ELLA Extended Live Load Analysis Guide
ELLA Extended Live Load Analysis Guide
SOFiSTiK | 2026
ELLA
Extended Live Load Analysis
SOFiSTiK AG
This manual is protected by copyright laws. No part of it may be translated, copied or reproduced, in any form or by
any means, without written permission from SOFiSTiK AG. SOFiSTiK reserves the right to modify or to release
new editions of this manual.
The manual and the program have been thoroughly checked for errors. However, SOFiSTiK does not claim that
either one is completely error free. Errors and omissions are corrected as soon as they are detected.
The user of the program is solely responsible for the applications. We strongly encourage the user to test the
correctness of all calculations at least by random sampling.
Front Cover
© Autobahn GmbH des Bundes | Photo: Christoph Seelbach
Contents | ELLA
Contents
Contents i
SOFiSTiK 2026 i
ELLA | Contents
ii SOFiSTiK 2026
Task Description | ELLA
1 Task Description
ELLA is a program for the analysis and evaluation of imposed loads that acts on spatial beam or
shell structures in the form of load trains which move along certain traffic lanes. The envelopes
of the scalar variables for the entire structure are usually the result. In the case of simple
statical systems and simple load cases a traditional analysis of the system can be faster for a
large number of imposed load cases under circumstances. The evaluation of influence lines
or influence surfaces becomes necessary for an accurate analysis of larger systems, as they
are usual particularly in the bridge construction. It is not so simple to determine either the
most unfavourable load position or the used parts of a load in order to maximize the certain
internal forces for important parts of the structure outside its midspan. In addition, different
load positions are critical for the different internal forces and moments. The evaluation of the
influence lines reduces the extent of the input very significantly.
The entire surface of a structure at which imposed loads act gets divided into straight or curved
lanes of variable width. Any load trains can be assigned to each lane. They can move indepen-
dently or in relation to each other. The traffic lanes are defined with axes which may contain
circular segments or clothoids. The loading refers always to the lane axis. An influence line
is generated for a vertical loading along the lane’s axis for each lane as well as for a moment
loading about the same axis due to the eccentrical parts. Additional influence lines for brake
or centrifugal loading can be considered upon request.
The program evaluates influence lines only for corresponding traffic lanes even in the case of
plane structures. However, a satisfactory representation of the two-dimensional effects can be
achieved for the available element distribution by partitioning the structure into multiple lanes.
The influence surfaces can be either transferred from the database or determined as a rule
with a calculation with an own static module of the influence line program.
In the second case, however, all the single element stiffnesses as well as the stiffness matrix of
the entire structure must be available. This requires that at least one load case has been anal-
ysed with the programs ASE or STAR2 with option CTRL SOLV 1 or 3, and the corresponding
data is still available. Load cases with settlement of support or non-linear load cases are not
suitable for this.
2 Theoretical Principles
When a large number of influence lines of the structure has to be calculated with this method,
an own load case will have to be analysed for each cross section and for each internal force.
This procedure exceeds usually the available CPU time and memory capacities of the static
programs.
In the case of plane structures the discontinuity is defined by means of a singularity. This
singularity does not result from a finite element analysis in principal. However we have a
smooth approximate influence function which may be used to derive the principal problem of
precision in a Finite-Element-Analysis very clearly (see Hartmann/Katz - Structural Analysis
with Finite Elements, Springer). This should be considered particularly in the case of shear
forces!
In order to accelerate the evaluation, the loads are simplified within a lane by a vertical loading
and a torsional moment due to the eccentric position of the vertical load. Thus the description
by means of an eccentric lane or an eccentric load is equivalent for single loads.
Most of the standard road loadings are internally partitioned into loads at the left-hand side and
at the right-hand side of the shear centre. The location of which is automatically determined
from the influence values for torsional and vertical loading.
In the case of two girders and articulated load distribution the result is the same, if the loading
is placed centrically at lane 1 or eccentric with respect to lane 2. The eccentricity moment
offers the necessary correction for the distribution of the loading at the longitudinal girders:
Lane 1 Lane 2
Lane 1:
Lane 2:
However, great differences are already available in this case for rigid load distribution. An
eccentricity should be then only relatively small. Here up to 20 percent deviations would result
in the distribution between the two variants.
In the case of three or more girders the results can be very different as the next two figures
show. It can be stated as a rule that both the load and the lane must be located in the same
span region in terms of the lateral behaviour of the structure.
Different results are always obtained in FE structures, when load and lane are not at the same
or neighboring elements. The deviations of the results are negligible in the neighbouring ele-
ment, but they increase of course as the distance between lane and load increases. As the
loading of single moments for planar shell or slab structures may lead to further errors, eccen-
tricities of loads must not be used for these types of the structure, instead an evaluation with a
moment M
load P
resulting force
must be interpolated. The program searches then for two series of nodes in the direct vicinity
of the lane and interpolates perpendicularly to its axis.
The selection of a neighboring node series is trivial for plane structures. For girders, however,
it is in general a problem for the artificial intelligence. It is extremely difficult for a program to
develop the human ability of the general overview. Therefore, this part of the program can not
execute in a uniquely determined way. A specific constellation not being recognized correctly
can occur again and again. The user must take a proper action in such cases by defining a
sequence of nodes or beam elements.
For the interpolation perpendicularly to the lane axis a distinction must be done very exactly
between three variations that result from different types of the load distribution:
In the first case (INT 0) an eccentrically placed loading generates directly a torsional moment.
The influence line of the shear force for the effect of the eccentric influence has no jump in the
loading location.
In the second case (INT 1) the load is replaced by two forces that act at the two girders.
The torsional moments result only as a secondary phenomenon due to the constraint of the
rotations. The influence line for the effect of the eccentricity may include jumps. This is usually
the most accurate method for plane structures analysed with Mindlin elements.
In the third case (INT 2) torsional moments result additionally from the fixed constraint of the
deck slab.
Following the inverse principle of influence lines it is also possible to view the different meth-
ods as different interpolation grades of the transverse influence line as shown in the following
picture:
In a first step an immediate evaluation of the maximum effects for a uniform loading and several
axle types with predefined distances of wheels is performed in the transverse direction.
• The values P are the sums of the influence values at the wheel positions for up to
three load directions (P = vertical, PT = transverse, PL = longitudinal)
• The value M is the influence value of a torsional load which is represented by a pair
of opposite forces acting at the wheel positions (marked green). It is calculated from
the vertical influence values at those positions.
• The given influence values p+ and p- are the integrals for a uniform vertical load with
the influence values across the lane width with positive and negative contributions.
• The given influence values r+ and r- are the integrals for a uniform residual vertical
load with the influence values on the right and left side of the lane.
• The first row shows the values for the most important configuration of wheels, the
following lines show the valueas of the alternate axles (here approx. 15% less)
• The obtained values will then become the values of the longitudinal influence line at
this position for the second evaluation along the lane.
• If the applied load train is smaller than the lane width, the influence values for a
uniform loading are not applicable. This case will be checked with an error message.
• If the applied load train has a different distance of the wheels, the influence values
will not match if the transverse influence line is not linear. Multiple wheel configura-
tions within a load train are possible. This case will be checked with a warning.
• All wheels are applied together, there is no possibility to split the wheels for differ-
ent signs of the influence line. If a load train is defined with single wheels, their
contribution is evaluated based on a central P and a M from the excentricity.
• It might be possible that the extreme load position is not the same for the maximum
vertical effect and the maximum torsional effect. This is neglected by using the most
adverse effect of the forces in all cases, based on the fact that any moment will be
transferred via forces of the wheels
Loads of the load trains are thus considered unfavourable, when a load in the positive gravity
direction (downwards) with the same eccentricity at the corresponding point of the influence line
would also act unfavourably. Loads that act at regions of the influence line with an opposite sign
are considered favourable. This classification decides about the appropriate dynamic factors,
the possible use of the reduction factors, and whether the load should be taken into account
at all. Horizontal loading will not be included in this decision. However all load directions are
considered for the determination of the most unfavourable load position and the final evaluation
of the results of cause.
In order to consider eccentricity effects (eg. torsional loading), the loads may be automatically
subdivided if the shear centre is within the width of the lane. However, despite all these fea-
tures the user should take care, that the real load situation may be mapped by the load train.
Sometimes either the lane axis must be displaced or a secondary lane must be added or the
loads must be given explicitly.
For the evaluation of horizontal loading by braking or wind one has to take care of some more
subtle effects. If for example the influence line for a bending moment is taken with respect to
the eccentric longitudinal force at the upper deck, we have areas with positive and negative
values:
• If the brake load is taken over the total length in the same direction, the contributions will
cancel mostly. This kind of loading should be defined as standard load case, but not within
ELLA.
• If the load train has a dedicated travelling direction, only those contributions will be taken
into account acting unfavourably, although the design code (e.g. Eurocode) may define
the load value by the total, but limited loading in the lane and an overall loaded length. A
modification of the load value depending on the loaded length will not occur.
• If the load train is allowed to travel in any direction, all contributions will be summed up with
their unfavourable sign. Thus the areas with the wrong sign will either have traffic in the
opposite direction or acceleration forces are effective. Although the Eurocode specifies this
possibility only for the total load value, this approach will also compensate for the missing
effect of the adoption of the load value to the loaded length from the remark above.
The most unfavourable load position is sought with a three-stage method. In a first step some
typical load patterns are positioned at different locations perpendicular to the traffic axis. This
may be done with a transverse influence line (LSEL INT>2) or based on simple mechanics of
a lever.
In a second step the load train is placed at selected positions which either the program selects
itself or the user prescribes with POSL. Variable lengths or convoys of loading are estimated
separately with a simple method. The lengths of the positive loads are set always equal to their
maximum value, whereas negative loads are placed in such a way that their influence value at
the end of the load is the largest possible.
