Alloy 825

[UKCats]

Where can I get some S-N data (allowable stress vs. number of cycles) for Alloy 825 material??? I have a fatigue failure and need to recommend replacment options.

 

[Metalguy]

Annealed material or cold-worked? Any corrosion-fatigue involved?

 

[UKCats]

What I have is Incoloy 825, which comes in an annealed condition. There is no sign of corrosion around the failure.

One problem I see is the support lugs are SA-285-C and they are welded directly to the pressure retaining shell. I would have at least put an Incoloy 825 pad between the CS and 825. My purposed fix is to rebuild it using 825 support lugs, to remove the dis-similar metals at a stress concentration point.

 

[unclesyd]

Where was the failure, in the weld metal, in the C/S., or
in the A286. Where did the failure originate, frontside, backside of the failed piece.

If you don't have welding problems I'd look at Differental Thermal Fatigue. Does the vessel cycle a lot? What is the heat input rate for the vessel if it cycles.

The ASM Handbook has some very good informaton on DTF.

I think you are on the right track, though with a little more information I might be able to give you more insite on this problem.

 

[TVP]

I looked through all of the sources that I have (ASM HANDBOOKS, MIL-HDBK-5, etc.) and could not find any fatigue data for Alloy 825. I would consult with one of the producers of this alloy for fatigue data:

Special Metals (Incoloy 825)

Krupp VDM (Nicrofer® 4221 - alloy 825)

Allvac (Nickelvac 825)

 

[UKCats]

The support lugs are attached by fillet welds. The crack is in the HAZ on the vessel wall. The crack goes all the way thru the vessel (i.e. the vessel is leaking right now).

I've look in ASM also and could not find any S-N data for A825. I think when I build a new vessel, I should make the lugs A825 using S-N data. Question... doesn't ripples in welds cause stress intensification??? I think grinding the welds smooth will reduce this and provide some benefit.

We inject 30# steam (266F) and the vessel operates as a once thru process...meaning, it's not batch. Fluid comes in the bottom, it's agitated, and 30# steam is injected by two nozzles 1/3rd of the way up. The steam nozzles are turned at 90-degrees to assist the agitator in swirling the fluid. There is some temperature differential below the steam nozzles vs. above the steam nozzles. The fluid is moderatly heated from 120F to 150F by the 30# steam.

 

[unclesyd]

Are the cracks in the base metal proper (tank) or are do they appear to originate in the toe of the filet weld.

What is the thermal history of tank?

If it is in the base metal I will have to change direction. You may have a base metal problem.

 

[UKCats]

The crack is in the toe of the filet weld. I think I should eliminate the carbon steel and grind the finished weld smooth. I don't know if PWHT will help????

 

[UKCats]

One more thing... the internal service has been ceramic shotpeened.

 

[unclesyd]

Is the "UK" United Kingdom or the University of Kentucky? I’ve loaned my code books out, maybe someone can help out with information as to what the code allows on 825.
I know we had to jump thru hoops on some 800 pipe.

Being at the toe of the weld puts this closer to being Differential Thermal Fatigue (DTF).
Are the support lugs insulated? We have found DTF in service with as little as 55̊ F temperature. This was the in mixing of two streams and it occurred just by injecting a hot stream (125̊F) into a larger stream at 70̊ F . Even though the velocity was very high we had DTF for about 2' from the injection point. They were no stresses other than hoop stress operating at 150 psig (8" Sch 40 S/S pipe). I think the introduction of the steam maybe a big factor. Take a look a see if you can figure out the temperature profiles around the supports. If you could do an internal the internal surface you might see a multitude of cracks. This is not a given with all DTF, I’ve seen both single and multiple cracks. The single cracking was on the in the lower temperature ranges. I’ve seen this DTF several times a low temperatures, quite a bit lower than most literature mentions.
If it’s from mixing it will originate on the inside. If its from the temperature differential at the supports it would have started on the at the toe of the filet weld.
All the thing you mention are viable and valuable considerations into account. We try to make any transition from support lug to vessel as smooth as possible (try for a 3:1 radius on the filet weld), not only for the stress relief but enabling future external inspections. Changing the lugs out is good idea. Just remember if the problem is a heat sink it will help to insulate the lugs.
I’m starting to bet on internal origin. The origin of a fatigue crack normally can be determined by which side is longer. I’ll gets some response to this.
I kinda rambled.
Hope this helps, but I do have some other suggestions of thing to do that might alleviate your problem

 

[UKCats]

UK is for Univ of Kentucky...and to add to my misery, I'm a bigger football fan than basketball.

