What is causing this miscalibration on my IIW block?
Using a code AWS shoe, I have been calibrating on an IIW Block. However, my depth readings on the .060 side drilled hole have been off. Here is how I have been calibrating, correct me if you see any mistakes.
First I place a piece of masking tape over the BIP line on the shoe (the majority of my work is with the 70 degree angle). I put sonotrace couplant on the BIP hash marks of the IIW block, then place the transducer on the IIW block. Using both hands, I move the transducer over to one side so that left side of the IIW block and left side of the transducer match up evenly. Then I move the transducer back and forth, making sure it is exactly parallel with the IIW block, to peak up on the signal from the large radius. When I get a maximized signal, I mark the masking tape on the side of the shoe directly above the main hash mark.
Keeping the main signal maximized, and keeping my mark on the masking tape aligned with the main hash mark on the IIW block, I put my A gate over what will be the 4" signal, and the B gate over the 9" signal. I keep my gates in flank mode. I usually put my A gate at 20% and my B gate at 10%" full screen heighth. Then I use the probe delay to dial in the A gate to 4", and the velocity to dial in 9" for the B gate. After going back and forth dialing those in several times, closing the distance until I am nearly exactly 4.000" and 9.000", I then check my angle on the perspex cylinder in the IIW block. When I find out what my true angle is, I go into the trig function of the USN 60 and change my angle to that.
Then I peak up on the .060 diamter side drilled hole in the block. The center of the hole is at .600". Subtracting the radius (.030) and accounting for the trigonometry of the angle I am using (another .010), I should get a depth of around .560". But this is very rarely the case. Most of the time I get at least .600". Sometimes I am as much as .100" off.
How do I correct this? If I am confident in my initial 4" and 9" calibration, and I feel that my angle and bip check are correct, what do I do?
16:41 Dec-15-2009 Phil Herman Sales, - Manufacture of NDT Reference Standards/Test Blocks PH Tool Reference Standards, USA, Joined Oct 1999 79
Re: What is causing this miscalibration on my IIW block?In Reply to Juan Amado at 15:40 Dec-15-2009 .
You have an IIW-Type 2 block, and are correct that the two radii are 2" and 4" from the exit mark. David is using an IIW-Type 1 block, which has only one the 4" radius on the block end, but also has the shallow 1" radius slot that allows calibration at 9" by accumulating two trips to the 4" radius together with one to the 1" radius, for a total of 9". Something else is going on here. Is the .060" diameter hole in the right location?
17:07 Dec-15-2009 Godfrey Hands Engineering, PRI Nadcap, United Kingdom, Joined Nov 1998 284
Re: What is causing this miscalibration on my IIW block?In Reply to Phil Herman at 16:41 Dec-15-2009 .
If you look at the calibrated range (4 to 9 inches), the instrument and probe will likely be quite linear within that range, and probably accurate to 1 to 2% of the total range, i.e. if we assume the range calibrated is 0 to 10 inches, 1% is then 0.1 inch.
This is the accuracy (or inaccuracy) that David is experiencing. Remember that this 0.6 inches measured is outside of the range that has been calibrated, so results are extrapolated.
Probably if a shorter range was used by calibrating with another block (example range 0 to 2 inches), then the errors should be smaller, even when we look at 1 to 2% accuracy, or 0.02 inches error.
15:10 Dec-18-2009 Roger Duwe NDT Inspector, API-510, 570, 653 MISTRAS, USA, Joined Jan 2009 148
Re: What is causing this miscalibration on my IIW block?In Reply to john at 20:25 Dec-16-2009 .
I use the same calibration sequence you are trying to use. There are two possible answers for your problem.
1) The IIW block you are using is dimensionally inaccurate. Easiest way to check is to get a new, high-quality shoe on a high-quality 'ducer. I use a new 70deg G-E [Krautkramer] shoe on a 1/2" dia [13 mm] round G-E 'ducer. Make sure that your show was cut for the material of your block [carbon steel or stainless steel] I prefer carbon steel. The Beam Index Point should be less than 1/32" [1mm] form the scribed line on the show. The angle should be less than 1/2 degree form the specified shoe angle. If either are off, your IIW block is probably badIf both are off, your block is bad. Confirm your resulta on a known good block -- Krautkramer, Panametrics, or PH Tool.
2) Your measured angle is off, skewing the trigonometric results, making the 0.600" deep hole appear off. After I have an excellent distance and velocity calibration from a good IIW block [auto-rated twice in a row] as accuately as possible,
and then look at the calculated depth of the .600 hole. If it is within .040 of actual, I leave the angle setting alone. If off more than that, I 'tweak' the shoe angle 0.1 degrees at a time. Small changes in the assumed angle of the shoe make big changes in the calculated depth of the .600 hole, especially at 70 degrees.
19:12 Dec-18-2009 Ed Ginzel R & D, - Materials Research Institute, Canada, Joined Nov 1998 1197
Re: What is causing this miscalibration on my IIW block?In Reply to David Bunch at 01:21 Dec-15-2009 (Opening).