In a third step the most unfavourable value of all analyzed positions keeps on being examined
for local extreme values. That means the load train positions and all lengths are varied until
A special hint should be given to the Culmann load position. For two adjacent loads the most
unfavourable moment is obtained when the larger force and the resultant of the two loads are
symmetric to the mid of the span. But the maximum moment is then not a the mid of the span!
For the special case of two equal loads P with a distance on a single span beam of length ,
the maximum moment is given by:
2
Mm = P · · 1 − (2.1)
2·
AGRL -1 Length values will be kept separately, each span is taken on its own. This is the mos
unfavourable approach.
AGRL 0 Length values will be combined within the same lane, but the load value will be
increased if the most important span would be more unfavourably
AGRL 1 Length values will be combined within the same lane, the load value is dependant
only on the total length (Default)
AGRL 2 Length values will be combined over all lanes
In any case only full spans or sections of the influence lines will be considered.
2.7 Limits
Some limits must be satisfied when using the program. The load direction of all load trains is
always in the direction of the gravity. The application of ELLA for plane frames or shear walls
provides so many ambiguities, that we do not support this feature currently.
Analysis time may become quite large. Thus the following points may be important:
• ECHO EXTR may create a really huge amount of print with high time requests.
• It is worth to select only those lanes with intended loadings.
• The influence lanes may be calculated only once, it is possibel to repeat an evaluation in a
separate input block.
• If only few influence lines are required, the analysis with unit dislocations may be faster
(see example ella1_threespan.dat)
• ECHO STAT prints the time of the individual steps.
The program uses a dynamic working memory which is used mainly for the storage of the
values needed for the influence lines. An equation for the necessary memory can not be
given, yet the in fact used size can be printed with ECHO STAT. Thus the amount of reserved
memory may be specified after PROG ELLA with a parameter Mnnn, where nnn is the size
of the memory in millions of 4-byte words. As a very high request will slow down the virtual
memory performance, the optimum value has to be found individually.
The program is executed mostly in a uniquely determined way. There are, however, two types
of problems for which a unique solution does not exist or can not be found by the program
under circumstances. These are:
• The algorithm for the selection of the nodes next to the lane It is to be expected that un-
expected results arise at irregularities (e.g. kinematic constraints, double defined beam
members etc.).
• Load trains with variable length In the case of load trains with loads of variable length the
whole problem may become very complex. One should use so few changeable lengths as
possible. In addition the user should define first the lengths which have the most significant
effects. In the case of doubt it is better to define more evaluation cases with fixed load
lengths.
3 Input Description
The geometric and statical model must have been already defined and stored in the database.
The geometric definitions of the lanes may be already available in the database.
The input is articulated in blocks which are decollated with the record END. Within every block
one can:
The influence lines and load positions remain stored in a separate database (.CDI) and are
available for further evaluations.
[mm] Explicit unit. Input defaults to the specified unit. Alternatively, an explicit as-
signment of a related unit is possible (eg. 2.5[m] ).
[mm] 1011 Implicit unit. Implicit units are categorised semantically and denoted by a cor-
responding identity number (shown in green). Valid categories referring to the
unit ”length” are, for example, geodetic elevation, section length and thickness.
The default unit for each category is defined by the currently active (design code
specific) unit set. This input default can be overridden as described above. The
specified unit in square brackets corresponds to the default for unit set 5 (Eu-
rocodes, NORM UNIT 5).
Record Items
ECHO TYPE VAL
CTRL OPT VAL
GAX NO S X Y Z R A
NX NY NZ TYPE TITL
LANE NO TYPE SA SE WR WL YCA
YCE YRA YLA YRE YLE L
LSEL NO INT DX DY DZ RSEL RSEQ
LSEL LSEQ
CALC TYPE LMAX LMIN FROM TO DELT SYST
LINF T1 L1 T2 L2 T3 L3 T4
L4 T5 L5 T6 L6 TITL
APPL TYPE TGEN
SIZE TYPE SC W H MARG FORM HDIV
C1 C2 C3 C4 C5 C6 C7
SHOW SNO TYPE NO X ETYP DSV YMAX
ACT TYPE GAMU GAMF PSI0 PSI1 PSI2 PS1S
PART SUP TITL
LC NO PSI0 PSI1 PSI2 PS1S GAMU GAMF
TITL
TRAI TYPE P1 P2 P3 P4 P5 P6
P7 P8 P9 PFAC WIDT PHI
PHI2 PHIS PHI4 V FUGA XCON DIR
DIRT
TRPL P PB PW PF PFAC DIST A
DPOS DMIN DMAX Y Y2 HW ZW
HF PHI B
TRBL P PB PW PF PFAC DIST A
L LMAX LDEL Y Y2 HW ZW
HF PHI B
SAVE LCB TYPE TITL
CASE NO GRP
POSL NO TRAI FACT
OPT SMIN SMAX DEL YEX V P
Table continued on next page.
Record Items
SYNC PLON PTRA FUGA IMPA EXCE EXTR
COMB V P SYNC
TYPE L1 F1 L2 F2 L3 F3
L4 F4 L5 F5 L6 F6
L7 F7 L8 F8 L9 F9
L10 F10
The records HEAD, END and PAGE are described in the general manual SOFiSTiK: ’FEA /
STRUCTURAL Installation and Basics’.
The record name ECHO must be input for every record so that the options are not confused
with other records with identical names. The volume of printed data is also controlled with the
record SHOW.
With ECHO IL the influence lines will be printed, in general as pictures, but with option EXTR
also as text. The small influence lines in transverse direction will be shown with the Option +4.
Without a SHOW selection the volume of pictures may become quite large.
With the option LPOS FULL the actual load positions of all evaluations will be saved to the
database and the graphical visualization of the influence lines will include the loading and will
be repeated for all evaluation cases.
ECHO EVAL YES prints only the generated load cases for the case of ECHO RES NO. For
the load train position ECHO EVAL FULL is needed and for the detailed statistical data ECHO
EVAL EXTR.
Hint
If no ECHO options are specified, the options LANE, IL, LOAD, LPOS and RES are set to
YES (regular output), while EVAL and STAT are set to NO (no output). If the ECHO option
is used without explicitly stating the OPT and/or VAL item, value FULL will be taken as
the default.
With CTRL SECT one can obtain evaluations only at certain beam sections. The input options
are:
CTRL SECT 0 only section with print-flag (see program SOFiMSHA record BSEC)
CTRL SECT 1 all beam sections (default)
With CTRL AGRL one can select how ELLA should cope with the aggregation of loaded
lengths. The load values as well as the impact factors may depend on those values.
CTRL AGRL -1 Length values will be kept separately, each span is taken on its own.
CTRL AGRL 0 Length values will be combined within the same lane, but the load
value will be increased if the most important span would be more un-
favourably
CTRL AGRL 1 Length values will be combined within the same lane, the load value is
dependant only on the total length (Default)
CTRL AGRL 2 Length values will be combined over all lanes
It has to be pointed out, that option -1 will be not be general applicable in case of elastic
supports, where strong discontinuities in the results may be obtained and thus is not a me-
chanically sound procedure.
A lane axis already stored in the database (by SOFiMSHC or SOFiLOAD or ELLA) can be
activated with the record LSEL.
Each lane has two coordinate systems. There is a local system (x) at the lane axis for real
distances. However, the inputs for the positions along the lanes are entered in a global system
of stations (s) which defines an alignment of the lanes to each other.
A general geometric axis with all details may be now defined in SOFIMSHC. For SOFiLOAD /
ELLA / TENDON only a simplified definition is provided as follows.
• Definition with a length resp. station in the plan view For the first record of an axis the
start point (X,Y,Z) and the start direction (NX,NY,NZ), the length S of the first segment and
optional curvature and clothoidal parameter is specified. All consecutive records contain
only the next S-value and the curvature parameters.
If a radius is defined, a transition element from the last radius to the current value is gener-
ated. Positive values for R cause a curvature to the right-hand side, whilst negative values a
curvature to the left-hand side. If A is defined, a clothoidal spiral transition curve is created
where the curvature increases for the positive values of A and decreases for the negative
values. If the radius R is not given it will be calculated automatically based on the length, if
it is given together with A the curvature of the start point will be calculated instead. Basket
arches with piecewice constant curvatures may be defined with A=0.0 and the selected
radius.
• Definition about 3D points at a curve The coordinates have to be input for all records. The
given direction (NX,NY,NZ) is the local z-axis and thus a defined circular arc turns around
this vector.
The described elements will be automatically adopted from WinGRAF as dynamic axis. A
graphical check is recommended with the programs WING or WinGRAF.
The options to define the elevation or cross falls are now only available in SOFiMSHC.
The total loading area will be subdivided in up to 99 individual traffic lanes relative to a given
geometrical axis. These may be created automatically by a predefined standard section or
with explicit coordinates. One may subdivide each segment with different widths and dynamic
impact factors. All lanes have to be defined together to an existing axis, any new definition will
delete all previously defined lane definitions.
The width of the lane is defined through the left and the right boundaries. The lane can be
widened or narrowed by different input data at different stations. If the lane edges happen to
intersect the lane, this should be defined only at distinct points to allow for correct treating of
torsional loading.
The width of the load train may be less than those of the lane. In that case additional basic
areas are established loaded by a constant load specified with the load train. If the load train
is larger than the lane width, reductions of the applied loading are possible.
The lane 0 has to be defined always, for the following procedures two cases have to be distin-
guished:
• There is only lane 0, the width of the moving loading is defined only by the load train. Only
then, the eccentricities of the load train are applied absolutely for all lanes.
• There are other lanes within the curb stone limits. Then lane 0 defines the total extend of
the traffic load including pathways. The width of the moving load is the maximum extend of
the other lanes.
Ó Ó Ó Ó Ó Ó Ó Ó Ó
Ó Ó Ó Ó Ó Ó Ó Ó Ó
AXIS.1
Ó Ó Ó Ó Ó Ó Ó Ó Ó Ó Ó
Y −4.00 −2.00
T 0.00
101 2.00 4.00 m
M 1 : 70
Default constant sections are available via literals as follows. If such type (without blank
spaces) is given at the main lane, secondary lanes will be automatically created:
For the following types the definition of the curbstone coordinates is always required, which is
constant along the total lane.
