Getting NDT Right in MT and PT

imeche.org

ENGINEERING TRAINING SOLUTIONSGETTING NDT RIGHT IN MT AND PT

Phil RawBSc (Hons) NDT

Improving the world through engineering imeche.org

CAREER HISTORY

• 27 years experience in the NDT industry

• Holder of PCN, ASNT & EN 4179 Level 3 certification in multiple methods also BSc Honours Degree in NDT

• Employed by Argyll-Ruane Ltd since 2001 as a Level 3 Consultant, Trainer and Examiner


GETTING NDT RIGHT MAGNETIC PARTICLE AND PENETRANT TESTING

As a trainer at Argyll-Ruane Ltd it is my job to teach NDT personnel how to “get NDT right”As a consultant I work for my clients as a Responsible Level 3 and then my job entails making sure NDT personnel are getting NDT right.

Over the years I have witnessed many occasions when NDT has not been done right and today I will discuss some of these incidents. Some of the consequences of these incidents have lead to individuals being dismissed, or been very expensive to correct.


Why is getting NDT right so important?

The British Institute of NDT states: Non-destructive testing and inspection are vital functions in achieving the goals of efficiency and quality at an acceptable cost.

In many cases, these functions are highly critical; painstaking procedures are adopted to provide the necessary degree of quality assurance. The consequences of failure of engineering materials, components and structures are well known and can be disastrous.



It is an increasing requirement of quality assurance systems that a company's technicians are able to demonstrate that they have the required level of knowledge and skill. This is particularly so since NDT and inspection activities are very operator dependent and those in authority have to place great reliance on the skill, experience, judgement and integrity of the personnel involved. Indeed, during fabrication, NDT and inspection provides the last line of defence before the product enters service, whilst once a product or structure enters service, in-service NDT is often the only line of defence against failure.



Basically the job you do is not an ordinary job – the job you do is critical and in many cases lives can depend on you doing your job correctly.Not getting NDT right is not an acceptable option, if we get NDT wrong it can have tragic consequences.

The next few slides show what can happen if NDT has gone wrong.



On 19th July, 1989, a United Airlines DC-10-10, experienced a catastrophic in flight failure of the No. 2 engine.

This engine failure led to the discharge of the stage 1 fan disk assembly parts from the No. 2 engine which led to the loss of all three hydraulic systems powering the flight controls.



The crew experienced severe difficulties controlling the airplane, which subsequently crashed during an attempted landing at Sioux Gateway Airport.

There were 285 passengers and 11 crewmembers on-board, 1 flight attendant and 110 passengers were fatally injured, 47 suffered serious injuries, 125 suffered minor injuries and only 13 people on board were uninjured.



During its 17 year service life the engines in which the accident disk was installed were overhauled, disassembled and inspected using fluorescent penetrant inspection in 1972, 1973, 1976, 1978 1982 and finally in February 1988 17 months before the accident.

The NTSB concluded that the fatigue crack was likely to have been between 12.0mm to 12.6mm surface length at the time of the last fluorescent penetrant inspection, the minimum detectable defect length in the critical disk bore area is 2.54mm when using fluorescent penetrant inspection therefore the fatigue crack should have been detected at the last inspection.



The NTSB reviewed the technique used to inspect the disk and one possibility for the failure of several inspections carried out after the fatigue crack had exceeded the minimum detectable defect size was the fact that the disk was suspended by a cable through the bore.

To fully apply the penetrant and later the developer and to fully inspect the bore including the area obscured by the cable, the operator has to physically rotate the part.



It is possible that the inspectors may not have correctly processed or viewed the bore area due to the cable, or that during rotation the cable obliterated the indication or caused loose developer to fall on to the crack and obscure the indication.

The NTSB also consider that there is a strong chance the inspectors were experiencing inattentional blindness. This type of disk tends to produce indications near dove tail posts and rarely in the bore, this could cause inspectors to inspect the bore with less attention.



The NTSB considers the probable cause of the accident was the inadequate consideration given to human factors limitations in the inspection and quality control procedures used by the engine overhaul facility.

Inspectors tend to work independently and have little or no supervision and hence there is no redundancy to prevent human error.

This resulted in several unsuccessful inspections which failed to locate a detectable sized fatigue crack located in a critical area of the fan disk.





In July 1996, a Delta Airlines flight suffered an uncontained engine failure during routine take off from Pensacola airport. The left hand side number 1 engine; a Pratt & Whitney JT8D-219 on the McDonnell Douglas MD88 was destroyed during the incident.

