Every inspection programme faces the same question: which method should we use? Non destructive testing offers several proven techniques, but each one sees certain defects clearly and others not at all. Choose the wrong method and a critical crack can pass inspection unnoticed.
The good news is that the choice is rarely a guess. Three questions narrow it down quickly: what is the part made of, what kind of defect are you looking for, and what does the part look like? This article walks through a simple decision framework built on those three questions, so you can match the right NDT solutions to your components and get reliable results every time.
The Framework: Three Questions in Order
- Material: Is the part ferromagnetic, electrically conductive, or neither? This rules methods in or out immediately.
- Defect type and location: Is the flaw a crack, a void or wall loss? Is it on the surface, just below it, or deep inside?
- Geometry and access: How big, thick and complex is the part, and which surfaces can you reach?
Answer them in this order. Material removes options, defect type picks the best of what remains, and geometry decides how to apply it in practice.
Step 1: Start with the Material
The material’s magnetic and electrical properties decide which methods can work at all.
- Ferromagnetic steels and cast irons (carbon steel, low-alloy steel, most tool and bearing steels): Magnetic Particle Inspection (MPI) is usually the fastest and most sensitive choice for surface and near-surface cracks. Ultrasonic, eddy current and radiographic testing also work.
- Non-magnetic conductive metals (aluminium, copper alloys, titanium, austenitic stainless steel): MPI cannot be used. Penetrant testing (PT) and eddy current testing (ET) are the main surface methods; ultrasonic testing (UT) covers the interior.
- Non-conductive materials (plastics, ceramics, composites): MPI and ET are ruled out. PT works on non-porous surfaces, and UT or radiography handle internal flaws.
- Porous materials (some sintered parts and rough castings): penetrant can bleed from the pores themselves and hide real indications, so MPI (if magnetic) or ET is often better.
For most engineering components made of steel, this first step points straight to MPI as the primary surface method, which is why well-specified NDT machines for magnetic particle testing are standard equipment in forging, automotive, bearing and railway workshops.
Step 2: Define the Defect Type and Location
Next, be clear about what you are trying to find and where it is likely to be.
Surface-breaking cracks
Fatigue cracks, grinding cracks, quench cracks and forging laps usually start at the surface.
- Steel parts: MPI, the most sensitive option for tight cracks.
- Non-magnetic metals: PT or ET.
Near-surface defects
Inclusions or cracks just below the surface that do not break through.
- Steel parts: MPI with DC magnetization reaches a short distance below the surface.
- Conductive metals: ET with a suitable frequency.
Internal defects
Internal cracks, laminations, inclusions and voids deep inside the section.
- UT for planar defects such as cracks and laminations, and for measuring depth.
- Radiographic testing (RT) for volumetric defects such as porosity and shrinkage in castings.
Wall thickness loss and corrosion
- UT thickness gauging for remaining wall thickness.
- ET for surface corrosion and coating thickness on conductive parts.
Crack orientation matters
Every method has blind spots tied to direction. MPI shows cracks that cut across the magnetic field, so parts are magnetized in two directions. UT finds cracks that face the sound beam, and RT can miss tight cracks that are not aligned with the radiation. Always ask which way the expected defect will run.
Step 3: Consider Geometry, Size and Access
The shape of the part decides how practical each method is.
- Simple shapes such as shafts, bars and axles: ideal for bench MPI with head-shot and coil magnetization, and for automated UT or ET scanning.
- Complex shapes such as gears, crankshafts and forgings: MPI and PT follow every contour; UT is harder because the sound beam must reach each area at a useful angle.
- Hollow parts such as rings, tubes and bushes: a central conductor in MPI tests inside and outside surfaces together; ET bobbin probes suit tubes.
- Large parts: portable MPI yokes, PT kits and UT instruments bring the inspection to the part.
- Thin sections: ET and PT work well; UT needs suitable high-frequency probes.
- One-sided access: UT and ET need only one surface; RT needs access to both sides for the source and detector.
