·Home ·Table of Contents ·General

Check of Purity Grade as a Statistical Process Control by means of an Automatic Ultrasonic Testing Equipment Dr.-Ing. Roman Koch, NUKEM Nutronik GmbH, Industriestraße 13, 63748 Alzenau, Germany Contact

Normally the ultrasonic check of purity grade is realised off-line of the production process by means of statistically taken laboratory samples scanned with a high spatial resolution in both scanning directions by means of a high frequency, high spatial resolving ultrasonic transducer. A check of the complete production of purity grade can be performed by integrating such a type of testing method into a rotating ultrasonic testing machine to get an on-line information. A comparable type of transducer with a long small focal tube is testing the material characteristic in the way of an open helix in addition to the normal flaw detection and wall thickness measurement. A quite constant sensitivity over the specified wall thickness range is reached due to a optimised transducer design. Special properties of evaluation are necessary to integrate the purity grade check into the standard ultrasonic testing procedure. Here, a special software was realised for a digital ultrasonic unit. A statistic evaluation for the recorded over threshold events is performed both for an individual tube and the complete tested lot. A continual and homogeneous inspection of the production is possible using such a procedure.
Additionally, for verification of the measurement results, an analytical and complete retest of the tubes with indications is performed by a laboratory equipment using the same technique.



Pic 1: Principle of Check of Purity Grade by Ultrasound

Principally the check of purity grade - as to say the detection of porosity, small flaws or micro inclusions - by means of ultrasound is realised using a normal incident, high frequency transducer with a high spatial resolution, as shown in picture 1. An additional requirement for the transducer will be set on the variation of the focal spot size with distance. It is necessary to inspect over a large area of wall thickness variation with the same sensitivity to get comparable results of the purity grade check with the same valuation of defect size over depth. But for a normal high focused transducer the wished small spot size can only be realised in a small area around the focal distance. For this application a transducer was especially optimised to produce a longer focal area, a so called focal tube. A 20 MHz transducer with a small crystal diameter and with the focal spot in 50 mm distance at a water path of 30 mm is used. The sound field and HF-signal characteristic of such a transducer is presented in picture 2. Picture 4 shows the amplitude signal of a natural defect over time-of-flight between the interference echo and the first back wall echo measured with such a transducer. In comparison also a 15 MHz transducer was checked on the same defect. The characteristic of this transducer is shown in picture 3 and the measurement result is presented in picture 5. Using such a transducer the time-of-flight resolution is reduced but there is a better signal-to-noise-ratio, a higher absolute amplitude value and due to the lower necessary gain the ringing of both interference echo and back wall echo is reduced. As a result such a 15 MHz transducer will be easier to handle in an automatic testing equipment. The differences between the two transducers on the inspection results will be shown later on the C-scan results of the laboratory equipment.


Pic 2: Sound Field, HF-Signal and Spectrum 20 MHz Transducer	Pic 3: Sound Field, HF-Signal and Spectrum 15 MHz Transducer
Fig 4: Defect Signal 20 MHz Transducer
Fig 5: Defect Signal 15 MHz Transducer

As a standard, a value for purity grade is received by a statistical evaluation of the over threshold events of a C-scan using one or two different threshold values to distinguish between different defect sizes. Sometimes also the defect length is taken into account.


Pic 6:Modular digital Ultrasonic Instrument with Function Setting by Software Down-Load

Fig 8: Testing Equipment ROTA 90Z including Bench

A modular ultrasonic unit (picture 6) where the hardware of each channel is the same and the individual function is determined only by the down-load of a special software was used to integrated such a measuring method into an automatic testing equipment. The block diagram in picture 7 shows the standard configuration of an ut-electronic including the additional channel for purity grade check with the normal incident transducer. This transducer is integrated into the rotational body additional to the normal standard arrangement of a ROTA 90Z (picture 8) with transducers for longitudinal and transversal defect detection and for wall thickness and diameter measurement. The transducer is build into a so called double probe holder beside the standard wall thickness measurement transducer and is realised in a ball housing. By means of the ball, the operator can adjust the incident angle in all directions to set an accurate normal incident to get a high signal to noise ratio. The central diameter adjustment principal of the ROTA 90Z guarantees now that the adjustment will be right for all tube diameters.


