Necessary to validate the welding strength of every joint in a large steel frame? 2

[Kai Wang]

Hi all,

Could anybody share experience when validating the welding strength of joints in steel frame what is the rule to choose the joint(s) to be validated? I was asked to validate welding joints with diheral angle lower than 30° in our steel frame which is composed of HSSs. I plan to extract the maximal loading of this kind joints in two suites of steel frame from global FEA, then validate the welding strength by hand calculation against the maximal loading. However, the joints of this kind may be more than 40 pieces in each suite.

I think it is tedious to extract loading of so many joints. Is there book, standard or rule which suggest how to choose the joint(s) to be validated? I have basic FEA capability and have learned the basic method to calculate stress in throat of weld which is mentioned in AISC Steel Construction Manual (SCM).

Any suggestion is appreciated. Thank you!

 

[Kai Wang]

The method I plan to use to extract loading of welding joint from global FEA is as follows. Maybe it is inefficient? Could anybody point out if my method is appropriate? Thank you!

(I am using ANSYS Workbench)
1. After solving the global FEA, obtain the following information of all beams

a. the axial force​

b. bending moment at two directions, respectively and​

c. torsional moment​

2. record the above items for each joint

 

[DrZoidberWoop]

I am not sure your interpretation of the issue is correct. Part of requesting you "validate the welding strength of the weld w/ dihedral angle less than 30 degrees" is related to whether you attributed any load to the HSS heel weld. Sometimes the strength of this weld is ignored in connection design because it is not AWS prequalified nor is it guaranteed the fabricator will have the qualified welding personnel or welding procedure specifications (WPS) to complete the weld. I've seen the heel weld <=30 degrees specified as a seal weld, with no strength attributed to it because it's a difficult weld to reliably make in many cases.

I am sure an FEA analysis could achieve whatever you need (I've done similar analysis with ABAQUS), but this problem may be more closely related to industry-standardized analysis and fabrication practices than a true numerical exercise. I'm not an expert in HSS connections, but those are my thoughts.

 

[Kai Wang]

Hi Dr Zoidber Woop,

Thank you for the suggestion on ignoring the weldment at heel for small angle weldment!

Sorry I have not explained very clearly. The point I am not sure is the method to choose the joint(s) in a large steel frame to validate. I think the critical joint should be checked. But I am confused on how to define "critical". I guess one method is to pick several joints, e.g. 3 joint, which suffers maximal axial force, and 3 joints suffered maximal bending moment, and also for shear force as well as torsional moment. But maybe there is a joint D which is not picked from this method, however the synthetic stress from axial, bending, shear, torsion on it is largest. Then it may be missed if I follow the above method which may lead to a potential risk.

I am wondering if there are books/standards or conventional rules to identify the joints to be validated. Or, maybe it mainly relies on the experience of the analyst?

 

[CANPRO]

The "method" you're asking about is experience. Technically, you should check them all. You'll eventually develop a feel for which joints are "critical" and which ones can pass by inspection. You should start with the most heavily loaded joints first - if the most heavily loaded joint passes your design check, then you can use cursory checks on similar joints with less load.

 

[L_K]

You should be able to decide the critical joints from stress distribution plot of the nominal load case(s).

 

[kingnero]

Are those butt welds you need/want to verify?

 

[Kai Wang]

Hi CANPRO,

Thank you for your suggestion! I think your idea is reasonable. So I have to check multiple joints to build my own's experience gradually...

 

[Kai Wang]

Hi L_K,

I also guess joints with the highest Von Mises stress may be the ones needed to be checked. But I do not know if this is a common practice and I worry critical joints may be missed since the tube is modeled as beam in FEA and it seems there is no direct derivation which shows the joint with highest Von Mises stress is just the most critical one.

Could you please share further idea? Thank you!

 

[Kai Wang]

Hi kingnero,

I think two kinds of weldment are involved in my structure, i.e. fillet weldment and flare-groove weldment which occurs at different regions of the joints. Only box cross section is involved in my structure but no circular section. The attached picture is an example which shows the regions of stepped box connection.

I am not specialized on welding. But by checking reference, I think the category of welding I mentioned above should be correct.

 

[kingnero]

I am not sure which codes you use, howwever using eurocodes you can assume that full penetration butt welds have the same "strength" as the adjoining member. This makes it not necessary to calculate those welds (providing the members are correctly dimensioned of course).
Using the same logic, if you use fillet welds with a throat height similar to the wall thickness of the (thinnest) adjoining member, you can assume that the weld is not the weakest link in the structure. Maybe this can help you setting a guideline in your work?