COMPARATIVE SEISMIC ANALYSIS OF OVERHEAD CRANE ON

STEEL FRAME CARRYING STRUCTURE: EVALUATION OF

ADEQUATE EQUIPMENT-STRUCTURE INTERACTION MODELING
Andrea GODOY1, Frdric BARBIER2, Charisis CHATZIGOGOS3, Nicolas BESSON4,
Alfred THIBON5 and Martin RAY6

ABSTRACT

Overhead travelling cranes are common equipment in industrial facilities and are usually designed to
resist dead and live loads. Seismic motion produces noteworthy horizontal loads that can cause
damage and jeopardize the stability of the crane or its components. Moreover, effects of seismic loads
on the crane depend strongly on the mechanical characteristics of the carrying structure. In some
cases, travelling cranes may represent an important mass with respect to the carrying structure and
their support arrangement may constitute a real constraint for the dynamic response of the carrying
structure. For steel frame carrying structures, due to their flexibility and light mass, dynamic
interaction with massive travelling cranes (especially for crane multi-support configurations) can
significantly influence seismic response. Adopting a proper seismic analysis technique is in such cases
an essential step for the seismic design of overhead cranes since it should describe accurately the
dynamic response of the crane on carrying structure.
Three numerical analysis techniques are examined in the present study: (a) uncoupled response
spectrum method; (b) inertial coupled response spectrum method; (c) dynamic coupled response
spectrum method. The aim of the study is to evaluate conservatisms and inadequacies associated to
each technique. The study concerns the seismic response of an overhead crane supported by a steel
frame building exhibiting linear elastic material behaviour and modelled using a three-dimensional
finite element mesh. In order to cover a wide range of configurations, different crane and trolley
locations and crane loading states are taken into account in the study.
Since structural integrity and stability assessment of travelling crane under seismic loads is the
final objective of the study, conservatisms and/or inadequacies of each analysis technique are
accounted for with respect to structural failure and standard collapse criteria.
The implementation of a suitable analysis technique may lead to a radical redefinition of seismic
capacity and stability margins of travelling cranes. The present paper outlines the key features of crane
and carrying structure that must be accounted for in order to carry out a concise seismic analysis.

1
Project engineer, Godynamique et Structure, Bagneux, [email protected]
2
Project engineer, Godynamique et Structure, Bagneux, [email protected]
3
Project engineer, Godynamique et Structure, Bagneux, [email protected]
4
Project engineer, EDF/SEPTEN, Lyon, [email protected]
5
Project engineer, EDF/CIDEN, Lyon, [email protected]
6
Team Manager engineer, EDF/CIDEN, Lyon, [email protected]

1
INTRODUCTION

Analysis of crane seismic response must represent adequately crane and carrying structure features, as
well as crane-carrying structure interaction when it has a significant influence in response.
Nevertheless, detailed analyses, such as time history on integrated model, often involve high time-
machine cost and may not lead to a great improvement in results. On this matter, modal-spectral
method can be an interesting approach to reduce time-machine resources. However, simplification
requires a preliminary examination of the system dynamic behaviour.
Inertial, kinematic or dynamic crane-carrying structure interaction may influence seismic
response, particularly when crane represents an important mass relatively to the carrying structure.
Modal features of crane and carrying structure as well as support arrangement of crane point out the
potential influence of interaction.
This paper presents a comparative seismic analysis of overhead crane on steel frame carrying
structure in which the influence of inertial and dynamic interaction phenomena is evaluated.

METHODOLOGY

Seismic response of overhead crane is calculated by a modal-spectral analysis, taking into account up
to six configurations (varying the location of overhead crane and trolley as well as the loading state
and height of travelling hoist), as shown in Fig.1.
Three seismic analysis techniques are implemented in order to carry out a comparative analysis
of results. The main differences between three methods are the seismic load representation and the
modelling of the structures.
Uncoupled response spectrum method (a) is carried out with a model of overhead crane apart.
Seismic load is represented by floor spectra at support level of overhead crane neglecting any
interaction effect between crane and carrying structure. Floor spectra represent the carrying structure
response at support points of overhead crane.
Inertial coupled response spectrum method (b) deals with a modal-spectral analysis on overhead
crane model introducing floor spectra at support level as the input seismic load. Floor spectra are
calculated from modal features of steel frame carrying structure introducing overhead crane as lumped
masses on rails to take account of inertial crane-carrying structure interaction. Obeying to standard
engineering practice, lumped masses are equally distributed on both tracks, with no regard to the real
distribution of mass that depends on bearing conditions and position of trolley. Lumped masses
representing crane are coherent with travelling hoist loading state for each configuration.
Dynamic coupled spectrum analysis (c) takes account of inertial and kinematic interaction,
using a complete model of the crane attached on the steel frame carrying structure, in order to evaluate
the influence of multi-support arrangement of crane.

