Reliability Engineering Snapshot TM

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

 

 

 

Material Properties - Case Study No. 112: The Necessity of Requiring a Welding Procedure Specification

 

Poor Weld Proceedure - Too Much Weld, Too Much HeatSomehow the phrases "I want a good weld" and "good welding practice" are not enough in a contract. If you want a good weld then you have to specify a welding procedure, otherwise your odds of getting a good weld are about the same as winning the million dollar jackpot on a Las Vegas slot machine.

This is what happens when there is no weld procedure specification, and no one supervising the job (picture left). This is what happens when Production personnel pressure Maintenance to finish the job when promised. This is what happens when Maintenance accepts a job that they should have never started in the first place. It would have been better for Maintenance not to start the job, unless Production agreed to allow the additional time required to do the job right. This monstrous weld would eventually cause additional problems and considerably more downtime later on. This weld would eventually cause a crack to form only two inches away, a few days later. The topography of that crack revealed that the failure was due to the welding procedure of the previous weld repair a few days before. It was not due to normal fatigue, as had been the case in the past on five other cracks.

Back in January, the stiffening ring on a rotary dryer had cracked all the way through the ring and into the reinforcing ring underneath it. The reinforcing ring is a thicker section of shell that handles the higher stresses developed by the tire. This serious structural failure had to be repaired. However, it was discovered by accident; other work had been going on around the dryer. This discovery would mean extra downtime. The contractor became a hero by finding someone to gouge out the crack and weld it back up within a 12-hour shift. The final weld profile was ground flush with the stiffening ring and the shell.

Crack View - Bottom Side

The stiffening ring cracked where the yellow line is drawn in the picture to the upper left. The fracture was recovered in the repair. A cross section of it is shown in the picture to the left. The outer diameter of the stiffening ring is at the top of the picture. The stiffening ring had failed over the years on several occasions. However, this failure was different. Normally the crack arrest rings, or beach marks, started as very small round circular rings (as they do here). As the cracks radiated outwards they always turned the corner on the sides of the stiffening ring, and eventually looked like horizontal lines across the width of the ring. This type of failure would be typical of a low-stress high-cycle fatigue failure. The crack initiation points were always located on the outer diameter of the ring (red dots, top of picture). In this fracture, there were four noticeable beach marks where the crack progression stopped for awhile (yellow dots); then the beach marks disappeared all together. The lack of horizontal beach marks wasn't right.

It appeared that this crack had already been in progress at the time that the other crack was repaired incorrectly. The crack had progressed radially outward from the red dots to the outer yellow dots. It was when the crack reached the last beach mark that the other crack was repaired incorrectly. When welding the other crack, nobody inspected the region for other cracks, otherwise this crack would have been discovered. The welder laid down a lot of weld in a short period of time. In order to do this, the interpass temperature had to be extremely high. See the article "Machine Design - Case Study No. 111: Welding Procedure Specification Key Parameters" for further details on the importance of interpass temperature. From knowledge we learned, from the correct repair (see the article), it took as little as three passes, with a high profile crown, to exceed the recommended interpass temperature for the material.

This joint was a highly restrained joint and because the stiffening ring got so hot during welding, the region near the weld expanded. When the weld metal cooled, it put higher tensile stresses in this region, and on the existing crack that lie dormant. The stresses were high enough to re-initiate the crack progression and maintain its progression to failure at a much higher rate.

 

View of Crack - Top Weld Side

 

Some tell-tale signs of a faster crack progression are shown in the three pictures below. Those pictures are close ups of the upper left, upper right and bottom right positions in the picture to the left. The chevron marks (enclosed in white squares) indicate the direction of the crack. In the upper left-hand corner of the stiffening ring (picture lower left) the chevron marks are pointing inwards toward the last series of beach marks (yellow dots). This meant that the crack path was travelling to the left in a horizontal direction. Normally, there is a beach mark in this region similar to that shown in the picture to the lower right. Remember, the normal crack progression is for the crack to turn the corner. It never did turn the corner in the picture to the lower left.

 

Close Up View of Top Left Corner Close Up View Top Right
In the upper right-hand corner of the stiffening ring (picture upper right) the chevron marks are again pointing to the beach mark located above them. In this case, the beach mark is turning the corner of the stiffening ring in a normal fashion. This crack started as a normal low-stress high-cycle fatigue failure. The bottom beach mark in this picture (bottom yellow dot) is where the beach marks usually become horizontal, and progress downward. The picture to the lower right shows the last traces of chevron marks. Below this point the chevron marks are missing because the progression of the crack is straight downwards. Again, what is missing in this fracture surface are the horizontal beach marks. They never got a chance to form because the stress increased significantly when the other repair was made and the heat affected zone of the weld, and the surrounding region, got too hot. When the hot region cooled, the old crack was put in a higher state of tensile stress. High enough to cause it to progress at a higher rate and without stopping.

 

 

We learned the hard way that welding is a science and not a crap shoot. Joint configuration and interpass temperature are two very important parameters to control in a highly restrained joint.

Close Up View Lower Right Side
All Pictures and Text Copyright © 2001 - 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.