Viljoen Investigation 2004 PDF
Viljoen Investigation 2004 PDF
Viljoen Investigation 2004 PDF
P. De K. Viljoen
December 2004
Stellenbosch University http://scholar.sun.ac.za/
I, the undersigned, hereby declare that the work contained in this assignment
was my own original work and that I have not previously in its entirety or in part
submitted it at any university for a degree.
Signature:
Date: .
Stellenbosch University http://scholar.sun.ac.za/
SYNOPSIS
The purpose o f this project was to study the deformations o f the top flange and web of
a girder panel resulting from loads, which have been imposed by an overhead
travelling crane. This was achieved by designing a crane girder that represents
dimensional ratios occurring in practice.
The first part o f this project attempts to determine the properties o f the crane girder.
The crane girder was built from drawings in the workshop of the University of
Stellenbosch’s Civil-Department. Positions were identified where the strains were to
be measured on the crane girder. The crane girder was subjected to loads according to
SABS 0160 - 1989 and measurements were taken. The same beam was also modelled
with finite elements. The numerical model was subjected to the same loads as the
experimental crane girder.
Comparisons were then made between the results o f the experimental investigation
and those of the numerical model. Good comparisons were achieved between the
results and the numerical model was assumed to be correct.
Other students could now use this model for investigating local stresses and strain
effects that might cause fatigue and other in-service problems o f electric overhead
travelling cranes support structures in practice.
Stellenbosch University http://scholar.sun.ac.za/
SAMEVATTING
Die doel van hierdie projek was om die vervormings van die boonste flense en web
van ’n kraanbalkpaneel te ondersoek, as gevolg van laste wat onderworpe was aan ’n
oorhoofse kraanbalk. Dit was bereik deur ‘n balk te ontwerp wat dimensionele
verhoudings in die praktyk verteenwoordig.
In die eerste gedeelte van die projek word die eienskappe van die kraanbalk bepaal.
Die kraanbalk was vanaf tekeninge in die werkswinkel van die Universiteit van
Stellenbosch se Siviele Departement gebou. Posisies was geidentifiseer waar die
vervormings op die kraanbalk gemeet sou word. Die kraanbalk was onderwerp aan
laste volgens SABS 0160 - 1989 en meetings was geneem. Dieselfde balk was ook
gemodelleer met eindige elemente. Die numeriese model was aan dieselfde laste as
die eksperimentele balk onderwerp.
Vergelykings was toe verkry tussen the resultate van die eksperimentele ondersoek en
die numeriese model. Goeie vergelykings tussen die resultate was behaal en die
numeriese model word as korrek aanvaar.
A nder studente kan nou hierdie model gebruik vir die ondersoek na lokale spannings
en vervormingseffekte wat moontlik vermoeidheid en ander in-diens probleme van
elastiese oorhoofse kraan ondersteunende strukture in the praktyk kan veroorsaak.
Stellenbosch University http://scholar.sun.ac.za/
CONTENTS
LIST OF TABLES IV
LIST OF FIGURES V
LIST OF PHOTOS VII
LIST OF SYMBOLS VIII
i
Stellenbosch University http://scholar.sun.ac.za/
LIST OF TABLES
TABLE 2.1: TYPICAL DIMENSIONAL RATIOS FROM CRANE GIRDERS AT SALDANHA STEEL
PLANT 2-6
iv
Stellenbosch University http://scholar.sun.ac.za/
LIST OF FIGURES
FIGURE 2.1: PLAN OF CRANE INDICATING TRANSVERSE WHEEL LOAD DIRECTIONS 2-4
FIGURE 2.2: MATERIAL PROPERTIES OF CRANE GIRDER STEEL 2-7
FIGURE 2.3: STRESS STRAIN CURVE OF ELASTOMERIC PAD 2-8
FIGURE 2.4: GENERAL CRANE CRAB AND CRANE BRIDGE LAYOUT 2-10
FIGURE 2.5: GENERAL ASSEMBLY 2-11
FIGURE 2.6: GIRDER CROSS SECTION 2-14
FIGURE 2.7: RAIL SECTION 2-15
FIGURE 2.8: COMBINED SECTION USED FOR VERTICAL DEFLECTION 2-16
FIGURE 2.9: SECTION USED FOR LATERAL HORIZONTAL DEFLECTION 2-18
FIGURE 2.10: EXPECTED STRESS DISTRIBUTION OVER DEPTH OF GIRDER AT MIDSPAN 2-20
V
Stellenbosch University http://scholar.sun.ac.za/
FIGURE 5.1: VON MISES STRESSES-VERTICAL LOAD ONLY OVER DEPTH OF BEAM AT
MIDSPAN 5-3
FIGURE 5.2: VON MISES STRESSES-VERTICAL AND MISALIGNMENT LOADS OVER
DEPTH OF BEAM AT MIDSPAN 5-4
FIGURE 5.3: VON MISES STRESSES-VERTICAL LOADS ONLY AT FLANGES OF BEAM AT
MIDSPAN 5-5
vi
Stellenbosch University http://scholar.sun.ac.za/
LIST OF PHOTOS
vii
Stellenbosch University http://scholar.sun.ac.za/
LIST OF SYMBOLS
X Factor
P2 Factor
Px Factor
C0 2 Factor
viii
Stellenbosch University http://scholar.sun.ac.za/
ix
Stellenbosch University http://scholar.sun.ac.za/
x
Stellenbosch University http://scholar.sun.ac.za/
W8 Same as W 5
W9 Factored W3
xi
Stellenbosch University http://scholar.sun.ac.za/
1-1
CHAPTER 1 - INTRODUCTION
1.1 GENERAL
The purpose o f this study was to establish and verify a finite element model for
current and future research.
1-2
According to SABS 0160-1989[1\ the following loads were generally applied to the
overhead travelling crane rail:
[l] SABS 0160 - 1989, South African Bureau of Standards, Private bag X191, Pretoria.
1-3
This study required the application o f analytical theory and numerical techniques, i.e.
theory o f mechanics and finite element method and experimental verification in order
to determine the following:
1 Web deformations and forces (strains and stresses) developing in a crane girder
due to concentrated vertical and horizontal crane wheel loads
3 The determination of the deformations and forces in (1) and (2) with analytical
numeric model, i.e. finite element method
2-1
Constraints:
The limitations o f the laboratory o f the University o f Stellenbosch Civil Engineering
Department and the availability of material placed restraints on the design. These
included the following:
Rail Size: The rail size used in other studies conducted at the University at
the stage o f testing was 30 kg/m. It was decided that the same rail size had to be used
since the material was available. The rail size has a direct influence on the top flange
width.
Design Loads: The maximum load that could be measured was 20 tonnes,
since load cells with a capacity exceeding 20 tonnes were not available. The design
loads will be discussed in more detail under Design Loads in Chapter 2.1.1.
Maximum Span: The test support structure could support a beam with a
maximum span o f 11.0 meters. It could have been possible to have a crane girder
exceeding this length, but not without designing a bigger support structure.
Wheel Size: A crane wheel with a diameter o f 265 mm was available. It was
decided that the same wheel should be used. See chapter on experimental setup.