Floor spectra calculation

Floor spectra calculations are carried out with code FSG (Igutsa and Der Kiureghian, 1995) using soil
response spectrum and modal parameters from steel frame carrying structure, such as modal
frequencies, modal masses and mode displacements at support points of overhead crane.
Floor response spectra for each direction of the seismic motion are combined by SRSS (square
root of the sums of squares) rule in order to obtain the response spectra of the combined three-
dimensional seismic excitation.
Floor spectrum introduced in crane seismic analysis is the envelope of floor response spectra
obtained for the six configurations of overhead crane and trolley. Floor spectra for each configuration
correspond to carrying structure seismic response at different sets of attachment points.
Modelling and material properties uncertainties for the seismic calculations are introduced in
floor response spectra by the broadening of the spectrum around frequencies of peak response on a
range of 15%fpeak. Final floor response spectrum implemented in analysis is the broadened envelope
of floor spectra for all configurations.

2
 A.Godoy et al. 3

Configuration 1 Configuration 2 Configuration 3

Crane at mid-span, loaded trolley Crane at mid-span, unloaded Crane on gantry, loaded trolley at
at mid-span trolley at mid-span end carriage
North
track

South
track

Configuration 4 Configuration 5 Configuration 6

Crane on gantry, unloaded trolley Crane on gantry, unloaded trolley Crane parked, unloaded trolley at
at end carriage at mid-span end carriage

Gantry 1 Gantry 2 Gantry 3 N

Figure 1. Overhead crane configurations

Runway
a beams b

Travelling hoist
and wire rope

Trolley

Principal
75m beams
65m

End carriage
beams

40m Gantry 3

40m Gantry 2
45
m
Gantry 1

Figure 2. Numerical model : a) steel frame carrying structure, b) overhead crane

Table 1. Mass of crane and carrying frame structure
Configuration Unloaded trolley Loaded trolley
m % m %
Element
[t] [-] [t] [-]
Principal beams 95.0 4.7% 95.0 4.5%
Overhead End carriage beams 16.6 0.8% 16.6 0.8%
crane Electrical and mecanical components 9.4 0.5% 9.4 0.4%
Trolley and travelling hoist 31.7 1.6% 106.7 5.1%
Carrying steel Runway beams 199.1 9.8% 199.1 9.4%
frame structure Steel frame 1685.7 82.7% 1685.7 79.8%
Mass of crane and trolley : 152.7 7% 227.7 11%
Mass of carrying frame structure : 1884.7 93% 1884.7 89%
Total mass : 2037.4 100% 2112.4 100%
Modal analysis
Modal analysis of overhead crane and steel frame carrying structure is carried out in order to identify
representative dynamic response of each structure. Illustrative crane eigenmodes represent more than
1.6% of the total mass of overhead crane with unloaded trolley, or 2.4t. In the aim to apply an
equivalent filter criterion, representative eigenmodes for steel frame carrying structure are those
involving more than 0.13% of the total mass of steel frame carrying structure.
Mass of overhead crane and trolley components and carrying structure are given in Tab. 1 and
fundamental eigenfrequencies are given in Fig. 3.

Spectrum analysis
Maximum seismic response in each direction is obtained by implementing a complete quadratic
combination (CQC) of principal eigenmodes of the studied structure, overhead crane apart (a, b) or
crane on carrying structure (c). Newmark combinations of spectrum response, according to
ASN/GUIDE/2/01 (2006) and given in Eq.(1), are implemented to take account of the effect of three
directions of seismic excitation (SX, SY, SZ).

SX 0.4SY 0.4SZ
0.4SX SY 0.4SZ (1)
0.4SX 0.4SY SZ

NUMERICAL MODEL

A three dimensional model of crane and steel frame carrying structure has been developed in ANSYS
(ANSYS, 2007). Both structures are represented by 3D beam elements, with linear visco-elastic
material definition. Numerical models of steel frame carrying structure and overhead crane appear in
Fig. 2.

