to the dimella we go!
We are building a large bearing hub assembly that is a cyclically loaded component which will see a lot of impact stress in the field. It is an A36, 2” base plate about 3’ long and about 10” wide with a heavy wall 7” A633 hallow heavy wall casting in the center. We are welding within the WPS we were given from the client and are running an E71T-1C, after a Tig root. The casting goes fully through the base plate and the welds on both sides are single bevel grove with additional 3/8” fillet reinforcement. There are 2 stainless 17-4 pins on both end of the plate that are also through the plate, as well as single bevel grooves and reinforcement fillet welds added according to the print.
The problem we are getting is a delayed cracking, not seen by the naked eye but we did our own in house Mag particle check and there are tiny tears here and there along the weld. We are pre heating to the WPS completely, we are making sure the heat is solid even 3+ inches outside the weld zone to be safe. It takes about 3 passes with our FCAW-G to finish the proper weld size for the center hub, and we don’t seem to have any cracking issues right away. We let them sit till they are room temp and still they are ok, when we take them to stress relieve however it seems they come back with these tiny tears. We have had some that have had small tears before they left for stress relieve also, just not as frequent. They are held at 1000 degrees F for 4 -6 hours then let to air cool in ambient air. When they get Mag tested again they have issues. We have had a slew of NDT people here from different companies, one guy passes all of them, one guy rejects all of them, one guys sees only a few issues. Confusion.
Now to the stainless pins… Al has helped me out here a little before this, but I think I have seen first hand now some welds stretching. On the back side of these pins is basically a single bevel groove created by two parts. A 90 degree edge on steel and a 45 degree bevel on the pin, with the pin being .100 above flush with the base plate. Using SMAW 309L-17 as per their WPS at 150 amps with proper preheat, we weld the circle. Once it cools, right at the start of the weld, maybe an inch into the path a center line crack begins to appear. Its not so much a crack at it looks like a tear, as if while one side is still solidifying the other side is pulling it back and the crevasse is created. When it appears, it is in the same spot every time. Mind you these are tacked very well in 4 places prior to welding, with the pins fitting near perfect into the hold with zero slop. Also the Mag guys who test the pins didn’t seem to pay too much attention to their teacher, or they are not familiar with stainless but there is always a ring around the stainless welds when they are welded to dissimilar metals, in this case A36. The guys who haven’t seen that before marked all our parts as fully cracked because they didn’t know what they were looking at.
Now my questions: On the main hub, do we need to brother in law that section to keep heat from pulling so much? Its tough because everything we are welding is a dissimilar metal, from the hub, to plate to the pins. We are exhausting our options on finding the cracking issue. Not sure if when it goes to heat treat that is messing with the expansion and contraction? Maybe throw all the parts in an oven and pull them out and weld them instead of localized heat? They gave us these parts because they don’t want to do them for obvious reasons. I don’t want these parts to kick our butt.
If i missed a detail ask away!
Thanks
Jordan
Jordan,
At first glance I am seeing several issues. You are dealing with a variety of thicknesses, materials, and joints that are going to yield a product that has parts with a large variety of expansion and contraction rates.
In my opinion you would definitely do well to heat the entire unit instead of using localized heating. But, it will also take making sure weld passes are done with a proper throat and try not to get too wide with the surface tension effect. You may need to change filler material in order to get something more forgiving as to it's own expansion and contraction in order to allow it to move with the parts and not pull thus tearing the base material.
I am no expert in metallurgical issues but it looks like you have a whole lot going on that really needs some expert input.
Have a Great Day, Brent
Brent,
Tell me about it haha, we have had engineers here to try and figure it out and were about to start our next batch and this will be the test to see if all these changes made a difference.
Thanks
Jordan
Sounds like you have a reheat cracking issue. Look up J=Factor and X-Factor on the internet and research the chemistry of your base metal and filler.
Generally with C-steel you don't get a reheat crack issue but if you have a weld configuration in high stress you may.
