Distribution of Forces in Weld Groups

1.
2 Distribution of Forces in Weld Groups
The design of weld groups (Figure 8) is tackled by considering the strength of

the individual welds.
For this, two approaches based on the design assumptions of Clause 6.1.4 of
Eurocode 3 can be followed:
1. Calculation based on the stresses in the parent material in the vicinity of

the weld.
2. Calculation based on the load acting on the welded joint as a whole.
The calculation based on the stresses in the parent material is easier and
quicker than the second method. The stresses in the vicinity of the weld can be
directly obtained from the structural design calculations. Because of the direct
link between the analysis for the parent material and that of the welds, it is
clear that the stresses in the weld are consistent with those in the parent
material. Of course, it is necessary that in determining the stresses in the
parent material, the stresses must be consistent with other parts in the
connection (Lecture 11.4.1 Section 2.2 and Figure 2). Use of the first method
is recommended.
However, there are cases where the first method cannot be applied because the
stresses in the adjacent parent material cannot be simply determined and the
second method must be used. Examples are:
 a lap joint.
 the connection of a brace to a gusset plate.
1. Calculation based on the stresses in the parent material
In a double fillet weld shown in Figure 9, the following stresses act on

the throat area (see also Lectures 11.2.2 and 11.2.3).
 =  = {0,5tz 0,52}/a = [t/(2a2)]z (1-4)
// = (t/2a)zy (1-5)
With the alternative method of Annex M of Eurocode 3 it follows:

(t/2a)[z2/2 + 3z2/2 + 3zy2]  fu/(wMw)
or:
(t/2a)[2z2 + 3zy2]  fu/(wMw) (1-6)
The second requirement   fu/Mw is only decisive if  is small, i.e. if

the resultant force is oblique to the plate.
If in the parent material only z is present then it follows:
  (z/fu)(wMw)t/2 (1-7)
For S235:   0,71(z/ fu)t (1-8)
For S355:   0,80(z/ fu)t (1-9)
If the theory of plasticity is used for the design of the structure and the
connection is located at a point where a plastic hinge may develop, then
the minimum throat thickness must be based on z = fy:
This gives for S235: a  0,46 t (1-10)
and for S355: a  0,55 t (1-11)
This requirement also applies for statically indeterminate structures that

are designed using the theory of elasticity. It is important to remember
that also in an elastic design, it is assumed (implicitly) that the
members and the connections have sufficient deformation capacity to
accommodate loads and stresses that usually are not explicitly taken
into account in the design calculations (e.g. stresses due to uneven
settlements of the supports; temperature loading; tolerance during
fabrication; local overloading by live loads, etc.) and further to allow
for the approximations inherent in the design models.
Therefore, it is necessary that the connected parts can yield before

rupture of the welds.
When the design formulae of Eurocode 3 [1] are applied with z = fy,
then the real rupturing strength of the weld is at least the real rupturing
strength of the plate. In other words, actual rupture occurs in the plate
and not in the weld.
Thus, for the above requirement (yield in the plate before rupture in the
weld), the design of the weld can be based on:
z = (fyr/fur)fy (1-12)
where fyr is the measured yield strength and fur the measured ultimate
strength of the plate material.
For the design values of the yield strength and the ultimate strength, it
follows for S355:
z = (1-13)
Because the actual value of fyr /fur can be higher than 0,70, it is required
that:
If deformation capacity is necessary, the welds must be designed to

transfer at least 80% of the yield force in the (weakest) connected
member.
This requirement gives the following values for the minimum throat
thickness of a double fillet as presented in Figure 9:
For S235: a  0,37 t (1-14)
For S355: a  0,44 t (1-15)
It should be noted that, using the mean stress method according to

