Ultrasonic Testing of FRP Thickness

June 12, 2017

If all you have is a hammer, everything looks like a nail.

— Abraham Maslow

In 1966, Abraham Maslow called this the “Law of the Hammer”. He observed that people will depend on tools or methods that are familiar before they make the effort to get new tools or learn new methods. I smile when I remember the number of times that I tried to use the wrong tools during a house renovation, before I broke down and got the right tools. Sometimes, using old tools in a new way can work. In this entry I am writing about some new methods for old tools.

Ultrasonic testing is one of the oldest non-destructive testing technologies, dating back to the 1940’s. To illustrate how it works, think about times when you have heard echoes of your voice. This might have happened in the mountains or in a large, empty building. The sound waves from your voice travelled away from you until they reflected, or bounced, from a surface. The reflected sound then returned to you and you heard it. The longer the time delay, the further away to the reflecting surface. If you don’t hear an echo, it could be because the reflecting surface is too close to detect the time delay, or the space might have materials in it that absorb the sound – like insulation, carpet, soil, and others – or there might be something between you and the reflector that blocks the sound or scatters it.

Ultrasonic methods work on the same principle, where high-frequency (beyond what we can hear – hence “ultra”sonic) wave pulses are used in solid materials. Ultrasonic testing instruments measure the time from when the pulse was created to when the reflection was received. One of the most common uses of ultrasound is to use the time to indicate thickness of the material. When the velocity that the ultrasound pulse travels from one side to the other of the material is known, the thickness can be determined, simply by multiplication. The accuracy of the thickness calculations depends on the sonic velocity – if the sonic velocity is not constant, the thicknesses calculated will vary.

For steel, and many metals, long experience with ultrasonic testing has shown that variations in sonic velocity are small. In statistical terms, the “confidence interval” is narrow. For this reason, it has become accepted practice for steel to calculate thicknesses using one (1) or two (2) calibration readings on a sample of the same type of steel with a known thickness.

Just like in the quote above, if this method works for steel, why not use it for other materials like fiberglass reinforced plastic (FRP)? In fact it works, but there is a major difference. The variations in sonic velocity that we have observed are much wider than they are for steel – often varying by 15%, or more, within a small sample. Reasons for the variations in velocity include changes from corrosion attack, variations in glass concentration, curing of the resin, porosity in the resin, impurities in the FRP, among others. So how do we work with these variations to use ultrasound for thickness measurement of FRP – what new methods are required?

The key is to understand how to use the variation of sonic velocity. If we use a single calibration reading – following the “old” method – how do we know if it represents average, or maximum, or minimum sonic velocity? The answer is that we don’t. We have to determine the actual variation of sonic velocity in the FRP and use that information to allow us to calculate thickness as a range. A typical example for a new FRP tank would be to use a nozzle cutout to take a number of measurements and corresponding ultrasonic calibration readings. This data is then used to determine the variation of sonic velocity in the sample – and from this, we can determine the sonic velocities associated with a specific “confidence interval”, or where we are confident to some percentage (usually 95%) that the average sonic velocity is within the range.

This becomes more complicated when a calibration sample is not available to determine the sonic velocity distribution, especially when the FRP has been in service. Long experience at working with ultrasound on FRP has shown that the sonic velocity through the FRP thickness will usually not remain as it was when new. In these cases, we at UTComp use a database of sonic velocities from many samples to determine a “minimum velocity” where 95% of average velocities are greater than this value, so that we can report that the thickness detected is a minimum and conservative value. When assessing FRP that has been in service, and then determining its remaining service life, UTComp will often evaluate the structural capacity of the FRP which includes the thickness of the material. In our case, using the minimum velocity value allows us to make conservative predictions and minimize risk to our customers. In all of this, we have found that thickness changes in FRP are very rare unless abrasion or aggressive oxidation are expected. Significantly, we have found that damage to the FRP from corrosion and structural loads are detected from features in the ultrasonic reading without need for the actual thickness of the material.

I have just described a transparent and straightforward approach to dealing with calibration for FRP sonic velocity.

Can we make it easy and reliable to use these simple tools for this straightforward task?

The answer takes us back to the start of this article. We at UTComp learned that new methods were required to use ultrasonic tools to get new results. To simplify this for our customers, UTComp has developed a number of tools that provide reliable results. We are also working on programs to be installed in existing ultrasonic equipment to ease challenges encountered by owners of FRP equipment. Bear in mind that “accurate” measurements can only be defined in terms of a confidence interval. For 95% confidence of the average thickness, the value is likely to be within about 10% of the calculated average.

If you are interested in updated ultrasonic testing equipment with some of these advanced tools for FRP thickness measurement, please email us at inquiries@utcomp.com. To see other ways that UTComp has been able to use ultrasound to improve FRP reliability and reduce costs for owners, please visit our website at www.utcomp.ca.

– Geoff Clarkson, P. Eng, FEC

President & CEO of UTComp

Want to know more? Email Geoff at G.Clarkson@UTCOMP.com or call 519-620-0772.

Follow UTComp on Twitter @UTComp_FRP and LinkedIn.

Questions about your FRP composite equipment?