Type o f Rail Pad and Clips: As for rail size, type and size of pad, as well as clips, it
was decided to use those that were available. The type o f rail pad used was the
Gantrex ® MK6 Rail Pad, while the clips used were o f the Stelcam ® 13 Boltable
Clip type.
2-2
Typical Dimensional Ratios: A crane girder that could represent crane girders used in
practice, had to be designed. Typical dimensional ratios (depth to thickness o f web,
top flange to bottom flange, etc.) were gained from an investigation into crane girders
in practice in South Africa. These dimensional ratios were presented in Appendix A.
M aterial Properties: The steel used in South Africa was normally mild steel with
yield stress of 300 MPa. Testing o f the materials was done on a Material Testing
M achine. The material properties were summarised in Chapter 2.1.3 - Material
Properties.
A lex Perez-Winkler[2] also conducted testing on the pad material and clips during
previous studies.
SABS 0160-1989[l1 was used to determine the design loads that would be used in both
the experimental and numerical models. Paragraph 5.7 - Loads due to Overhead
Cranes takes the designer through the following design steps:
The classification of the crane was done in accordance with SABS 0160-1989[1]
Paragraph 5.7.2 - Classification o f EOTCs. The different classes were numbered from
1 to 4, where 1 has the lowest service classification to 4 having the highest service
classification. It was decided, for the purpose o f this research, that a crane with a
classification o f 4 should be used. This resulted in higher transverse loads.
For the determination o f the test loads reference to SABS 0160-1989[1] Paragraph
5.7.3 - Vertical Loads were made. This paragraph in SABS indicates that the owner
m ust specify the vertical wheel load. In the absence of an owner, it was decided to use
111 SABS 0160 - 1989, South African Bureau o f Standards, Private Bag X191, Pretoria.
(21 An Investigation of Overhead Crane Wheel/Rail/Girder Interaction, A.R. Perez Winkler, 2003.
2-3
SASCH[3] to determine the allowable crane wheel forces to be used as the test loads.
SASCH uses Crane Class, rail size and wheel diameter to determine the maximum
load for each wheel. With all o f the above parameters fixed (from previous studies),
the maximum wheel load was calculated.
Also, according to SABS 0160-1989[2]; Paragraph 5.7.3 - Vertical Loads, the impact
factor for Class 4 cranes was equal to 1.3.
W = Cdlf
where
W = Vertical wheel load
C = 0.29 (for class 4 crane with rail size o f 30 kg/m)
D = 265 mm
If = 1.3
thus
W = lOOkN
131 SASCH, South African Steel Construction Handbook, Third Edition, 1997.
2-4
For a Class 4 crane, the load scale factor for misalignment o f the crane wheels o f the
rail and for skewing of the crane was 0.20. The direction these loads were imposed in
can be seen in Figure 2-1: Plan of crane showing direction o f transverse wheel loads.
□ □ □
□
□ □ □ □ -
M ISALIG N M EN T
□ □— - -— □ □
□ □
□
□
S K E W IN G
where:
P = Transverse wheel load
W = Vertical wheel load
Xj = 0.2 (Scale factor for misalignment and skewing)
Thus
P = 20 kN
2-5
2-6
Typical EOT crane girder dimensions from the Saldanha steel plant were used for the
preliminary sizing o f the experimental crane girder. Since this plant has more than
forty EOT cranes, a good correlation between the typical relationships of the
following could be drawn.
i - Height o f web to thickness o f web
ii - Top flange width to thickness of top flange
iii - Bottom flange width to thickness o f bottom flange
iv - Span to spacing o f web stiffeners
v - Span to height o f web
Plans from the Saldanha steel plant were studied and the following information was
recorded. See Appendix A for the complete table. As can be seen in the table in
Appendix A, cranes with surge plates were also included in the calculation o f these
ratios. The averages were affected in a minor way than when the surge plates were not
used in the calculation o f the averages. This is due to the small number of cranes with
surge plates in the sample investigated.
These ratios were used to determine the preliminary crane girder dimensions. As can
be seen in Appendix C, not only these dimensional ratios were used in the calculation
o f the final element sizes.
2-7
The steel used in the manufacturing of the crane girder was tested with the material
testing machine at the University o f Stellenbosch, and the results were shown in the
figure below. The same properties were used in the numerical model. See Chapter 4 -
Numerical Model. Only the elastic properties were identified, since all the girder tests
had to take place in the elastic zone o f the steel.
Strain (mm/mm)
Elastic:
M odules o f Elasticity 203.5 GPa
Poison ratio: 0.3 (This ratio was not determined, but was used in the
numerical model.)
Yield Stress 310.5 MPa
The test procedures for determining the material properties were given in Appendix B.
2-8
Elastic:
Modulus o f Elasticity: 20 MPa
Poisson ratio: 0 (This ratio was not determined, but was used in the
numerical model. A more accurate value would have
been 0.5. This would only complicate the model, and no
more accurate results (in the region o f inspection) was
expected.)
Plastic:
1 0 .8 0.04
2 2 .1 0.09
3 2 2 .8 0.39
Strain (mm/mm)
2-9
2.1.3.3 Rail
Material properties relating to the rail were also determined by Alex Perez-Winkler,
and were as follows:
Elastic:
Modules of elasticity: 200 GPa
Poisson ratio: 0.3 (This ratio was not determined, but was used in the
numerical model.)
Yield Stress: 360 MPa
2-10
2.2 DESIGN
W ith the dimensional ratios as discussed in the previous section, and with other
criteria, for example classification, a preliminary crane girder was designed. The
com plete design was shown in Appendix C.
— CRANE G IR D E R W ITH R A IL
y— END C A R RIDGE
y / y -------- C R A N E B R ID G E
— 1 1
■ ■ ■
_ 1 1
" \
'- - C R A B
Class o f Crane: A Class 4 Crane was used. See Chapter 2.1.1 Design Loads.
Rail size: A 30-kg/m rail was used. See Chapter 2.1. General Design
Approach.
Wheel diameter: A 265 mm diameter wheel was used. See Chapter 2.1. General
Design Approach.
Wheel spacing: A wheel spacing o f 900 mm was used. See Chapter 2.2.2.
General Assembly
2-11
The rail was connected to the top flange of the crane girder through rail clips. A
bearing pad (Gantrex ® MK 6 ) was installed between the rail and crane girder. The
purpose o f the pad was to:
- Distribute the wheel load over a large surface area;
Eliminate load concentrations and the resulting fatigue stresses;
- Compensate for the uneven surface between the rail and its
support;
- Reduce impact, vibration and noise;
- Eliminate fretting corrosion (wear) o f the support surface under
the rail.
The clips used were Stelcam ® 13 Boltable Clips and the purpose o f the rail clips was
to,
- Connect and align the rail to the crane girder;
Give resistance to lateral force.
The clips were spaced according the clip supplier, at 720 mm for the internal clips and
620 mm for the end clips. The required torque to the bolts connecting the clips to the
top flange of the girder was also according to the supplier’s specifications and was
equal to 280 N.m.
2-12
The following section gives a summary o f the general layout and sizes o f members.
The calculation of the crane girder sectional properties was shown in Appendix C.