Steel frame carrying structure

Steel frame carrying structure is composed of three gantries, orientated along N-S vertical planes,
separated of 40m each. Each gantry is 75m high and 45m large. Horizontal and diagonal metallic
beams are located in the plane of gantries for the transmission of horizontal loads.
Travelling runway beams are I-shaped metallic beams located at 65m height from the basis of
the steel frame carrying structure. North and South crane long travel tracks are located on top of I-
shaped runway beams and extend between two gantries, along 40m. Runway beams are continuous
beams simply supported on gantries, with two spans of 42m.

Overhead crane
Overhead crane extends across 42m and lies on two rails. Long travel rails, parallel to E-W direction,
are connected on top of the I-shaped runway beams. Two box girder metallic beams constitute the
principal beams of crane. Trolley runway, parallel to N-S direction, is located on top of these beams.
Principal beams are supported on two box girder end carriage beams, located on North and South
track. Each end carriage contains a motor wheel and two free wheels (as shown on Fig. 4).
The main structure of crane is modelled by 6 degree of freedom 3D-beam elements. Beams are
represented at the cross section gravity centre fibre and connexion between beams is assured by rigid
link elements. The trolley is represented by rigid beams of equivalent mass that are linked to principal
beams. Travelling hoist and wire rope are modelled as a spring-mass system with equivalent frequency
of oscillation. The length of the wire rope is therefore taken into account in the modal analysis.

Crane bearing conditions

Bearing conditions of overhead crane represent the link between crane and rails. These conditions
introduce some simplifications inherent to a linear approach since sliding or uplift, as a non-linear
phenomena, cannot be reproduced. Likewise, the complex link of crane on cross travel direction, with

4
 A.Godoy et al. 5

a gap between doubled flanged wheels and rail (clearance) may result in a non-linear pounding
response that cannot be captured with a linear analysis.
Concerning long travel direction, motor wheels are rigidly linked to rails while free wheels are
allowed to slide. On cross travel direction, it is considered that only North wheels assure the stability
of the crane, while South wheels are free. In vertical direction, all wheels are rigidly linked to their
rails. All wheels are considered as an articulated link.

MODAL ANALYSIS

Overhead crane eigenmodes

Crane fundamental modal shapes are bending around weak and strong axis of principal beams.
Traveller hoist horizontal vibrations are uncoupled from crane vibrations due to their low frequency
compared to the crane eigenfrequencies. Nevertheless, there is a contribution of hoist load on vertical
vibration of crane principal beams.
For isolated crane numerical model (a, b), vertical bending of crane principal beams is coupled
with a translational motion along cross travel direction. This coupling is caused by bearing conditions
imposed at crane support points, with South wheels modelled as free wheels on cross travel direction.
This coupling of vertical and cross travel oscillations is imposed by numerical analysis and does not
represent a real dynamic response of the crane.

Crane-support structure interaction modes

For model including both crane and steel frame carrying structure (c), additional eigenmodes appear,
that represent crane-carrying structure interaction.
On long travel direction, the fundamental mode involving overhead crane is the fundamental
mode of steel frame carrying structure that induces a translational rigid body motion of the crane.
Eigenfrequency and modal shape are the same for all crane configurations: there is no coupling of the
response of crane and support structure along long travel direction.
Vertical and cross travel interaction modes show the coupling of crane and support structure
response. Fundamental mode along the cross travel direction is a bending of North runway beam
around weak inertia axis that induces a rigid body translational motion of crane. Eigenfrequency
associated to this mode depends on the crane trolley configuration. Moreover, if we compare modal
analysis results of steel frame carrying structure without the crane (a) and with the crane (b, c), a shift
in frequencies can be observed. Overhead crane modifies steel frame modes and vice-versa.
For modes along vertical direction, interaction phenomena are observed. There is a coupling of
crane and runway beam response. Bending around strong inertia axis of runway beams induces a
vertical motion of crane supports. Some modal shapes show that vertical bending of runway beams
induces differential vertical motion of North and South long travel tracks. These phenomena are
amplified for configurations with crane located at mid span of runway beam.

