Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

[SEAOC] welded matched box connections in bridge

[Subject Prev][Subject Next][Thread Prev][Thread Next]
I'm looking for a consultant who is intimately familiar with the AWS
provisions for the strength of welded matched box connections.  The issues
are presented below.  If you know someone who fits the bill, please let me
know.  Reply to tnhanc(--nospam--at)alaska.net.

Thanks in advance,

Dave Evans, P.E.
tnhanc(--nospam--at)alaska.net


The sorry story:

I'm checking submitted calculations for an existing pedestrian bridge:  80'
span, 10' wide.  Top and bottom chords TS8x3x3/16, long dim vert.
Diagonals, or branch members, TS3x3x3/16; therefore matched box connections.
Not overlapping, so we have gap K (really N) connection (combination T and
Y).  Sloping diagonals are oriented so they're normally in compression;
theta = 57 degrees.  Verticals are in tension.  Design load is a heavy snow
load.  Also H5 truck.  Remote area, little traffic.  Occasional biker,
skier, hiker.

1.	The designer used 1984 AWS D1.1 (F1 + F2) to check matched box conn
capacity: worst case for sloping diags is 33 k load vs. 48k
allowable...ratio is 0.69.     However, the connection geometry  exceeds
some of the AWS limitations:
	A.	H/tc = 43, 23% greater than 35 upper limit
	B.	H/D = 2.67, 34% greater than 2.0 upper limit

	Q: Practically speaking, are the geometry busts significant?  If so, or if
it is impossible to say, what should I do?

2.	AWS 94 10.5.2.3 requires welds either to be prequalified complete pen or
partial pen, or to be fillet welds...but...there's no prequalified joint for
fillets if beta is greater than 0.8 per limitations in 10.13.3.  (Sides of
branch must fall on face of chord)

	Q:    For beta = 1, matched box, practically speaking, then is the only
acceptable connection  complete pen or partial pen, per prequalified joint
details (1996 AWS D1.1 Fig 3.5 or 3.6)?

	Q:      Suppose the fabricator claims that the branch connection to the
chord is merely a combination of prequalified skewed fillet weld (Fig. 3.11)
and prequalified flare-bevel weld joints (Fig 3.3).  Would that claim be
valid?   If not, why not?

	Q:     For this pedestrian bridge, how important is it that the portions of
the weld at the corners of the branch be welded per the prequalified joint
details?   That is the area that I suspect the fabricator did not place much
emphasis, but the AWS certainly puts a lot of emphasis on the corners.

3.	The shop drawings show 3/16" fillet to face of chord and 3/16" size
(sic) flare-bevel groove weld to sides.  There was no verification
inspection, so we don't know for certain whether the branch ends were
tediously profiled per the prequalified joint detail for partial pen welds. 

	Q:      Is there any non-destructive way of determining whether the ends
were profiled per the prequalified joint detail?   

	Q:       If the branch ends were not profiled per the prequalified joint
detail, or if there is no practical way to verify, how can we determine if
the joints are acceptable?   Can a weld inspector make that judgment based
on visual inspection?

4.	The designer did not check the weld capacity using the effective lengths
of AWS 94 10.8.5.1; but I did, and in the worst case, the load exceeds the
allowable by 3%.   I assumed that the effective throat of the flare bevel
welds was 5/8 * chord wall thickness per AWS and AISC (ignored the erroneous
3/16" size shown on the drawings).  Normally, I'd consider 3%  overstress
acceptable.

	Q:      However, considering that the joint prep is unknown, is there
reason to be concerned about the  weld strength?  If so, what to do?

...