The length o f the girder was fixed at 4.5 m (4500 mm). This length was less than the
m axim um allowable length of the test structure o f the support system.
The width o f the top flange (bft) was 300 mm. The following criteria was used in
calculating the width of the top flange o f the plate girder:
- Minimum width for the 30 kg/m rail and Stelcam ® 13 boltable
clips.
The thickness o f the top flange (tft) was 20 mm. The thickness was calculated using
the following criteria:
- Dimensional ratios (see Chapter 2.1.2 Typical dimensional
ratios).
- Flange classification criteria (class 3 flange was used).
Serviceability criteria (limits to horizontal deflection caused by
misalignment using the top flange and rail only for resistance).
The w idth o f the bottom flange (bfb) was 200 mm. The following criteria were used in
calculating the width of the bottom flange.
- The minimum support width o f crane column had to be greater
or equal to h alf the top flange width.
- The correct second moment o f the area o f the bottom flange, as
determined by the maximum vertical moment, to satisfy the
required second moment o f the girder had to be used.
2-13
The thickness o f the bottom flange (tfb) was 10 mm. The thickness was calculated
using the following criteria:
- Dimensional ratios (see Chapter 2.1.2 Typical dimensional
ratios).
- Flange classification criteria (Class 3 flange was used).
2.2.3A W eb
The depth o f the web (hw) was calculated to be 450 mm. The following criteria was
used in calculated the depth of the web.
- Vertical deflection limit.
The thickness o f the web (tw) was 10 mm. The following criteria were used in
calculating the thickness o f the web.
Dimensional ratios (see Chapter 2.1.2 Typical dimensional
ratios).
Slenderness limit (web classification)
The thickness o f the intermediate and bearing stiffeners was 10 mm. SABS 0162 - 1:
1993[4], par 13.4.11 was used to calculate sizes and spacing o f the intermediate
stiffeners. The spacing o f the stiffeners was 900 mm internally and 815 mm for the
end stiffeners.
[4] SABS 0162 - 1: 1993, South African Bureau o f Standards, Private Bag X I91, Pretoria.
2-14
2.2.3.6 Summary
y
y
FIGURE 2.6: GIRDER CROSS SECTION
Section properties:
A = 12,50 x lO 6 mm 2
yc = 311,00 mm
I c ir d e r ,x x = 1650,00 x 106 mm 4
lG ird er,y y , T op Flange — 45,00 X 10 mm
2-15
As discussed in Chapter 2.1 General design approach, a 30 kg/m rail was used. The
sectional properties of the 30 kg/m rail were shown in Figure 2.7: Rail section.
109.54
Section properties:
A = 3,85 x 106 mm 2
yc = 53,13 mm
I R ail,xx
= 6,27 x 106 mm 4
4
lR a il,y y = 1,57 x 106 mm
2-16
No slip between the rail and the test girder was expected to take place at the test loads.
The moment o f inertia of the combined section:
(
L =
2-17
where: P = 100 kN
a = 1800 mm (Distance from support to applied load)
I = 4500 mm (Span)
E = 203.5 GPa (Modules o f Elasticity)
I, = 592.6 x 106 mm 4 (Moment of Inertia o f the combined
section)
ISlip = 447.3 x 106 mm 4 (Moment o f Inertia o f the “slipped” section)
a a/1
This range was used for the calculation o f the deflection meter instrumentation in the
measurement o f the deflections during the experimental setup. See Chapter 3 -
Experimental Investigation.
2-18
It could be assumed at this stage that only the top flange o f the crane girder and the
rail would provide stiffness against the lateral horizontal deflection.
where: P = 20 kN
a = 1800 mm
I = 4500 mm
E = 203,5 GPa (Modules o f elasticity)
lyy ,,
= IrGirder
j yy Top
t Flange
fi Rail,yy 5 x 106
= 46,57
a a/1
2-19
The stresses in the flanges were calculated next. These forces and force effects can be
summarised as the following:
(1) Vertical wheel loads
(2) Transverse wheel loads due to Misalignment
With these and the maximum moment due to the load in position 1 (refer to Chapter
3.2 Test Loads for the definition o f load position 1) calculated to be equal to 180
kN.m known, the stresses in the top and the bottom flange could be calculated using
the formulas below:
^ , - = ” = -110,35 MPa
’
_ M Jh y) _
Top T
The stresses due to misalignment were calculated, using only the top flange o f the
crane girder and the rail. The maximum moment due to the lateral horizontal load
(misalignment) in load position 1 was equal to 36 kN.m.
M wy
a Left = a Top "* ~ 151,39MPa
yy
M y
<JRight = 11Top = -80.5 IM Pa
yy
2-20
The following results were used in determining the required type and size o f the
deflection meters (LVDTs1) as well as the required capacity o f the strain gauges.
These results were obtained from the vertical load case in load position 1.
Maximum vertical deflection: 2.97 mm
Maximum compression stress in top flange: 35.44 MPa
Maximum tensile stress in bottom flange: -110.35 MPa
------------------------------------------------------------------------------------------------ 4 6 0 -
/
- - 440
/
/
f
.
■ ' - •
/ ■
E
a
a> . . . . ...................................................... ......................................................../ OflA ... , .................................
OQ /
. . . -
4 -
o / 240 :
■C
/ -0 0
a
a> „ '■ ' ” ■ ' ” .
a S 160
/
.
/ io n
/
/ o n
/ 80
/ 40
/
y
-----------T-------------------- 1---------------------1------ -------------- --------------------
\
- 1 5 0 .0 0 - 1 2 0 .0 0 - 9 0 .0 0 - 6 0 .0 0 - 3 0 .0 0 0 .0 0 3 0 .0 0 6 0 .0 0 9 0 .0 0 1 2 0 .0 0 1 5 0 .0 0
S tress (MPa)
The deflection and stresses in the top flange due to the lateral horizontal load case o f
misalignment were:
Maximum horizontal deflection: 7.56 mm
Maximum compression stress in top flange: 151.39 MPa
M inimum tension stress in top flange: -80.51 MPa
3-1
3.1 INTRODUCTION
The purpose for the experimental investigation was to verify the Finite Element
Model With the selected load cases and combinations sufficient results were available
to verify the Finite Element Model.
Two identical test beams were constructed to eliminate construction defects. By using
the averages between the two beams, a more accurate representation of a beam
without defects could be achieved.
The design drawings o f the test beam were presented in Appendix D - Design
drawings fo r the construction o f the experimental crane girder.
3-2
The test beams were assembled by placing the bearing pad on the girder and then
placing the 30-kg/m rail on top of the bearing pad. The clips were positioned and
fixed as indicated on the photos below. The bolts were tensioned to a torque of 280
N.m. The following photos were from the web page of Gantrex ®,
www.gantrex.co.com. For the spacing o f the clips, refer to Chapter 2.2.2: General
assembly.
i . Position clip on bolt or stud, 2. Position self-locking cam. Install and lightly tighten nut.
3-3
Position 1
In the first position, the maximum moment affect was measured:
• Vertical loads at midspan
• Vertical loads at midspan with horizontal loads representing
misalignment
• Vertical loads at midspan with horizontal loads representing
skewing
Figure 3.1: Position 1 - Layout and force effects, indicates the layout of the loads in
position 1 and also the load effects.
