Mathematical model for the determination of thermal spalling in refractory material on basis of the practical relationship of the appearance of rupture, physical properties and physical conditions

Directorate-General For Energy

Description

Industrial research and development

Informations

• Publié par: DIRECTORATE-GENERAL-FOR-ENERGY
• Nombre de lectures: 28
• Langue: English
• Poids de l'ouvrage: 5 Mo

Extrait

Commission of the European Communities
technical coal research
MATHEMATICAL MODEL FOR THE DETERMINATION
OF THERMAL SPALLING IN REFRACTORY MATERIAL
ON BASIS OF THE PRACTICAL RELATIONSHIP
OF THE APPEARANCE OF RUPTURE,
PHYSICAL PROPERTIES AND PHYSICAL CONDITIONS
Report
EUR 10068 EN
Blow-up from microfiche original Commission of the European Communities
technical coal research
MATHEMATICAL MODEL FOR THE DETERMINATION
OF THERMAL SPALLING IN REFRACTORY MATERIAL
ON BASIS OF THE PRACTICAL RELATIONSHIP
OF THE APPEARANCE OF RUPTURE,
PHYSICAL PROPERTIES AND PHYSICAL CONDITIONS
HOOGOVENS GROEP BV
Research- en bedrijfslaboratoria
Vuurvast, metallurgie en staalfabricage
IJMUIDEN (Nederland)
FINAL REPORT
Contract N° 7220-EB/603
Directorate-General Energy
1985 EUR 10068 EN Published by the
COMMISSION OF THE EUROPEAN COMMUNITIES
Directorate-General
Information Market and Innovation
Bâtiment Jean Monnet
LUXEMBOURG
LEGAL NOTICE
Neither the Commission of the European Communities nor any person acting on behalf
of then is responsible for the use which might be made of the following
information
© ECSC- EEG— EAEC Brussels - Luxembourg, 1985 Hoogovens IJmuiden
- 2 -
SUMMARY
Finite element calculation methods are in common use for calculations of stres­
ses in steel constructions and thermal calculations of other materials as well.
However refractory materials show important differences of material properties.
So application of FEM-methods for study of spalling behaviour of refractory
materials needs adjustment and testing of the method.
In this research simulative tests and mathematical calculations are executed
parallel. So model calculation results can be evaluated with help of the prac­
tical appearance of rupture.
Material properties are used for the model calculations. These methods are
screened with help of a statistical analysis and tested in order to establish
the physical utility.
The statistical analysis showed that the determination of the elastical pro­
perties needed improvement. That is why a new deformation measuring device has
been developed for service at high temperatures.
Literature showed that the elastic properties tested by different tests could
not be brought in agreement. So a number of tests were done to record the elas­
tic behaviour in such a way that the results of the test could be combined to
a stress-strain diagram which shows the differences between tensile conditions.
Ultrasonic test methods appeared to be not usefull for this application.
While the standard FEM~programs do not meet with the above mentioned stress-
strain diagram a computer program is implemented into the FEM-sequence to rea­
lize, a physically correct calculation. The yield hypothesis has als been chosen
in accordance with the established stress-strain diagram.
The results of the calculations were in good agreement with the simulation
tests except for a systematic deviation.
The final model has been applied to a industrial situation at the IJmuiden
coke plant II. This has resulted to a replacement of a part of the construction
(silica) by another material.
Hoogovens Groep Hoogovens IJmuiden
CONTENTS
pagj
1. INTRODUCTION 5
2. THEORETICAL STUDY 7
2.1 Thermal shock
2.2 The influence of thermal conditions on thermal stress
s ituations
2.3 The influence of brick geometry on thermal stress
situations 10
2.4 Analytical calculations methods
2.5 The fracture mechanics approach7
3. METHODS OF TESTING 21
3.1 Methods of testing necessary for the thermal model 21
3.2s ofgy for stress distribution
calculations3
3.3 Statistical consequences of the accuracy of the test
methods 32
3.4 Improved deflection measuring device for hot bending test 3
4. MATERIAL PROPERTIES6
4.1 Properties of some brick qualities 3
4.2 Experimental properties to determine the shape of the stress
s train diagram 3
5. THE CRITERION FOR YIELDING 43
6. SIMULATION TESTS5
6.1 Simulative test results of silica
6.2 'Simulative tests of chamotte bricks 49
7. MODEL DEVELOPMENT 51
7.1 Definitions
7.2 Development scheme3
7.3 Blockgenerator
7.4 Transient heat transfer analysis for non-linear model using
the finite element method
7.5 Test results of the heat transfer analysis 56
Hoogovens Groep Hoogovens IJmuiden
- 4 -
page
7.6 Calculations of thermal stresses for a model with linear
elastic material properties 58
7.7 Development of non linear elastic model for laminated
materials 66
7.8 Results of some non linear calculations 67
7.9s of calculation with equivalent stress presentation .. 6
7.10 Handling of computer output data
8. PHYSICAL SIGNIFICANCE OF STRESS RESULTS 71
9. APPLICATION OF THE MODEL FOR THE CERAMIC BURNER INLET OF A COKE
OVEN 73
10. CONCLUSIONS 88
11 . ACKNOWLEDGEMENT9
12. LITERATURE , 90
13. LIST OF SYMBOLS4
APPENDICES
oogovens Groep Hoogovens IJmuiden
- 5 -
I. INTRODUCTION
The development of refractory materials has been dominated for a long time
by improvement of refractoriness and strength. This was done by searching
for better refining methods for the raw materials and fabrication methods
which cause good densification. Corrosion resistance was also improved.
When these improvements by these efforts reached a acceptable level an other
property of refractory materials was emphasized; the spalling behaviour.
This phenomenon occurs principally in all ceramic materials. The reason why
it happens mainly in the steel-industry is caused by the big temperature
changes in these production processes. Expansion variations between 0.3 and
1.5 % are not outstanding. The deformability does contrast with these ex­
pansion variations. The usual potential is only 0.1 till 0.3 %.
In fact the inability of they to cope with the big expansion
variations is the source of spalling problems. As a consequence materials
which do not expand will not show spalling problems.
In the last decennia several attempts have been made by investigators in
Europe and the USA to describe spalling phenomena of refractory materials
in service. It was well understood which material properties were of in­
fluence. So for the material researcher possibilities were present to im­
prove the refractories. However thes to improve the material
are restricted and in some cases process conditions or construction methods
play a more important role.
At this point there is a need to describe the spalling process as a whole.
The relative material improvements can not be translated to industrial situ­
ations without including brick dimensions and process influences.
Two approaches are possible:
An analytical description and a numerical description.
The first having the disadvantage that the description is very complex and
restricted to certain special cases. The second having the disadvantage that,
although the calculations are relatively simple, because of the iterative
character the work is enormous. This caused a delay in the developments.
Besides developments were also delayed by unsatisfactory test methods at ele­
vated temperatures.
Recent developments in computer technology take away the disadvantage of nu­
merical description (big calculation jobs) at a quick rate. The "computer
second" has become more powerfull and cheaper lately and this tendency will
proceed.
Hoogovens Groep Hoogovens IJmuiden
That is why a research has been started to realize a simulation method by
numerical methods in order to handle a large variety of thermal stress
problems in coke ovens and other steel plant furnaces.
F
Hoogovens Groep