ of 70 ° or more. Ultrasonic flaw detection can be performed in the state.
[0053]
Even in the synchronous addition averaging circuit, the signal level of material noise and electrical noise having no random periodicity can be obtained by averaging the echo signal over a plurality of periods in synchronization with the output period of the chirp pulse signal. The signal level of the defect echo which decreases and has a high periodicity increases.
[0054]
Therefore, similarly to the digital filter described above, since the echo signal can be amplified by an electrical amplifier circuit with almost no electrical noise amplification, ultrasonic flaw detection is performed while maintaining a sufficiently high S / N ratio. Can be implemented.
Still another invention is formed by bending a strip-shaped steel plate into a cylindrical shape, and then manufacturing a steel pipe by welding the side surfaces of the steel sheet, and the above-described ultrasonic flaw detection method for a steel pipe, And a second step of flaw detection of the steel pipe manufactured in the first step.
[0055]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
[0056]
FIG. 1 is a schematic configuration diagram of a steel pipe ultrasonic flaw detection apparatus to which the steel pipe ultrasonic flaw detection method of the embodiment is applied. The same parts as those in the conventional ultrasonic flaw detection method shown in FIG. Therefore, the detailed description of the overlapping part is omitted.
[0057]
The steel pipe ultrasonic flaw detector according to this embodiment has a surface caused by poor penetration occurring at the center in the welded portion of a UOE steel pipe having a wall thickness t = 22.2 mm and an outer diameter D = 1320.8 mm, for example. The defect is detected with ultrasonic waves.
[0058]
In FIG. 1, the probe selection calculation device 11 includes a thickness direction and a tube axis direction in the welded portion 3 in accordance with the wall thickness t and the outer diameter D of the steel pipe 1 that is the object to be inspected acquired from the host computer 13. An appropriate flaw detection refraction angle θ for flaw detection of the planar defect 4 spreading to the surface is calculated, and selection of the oblique probe 5 to be used to enable flaw detection at the refraction angle θ, that is, the ultrasonic wave 6 The incident angle i with respect to the outer peripheral surface 2 of the steel pipe 1 is calculated. In addition, when calculating the incident angle i, when the material of the steel pipe 1 has acoustic anisotropy, it considers that.
[0059]
In the probe selection calculation device 11, the flaw detection refraction angle θ is calculated from the thickness t and the outer diameter D using the above-described equation (1).
[0060]
θ ＝ Sin^-1[1- (t / D)] (1)
Further, the propagation direction of the ultrasonic wave 6 in the oblique probe 5 for performing the flaw detection at the flaw detection refraction angle θ, that is, the incident angle i with respect to the outer peripheral surface 2 of the steel pipe 1 is expressed by the equation (4) from Snell's law. Calculate using.
[0061]
sini / V_w= Sinθ / V (4)
However, V_w, V are the longitudinal acoustic velocity of the ultrasonic waves in the wedge and the transverse acoustic velocity in the steel plate (steel pipe 1).
[0062]
However, when the material of the steel pipe 2 has acoustic anisotropy, the transverse wave sound velocity of the object to be inspected (steel plate) is not uniformly represented by V, but varies depending on the propagation direction, and thus as described in FIG. Is represented by a function V (θ) of the refraction angle θ. Therefore, the calculation of the incident angle i uses equation (5) instead of equation (4).
[0063]
sini / V_w= Sinθ / V (θ) (5)
When the necessary incident angle i and refraction angle θ are calculated, the width W of the vibrator necessary for flaw detection is calculated.
[0064]
The width W of the vibrator necessary for flaw detection is determined by how much the directivity of the ultrasonic wave 6 is expanded. When looking at the vibrator in the oblique probe 5 from inside the steel pipe 2, the actual width W of the vibrator is measured. Instead, since it looks with the apparent width Wa of the vibrator, the spread of the ultrasonic wave 6 in the steel pipe 2 must be calculated with the apparent width Wa of the vibrator.
[0065]
The apparent transducer width Wa and actual transducer width W wait for the relationship of the following equation (6) when the transducer reduction ratio F is used.