If the lanes have all the same width or the number of lanes is exceeding 6, the lanes starting
with no. 1 will be aligned from the right edge (positive y-values). If this is not the case lane 1 is
always the central lane in the mid of the total traffic area, more lanes with numbers 2 to 7 will
be created on alternating sides aside to lane 1.
In all cases right adjusted lanes starting with number 10 and left adjusted lanes starting with
number 20 will be available. However they will be created as own elements only if the second
case is given and the eccentricity is larger than 0.30 m.
For TMH7 lanes 11,12,13 resp. 19,18,17 with extreme eccentricity will be created.
Ó Ó Ó Ó Ó Ó Ó Ó
Ó Ó Ó Ó Ó Ó
AXIS.11Ó Ó Ó Ó Ó Ó Ó Ó
AXIS.10
Ó Ó Ó Ó Ó Ó
Y −4.00
T 102
−2.00 0.00
T 2.00
101 4.00 m
M 1 : 70
Figure 3.5: Two lanes adjusted to the right loaded and residual area
Ó Ó Ó Ó Ó Ó Ó Ó Ó
Ó Ó Ó Ó
AXIS.20
Ó Ó Ó Ó Ó Ó Ó Ó Ó
AXIS.21
Ó Ó Ó Ó Ó Ó Ó
Y −4.00
T−2.00
101 0.00
T 1022.00 4.00 m
M 1 : 70
Figure 3.6: Two lanes adjusted to the left loaded and residual area
Residual areas are the areas between the edges defined by the width of the load train and the
edges of the lane itself. The pathways are thus only included in lane 0. Gaps between the
lanes are NOT treated as residual loading areas.
For the loading of the lane in SOFiLOAD it will be subdivided in the longitudinal direction
into segments (e.g. each span). If several lane-definitions should be combined to one single
segment, the definition of L is only possible for the first of the definitions, while specifying for
all the others the literal CONT at this position.
For railway bridges there are detailed tables available within EN 1991-2 or DIN FB 101 how
to select the governing length. Further there are detailed descriptions how to calculate the
effective pressures below the ballast. For the latter the input values HEFF to INCD are needed,
however applicable and definable in SOFiLOAD only.
+ INC
s
BEFF HEFF HS
INCD INCD y
YL z YR
Figure 3.7: Ballast and sleepers
The height HSA of the tracks or road above the lane and the cross fall INCA/INCE of the lane
will create additional moments within SOFILOAD and ELLA (INT<3).
An already defined lane gets activated for evaluation with the record LSEL. At NO one has to
specify the axis identifier (4 chars) optionally appended by a dot and a digit for the lane number
(e.g. A1.3 but also with wild cards like A1.? or A1.1?). Without a lane number all lanes of that
axis are selected, which is not always a good choice to minimize analysis time. At the same
time parameters can be specified for the selection of the neighbouring nodes.
Lanes which are defined with the record GAX in the same block are selected automatically with
the standard values for DX, DY and DZ. A special feature can be achieved with an input of a
negative value for INT. In this case all geometric parameters and influence lines are deleted for
this lane.
If the user does not define an explicit sequence of neighboring nodes, Ella will try in a first step
to look for all nodes which are adjacent to the lane and tries to link them to one (INT 0) or two
(INT 1,2) nodal sequences.
For this search DY and DZ describe a region in which the corresponding nodes must be situ-
ated. DY is measured in this case at the right-hand side and at the left-hand side of the lane.
A value of about 90 percent of the distance between the main girders is a sound definition.
The parameter DZ defines the limits of the region under the lane which is to be examined. This
is for example necessary for three-dimensional structures at which parts of this structure lie
sometimes beneath the lane. Nodes above the lane axis are never considered.
For INT>2 DX specifies a maximum distance of the load points along the axis. While the default
is to use a load point at every edge crossing the lane, a smaller value will insert additional points
between.
For INT 0 to 2 the parameter DX specifies, how large the angle between the lane axis and
an edge of the nodal sequence is accepted. An input is necessary, if transverse girders and
longitudinal girders have similar angles of intersection to the lane. If the node sequences may
intersect the lane, DX has to be input with a negative value.
If ELLA is not able to find suitable sequences of nodes, then it is possible to specify
two sequences explicitly by boundary or beam elements containing the desired nodes with
RSEL/RSEQ and LSEL/LSEQ.
For plates and shells structures it is recommended to make an evaluation of the influence area
in transverse direction before establishing the longitudinal influence values. With types 3,5,7
and 9 these numbers of load points in transverse direction will be examined based on the
existing and selected QUAD elements.
QUAD elements are searched only based on the plan view. If a secondary group or a range
of primary groups is specified, no further checks about the elevation will be made. However
if no such definitions are given, only those QUAD elements will be considered situated with a
maximum of DZ below the loading or where the loading is within the element. Thus it is strongly
recommended to use a group definition to specify the traffic area properly. One secondary
group or a range of primary groups may be specified.
For the evaluation the transverse width of the traffic lane, the distance of the wheels and the
minimum and maximum eccentricities are required. For this purpose a loadtrain has to be
specified with TRAI providing the necessary information. For the evaluation in the longitudinal
direction these values may not be changed any more. Therefore the width of the traffic lane
of the given load train must match the values of the load trains positioned in the lanes. The
distance between the wheels should also match, but differences often result only in small devi-
ations, which are flagged by a warning message only. If several load trains with different wheel
spacings are used, a load train must be selected containing all possible wheel spacings. In
the case of an LM3 with accompanying LM1, the LM3 must be selected for the evaluation. It
also contains the information about the LM1 wheel distance, but not vice versa. For special
cases a USER defined train containing all necessary wheel spacings may be applied. In ex-
ceptional cases an explicit definition of the values with B, BWH, YMAX and YMIN in LSEL is
also possible.
A graphical control of the transverse influence lines may be obtained with ECHO EL +4.
For each extreme value of any scalar variable a load case can be generated and stored in the
database. The corresponding values of the rest of the internal forces are calculated automati-
cally too.
As an alternate way one may read influence lines or areas which were stored as load cases in
the data base with the input option LINF. The values are then interpolated to the defined lanes.
The results, however, do not get stored in the database.
TYPE Scalar variable for which the influence lines should be LT -
evaluated
LMAX Load case number for the maximum − *
0 no evaluation
LMIN Load case number for the minimum − *
0 no evaluation
FROM Smallest element number −/ Lt4 1
TO Largest element number −/ Lt4 FROM
INC Increment of the element numbers − 1
SYST Subsystem for the analysis LT *
SPAC / GIRD
PHI Impact factor or characteristic length for impact factor − *
if given with a length unit
With CALC the elements will be selected for which the influence lines will be established. The
parameters specified with LSEL will be applied for this calculation. Without any CALC definition
the last influence lines in file [Link] will be reused.
If no input occurred for FROM, TO and DELT, all elements of the structure will be selected.
Multiple overlapping definitions are possible, the latest positive definition is valid. Instead of
the element numbers a group may be selected via or FROM ’GRP’ group.
The parameter SYST can define a reduced influence line evaluation. This may result in a
significant reduction of the CPU time and the memory requirement e.g. in the case of girders,
if a three-dimensional model was defined due to the prestress and the temperature, while the
imposed loads are acting only at a two-dimensional girder. The influence lines for the axial
force as well as for the transverse bending are not calculated or evaluated with the input SYST
GIRD.
1. Beam Elements
N, VY, VZ, MT, MY, MZ, MB, MT2 Forces and moments
4. Nodes
5. Result sets
Result sets defined in SOFIMSHA will be selected with TYPE RSET:xxxxx, where
xxxxx is either the ID of the item within a set to be evaluated or a number if the x-th
item of all result sets should be evaluated. Supported are all forces and moments of
beam, truss and cable elements, forces and stresses in the center of QUAD and BRIC
elements and nodal displacements, support forces and spring forces (but not resultant
PT) Also any formula expressions within a result set definition will inihbit to select this
item as an influence line for ELLA.
At FROM and TO the idents of the result sets may be specified in lexical order (FROM
may also contain wildcards, e.g. "SU*") .
The evaluation of the influence lines for nodal support forces is possible by an import of a
deflection area caused from a nodal displacement, as force of a supporting spring or via the
result sets.
The evaluation of the influence lines for pile elements does not include the effect of the bedding,
thus only the results at the nodes are correct.
The application of ELLA for plane frames or shear walls provides so many possibilities
for miss-understandings, that we do not support this feature.
Influence lines or areas which were stored as load cases can be taken over for the evaluation
along the defined lane axis. However, results are only printed and do not get stored in the
database.
Positive load case numbers are processed with the first set of impact factors (bending), while
negative load case numbers use the second set (shear).
The CPU time for obtaining the influence lines can be reduced by using the default value of
APPL. The options TRAV and FULL are only useful for three-dimensional systems with the
corresponding loads.
The item TGEN specifies at which cross sections within a beam element the additional loading
locations are defined. A loading location is inserted, when the difference of the reciprocal
stiffness of two neighbouring cross sections is larger than a multiple of TGEN.
1 1 1
− > TGEN · (3.1)
E E+1 E
The input of TGEN is recommended for structures with many similar cross sections, because
the CPU time and the memory allocation are reduced then significantly. For instance, TGEN
1.E10 suppresses the insertion of any loading locations at internal beam sections.
Graphic representations are only produced, when the record SIZE is input. If a pen number is
defined negatively, the corresponding line type does not get plotted. Coding of lines see GPLI
The lines to be plotted are prescribed with the input of the record SHOW. Without ECHO LPOS
only the influence lines will be shown, with ECHO LPOS the actual load position will be also
visualized, which will need more pictures in general.
The item causes a subdivision of the sheet into some strips, so that the corresponding influence
lines can be plotted at the same drawing. Positive values draw the lane axis along the paper
width and subdivide the paper height; negative values have a contrary effect. The default is
that three lines get plotted parallel to the largest dimension of the paper.