SEISMIC RESPONSE

Floor response spectra

Accelerations at overhead crane supports are represented by floor spectra at runway level on Fig. 5.
Floor spectra show the influence of crane on steel frame carrying structure response as well as the
anisotropic character of response of crane-carrying structure system.
Horizontal response at runway level is reduced compared to ground input motion. Peak floor
acceleration (PFA) at runway beam is between 0.07g and 0.08g along long travel direction, compared
to 0.12g peak ground acceleration (PGA). Carrying steel frame structure, with horizontal
eigenfrequencies below 1.0Hz, acts as a low-pass filter of horizontal ground motion. On the contrary,
vertical seismic response at runway level shows an important amplification of vertical PGA, with an
amplification factor, PFA/PGA, between 2.0 and 3.4.
Long travel direction response spectra values are similar for all methods since crane and
carrying structure can be considered as uncoupled systems. Multisupported nature of overhead crane
has little influence in response since crane bearing conditions are symmetrical in this direction and
motion of runway beams is in phase.

Long travel direction Cross travel direction

10.0 10.0 7.0
6.3
4.0 3.7
2.9 3.1
1.6 2.1 1.9
1.5 1.6
1.7
1.0 1.0 0.8
0.7
0.5 0.7
0.3
Eigenfrequencies [Hz]

0.1 0.1
Carrying Carrying 1st Principal 1st 2nd Carrying 1st 2nd
structure global structure torsion beams horizontal Runway Runway structure Principal Principal
deformation bending beams beams torsion beams beams
horizontal horizontal vertical vertical
bending bending bending bending
Vertical direction
10.0
5.3
5.0 3.7
3.8 3.1 Carrying structure apart (a)
1.9
1.7 Crane apart (a, b)
1.0
Crane on carrying structure (c)

0.1
1st Runway beams 1st Principal beams
vertical bending vertical bending

Figure 3. Eigenfrequencies for different models

UX=UY=UZ=0 UY=UZ=0

NORTH TRACK
Motor wheel
Free wheel

Principal
beams

UX=UY=UZ=0
UX=UY=UZ=0

Trolley
UX=UZ=0
UX=UZ=0

Cross travel direction

Y N
U is the relatif displacement between
Long travel
X direction crane (or trolley) and its support
SOUTH TRACK

GANTRY 1 UX=UZ=0 UZ=0 GANTRY 2

Figure 4. Bearing conditions of crane and trolley in numerical model

6
 A.Godoy et al. 7

Cross travel floor response spectra show a shift in amplification frequencies and in response
values. Floor response spectrum including dynamic interaction (c) is the highest for all frequencies.
Crane multisupport arrangement has an important influence in tranversal response. Assymetrical
bearing conditions on cross travel direction (North wheels are blocked along cross travel direction
while South wheels are free) is equivalent to a loaded track with the entire crane mass and an unloaded
track. This configuration is not well captured by a classical inertial coupled model (b), since it
represents an equal repartition of loads from crane on both tracks.
Vertical floor response spectra depend as well on the modelling of interaction. Taking account
of inertial interaction induces a shift on amplification frequencies. The system including crane mass
(b) is globally more flexible, hence peak frequencies are lower that the ones for a system with no
interaction (a). Taking account of dynamic interaction (c) induces a reduction of spectral response
values compared to uncoupled (a) and inertial coupled responses (b).

Displacements and efforts

Differential displacements at support points of overhead crane illustrate crane-carrying structure
dynamic interaction. Only results with dynamic coupled method (c) capture these displacements while
with uncoupled (a) and inertial coupled methods (b), motion of support points is of same amplitude
and phase.
Under seismic action, maximum relative displacement between tracks is equal to 13mm along
long travel direction, to 75mm along cross travel direction and below 2mm along vertical direction.
Relative displacements point out the coupling of long travel and cross travel direction of response due
to torsion of the carrying frame structure.
Supports displacements, as introduced by dynamic coupled method (c), increase flexibilty of the
crane system and induce rigid body motion of crane. Thus, seismic level associated to principal modes
of crane system is lower and deformations of crane components (such as principal beams or end
carriage beams) are reduced.
Displacements at mid-span of principal beams caused by their deformation are significantly
reduced when dynamic interaction is taken into account. Vertical displacement at mid-span of
principal beams is around 9mm for uncoupled method (a), 18mm for inertial coupled method (b) and
3mm for dynamic coupled method (c).
These results are directly correlated with efforts in crane components, such as bending moment
on principal beams. For instance, under a vertical seismic action, maximum bending moment around
strong inertia axis of principal beams (vertical bending) is equal to 1.2MNm for uncoupled method (a),
2.1MNm for inertial coupled method (b) and 0.8MNm for dynamic coupled method (c), which is
equivalent to a reduction of maximum seismic vertical bending moment between 43% and 62%.
Under a horizontal seismic action parallel to long travel direction, maximum bending moment
around weak inertia axis of principal beams (horizontal bending) is equal to 0.28MNm for uncoupled
method (a) and for inertial coupled method (b) and 0.12MNm for dynamic coupled method (c), which
is equivalent to a reduction of 57% of maximum seismic horizontal bending moment.