Pic 7: Block Diagram Testing Equipment including Channel for Purity Grade Check

The transmitter and receiver board for the purity grade channel have been installed as in situ electronics next to the inspection unit to avoid a drop down of the signal amplitude of the high frequency transducer due to long cabling to the ultrasonic electronics located in the cabinet. A high shot density on circumference is reached given by the high pulse repetition frequency of the ultrasonic unit in spite of the high rotational speed of the ultrasonic inspection unit. The density in tube length axis is given by the focal spot size of the transducer and the transversal transport speed of the conveyor. So, with full inspection velocity only a small percentage of the tube length is checked for purity grade. But the result is registered continuously and homogeneously over the complete tube and each individual tube is checked. That's why the described inspection method is called statistical process control.


Pic 9: Main Menu with Lot Statistic for Purity Grad Check

selection of purity grade check option mean value of purity grade for total lot number of tubes with purity grade alarm signal percent value of tubes with purity grade alarm signal total tube length of tubes with purity grade alarm signal alarm signal (over threshold event of the weighted sum value) for each individual tube maximum value of the weighted sum value for the complete tube for each individual tube

A special software was developed to evaluated the signals received by the purity grade transducer. In correspondence to the method of the C-scan evaluation, in this software the over threshold events of a gate are counted and evaluated. Hereby the start position of the gate can be controlled in a so called entrance echo trigger mode by the position of the interference echo and the stop of the gate by the position of the back wall echo due to the use of a digital ultrasonic unit with complete digitised A-scan. So the evaluation is possible close to surface and back wall without an influence of water path or wall thickness variation. A lower and upper threshold are provided for this gate and for these thresholds the number of amplitude over threshold events are counted. The number of events is divided through the total number of ultrasonic shots per unit to get a percent value. The setting of the recorder resolution in mm or ms is used as unit for this calculation. A third output is calculated as a sum of both values whereby the value of the second threshold is multiplied by a free selectable factor to get a weighted sum. This provides the possibility to give a different weight for defects with higher amplitude on the statistic result. For the automatic inspection the described three values calculated for the set recorder resolution will be shown on a separate recorder trace display (picture 10) additionally to the normal amplitude output. The minimum, the mean and the maximum value of the weighted sum of the event counters are determined at the end of the tube and shown on the right side of the recorder trace. The maximum value is transmitted to the master control (MCU) software (picture 9) for a statistical lot evaluation. Additionally an alarm threshold for the sum value exists which is shown on the recorder trace, too. If the sum value is over this threshold a corresponding mark appears on the recorder output and at the end of the tube also an over threshold alarm is transmitted to the MCU. With this alarm the MCU provides a lot statistic with the number of tubes with purity grade check alarm. Additionally a mean value for the purity grade is calculated from the sum of the values of each tube divided by the total number of tubes for which the purity grade check was switched on.


counter for over-threshold event lower threshold counter for over-threshold event upper threshold weighted sum of both event counters threshold for weighted sum value alarm signal maximum value of weighted sum over complete tube mean value of weighted sum over complete tube minimum value of weighted sum over complete tube

Pic 10: Recorder Output for Purity Grade Check including Event Statistic over Tube Length

The advantage of this method is now that the purity grade values for each individual tube and the complete tested lot are available on-line.


Pic 11: Block Diagram Laboratory Equipment

A tube segment which shows on the recorder output a lot of indications can by selected for a verification of the measurement. The same hardware (picture 11) with a special C-scan software feature is used to create a C-scan of the selected tube segment by means of a one-channel immersion technique equipment (picture 12) with rotating and forward moving tube and a steady transducer in a tank. Picture 13 and 14 present the results of such a C-scan performed with the two different transducers discussed at the beginning. Comparing the two C-scans it is evident that the resolution of the 20 MHz transducer is better but the relevant indications are found with the 15 MHz transducer, too. So, if the reference size of the defects is not defined to small it is also possible to work in the automatic equipment with a 15 MHz transducer for which the setting is easier to handle.

Pic 12: One-Channel Immersion Technique Equipment	Pic 13: C-Scan 20 MHz Transducer
	Pic 14: C-Scan 15 MHz Transducer

As a summary, the presented continual statistical check of purity grade for the complete tube production gives the possibility to optimise the production process immediately by using the on-line inspection results. The quality can be evaluated and the production process can be improved by giving the testing data to the steel making plant.

© AIPnD , created by NDT.net

|Home|    |Top|