With SHOW the lines which should be printed (record ECHO IL) and / or plotted (record SIZE)
are selected. The item DSV controls the distance between marked positions in the output.
Available values are shown then at least in this distance. However, extreme values are output
additionally in all cases. There will be up to four lines for three force components and the
torsional moment. The presentation may be changed with record SIZE.
PART Partition of the action group, defines which part of the LT *
equation the action belongs to:
G permanent action group
P action group for prestress and creep
Q variable action group
load groups 1 - 99: act only for categories
within an action
Q_1 variable action load group 1
...
Q_99 variable action load group 99
SUP Handling of the load cases within the action / cate- LT *
gory:
PERM permanent with uniform factor
PERC permanent with individual factors
COND conditional
EXCL mutually exclusive within category
EXEX mutually exclusive within action
UNSI COND with unfavourable sign
USEX EXCL with unfavourable sign
ALEX permanent but exclusive within action
As type of the action one may use any literal with up to 4 characters, however some combina-
tions are reserved for special purpose. They are defined in the corresponding INI-files and will
follow the designation of the selected design code as close as possible. An overview is given in
the table below. Furthermore, it is possible to subdivide each action with up to two characters
in categories, designated with a character from A to Z, each having distinguished combination
coefficients or partial safety factors. The category will be appended with an underscore to the
name of the action.
• Each category has its own combination values and its own load cases and a default super-
position rule within the action category.
• Each action comprehends all its categories, i.e. one may select either a sub category on
its own or all categories by the generic name of the category. (Q selects all Q_?)
Figure 3.9: Structure of the actions, their categories and their load cases
For example there are for road bridges in EN 1990 for the uniform life load UDL and the movable
double axle different combination coefficients. For simple cases one will define a load case for
every span of type L_U (UDL), to be superimposed conditionally, while the discrete load cases
of different load positions of the double axle will be defined as type L_T and exclude each
other. All load cases however will be applied together as leading variable action.
Finally any individual load case may have its own combination factors assigned in SOFiLOAD.
(see LC). This should be avoided however and a definition of combination coefficients in MAX-
IMA is then at least ambiguous and of little use therefore.
The actions indicated in the table are used only as preset action, provided that they are avail-
able in the INI-file of the selected design code.
The entries within one row with preceding Z for additional loading and S for special loading
may be treated in AQB as belonging to the same action.
For the seiesmic action there is in EN 1998 an importance factor γ depending on the impor-
tance class to be selected by the user explicitly:
Action ψ0 ψ1 ψ2
Q: Payload for buildings [0.70] [0.50] [0.30]
- Q_A Housing [0.70] [0.50] [0.30]
- Q_B Offices [0.70] [0.50] [0.30]
- Q_C Assembly rooms [0.70] [0.70] [0.60]
- Q_D Sales rooms [0.70] [0.70] [0.60]
- Q_E Storage rooms [1.00] [0.90] [0.80]
Live load for buildings
- Q_F truck load < 30kN [0.70] [0.70] [0.60]
- Q_G 30kN < truck load < 160kN [0.70] [0.50] [0.30]
- Q_H Roofs [0.00] [0.00] [0.00]
L: Live loads according EN 1990
- L_T, Tandem Axle of LM 1 (γ = 1.35) [0.75] [0.75] [0.00]
GR_T
- L_U, UDL loading of LM 1 (γ = 1.35) [0.40] [0.40] [0.00]
GR_U
- L_C Load model UIC 71 (γ = 1.45) [0.80] [0.80] [0.00]
- L_D Load model SW/0 (γ = 1.45 [0.80] [0.80] [0.00]
- L_E Load model SW/2 (γ = 1.20) [0.00] [1.00] [0.00]
Live load for buildings
- L_F truck load < 30kN [0.70] [0.70] [0.60]
- L_G 30kN < truck load < 160kN [0.70] [0.50] [0.30]
- L_H Roofs [0.00] [0.00] [0.00]
S: Snow for buildings
- S_L = buildings up to 1000m elevation [0.50] [0.20] [0.00]
- S_H = buildings above 1000m elevation [0.70] [0.50] [0.20]
- For SIA 260 with PSI0 as elevation h 1-60/ h 1-250/ h 1-1000/ h
Action ψ0 ψ1 ψ2
W: Wind for buildings [0.60] [0.20] [0.00]
T: Temperature (no fire) for buildings [0.60] [0.50] [0.00]
F: Settlements [1.00] [1.00] [1.00]
Other actions [0.80] [0.70] [0.50]
A definition of ACT will reinitialize the action with the defaults defined in the INI-file depending
on the category of the design code. Values GAMU to PS1S replace the predefined values.
The user should check the defaults in all cases. Deviations of the PSI-values for wind and
temperature especially for bridges or other non buildings are to be expected as all values are
"boxed values" depending on individual or country dependant settings. For example you will
find in OENORM B 4750 categories with trucks between 30 and 60 kN with values augmented
by 0.1 for ψ1 and ψ2 .
For all EN codes as well the SIA codes there are two defaults for the live load for buildings with
the categories F,G and H. The user can choose, whether the live load for buildings should be
in an own action L (with categories L_F, L_G and L_H) or whether these categories should be
a part of the action Q in addition to the payloads for buildings (categories Q_A till Q_H).
If the design code NORM ot its category CAT is changed in AQUA, all actions should be rede-
fined. ACT INIT deletes all defined actions and superposition rules if specified in SOFiLOAD.
With the specification of ACT all subsequent load cases in SOFILOAD will then have this action
type as default.
The item PART defines, to which action group of the selected superposition equation the input
action belongs. The actions in the superposition equations are distinguished according to their
temporal variances:
Thus the item PART allows the correct classification of user-defined actions, but also of special
cases as for example PART G SUP ALEX (always but exclusive, it means, that only one load
case is always used). Exclusively the item SUP controls, how the load cases which are a part
of the actions are used for the superposition (always, conditional, mutually exclusive).
Meant for example PART P, that the input action is used in the equation part Pk for superposi-
tions for ultimate and serviceability limit state (EN,DIN, DIN-FB, see manual MAXIMA, chapter
2).
The item PART depends on TYPE for preset actions. For example PART G is the default for
the actions G, G_1 and G_2, PART P for the actions P and C, PART Q for the actions Q and L,
PART A for the action A and PART E for the action E.
The item PART has to be input for user-defined actions. Otherwise the user-defined actions
get PART Q (default).
Hint
All categories of an action must have the same definition of PART. E.g. categories Q_A
and Q_B of the variable action Q get PART Q. Only the load groups of a variable action
PART Q_1 till Q_99 can be different.
The item SUP defines the default treatment of multiple load cases within an action. The default
will be taken from the INI-File, but for free defined types of actions it is EXCL in general, if the
type of the action has a leading Q it will become COND and for a leading G it will be PERM.
In MAXIMA one may overwrite this for any particular load case of an action or recombine
load cases (e.g. 0.7*LC_101(Ex)+0.5*LC_102(Ey)). Due to historic reasons there are special
literals at record LC in MAXIMA in addition to the literals of SUP:
SOFiLOAD MAXIMA
Explanation
ACT SUP LC TYPE
PERM PERM,G always (permanent), partial safety factor action-wise
PERC PERC always (permanent) with variable factors, partial safety factor load-
case-wise
COND COND,Q conditional (only unfavourable)
EXCL Ano mutually exculsive, MAXIMA record LC different alternative load
case group numbers per category
EXEX Ano mutually exclusive but conditional inclusive categories (only within
an action), MAXIMA record LC only one alternative load case
group number for the categories of an action
UNSI W changing sign (e.g. seismic actions)
USEX Xno changing sign, mutually exclusive with input of an alternative load
case group number in MAXIMA record LC
ALEX AGno always but exclusive (only within an action) with input of an alter-
native load case group number in MAXIMA record LC
F follow-up load case to be combined
Using SUP PERM the same partial safety factor is used for all load cases of a permanent
action (e.g. ACT G). In this case the sum of the values of the single load cases is decisive,
whether the unfavourable (GAMU) or the favourable (GAMF) partial safety factor is applied. In
the contrast to that the partial safety factors are considered load-case-wisely, if SUP PERC is
defined. Here one load case of the permanent action can get GAMU and another GAMF.
SUP EXEX excludes only categories against each other within an action (no exclusion
of different actions). If for example the categories X_1 and X_2 of the action X are defined
with SUP EXEX, then only the most unfavourable load case is used either from X_1 or from
X_2 and also only, if it has an unfavourable contribution. In the contrast to that an unfavourable
load case from X_1 and an unfavourable load case from X_2 are used for the input of SUP
EXCL.
SUP ALEX defines a permanent alternative group (MAXIMA record LC TYPE AG1). Only
the load cases of an action or all load cases of the categories of an action are used for the
formation of a permanent alternative group. Here only a load case is considered for the super-
position, also if it has a favourable contribution. Load cases of different actions do not exclude
themselves mutually. If SUP ALEX is input for some actions, a load case is considered per
action.
The total mechanism of the defaults for the action types is governed by the following rules:
• By selecting a design code within AQUA all definitions given in the associated INI-File in the
[action] -section will become available. If no such file is available, defaults in the program
are available for all EC and newer DIN (since year 2000). For all other cases there are
exactly two predefined actions G and Q without any coefficients or factors.
• When addressing an action, a copy of this definition is created in the database which has
then the possibility of deviating factors if explicitly defined with this record or via the SOFi-
PLUS menus.
• When creating a load case, the factors are copied from the action definition to the load
case. However, those factors may be modified for any individual load case.
• For special design purposes the factors may be changed temporarily when using the load
cases in AQB.
The load groups PART Q_1 till Q_99 can be used for a better formation of combination without
intermediate superpositions in MAXIMA e.g. for the load models of the bridge design (EN
1990/A1).