The front compressor front hub (fan hub) suffered a catastrophic fatigue failure.

Uncontained engine debris penetrated the rear fuselage on the left hand side and tragically two people were killed, one person on the other side of the aisle suffered serious injuries and one passenger sustained serious injuries during the cabin evacuation.



The NTSB investigation revealed that the failed titanium fan hub had fractured radially in two places; one of the radial fractures contained a fatigue crack that originated at two locations on the inboard side in a tie rod hole, at 7.79mm and 14.04mm from the aft edge of the tie rod hole respectively.

The fatigue fracture extended 38mm radially inboard towards the centre of the engine, the rest of the fracture surface was considered to be the catastrophic final overstress failure.



The last inspection that the accident hub received was a fluorescent penetrant inspection carried out in October 1995 by a Delta airlines level 1 certified penetrant inspector with 11 months experience.

The FAA inspected the Delta airlines testing facility and determined that there was no assurance that parts received by penetrant inspectors were “clean enough for an adequate penetrant test”.

The hubs were not pre cleaned prior to penetrant testing by qualified NDT inspectors; they were cleaned by shop floor cleaning operatives with no NDT training and who were not made aware of the “criticality of the engine components and the end purpose for which the components were being cleaned”.



The final cleaning operations involved immersing parts in a hot water rinse and then flash drying.

For flash drying to work effectively and leave no water trapped in any discontinuities that would prevent penetrant entering the discontinuity, the part must reach the temperature of the water which must be between 65°C to 93°C.

Cleaning operatives did not measure part temperature and used only their sense of touch to feel if the part was hot enough; the water temperature was only checked on a weekly basis.



The report also looked at human factors as a potential cause of the crack being missed; these hubs are very large and are a very complicated shape and can take 40 minutes to two hours to inspect fully.

Any distraction during this period could cause the inspector to fail to resume the inspection where he left off, and with penetrant inspection unless you physically mark an area you do not know if it has been inspected or not.



Delta inspectors did not mark inspected areas or use a grid pattern to assist inspection.

The NTSB also considered that there was a low expectation amongst inspectors of finding a crack; the supervisor stated that he was not aware that any cracks had ever been found on this type of hub at Delta.

The conclusion was that in-attentional blindness could have caused the inspector to overlook or minimise the significance of an indication.



The NTSB concluded that the probable cause of the accident was

• the failure of the fluorescent penetrant inspection process to detect a detectable fatigue crack initiating from an area of altered microstructure that was created during the drilling process and

• lack of sufficient redundancy in the in-service inspection program.



Next lets have a look at some examples of how I have seen NDT not done correctly.



Ensure you are correctly certified for the work you are being asked to do.

Case Study: An aerospace manufacturer supplying to an engine manufacturer uses an EN 4179 certification scheme – (employer based certification) used NDT contractors who held EN 4179 certification issued by their employer. The written practice of the aerospace manufacturer has not been complied with and the contractors were not correctly certified to test parts.



• The parts they had penetrant tested underwent further manufacturing operations which meant the parts could not be re-tested by correctly certified NDT personnel.

• The engine manufacturer would not accept the £200k of parts tested by the contractors and one individual left the company over the issue.



Ensure your eye sight is tested annually.

Case Study: During a routine audit of an NDT facility the vision test certificate of one individual had expired some weeks previously and the operator had tested many parts that were supplied to an aerospace facility.

As his vision test was overdue his certification should have been suspended and he should not have tested any parts until he had passed an eye test.



When the customer was informed, they sent back all the parts tested by the inspector from the date his vision certificate had expired and demanded it was re-tested at the suppliers’ expense.



Ensure you understand and comply with NDT techniques and procedures.Case Study: A client conducted an audit of a suppliers NDT facility and watched an operator inspecting some of their parts. The operator did not follow the approved NDT technique and therefore did not correctly test the part.

The auditor twice asked the operator at the conclusion of the inspection if he was happy that he had fully completed the inspection and the operator said yes.



The client demanded the operator’s certification was immediately suspended and that all material he had tested for them had to be re-inspected. The inspector left the company over the issue.



Case Study: Many penetrant testing process specifications specify using a solvent wipe technique to evaluate indications. The process specifications usually specify wiping once and only once across the indication with a solvent damped cotton swab and then applying developer.

