Technical

In-plane stability
of portal frames
David Brown of the Steel Construction Institute offers a reminder of the
guidance covering in-plane stability of this very common form of construction.
Judging by recent questions received by the SCI, the topic is not as clearly
understood as it should be!

The problem(s) identified design, backed up by many decades of successful application, so

In March and April of 2020, to offer some light relief during one might expect definitive guidance in our previous standard.
COVID-19 lockdown, the SCI ran a number of free webinars Portal frames are one example of a continuous frame, and
covering the analysis, member verification and detailing of portal may be designed elastically or plastically, so we need to look
frames. The webinars were significantly over-subscribed, and carefully at the relevant clauses.
were repeated four times. Two topics gave rise to the most Within the “Continuous structures” section, clause 5.2.3.1
questions – in-plane stability and restraints to the inside flange. discusses plastic analysis. The second paragraph should be
Many delegates wanted to know what in-plane effective length sufficient to clarify the in-plane verifications needed:
should be used when verifying members – particularly the
column. Others wanted to apply Annex E of BS 5950 to determine The in-plane stability of the members in a continuous frame
an in-plane effective length. There was no problem with out-of- designed using plastic analysis should be established by
plane lengths – no-one questioned that out-of-plane, members checking the in-plane stability of the frame itself, see 5.5.4.
should be verified between restrained positions.
Perhaps the problem is highlighted if designers are using Designers should note that according to this clause in-plane
general elastic analysis software to determine the design forces checks of individual members are not required.
and moments around the frame and then to verify the members Portal frames are addressed in section 5.5. Clause 5.5.2 covers
within it. Such software expects to complete both in-plane elastic design:
checks and out-of-plane checks, which naturally demands an in-
plane buckling length. Portal frames are a special case, with If elastic global analysis is used for a portal frame, the cross-
particular rules discussed in this article. section capacity should be checked… and the out-of-plane
buckling resistance should be checked….
What does BS 5950 say? For portal frames with no in-plane bracing…the in-plane
Before opening the Eurocode, it is valuable to look at the stability of the frame should be verified by checking the cross-
particular rules for portal frames given in BS 5950. The UK would section capacity and the out-of-plane buckling resistance of the
claim to have developed most of the rules for portal frame members (amplified if necessary)

Nationwide delivery of all Structural Steel Sections

RAINHAM
Phone: 01708 522311 Fax: 01708 559024
MULTI PRODUCTS ARRIVE ON ONE VEHICLE

NSC
26 Jun 20
 Technical

Plastic design is covered in clause 5.5.3:

Plastic global analysis may be used for a portal frame provided

that the conditions in 5.2.3 are satisfied (which is a reference
back to the clause previously quoted).

Checking the in-plane buckling of individual members in a Figure 1; Multi-span portal with internal column
portal frame is inappropriate – the frame buckles as a single
entity, and therefore the standard demands that stability is
verified by checking “the in-plane stability of the frame itself”.

Multi-span frames
One potential exception to the preceding general rule is an
internal column in a multi-span frame (Figure 1). In the so-called
gravity combination, the bending moment in the internal column
may be very small. The in-plane buckling of this member should
be checked. P292 recommends an effective length factor of 1.0
for truly pinned bases, 0.85 for nominally pinned bases and 0.7
for nominally fixed bases.
Internal columns probably have no restraint at any level Figure 2; Orientation of internal column
below the haunch. If the internal column is orientated in the
orthodox direction (major axis in the plane of the frame) then the considered. Satisfying the sway-check method means that
minor axis resistance will of course be critical, not the in-plane second-order effects are small enough to be ignored. The
buckling. If the internal column was turned 90°, such that its amplified moments method allows for second-order effects with
weak axis was in the plane of the frame (Figure 2), or if the an amplifier unless the effects are small enough to be ignored.
internal column was a fabricated section with a larger inertia out Second-order analysis will always allow for those effects.
of the plane of the frame, then in-plane buckling could be critical,
but it seems most unlikely. Member checks in BS 5950
Having completed the in-plane buckling checks of the frame in
In-plane buckling of the frame its entirety and allowing for second-order effects if necessary, the
According to BS 5950, in-plane stability of portal frames can be cross section has to be checked and then out-of-plane checks
verified by three methods: completed. BS 5950 has a range of clauses covering different
conditions – next to plastic hinges, with intermediate restraints
1. The sway-check method – commonly known as the h/1000 to the tension flange only, tapered sections etc.
check, with a limited scope (and a snap-through check for
multi-span frames); Why not Annex E?
2. The amplified moments method, requiring the The introduction to the Annex seems to offer opportunities for
determination of λcr and an amplifier if necessary. No use, describing “the effective length LE for in-plane buckling of a
amplifier is required if λcr > 10; column or other compression member in a continuous structure
3. Second-order analysis. with moment resisting joints should be determined using the
methods given in this annex.” That sounds appropriate for
In each method, the impact of second-order effects is portals, but as one reads further, it becomes abundantly clear 28