« K UI 1( 0 Ml
+500
20 kN 20 kN
36 kILra
Horizontal l I Y // f/ T 77 \
Y/////T7\
^ y - t t t T T T T I 11 111 n T T r m r . ---
(Misalignment) , 1800 .900 , 1800
-20 kN
20 kN „ 1800
3-4
Position 2
In the second position, the high shear effect was measured:
• Vertical loads at endspan
• Vertical loads at endspan with horizontal loads representing
misalignment
• Vertical loads at endspan with horizontal loads representing
skewing
100 kN 100 t)
+500
20 kN 2C kN
m “ j S ___ 14.4 kNjn
Horizontal 1 ^^r-r-n T T T T I iTTTTTTTrt-^-
+500
f
Horizontal
(Skewing) 3150
3-5
All the equipment used was available at the University of Stellenbosch, Civil
Engineering Department. The following figure indicates the general layout o f the
experimental setup, with reference to the apparatus and loading equipment.
(The above drawing can also be viewed on the attached CD, under the file name
EXPERIMENTAL SETUP.dwg)
A support frame that can be used for the purpose o f testing different systems was in
place in the laboratory. This consists of rails, beams and columns. The support frame
can be tied back to the main building.
3-6
Two stub columns were positioned 4500 mm apart. On top o f these columns glacier
bearings were used to represent pinned support boundary conditions.
Columns were then positioned on either side o f the stub columns. These columns
were fixed back to the main building and were used for lateral support of the top
flange o f the test girder at the support points.
Columns were also fixed on either side of the sway frame (see Section 3.3.1.2, Sway
Frame). These columns were fixed back to the main building, and were used to apply
horizontal forces at the position o f the vertical loads.
Even though the support frame was fixed back to the main building, no deflection
measurement equipment was fixed to any member o f the support frame.
3-7
The Sway Frame was used to make sure that the vertical load was always vertical,
independent of the possible rotation or sway o f the test beam. This needed to be used
since the theoretical design (Chapter 2) and the Numerical Model (Chapter 4) both
used a constant vertical force, and did not take rotation into account. The sway frame
also follows any horizontal deflection o f the beam, thus always applying the load
through the horizontal shear centre.
INDICATING SETUP
3-8
The load was transferred onto the test girder with a pump and load actuator system.
The pump with load actuator was connected to the sway frame. The load cell was, in
turn, fixed to the load actuator. A rod fixed to the wheel via a cross-and-rod system
(see Figure 3.5: Cross fo r distributing load to top o f beam) was connected through the
load actuator and load cell.
• For the vertical loads, the Enerpac 60 ton RCH 603 actuator with
the Enerpac Type P39 pump was used.
• For the horizontal loads, the Enerpac 20 ton RCH 202 actuator with
the Enerpac Type P228 pump was used.
3-9
3.3.2.1 Amplifier
For the amplification of the electronic signal received from the measurement
equipment, the Spiders with analogue/digital interface was used. The spiders come
with computer software that enables the user to take readings o f the change in the
electronic signal directly in the required units, for example, taking measurements from
load cells directly as kN. Once the test has been completed, the user can save the
information in electronic format for later use.
A total o f 64 amplified channels were available on the Spiders. All 64 were used.
A total o f four channels were used for the load cells - refer to chapter 3.3.2.2: Load
cells.
A total o f four channels were used for the LVDT’s - refer to chapter 3.3.2.3: LVDT.
A total o f 48 channels were used for the strain gauges and strain rosettes - refer to
Chapter 3.3.2.4: Strain gauges.
For the vertical loads, the HBM U2/20-ton load cells were used, and for the horizontal
loads the ULP/S 5-ton load cell was used.
3.3.2.3 LVDT
The positions o f the LVDTs were fixed where the maximum deflection, vertical and
horizontal, was expected. The positions can be seen in Figure 3.6: Positions o f LVDTs
on test beam.
3-11
Strain gauges measure the change in strain on the surface o f the member/structure
being subjected to loads.
The following questions were asked to determine the positions o f the strain gauges on
the test girder:
• Where will the maximum stresses occur in the girder?
• What will the stress distribution over the depth o f the beam look like?
• What will the stress distribution look like over the width o f the top and
bottom flange under the load cases for skewing and misalignment?
The positions o f the strain gauges were shown in Figure 3.7: Position o f strain
gauges.
3-13
The measured strains were converted to Von Mises Stresses. The following formulas
indicate the process of manipulation to convert from the strains (as measured) to the
Von Mises stresses.
— J te /
^ .. 1' ' k
1r
If angles 0i, 02 and 83 , together with the corresponding strains sei, S0 2 , and 803 , were
known from measurements, three simultaneous equations can be written.
£ e\ = £ x c o s 2 ^1 + s y s i n2 ^1 + Yxy s ^n ^1 c o s ^1
S 63 = e x C O s2 ^ 3 + e y S i n 2 ^ 3 + Y xy S*11 ^ 3 C O S ^ 3
3-14
By direct substitution into previous equations, it was found that, for this rosette;
The maximum normal strain was 81 ; the minimum was £2 . These were the principal
strains, and no shear strains were associated with them. The directions o f the normal
strains coincide with the directions o f the principal stresses. Thus, from Mohr’s circle,
the analytical expression for the principal strains is:
In m ost problems where strain rosettes were used, it was necessary to determine the
principal stresses at the point of strain measurement. In this problem, the normal stress
on the surface was zero, i.e. crz = 0. Therefore, this was a plane stress problem. The
principal stresses were:
cr, E I \
-—
\-v
r t a +V£>)
The Elastic constants E and v were determined earlier (see Chapter 2.1.3: Material
Properties. Von Mises stresses could be calculated from the principle stresses using
the following formula:
Where
( 1 ^
7 — y.
and
3-15
mm mm mm
3-16
With the strain gauges and rosettes and the use o f the amplifiers, the change in the
strain against the applied load (also continuously measured) could be plotted for each
strain gauge. The strains were converted to stresses and the Von Mises stresses were
calculated for comparison with the results from the finite element analysis.
Some o f the results from the experimental investigation can be seen on the following
pages. More results were available in table form, and can be seen on the attached
compact disc. (The file name on the disc was Von Mises Stresses.xls.)
The following two graphs were from the load in the first load position (representing
maximum moment). The first graph indicates the stress distribution over the depth of
the beam at midspan, while the second graph indicates the stress distribution over the
width o f the flanges.
3-19
The following results were the results from the second load position. As can be
expected, the stresses due to bending were much lower than the stresses in the
previous graphs for load position one.
4-1
4.1 INTRODUCTION
A numerical model o f the crane girder, pad and rail system was modelled using finite
elements. The program used was the commercially available ABAQUS[6].
The purpose o f this model was to serve other students in numerically analysing
different layouts and sizes of the crane girder. This numerical model needs to be
verified so that further research can be done with confidence.
[6] Abaqus, Version 5.8, Habbitt, Karlsson & Sorensen, Inc. Finite element analysis services (Pty) Ltd,
The Waverley, Mowbray 7700.