Are you saying if the cracks are longer on the inside, DTF could be the problem... and

Are you saying if we have DTF, we should insulate the lugs??? Usually you don't see lugs insulated, do you???... please advise.

Where can I find more info of DTF???

I was going to feather out the filet welds after fabrication to remove stress concentrations at the toe. I was also going to caution the fabricator about undercutting in this area.

Although DTF could play a roll, the biggest stress of the vessel is the 30# steam. We steam sparging and the vessel makes a lot of noise. Attached piping, etc. shakes, so I know these high frequency loads are being transferred to the vessel...and it's anchor points (i.e. We put the vessel on springs 2 years ago, but the spings broke.) I looking for ideas if there's anything thing I can do about that, besides make the shell/lugs thicker.

Thanks for your input...

 

[unclesyd]

If your problem is DTF none of the operating stresses are going to be the primary cause of the cracking. The operating stress may contribute some, though it will be only a very small portion of the stress that initiates the cracking. The operating stress and any stresses introduced by a physical parameter, such as a cold support lug induced stress will become additive as the crack progresses. They operating stresses are the not the driving force initially. The stresses in the surface of the metal where DTF starts, say from a mixing problem that is causing localized heating and cooling, are going to be extremely high and the crack is going to initiate there.. This is why you can have multiple cracking on the side where the surface temperature differential is occurring with only one or two wall penetrations. You can visualize this as water dripping on a hot pipe, which happens and causes DTF. As I mention before we saw DTF in places where mixing went wrong.

Let me withdraw the crack length statement. It’s not general enough to be applicable in your case.

We insulted all support lugs on S/S vessels under going cyclic stresses. If we think that there is a difference in thermal expansion or heat flux is going to be a problem in a vessel we essentially insulate every thing to prevent any heat sinks. I should have noted in the previous post that this is usually refereed to as Thermal Fatigue. We see this were we have a heat sink that causes different heating rates in the components. We still have cracking around the welds, both on the shell side heal and toe of the filet weld. We monitor the cracking by periodic inspection both external and internal. Dye penetrant and ultrasonic. We have made some design changes also on the worst offenders. As you stated we don’t allow C/S near a S/S vessel. We do use C/S when we have an integral pad or boss.

Both of these cases are driven by temperature (cycle). I maybe amiss in calling both DTF. But if you can get ride of the temperature differential (cycle) the cracking will go away. Again you are on the right track in CID of the support lugs. I still think an internal inspection or UT of the shell at the support lug is advisable.

You can information in the Metals Handbook (Failure Analysis and Prevention). They primarily concerned with the mixing connection.

I understand the ASME Code is now addressing thermal fatigue. I don’t have any newer code books available so I have to defer this information to someone else.

Again I rambled.

 

[unclesyd]

UKCats,
I posted my last response without completely reading your posts. !!!This was bad on my part.!!!!
I have a excuse. I was on my way to see a friend that had just reworked the bottom of his old wooden boat with electro-galvanized screws instead of stainless.

What is the source of your vibration? If it’s the steam spargers I may have another problem to work on. We had some fairly large solution tanks that were built with 3 spargers per tank to heat and also help agitate the solution. We also had similar problems with the supports when they raised the rates. Let me refresh my memory,that means look at my notes, as there were 3 tanks and 3 different approaches to resolving the problem. I know there were internal nozzles and inline heaters involved.

On the my DTF posts the information is still maybe applicable even when you get the vibration problem resolved. Get more information on DTF/Thermal fatigue as it will be very useful in the future.

Let the Forum know where you think the vibration is originating if it’s not the steam.

Good luck!
Tidbit: I was in the service (1957) with a UK Basketball fan that could give one the names and statistics of all the players and teams back to the thirties. We even checked his spell, it was correct.

 

[UKCats]

What is CID???

 

[unclesyd]

CID is "Change In Design". Probably dated myself.
It was a package put togather by the engineering group any time you changed the design or altered any piece of equipment physically. It stopped with the new as-built drawings.

Still looking for the steam sparger/vibration issue.