- Rough or coated surfaces: coatings reduce MPI and ET sensitivity and block PT, so clean to bare metal where the procedure requires it.
Matching geometry to the right NDT equipment, such as a bench unit, portable yoke or automated system, is often what separates a reliable inspection from a slow or incomplete one.
Other Factors That Shape the Decision
- Production volume: high-volume lines benefit from automated or semi-automated MPI and ET systems that keep inspection time per part low.
- Required sensitivity: safety-critical parts in aerospace, railways and automotive need methods with proven detection reliability.
- Applicable standards and customer specifications: codes such as ASTM E1444 and ISO 9934 for MPI, or customer drawings, may define the method for you.
- Records and traceability: RT gives a permanent image; digital MPI, UT and ET systems can store settings and results.
- Safety and environment: RT needs radiation controls; MPI and PT need proper handling of particles and chemicals.
- Cost and speed: surface methods such as MPI and PT are quick and low cost per part; volumetric methods take longer and need more skill.
- Post-inspection steps: MPI parts usually need demagnetization; PT parts need cleaning.
Putting the Framework to Work: Examples
Steel crankshaft, checking for fatigue and grinding cracks Material: forged steel. Defect: surface cracks at fillets and oil holes. Geometry: complex, medium size. Choice: wet fluorescent MPI on a bench machine, with head shot and coil shot.
Aluminium housing, checking for surface cracks Material: non-magnetic. Defect: surface-breaking cracks. Geometry: complex casting. Choice: fluorescent penetrant testing.
Steel railway axle, checking in service Material: steel. Defect: fatigue cracks at wheel seats. Access: wheels fitted. Choice: UT from the axle ends in service, then MPI at overhaul when the wheels are removed.
Cast steel valve body, checking for internal porosity Material: cast steel. Defect: internal shrinkage and porosity. Geometry: thick, complex. Choice: radiographic testing for the interior, MPI for surface cracks.
Bearing rings in mass production Material: bearing steel. Defect: grinding cracks. Geometry: rings, high volume. Choice: automated MPI with a central conductor, or ET for fast scanning.
Common Mistakes to Avoid
- Using MPI on non-magnetic parts: it will not work on aluminium, titanium or austenitic stainless steel.
- Relying on PT for subsurface defects: penetrant only finds flaws that are open to the surface.
- Magnetizing in one direction only: cracks parallel to the field are missed.
- Skipping surface preparation: paint, oil and scale hide indications in almost every method.
- Choosing by habit, not by defect: the method you always use may not be the one that sees the defect you are worried about.
- Ignoring equipment checks: field strength, UV-A intensity, probe calibration and bath concentration all affect what you find.
Frequently Asked Questions
What is the best NDT method for steel parts? For surface and near-surface cracks in steel, Magnetic Particle Inspection is usually the most sensitive and cost-effective method. For internal defects, ultrasonic or radiographic testing is added.
Can one NDT method find every defect? No. Each method has limits based on material, defect type, orientation and depth. Critical parts are often inspected with two complementary methods.
Which NDT method works on aluminium? Penetrant testing and eddy current testing for surface defects, and ultrasonic or radiographic testing for internal defects. MPI does not work on aluminium.
How do I choose between MPI and penetrant testing? If the part is ferromagnetic,
is generally faster and finds tighter and near-surface cracks. If it is non-magnetic, use penetrant testing.
Which NDT method is best for high-volume production? Automated MPI and eddy current systems handle large volumes with consistent results and short cycle times.
Conclusion
Choosing an NDT method does not have to be guesswork. Start with the material, define the defect you need to find and where it lies, then check the geometry and access. Add production volume, standards and cost, and the right method, or combination of methods, becomes clear.
Magkraft Integrated Pvt. Ltd. designs and manufactures NDT machines and NDT solutions for reliable crack detection, from bench MPI systems to automated inspection lines. Talk to our team to find the right NDT equipment for your components.