Hint
Load groups act only within an action. Therefore, the categories are only mutually exclu-
sive!
If different load groups are defined for categories with corresponding inputs for the item SUP,
then the intermediate superpositions and the following superpositions are done in one MAXIMA
run.
Example:
PROG SOFILOAD
HEAD
ACT XY_A gamu 1.4 0.0 0.6 0.4 0.2 0.6 PART Q_1 SUP excl
ACT XY_B gamu 1.3 0.0 0.5 0.3 0.1 0.5 PART Q_1 SUP excl
ACT XY_C gamu 1.45 0.0 0.6 0.3 0.0 0.6 PART Q_2 SUP exex
ACT XY_D gamu 1.35 0.0 0.5 0.2 0.0 0.5 PART Q_3 SUP exex
...
PROG MAXIMA
HEAD
COMB 1 desi TYPE desi BASE 2100
ACT XY_A
LC -1
ACT XY_B
LC -1
ACT XY_C
LC -1
ACT XY_D
LC -1
or
First of all three load groups Q_1, Q_2 and Q_3 are defined with PART in the program
SOFiLOAD. The categories XY_A und XY_B are a art of the load group 1. Their load cases are
mutually exclusive only within the respective category about the input SUP EXCL. The cate-
gories XY_C and XY_D are a part of the load groups 2 and 3. All load cases of both categories
are mutually exclusive about the input SUP EXEX.
With one of the MAXIMA combinations then the superposition is done as follows: First of all
MAXIMA searches which load cases of the load group 1 are decisive (possible one ore two or
no load case = first intermediate superposition).Then the decisive load case of the load groups
2 and 3 is determined (possible one or no load case = second intermediate superposition). In
the last step the final superposition is done. Either the decisive load cases of the load group 1
or the decisive load case of the goups 2 and 3 are here the final result.
It should be noted here that load groups are mutually exclusive. If, for example, the superpo-
sition of the load cases for two load load groups are specified with SUP COND - conditionally,
only the load cases of the decisive load group are taken into account.
Example:
PROG SOFILOAD
HEAD
ACT XY_1 gamu 1.4 0.0 0.6 0.4 0.2 0.6 PART Q_1 SUP cond
ACT XY_2 gamu 1.3 0.0 0.5 0.3 0.1 0.5 PART Q_2 SUP cond
The load cases of category XY_1 belong to load group 1 and those of category XY_2 belong
to load group 2. Thus, either the load cases of XY_1 or of XY_2 are taken into account in the
superposition depending on which brings the more unfavorable result.
The input of the loading occurs in the form of load trains identified by a load case number,
which must not be used for other cases, and the position of load trains within lanes.
The combination values may be taken from the action selected by the TYPE or will be default to
the values according to EC1 Part 3 Appendix C, or may be specified explicitly. For the resulting
load cases of the evaluation the maximum value of all applied trains will be saved. These
values will then be accepted from the program MAXIMA.
Deviations of the PSI-values for wind and temperature especially for bridges ot other non build-
ings are to be expected as all values are "boxed values" depending on individual or country
dependent settings. For example you will find in OENORM B 4750 categories with trucks
between 30 and 60 kN values augmented by 0.1 for ψ1and ψ2.
Furthermore the DIN-Fachberichte supply different combination values for the constant load
an the heavy vehicle.
XCON Minimum distance for repeated load train (convoy) [m] 1001 0
YEX possible eccentricity of train (+/-) [m] 1001 0.0
DIR Direction of train LT B
B = in both directions
R = only to the right (positive)
L = only to the left (negative)
N = do not apply
DIRT Direction of transverse loading LT B
B = in both directions
R = only to the right (positive)
L = only to the left (negative)
N = suppress this loading
The description of all available load trains can be found within the next pages. If a value for
the convoy distance is given (XCON), the specified train will be applied multiple times with
that given minimum distance between the last load ordinate and the reference point of the
consecutive load train.
The the residual area loading P5 is applied without dynamic impact factors and without cen-
trifugal forces. It is applied only for:
• Areas between the constant width of the load train and the polygonal bounds of the lane.
If the load train has no width (WIDT 0) there won’t be a residual area either (Railways)
• Areas between the borders of lane 0 and all other lanes (e.g. pathways)
• The total lane area for the load group GR4 (e.g. pedestrian bridges)
The factor PFAC is a multiplier in general for all load elements of the load train, it may be used
for classified LM71 or SW/0 loadings. (e.g. a value from the list 0.75 - 0.83 - 0.91 - 1.00 - 1.10
- 1.21 - 1.33 - 1.46 from EN 1991-2 6.3.2(3)).
The factor PFAV specifies the factor for favorable loadings (see e.g. EN 1990-2 GRP 13/14
and 23/24). For road train loadings this factor will be applied only for the wheel loadings. The
default for railway load trains is 0.0 for the LM71 and 1.0 for all other load trains. The default
for the majority for most other road trains is 0.0. For load trains for roads, those factors may be
specified only with individual load train definitions.
The eccentricity of the load train within the lane should only be defined when applying the load-
ing. Only for the wide heavy load trains (LM3, HB etc.) an explicit eccentricity of that loading
may be defined a priori. The definition of YEX in TRAI is intended to describe the eccentricity
for railway loadings according EN 1991-2 6.3.5., predefined with r/18 in any direction.
The load trains have a moving direction which is not only important for some of the train loads
but also for the direction of the braking load. Positive load values are considered to be braking
loads in traveling direction, negative load values are accelerating loads in reverse direction. If
DIR N is specified, no impact factor will be applied and the brake and velocity and wind load
will be preset to zero. Also in this case some filters are available in ELLA.
DIR controlls the sequence of generated loads for a standard train, but it cannot be used to flip
the sequence of a user defined load train.
Wind loading may be defined in general by a wind attack area with the section definition. If there
will be no traffic for strong winds, this is the maximum wind load on the structure itself. But a
combined variabel wind and traffic loading an area load P6 may be defined, acting horizontally
upon a traffic band with height P7. However, AASHTO uses a line load P6 at the height P7
which have to be defined instead. The sign of the wind load is defined as positive if coming from
the right in the moving direction, but option DIRT allows a more specific definition. ELLA has
the possibility in record POSL via OPT to apply the wind load in the most unfavourable direction
at its own, if DIRT has allowed this. Both options are combined with a logical ”and”-operation.
The meaning of value P8 is different for each loading type. The sign of the braking load will be
taken from the moving direction. A changing sign (DIR B) will be only possible for ELLA to take
care of. ELLA and SOFILOAD allow to change the direction or deactivate the loading.
All load trains may be extended with TRPL and TRBL definitions. If none of the standard load
elements of the rain should be used, but other properties like impact factors should be retained,
the literal USER may be defined for P1. The user should take care of:
• The default live load P5 should be deducted for the areas loaded by the defined loading.
• The distance to the last load of a standard train may be not as expected due to automatically
generated loads of those type and should be checked.
The loading may be applied with a variable eccentricity YEX. For a railway loading (default
±r/ 18) additional moments are created. For road loading the total train will have an offset
(default 0.0) to the right or left.
general: 1.33
fatigue load models: 1.15
• CAN Impact factor according to CAN/CSA
between 1.25 and 1.40, depending on the available length for one, two or three
axles. (A depencency for the true number of applied axles is not available)
• IRC Impact factor according to Clause 211 Fig. 5. for Class A, B and AA loading
for concrete and steel bridges.
• JRA Impact factor according to JRA are only distinct for T-Loading, for L-loading
there are deviations for standard concrete and steel bridges.
• AS Impact factor according to AS 5100 are defined as (1+α ). For roads fixed val-
ues are given in Table 6.7.2, however as M1600 requires a greater allowance
if only a single three- axle component controls and there exists also a station-
ary loading S1600 without dynamic allowance, a user definition is required in
some cases.
Railway loading distinguishes between ballasted and direct fixing (input with
LANE) and account for the reference length as follows:
p
PHB = 1.0 + 2.16/ ( LPH − 0.20) − 0.27 ( > 3.6) (3.10a)
p
PHB = 1.0 + 2.16/ ( LPH − 0.20) − 0.17 ( > 2.0) (3.10b)
LPHI should be defined for the axis via LANE. However ELLA may also use the distance of the
zero values of the influence line as a good approach to the effective length. If no input for LPHI
is given, the dynamic factor will be set equal to the value with a minimum length and the width
of the lane will be set equal to 0.0.
Dynamic coefficients according to appendix E of DIN FB 101 are dependant on the frequency
of the structure and should be specified explicitly therefore.
Dynamic coefficients according to the French Fascicule No. 61 section 5.5 depend on the ratio
G/S of the permanent to the maximum loading. If this formula should be taken this ratio has to
be specified instead of the impact factor PHI.
load components.
The forces calculated according to the laws of physics with vin [ m/sec] are:
2
Pƒ = P · ; g = 9.80665 m/ sec2 (3.11)
Rg
Sometimes there is an equivalent formula in the design code for a speed in [ km/h] with a
factor of v2 /(127R).
For FUGA you may however specify an additional factor explicitly. Negative factors will enforce
the laws of physic to all type of load trains.
In those cases where a constant horizontal centrifugal loading or horizontal bumping has to be
considered a horizontal loading PW/HW may be specified with TRPL/TRBL.
The double axle should be applied only in total. Thus all loads of the axles will be applied even
if one is favourable.
Load Model 1
There is one double axle with a spacing of 1.2 m. For the different lane classes it is generally
0.5∗
2.00 1
0.5∗
recommended to use three different load trains with individual adjustment factors αq. These
factors are specified in national annexes and are defined in the INI-files. (e.g. DIN Fachbericht
with a value of αQ = 0.8, SIA with a value of αQ = 0.9). Other values may be specified with a
direct load value if needed.
The brake loading is calculated according to the length of the total lane if the item DIR is not
defined as N. It will be distributed as uniform load along the whole loaded lane if the axle load
exceeds 200 kN:
The upper limit is defined by the requests of the STANAG, it may be redefined by a national
annex or other means. This may be specified via a global variable EN1991-2_QLMAX in the
INI-file or via TVAR in the section loading.
