Back to Basics
[ click here for the Back to Basics Archive ]
Effects of Temperature on High Density Polyethylene Piping and Accuracy of Ultrasonic Thickness Gaging
by Kevin Harkreader*
What can be more basic to ultrasonic thickness gaging than using the correct sonic velocity? Here is an interesting and useful story that points out the importance of this basic information and how to get the correct velocity. Enjoy this one. I did.
Frank Iddings
Tutorial Projects Editor
Introduction
Ultrasound provides a very accurate and controlled way of determining the thickness of a wide variety of materials. Ultrasonic thickness data is useful when determining the rate of material loss from erosion or corrosion and can also be used for calculating remaining life predictions. Ultrasound can be applied to high density polyethylene pipe for the same purpose. High density polyethylene pipe is made from petroleum products. Its most common use is in milk and water containers, but it is also used for piping. This easily recognizable black plastic piping has become a very popular choice for use in industrial applications such as in mining, power plants and petrochemical plants. Some of its uses include the transfer of acid, water, coal slurry, hazardous waste and organic chemicals.As part of an industrial plant's predictive maintenance program and with concerns over reliability and safety issues, you may be required to record the thickness of a high density polyethylene pipe line and possibly set up a monitoring program to determine the rate of wall loss. For this purpose, accuracy of the thickness measurement is going to be very important. Given the fact that actually taking thickness readings is relatively simple, requiring only the expertise of a Level I, special attention should be given if you find yourself with this responsibility. There is a fundamental difference between the ultrasonic testing of these plastics and that of metals and a few more variables that can affect the accuracy of the ultrasonic measurement. The most important element is the effect of temperature on the accuracy of thickness testing.
Velocity, impedance and attenuation are all factors that will influence the ability to capture a thickness.
Properties and Effects
Because of the density and elasticity makeup of high density polyethylene, sending and receiving ultrasonic pulses through the plastic is going to be highly attenuated and this factor increases if you take into account any fluid, such as water, flowing through the pipe. If we were to compare attenuation or loss of acoustic energy from scattering and absorption, it is absorbed about six times more than in carbon steel. Reflection of the ultrasonic signal from the pipe backwall is also of significance. Since the acoustic impedance constants of water and the piping are closely related, the reflection coefficient or the amount of ultrasound that returns from the pipe/water interface (backwall), using a 6.4 mm (0.25 in.) diameter 2.25 MHz transducer, is only 8 percent. The remaining 92 percent reflection coefficient is absorbed into the water.Velocity, impedance and attenuation are all factors that will influence the ability to capture a thickness. The signal strength will be greatly diminished, but with some adjustment of the gain and proper selection of transducers, capturing a thickness from a 51 mm (2 in.) wall is probable. Keep in mind that any internal geometric changes from erosion rendering the pipe surfaces nonparallel will create some diffraction and will contribute to an even lower signal response. With this information in mind, we can now understand some of the inherent problems that are encountered which should help in the proper selection of transducers and instruments.
The difficulties discussed above are going to be important when considering instrument and transducer selection. You will need to experiment with transducer diameters and frequencies to ensure that proper surface contact and a strong enough pulse exist to punch through the ranges of pipe thickness. Using an A-scan instrument with gating capabilities along with a digital readout to three decimal points is really going to become important if you're doing any material loss calculations. An A-scan thickness meter with a digital readout and dual transducer will not provide a strong enough pulse for any pipe wall thicker than 9.5 mm (0.4 in.). A conventional off the shelf transducer of 2.25 or 5 MHz will be suitable for most pipes up to 16 mm (0.63 in.) thick. As thickness increases, larger diameter and lower frequency transducers will be needed.
