ZHUOJIN-LOGOZhuojin Pipe Fitting Co., Ltd
Zhuojin Pipe Fitting Co., Ltd
STEEL PIPE · FITTING · WELDING  |  NOTES FROM A 30-YEAR INDUSTRY VETERAN

Internal Magnetic Array Eddy Current Testing for Small-Diameter Steel Pipes

Why This Paper Matters on the Shop Floor

I have spent thirty years watching inspection crews struggle with small-diameter pipe. The problem is not the pipe; it is the access. When you have a 22 mm outside diameter line carrying process fluid, you cannot get a conventional eddy current coil inside it, and external probes lose sensitivity fast because the wall is thick relative to the coil diameter. This paper from Hubei University of Technology and the Hubei Special Equipment Inspection and Research Institute addresses exactly the gap I have seen on countless in-service lines: how do you reliably find a 0.2 mm deep crack on the internal wall of a 18 mm bore pipe without cutting it open? The answer they propose is an internal through-type magnetizing array eddy current sensor, and having reviewed the test data, I can tell you this is a serious piece of work that deserves attention from anyone running petrochemical, power or gas distribution lines.

Sensor Design and Simulation Approach

The authors used COMSOL to build a magnetizing array eddy current model before making hardware. This is the right sequence. I have seen too many probe designs go straight to machining without a field simulation, and the result is a probe that works in the lab and fails on the first real pipe because the magnetic circuit was never balanced. The sensor uses a magnetization array that creates a controlled magnetic field in the pipe wall, and the eddy current coils are arranged in three main channels. The key design insight is that the array configuration is intended to compensate for the blind zone that normally occurs when a crack is oriented parallel to the magnetization direction. The simulation allowed them to vary the crack angle and observe the relative permeability and voltage signal response, which gave them the confidence to build the actual probe.

Parameter Value
Pipe material 45# steel
Outside diameter 22 mm
Inside diameter 18 mm
Internal wall defect size 15 mm x 0.2 mm x 0.5 mm
External wall defect size 15 mm x 1 mm x 1 mm
Number of main channels 3

Experimental Validation and Key Results

The experimental work used real 45# steel pipe with machined defects. The internal defects were 15 mm long, 0.2 mm wide and 0.5 mm deep, which is a realistic stress corrosion crack or fatigue crack geometry. The external defects were 15 mm x 1 mm x 1 mm, simulating a through-wall or near-through-wall external damage. The results are what I would call genuinely useful. The three-channel array configuration successfully eliminated the internal wall blind zone, and after applying a signal processing algorithm, the synthesized signal produced similar amplitude for cracks at any angle. This is a significant achievement because conventional single-coil probes lose sensitivity dramatically when the crack orientation approaches the magnetization direction. For external wall cracks, the sensor detected cracks in any direction, but the signal amplitude decreased as the crack orientation angle relative to the magnetization direction decreased. This is expected physics, but the paper quantifies it, which is exactly what an inspection engineer needs to set acceptance criteria.

Engineering Practice Insights

What does this mean for a plant engineer? First, it means you can now consider internal inspection of small-bore lines without pulling them out of service for hydrostatic testing or cutting coupons. Second, the signal processing algorithm is critical. Do not expect raw eddy current signals to give you a clean answer; the array data must be fused and interpreted. Third, the sensitivity limit matters. The internal defect depth of 0.5 mm was detected, but I would want to know the probability of detection for shallower defects. In my experience, a 0.2 mm deep crack in a 4 mm wall is already significant, but a 0.1 mm crack may be the threshold for initiation. The paper does not address POD curves, and that is a gap I would flag for the next phase of research. Finally, the probe is designed for 22 mm OD pipe. Scaling it to larger diameters is not trivial because the magnetic circuit changes. If you are thinking of applying this to 60 mm or 114 mm pipe, you need to re-run the simulation and re-validate. Do not assume the array geometry transfers.

Key Questions and Reflections

The paper raises one question I have asked many times in my career: what is the inspection interval? If you can detect a 0.5 mm internal crack, you still need to know how fast it grows to set a safe re-inspection period. The authors do not address crack growth rates, but that is a materials and fracture mechanics problem, not an NDT problem. The NDT side is now solved for this pipe size. The other question is calibration. How do you calibrate a probe like this on site? You cannot easily machine defects into an in-service pipe. The answer is to use a calibration block with known EDM notches, but the block must be the same material and wall thickness as the pipe under test. I would insist on that in any procurement specification. The paper is a solid contribution to the NDT literature. It solves a real problem, it validates the solution experimentally, and it leaves a clear path for further work on POD and larger diameters. I would recommend it to any engineer responsible for in-service inspection of small-diameter pipework, and I would keep the signal processing algorithm in mind when specifying software for the inspection system.

Study Insights and Outlook

The takeaway from this study is that array eddy current with magnetization is no longer a laboratory curiosity for small-diameter pipe. It is a deployable technique that can find both internal and external defects at any orientation, provided you use the right array configuration and the right signal fusion. The next step I would like to see is a field trial on an actual operating line with known defects, followed by a comparison with ultrasonic and radiographic results. If the field trial confirms the laboratory data, this method will become a standard tool for in-service inspection of small-bore pipework, and it will save the industry a lot of money by avoiding unnecessary cut-outs and hydrostatic tests. For now, I would treat this paper as a strong technical foundation and a reminder that good NDT is always a combination of physics, signal processing and practical validation on real pipe.