Load Model 2
This load model is a single axle taken from load model 1, but with a dynamic impact factor and
the wheels may act independently. As this load model is only for local elements it should not be
selected within ELLA, especially as it does not work with influence areas (INT>2). A possible
application is the accidential load case of a vehicle on the pathway:
1 2 1 2
0.40
2.00
3
2.00
αQ2 Q2k
0.40
3
0.50
Load Model 3
This load model is for special cases of heavy load vehicles. We have to distinguish between the
narrow version fitting within one lane (2 axles) and the broader version for P2=240 (3 axles).
For the latter case the loading of the same load train in two synchronized lanes with special
provisions of YEX should be used. Parts of the load train not within the lane imits will be clipped
automatically.
1.50 1.50
X X
4.20
2.70
1.50 1.50
1.50 1.50
1 2 1 2
3.00 3.00 3.00 3.00
The variants with an axle loading of 240 kN are wider than the default lane. Thus, it is in
general necessary to specify a transverse offset when applying this load. This will be +/- 0.75
m for the main lane, +2.25 for the left adjacent lane and -2.25 for the right adjacent lane.
The speed of the load train is a very important parameter, it does not only control front and rear
for some load trains, but it defines also if the load train is moving slowly (v = 5 km/h) or with a
regular speed (v = 70 km/h). This triggers if an impact factor is applied or not.
Further there will be no loading witin a distance of 25 m before and behind that vehicle, but then
a load train LM1 with ist frequent load values (ψ1 ) may be applied, by specifying the load case
number of a correspondig LM1 load train at P4. To select the relative position to the vehicle
before or behind the sign of the P4 definition will be taken. A positive value will select a vehicel
before, while a negative value will select a value behind. The implemented algorithm does not
allow to select this automatically.
As the coefficients ψ1 are different for UDL and TS, we use from this load case the value ψ1 for
TS, but for UDL and the residual loding the default value of 0.4 of the EN will be taken.
Load trains of the Special Types General Order (STGO) will be selected with TYPE SV, where
P1 defines the class (80, 100 or 196), P2 is not used and P3 defines the central distance
(values of 1.2, 5.0 and 9.0 m are common). All other remarks especially about P4 remain
valid:
165 165 165 165 165 165 165 165 165 180 180 180
kN kN kN kN kN kN kN kN kN kN kN kN
1.2m 1.2m 1.2m 1.2m 1.2m 1.2m 1.2m 4.0m 1.6m 4.4m
0.35 m 3
0.35 m
3.0 m
Load trains of the Special Order (SO) will be selected with TYPE SOV:
Load Model 4
This is a pedestrian load model with a basic load area with a maximum value of 5.0kN/ m2 .
For a loaded length greater 10 m the load value will be decreased:
120
qƒ k = 2.0 + > 2.50 kN/ m2 (3.16)
L + 30
QSV1 QSV2
3.00 m
0.20 m
0.20 m
1.30 m
For the application of this load model the definition of the loadgroup is mandatory: GR3 or GR4
for the area loading and GR0 for the service vehicle.
P1 = 1 90 / 190 kN
P1 = 2 80 / 140 / 140 kN
P1 = 3 90 / 180 / 120 / 120 / 120 kN
P1 = 4 90 / 190 / 140 / 140 kN
P1 = 5 90 / 180 / 120 / 110 / 110 kN
0.40 m
2.00 m 0.40 m X 1
This is a complete truck with two double axles P1 with a distance of 6.0 m designed for the
fatigue design.
P1 = 1 70 / 130 kN
P1 = 2 70 / 120 / 120 kN
P1 = 3 70 / 150 / 90 / 90 / 90 kN
P1 = 4 70 / 140 / 90 / 90 kN
P1 = 5 70 / 130 / 90 / 80 / 80 kN
Those classical load trains are used in several countries in Europe (DIN 1072, OENORM B
4002, Spanische IAP and Brasilian NBR). National variants may be selected with a country
code. The width of the moving loading is 3.0 m in general, for the OENORM it is 2.5 m and for
the Spanish IAP a minimum distance to the secondary truck suggests a width of 4.0 m. The
rest of the traffic band is loaded by P3, the remaining areas inside the lane are loaded with
the basic area loading P5. With DIR N one may select a secondary lane without any dynamic
factors.
The definition of a second load train beside with P2 und P4 can not be supported for some
of the recent features. It us strongly recommended to use synchronized lanes instead, it is
required to use two lanes with individual load trains instead. The loading in the second lane
requires the value of P5 defined for P3 and the impact factor to be deactivated with 1.0.
All loads are divided to portions to the right and to the left of the lane axis, which can act
independently.
∗ The IAP (Instruccin sobre las acciones a considerar en al proyecto de puentes de carretera,
Ministerio de fomento, 2000) has a constant value of 4.0 kN/m2 on the total area of the bridge,
but without any deduction for the SLW. For fatigue a SLW with 390 kN and an impact factor of
1.2 = 468 kN, to be obtained with PFAC 0.78.
DIN 1072 defines a total brake loading with a maximum of 900 kN and a minimum of 30 % of
the heavy load trains within the principal and secondary lane or 25% of the total principal lane.
As the brake loading is an other load class than the standard life load it is not always possible
to apply both loads within one loadcase. IAP has a minimum of 140 kN and a maximum of 720
kN.
The loading HA and NA is described by a uniform load and an additional concentrated single
or tandem load with spacing P4.
The loading HA (Departmental Standard BD 37) is represented by a uniform load P3 · pu, with
the magnitude pu depending on the length of the loading, and either a concentrated load (KEL)
P1=120 kN or a tandem loading with P1=148 kN (with distance P2, default 1.2m). The value
of P4 defines the loaded length L. If P4 is not given, L will be calculated in ELLA based on the
sum of the distances of the null points of the influence line.
• up to L = 50 m:
p = 36 · L−0.1 (3.18)
It is then possible that shorter loaded lengths could lead to higher forces. As this can not be
treated easily, ELLA uses an approximation by evaluating an effective length according to BD
37/01 figure 11. “Base lengths for highly cusped influence lines”. In a first step the integral
of the total positive and negative contributions of the influence line will be evaluated, then the
effective length is caluclated:
In Hong Kong (BS + Country code 852) the formulas are slightly different:
• up to L = 70 m:
• from L = 70 m:
the old definition of BS may be selected if P4 is specified explicitly negative or the country code
has been set for Israel (972). it is defined as:
The brake loading is also calculated according to the loaded length of the lane:
The loading NA (Bridge Design Code TMH7, South Africa 1981) is represented by a uniform
load P3 · pu, with the magnitude depending on the length of the loading and a concentrated
p
load P1=144/ n kN (KEL). The value of P4 defines the loaded length L as with the HA loading
described above. P2 may be used as with HA to generate a tandem loading.
To accurately determine the loaded length according to Bridge Design Code TMH7, the control
setting CTRL AGRL in the ELLA module must be set to -1. This setting enables the evaluation
of each span within a lane individually. The influence of the loading sequence number n when
p
calculating the concentrated axial load P1=144/ n kN needs to be defined explicitly, using the
option Intermediate Combination of Evaluations in the ELLA module. For more information,
please see the record COMB in the ELLA module.
• up to L = 36 m:
p = 36 (3.23)
• from L = 36 m:
The brake loading is also calculated according to the loaded length of the lane:
The loading HFUG allows centrifugal loading according BD 37-01 6.9, It contains a concen-
trated transverse load and a equal but distributed vertical loading P1. The load train may be
repeated every 50 m. It shall be applied for radii less than 1000 m.
The load value P1 defaults to 400 kN (300 kN for Israel’s IS 1227). The length of the distributed
loading is preset with P2 = 6.0 m.
The loading HB and NB consists of a heavy loading and a simultaneous traffic band load.
In some countries very similar load trains are used (South Africa, Israel). These Variants will
be selected with type NB or the selected country code:
COUN = GB BS 5400 with distances 1.8 and x = 6.0, 11.0, 16.0, 21.0 or 26.0 m
COUN = ZA TMH7 NB with distances 2.0 and x = 6.0, 11.0, 16.0, 21.0 or 26.0 m
COUN = IL IS 1227 with distances 1.5 and x = 8.0, 13.0, 18.0 or 23.0 m
BS 5400 makes a distinction between four positions of the HB loading, which affect the magni-
tude of the traffic band load. These are controlled by the factor P3 and the eccentricity P2:
The value P3 of the latter load varies with the length P4 according to the loading HA.
The Departmental Standard BD37/88 has other factors (Table 14). The default value for P3 will
be selected based on a nominal width bL of 3.048 m and the first lane.
TMH7 applies the NB vehicle not combined with other life loads. Values P2 to P4 will be
ignored.
Brake loading are % of the total load. If the train has a unique traveling direction it is applied
between the first and second axle. If the train travels in both directions it is distributed in two
separate blocks at both ends.
P8 = 0.25 (HB)
P8 = 0.20 (NB)
Hint
The basic area loading P5 is not deducted in the case of HA and HB loading. It always
acts upon the full width of the lane.
The IS 1227 has a load train HC, which is very similar to the HB. The axial load of the last two
axles is augmented by a factor of 2, the distance x is exactly 13.0 m.
Old loading according AASHTO from 1944 may be selected with three loading classes 10, 15
and 20.
The width of the train is limited to 10’ = 3.048 m or for the SI version to 3.000 m. Thus, uniform
loading is modelled with two stripes with a fixed width.
The loading HS is described by a uniform load and an additional concentrated single or tandem
load with spacing P4. The value P1 specifies the load train class according to AASHTO. The
value P2 of the old load definitions should be selected for shear explicitly.
