Professional Documents
Culture Documents
Subject Second Order Calculations Date 13 Aug 2018 Page 1/11 Sap nr. 6000167704
Doc. nr. 050200003 Region Europe Rev. 00
This best practice gives a guideline when to apply a second order (SCIA) calculation and how to determine
the initial imperfection to be used for the strength calculation of (slender) Mammoet structures.
Within Mammoet a large variety of (support) structures is used. Quite often these structures consist of bulky
members but also tall and slender systems like a high lifting gantry or a long spreader beam are used.
Note:
For this Best Practice it is assumed the reader has sufficient background knowledge in the field of structural
calculation and the use of FEA models to properly understand the technical content of this Best Practice. As
such all explanations are kept very brief. If things are unclear please consult your teamleader or technical
advisor.
Content
On the next page a flowchart is given to select the method of analysis for the structure, followed by
explanations and methods to determine the initial imperfection.
NOTE
THIS BEST PRACTICE GUIDELINE IS PREPARED BASED ON THE BEST KNOWLEDGE AT THE TIME OF
WRITING AND SUBJECTED TO NEW INSIGHTS. FOR ALTERATIONS CONTACT THE OWNER.
Structure (1)
No No
Yes
αcr > 10 (9)
Calculation handbook
Excel 3.4ª or
EN 1993-1-1 (6)
Note:
Due to the very large possible variety of structures, above flowchart shall be considered being a guideline
only. In case of doubt on how to analyse the structure for buckling, please discuss with teamleader or
technical advisor.
(1) Structure
Even if a (SCIA) computer model of the structure is not yet available, some simple questions can be
answered to get a first idea if the structure is critical and needs checking for (global) buckling:
Q: Are these compression members very long and thin (= slender)? (yes = unfavourable)
Q: Are braces used to stabilize the structure for side loads and deformation? (no = unfavourable)
Note:
Long I/H-beams and narrow box girders (H/B > 3.0) loaded in bending typically are sensitive for lateral
torsional buckling (LTB). While not being the main interest of this best practice, LTB needs to be taken into
account in the strength assessment of the structure.
Figure 3: Sway buckling of an MSG gantry Figure 4: Sway buckling of a portal frame
For structures sensitive to sway buckling in most cases the exact buckling length of the compression
members is difficult to determine by manual calculations -> Use SCIA model.
If a structure is not fully braced, local sway effects can occur which influence the structure bucking strength:
Taking such effects properly into account in a manual calculation is difficult -> Use SCIA model.
Refer figure 6, due to the imperfection δ and the column loadings P the force in the brace is much more than
would be expected by taking only the effect of the horizontal load ‘H’ into account -> Use SCIA model.
(4) Is the compression member influenced from significant bending due to deadload?
If a member is already bended severely due to the deadload, the buckling capacity of the member is
reduced. An example is given below for a long spreader beam:
As a practical approach calculate the (ULS) bending stress in the member from deadload. If this stress is
below 20% of Fy the Mammoet Excel sheet 3.4A can be used to assess the buckling, but the bending needs
to be included in the unity check, using EN 1993-1-1 part 6.3.3:
For most cases (no lateral torsional buckling, class 3 section or better) this complex equation can be
simplified to:
N ed M
k
N Rk M Rk
The value of the interaction factor ‘k’ shall be calculated using EN 1993-1-1 annex A or B. As an alternative
the ‘steel check’ from SCIA can be used to check this buckling + bending interaction.
If the bending stress due to deadload exceeds 20% of Fy a second order calculation shall be used.
Use a manual calculation to calculate the internal (ULS) forces in the members. Second order effects do not
have to be taken into account.
For standard RHS, CHS and H-profiles the Mammoet Excel sheet 3.4A from the calculation handbook can be
used. For other types of cross section EN 1993-1-1 chapter 6.3 can be used. The increase of the unity check
due to additional bending needs to be taken into account, refer (4).
If many different loads and load combinations are applicable, manual calculation becomes very time
consuming, In such cases a making a SCIA model generally will take less total time.
In some specific cases EN 1993-1-1 part 5.2.2 (5)B can be used. In this method the moments / deflections
are increased using:
For a proper use of this specific method please consult your technical advisor.
A ‘stability’ analysis is performed in SCIA using ULS load combinations. As per EN 1993-1-1 part 5.2.1 (3)
only when αcr > 10 a first order analysis is allowed. For lower αcr values use a second order calculation and
include an initial imperfection.
The second order calculation with initial imperfection is only used to take the global buckling of the complete
structure properly into account. Individual member buckling typically is assessed using the SCIA ‘steel check’
which verifies the individual members on buckling following EN 1993-1-1 chapter 6.3.
For example a load spreader with large influence from bending due to deadload.
pinned-pinned cantilever
Refer EN 1993-1-1 table 5.1:
free
L L
fixed
For cantilever buckling, the imperfection shall be based on 2x the member length.
For critical cases, when required, a more accurate calculation of the imperfection can be applied. Refer EN
1993-1-1 part 5.3.2 (11):
e0 0.2
M Rk
N Rk
It is noted using this value of imperfection, for the case of axial force only, results in exactly the same
buckling unity checks as calculated with the buckling equations as given in EN 1993-1-1 part 6.3.1.2.
In a continuous overall truss / build up member there is no significant play in the connection of the sections.
The connections can be either bolted or welded. For such structure the imperfection 1:500 is directly given by
the code, refer EN 1993-1-1 figure 6.7:
Note: for the fabrication of this type of structures EN 1090 applies; maximum deviation is limited to 1:500:
With pinned chords and/or pinned braces, the geometrical imperfection is mainly determined by the play in
the pinned connections. To determine the total initial imperfection to be used in the analysis please consult
team leader or technical advisor.
Note:
For the New York Wheel mast sections calculation report 15015640-C2B presents an analysis of the effect of
play in the pinned connection on the initial geometrical imperfection.