[0066]
Wa = F · W (6)
Further, the reduction rate F of the vibrator can be calculated by the equation (7) from the relationship between the incident angle i and the refraction angle θ.
[0067]
F = cosθ / cosi (7)
On the other hand, in the steel pipe 2, the ultrasonic wave 6 is diffracted as if it was generated from a vibrator having an apparent width Wa, and has a directivity angle represented by the following equation.
[0068]
φ_-6= Sin^-1[0.433λ / Wa] (8)
However, φ_-6Is the direction (directivity angle) in which the echo signal detected by the single probe method becomes half of the maximum value. Λ is a wavelength.
[0069]
By conversely considering these, the actual transducer width W for expanding the ultrasonic beam necessary for flaw detection can be calculated.
[0070]
3 shows a -6 dB directivity angle φ from the maximum value in the steel pipe 2._-6The figure shows the transducer width required to set the angle to 4.6 °. The abscissa represents the refraction angle θ and is illustrated at a sound velocity of 100 m / s.
[0071]
The probe selection calculation device 11 displays and outputs the calculated refraction angle θ, incident angle i, and actual transducer width W to the display unit 12 and sends it to the PWD (distance between welded probe) calculation device 14. To do.
[0072]
Since the steel pipe 1 to be inspected in this embodiment is t = 22.2 and D = 1320.8, the flaw detection refraction angle θ is 79.5 °. Further, as shown in FIG. 2, the sound velocity V (79.5 °) of the ultrasonic wave 6 propagating in the direction of the refraction angle θ in the steel pipe 1 due to acoustic anisotropy is 3100 m / s. Since the speed of sound of the wedge (made of acrylic) incorporated in the oblique angle probe 5 is 2730 m / s, the incident angle i = 60 °. Furthermore, the directivity angle φ_-6The width W of the vibrator for achieving 4.6 ° is about 12 mm from FIG.
[0073]
When the incident angle i = 60 ° and the transducer width W = 12 mm are displayed on the display unit 12, the inspector displays a plurality of oblique angle probes having various incident angles i prepared in advance. 5 having an incident angle i = 60 ° and a transducer width W of 12 mm is selected and mounted on the holder 17 of the oblique probe 5.
[0074]
The operation of selecting and mounting the incident angle i may be automated by adopting an automatic variable angle probe as the oblique angle probe 5.
[0075]
Next, the PWD calculation device 14 determines the position of the oblique probe 5 from the values of the flaw detection refraction angle θ, the wall thickness t, and the outer diameter D transferred from the probe selection calculation device 11 as follows. .
[0076]
The position of the bevel probe 5 when the bevel defect 5 is used to detect a surface defect 4 caused by insufficient penetration in the weld 3 of the steel pipe 1 made of UOE material is determined by the weld probe. It is expressed by the inter-child distance PWD and is the above-described equation (3).
[0077]
PWD = [(90−θ (degree)) / 360] Dπ (3)
If each value of the steel pipe 1 to be inspected in this embodiment, D = 1320.8, θ = 79.5 ° is substituted into this equation, PWD = 121 mm.
[0078]
The value of PWD calculated by the PWD arithmetic unit 14 is transferred to the manipulator 15. The manipulator 15 positions the holder 17 so that the ultrasonic wave 6 enters the steel pipe 1 from the oblique probe 5 at a position of PWD = 121 mm from the center position of the welded portion 3 detected by the weld seam detector 15. And fix.
[0079]
In this way, by determining the position of the oblique probe 5 with respect to the welded portion 3, the ultrasonic wave 6 propagated in the steel pipe 1 is perpendicularly incident on the planar defect 4 existing in the welded portion 3. .
[0080]
As can be understood from the reciprocal transmittance characteristic diagram shown in FIG. 4, in the steel pipe 1 to be inspected according to the present embodiment, the flaw detection refraction angle is as large as θ = 79.5 °, and therefore − There is a sensitivity reduction of 3 dB.
[0081]
For this reason, in the embodiment apparatus, the correlation calculation technique using the chirp pulse signal and the synchronous addition averaging technique are used to amplify the echo signal by the amplifier circuit without increasing the noise, thereby preventing the sensitivity reduction.