4-2
The same load cases as discussed in Chapter 3 for the experimental model were used
in the numerical model. The purpose was to compare the results from the different
models.
The load cases were normally applied with two or more steps. The first step was the
“assembly o f the model”. In this step all the parts were put together, and the load for
the clips (refer to Chapter 4.3: Model Description) were applied. In the second step,
the vertical load was applied. And in the third step (if applicable) the transverse loads
were applied.
4-3
4.3.1 GENERAL
Classical beam or plate theories were not suited for modelling the crane girder-pad-
rail interaction problem. For this reason, it was decided to model the whole system in
3D. The model consisted of a rail, elastomeric pad, and the crane girder. The clips
were omitted (a load representing the pressure exerted by the clips onto the rail were
used - this reduced the size and complexity o f the model considerably).
Also, instead o f modelling a wheel through which the loads were to be applied, only
point loads were applied onto the rail. It was believed that the difference in the
stresses on the top flange, when the loads were applied with the wheel, versus when it
was applied with a point load, was negligible. The reason for this assumption was due
to the path that the load travels (Rail to pad to top flange).
The model was constructed with 20-node solid hexagonal brick elements. These
elements had mid-side nodes and were based on quadratic shape functions. Most 3-D
structural problems that involve bending can be modelled very accurately with a
modest number o f elements that deform quadratically. Other quadratic elements, like
the 10-node tetrahedron and the 15-node wedge, give similar results but lead to larger
computational workloads.
These elements made up the different parts (crane girder, Gantrex pad and rail). These
parts were held together by constraint functions along the interfaces.
4-4
The crane girders, as used in the experimental investigation, were modelled using the
finite element method. These measurements were given in Appendix D - Design
Drawings fo r the Construction o f the Experimental Crane Girder. For the fillet welds,
a 10 mm chamfer was included in the steel layout. The drawing below indicates the
finite element mesh of the crane girder with web stiffeners.
The beam consisted o f a total o f eight (four on each size) intermediate stiffeners and
four (two on each side) bearing stiffeners. The model size can be summarised as
follows:
4-5
Between the rail and the girder was an elastomeric rail pad (Gantrex MK6). This pad
was also modelled to give a more accurate representation o f the real setup. The size of
the pad in the model was:
Thickness: 7 mm
Width: 98 mm
Length: 4500 mm
The finite element mesh for the elastomeric pad was shown in Figure 4.2: Gantrex
M K6 elastomeric rail pad mesh layout.
J-
FIGURE 4.2: GANTREX MK6 ELASTOMERIC RAIL PAD MESH LAYOUT
4-6
4.3.4 RAIL
It was also decided to model the rail to ensure a representative distribution of the point
loads through this member, The rail used in the model was a 30 kg/m rail and the
dimensions were as follows:
Length: 5000 mm
Size: See figure below.
4-7
The finite element mesh of the rail was shown in Figure 4.4: 30 kg/m rail mesh layout.
4-8
It was necessary to limit the size o f the model so that no unrealistic computational
times were needed to solve a model without compromising the accuracy o f the results.
The size o f a finite element model was usually defined in terms o f the total number of
degrees o f freedom. The size o f the model that was constructed had 3786 degrees of
freedom. This translated to about six hours CPU time for a machine with 733 MHz
clock speed and 1000 Mbytes Ram.
PROBLEM SIZE
4-9
The following figures indicate the Von Mises stresses over the flanges and depth of
the beam, as shown in Chapter 3: Experimental Investigation, but also included in this
section were figures o f the different loaded beams. See Figure 4.7 to Figure 4.9.
Firstly, the deflections determined during the numerical investigation in table form
are:
Horizontal Horizontal
Vertical Deflection
Deflection of Deflection of
at midspan
Top Flange Bottom Flange
mm mm mm
Vertical - Position 1 3.86 0.01 0.01
Misalignment - Position 1 3.97 -8.55 1.66
Skewing - Position 1 3.85 0.01 0.01
Vertical - Position 2 1.81 0.01 0.01
Misalignment - Position 2 1.80 -3.57 -2.90
Skewing - Position 2 1.79 -1.18 1.39
T A B L E 4.4: DEFLECTIOlV DETERMINED DlURING NUMERICAL
INVESTIGATION
More results than those shown below were available in table form, and can be viewed
on the attached compact disc under the file name Von Mises Stresses.xls.
4-10
Figure 4.5 and Figure 4.6 were the results from the load in position 1 (refer to chapter
3.2: Test loads and load positions). The first graph indicates the stress distribution
over the depth o f the beam at midspan, while the second graph indicates the stress
distribution over the width o f the flanges.
4-16
Figure 4.10 shows the results from the load in position 2 (refer to Chapter 3.2: Test
loads and load positions). As could be expected, the stresses due to bending were also
much lower than the stresses in the previous graphs for load position one.
5-1
5.1 INTRODUCTION
The numerical model has to be verified for future research. While the theoretical
investigation was used to size the test beam and determine expected results for the
experimental investigation, only the experimental investigation will be used to verify
the numerical model. By comparing the following results, the numerical model can be
assumed to be correct. The data to be compared were as follows:
5-2
For purposes o f clarity, the comparisons were presented in graph form. (More data
were available on the attached compact disc; refer to file Von Mises Stresses.xls).
Deviation between
Experimental Numerical Experimental and
Numerical
mm mm %
The following three graphs were from the load in the first position (refer to Chapter
3.2: Test loads and load positions). Figures 5.1 and 5.2 indicate the stress distribution
over the depth o f the beam at midspan, while Figure 5.3 indicates the stress
distribution over the width o f the flanges. These stresses were for both the
experimental and numerical models and it can be seen that the results o f the different
models compare well. Additional graphs and tables showing these results can be
viewed on the attached compact disc (refer to the file Von Mises Stresses.xls).
6-1
6.1 INTRODUCTION
The purpose o f this project was to design and verify a numerical model of a crane
girder for further research.
Warren'7J was currently doing another research project through which the fatigue
stresses due to eccentric loads were investigated. These stresses were determined
theoretically, but needed to be determined numerically as well. The model that was
created assisted in determining stresses at the required positions using the finite
element method.
It should however be noted that the type o f elements chosen and the node layout
worked well for the purpose o f the model. But it is the responsibility o f the researcher
using this model to make sure that it is also applicable for his/her requirements.
171 Juliet Warren, Fatigue research o f Crane Girders when subjected to Loads imposed by Overhead
Travelling Cranes.
6-2
The numerical model was loaded with two additional loads as can be seen in the
following figure:
The purpose of these load positions was to determine the effect on the stresses in the
weld connecting the top flange and the web of the crane girder. Both these point loads
were applied at midspan.
The researcher would need to be familiar with Abaqus should he need to do farther
analyses. The model, however, was implemented in such a way that the researcher can
easily create further steps for additional load combinations.