32 kip 32 kip
(145 kN) (145 kN)
8 kip
(35 kN)
P3
14' 0" to 30' 0" 14' 0"
(4.3 to 9.0 m) (4.3 m)
Figure 3.25: Load train HL / HST (three axle truck, ed. 2002/2005)
The load train has a width of 10 ft (3.048 m), the variant distance can be between 14 and 30
ft (4.30 to 9.00 m). The special version US2 resp. SI2 selects the two compact vehicles with a
mimim distance of 50 ft (15 m) with 90% of the loading for negative moments.
The total weight P2, the lane load P3 and the distance P4 are preset as follows:
P1 Load Train P2 P3 P4
10 H10-44 20 kip 0.32 klf -
15 H15-44 30 kip 0.48 klf -
20 H20-44 40 kip 0.64 klf -
US HL 93 72 kip 0.640 klf 14 ÷ 30 ft
SI HL 93 325 kN 9.3 kN/m 4.30 ÷ 9.00 m
US2 HL 93 72 kip 0.640 klf 14 ft
SI2 HL 93 325 kN 9.3 kN/m 4.30 m
USF HL 93 (fat) 72 kip 0.640 klf 4 ft
SIF HL 93 (fat) 325 kN 9.3 kN/m 1.20 m
The traffic band load P3 has been specified in the 2002 edition as an alternate loading, which
can be obtained by setting P3 equal to zero. With the load model HL 93 the loading P3 is active
on the full longitudinal length.
The variable distance between the two axles of the HL may be overwritten with a fixed value
by P4. The detailed load model for local and fatigue design may be selected with the load train
types USF and SIF switching the dynamic allowance impact factor to 1.15 and a fixed distance
of 30" (9.0 m). P4 is then the spacing of the double axis, thus P4 0.0 will create the simplified
modell.
Axle no. 1 2 3 4 5
0.04W 0.1W 0.1W 0.14W 0.12W Wheel loads
CL-W
0.08W 0.2W 0.2W 0.28W 0.24W Axle loads
25 62.5 62.5 87.5 75 Wheel loads, kN
CL-625
50 125 125 175 150 Axle loads, kN
18 m
0.25 m
(Typ.)
The value P1 defines the level of the load train (CL, CL4 = only 4 axles, CL3 = only 3 axles)
The definitions ON, ON4 and ON3 select the variants for Ontario according Annex A14.4.
The definitions BC, BC4 and BC3 select the variants for Britisch Columbia. The definition MV
selects the maintenace vehicle according 3.8.11.
The value P2 defines the class of the load train (default 625).
If the value P3 is defined ( 9.0, 8.0 or 7.0 for class A, B or C and D) a constant lane load will
be defined with reduced load values of the load train without any dynamic allowance. For this
loading a braking load P8 of min(700,180+0.1*L) will be applied. So P8 should become zero
for secondary lanes.
The height of the wind attack area P7 is 3.0[ m] , the zentrifugal force is acting at P9=2.0[ m]
aove the lane.
The value P4 defines the fixed distance to the 4th load of the vehicle, it is 6.6 in general, but
up to 18 m in British Columbia.
Further wit Literals EP1 and EP2 the special load trains EPLL1 and EPLL2 may be selected.
The value P8 defines a factor on the braking load which is applied as 180+0.1*P3*L but not
larger than 700 kN distributed along the total length. P8 0.0 deactivates this type of loading.
• IRC AA - Heavy Loading for certain areas or highways This is a bogie of two axles with a
total weight of P2 (40t). There is only one lane loaded with this train.
• IRC AAT - Loading for certain areas or highways This is a tracked vehicle with a total weight
of P2 (70t). A convoy with a minimum distance of 90 m is foreseen. There is only one lane
loaded with this train.
• IRC A - Loading for all permanent bridges This is a train with two trailers, repeatable with a
minimum distance of 18.5 m. The axle loading is P2 (27 kN), P3 (114 kN) and P4 (68 kN).
• IRC B - Loading for timber bridges and temporary structures This is a train with two trailers,
repeatable with a minimum distance of 18.5 m. The axle loading is P2 (16 kN), P3 (68 kN)
and P4 (41 kN).
• IRC nnR
All numbers refer to the hypothetical vehicles of appendix 1. An appended R will select the
wheeled trains, an appended S the four wheelers (only 3 to 24) and an appended T will
select the tracked vehicle:
3, 5, 9, 12, 18, 24, 30, 40, 50, 60, 70
• IRC PD for loading on footways is just a selectable type. The reduction according to clause
209.4 dependant from width and length has to be done by the user explicitly via parameter
P5.
For the longitudinal effects of the loading the clause 208 establishes reduction factor for multiple
lanes. This effect should be taken into account during the superposition of load cases.
The brake load with P8 (0.2 = 20 %) is defined as the fraction for the first load train, while the
other trains have only half of this value. P8 = 0.05 should be applied in the case of more than
two lanes to the other ones.
JRAT JRAL
[Link]
Loads for pedestrian lanes according to Tab. 2.2.4. have to be defined with type JRAL and
P3=0.0.
AS_A 160 A single axle, including the special case of a single wheel W 80 for local ef-
fects.
and further a system with 12 axles and a variant length to be used with influence lines.
ELEVATION
1.25 1.25 3.75 1.25 1.25 Varies 6.25 min. 1.25 1.25 5.0 1.25 1.25
0.6
0.2 2.0
If more than one lane is loaded, the second lane shall be loaded with a lane factor of 0.8 all
subsequent lanes with 0.4. For fatigue 70 percent of the A160 or M1600 load without uniform
load value P3 should be used.
The "heavy load platform" is given in AS 5100.7. It is a load with 16 axles with 200 kN each for
the HLP 320 and 250 kN for the HLP 400. To obtain the variable distance between 6 and 15 m
between two groups of 8 axles each of the old design code, speciy P2 -1.0.
The Swedish load trains have the special feature of variant distances between the loads which
are best treated using influence lines. However the short distance of 1.5m will be fixed to that
minimum value, as there are hardly cases to be expected not yielding the most unfavourable
value with that short distance.
Most train types have different load values for primary and secondary lanes. The parameter
P1 is used to specify an enumeration for the most important lane, the secondary lane etc.
P1 Lane index 1 2 3
P2 Axle load [ kN] 310 210 155
BRO5 2 x 3 axle load train of the Swedish BRO type 5
P1 Lane index 1 2 3
P2 Axle load [ kN] 250 170 0
P3 Line loading [ kN/m] 12 9 6
P4 explicit distance (minimum 6.0 m)
P5 area loading [ kN/m2 ] 4 3 2
The second train has a distance with minimum of 50 m, the individual loads with
a minimum of 10 m
BROF Fatigue relevant load train of the Swedish BRO
First there is an area load of type NFA, where the load value is dependant on the loaded length.
360
A = 2.3 + ≥ 4.0 − 0.002 · L kN/ m2 (3.25)
12 + L
While this equation is used with this loading, no action will be taken to check if a shorter loaded
length has a more severe effect.
The lane factors according paragraph 4.2 of the document can be specified with the global load
factor PFAC.
The second load train NFB describes the load trains Bc and Bt, applicable for local checks.
According to the document only one vehicle should be loaded with braking loads, this will be
the first vehicle.
The third load train NFC describes exceptional heay load trains.
Type of loadtrain P1 P2
P.E.B. Leclerc (1999) L 96
CEREMA Semi-Remorque (2016) R 72
CEREMA Semi-Remorque (2016) R 94
CEREMA Semi-Remorque (2016) R 120
CEREMA Grue Automotive (2016) G 72
CEREMA Grue Automotive (2016) G 94
CEREMA Grue Automotive (2016) G 108
The type LM71 is considered for a train within a excellent maintained track. If that is not the
case, it is possible to select with type UICB a variant with impact factors for the standard
maintenance.
Loadings are defined with a possible eccentricity of 1/18 of the gauge according to EC/DIN-
Fachbericht. The direction of this eccentricity is selected according to the centrifugal force. An
explicit definition with YEX 0 deactivates this feature allowing to specify explicit values in COPY
YEX.
For a sleeper distance up to 0.8 m each load is converted to 3 single loads equal to 0.25 · PA,
0.50 · PA and 0.25 · PA at distances of P1 from each other. A value greater 0.8 will yield a
constant block load as follows:
Figure 3.35: Load train UIC = LM 71 = UIC 71 = RU (BS5400) - Constant Block Load
Brake loading: P8 = 20.0 kN/m (constant brake load, but in total less than 6000 kN)
(-33.0 kN/m for traction, but in total less than 1000 kN)
Centre of mass: P9 = 1.8 m
Aerodynamic loading due to trains passing and derailing loading are not available.
Hint
load trains SW will act also if their contributions have favourable effects.
Load values P2 to P3 are informative values only. These load trains will be applied as con-
stant block load in general, but if you specify P4 = 0.0, all individual axles (up to 96!) will be
generated. The speed of the trains will default to the values specified in the EC.
Defaults: P1 = 200 kN
P2 = 50 kN/m
P3 = 25 kN/m
SIMULATED
VEHICLE COUPLER
LOCOMOTIVE
300 kN 300 kN 300 kN 300 kN 360 kN
SIMULATED
VEHICLE LOCOMOTIVE
12 to 20 12 to 20 12 to 20 12 to 20
Repeat number (Varies) (Varies) (Varies) (Varies)
of vehicles as
necessary
Figure 3.39: Load train ASRT = Railway Loading 300LA according AS 5100
The load train 300LA of the AS 5100 consists of a locomotive and as many vehicles as nec-
essary, each of them with a variant length between 12 and 20 m. Unfortunately even with
influence line it is hardly possible to deal with an infinite number of variables, thus the load train
has always a finite number of vehicles.
For more than tow tracks, the loading in the third track is applied with a factor of 0.85, the fourth
track with 0.70 and all other with 0.60.
For this load train different coefficients for dynamic allowances are specified, depending on
the type of the rail fixing. (Ballast or direct fixing). These information is taken from the LANE
definition.