[0082]
Next, these signal processing techniques will be described.
[0083]
In FIG. 1, the synchronization signal generator 18 has a transmission period T₀Is sent to the transmission memory 19, the FIR filter 26 and the synchronization weighted averaging circuit 27. As shown in FIG. 5, the waveform calculator 20 generates a chirp wave that sweeps the frequency within a predetermined pulse width T, and supplies the chirp wave to the transmission memory 19 and the FIR filter 26. This chirp wave is calculated using a computer and is expressed by equation (9).
[0084]

Where f_CIs the center frequency, B is the frequency sweep bandwidth, and T is the pulse width.
[0085]
The chirp wave stored in the transmission memory 19 is transmitted from the synchronization signal generator 18 by the transmission period T.₀Is read in synchronism with the transmission timing signal a having the following structure, converted into an analog signal by the D / A converter 21 and further amplified by the transmission amplifier 22,₀Is applied to the oblique probe 5 through the diplexer 23 as a chirp pulse signal b having
[0086]
The ultrasonic wave 6 incident on the inside of the steel pipe 1 as the object to be inspected through the acoustic contact medium such as water from the oblique angle probe 5 is reflected by the planar defect 4 in the welded portion 3 and again is inclined. Received by the probe 5. The echo signal c output from the oblique angle probe 5 is amplified by the receiving amplifier 24 via the diplexer 23, converted into a digital signal by the A / D converter 25, and is an FIR filter (finite) as a digital filter. It passes through the impulse response filter) 26 and is input to the synchronous addition averaging circuit 27.
[0087]
As shown in FIG. 6, the FIR filter 26 includes 128 multipliers 26a, 128 adders 26b, and 128 delay units 26c. Each coefficient C applied to each multiplier 26a of the FIR filter 26_o~ C₁₂₇The signal of the chirp waveform generated by the waveform calculator 20 is set as reference signal data. Each delay unit 26c transmits a transmission period T to the input signal._oA time delay corresponding to is output.
[0088]
In the FIR filter 26 having such a configuration, the waveform x (τ) of the echo signal c discretized into a digital signal and a reference waveform for performing a correlation operation are sampled (discretized) at a certain sampling frequency. ) In this example, each discretized data value has 128 coefficients C_o~ C₁₂₇Are respectively input to one of the multipliers 26a.
[0089]
On the other hand, the transmission cycle T from the input end₀Discretized reception data x (τ) input every time is directly supplied to the other input of each multiplier 26a, and the reference data (coefficient) C_o~ C₁₂₇Are multiplied individually. Leading coefficient C₁₂₇Each multiplication result excluding the multiplication result is supplied to one of the inputs of the corresponding adder in which 127 delay devices 26c and adders 26b are alternately connected in series.
[0090]
And the leading coefficient C₁₂₇Is directly supplied to the first delay device 26c alternately connected in series, and the coefficient C is added to the other input of the adder 26b connected in series in the subsequent stage of the delay device 26c.₁₂₆The multiplication result is supplied. Then, the output of the last adder 26b in the serial combination becomes the correlation calculation output y (τ).
[0091]
Now input signal x_{i (j)}, Coefficient (reference data) as C_(k), Output signal y_{i (j)}, Tap number v_c, Where i is the repetition of the flaw detection signal and n is the number of flaw detection signal data points in one cycle, the output signal y_{i (j)}Is obtained by performing the convolution operation of equation (10).
[0092]
[Expression 1]

[0093]
When the reference signals are reversed in the equation (10), the following equation (11) is obtained.
[0094]
[Expression 2]

[0095]
As a result, the correlation calculation for the echo signal (received signal) c can be performed using the FIR filter 26. In this correlation operation, the cross-correlation is calculated while shifting the reference signal and the received signal by j.
[0096]
The specific operation of the FIR filter 26 will be described with reference to the waveform diagram shown in FIG.