7-1
CHAPTER 7 - CONCLUSIONS
The investigation into the top flange and web deformation in a crane girder panel was
a success. This can be stated due to the good comparisons between the experimental
and numerical results. These comparisons were however not achieved without some
difficulty. Some o f the problems experienced during the experimental investigation
and how they were overcome, can be briefly noted:
7-2
Other researchers can use the numerical finite element model with confidence. This
will enable the researchers to do relative quick and inexpensive investigation in the
behavior o f a crane girder panel when subjected to loads imposed by an overhead
traveling crane. Examples o f these can be summarized as follows:
8-1
CHAPTER 8 - REFERENCES
1 SABS 0160 - 1989, South African Bureau o f Standards, Private Bag X191,
Pretoria.
4 SABS 0162 - 1: 1993, South African Bureau o f Standards, Private Bag X191,
Pretoria.
6 Abaqus, Version 5.8, Habbitt, Karlsson & Sorensen, Inc. Finite element analysis
services (Pty) Ltd, The Waverley, Mowbray 7700.
7 Juliet Warren, Fatigue research o f Crane Girders when subjected to Loads imposed
by Overhead Travelling Cranes, University o f Stellenbosch, 2004.
8 J Collins, R Mackenzie, K Pospisek and W James. “The design and Long Term
Performance o f Heavy Duty Crane Beams”. SA institute o f steel construction. P
473-487.
8-2
10KH Frank and JW Fisher. “Fatigue Strength o f Fillet Welded Cruciform Joints”.
Journal o f the Structural Division. September 1979. P 1727-1741.
M AG Senior and TR Gurney. “The design and service life o f the upper part o f
welded crane girders”. The Structural Engineer. October 1963. P 301-312.
9-1
CHAPTER 9 - ACKNOWLEDGEMENTS
Firstly, I should like to thank the University o f Stellenbosch for the opportunity to do
this research.
I should like to thank Prof. P.E. Dunaiski o f the Department o f Civil Engineering at
the University o f Stellenbosch for his guidance and support.
I should like to thank Messrs A. Rossouw, L. Frederiks and A. Layman for the
physical help during the construction and setup o f the experiments.
Finally, I should like to thank my parents for their continuous support and motivation.
A-1
The following pages contain a summary of the typical dimensions o f cranes at the
Saldanha Steel plant. These dimensions were used to determine representative
dimensional ratios for the design o f the experimental girder.
B-1
Test pieces were manufactured from the top flange, bottom flange and web o f the
beam. The dimensions and test procedures were in accordance with SABS ISO
6892:1984 . The dimensions o f the test pieces were as follows.
Section Dimensions
L (mm) b (mm) t (mm)
Top Flange 300 40 20
Web 300 40 8
The results of these tests were given in Figure B2: Stress Strain Curve. In summary
however, the following can be noted:
E-Modul es 203.50 GP a
Yield Stress 310.50 MPa
B -2
<r'
C-l
The following design was created using commercially available software called
MathCAD 2000 Professional.
Crane Data
NumWheels := 4
NumSideWheels := 2
CraneClass := 4
r.,,,3 0 (S)
C := 0.29
Vu i:= L L F I F V V max (Factored wheel force with impact) vui= 99905 (N)
Rail Size:
Size:= 30
Rail Pads:
15
TrClipBltW := FRailBase + 25.4 TrClipBltW = 163.929 (mm)
17
MinRSW := F RailBas<. + 25.4 MinRSW = 217.45 (mm)
Stellenbosch University http://scholar.sun.ac.za/
Desiqn Parameters:
O
CO
v := 0.3
>
II
E := 203 .5 -103 (M Pa) E= 203.5E +003 (M Pa)
E
G = 78E +003 (M Pa)
2 .( 1 + v)
tfb
tft, (Avarage from dimensional ratios) tftpreUn,] = 18.182 (mm)
16.5
tft,prelim2 •
(tfb-V^) ‘ftprelim2 = 12.990 (mm)
2 •WTRpiangc
(3.1) Misalignment:
Lcg Wheelbase,,,
a= 1800 (mm)
a '= ~2 2
2 ___a__
SASCH Table 5.7 Am := Pt2s'LcG
24-E 3 -4f c )
A „, = 352260442
(Vmm5)I
Leo
A M a*H := A M axH = 7.500 (mm)
(3.2) Skewing:
Leo . ,
a3 := - y *>3 := I- c g ~ a3
2 2
Pi3s-a3 b3
SASCH Table 5.7 Am = 186578624 ( m5)/
\m
3-e-Lcg
^H 2P relim ' 1 2
tft,p re lim 4 : tftprelim4 = 1 1 - 0 5 7 (m m )
tfb 3
tftprelim 4 otherwise
1 8 if 1 6 < t f t prelira5< 1 8
2 0 otherw ise
Stellenbosch University http://scholar.sun.ac.za/
b ffy n in P re lim 150 Choose 150 because: Estimate support bfb,„lnprctjm = 1 5 0 (mm)
width of crane column, bfb greater or
equal to half of top flange width.
(2) Use Flange classification criteria (class 3 flange) with web stabilisation
requirements to determine initial (preliminary) flange properties:
Determine load case that provides
worst case of concentrated force:
-----------------------------------------------------
f
- : :
Lcg Wheelbase,,,
b, := a,
2 2
Lcg Wheelbase^
a2 := b 2 ~ a2
Vui-(b; + bi)
Rl~ Rl = 1 0 0 0 0 0 (N)
Log
M, := RLai Mi = 1 8 0 0 0 0 0 0 0 (N m m )
M2 otherwise
4-MMax
^ E q u iv a le n t PV Equivalent = 1 6 0 0 0 0 (N )
Lcg
(Rosswell vormula for single load P.
Stellenbosch University http://scholar.sun.ac.za/
L eo
"—OPFA B Ratio A0 = 3.600 (mm)
SABS 0160: 1989 SHDL := 6 0 0 Span to horizontal deflection limit SHDL = 600
Tabel 1.1
Lcg
^BotFlangeLim ~ ' '5 (mm)
SHDL
Lcg
M inGirderD:= ------------- MinGirderD = 436.469 (mm)
SGDRAvr
PvEquivalent
PBotFli
;2 n g c MinGirderD \
(l+ — — ---- ) PBotFlange —542.535
(Vmm
- )J
^B o tF la n g e L im /
! (24-WTRnange-lBotFlangeRcq)
bfbpreiimi •“ bfbpreHml = 193.524 (mm)
JTy
bfoprelim l o th e r w is e
350 otherwise
Stellenbosch University http://scholar.sun.ac.za/
bfb-JFy
bftprelim1-- bftprdiml = 8 .6 6 (mm)
2.WTRHange
bfb
bftprelim 2 ■“ bftprelim 2= 5 .6 0 2 (mm)
W TR fJotFlangeM e
bftprelim2 otherwise
Determine the depth of the web from a calculation of the vertical deflection
considering only the flanges when calculating the second moment of area.
V ui .
