I got the task to find out if there`s a difference in mechanized UT of pipes under unfilled and filled conditions (by water) as a 1st measurement (basis) for the in-service inspection of Nuclear power plants. Who knows some literature and where were made such examinations? I hope You, the ndt.net users can help me. Thanks in advance. Joerg (e-mail please see above) phone: +49 9131 18 7895 fax: +49 9131 18 2053
04:47 Feb-25-1999 Rainer Meier R & D retired from intelligeNDT Systems & Services, Germany, Joined Nov 1998 15
Re: UT of filled or unfilled pipes : I got the task to find out if there`s a difference in : mechanized UT of pipes under unfilled and filled : conditions (by water) as a 1st measurement (basis) for : the in-service inspection of Nuclear power plants. : Who knows some literature and where were made such : examinations? : I hope You, the ndt.net users can help me. : Thanks in advance. : Joerg (e-mail please see above) : phone: +49 9131 18 7895 fax: +49 9131 18 2053
05:27 Feb-25-1999 Rainer Meier R & D retired from intelligeNDT Systems & Services, Germany, Joined Nov 1998 15
Re: UT of filled or unfilled pipes Dear Mr. Neubert,
the echo of a backwall or a reflector is influenced by the media behind the backwall/reflector. In practice this effect can normally be neglected. In the case of a longitudinal wave in a steel pipe wall, the reflection coefficient changes from nearly 1 (air filling) to 0.937 for the water filled pipe. That means, that the loss of backwall amplitude due to the water filling is less than 0.6 dB. This can be neglected compared with other effects to the signal amplitude, e.g. changing coupling conditions.
In the case of shear waves, the amplitude loss due to water filling might be slightly higher, but does not exceed 1 dB. (The shear wave impedance is smaller then the long wave impedance).
Additional, water film or water drops can cause unwanted echos. If there is a waterfilm on the backwall of a component, the echo from the waterfilm surface may interfere with the components backwall echo. But due to the fact, that only 12 percent of the sound energy penetrates the backwall, this effect is also not significant.
The normally used UT-waves are only minor influenced by the water filling. This also includes creeping waves and long-shear wave conversion. Rayleigh waves are much more influenced, through a high damping caused by the water and through reflections caused by water drops.
The above mentioned facts are valid for steel pipes. If you think about other materials, e.g. aluminum pipes with a much lower acoustic impedance, the water filling will have a higher influence!
: I got the task to find out if there`s a difference in : mechanized UT of pipes under unfilled and filled : conditions (by water) as a 1st measurement (basis) for : the in-service inspection of Nuclear power plants. : Who knows some literature and where were made such : examinations? : I hope You, the ndt.net users can help me. : Thanks in advance. : Joerg (e-mail please see above) : phone: +49 9131 18 7895 fax: +49 9131 18 2053
So shear wave weld inspection of a filled water vessel shouldn't be an issue? You say that 12 percent of the amplitude is lost off the back wall reflection but the say only 1 db is lost for shear wave. I don't understand how that is....if 12 percent is lost and your running at 50 db it would be 12 percent of 50db right? That's alot mor then 1 db.
Joerg, we had a similar task 30 years ago, to assess the effect of water in the vessel when doing periodic inspection. The easiest option was to evaluate the effect on our DAC curve. Of course in the first half skip there is no effect, only after reflection off the steel water interface does the problem show itself. To determine how significant this is there is an easy setup. We placed our calibration block in a tray (about 50mm deep) large enough to hold the calibration block. Under each end of the block we placed a square bar that allowed us to raise the block off the tray surface about 10mm. Then draw your DAC. You can draw it first with no water then again after adding a small amount to contact the skip surface and see the difference in effect. There will be a difference based on the angle used!
What was the loss of sound due to transmission in the water? Would you consider the effect unsignificant? Does it exceed 1 dB for shear wave inspections? When i calculate the reflection coeffecient of a shear wave through steel into water nearly 20 percent of the sound is transmitted into the water, I would think this would be significant due to the fact that when the sound beam strikes a defect it will propagate back to the transducer in the same path it came, meaning a defect in the second leg would have a sound beam that contacts the water surface twice ( Lossing 40 percent of its sound, or nearly 6 db). Is this safe to assume Ed?
You say roughly a 2 db loss, with a 40 percent transmission loss in the water wouldn't it be closer to a 5 db loss? Can you explain? Also, I'd your calibrating for a weld that is 3/8" thick on a 3/4" block how can you compensate for the water loss in the second half skip due to the fact you will not get to the water on the calibration block until 3/4" deep but will be in the second leg of the weld after 3/8"???
Mark, have you tried the simple setup I suggested? It took me all of 5 minutes to verify on a single point. You will note from Krautkramer's echo-transmittance curves that the equation in the third English edition (A.17) is for the incidence from the liquid. But the result can be derived for incidence from the solid. In the simple perpendicular incidence case the Java applet on http://www.ndt-ed.org/EducationResources/CommunityCollege/Ultrasonics/Physics/applet_2_7/applet_2_7.htm predicts a 2dB for the one-way path for the 0 degree condition reflection! In pulse-echo the beam returns along the same path so another 2dB would be lost. The setup I used was for a 60° on the IOW block so the loss is not the same as for 0 degree. I peaked the signal at 100% and held the probe in place as I placed it on the water...the signal dropped to about 81% (near enough to 2dB). The math is nice, but it is always useful to validate by actually doing the hands-on.
Andrey, good idea! I uploaded the model image summary. I ran 2 probe angle models, 45° and 60° with a direct path to a 3mm SDH and a skip path using a 50mm steel block model. When plotting the echo dynamics as the probe is moved past the SDH, the peaks indicate the points we would use to draw a DAC. When water is used at the skip point we see the amplitude drop 1.6dB for the 45° beam and 1.4 for the 60° beam. In my earlier post I made a physical measurement with the 60° beam and said the signal dropped from 100% to 81%. That is 1.8dB from the physical measurement. Civa predicts 1.6dB. Further to Joerg's concerns for literature search, there may be a report on a similar effect in CANDU feeder tube inspections. There they made provision for amplitude corrections at each skip point on the steel-water interface. But I think the Civa results could be used in technical justifications.