final-report-of-the-advisory-committee-on-falsework-bragg-report - Flipbook - Page 61
One suggestion we received was that in fixing the
location of a strut an error of not more than 30 mm
in any direction is admissible provided this does not
result in much greater eccentricity in load application.
Similarly we would suggest that the offset from the
vertical of any unsupported length-of strut should not
be more than 1/5th its width in that direction, e.g,
10 mm per lift height between lacing levels of ordinary
scaffold tube. In circumstances where there are several
such lifts the cumulative offset should not be more
than three times this value at any point, and checks
that the erectors have achieved this accuracy should
be made. It is, however, for the British Standards
Institution Committee rather than ourselves to recommend exact details of allowable tolerances.
Allied to a lack of specified tolerances is the uncertain
eccentricity of the load applied to struts, on which
we have already commented. Eccentricities of load
on a strut may be due to misplacement, misalignment
or lack of verticality: or to lack of suitable packing
or wedging at either end. They may be avoided by
good workmanship, which is ensured by proper
training. For example, concentric loading on a forkhead can be achieved by turning the forkhead so that
its sides impinge upon opposite sides of an undersized
timber by using properly matched taper wedges driven
in and nailed to the bearer to ensure that the central
position is maintained. Known eccentricities must be
allowed for in calculating bending, thrust or tension
stresses. But in every case the designer should also
allow in his calculation for eccentricities and nonaxial effects up to the limit of the tolerances he
specifies. Every effort should be made to ensure that
eccentricities are as small as practicable.
Factors of safety
In the classic method of elastic design of structures
sufficient members are put in to ensure that the
stresses - calculated on the assumption of elastic
behaviour - nowhere exceed the material strength
divided by a factor of safety. This factor is often
taken as about 1 ยท65, and gives a safe working load
which is about two-thirds of that which would
cause failure.
The factor of safety covers a number of unknowns material variability, load variability, errors in construction and so on. More sophisticated analyses,
such as the limit state design, have suggested that
these effects should be allowed for separately. Thus
a characteristic strength, which is possessed by 95 %
of the elements of a given type, is determined and
divided by a partial safety factor to obtain a working
strength. Similarly a characteristic load, which is
62
unlikely to be exceeded, is determined and multiplied
by a partial safety factor to obtain a design load. A
further ratio between strength and load can then be
introduced to allow for particular circumstances - for
example the seriousness of failure.
We believe that our knowledge of the variabilities
encountered in temporary works does not yet make it
possible to apply the individual factors required in the
limit state design. Indeed the use of such a detailed
analysis might introduce a feeling of confidence in
the figures that was quite unjustified.
We therefore recommend that temporary structures
should be designed to accommodate all calculable
loads with a minimum safety factor of two, and that
the calculations take full account of the tolerances
adopted. Consideration should be given to increasing
the factor of safety towards three if any of the following circumstances apply:
Design If the design itself or the equipment used
has novel features. If there are special uncertainties
about the loads to be encountered in practice. If
some members of the design staff lack training and
experience. Where tolerances are not explicit.
Where the structure does not permit accurate satisfactory calculation.
Construction If novel methods of erection are to be
used. If novel materials are to be used.
Situation If working conditions are poor. If the
consequences of failure, either as regards injury or
damage, are serious.
Use of an overall factor greater than two should
produce a "fail-safe" structure in which there is a
duplicate path for every load. Having produced such
a design it is instructive for the designer to check the
effect on the rest of the structure of severe distortion
or failure of any one member. Such an exercise is
valuable in drawing attention to critical areas and to
possible weaknesses which had not previously been
appreciated.
It is important to remember that safety factors are
introduced to deal with the unknowns. They are not
supposed to cover variables such as wind loads or
vehicle loads which should have been foreseen and
allowed for in the original design brief.
It is also important that suppliers of proprietary
materials should recognise that the existence of safety
factors does not give them a licence to neglect quality
control. It is not sufficient merely to publish nominal
or average ratings. Failure loads under particular
conditions of test should be given. We recommended
in an earlier section that suppliers state both their
safe working loads and their failure loads.