Eurocode 3, Chapter 6, greater throat thicknesses are found for end
fillet welds. The difference is a factor of 1,22!
Thus the application of the mean stress method results in 1,222=1,5

times more weld metal than necessary.
2. Calculation based on the load acting on the welded joint as a whole
This method must be applied if the first method is not applicable. For the
determination of the strength of a weld group, the design values for the
strength of the separate welds may be added, provided that the equilibrium
requirements are fulfilled.
This approach is based on the assumption that the welds can yield to permit
the redistribution of stresses necessary to accommodate local overloading. In
other words, the welds must posses sufficient deformation capacity.
To gain some idea of the deformation capacity of welds under various loading
combinations, tests [3] have been carried out as indicated in Figure 10. In
these tests, the welds were thin compared with the plates in order to ensure
yielding in the welds and not in the plates. The measuring length lo is given in
Figure 10. The deformation of the plates was subtracted from the measured
values, to obtain the deformation of the weld and the parent material in the
direct vicinity of the weld, see Figure 4a.
As already discussed, it appears that the deformation (mm) at the same stress
in the weld is proportional to the throat thickness. Thus, when the thickness of
a weld is doubled, not only is its strength doubled, but also its deformation
capacity. This is the reason why the deformations are given as l/a on the
horizontal axis in Figure 11.
To demonstrate the importance of the above findings, the lap joint of Figure
12 is analyzed. Suppose that the plates are infinitely stiff compared with the
stiffness of the welds. When the thicknesses of the end fillet weld and the side
fillet welds are about the same, then, at the start of rupture (in the welds), the
forces in all welds are practically equal to their ultimate load. This can easily
be seen when the lines for 11 and   are compared. With aside = aend and l
for the side weld and end weld about the same, l/a is the same for side weld
and end weld. Therefore, the ultimate strength of both welds may be added.
This may not be true if one of the welds is very small, compared to the other.
It can be concluded therefore, that the ultimate strength of the lap joint is
equal to the sum of the ultimate strengths of the separate welds.
Now suppose that the throat thickness aend of the end fillet weld is only 10%
of the throat thickness aside of the side fillet welds. At the start of rupture of the
end fillet weld, the elongation l=100.10-3. aend=10.10-3 aside, see Figure 11.
The stress 11 corresponding to l/aside = 10 . 10-3 is about 230N/mm2, whilst

the rupture strength for 11 is about 350N/mm2.
In this case, the ultimate strength of the lap joint is less than the sum of the
ultimate strengths of the separate welds. As a result, it is recommended the
following design rule is used:
Design recommendation
Try to give the end fillet weld and the side fillet weld the same thickness, and
never design the end fillet weld to be less than 0,5 times the thickness of the
side fillet welds.
The use of a thin weld at the front of the lap joint (point A in Figure 12), e.g.
to prevent corrosion, must be avoided. If such a weld is necessary, then it
should be given the same thickness as the other welds. This is particularly
important because the plates are in reality not infinitely stiff compared with
the welds. The required deformation capacity therefore is larger at the front
(point A) than at the back of the lap joint (point B).

Distribution of Forces in Weld Groups

Uploaded by

Copyright:

Available Formats

Distribution of Forces in Weld Groups

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Distribution of Forces in Weld Groups

Uploaded by

Copyright:

Available Formats

How are stresses in welds calculated based on stresses in the parent material?

How are stresses in welds calculated based on stresses in the parent material?

What factors affect the deformation capacity of welds?

What factors affect the deformation capacity of welds?

1.

2 Distribution of Forces in Weld Groups

The design of weld groups (Figure 8) is tackled by considering the strength of

1. Calculation based on the stresses in the parent material in the vicinity of

1. Calculation based on the stresses in the parent material

In a double fillet weld shown in Figure 9, the following stresses act on

// = (t/2a)zy (1-5)

With the alternative method of Annex M of Eurocode 3 it follows:

(t/2a)[2z2 + 3zy2]  fu/(wMw) (1-6)

The second requirement   fu/Mw is only decisive if  is small, i.e. if

If in the parent material only z is present then it follows:

For S235:   0,71(z/ fu)t (1-8)

For S355:   0,80(z/ fu)t (1-9)

This gives for S235: a  0,46 t (1-10)

and for S355: a  0,55 t (1-11)

This requirement also applies for statically indeterminate structures that

Therefore, it is necessary that the connected parts can yield before

If deformation capacity is necessary, the welds must be designed to

For S235: a  0,37 t (1-14)

For S355: a  0,44 t (1-15)

It should be noted that, using the mean stress method according to

Thus the application of the mean stress method results in 1,222=1,5

2. Calculation based on the load acting on the welded joint as a whole

The stress 11 corresponding to l/aside = 10 . 10-3 is about 230N/mm2, whilst

You might also like