Adventures in titanium—Why is my stovepipe hat assembly warping, turning blue, and cracking?
Fabricating a titanium stovepipe hat assembly had one shop perplexed. What was causing the material to turn blue, warp, and crack? Why was it so difficult to make this part fit into the next assembly?
Editor’s Note: In 2004 David Hass wrote an article about titanium welding that has become one of the most widely read articles on thefabricator.com. The article generated many questions from readers about the process. Hass answers some of these in his new column, Adventures in Titanium.
“Who me, author an article?” After e-mailing an author of a welding article posted on thefabricator.com to clarify what I thought might have been an error, I found a new friend in the author and a new mentor in the Web site’s content manager. Being past the half-century mark in my own life, gaining a mentor was unexpected.
My mentor said, “Write about what you know.” So I wrote about titanium, and the article, “Titanium—you can weld it,” was published online. That was seven years ago, and I’ve received many inquiries from readers about various titanium welding projects since. If you think you know all there is to know about a subject, I have news for you. Until you publish online and take questions from the audience, you don’t know what it is you don’t know.
What’s Wrong with My Assembly?
Aimee sent this question, not expecting anything, but she was at her wit’s end. A titanium stovepipe hat assembly her shop was fabricating was warping and turning blue and wouldn’t fit into the next assembly without hand-fitting and a lot of muscle to wrestle it into place. Every time her welders bent the welded parts, the parts broke or cracked. Coworkers had no idea what the problem was, much less how to fix it. After a quick e-mail introduction, we resorted to a telephone call.
“It’s blue! But the raw material looked so good – all nice and shiny.”
“I’m sure it did. Are you preparing the edges?”
“We grind them, but it may be a couple of days between grinding and welding.”
“Clean atmosphere in the shop?”
“Pretty clean … the grinding dust sometimes blows around, but we clean up every day.”
“Do you use air filters on the shop air that powers the grinders?”
“Do you skip-weld?”
“Oh, the pictures you sent made me think you were welding the cylinder to the circular plate in one weld.”
“That’s right, we do!”
“That’s why it warped. Heat treating the parts before the weld?”
“We do a solution treatment before and then an age afterward.”
“OK, I think I have a couple of ideas to double-check in your process.”
If the atmosphere during heat treating is way off, you will see the colors. If it is off by only a couple parts per million (PPM), the color may not show up after the heat treatment, but the oxide already exists on the surface. Welding heats the oxide again and allows it to grow, changing from clear to a thickness where interference effects color the surface with beautiful iridescent colors.
Organics (like grease and oil) have a nasty habit of showing up on one part as a stain (which is easy to sort out), but also can contaminate all of the pieces in a furnace load with an invisible oxide layer.
Over many years of use compressors can leak oil and fail to filter out moisture. While I’m all in favor of replacing worn-out equipment with new, my managers over the years have been reluctant to wave a magic wand and fund my entire wish list. The option is to keep using the old compressors, but for critical operations, filter the air so oil and water are not expelled through the hand grinder head.
These filters never seem to get much maintenance, so my preference is a dual-stage, self-cleaning filter. Filters that are fed from the inside diameter have worked best for me and the sales guys tell me it is because the filter element expands as the pores clog. Filters fed from the outside diameter tend to collapse the pores as the filter becomes loaded.
Your shop’s maintenance regime will dictate the best choice. Preventing oil and water from blowing out of a grinder head is a good idea, especially with titanium.
Shops that grind a lot of titanium use downdraft booths to control the dust and ensure it does not become a fire hazard. Smaller shops may not have the overhead to afford this technique, so grinding in a specific room might be an alternative.
Welding curtains can help somewhat, controlling both the air currents and providing a safety screen for passersby, but dust usually is able to pass over or through them.
Any exhaust filters in use should be monitored frequently to ensure titanium dust has not clogged them or penetrated a motor mechanism. Any titanium dust on floors should be cleaned up expeditiously.
Titanium doesn’t conduct heat well, so when it is heated, it takes longer to cool than other metals. By welding an inch of bead on one side of the part, stopping, and skipping (Figure 1) to the other side of the part, the welder is working on cold metal throughout the welding process. Like tightening the lugs on a tire, no one side is stressed more than another. For the part size Aimee was welding, a five-step star pattern looked like a good balance to let the metal cool between welds.
So to recap: Clean the parts before they go into the furnace; watch the atmosphere and keep it pure; grind some distance away from where you weld; and weld in a clean area very soon after grinding. All this just seems like something you’d do to get a Good Housekeeping® seal of approval. None of these suggestions are very technical and appear to me as commonsense machine shop procedures.
One of the quality checks I have used over the years has been to keep my own scorecard to track projects—which projects remain in place six months later and which disappear within a year. So time passed, and I called Aimee a day or two after sending an e-mail to find out the status of her application.
There was a problem in their heat-treat furnace—one of the gauges was out of calibration. The welders had told her there was less slag when they ground and welded in quick succession. They still had a problem with warped brims on the top hats with a single-pass weld, but there was less cracking, and they could live with the arm wrestling to fit the next assembly. OK, not a home run, but a 0.660 average is better than Babe Ruth’s. Anything else I could do to help? No, well, keep in touch.
Nothing like a satisfied customer. How are yours?