Fillet weld for beam to beam

[PrashantB]

Hi,
I am looking for the code reference if any or the best practices to define the weld for beam web to web joint (two beams connected at 90 deg.)
Considering this beam to beam joint is primary/ critical and considering flange to flage CJP,for web to web joint my understanding is instead of over weld design to have CJP , both side tothe web, fillet weld should work. However refering to AWS D1.1 my understanding is the minimum fillet speciffied in AWS is good enough to considered for this web to web both side fillet? Is this assumption right? If not should I consider fillet side =75% of web thickness on each side of web(assuming both beam are same type and web thickness are same), For ex. if web thickness is 20mm then fillet size of 15mm each side of the web? or this 75% should be in total, means 35% fillet on each side (same ex. 7.5mm on eachside of web?

My interpritation is that this 75% on each size is too much and over design. oversize of weld will impact on welding and NDE cost, number of runs will increase, increase in risk for weld defects, and also overall weight of structure will increase in total. from weld optimisation point of view wht is the best approch to follow rather simply follow the 75$ thumb rule?
I am not able to get any code reference for this 75% consideration what my superier is asking to follow as a thumbrule. is thjis specified in any code?
I have attached the image for beam to beam joint and also another image for the generic T joint between two plates. my doubt remains the same for plate to plate fillet weld as well. Shoud I consider 75% on each side or total 75% so approx 37% fillet on each side?

Thanks in advance and looking forward to get the valuable suggestons.
PrashantB

 

[KootK]

1) I'm going to assume that we're talking about the welds that connect flanges to webs in built-up beams.

2) In all applications, it is common to design the welds for the real, horizontal shear demand rather than to "develop" the web plate thickness. This is usually VQ/It along the beam and either MQ/I or full flange development at the ends of the beam.

2) In non-fatigue applications, it is common to use stitch welding, following code limits on maximum and minimum fillet weld sizes and stitch lengths.

3) In addition to the issues that you mentioned, excessive welding of these joints tents to great an excess of beam distortion which then needs to be either rectified or lived with.

4) Other than the basic maximums, minimums, and fatigue requirements, I know of no code limitations on this.

 

[SandwichEngine]

This is how PEMB tapered I-shapes are made. The only check you need to do is for shear flow.

Shear flow demands get much smaller the farther you get away from the centroid. For PEMB beams this means they are over-designed in most cases even with a very small one-sided fillet weld.

If you're actually doing a T-section like your image suggests, the welds will be much closer to the centroid and may require double sided welds.

 

[JStephen]

He says he's welding one beam to another beam at right angles, so the weld in question is web of one beam to web of the other, not the web-to-flange welds.

 

[Agent666]

Maybe in your code there is a minimum percentage of the full shear capacity that any connection needs to meet. If your code has one just design for that (it will be lower than the 75% of web thickness value), or the actual load if that is higher.

Edit - - In NZ steel code this minimum is 15% of the full shear capacity, though generally 30% minimum is adopted by industry standard/pre-qualified connections.

https://engineervsheep.com