Also, the Euroweld site has good info and I believe Sperko's site does too.
Can you expand on this part of your OP?
"It takes about 3 passes with our FCAW-G to finish the proper weld size for the center hub"
Tim
It requires a 3/8" reinforcing fillet weld around the center hub, using 1/16th FCAW-G... take us three passes to achive the proper size weld and we switch our starts and stops in a 90 degree rotation each layer of weld so that everything is even.
Jordan
Ok, so just to be clear, you're using E71T-1C filler metal with CO2 shielding gas, not innershield flux core, right?
Probably a stupid question, but I'm only familiar with the "FCAW-G" expression in relation to gas-less innershield.
Also, I'm curious about the groove width to depth ratio, which if too deep and narrow can lead to cracking. This would be in the groove portion however and not the reinforcing fillet, unless it was somehow transferring through eratically. How deep is the groove and what bevel angle?
A picture is worth a thousand words...
Tim
Tim
FCAW-G is a gas shielded flux cored wire, and yes that is the correct wire and shielding gas. I cant post up a print but I can tell you its a .25 single bevel groove weld with a .38 reinforcing fillet on top. I believe the bevel angle is 45 degrees.
Jordan
Hello Jordan, has this ever been done before or is this a first time item? I am asking because it's often easy to draw-up an item to build, yet it sounds as if some of the materials engineering may be lacking and it shouldn't necessarily rest on your shoulders to fix it. So the jist of my question relates to whether this is an in-house item or is it something that has been determined on the outside and your company has taken it on. If it's from the outside I would be having some conversations with them to address the issue as well. Good luck and best regards, Allan
Allan,
Yea they build these themselves, but the come to tell us that out of 105 welders in their shop MAYBE 3 guys can build these without issues, where I am at we have two welders. This job has really put on the stress tell me about it. When the clock strikes 3:30 I put it away. But yea I can see there being many material issues here. I think the company has ran into so many problems that they are considering making the plate and the center hub one giant casting and then just dealing with the pins. Whatever the deal is its in our lap until furthur notice.
Thanks
Jordan
Hello again Jordan, with their own odds stacked so far against success on these in the first place this sounds like a case of a pass-off. I know that my comments don't help to change your issues here, yet I still question the level of your companies liability also. I do hope that you are being able to pass on the cost of all of your trials in this process, otherwise, I believe that I might be looking for a way to exit stage left.
I also wonder what the long-term success of these items has been with the poor rate of successful completion by it's initial producers. Are these a constantly replaced item due to failures? Do they inherently require replacement on a regular basis due to design/service issues? Are they trying to hold you to quality standards that simply don't exist? or cannot? Thinking out loud Jordan, I hope that you get some answers that will allow you to get through this before you develop a bunch of gray hairs. Good luck and best regards, Allan
Allan,
Indeed I do not need grey hairs at 28, that would not be a good sign haha, I have indeed considered the liability and am not in control of what my company does but I have emphasized on more than many occasions that we might not be capable of doing this. Also it is well beyond the code, we surpassed the code many months ago. we are welding now to an "opinion" standard, set by some magical veteran welder one guy once saw in a movie for all I know, asthetically that is. Now I have been welding for about 10 years as a combo welder and jack of all welding forms, doing the CWI thing for almost 2 years and Ill tell you, the welds on these are clean as a whistle. Layerd smooth, and the SMAW is some of the nicest I have seen. But it seems we are welding again to the opinion of a man in Italian loafers. I dont think the items are constantly replaced , but they are a core component in a seriously big money making machine. If these parts were to break fully they are easily in a position to endanger people. I couldnt have been more clear to the higher ups about my position on taking on a job like this but again, they run the show and I think the look at my age like I know nothing. Tis life till those gray hairs show up I guess.