Resultant forces
Resultant seismic forces, quantified in terms of acceleration, illustrate the seismic level effectively
induced in the overhead crane. Under long travel direction excitation, resultant force in this direction
for methods (a) and (b) goes from 0.08g to 0.10g, which prooves the poor influence of inertial
interaction in response. For dynamic coupled method (c), resultant force is 0.06g.
For cross travel direction excitation, resultant force for methods (a) and (b) varies between
0.14g to 0.20g, while for dynamic coupled method (c), it goes from 0.29g to 0.46g. The high value
obtained with method (c) illustrates the importance of differencial displacements between crane
wheels and track, that cannot be appropriately estimated with methods (a) and (b).
In the case of vertical excitation, resultant force for method (a) is between 0.15g and 0.25g
while for method (b), it goes from 0.17g to 0.40g. Such difference comes from the variation of floor
spectra values for uncoupled and inertial coupled models in the range of vertical fundamental
frequencies of crane (Fig. 5). Vertical resultant force for dynamic coupled method (c) is the lowest,
ranging from 0.09g to 0.14g.
 1.00
a

Pseudo-acceleration [g]

0.10

Ground response spectrum

Floor spectrum (no interaction)
Floor spectrum (inertial interaction)
Floor spectrum (dynamic interaction)
0.01

1.00
b
Pseudo-acceleration [g]

0.10

Ground response spectrum

Floor spectrum (no interaction)
Floor spectrum (inertial interaction)
Floor spectrum (dynamic interaction)
0.01
1.00
c
Pseudo-acceleration [g]

0.10

Ground response spectrum

Floor spectrum (no interaction)
Floor spectrum (inertial interaction)
Floor spectrum (dynamic interaction)
0.01
0.10 1.00 10.00 100.00
Frequency [Hz]

Figure 5. Floor spectra at runway level for 7% critical damping: a) response along long travel direction, b)
response along cross travel direction and c) response along vertical direction

8
 A.Godoy et al. 9

ULTIMATE LIMIT STATES OF STRENGTH AND STABILITY

Stability assessment
Stability assessment of crane and trolley concerns sliding along track, uplift and derailment. Fig. 6
show safety factors for different crane configurations and analysis methods. Results show that crane
and trolley sliding along track is the most common instability during earthquake loading while uplift is
rare. For these instability mecanisms, dynamic coupled method (c) turns out to be as the most
favourable analysis method giving source to the highest safety factors.
Safety factor for uplift is calculated as specified in Eq.(2), where FZ,{G} is the vertical resultant
force (FZ<0 for uplift) of crane or trolley under dead loads and FZ,{G+E} under the combination of dead
loads, G, and earthquake loads, E. The most unfavourable results concerning uplift of crane and trolley
are obtained with inertial coupled method (b). The implementation of uncoupled method (a) represents
a maximum increase in safety factors for crane uplift of 9% (20%, for trolley uplift) while the
implementation of dynamic coupled method (c) induces an increase in crane uplift safety factors of
16% (44% for trolley uplift).

FZ ,{G + E }
SFuplift = 1 + (2)
FZ ,{G}

With regard to crane sliding, dynamic coupled method results (c) are the most favourable since
the seismic effect on long travel and vertical direction is reduced. Consequently, decompression of
blocked wheels is limited and friction force is sufficient. The implementation of dynamic coupled
method (c) results in an increase of safety factors ranging from 37% to 48%.
Derailment safety factor is evaluated as specified in Eq.(3), with {G+E} the relative displacement
along cross travel direction between South track and South crane wheels and d the mean of the width
of rail and wheel. Unlike for uplift and sliding, derailment safety factors are the lowest for dynamic
coupled method (c).

d
SFderailment = (3)
{G + E }

Relative displacement between South track and South crane wheels is partly due to crane
deformation and partly due to track displacement. For methods (a) and (b), results introduce the part of
displacement due to the deformation of crane only, while tracks relative displacement cannot be taken
into account since tracks are not represented in the numerical model. Therefore, estimated relative
displacements out of dynamic coupled method (c) are higher and more precise than for the two other
methods. Derailment risk assessment is hence not accurate when dynamic coupling is not taken into
account.