MLC Load trains of the STANAG are given in the sixth edition in appendix A. There are the
classes P1 = 4, 8, 12, 16, 20, 24, 30 up to 150. For every class we have available:
A freely defined loading by the user comprises an arbitrary number of loads defined relative to
each other. The first load determines the load reference point (x=0.0). Each further loading
is defined with a distance to the previous [Link] distance of the point load to the last load
point may be either constant or variable with a minimum value. In the latter case and additional
minimum distance to the load reference point may be defined with DPOS. For the evaluation
in ELLA two additional values DMIN and DMAX may be specified for special cases describing
an influence region before and behind the load.
For the evaluation of the load trains in ELLA symmetric load trains are expected. Thus, any
definition of Y should be complemented with a second definition at the same position (A 0) with
a negative Y-value. Thus, a second value of Y2 is no longer supported. An eccentricity may be
specified with YEX with the load train. NW allows to specify directly a number of wheels with
a distance B for that load element. They loads with the same load value will then add up to a
total axle distributing the loading on all wheels of that axle.
If NW = 0 the load values will be converted to a distributed load based on the width of the load
train WIDT. (120kN distributed on 3m yield 40kN/m) and will be then clipped at the boundary
of the the current lane width. But there are also cases (KEL of HA in BS), where the load is
distributed over the lane width, which is controlled via PHI.
A definition of BW and LW will allow to create also area loadings for every wheel.
x > DPOS(4)
A1
A2 A3 A4 A5
Figure 3.41: Freely defined load train
The loading has a vertical component P, a longitudinal component PB and two components in
transverse direction. The load may be defined either as single point load, as a pair of wheels
(Definition of B < WIDT) or as distributed load (Definition of BW only or B equal to WIDT) The
true dimension of the wheel contact area is only used for COPY within SOFiLOAD, ELLA uses
only the resultant forces for the evaluation of the influence lines.
Positive loadings of PB are braking loads acting in travelling direction, negative values are
intended as accelerating loads in the reverse direction. Loads are applied at the height of the
lane.
The centrifugal loading acts in the height hs, the constant horizontal force in the height zw,
thus creating torsional moments. If the load has a width B, this moment will be applied as a
pair of forces on the lane surface, a torsional moment along the lane axis will be generated
otherwise.
PW hw
PFSv2/gR
hs
As an alternate method, it is possible to associate a load train to a user defined structural model
of the load train. In that case CONT specifies the number of a node of the structural system,
which is located at the current position of the load train. If CONT is specified with a positive
number, the node will get the deformations of the structure at this point. If defined negative,
the node will get only the offset of the current position for the visualization however. This has
to be done for all nodes of the train model having no direct contact to the bridge.
A freely defined loading by the user comprises an arbitrary number of loads defined relative to
each other. The general parameters and remarks are given at record TRPL .
All load values will be treated as line loads in general. If NW = 0 the load values will be
converted to a distributed load based on the width of the load train WIDT. (12kN/m distributed
on 3m yield 4kN/m2) and will be then clipped at the boundary of the the current lane width. For
standard load trains, the definition of the code is often in that format. But there are also cases,
where the load is distributed over the lane width, which is controlled via PHI.
For NW>0 one or more line loads of the wheels or tracks will be created with the distance B. If
BW is defined, an area load may be created instead of the line load.
If only HW is specified a value for PW will be evaluated from the wind load given with the load
train wind definition.
Input values for multi-body-dynamics of line loads TRBL for record TREX will not be used
currently.
The input of the loading occurs in the form of saving and evaluation cases. Each saving case
includes several evaluation cases CASE and optional superposition rules of those evaluation
cases COMB.
The value given here at LCB is added to the number of the case defined with the record CALC.
Example:
CALC defines LMAX 12 and LMIN 11
1. Evaluation of the imposed loads in the construction stage (crane etc.)
SAVE 200
The load cases 211 and 212 are the result.
2. Evaluation of the imposed loads in the final stage
SAVE 250
The load case 261 and 262 are the result.
The action type and the designation of the resulting load cases may be specified explicitly or
be derived from the data saved with the load cases of the used load trains automatically.
NO Evaluation Case − 1
GRP Group of Loading LT ALL
The input of the loading within a saving case occurs in the form of evaluation cases. Each case
gets an identification number and contains the assignement of load trains to individual lanes.
With the input of GRP one may specify literals for load groups. This definition will then select
factors according EN 1991-2 or DIN Fachbericht.
The loading is generally done within the lanes only. But the load train may be smaller or wider
than the lane or not all lanes will be loaded. Then we have the following rules:
• A loading with group GR4 will apply the residual load on the total traffic area.
With this record the loading within a lane for an evaluation case is defined. You have to specify
at TRAI the load case number of the defined load train and/or the residual loading at P. Load-
ings outside of the defined lane width will not be applied. If only a residual loading is defined,
the loading is applied to the right and/or the left of the lane axis without any consideration of
other zeros in the transverse direction.
To specify the superposition of the loading for primary and secondary lanes there are now the
following possibilities:
The record CASE must be followed by the corresponding POSL records. Lanes without load
train remain unloaded.
The program places automatically the loading at the most unfavourable position inside the
above region. To achieve this, all selected points of the load train are placed at up to four
extremes of the influence line. The maximum value of this initial position is used for a detailed
iteration. However, the region can be input for special cases. This is the case, if:
In all other cases it is reasonable not to input any values for SMIN, SMAX, or DEL. Only a
single intervall may be used.
Load trains will be positioned at parallel locations in general. If independently moving trains
are required, this needs separate lanes in every case and either separate geometries (eg.
AXA.1 and AXB.1) or a special SYNC OFF definition. Variable load distances will then be
applied as in the main lane. It should be noted that for the case of favorable contributions of
the synchronized load trains, results may be on the unsafe side, especially for a classical load
position analysis.
For a pure beam system, a transverse shift of the loading during the evaluation might be
applicable, for FE structures such an eccentric loading will be produce reasonable results only
in rare cases. Thus it is strongly recommended to define an own lane for each eccentric
loading. For wide heavy load trains YEX may be used to shift the position of the load train, that
only parts of the loading which are within the lane limits will be applied, but then a synchronized
loading has to be assigned in the adjacent lanes. If there is only a single lane 0, YEX is still
used for historical reasons as a global eccentricity for the total load train.
ELLA tries to find the maximum forces. The influence of wind on the vehicles is considered
therefore with a simplified manner: The wind acts always in the most unfavorable direction. The
wind on the structure itself is not handled and detailed effects like partial loading (consideration
of lee), wind on partial vehicles, coherence or unified wind directions at all parts of the structure
are also not considered. However, in some case the wind may be defined with a fixed sign.
Thus one may describe a few of these effects.
In the case of a bow of the lane the masses of the load train, usually calculated from the
vertical loads, are used together with the velocity and the curvature of the lane in the plan view
for the calculation of the horizontal loads. Depending on the type of the load train, reductions
or simplified rules may apply.
For the eccentricity there are the following effects, which might be influenced by the EXCE
value:
• Any loading may be eccentric within a given range within the lane. Especially for railway
loadings this is a very important option to select the most unfavorable eccentricity.
• A loading may be subdivided in a right and left part based on the torsional effect. For some
evaluations (e.g. SLW) a single wheel may become effective in those cases.
The envelope of extreme values is calculated after the end of evaluation of all the evaluating
cases. Without COMB definitions each evaluation case is mutually exclusive. However for
special cases it is possible to define intermediate results with the COMB record which are then
combined with variant superposition conditions.
Each intermediate result has a superposition rule assigned and is created from up to ten basic
evaluation cases with factors. If more than ten cases should be combined, additional records
with COMB F may be used for that purpose.
For the common case that a combination of multiple lanes is required with either different type
of load trains or with different global factors, an internal sorting may be performed by specifying
a range of case numbers "cmin:cmax". In that case only loadings are considered not having
loads in any already loaded lane.
Examples:
COMB A0 F1 1.0 F2 0.6 F3 0.4 combines any 3 cases such that the
most unfavourable will be taken with
100 %, the second with 60 % and the
least unfavourable with 40 %
COMB A0 L1 1 1.0 2:9 0.6 2:9 0.4 combines case 1 to 100 % with 60 %
of the largest of the cases 2 to 9
and 40 % of highest remaining case.
Example:
A railway bridge with 3 tracks and 2 spans with 8.00 and 12.00 m length has to be ana-
lyzed. According to DS 804 are to be examined:
The following input is possible for the definition and the superposition of the four combinations
based on nine evaluation cases:
4 Output Description
The output of an influence line evaluation can become very extensive. Therefore the use of the
various ECHO options may be very helpful.
With ECHO LINE FULL one obtains a list of the neighbouring nodes at the left-hand side and
to the right-hand side of the lane.
The table of the influence lines contains in its headline the application point of the unit load and
the scalar variable. The columns are as follows:
Influence lines are plotted with a predetermined scale in the longitudinal direction and with a
scale for the influence line values, which is determined by the height of the drawing bands.
Maximum and minimum values are printed at the plot.
For the graphics of the transverse influence line, the sum of the influence values of the the two
wheels of the axle (Pv) and the integrals of the positive (p+) and negative (p-) influence values
are also given. The values r+ and r- are the integrals of the residual areas outside the traffic
lane.
For loads with variable length the output of a second line with the maximum and the minimum
length and a possible subdivision length value in the column LDEL follows.
The dynamic factors and the lane widths are output in an additional table. The parameters
therein mean:
When one of these two options is active, a line which contains the identification of the position
is printed for each position with corresponding influence lines.
As long as EVAL is active, the position of the load train as well as the possible convoy distances
or the variable lengths are output in a succeeding table.
If EVAL is set to FULL, the regions with different signs are printed then for each evaluation
case and for each lane together with the applied dynamic factors PHI1/PHI2 for unfavourably/-
favourably acting loading and the positions of extreme values.
If RES is active, the extreme values of the internal forces or other parameters are output in
an additional table. The first columns contain the number, with which the load case has to be
stored, and a text for the designation the evaluation rule.