Jordan
Out of curiosity, what is the complete nomenclature and manufacturer of the E71T-1C your using. Also what is the tig filler metal your using?
ran with CO2------- C- .02%..... Mn 1.3%.... Si .5%...... P .010%..... S .013% yield at 75 KSI tensile 85 KSI Elongation in 2" at 26% ESAB 7100 Ultra, runs very nice. Tig wire is an ER70S-1 Everything is withing their WPS given to us to build these.
Jordan
I think it would not be hard to get a welding engineer from the filler manufacturer, they should be able to double check the chemistries and WPS. somthing changed if the process was qualified
Tim,
Just for the record, FCAW-G is always Flux Core Arc Welding with Gas Shielding.
Some electrode manufacturers use the G for designating their gasless electrodes also called innershield wire. It can be more than a little confusing for those not versed in the designations.
But, under all circumstances, when designating the welding processes, FCAW-G is defined in all AWS documents as being Flux Cored Arc Welding-Gas Shielded.
Have a Great Day, Brent
OK, thanks.
I used to think that once I became old and wise I would stop making stupid assumptions.
Now that I'm older, I've learned that I can blame making stupid comments on becoming senile...
Tim
Hey Jordan,
"...there are tiny tears here and there along the weld."
Where exactly? Toe? weld metal?
I think you have a hydrogen cracking issue. You can bump up the preheat on the casting (use oven), and/or reduce the high stress levels created by circular welds (but I don't think redesign is an option).
Tyrone
The thought of hydrogen cracking occurred to me as well.
Although I have not used the FCAW process in decades, I seem to recall hearing or reading something about flux contamination with moisture. Hopefully someone with more practical experience in this area can chime in to address this concern. Does FCAW wire need to be stored in a rod box like low hydrogen electrodes, or is the tubular wire sealed sufficiently to block out all moisture? I have heard of rust causing issues, but what about contamination of the wire core?
I suppose I could Google for some report data, but I don't have time at this particular moment.
My other suggestion is to draw a traditional fishbone diagram problem solving chart to list all ideas from the brainstorming effort. Then, the concerns that are not bad actors can be crossed out one by one. The fishbone needs to address: methods, materials, machinery, and man/skills as a minimum. I think this thread is off to a pretty good start though.
You need to know what type of crack you are dealing with first. Then fishbone it to determine the root cause of that type of crack. Different types of cracks may have different fixes.
Tyrone
The crack types could be part of the fishbone diagram process as well, maybe even on a separate diagram.
We really don't know what their met lab capability is yet. Regardless of the type of crack, due diligence should be exercised to eliminate all possible causes.
Your point is well taken though - the root cause should be identified and eliminated.
Tyrone,
I would say MAYBE longitudinal cracking, in the weld metal, but honestly the cracks are never consistant, They are usually 1/4" long but vary in direction in accordance to the weld itself. Some are at a 45 to the weld direction some run linear. The welds on the back side of the stainless were longitudinal, when you looks close at them even with a magnifying lens the metal is actually pulling part from the other. Its not a clean crack, and whats really odd that I have been trying to wrap my head around is that this tear on the pin comes after the opposite side is already fully welded out, were talking a .300 fillet weld 2 inches behind the plate, so I cant see the pin itself moving. It fits into the base plate like a glove with no room to move anyways.
Could it be that maybe the large A36 plate sucks so much heat from the weld that it could literally suck the weld apart?
We have an oven, and yes I plan to use it. It just seems after all the trial and error we are still at square one in some ways.
Jordan
I think this would be a good opportunity to try out vibratory stress relief, not so much as an after the fact replacement of thermal stress relief, but for in-process weld conditioning to prevent stress build up.
You might consider giving Paul at Bonal a call to set up a trial run.
Tim
Paul Pieprzyk
Bonal Technologies, Inc.
