Welding Tube Steel

[cborger]

I am checking a welded tube steel connection with TS 6x6 vertical verendial truss members and TS 6x6 chords. They are using a fillet weld all around and are counting all four sides as effective. I have heard that some codes do not allow the full perimeter (i.e only 2 sides effective) due to the radii of the tube steel corners. Anyone know if this is true and which code says this (AWS possibly?)

Thanks,

Craig

 

[CWIC]

It depends if the members are matched or stepped as noted in AWS D1.1-98, Figure 2.14. If the tubes are matched, then technically what you have is 2 T-joints w/fillet welds and 2 flare bevel groove welds.

 

[cborger]

I believe the tubes are considered matched. The weld is detailed as fillet on 2 sides and flare bevel groove on the other 2 sides. I have been told that only the 2 sides with fillet welds and a small dimension of the other side (2 x the thickness of the material) can be considered effective (i.e. the groove welds are not effective).

The effective weld area in plan view looks like this:
__________
| |

<-- sides with rounded edge (not effective)


|__________|

Craig

 

[CWIC]

In your original post you referenced AWS.

You may want to research the requirements (effective weld size of flare-bevel grooves) in AWS D1.1:2002, 2.3.1.4 and Table 2.1 if you have a structural application and the thickness of your tubing is 1/8&quot; or greater.

 

[RHK]

The best information available on the design of tube/tube joints is the CIDECT Design Guide for Rectangular Hollow Section joints under predominantly static loading (ISBN 3-8249-0089-0). This is based on test results from Japan, Holland et al. If you are in the USA, then the nearest author is Jeffrey Packer of the University of Toronto, Canada.

The design guides give the ultimate strength of the joint. From this you can back-calculate the stress in the side walls, and so calculate the fillet weld size.



Russell Keays

 

[cborger]

Thanks for the replies.