[0097]
In FIG. 7, time τ₁This point corresponds to the position of j = 0 in the equation. First, τ_iThe correlation calculation between the received signal (echo signal) and the reference signal is performed at 0 to N._c-1Only the amount of data is performed. Where N_cIs the tap length of the FIR filter 26, that is, the coefficient C_o~ C₁₂₇Is the number of The correlation result is output as the lowest signal in the figure, that is, as a new echo signal d. This τ_iAt this point, the reference signal and the received signal (echo signal) are not similar, so the output is almost zero.
[0098]
Next, increase j by 1 and τ in the figure₂, Τ_Three, ... are performed sequentially. As a result, the correlation signal having the maximum peak is obtained at the point where the phase of the echo in the received signal (echo signal) matches the phase of the reference signal (approximately the center of the time axis in the figure). As a result, the pulse width of the chirp wave is compressed, and the electrical noise signal uncorrelated with the reference signal is greatly reduced.
[0099]
The echo signal d pulse-compressed by the correlation processing in the FIR filter 26 is subjected to synchronous addition averaging processing in the next synchronous addition averaging circuit 27.
[0100]
FIG. 8 is a detailed configuration diagram of the synchronous addition averaging circuit 27.
[0101]
The synchronous addition averaging circuit 27 is roughly divided into a first memory 28, an adder 29, write / read address generators 30a and 30b of the first memory 28, a subtractor 31, a second memory 32, The second memory 32 includes write / read address generators 33 a and 33 b and a subtractor 34.
[0102]
The first memory 28 has a transmission cycle T of the transmission timing signal a from the synchronization signal generator 18.₀Each of the received signals (echo signal d) received from the FIR filter 26 has a capacity capable of storing data for N + 1 cycles or more. In this example, the received signal (echo signal d) for each cycle contains M pieces of data. Each data is composed of 8 bits.
[0103]
Further, the second memory 32 stores each reception cycle T.₀In this example, it is assumed that M pieces of 16-bit data are stored for each of M pieces of data.
[0104]
Next, the operation of each part will be described. Initially, nothing is stored in the first memory 28 and the second memory 32 in an initial state where no reception signal (echo signal d) is input from the FIR filter 26. First cycle T₀When the received signal (echo signal d) is input, this received signal is stored in the first area of the first memory 28 and is added to one of the inputs of the adder 29. In this first cycle, the signal read address RA of the first memory 28 is₁Since nothing is stored in the area designated by the second memory 32 and the second memory 32, the received signal (echo signal d) applied to one of the inputs of the adder 29 passes through the adder 29 as it is, and is further subtracted. 31 is passed through and stored in the second memory 32.
[0105]
Second cycle T₀When the received signal (echo signal d) is input, the received signal of the second cycle and the received signal (echo signal d) of the previous cycle stored in the second memory 32 are added by the adder 29. Is added. This addition result passes through the subtractor 31 and is stored in the second memory 32. The second received signal (echo signal d) is also stored in the second area of the first memory 28.
[0106]
Thereafter, until the reception signal (echo signal d) for the average number N cycles is input, the reception signal (echo signal d) is sequentially added to the second memory 32 in this manner. The operation of sequentially storing received signals in the corresponding area continues.
[0107]
When a reception signal (echo signal d) having an average number of N cycles or more is input, a reception signal (echo signal d) N cycles before the newly input reception signal (echo signal d) is stored in the first memory 28. Signal read address RA_iAre read out from the designated area and supplied to the subtractor 31 as a subtraction signal.
[0108]
For this reason, the adder 29 adds the signal stored in the second memory 32 and the newly input reception signal (echo signal d) and outputs the addition result. Since the subtractor 31 subtracts the reception signal N cycles before read from the memory 28, the output signal of the subtractor 31 is always the sum of the reception signals for N cycles.
[0109]
The addition result of the reception signals for N cycles is stored in the second memory 32 and is also input to the divider 34 and divided by N. Therefore, the output of the divider 34 is an ultrasonic pulse transmission cycle. T₀Each time, the synchronous addition average result of the reception signals (echo signal d) for N periods is output to the next display unit 35 as a new echo signal e.
[0110]
Here, the timing signal supplied to the write address generators 30a and 33a and the read address generators 30b and 33b for storing and reading the received signal (echo signal d) is supplied from the synchronization signal generator 18. A transmission timing signal a is used.