1300 North Campbell Road
Royal Oak, MI 48067
Tel: 248 582-0900 x245
or 800 638-2529
Fax: 248 582-0901
www.bonal.com
paul@bonal.com
Yep, solidification cracking. Rapid cooling of the weld and high stress (circular weld + no plate movement). You'll get tiny fissures immediately after welding, then they grow over time due to the residual stresses.
Gonna have to increase preheat and slow down your travel speed when welding. Do your homework to figure out the max preheat temp you can safely go on the material you are welding.
Tyrone
I tend to agree, but what about the delay factor? My experience has nearly always been that solidification cracking occurs very soon after welding.
Maybe there is a perfect storm of multiple factors working here.
My approach would be to just eliminate all possible contributors through the use of best practices.
The first two of the new batch are just wrapping up then MAG at noon. I will keep you posted
Jordan
Not sure you can use the "other companys" WPS and remain within the code without capital "E" engineers approval
We have the engineers approval, he was here last week trying to help us.
Solidification cracking is a hot cracking caused by tramp elements in the grain boundaries during the 'mushy' period of solidification. It happens immediately, is predominant with higher alloys, and is intergranular. Increasing preheat will not help and will probably aggravate the problem.
Hydrogen cracking is generally delayed because it is caused by the migration of H into voids in the microstructure creating a pressure that initiates cracking and grows over time and is transgranular..
Reheat cracking will take place subsequent to PWHT as J/X factor elements migrate to grain boundaries causing cracking. It can also happen with long periods in sensitive temp ranges in service and is clearly intergranular. Preheat will not help here either and may aggravate.
Any ideas on correction or prevention?
Jordan
Do you have actual chemistry on all the base metals, and electrode? Pix/sketch woul be helpful. Pix of the cracks, especially micro will assist in characterizing what is going on.
It looks like the 633 is about 2x as strong as A36, are your tears more on the A36 side
Try a single pass on the 633 joint and let it sit, eliminate reheat cracking due to multipass. If it cracks after PWHT, reheat is suspect
As Jeff says above, the SS welds are hot cracking, centerline cracks indicate solidification cracking, probably a chemistry issue aggravated by restraint
Yesterday we had a mag test again and the center hub(633) and the base plate (a36) didnt have any cracks. The only cracks we had was an A36 to A36 fillet weld which was easy to fix, that crack however was longitudinal and two centerline cracks each about 1/4" long. We think that was just bad technique. One pin had a crack also much smaller than before. The crack was right at the beginning of the weld. This is a small circular weld on the back of the plate between the 17-4 pin and base plate. About a Print... prob cant do that, not sure if thats ok with my company even though I agree it would help a lot.
Come Monday I will take photos and post them to this thread and you can take a look
Jordan
You have to determine if its cold cracking or reheat cracking. You said that you noticed it after PWHT which of course would indicate reheat cracking, except reheat cracking is rare with C-steel. You may just not have noticed it if it was small.
If it is cold cracking (H cracking) you have to use very stringent low hydrogen methods with highly stressing weld configurations.
If it is reheat cracking you have to use low X/J factor weld materials and decrease the time in the sensitive range.
Here are two photos of the back side of the HUB so the center section is the A633 and the plate is the A36. Also there was a bit of an optical illusion with the camera, that bottom weld on the outer edge of the ring is flush or slightly lower than the weld about and behind it. It looks by the photo that that weld is high and standing out, but I assure you it is not.
Also we lay welds this clean and they come back "not asthetically pleasing" they will most likely say the toe edge isnt a perfect blended transition or the whole stack of welds isnt perfectally flat.
Also, just to add fuel to the fire, is the MT inspection finding all the cracks that exist. Might be interesting to weld up the single bevel (presumably it is cut in the A36) and have a look, make sure you don't have a real deep, narrow weld, which promotes solidification cracking due to grain boundary planes and restraint, I think you have more going on than this, but another thing to check
You have to determine if it is H cracking or reheat cracking, which means you have to know for sure if it is cracking prior to HT.
If the MT is proving inconclusive you may have to go to either wet fluorescent mag or PT which will give more defined results.