Structural integrity
Structural integrity of overhead crane critical sections is examined, comparing yield stress of steel to
maximum stress under the effect of dead loads and earthquake loads. Safety factor is defined in
accordance with FEM 1.001 (1998). Parameters on Eq.(4) are yield strength of steel, e; maximum
axial stress on critical section, , and maximum shear stress on critical section, .

3 e
SFstrength = min e ; e ; (4)
2 + 3 2

Dynamic coupling method (c) produces the most favourable results for principal beams and end
carriage beams critical sections, as displayed in Fig.7.
Regarding principal beams, inertial coupled method (b) is the most conservative, with safety
factors between 1.9 to 3.5 for all configurations. Results obtained with uncoupled method (a)
respresent an increase in safety factors between 5% and 16%. The implementation of dynamic coupled
method (c) represents an increase in safety factors varying from 9% to 33%.
Concerning end carriage beams, uncoupled method (a) and inertial coupled method (b) results
are alternatively the most conservative, depending on the configuration. Dynamic coupled method (c)
results represent an increase of 13% to 41% on safety factors.

Crane uplift Trolley uplift

2 2

Safety factor [-]

1.5 1.5

1 1

0.5 0.5

0 0
1 2 3 4 5 6 1 2 3 4 5 6
Crane sliding Trolley sliding
2 2
Safety factor [-]
Safety factor [-]

1.5 1.5

1 1

0.5 0.5

0 0
1 2 3 4 5 6 1 2 3 4 5 6
Crane derailment
7.0
6.0
Safety factor [-]

5.0 Uncoupled method (a)

4.0 Inertial coupled method (b)
3.0
Dynamic coupled method (c)
2.0
1.0
0.0
1 2 3 4 15 2 6 3 4 5 6
Configuration

Figure 6. Stability criteria safety factors for all configurations

Crane principal beams End carriage beams

6.0 14
5.0 12
Safety factor [-]
Safety factor [-]

Safety factor [-]

4.0 10
8
3.0
6
2.0
4
1.0 2
0.0 0
1 2 3 4 1 5 2 6 3 4 1 5 2 6 3 4 5 6
Configuration Configuration

Figure 7. Structural integrity safety factors for all configurations

10
 A.Godoy et al. 11

INTERACTION DOMAIN

An interaction domain can be defined in terms of mass and frequency ratios between crane and
support structure (ASCE 4-98, 1998 and ASN/GUIDE/2/01, 2006). Interaction domain establishes the
features of crane and carrying structure for which a coupling of response is observed. Interaction
domains are well documented to one degree of freedom (DOF) equipment with a unique support point.
In the case of overhead crane, interaction domain is properly fitted to dynamics features when the
response is strongly monomodal.
Dynamic features of crane and carrying frame structure along different response directions are
plotted on Fig. 8. Mass ratio, Rm, is the ratio of effective mass of crane and carrying structure
fundamental mode in each direction. Frequency ratio, Rf, is the ratio of crane and carrying structure
fundamental eigenfrequencies.
It appears clearly that long travel direction motion of crane and carrying structure are uncoupled
since fundamental frequencies of crane and carrying structure are different and effective mass of
fundamental mode of crane represents less than 10% of the effective mass of fundamental mode of
carrying structure. In this direction, fundamental mode of carrying structure mobilizes the steel frame
globally.
100.00
Limit of interaction domain
Long travel features
Cross travel features
Vertical features
Frequency ratio Rf [-]

10.00

1.00

NO INTERACTION INTERACTION
0.10
0% 1% 10% 100%
Mass ratio Rm [%]