On many occasions during NADCAP compliance audits NDT inspectors have been seen to wipe across indications more than once thus incurring a non-conformance.



Ensure you carry out shift checks correctly.Case Study: During an audit of the NDT facility of a company that produces motorsport racing engines, the shift check log was reviewed and the result recorded for UV-A inspection lamp irradiance was exactly the same figure for every day for six months.

The operator was then asked to demonstrate the test and the result was approximately half the value of the recorded figure for the test “carried out” only an hour earlier.



Further investigation revealed other clear instances where the operator was not carrying out the shift checks just writing down numbers in the log sheet.The operator was dismissed by his employer for unethical behaviour.



Ensure you understand how to use NDT equipment.

Case Study: A manufacturing company had numerous customer complaints of defective material being delivered.

The company blamed the NDT inspectors and called me in to carry out refresher training.



The inspection area on the shop floor was dimly lit from high overhead lighting, when this was checked with a photometer a reading of only 110 lux was observed, the process specification required a minimum of 1000 lux.

The technician was misreading the photometer by a factor of 10 and claimed how he was reading it was how he had been instructed to do so by the previous quality manager. The technician was then instructed how to use this equipment correctly.



Once new lighting was installed the frequency of customer complaints was reduced significantly.



Ensure you never forget NDT method basics.

Case Study 1: A magnetic particle testing technique using bench equipment was raised by a Level 2 and submitted for Level 3 approval.

The technique stated testing the part using magnetic flow followed by a coil shot, this of course is basically testing the part twice using longitudinal magnetisation and the part would not be tested for defects in all possible orientations.

Had this technique been used defective parts could have ended up on an aircraft.



Case Study 2: A Level 2 NDT inspector raised a technique for magnetic particle testing and submitted it for approval.Upon review it was noted that the part was made of titanium! The data card was rejected.



Case Study 3: During an audit a magnetic particle testing data card for bench equipment was reviewed, it was noted that the parts tested with this data card were made of austenitic stainless steel.The data card had been used to “test” these parts for over two years!Operators were using correct test pieces to carry out shift checks but were not using flux indicators on the parts under test.



Case study 4: An audit was carried out on a facility manufacturing car engine parts that was using an automatic magnetic particle testing machine.It was noted that the current flow shot stopped before the automatic flow of ink stopped and was basically washing any indications away as demonstrated when a flux indicator was applied.No defects had ever been found whilst this equipment had been in operation.The timing of ink application was amended and the inspectors started finding defects for the first time.



Case study 5: During a performance review of an NDT inspector using fluorescent penetrant it was noted that he was wearing photochromatic spectacles in the UV-A inspection area.He had been using these spectacles for several months.



Keep an eye out for unethical behaviourCase Study: During an audit of an NDT facility completed follower cards were reviewed. The NDT penetrant testing operation had been stamped off with the stamp of an NDT inspector that had died several months previously.The investigation found that a non NDT certified member of staff had obtained the stamp and was using it to speed up production.Considerable quantities of parts had been despatched to customers without any NDT inspection!



During NDT audits don’t panic just do your job normallyCase Study: As an observer during a NADCAP magnetic particle testing compliance audit, it was observed that a young inspector was very nervous.He was testing some small parts in the heads of a bench unit and made the mistake of handling the parts before he had inspected them under the UV-A inspection lamp.The auditor wrote this up as a non-conformance and the inspector had to have a training session as part of the corrective actions.



Ensure non NDT individuals do not abuse or use NDT equipment.

Case Study: During a summer shutdown of a factory, the factory floor received a shiny new coat of paint, certain shop floor individuals who afterwards refused to identify themselves used the fluorescent ink bath of the magnetic particle testing machine to clean their brushes.

200 litres of fluorescent ink had to be replaced and a full shift had to be spent cleaning the machine.



Case Study: A large forging was tested with penetrant and a large crack was found. A shop floor handyman was assigned to re-work the forging by grinding with an NDT inspector retesting the ground area to ensure it was cleared.After several sessions of grinding and retesting without clearing the defect the NDT inspector went for a lunch break. When he returned from lunch the forging was on the back of a lorry.The handyman had “tested” the ground area and decided the defect was clear.He had in fact buried the defect under a thick layer of developer and further grinding was required.



Keep an eye open for the unexpected.

