Back to Basics

[ click here for the Back to Basics Archive ]

Monitoring Capability of Penetrant System Performance Panels

by George Hopman*

Ever since the TAM panel was designed, its use has been written into countless specifications, and sophisticated equipment has been designed to photograph and measure the indications produced on the panels. However refined the panels and the ancillary equipment have become, little work apparently is documented verifying their ability to actually do what the specifications require. This paper gives the results of some tests made to determine whether the panels are able to detect when a penetrant system is operated outside of the specification limits. It should be noted that the tests were made with a limited number of panels and a limited number of tests, and may not be representative of all panels. Nevertheless, the conclusions should be of great interest to penetrant system operators, auditors and specification writers. William E. Mooz Associate Technical Editor

 

INTRODUCTION

One of the more controversial issues within the aerospace industry is the use of starburst panels for monitoring penetrant system performance. Countless hours have been spent in performance of this test, managing the cleanliness of these panels and responding to related audit findings. But what is the capability of these panels to detect variability within the penetrant process? There does not appear to be any published studies on this issue.

There are process controls in place for each step of the penetrant process. The stated intent of the system performance test is to detect sudden changes in the systems that are not detected in the tests already performed. Regarding this test, one major aerospace engine manufacturer mandates that, "The defects in the standard will be capable of demonstrating unsatisfactory system performance" (Allison Engine Company, 1999). What is this standard? Is it the starburst panel?

---

The results would seem to indicate that starburst panels are not very discriminating in detecting process variability.

---

This paper documents experiments not performed under laboratory conditions, but done in a real-world shop atmosphere where the process was deliberately violated. A baseline was established and documented with a black light photograph. Then mandated process variables were changed while running the panel and pictures again taken to determine if the panel could detect the change. Nadcap HB7114/1 (2008) states that, "Results of the system performance test, utilizing the in-use materials, must indicate the same number and appearance (e.g. size, etc.) of the flaws detected originally when the baseline was established." In practice, this "standard" is subjective, yet this was the criteria used to judge the acceptability of these tests.

BASELINE PROCESS

Level 3 water washable and post-emulsifiable penetrants from the same manufacturer were chosen to baseline the process in this limited experiment. It is generally recognized that there is significant variability between brand new starburst panels. A more expansive study would have included variability between panels and other brands/sensitivities of penetrant material. The panels were stored in acetone and cleaned with acetone before each use. The panels were checked under ultraviolet illumination before each use. If any fluorescence was noted, the starbursts were sprayed with nonaqueous developer and baked in the penetrant drying oven. If any fluorescence persisted, the panels were sprayed with replica transfer coating to pull any remaining penetrant out.

The clean panel was brushed with virgin penetrant and allowed to dwell exactly 10 min. Since ASTM E 1417 (2005) does not require a washability or emulsifier removability test unless the starbursts fail, only the starbursts were coated with penetrant. The panel was rinsed at no closer than 0.3 m (12 in.) under proper illumination with water at 241.3 kPa (35 lbf/in2) with a temperature of approximately 304 K (88 °F). The panel was blown off to minimize water spots, then dried (approximately 5 min) in an oven at 333 K (140 °F). Virgin dry developer was applied by puffer bulb and allowed to dwell 10 min. The panels were photographed under consistent, proper ultraviolet illumination using a 10 megapixel digital camera with a yellow filter. The camera was affixed to a stand for consistent results. For Method D removal, a pre-rinse was performed. The emulsifier (a different brand) was applied by spray at a concentration of 3% for approximately 10 s, and a standard post-rinse was performed.

CONTAMINATION

Multiple studies have documented that water and other contaminants will negate a proper penetrant test. That is not debatable. Perhaps the worst contaminant is acid. For those who work in a foundry, pre-penetrant etch is often a necessity. If a proper cleaning is not performed before the penetrant process, the potential for system degradation emerges. 5 mL of water washable penetrant was contaminated with approximately 1% volume of nitric acid and a sample painted on the starbursts. The penetrant itself underwent noticeable physical changes. Several certified inspectors remarked that the penetrant didn't look right when dwelling on the panel. Duplicating the baseline test with this contaminated penetrant yielded a satisfactory system performance test. Figure 1a shows a unique panel used for the baseline and the nitric acid contamination run. The picture obtained after the resultant test did not yield any noticeable difference from the baseline photograph.

Figure 1 — Starburst panel comparison; (a) baseline and nitric acid contamination; (b) baseline and 20% water contamination.

Water is a more common contaminant within the penetrant process. Sources of water could include rinse overspray, a leaky roof, and parts that were not properly dried in the pre-clean process. Per ASTM E 1417 (2005), recycled penetrant is checked monthly for a maximum of 5% water contamination. Water washable penetrant was contaminated with 20% water. The penetrant became cloudy with the contaminant. Figure 1b shows a unique panel used for the baseline and the contamination run. The contaminated panel was under-washed, but did not yield any noticeable difference from the baseline check.