A v i '= T ~ A H1 Av i= 1761302211.3 (mm5)
* t2s
SABS 0162:1993 ALimv := 600 Crane capacity is < 22.5 Tonne ^ L im V = 600
Appendix I
Lcg
A MaxV • - ' AMaxV= 7.500 (mm)
^ L im V
T ^vi
WPrelimRcq •“ . IvPrelim Req = 234840294.8' (m m 4 )
A MaxV
Yn a i tfb * ” + b f b - b f t- ^ tf t + h Wprc]jm ] + ^ ^
(Without web and rail)
Y n a 2 := tfb - tft + b f b + b f t
( tf b - t f t 3 ) (b fb -b ft3)
IV2Prelim — ^ ^ •v 2 P rd im = 21 6666.67
Iv4Prelim b fb b ft ^ tf t + h wpre|Irn i — ^ ^
lv 4 P re lim = 183843757.5
"Fail” otherwise
Stellenbosch University http://scholar.sun.ac.za/
5 0 0 otherwise
(hw^y)
•wpreliml • ^wpreliml = 4 .1 0 2 (mm)
W T R Web
S := L cg S = 4500 (mm)
S s := I hw if hw < S Sg = 4 5 0 (mm)
| S otherw ise
3
SABS 0162:1984 FwebResMin
twprelim2 = 8 .6 6 4 (mm)
SS E
1 +•
1.5
tw := 1 0 (m m )
Stellenbosch University http://scholar.sun.ac.za/
hw
Ay2 := hw-tw- tft Ay2 = 1 1 0 2 5 0 0 (m m 3)
A = 12500 ( m m 2)
(Ay] + Ay2)
Y n a := YNa = 1 6 9 . 0 0 0 (m m )
( tf b tf t3) ( tft
= — 12— + I A- ~2
(bfb-bft3) .. b ft^ 2
Ixx2 := -----— -------h bfb bft l h - YNA + —
( ) f __ hi
Ixx3:= — — — + bfb-hw-l tft - YNA + —
Zx,:= Zj , = 2 5 7 0 3 6 7 .8 5 (mm3)
Y na
( t f t t f b 3)
1^ = 45 0 0 0 0 0 0 (mm4)
12
(bft-bfb3)
lyb iyb= 6 6 66 66 6 .6 7 (mm4)
12
Zyb lyb
Zyb = 6 6 6 6 6 .6 7 (\mm3)/
bfb
2
ly = 5 1 6 6 6 6 6 6 .6 7 (mm4)
Y „:= Y0 = -8 9 .0 0 0 (mm )
, tft bft 1 3 „ „
h ----------------tfb tft-B
2 2)
cV-w •=
• Cw = 1.256E +012 (mm6)
12
Stellenbosch University http://scholar.sun.ac.za/
Y300WA kg
™CGwpm•“ Q Wcowpm = 98E -003
10
30
W RaillO -'= ------ I WRaillO= 0 .0 3 0 [ ——
Vi—
10
bpad •—63.5 Elastrometric Pad Width bpad = 63.5 (mm)
g := 9.81 g= 9.81
W CGservkg = 0 .1 2 9
kg
W cG scrvkg “ ^ R a il 10 + W CGwpm + W Pad
mm
N
Wcoul. := DLF WcGserv w cGuit = 1 -393
mm
Lcg
Rcgsctv= 2850.11 (N)
rccW = W cg~ v —
Leo
R CGult := W CGult' Rcoui. = 3 1 3 5 .1 2 (N)
Stellenbosch University http://scholar.sun.ac.za/
Vertical Bending:
Lcg
WcGult' + Vui-(bi + bj)
Rviuim = 1 0 3 1 3 5 .1 2 (N)
Lcg
MRultm otherwise
Mvultm
Mvultm
FvBotultm= 1 3 2 .9 0 9 (MPa)
Zxb
Lcg
W CGult---- ------ + V ui-b3
Myults
FvTopults= 4 5 .1 4 0 (N)
Transverse Bendina:
Case 1: Misalignment:
:= Pt2s Hm = 20000 (N )
100000
>6
>3
Vm = (N)
II
C ase 2: Skewing:
Hs := P,3s Hs = 20000 (N )
Vs := v ul v s= 100000 (N)
Calculate horizontal deflection, flange forces, moments, and stresses due to the following
applied forces and force effects at crane wheel positions:
(2) Transverse displacem ent due to transverse wheel forces at rail head ievel.
(3) Transverse deflection due to the force effect:torsion about the shear centre
caused by the transverse wheel forces.
(4) Transverse deflection due to the force effect torsion due to vertical wheel forces
offset by m isalignm ent of the crane rail.
(5) Transverse deflection due to the force effect: stabilisation forces for the web,
which acts as a column under the vertical wheel forces.
Note:
(a) Transverse thrusts / forces are resolved as equivalent horizontal forces and
torsional m om ents about the centroid of the girder section.
(b) The equivalent horizontal force at the centroid o f the girder section is resolved
into equivalent horizontal forces at the top and bottom flange (by taking
m om ents about the centroid of the top and bottom flange).
(c) Torsional m om ents are resolved as equal horizontal force couples. The lever
arm is the distance between the top and bottom flange centroids.
(d) Horizontal flange forces resolved from torsional m om ents are adjusted to obtain
the warping torsion com ponent, which contributes to the longitudinal stresses
and deflections.
Stellenbosch University http://scholar.sun.ac.za/
(2) Transverse displacements due to transverse w heel forces applied at rail head level:
tft bft
hff := h ■ hff = 4 6 5 (mm)
2 2
bft ,
^scbfh ~ ^ —Yna + Y0 h Scbf = 3 9 5 .0 0 0 (mm)
HTFm,:= Hn-
hscbf HTFm, = 1 6 9 8 9 .2 5 (N)
hff
HTFs, := Hs-
hscbf HTFst= 1 6 9 8 9 .2 5 M)
HBFm,:= Hm -
hsctf HBFmt= 2 5 8 0 .6 5 (N)
h ff
HBF,. := Hs-
^sctf h b f s1= 2 5 8 0 .6 5 (N)
"hfT
HTF«
H B F mt <
Stellenbosch University http://scholar.sun.ac.za/
(3) &(4) Transverse Deflection due to the force effect: Torsion about the
shear centre caused by the transverse wheel forces and transverse
deflection due to the force effect: Torsion due to vertical wheel forces
offset by m isalignm ent of the crane rail.
— if — < 0 .5 “ ml = 0 -4 0 0
Lcg Lcg
1 -------— I otherwise
Lc g J
a m2 •“ — if — < 0 .5 “ m2 = 0 .4 0 0
Lcg LCg
1 ------ —- otherwise
LcgJ
a m2 otherw ise
G-J
X :=
ECW
X L:= X -L co XL = 2 . 5 1 1 (m )
Case 2: Skewing
a* = 0 .5 0 0
L cg
HTFmhft:= HTFmhft= 5 9 1 6 .1 3 (N )
hff
HTFshft := HTFshft = 7 4 3 7 .4 2 (N )
hff
HTFmvft:= ^ . p n HTFmvft= 1 4 3 0 .1 1 (N )
hff
Tsv
HTF svf, := — ps HTFsvft= 1 7 9 7 .8 5 (N )
hff
HBFmvft:= —
HTFmvft HBFmvft = - 1 4 3 0 .1 1 (N )
H B F := —HTFsvft HBFsvft = - 1 7 9 7 .8 5 (N )
H T F m h ft
H B F m h ft
Stellenbosch University http://scholar.sun.ac.za/
Lcg a. ( HTFmhft
STFmhft := 3 -4
24 Lcg I E Iyt
8TFmhft= 2 .3 1 6 (mm)
8BFmhf, = -1 5 .6 3 0 (mm)
. 2
f HTFmvft
Lcg ai
5TFmvft •- 3 -4
24 Lcg I ^
8TFmvft= 0 .560 (mm)
. 2
Lcg ai f HBFmvft
5BFmvft 3 - 4-!
