Reliability Engineering Snapshot TM

Illustrated Case Studies in the Maintenance Engineering World of Failure Analysis, Predictive Maintenance, and Non Destructive Evaluation

When considering the affects of galvanic corrosion, don't forget the gaskets.

The elbow pictured to the left was removed from a pump volute. The pump was a cantilevered pump in submerged service and it had been in service for six years. During that period of time it had been removed for various reasons. However, this time there appeared to be a great degree of metal missing in the gasket region. The gasket specification recommended by the OEM was a fiber gasket, but the service it was used in called for a spiral wound gasket shown in Figure 2. The gasket that came out of service was never recovered and therefore a critical piece of evidence was lost. However, the corrosion pattern was pretty obvious and a new gasket was positioned exactly as the corrosion pattern indicated. The position of the old gasket is shown in Figure 2 with the new gasket. When looking at the corrosion pattern in Figure 3 and then looking at the offset position of the gasket in Figure 2 one might be quick to think that turbulence eroded the region of greatest metal loss.

However, when a ruler was placed over the flange face (Figure 4, left) it revealed that the flange face nearest the inner diameter of the pipe elbow was still there. Only the metal directly underneath the spiral wound portion of the gasket was missing. Erosion was ruled out.

The next possibility was crevice corrosion. Crevice corrosion requires very tight clearances where stagnant conditions can exist. It is not uncommon for some gaskets to wick and allow favorable conditions for crevice corrosion to initiate. When looking at Figure 5 there were deposits that accumulated between the retaining ring and the chrome-steel flange face. The imprint of the retaining ring remained on the deposits. There were clear signs of green oxides where chromium reacted. When the deposits were removed, as shown in Figure 6, the flange face was still in good condition. Crevice corrosion is not preferential in where it attacks, yet there was a clear demarcation line between the attacked region (under the spiral windings) and the retaining ring that did not come into contact with the flange (as evidenced by the accumulation of deposits shown in both figures).

If the cause for the metal removal on the flange face wasn't erosion or crevice corrosion, then what was it? It was certain that there was evidence of a corrosion mechanism. There were definite signs of corrosion byproducts in Figure 3. The green regions were chromium oxides. A close up view of the lower right region in Figure 3 can be seen by clicking on it. The elbow was a 29% chromium-iron material. The gasket was a 316 stainless steel spiral wound gasket with a 316 stainless steel retaining ring. If it were galvanic corrosion then a clear circuit path would have to exist between the flange face and the gasket.

The way galvanic corrosion works is quite simple. It behaves the same way as what happens inside the common household battery. The most important ingredient is the electrolyte which can be anything that will support an electric current. Since this pump was submerged in liquid elemental phosphorous, a readily available electrolyte was present. The next thing that was necessary was the close proximity of two dissimilar metals that when placed in close proximity to each other, in the presence of the electrolyte, would exchange electrons. One metal would give them up (anodic) while the other would accept them (cathodic). When this exchange of electrons takes place, its called electricity. The more dissimilar the two metals, the greater the exchange rate of electrons, and the greater the corrosion rate.

The 29% chrome steel and the stainless steel are very close to each other on the galvanic series chart, with the chrome steel being slightly more anodic, or willing to give up electrons. The design engineer might have been willing to shrug this one off and call it good. However, stainless steels, and in particular chromium, show a tendency to go both ways. In other words one minute the chromium alloy can be anodic and give up electrons and the next minute it can be cathodic and accept electrons. All of this depends upon the electromotive potential between the two dissimilar metals. It can also depend upon temperature and pH, and most importantly, the local electromotive potential (e.g. area affect and/or ion concentration). When stainless steel is anodic it's ability to give up electrons is close to that of the 29% chrome-steel and so there isn't enough difference in the two to develop a current. However, if stainless steel decides to go cathodic then the electromotive potential is very significant and a powerful battery is created. Additional clues to the galvanic affect were to be found on the other flange face.

Figure 6 shows the overall view of the 2nd flange face. The winding pattern from the gasket is clearly indented into the softer chrome-steel face. Figure 7 shows a close up view of the 2nd flange face. In the case of the electromotive force, the stainless steel went cathodic with respect to the chrome-steel. There was a high potential difference and a high current was created. The electric circuit was completed through the flange bolts, and the current traveled in a circle from the flange face to the stainless steel windings and back again through the flange bolts. The white dots denote specific winding locations. It is obvious that this flange should have been retired long ago, but as economics would have it, the flange was saved. In all likelihood this flange leaked to some degree but never was detected in the submerged service. The leakage provided the necesary electrolyte. The pit shown in the figure was an obvious location of high electromotive force, or a good contact area.

No part of this article or any article archived in this website, or any part thereof, may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, without the prior written permission of the copyright holder R. H. Adler. Nothing contained in this article or any article archived in this website shall be construed as a grant of any right of manufacture, sale, use, or reproduction, in connection with any method, process, apparatus, product, composition, or system, whether or not covered by letters of patent, copyright, or trademark, and nothing contained in this article or any article archived in this website, shall be construed as a defense against any alleged infringement of letters of patent, copyright, or trademark, or as a defense against any liability for such infringement.