Stress intensity factor at "T" fillet weld 3

[DAS1001]

Hello, new member here. I have been asked to analyze the welded assembly shown in the attached pictures, and determine ultimate load P. The problem is, I have spent my career dealing with aluminum aircraft structures, and have basically zero knowledge about welds.

At the area of interest, the stress in the bracket will be mostly tension. The area where the two parts contact each other is not welded, only the fillets. It seems to me that this is similar to a member, loaded in tension, with an interior crack.

Can anyone give me some wisdom about how to calculate stress intensity factor for this configuration ? The literature I've been able to find online so far does not go into much detail.

Notes:

1. Parts are all low carbon steel. The bracket is AISI 1025, the beam is ASTM A36. I assume the weld rod has similar properties to these.

2. I have calculated stresses at the fillet throat as shown here:

http://www.gowelding.com/calcs/c2.html,

however this does not account for stress concentration at the fillet heel due to the crack tip. I have also analyzed using FEA, both linear static and nonlinear with yielding, and got widely diverging results.

 

[dhengr]

DAS1001:
I can’t offer much ‘wisdom about how to calculate stress intensity factor for this configuration,’ but I might offer some insight about the complexity of that region of that welded joint. Some one in their infinite wisdom decided to use .375” fillets where the welds are in compression and shear, but then only .25” fillets where the welds are in tension and shear. Can you explain that? Those are some pretty small welds on some pretty thick pls., without knowing the value of “P”. If this is not a done deal, I would apply 5-6” wide by 4” high end pls. to your vert. pin pl., and then slope the pin pl. down to the top of these end pls. That way, you can weld up the end pls. and across the beam pl., and protect the tip of the weld in question. The tip of the weld in question is really a nasty condition however you cut it. It is the termination (or starting point) of the weld, always a weak/sensitive spot. The welders tend to leave craters, crack sensitive spots, and the like at starts and stops. There will be some question about fusion at these spots, particularly at the root of the weld. And, the root, in general, is a weak spot, and crack sensitive too, right where you see an internal crack starter. Then, you have the stress concentration and tri-axial stress condition right there too, max. tension due to the moment, plus the distributed shear loading. Thus, you just want to protect these types of weld conditions with something you can actually put some numbers on, to hell with some stress intensity factor which can change with the spark of a welding rod.

If you are going to be doing some welding design, AWS has a set (6-7 vols.) of Welding Handbooks which answer many welding questions, and their welding code is a must-have. Lincoln Electric/Lincoln Arc Welding Foundation has a bunch of very good literature, which is very economically priced too. In particular, a couple books by Omer W. Blodgett; “Design of Weldments” and “Design of Welded Structures.” Also, any of their more current books, papers, etc. by Duane Miller, Blodgett’s protégée.

 

[Tmoose]

What is the value of "P"?

The backstory behind the two different sizes of welds is interesting, I bet.

If fatigue is a concern, weld "returns" are required for the reasons dhengr mentioned plus a few more.

https://app.aws.org/technical/d3/D1.1_2000_Section2_Design.pdf

2.28 Fillet Weld Terminations

 

[DAS1001]

Thanks, guys. You are confirming my initial suspicion that this is not a very good design.

Tmoose, The maximum value for P is what I am trying to calculate. I have typically set P = 1,000 lbs, calculated stresses, then scaled P up until stresses reach allowable values. I don't know the rationale for the different fillet sizes.

dhengr, please forgive my ignorance, but what does "pl" and "pls" mean ? Thanks for the book references, I will check them out.

My planned response is to present the numbers I have calculated so far, but then STRONGLY suggest they change the design to a full penetration weld. I realize I will still have stress concentrations where the weld bead meets the base material, but it's still a much more benign configuration than what I have now.

They also want me to calculate fatigue life. My answer is going to be "Really short".

 

[WinelandV]

Whew, DAS1001, you're really in the middle of this. You should go ahead and pick up a copy of ASME's BTH-1-2014 (or whatever the latest edition is). They have factors to limit capacities based on the expected cycles of loading.

As to your padeye itself, I've never seen one with a notch at the connection interface. Not saying it can't be done, just that it's not common.

 

[SnTMan]

Let's be clear: Lots of things are held together successfully (and economically) using only fillet welds.

Yes, eliminate the notch. Fillet weld 100%, sized appropriately for the load. This technique is in our college textbooks. Mind fatigue if that's a factor.

Regards,

Mike


The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[JGard1985]

Lets be really clear here....this problem is exactly...word for word...the sample problem for a USAF government structural analysis solicitation for engineering services that is due next week....

https://www.fbo.gov/index?s=opportu...1eefd6843ac97f46fa6323a7d89&tab=core&_cview=1

Attachment 5

 

[SnTMan]

Well, can't just eliminate the notch then

The problem with sloppy work is that the supply FAR EXCEEDS the demand