Bolts
Bolt slip in connections
The effect of bolt slip in truss connections is an issue that is raised with SCI from time to time
in various contexts. Richard Henderson discusses some of the issues.
Introduction
The deflection of a truss can be estimated using various analytical methods
and often a stick finite element (FE) package will be used to determine
the member forces and the deflections under the different load cases. The
calculated deflection depends on the assumptions made in the analysis
about the nature of the joints – whether pinned or rigid.
Truss Joint types
In BS EN 1993-1-8, three categories of bolted connection loaded in shear are
identified:
• Category A: bearing connections where the bolts act in shear and
bearing;
Connections made with preloaded bolts:
• Category B: slip-resistant at serviceability limit state;
• Category C: slip-resistant at ultimate limit state.
Connections in category B must also be designed for shear and bearing
in the ultimate limit state and Category C for bearing and net area. Fewer
bolts will be required in Category B connections than in Category C ones.
SCI recommends adopting joints made with preloaded bolts where
members are spliced and deflection is of concern because this allows
the deflection of a truss to be better controlled. Category B joints are
usually sufficient but Category C joints maybe specified in special cases
(eg with oversize or slotted holes). In theory, once the joints are made, the
subsequent deflection of the structure is due only to the elastic deformation
of the members.
Predicting deflections in trusses
As discussed in the introduction, an FE model of a truss will deliver the
deflections of the structure as well as the member forces for a given load
case. The actual deflection of a truss made with Category A bolted joints
may well be greater than the predicted deflection, because the joints may
slip when the load comes onto the structure and the bolts take up their
loaded position. The deflection will be more significant if holes are oversize
or slotted. This effect may be predicted by using virtual work methods
which assume a pin-jointed model and adding an allowance for the slip at
each bolted connection to the extension of the member due to the internal
forces. This can be illustrated by example.
Example 1
Consider a two element pin-jointed bracket connected to rigid supports
as shown in Figure 1. Estimate the total deflection if there is a 2 mm slip in
each bolted connection.
Considering the elements separately for displacements that are
small relative to the lengths of the members, if there is a change in
length in the elements of 2 mm due to bolt slip, the vertical deflection
in millimetres resulting from the extension of the diagonal is 2/sinθ and
2/tanθ from shortening of the horizontal member. The total deflection is
therefore 2 × (1/sinθ + 1/tanθ) = 4.8 mm for θ = 45°.
The same calculation by virtual work is given in Table 1.
26 NSC
Technical Digest 2019
Figure 1: Bracket arrangement
Element Diagonal Strut Total
Area (mm2) A 470 667 (mm)
Length (m) L 4√2 4
Member forces (kN) p100√2 100
1 Member forces due to unit
p√2 1
2 load
Member flexibility (mm/
kN)
L/EA 0.0573 0.0286
Member deformation
(mm)
p1 L/EA 8.1 2.9
Deflection due to member
deformation (mm)
p2 p1 L/EA 11.5 2.9 14.4
Slip (mm) s 2.0 2.0
Deflection due to slip (mm) p2 s 2.8 2.0 4.8
Total deflection: Σp2 (p1L/AE + s) 19.2
Table 1: Bracket deflection
Both methods give the same deflection due to bolt slip.
Example 2
To illustrate the effect of bolt slip consider a pin jointed Pratt truss (N frame),
shown in Figure 2. Member areas are based on a tensile stress of 350 MPa
and 150 MPa in compression, with the area limited to a minimum value.
The deflection of the truss centre under the total design load is
estimated to be 175 mm (span divided by 230), calculated by virtual
work. An FE model gives a deflection of 179 mm. The deflection can be
apportioned to 110 mm of bending deflection (deformation of the truss
booms) and 65 mm of shear deflection, from the bracing members. Making
this distinction is useful if the deflection is to be reduced because the
elements making the greatest contribution to the total deflection can be
identified.
In estimating the effect of bolt slip, it is assumed that with automated
saw and drill lines, the accuracy of holing is such that slip can occur in all
holes simultaneously. If all the members are bolted with 2 mm oversize
holes and 1 mm of slip is assumed at each end of a member, a total of 2 mm
per member, the deflection increases by 43% to about 250 mm. The effect