Experience Speaks
I want to share a case of how we dealt with taking accurate thickness readings on high density polyethylene pipe for a mining company transferring tailings slurry. Originally, steel pipe was being used for the purpose of transporting the slurry, but the abrasive and erosive effects caused premature failure. The high density polyethylene pipe alone could not withstand the pressures involved with moving the slurry and there was no real data to support its holding up against this type of abrasive action. Eventually a decision was made to use the high density polyethylene pipe as a liner inside the steel pipe. To access the polyethylene piping for thickness testing, 13 mm (0.5 in.) diameter holes were cut out from the bottom center of the steel pipe and a threaded nut was welded into place. A 13 mm (0.5 in.) diameter bolt was then inserted that could be easily removed when taking thickness measurements. This was now a numbered testing point for monitoring.Baseline readings were to be taken and subsequent followup readings gathered every quarter (three months) to monitor any liner material loss from possible erosion. The sonic velocity in the polyethylene pipe is about 2.5 times slower than in steel, so for accurate thickness testing, a stepped calibration block meeting the range of thickness of the same material was machined. Machining high density polyethylene pipe accurately is much more difficult than machining steel, but with the use of a micro thickness gage, each step can be measured and recorded (Figure 1).
It was not until after the third quarterly tests were done and the thickness data collected that noticeable inaccuracies in the data were found. Some of the thickness numbers increased while others decreased for no given reason and at random data points. The calibrations were made throughout the day during the entire process and curiously enough we were continually adjusting parameters on the instrument. Unfortunately, our data were so out of line that they became useless. Our baseline readings were in question and useless for calculating erosion rates. It became evident that we were uncertain as to which numbers gathered were even correct.
Temperature and Velocity
A closer look into the variables involved led us to discover that small differences in the temperature of the polyethylene had a direct relation to material velocity, much different from that of steel. As the temperature of the pipe increased throughout the day, the velocities became increasingly slower and the reverse happened as it decreased in temperature. So depending on the season or even throughout the day, temperature variations at each of the 200 testing points had an effect on the thickness recorded. Even partly cloudy days which produced shadowing effects were sensitive enough to influence the outcome of data.The data in Table 1 show examples of the change in thickness when temperature increases 5.5 °C (10 °F) and 16.7 °C (30 °F) without adjustment for material velocity. A mere 5.5 °C (10 °F) change in temperature might represent the difference in morning and afternoon while the 16.7 °C (30 °F) change could easily represent the difference between winter and summer. As noted below, temperature effects are more pronounced as the thickness increases.
To compensate for the changes in temperature and the changes it has on material velocity, a velocity and temperature correlation table was developed (Table 2). To ensure that this data was accurate, a piece of high density polyethylene pipe was sent to a manufacturer's NDT laboratory familiar with our needs.
Calibration and Testing
Instrument calibration is first done on the high density polyethylene pipe step wedge with a temperature gage (Figure 2) and the velocity on the instrument adjusted according to Table 2. The zero offset of the transducer is also entered into the instrument setup. A transducer is then placed on each step of the block and checked against the measurements taken with the micro thickness gage. This is a simple two step calibration check: take the temperature and adjust the velocity accordingly. With the use of a digital A-scan instrument, always record thickness data at the same signal height as that of the calibration. If you calibrate at 80 percent screen height, record your thickness data at the same screen height. Equally important is the placement of the gate height. It should always remain at the same percent screen height level when calibrating and recording thickness data and this information should be included in the procedure.Each test point will require that a temperature measurement be taken with the use of a digital meter. The temperature can then be compared to the correlating velocity in Table 2 and the instrument velocity can then be adjusted. If temperature readings are not taken, the accuracy of the thickness data will fluctuate from winter to summer as well as throughout the day. It is not uncommon to see the temperatures change throughout the day since the pipe is black and absorbs the sun's heat even through the steel. As noted on the chart, a change of 16.7 °C (30 °F) on a 13 mm (0.5 in.) pipe wall could give you an error of 0.6 mm (0.02 in.) if no velocity adjustments are made.
One variable that has not been determined is the temperature difference of the outside diameter of the pipe versus the inside diameter of the temperature. However, thickness readings have been conducted for four consecutive years and found to be very accurate by simply taking temperature readings on the pipe exterior at each monitoring access point and disregarding any internal wall temperature difference. A range of thicknesses from 6.4 to 57 mm (0.25 to 2.25 in.) has been taken with this method. The temperature of the fluid through the pipe can change the internal wall of the pipe lining as well. However, our experience shows that this has not dramatically changed the accuracy of the readings over a time span of four years.
