Combined stress for fillet weld 3

[EngrRC]

Hi all,

I am referring to the combined stress equation for fillet welds as per BS5950 which reads:
(F_L/P_L)^2+(F_T+P_T)^2 < 1

where F_L/P_L = utilization in the longitudinal direction and
F_T/P_T = utilization in the transverse direction

Shouldn't there be a square root in the formula?

 

[Lomarandil]

I don't know anything about BS5950 in particular, but that's a common form of stress interaction formula (with the exponent varying, usually 1.67 or 2).

----
The name is a long story -- just call me Lo.

 

[EngrRC]

Hi Lomarandil,

Isn't the usual formula square root of the sum of squares?

 

[Agent666]

That eqn is for use only when you are using the alternative directional method, refer to the standard clause 6.8.7.3. Using this method is not the same as taking the vectorial sum of the forces acting on the weld (like described in the previous 6.8.7.2 clause {simplified method}, which is what you are probably thinking of/describing). Note that F_L & F_P are not necessarily the applied forces, they are the forces on the critical plane of the weld as I read it (not being a BS5950 user).

 

[EngrRC]

Hi Agent 666,
So you mean that the directional method formula is correct? If so, the simplified method/vectorial sum appears to be the SRSS resultant I was talking about.

 

[Agent666]

Yes, I believe so.

 

[JStephen]

If you say a quantity is less than one or the square root of the quantity is less than one, it's the same thing, is it not?

 

[BridgeSmith]

Bingo JStephen! Took me a second to realize it, but of of course you're right.

 

[EngrRC]

Jstephen, yes you are right. However, we would not want an interaction ratio almost equal to 1. For example, based on the interaction formula above, the interaction ratio is 0.9 and you might think you have a comfortable level of safety. But if you take the square root, it's already at 0.95! 5% does not sound much but to me it's a big deal.

 

[BridgeSmith]

The limit is 1.0; not 0.9 or 0.95, so it doesn't really matter how much less than 1 it is.

 

[MIStructE_IRE]

I’m curious how many of you guys are still using BS5950. I wish I could as it made way more sense than this bloody Eurocode!!

 

[CANPRO]

Agree with HotRod10. If you’re comfortable with your design procedure, it checks out with the governing code, and you’re at 0.99 utilization, then give yourself a pat on the back for an efficient design and move on to the next problem.