Case Study: During an audit of a fluorescent penetrant testing facility, a Level 2 inspector was demonstrating how to carry out a test for fluorescent penetrant contamination by taking a sample of penetrant in a glass beaker and viewing this sample.When I looked into the penetrant tank I saw something on the tank bottom, the inspector put his hand in and retrieved a broken tube from a fluorescent strip light.



As these lights have coatings on the inside of the tube the penetrant had to be discarded, the tank cleaned and replenished with fresh penetrant.



Ensure you review equipment calibration certificates.Case Study: It is fairly common that photometers and radiometers have to be adjusted by calibration engineers to bring them within tolerance of the relevant standards. Review calibration certificates and see if they have an “as received condition” which exceeds mandated tolerance requirements i.e. ± 5% and then an “after adjustment figure” which does comply with required tolerances. If this is the case a risk assessment is to be carried out.



For example in the as received condition a radiometer has an accuracy of + 10%, this means that it is reading the UV-A output from inspection lamps 10% higher than the actual lamp output; 1000µW/cm² as read is actually only 900µW/cm². A risk assessment would require reviewing the shift check results for UV-A lamp output since the previous radiometer calibration and find the lowest lamp output reading, subtract 10% from this figure and if this is still above the minimum required by the relevant standard there is no impact on product.



Ensure you use equipment correctly.

Case Study 1: Electromagnetic yokes should be checked with a test weight to determine if the lifting power of the yoke is sufficient. During audits it is common to see NDT inspectors using flux indicator strips with electromagnetic yokes and neglecting to use the test weight.



The problem is that if you put a flux indicator on any surface even non ferromagnetic materials such as a piece of wood or a slice of bread, spray ink and energise the electromagnet the magnetic flux between the poles will form an indication on the flux indicator – clearly we can’t test a plank of wood!

A flux indicator with an electromagnet can only show flux direction, which we should know anyway but will not determine if the flux density is adequate.



Case Study 2: During a performance review an operator was asked to test a part in a magnetic particle testing bench. He set up a current flow shot and placed a flux indicator on the part but in the wrong direction, when no indication formed he stated the machine was not working.He was then embarrassed when instructed to turn the strip at 90 degrees to the first orientation and three strong indications suddenly appeared.



Avoid distractions during testing Case Study: An operator was distracted by an urgent telephone call halfway through magnetic particle inspection of a large forging. When he returned to the job he thought he had completed the inspection and released it to the next operation.Another member of staff spotted a crack in the untested half of the forging some time later.Marking tested areas on large parts can help stop this type of mistake.



Don’t just look for small discontinuitiesCase Study: An aero engine shaft was returned from a customer that had been in-service in an engine for some time.During a routine engine strip down the shaft was tested using magnetic particle techniques and a crack approximately 100mm long was found.Metallurgical analysis determined that this was not an in-service defect it was a manufacturing defect and fortunately had not propagated during service.



The company that made the shaft concluded that NDT inspectors seldom found defects on these components and when they did they were only 2 – 3mm long maximum. The crack was running circumferentially and may have been mistaken for a machining line as the inspectors never expected a crack that big to be found.



Ensure you are correctly trained to operate NDT equipmentCase Study: A performance review was carried out on a NDT inspector who was using a magnetic particle testing bench unit.The inspector put a part in the bench sprayed ink on it and started inspecting it with a UV-A inspection lamp. He had not pressed the button to energise the current!When asked why he said he had never been told to press the button and did not know what it was for.Worryingly he had been “testing” work like this for over a year!



Be aware of human factors in NDTThe reliability of NDT can be significantly influenced by the environment in which parts are processed and inspected. Consideration of human factors is an important element in achieving process capability. Human factors are typically dependent on a large number of influences such as fatigue, environment, stress and complexity of task.



How can these human factors be addressed?

• Fatigue – rotation of tasks, duration of inspection times• Environment – working space, inspection booth

extraction, comfort level• Stress – improved working practices, minimise

inspection interruption• Complexity of task – appropriate equipment, fixtures

and training



ConclusionsMT and PT are in general easy methods to get right but as we have seen in this presentation there are many pitfalls we need to avoid, unfortunately it can be all too easy to get NDT wrong.• We must never be complacent, • We must never cut corners, • We must guard against making mistakes.

We must never forget how important NDT is; ultimately peoples lives depend on us to get NDT

right.


THANK YOUFor more information please contact:

Phil [email protected]

imeche.org/arl

Getting NDT Right in MT and PT

Engineering