[0111]
By adopting the synchronous addition averaging circuit 27 configured as described above, the echo signal is averaged over a plurality of periods in synchronization with the output period of the chirped pulse signal, thereby not having a random periodicity. The signal level of noise or electrical noise is lowered, and the signal level of defect echoes having periodicity is raised.
[0112]
More specifically, in the ultrasonic flaw detector of this embodiment, since the pulse compression of the chirp wave is used, the wide-angle oblique probe 5 that enables transmission / reception of the ultrasonic wave 6 in a wide frequency band is provided. Used. The wide-angle oblique probe 5 generally has low sensitivity, but since the transmission pulse width is short because of the use of chirp pulse compression, each pulse of noise echo reflected / interfering due to surface properties is per pulse. Can be made larger.
[0113]
As a result, it is possible to reduce noise echoes compared to the case of the narrow-band probe by averaging the signals synchronously. Therefore, the S / N ratio of the defect echo included in the echo signal can be greatly improved.
[0114]
FIG. 9 is an enlarged view of the noise echo waveform caused by the surface texture. FIGS. 9A to 9D show the narrow band probe when the flaw detection position is changed to 1 to 4. FIG. Each waveform example by a broadband probe is shown. Further, (c) in FIG. 9 is a waveform obtained by averaging the four waveforms (a) to (d). Thus, the S / N ratio can be improved by averaging the echo signals d.
[0115]
【The invention's effect】
As described above, in the steel pipe ultrasonic flaw detection method and the steel pipe ultrasonic flaw detection apparatus of the present invention, the bevel probe is used so that the ultrasonic waves are perpendicularly incident on the surface defects in the welded portion of the steel pipe. It is arranged.
[0116]
Therefore, the echo from the defect can be detected strongly, and it is difficult to generate the echo from the weld overlay, and the path length of the ultrasonic wave propagating in the steel pipe is relatively long. Therefore, the S / N ratio of the defect echo included in the echo signal output from the oblique angle probe can be improved.
[0117]
In addition, if the ratio of wall thickness to outer diameter (t / D) of the steel pipe is small, it is necessary to use a probe with a large refraction angle. Pulse compression technologyUsingSignal processing is performed on the echo signal output from the oblique probe.
[0118]
Therefore, the signal level of the echo signal can be recovered by the amplifier circuit without increasing the electrical noise.
[0119]
Furthermore, even when acoustic anisotropy is observed in the material of the steel pipe to be inspected, it is possible to normally detect flaws by selecting a diagonal probe satisfying the required flaw detection refraction angle by obtaining the sound velocity distribution.
[Brief description of the drawings]
FIG. 1 is a diagram showing the configuration of a steel pipe ultrasonic flaw detector to which a steel pipe ultrasonic flaw detection method according to an embodiment of the present invention is applied.
FIG. 2 is a graph showing the velocity distribution characteristics of ultrasonic waves of an acoustically anisotropic material.
FIG. 3 is a diagram showing a relationship between a refraction angle and a necessary transducer width in an oblique probe.
FIG. 4 is a diagram showing the reciprocal transmittance at an oblique incidence of ultrasonic waves.
FIG. 5 is a waveform diagram showing a chirp pulse signal applied to an oblique probe.
FIG. 6 is a block diagram showing a schematic configuration of an FIR filter.
FIG. 7 is a diagram for explaining the operation of the FIR filter.
FIG. 8 is a block diagram showing a configuration of a synchronous addition averaging circuit
FIG. 9 is a diagram showing an example of a noise echo waveform and an averaged waveform by narrow band and wide band probes;
FIG. 10 is a diagram showing a general flaw detection method for a steel pipe using a bevel probe.