Another way is crack configuration. Keep in mind H cracking will be transgranular and reheat cracking is intergranular, which means the reheat cracking will be more 'spider webby' and (I think you can find this term in AWS A3.0), and branched.
It is important you determine this because the cures are exactly opposite.
And by the way, your weld configuration is so notorious for high stress and cracking tendencies it is actually used in a weld crack sensitivity test for filler metals.
I agree and to add to the poor choice of bead sequence Jeff pointed out, I would have used a different weld sequence because it clearly shows a stress riser that is the undercut into the weld closest to the the OD of the part welded to that face plate which is probably why there was the comment of not having a smooth transition @ the toe of that weld...
I see that the base plate is 2" thick correct? Well, then how many passes are being used in order to completely fill the joint minus the final cover pass???
Did anybody suggest using a temper bead technique and/or was there a buttering of transition weld layer of 309L SS deposited around the OD of the 17-4 SS pin prior to fitting the pin into the hole that mates it to the 2" A-36 plate along with widening the joint configuration to take this into account by any chance???
I'm just think out loud Jordan but, I believe Jeff is giving you some excellent advice and yet like he said earlier, there must be more to this dilemma that what may be involved.
Good luck on this problem.
Respectfully,
Henry
As of right now and the last batch of 6 we made, we havent had any cracks on the main hub prior to heat treat. Now today we are getting them back and ill check them again. I think I have the 17-4 pins situation figured out. We took the entore plate and hub and put them in our oven to 300 degrees and take them out and they get welded, with a applied local heat to maintain a high preheat around the back of the pin loactions. So far I havent seen a split in a bead in the last 12 pins.
With how many passes, I am not sure if you are asking about the pins or center hub. The center hub is 3 passes including the root, and the pins are also 3 passes ((6) 5/32 309L electrodes). Just this last weekend to try and irradicate the cracking we tried to do more than just 180 degrees passes on the center hub and I hooked it up to a rotator and then preheated it to about 350, and ran a full 360 degree bead and swapped starting points. So far so good. The engineer for thier company approved everything we are doing. Just waiting on MAG particle today to see if there is post heat treat cracking.
Thank you Jeff and Henry for the input.
Jordan
So it is official-----
Parts passed MT and UT and were accepted. I think we got all the kinks worked out now.
Thank you to all who chimed in with their comments.
Jordan
Jordan,
Glad you worked together to solve your problem.
Put this one in your experience portfolio.
Tyrone
With those two materials it is most likely caused by a mismatch in thermal expansion with a pinch of centerline solidification cracking in the SS. When you weld a fillet circumferentially you are generally putting the fillet face in tension, since it's a mixed alloy weld you are adding that much more residual stress due to the relatively large difference in thermal expansion. Fixes
The easiest would be to modify your preheat temperatures you didn't mention them, but perhaps that large plate could be soaked at temp longer or have it's temperature increased. I would also advocate or at least try overmatching the filler. Depending how tight your weld specs and in-service specs and code is I would experiment with 316L. The overmatched filler may help with interstitials causing centerline cracking in the weld zone. Which also reminds me, cleanliness will be king on this weld, the sulfur and manganese in the A36 alone with slag and oxides will love to head to the center of the stainless steel weld and cause cracking. Obviously there's only so much you can control the A36 spec, but you can do your best to have it cleaned to bright metal in the vicinity of welding.
Wish I could add more on the other one but I'm having a hard time picturing it in my mind. IF the problem only shows up after stress relief anneal it either means it's a residual stress related problem or the time at temperature is growing micro cracks.
I think you are right. We just got all the circular welds down and the cracking undercontrol. We did get a few hairline cracks in straigt fillet welds where thick metal was welded to thinner metal. We are now doing a post weld slow cool down in an oven. Putting them in at 350 and taking them to room temp over the course of 6 hours most likely. Keep you posted.
Jordan