GRADES S355JR/J0/J2 Beams • Columns

STEEL
Channel • Angle
Flats • Uni Flats
Saw Cutting
Shot Blasting
Painting • Drilling
Head Office: 01708 522311 Fax: 01708 559024 Bury Office: 01617 962889 Fax: 01617 962921
Hot & Cold Structural
email: [email protected] www.rainhamsteel.co.uk Hollow Sections
Full range of advanced steel sections available ex-stock

NSC
Jun 20 27
 Technical

27 that this annex is limited to columns in rectilinear multi-storey in Annex BB, which is simply the guidance from BS 5950
frames. The annex describes columns in multi-storey beam-and- ‘translated’ into Eurocode nomenclature.
column framed buildings with …. concrete or composite floor
and roof slabs. Hardly the description of a portal frame! Member verification in section 6.3 of BS EN 1993-1-1
If (and only if ) the interaction factors in expressions 6.61 and 6.62
Eurocode rules are taken from Annex B of the Eurocode (very strongly
One would not expect the fundamental physics to change simply recommended by SCI), it can be concluded that expression 6.61
because the Eurocode was introduced. On that basis alone, one deals with in-plane effects and expression 6.62 deals with out-of-
should be confident that the same rules apply to orthodox portal plane effects. Since we have concluded that no in-plane member
frames – that in-plane, the stability of the entire frame as one unit checks are needed (other than the possible internal columns
is critical, followed by checks of the cross section and only out-of- mentioned earlier), we can dispense with expression 6.61
plane buckling checks. altogether.
The key clause is 5.2.2(7)a in BS EN 1993-1-1: As there is no minor axis moment in a portal frame,
expression 6.62 reduces to a rather simpler form:
If second order effects in individual members and relevant NEd M
member imperfections are totally accounted for in the global + kzy y,Ed
Nb,z,Rd Mb,Rd
analysis of the structure, no individual stability check for the
members according to 6.3 is necessary. The numerators are the design force and major axis moment.
The denominators are the minor axis flexural resistance and the
In-plane second order effects are allowed for by determining lateral torsional buckling resistance, which with some judicious
αcr (directly equivalent to λcr in BS 5950), and using an amplifier in interpolation can generally be obtained from look-up tables if
the global analysis if necessary. Frame imperfections are allowed required. In all cases, the lateral torsional buckling resistance
for by always including the equivalent horizontal forces (EHF) in depends on the shape of the bending moment diagram over the
every combination. The only in-plane effects that are not length being considered, reflected in the value of the factor C1.
included in the global analysis are the individual member Resources are readily available to determine the C1 factor for
imperfections, such as an initial lack of straightness. To consider different shapes of bending moment diagram. The interaction
the impact of in-plane member imperfections, colleagues at the factor kzy is painful to compute, but in portal frames is generally
SCI spent (very) many hours analysing a wide range of frames around 0.97 – there is not much loss in manual calculations if kzy
with and without in-plane member imperfections. Imperfections is assumed to be 1.0.
were modelled in both directions, in each member, to produce
the most onerous effect. The study concluded that the value of αcr Conclusions
changed less than 0.3%. Two conclusions can be made. Firstly Portal frames are special in many ways, despite their frequent use
that the effect of in-plane member imperfections on the stability in the UK. They are slender, have significant axial forces in the
of the frame is small enough to be ignored – or presented members, generally are sensitive to second-order effects,
another way, we can say that all relevant in-plane effects have experience reversing bending moments and demand very careful
been allowed for in the global analysis. We therefore do not need restraints to otherwise unrestrained flanges. The objective of this
an in-plane stability check of individual members. The second article was to confirm one special design feature – that in-plane
conclusion is that as expected, BS 5950 was correct – “The in- buckling is an concern for the frame as a whole, not for individual
plane stability of the members in a continuous frame …. should be members.
established by checking the in-plane stability of the frame itself”
The global analysis has not verified the out-of-plane 1 King, C, M.
resistance – members still must be verified between restraints, In-plane stability of portal frames to BS 5950-1-2000 (P292)
using section 6.3 of the Eurocode, aided perhaps by the guidance SCI, 2001

Search for Advisory Desk

articles on newsteelconstruction.com
Use the search bar at the top of every page of
newsteelconstruction.com to search out Advisory Desk articles
by name, number or subject, or list them (most recent first) by
hovering over Technical in the main menu and selecting Advisory
Desk from the resulting pop-up menu.

NSC
28 Jun 20