Accuracy in UT is a much debated topic (especially when it is in reference to sizing). But the accuracy being discussed here is relative to position and this is a bit different from sizing (although for max-amp sizing techniques this accuracy would be pertinent).
There are several factors to consider when looking at the degree of accuracy that may be expected. Not the least of these is the process of digitisation. You noted that you are using a USN60. Assuming this has a clock rate of 100MHz, at an assumed acoustic velocity of 3230m/s the positional uncertainty will be the distance equivalent to the 10ns sampling, or about 0.006" in pulse-echo. You are probably relying on a digital gate readout to determine values. Typically these may be set using either the peak or shoulder (leading edge) of the pertinent signal. Using the leading edge may introduce variations depending on the smoothing, filtering and pulse-shape characteristics of the reflected pulse and the threshold set.
Your comparison of positions (to the SDH) and calibration of time-base and delays uses a process of maximising a response. The range is set using an integrated response to a large target (i.e. the 4" radius at 4" and 9" distances). To make things work out on this display you have manipulated the acoustic velocity! There is no guarantee that the "adjusted" velocity is the true velocity. (A 1999 NIST report measured a large selection of IIW blocks and found them to have a nominal transverse velocity of 3230m/s +/- 30m/s). You then move to a small diameter cylindrical reflector very near the probe exit point. The returned signal is no longer a large integration of the spreading beam by a focusing reflection. It is instead a divergent reflection from a small region within the beam. This small cylinder is now able to exhibit the effects of D/lambda (i.e. probe edge diffractions) so you could be measuring a maximised response from a lower angle portion of the beam as a result of echo-transmittance effects.
Another factor related to the process of using the maximum response on a relatively close target is the influence of the near field. You provided no details of the probe you are using but if most work is at 70° I would expect a 5MHz probe is useful and a 12.5mm diameter element common. With a typical wedge this could place the 15mm (0.6") depth SDH in the IIW block inside the near zone.
You made another statement about the estimated depth: "Subtracting the radius (.030) and accounting for the trigonometry of the angle I am using (another .010), I should get a depth of around .560"."
The depth read by the gated signal is a result of the computation that equates the soundpath to the hypotenuse of a right-triangle. A 70° ray impinges on the cylinder at some point slightly lower than the apex of the arc nearest the surface. That would suggest you should add the 0.01" to the 0.570" instead of subtracting it. In fact, the correction for perpendicular incidence of the 70° beam on the cylinder 0.06"diameter would be 0.03-(0.03*sin20°) = 0.02. i.e. if your upper edge of the SDH is 0.570" from the surface, the point of impingement of a 70° beam is indeed 0.590" down. If most of the time you are getting 0.600" I would think that your accuracy is adequate, especially considering the potential collection of errors that could be causing problems.
Godfrey's response was perhaps the most relevant to the application. Instruments are built to operate within a tolerance. Godfrey pointed out what must be the most obvious first assumption, i.e. that a 10 inch range would be expected to have a basic tolerance of about 0.1 inch (i.e. 1%). The best advice was the suggestion to calibrate the range comparable to that to be used. You noted that MOST of your work is with a 70° angle beam. Therefore we might expect most of the sound-paths of interest to be short. The V2 block with its 25mm and 75mm (or if you prefer the Imperial 1" and 3") radii would be a good start.
Re: What is causing this miscalibration on my IIW block?In Reply to Ed Ginzel at 19:12 Dec-18-2009 .
Wow! Great response. I am definitely keeping a copy of this.
I just went into AutoCad and drew out the .060 .600 deep hole, and measured the depth. It came to .5896 depth using a 70 degree angle. So actually my calibration has been fairly decent. I just wish I could say the same about my math haha.
I have been using the the large 3/4" x 3/4" (2.25 mhz) AWS transducer and wedges. The small 3/8" and 1/2" ASME transducers and shoes calibrate very easy. My IIW block is unfortunately in horrible shape. There is not a single spot on the block that doesn't have a scratch. I am sure that has something to do with it too.
11:10 Dec-19-2009 Nigel Armstrong Engineering, - Specialist services United Kingdom, Joined Oct 2000 1094
Re: What is causing this miscalibration on my IIW block?In Reply to David Bunch at 00:24 Dec-19-2009 .
First, I echo David - what a great response from Ed! And I agree with Ed re Godfrey's initial practical hitting the nail on the head - wrong calibrated range for the test requirement.
Ah David, the travails of being a modern NDT tech - expensive software such as AutoCad and hardware contrast with the calibration block that should be assigned to the scrap-heap! I cant wait for the arrival of "virtual" calibrations and wireless NDT!
It used to be so simple - pencil, paper and probe beam axes on a clear RT acetate for planning, 25mm and 100m on-screen from the V2, check the 50mm response maxed in the right place for caling, set the grass to 2 - 3mm height for scanning. Its just been an exponential curve of increasing complexity ever since those halcyon days!