[Explanation of symbols]
1 ... Steel pipe
3 ... welded part
4 ... Surface defects
5 ... Bevel probe
6 ... Ultrasonic
11 ... Probe selection calculation device
12, 35 ... display section
13 ... Host computer
14 ... PWD arithmetic unit
15 ... Manipulator
16 ... Weld seam detector
17 ... Holder
18 ... Synchronization signal generator
19: Memory for transmission
20 ... Waveform calculator
21 ... D / A converter
22: Transmitter amplifier
23 ... Diplexer
24 ... Receiving amplifier
25 ... A / D converter
26 ... FIR filter
27. Synchronous addition averaging circuit
16 manipulator
28 ... First memory
29 ... Adder
30a... First memory write address generator
30b ... Read address generator of the first memory
31 ... Subtractor
32 ... Second memory
33a ... Second memory write address generator
33b ... Second memory read address generator
34 ... Divider

Claims (6)

A planar defect existing in a plane including a radial direction and an axial direction in a welded portion existing in the axial direction of a steel pipe manufactured by bending a strip-shaped steel plate into a cylindrical shape is attached to the outer peripheral surface of the steel pipe. In an ultrasonic flaw detection method for steel pipes that are flaw-detected by using an oblique probe,
Based on the ratio of the wall thickness to the outer diameter of the steel pipe, the ultrasonic wave incident from the oblique probe into the steel pipe is perpendicularly incident on the planar defect. Set the flaw detection angle of ultrasonic waves entering the steel pipe,
Based on the velocity characteristics in the propagation direction of the ultrasonic wave and the set flaw detection refraction angle, the incident angle of the ultrasonic wave in the oblique probe to the steel pipe is set,
A steel pipe characterized in that a width of a vibrator in the oblique probe is determined based on a reduction ratio of the vibrator obtained from a relationship between the set incident angle and flaw detection refraction angle. Ultrasonic flaw detection method. The incident angle of the ultrasonic wave on the steel pipe in the oblique probe is determined based on the velocity characteristic in the direction corresponding to the propagation direction of the ultrasonic wave in the acoustic anisotropy of the steel pipe and the flaw detection refraction angle. An ultrasonic flaw detection method for a steel pipe according to claim 1. A planar defect existing in a plane including a radial direction and an axial direction in a welded portion existing in the axial direction of a steel pipe manufactured by bending a strip-shaped steel plate into a cylindrical shape is attached to the outer peripheral surface of the steel pipe. In a steel pipe ultrasonic flaw detector that performs ultrasonic flaw detection using an oblique probe,
Based on the ratio of the wall thickness to the outer diameter of the steel pipe, the ultrasonic wave incident from the oblique probe into the steel pipe is perpendicularly incident on the planar defect. A flaw detection refraction angle calculation means for calculating a flaw detection refraction angle of the ultrasonic wave incident on the steel pipe, and a supersonic wave in the oblique angle probe based on the velocity characteristic in the propagation direction of the ultrasonic wave and the calculated flaw detection refraction angle. An incident angle calculating means for calculating an incident angle of the sound wave to the steel pipe;
Vibrator width calculation means for calculating the width of the vibrator in the oblique angle probe based on the reduction ratio of the vibrator obtained from the calculated relationship between the incident angle and the flaw detection refraction angle. An ultrasonic flaw detector for steel pipes characterized by comprising: The incident angle calculation means is based on the velocity characteristic of the direction corresponding to the propagation direction of the ultrasonic wave in the acoustic anisotropy of the steel pipe and the flaw detection refraction angle, and the ultrasonic wave in the oblique angle probe with respect to the steel pipe. 4. The ultrasonic flaw detector for steel pipes according to claim 3, wherein an incident angle is calculated. Probe driving means for applying a chirped pulse signal whose frequency is swept within a predetermined pulse width to the oblique angle probe at a constant period;
An A / D converter for A / D converting the echo signal output from the oblique angle probe, and performing a correlation operation on the A / D converted echo signal using the chirp pulse signal; The ultrasonic flaw detection apparatus for a steel pipe according to claim 3 or 4 , further comprising a digital filter that outputs a calculation result as a new echo signal. A first step of producing a steel pipe by bending a strip-shaped steel plate into a cylindrical shape and then welding the side surfaces of the steel plate;
A second step of flaw detection of the steel pipe manufactured in the first step by the ultrasonic inspection method for a steel pipe according to claim 1 or 2.
The manufacturing method of the steel pipe characterized by having.

Images