Figure 8. Interaction domain according to ASN/GUIDE/02. Crane and carrying structure dynamic parameters for
all configurations
Vertical and cross travel features show that the response of crane and carrying structure is
coupled, since carrying structure in these directions involves essentially runway beams, as the
effective carrying structure of overhead crane. The mass of crane represents more than 10% of the
effective mass of carrying structure.
Cross travel features illustrate crane as a rigid system compared to carrying structure due to the
important flexibility of runway beams bending around weak inertia axis. However, inertial coupled
method (b), as well as uncoupled method (a), leads to inaccurate results since it is considered that
motion of both tracks are of same amplitude and phase. Yet differential displacements between both
tracks are relevant, as is pointed out by dynamic coupled method (c) displacement results.
Vertical features of crane and carrying structure point out to a partial frequency uncoupling,
since fundamental frequency of crane happens to be lower than carrying structure eigenfrequencies.
When an equipment (i.e. crane) is flexible compared to its carrying structure, the influence of mass on
the global system response is neglectable. Dynamic coupled (c) and uncoupled (a) vertical response
spectrum at runway level in the range of crane eigenfrequencies (from 1.9Hz to 3.1Hz) are strikingly
similar (Fig. 5). Furthermore, inertial coupled (b) vertical response spectrum values in the same
frequency range appear to be significantly higher. Hence inertial coupled method (b) leads to over-
conservative results on vertical direction.

CONCLUSIONS

Evaluating the seismic response of an equipment (i.e. overhead crane) attached to a carrying structure
(i.e. steel frame) requires a preliminary analysis of their modal features and the definition of the
effective carrying structure for each direction. The influence of crane-carrying structure interaction
in seismic response of overhead crane is strongly correlated to the response direction, as a result of the
anisotropy of bearing conditions and carrying structure behaviour.
Long travel direction seismic responses of crane and carrying structure are uncoupled since
modal features remain constant for all models. In this direction, the entire steel frame carrying
structure acts as a 1-DOF flexible oscillator, with low fundamental frequency of vibration (around
0.7Hz) compared to overhead crane fundamental mode.
Cross travel seismic response is complex, since significant inertial and kinematic interaction is
produced. The effective carrying structure depends on the position of overhead crane. When the crane
is located at mid-span of runway beams, effective carrying structure is composed mainly by North
runway beam (bending around weak axis) and when the crane is located above gantry, effective
carrying structure is composed by runway beam and gantry. Furthermore, dissymetric bearing
conditions with regards to crane tracks introduce differential displacements between tracks that can
only be determined when carrying structure is modelled.
Vertical seismic behaviour of crane-carrying structure system depends on excitation frequency.
For fundamental frequencies of crane, in the range of 2Hz to 3Hz, only kinematic interaction takes
part, while for higher frequencies, above 4Hz, inertial interaction controls the response. Effective
carrying structure concerns runway beams bending around strong inertia axis.
Results attest that a correct implementation of crane-carrying structure interaction leads to a
reduction of seismic efforts and hence an increase of safety factors regarding integrity (up to 16% for
end carriage beams and 33% for principal beams) and of safety factors regarding stability (up to 16%
concerning crane uplift and 48% concerning crane sliding). The quantified differences between
methods include the conservatism introduced by the broadening of floor spectra for uncoupled method
(a) and inertial coupled method (b). Furthermore, relative displacements between wheels and track on
cross travel direction can only be obtained with a model including carrying structure (c).
A preliminary analysis of modal features of primary system and subsystem is useful to identify
the importance of interaction as well as to define the effective carrying structure at each direction. A
dynamic coupled spectrum analysis carried out on a model including the equipment and the effective
carrying structure (i.e. runway beams) should lead to satisfactory results with efficient time-machine
resources.
Interaction domains defined for 1-DOF subsystems can be useful to characterize the
fundamental mode of vibration yet they are inadequate to describe accurately the seismic behaviour of
the system.

REFERENCES
ANSYS (2007), Release 11.0. ANSYS, Inc.
ASCE 4-98 (1998), American Society of Civil Engineers, Seismic Analysis of Safety Related Nuclear Structures
and Commentary
ASN/GUIDE/2/01 (2006), Autorit de sret nuclaire, Guide de lASN ; Prise en compte du risque sismique
la conception de gnie civil dinstallations nuclaires de base lexception des stockages long terme des
dchets radioactifs
Igutsa T and Der Kiureghian A (1995) FSG-Floor Spectrum Generator, last revision, April 1995
FEM 1.001 (1998), Rgles pour le calcul des appareils de levage, 3rd edition, 1998.10.01

12