24 Lcg I E'lyb
8BFmvft = -3 .7 7 8 (mm)
f 2 2) / \
a 3 -b 3 f H T F d ,^
8TFshft := 8TFshft= 1.542 (mm)
v3-Lcg ^ I E Iy. J
f
a3
2
b3
2) __
(/ HBFshftA \
8BFs),ft = -1 0 .4 0 7 (mm)
SBFshf, :=
^3-LcG/ I E-lyb J
( 2 2 s]
a3 b3 ( HTFsvfA
8TFsvft:= STFshft= 1.542 (mm)
v3-LcG/ I E Iyt J
( 2 2 \
a3 -b3 ( HBFsvlA
8BFsvft : - SBFgyfj = -2 .5 1 6 (mm)
V 3 'Lc g , I E Iyb J
Stellenbosch University http://scholar.sun.ac.za/
(5) Transverse deflection due to the force effect: Stabilisation forces for the web,
which acts as a column under the vertical wheel forces.
• 2
l CG al f H T F mwebst
STF, 3 -4 -1 -5 -
24 l cg \ ~ eTiyt
,Leo2-a| ( HBFmwebst
3-4-
24 l cg I E-Iyb
5BFmwebst = -2 .6 4 2 (mm)
( 22 '] / _____ \
j H TFgwebst j
a3 V
5TFswebst: STFswcbst = 0.207 (mm)
V3 l Co J I Ely. J
Iteration 1:
AmTiit = 2 8 .7 4 9 (mm)
Iteration 2:
.L 2
cg 'a i H^mwebst
STFmwcbs,
24 I Ely,
STFmwebst= 1 . 6 9 4 (mm)
, 2
J HBF mwebst
Lcg ai
8 BFmWebst 3-4| —
24 Lcg V E-I,
'' Ayb
SBFmwebst = - 1 1 . 4 3 4 (m m )
HTFmwebst H T F mWebst = 5 8 4 7 . 3 4 (N )
h ff
( 2 2 s) / ____ \
a3 t>3 j H T F swc),st |
8 T F SWebst :~ ST F sw ebst = 0 . 7 8 3 (m m )
^ 3 -L c g ^ I E-1^ J
( a32 b 32s]
SBFs,
f HBFSwebst |
SBFswebst = - 5 . 2 8 5 (m m )
^ 3 -L c Gy I EIyb J
^s’^sTilt'Ps
HTF sweb s t := H TF swebst = 4 6 2 0 . 6 9 (N )
h ff
Iteration 3:
Lcg f HTFmwej,st
24 I E-Vt
STFmwebst = 2 .2 8 9 (mm)
, 2
8BFm
l cg “i
3 -4 ai
24 lcg V E I‘yb
:
5BFmwebst = -1 5 .4 4 8 (mm)
^ m ' ^ m T ilt'P iT
HTFmwe5st := H TFmwebst = 6541.23 (N )
hfT
( 2 2 ) j/ H___ X
a3 V T Fgw ebst j
8TFswebst •- 5TFswebst = 0 .9 5 8 (mm)
^3-LcGy I E'ty. J
( a32 b /2 s) /■____ \
| HBF swebst j
8BF*, 8BFswebs, = -6 .4 6 6 (mm)
^ 3 - L cg J I E I yb J
V s ’^ s T il f P s
HTFswebst = 4 8 7 7 .3 8 (N)
h ff
it-px H-L
p! = 12.282
2-K Lcg v GJ
2 Cr
7t - E -C u
P2 := p2 = 156.490
(K -L cg ) GJ
y := 0.67 y = 0.67
<t>MrxTop
^rxTop FrxTop = 310.15 (MPa)
Zx,
<l>MrxBot
FrxBot FrxBo.= 270 (MPa)
Zxb
Stellenbosch University http://scholar.sun.ac.za/
Combined Stresses:
W eb Shear Strength:
W eb Buckling:
hw
HTDActllfli:= — HTDActua] = 4 5 .0 0 0
1100
HTDLimi, = 3.667
S = 900 (mm)
5.34
SABS 0162-1:1993 kv:= if - L s l kv = 7.340
Par. 13.4.11
5.34 + otherwise
( 2 9 0 / f7Tv)
fcri- fcri= 3 0 2 .4 09 (MPa)
HTD^ctual
1 +
Ay := hw-tw Av = 4500 (m m l
"FAIL" otherwise
Stellenbosch University http://scholar.sun.ac.za/
1.5
B c r := 3 0 0 d .b C - 1 + 3-
N i l 1' tft
B c r= 1 0 8 6 9 2 7 .1 1 ( H )
hw ) Vtft,
BcrCheck = "OK"
"FAIL" otherwise
S s := hw if hw S S Ss = 4 5 0 (m m )
S otherwise
1 + F w cb R e s = 4 5 2 2 2 2 .2 2 (M P a )
"FAIL" otherwise
Stellenbosch University http://scholar.sun.ac.za/
2
A BearStiff == tB e a rS tifrM b + 1 2 - t w A BearStiff = 2000 (mm2)
( l2 -tw)-tw3 tBearStiff'bfb3
^BearStiff •— lB e a rS tif f= 2 6 7 6 6 6 6 .6 7 (mm4)
12 12
^BearStiff
ABearStiff = 3 6 .5 8 3 (mm)
A BearStiff
hw I Fv
^-BearStiff = 0 . 1 5 0
r BearStiff .1 n ^ . £
0 otherwise
F B e a rS tiff= 2 6 9 .9 (MPa)
"FAIL" otherwise
bft)- y J T y
^BearStiffMin — 8 .6 6 0 (m m )
2 WTR
^BearStiffMin o th e rw is e
10 i f 8 < tBearStiffPrelim ^ 1 0
12 o th e r w is e
Stellenbosch University http://scholar.sun.ac.za/
Welds:
•^material -= 0 .9 5
4*weld := 0 .6 7
fy.Mat:= 3 0 0
fy.Weld:= 4 1 0
( 71
^r.Weld 0 .67 -(J)wcid-[ cod j I'C'fy.Weld Vr.Weld = 6 5 0 .7 1
Vui'Qbot
V\veld.bot Vweld.bot = 7 0 .4 4 3
2 -Ixx
Q to p tft-tfb-j Yna 2
Vui'Qtop
Vweld.top -= V w e .d ,o P = 1 0 9 .8 0 9
2 -Ixx
Vweld.top
e:= 3 . 15- e = 0 .8 4 4
•y.Weld
tfb = 3 0 0 (mm)
tft = 2 0 (mm)
hw = 4 5 0 (mm)
tw = 10 (mm)
bfb = 2 0 0 (mm)
bfl = 1 0 (mm)
tBrarStiff = 10 (mm)
S = 900 (mm)
Leo = 4 5 0 0 (mm)