 Remember what you learned in class about sharp corners acting like stress risers? That problem doesn't end with machined components. High profile welds can act like sharp corners that create stress concentrations; and it's no surprise to see them fail in fatigue at the toe of the weld.
Refer to the Material Properties - Case No. 66: Weld Toe High Profile Stress Concentration article for the background history of this failure. The weld that attached the skirt support ring to the vessel failed in shear. It failed along the weld toe. The face of the weld bead made a sharp angle with the vessel cone section, and this created a stress concentration. It took many years for this weld to fail, but fail it did. So what did we do about this the second time around? ... read on ... When this vessel was raised it became obvious that the cone section of the vessel had flared the skirt support radially outwards by 1/2" (Fig. 1) as it wedged into the skirt by as much as 3" (Fig. 2). This was anticipated. The skirt would be salvaged by slicing it vertically in several places and drawing it in to meet the cone section. These slices would have to be welded with full penetration beads. |
|
 |
 |
|
The attachment weld between the skirt support and the cone section of the vessel needed to be of a higher quality than the original weld. To insure that this happened this weld was going to be radiographed. The surfaces were prepared. The old weld was removed, the skirt was squared to accept the new attachment weld, and all other surfaces were ground down to bright metal (Fig. 3). |
|
 |
 |
| The first weld pass was laid in. It was a good example of a high profile weld bead, especially on the bottom toe with respect to the skirt. Both weld toes were sharp corners, and it mattered. The first weld pass was checked with dye penetrant for surface defects (Fig. 5). The dye penetrant easily picked up the surface defects. All surface defects were ground out, after which, two more passes were laid in on either side of the first bead. The profile was ground smooth (Fig. 6). After grinding smooth, the weld was checked with dye penetrant for any surface defects (Fig. 7). All surface defects were ground out and rechecked. | |
 |
 |
 After the three-pass weld was dye penetrant checked and the defects corrected, it was radiographed. The radiographs were interpreted according to AWS D1.1 specification, and did not detect any defects. At this point the stiffening rings were attached (Fig. 8). The last part of the job was to check the vertical splice welds on the skirt (Fig. 8 circled) using the ultrasonic shear wave technique (Fig. 9). The welders were very curious about how the defects were detected. Any defect reflects part of the wave back to the transducer, and shows up on the screen as a spike. Depending upon the size and configuration of the spike, the inspector can assess the exact nature of the defect. In one instance, a defect is detected (Fig. 10 orange spike) and subsequently it was ground out and rewelded.
The new weld profile had been improved over the old one. So what? So maybe this weld lasts three to five years longer than the old one. Or better yet, it never fails. |
|
 |
 |
|
All Pictures and Text Copyright © 2000 - 2016 Contact Mr. Adler Adler Engineering LLC of Wyoming USA
Great care has been taken in the compilation of this article. However, no warranty, expressed or implied, including without limitation, warranties of merchantability or fitness for a particular purpose, are given in connection with this article or any article archived on this website. Although this information is believed to be accurate by the author, the author cannot guarantee favorable results will be obtained from the use of this article alone. This article is intended for use by persons at their sole discretion and risk. Since the conditions of product or material use are outside of the author's control, the author assumes no liability or obligation in connection with any use of this information. The author is not liable for special, indirect or consequential damages resulting from the use of this material. 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. |