Writing Procedures for Variables Associated with the Test
Many of us believe that once the choices of the instrument and transducer are made, calibration of the instrument for the most part really only involves using a correct velocity and adjusting for the transducer zero offset. Others believe this is incorrect. This is where a procedure is necessary. This is a set of instructions for repeating and duplicating all aspects of the testing from calibration to gathering information and sharing this information from one technician to the next. Writing a procedure may seem unnecessary for merely taking thickness readings, but the whole point is that one technician may do it differently from another and no one wants to memorize data setups, especially if the test intervals are three months apart. When writing your procedure, keep it in a language that everyone can understand — short and simple, but specific.Knowing the many variables, such as those listed below, and adjusting for them, will allow one to ensure an accurate data set. One should recognize the importance of this information and consider it part of the procedure. Here are some variables to think about for a procedure:
selection of instrument (should remain the same throughout the testing time span)
selection of transducers (should remain the same throughout the testing time span)
range settings
gate settings (especially recording the percent screen height)
percent screen height of backwall before collecting data (60, 70 or 80 percent)
signal flank versus point recording
temperature/velocity adjustments
whether there are fluids going through the pipe
pipe temperature variation due to weather conditions
accuracy of temperature at each test point
calibration of the temperature instrument gage at the beginning of each testing day
whether thickness data is being taken in the same location each time.
Training of the Technician and Full Understanding of Procedure
Always record from the flank of your signal if that is how you calibrated it and make it a habit to record the lowest reading on the instrument. Make sure to use a micro thickness gage for measuring the step wedge. A machinist will not be able to get as accurate a tolerance as a steel step wedge. Mark your step wedge with a permanent marker no larger than the diameter of the transducer at the same location as the thickness gage reading and be exact to eliminate any room for error. This is where the temperature probe and transducer will be placed each time for calibration. Take temperature readings at each thickness monitoring location. Do not assume that the temperature at the top of the pipe is the same as that along the side or bottom. When you find an area that you want to monitor, scan the circumference of the pipe after it has been in service for a while and use the lowest reading as the monitoring location. Make your test points permanent. Pick your point then draw a circle with a permanent marker, but do not make it any larger than the diameter of the transducer. You
do not want to be scanning a 25 mm (1 in.) diameter area with a 6.4 mm (0.25 in.) transducer. Record your thickness reading only after the mark is in place.Typically, the erosion is going to take place on the bottom in the center and in the case of the tailing slurry it was within a range of 127 mm (5 in.) of the bottom center for a 0.5 m (20 in.) pipe. Therefore, try to determine closely where the testing points should be. Bends create erosion at different locations. If the pipe takes a steep drop or if the velocity of flow increases and has a turn, the worst areas may be on the side of the pipe.
A flanged pipe can create turbulence and may increase erosion within a 0.3 m (12 in.) distance. It would be best to take readings in several locations at least 0.3 m (12 in.) before and after (for 0.5 m [20 in.] pipe) the flange; or better yet, try to scan for the lowest area and mark it as one to monitor. We found that the highest number of failures occurring from erosion were within this flange area. Within 0.3 m (12 in.) of the flange, visual testing revealed unusual wavy wear patterns 25 to 51 mm (1 to 2 in.) apart, which is where it can become difficult to capture a thickness if parallel surfaces do not exist.
* HARSCO Track Technologies, 3434 E. 7800 S. #220, Salt Lake City, UT 84121; (801) 517-7431; e-mail <fishing@icw.com>.
Copyright © 2001 by the American Society for Nondestructive Testing, Inc. All rights reserved.
Copyright © 2008 by the American Society for Nondestructive Testing, Inc. ASNT is not responsible for the authenticity or accuracy of information herein. Published opinions and statements do not necessarily reflect the opinion of ASNT. Products or services that are advertised or mentioned do not carry the endorsement or recommendation of ASNT.
IRRSP, NDT Handbook, The NDT Technician and www.asnt.org are trademarks of the American Society for Nondestructive Testing, Inc. ACCP, ASNT, Level III Study Guide, Materials Evaluation, Nondestructive Testing Handbook, Research in Nondestructive Evaluation and RNDE are registered trademarks of the American Society for Nondestructive Testing, Inc.
ASNT exists to create a safer world by promoting the profession and technologies of nondestructive testing.