Welding Standards of Calculation

[JohnnnyBoy]

I'm a recent grad in civil engineering and confused about how exactly to calculate welding capacities. The issue i'm having is in school we learn to use the handbook of Steel construction cisc. How within this code shear is calculated as Vr=0.67*Θw*Aw*Xu(1+0.5sinθ^1.5)*Mw In this case it can be simplified to Vr=0.67*Θw*Aw*Xu assuming shear acting on weld axis and no strength reduction factor. Further more in the code Groove welds are define the Vr=0.67*Θw*Aw*Xu as well.

In speaking with a few mechanical engineers they had a very different way of calculation the strength in which I've attached to this thread. It uses classification of welds from Category A - Category F along with a cycle life and gives a strength to use and multiplied by the effective throat.

I've used fillet weld of 1/4" 5" long for this example. If you were to look at fillet weld in shear diagram 8(a) the weld metal would be considered category F and would have a ultimate strength of 15 ksi (103 Mpa) while using the method above would give me a strength of 220 MPA by Ultimate Stress = 0.67*Xu*Θw = 0.67*0.67*490 = 220 MPA

Sorry if this is a little unorganized but if anyone can help me see where I am going wrong or why there is such a difference in the methods of calculation. Any good document I can read would be very helpful too. Thanks in advance.

 

[Agent666]

You didn't attach the attachment.

Certainly if the mechanical engineering solution includes the effects of fatigue via the categories you will end up with a larger weld to limit the weld stresses to acceptable levels. If you did the same structurally, you will also possibly end up with a larger weld as well.

The structural weld solution doesn't make any allowances for fatigue.

 

[JohnnnyBoy]

Sorry here is the attachment.

The mechanical solution does account for fatigue where as the structural seems to ignore altogether or so it seems. In the equation there is a reduction factor of 0.67 as well as a resistance factor that is derived from reliability analysis of welds. Now when it comes to fatigue I don't believe a building will be fully loaded very often in its lifetime compared to many mechanical situations. Obviously fatigue still needs to be accounted for but it may have been accounting for in the calculation of the factors associated with the structural calculations of weld loading. My big issue I'm having is there seems to be a total lack of changes in strength based on how the weld is loaded in the structural analysis that is included in the mechanical.

 

[sandman21]

What are you trying to do? Design a building? A bridge? What code are you using other than the AWS?

A building will not typically be designed for fatigue, as the cycles on a building rarely have the frequencies needed to be a concern.

 

[Tmoose]

Hi Johnnyboy,

What do you make of Figures 2.8, 2.9, and 2.10 ?

regards

 

[JohnnnyBoy]

I'm using S16-09 now I realize the PDF I posted is AWD although S16-09 does have the same curves and diagrams I just couldn't get the actual pdfs of S16-09.

I agree with your statement that building are not and probably for good reason are not designed for fatigue. In this case I would be designing a weld for a steel column to beam. Very simple design (or should be). Perhaps I'm way over thinking this.

Figures 2.8-2.10 are fatigue curves and diagrams explaining the application of Categories of welding. Now as mentioned above since there is very little cyclical loading on a building (in comparison) then I think we should be using the left most section of the chart although this is also starting at a cycle life of 100,000.

Cycle life of a building is between 20-100 years. Assuming designed loading occurred 5 times a day (most probably less), would be 91,200 for an assumed 50 years.

 

[SectionIX]

Why are you using the 2000 edition of AWS D1.1.

A few changes have been made in the last 16 years...

 

[JohnnnyBoy]

I am using the latest version of S-16-09 not AWS although AWS was the only example of the methodology that I could obtain in PDF.