PENETRANT DWELL

Depending upon the discontinuity morphology, penetrant dwell time may be important. The starburst panels contain open cracks that are supposed to be clean. They are not tight fatigue cracks, forging laps or cold shuts that would justify a relatively longer penetrant dwell time. The second edition of ASNT's Nondestructive Testing Handbook on liquid penetrant testing (1982) states that penetrant takes only 2 s to enter an open crack. Figure 2a documents the baseline and Figure 2b shows the panel's appearance with only a 30 s penetrant dwell time. The 30 s penetrant dwell panel was obviously under-washed, but the starburst results are essentially identical to the baseline.

Figure 2 — Starburst panel comparison: (a) baseline; (b) 30 s penetrant dwell; (c) 14 h penetrant dwell.

How about violating maximum dwell time? Penetrant will not dry, even with lengthy dwell times at 325 K (125 °F; Hopman, 2003). The water washable penetrant was allowed to dwell for 14 h. The penetrant did not dry, washed off easily and did not appear any different than the baseline photograph (Figure 2c).

RINSE PARAMETERS

The rinse water is already monitored with a temperature and pressure gage, but can the panel pick up temperature violations? The water washable panel was wiped with water that was heated to 358 K (185 °F) in a microwave. Water wiping with a clean cloth is an authorized form of removal in accordance with ASTM E 1417 (2005). Figure 3a documents the water washable baseline and Figure 3b shows the panel's appearance after wiping the panel with water heated to 358 K (185 °F). The results did not appear any different than the baseline photograph.

Figure 3 — Starburst panel comparison: (a) baseline; (b) 358 K (185° F) water; (c) 274 K (33° F) water; (d) overwash.

The water washable panel was wiped with water that was cooled in a freezer to 274 K (33 °F_. Figure 3c documents the results of the penetrant removal deviation. The results did not appear any different than the baseline photograph.

The principal variable within the rinse booth is the inspector. The inspector controls the distance and the time of the wash. The water washable panel was washed at a distance of 102 mm (4 in.) for 2 min. The results (Figure 3d) show a degradation in the largest starburst but not the smaller ones. Other iterations of this test showed better results with the largest starburst.

Figure 4 — Starburst panel comparison: (a) baseline; (b) 3 min spray emulsification at 102 mm (4 in.).

A unique starburst panel was used for the post-emulsified process. Again, the principal variable within the rinse booth is the inspector. The Method D emulsifier was sprayed at a distance of 102 mm (4 in.) for 3 min. Figure 4a documents the post-emulsified baseline and Figure 4b shows the panel's appearance with an abusive emulsifier dwell time. The results are slightly degraded but within acceptable variance.

The drying process may be critical to the penetrant operation. Multiple findings have been written against users for leaving parts in the oven longer than necessary. Depending upon the temperature, panels only take a few minutes to dry completely. The water washable panel was dried six times longer than necessary (30 min). Figure 5a documents the water washable baseline and Figure 5b shows the panel's appearance after drying in the oven for 30 min and completing the proper developer dwell time. The test panel was under-washed, but with the exception of the smallest starburst, the results did not appear any different than the baseline photograph. No panel was run without developer since the lack of wet or dry developer is easily detected.

Figure 5 — Starburst panel comparison: (a) baseline; (b) 30 min drying at 333 K (140° F).

CONCLUSION

The experiments conducted were deliberately extreme in order to demonstrate sudden drastic changes rather than gradual changes. The results would seem to indicate that starburst panels are not very discriminating in detecting process variability. In fact, the originator of the panel, Sherwin, Inc., has stated, "The PSM 5 or any TAM panel is incapable of demonstrating system degradation."

In light of this, consideration should be given to the expectations for this process control. It is doubtful that this test has ever prevented a defective part from escaping the penetrant testing process. Perhaps this test should be de-emphasized by the industry, and more attention should be focused on the primary causes of escapes (parts that aren't clean) as well as inspectors who don't take the proper time and attention to inspect the component.

REFERENCES

Allison Engine Company, EIS 1169 - Fluorescent Penetrant Inspection, Indianapolis, Allison Engine Company, 1999.

ASNT, Nondestructive Testing Handbook, second edition: Volume 2, Liquid Penetrant Testing, Columbus, Ohio, American Society for Nondestructive Testing, 1982.

ASTM, ASTM E 1417: Standard Practice for Liquid Penetrant Testing, West Conshohocken, Pennsylvania, ASTM International, 2005.

Hopman, G., "Extended Dwell Times and Drying of Penetrant," ASNT Fall Conference and Quality Testing Show, Pittsburgh, Pennsylvania, 13-17 October, 2003.

Nadcap, HB7114/1 - Nadcap Requirements for Nondestructive Testing Facility Penetrant Survey, Warrendale, Pennsylvania, National Aerospace and Defense Contractors Accreditation Program, 2008.

 

* NDE Solutions, Inc., PO Box 30085, Phoenix, AZ 85046; (602) 595-1033; e-mail george@ndesolutions.net.

Copyright © 2009 by the American Society for Nondestructive Testing, Inc. All rights reserved.

[ Back to Materials Evaluation ]