Beruflich Dokumente
Kultur Dokumente
142
25th International Conference on Ground Control in Mining
beyond the point of failure. From laboratory studies on welded wire
Transverse wire Longitudinal wire mesh, three types of strength failures were noticed [2]. These were
weld, Heat Affected Zone (HAZ) and wire failures. When the loads
imposed on the screen produce shear or tensile stresses exceeding
the strength of weld spots, then the weld failure occurs. Similarly,
when the loads exceed the strength of wires in areas affected by the
heat and pressure during welding, or the tensile strength of wires in
Sheet width
the remaining areas, the HAZ and the wire failures occur,
respectively [2].
In underground mining application, typically, the transverse and 4.0 WIRE MESH DESIGN APPROACHES
longitudinal wires have the same diameter and spacing though
occasional use of rectangular or non-uniform spacing is also seen. As pointed out before, the two important parameters that need
8-gauge screen with 4” x 4” spacing is the most commonly used careful evaluation for wire mesh design are: the gauge and wire
configuration in the US coal mines. Occasionally 6 or 4-guage spacing. The selection of these two factors is primarily dependent
mesh or 3” x 3” aperture is also used. Non-uniform wire spacing is on the expected rock load that the screen should carry without
preferred by a few mainly as a convenience to let roof bolter strength failure or excessive sag. Unfortunately, there are currently
operators know the exact location of roof bolts with respect to the no accurate ways to estimate the rock load on the mesh or for that
placement of the sheet. matter even to decide under what conditions any skin control
measures are necessary. As far as the authors are aware, the only
equation available to estimate the expected load on a wire mesh is
3.0 MODES OF MESH FAILURE given by Coates [3] as
AND THE DESIGN CRITERIA
Pv 0.727 s (1)
Functionally, a welded wire mesh can fail in two different modes:
strength and bagging. In the strength failure, the mesh suffers
with Pv being the vertical load per unit area, the specific weight of
structural failure and will not be able to handle loads added to it
the roof rock and s is the spacing between bolts in a row. Equation
143
25th International Conference on Ground Control in Mining
(1) is based on the pressure-arch theory and has some assumptions Owing to these limitations, all the published physical testing works
that are not readily verifiable through field observations and hence considered only a few of the factors that can affect a wire mesh’s
its validity can not be assured. Further, at the upper limit, the rock behavior and that too within a very limited range of their possible
load estimated by equation (1) is extremely high for the mesh values.
application. Additionally, equation (1) considers only stress related
failure and can not account for weathering related slaking, which is Since the structural behavior of the screen is a 3D problem, no
a major contributor for skin failures in thinly laminated weak analytical solution could be obtained that accurately satisfies all the
shales. boundary and loading conditions seen in the field. As a result,
extremely simplifying assumptions were made in the past to reduce
Since there are no reliable ways to estimate the probable rock load the problem to a 2D situation to develop analytical solutions. While
on the screen, engineering judgment based on past experience at the physical testing is the most accurate way, as mentioned above, it
mine or from nearby mines is necessary. Even though not based on has several practical limitations and hence can not be used for
a systematic study, experience shows that roofs with Coal Mine routine designs. For these reasons, Thompson et al [7] proposed to
Roof Rating (CMRR) [4] less than 35 may need some form of skin use numerical modeling for the mesh design as it can incorporate a
control, especially, in openings that need to stay open over long large number of factors in the analysis. Although their paper talks
periods of time. Skin control is also necessary in highly laminated about a methodology to achieve this goal, at the time they
weak shale roofs that are prone to weathering related disintegration. published their paper the approach was not fully developed and
In addition to the CMRR value, the Unit 1 rating used to derive the hence no results were provided there [7]. As far as the authors of
overall CMRR also provide additional insights on how prone the this paper are aware, there is no published work in the literature
immediate roof is to skin failures. A systematic research is required that tackles the wire mesh design problem using 3D numerical
to estimate the extent of slaking and stress related surface failures. modeling.
If the expected load on the mesh is estimated by some means, then 5.0 NUMERICAL MODELING METHODOLOGY
the required wire gauge and spacing could be determined by
different approaches. Literature review [2,3,5-9] shows two
Ideally, a wire mesh design approach (physical or analytical)
approaches were used in the past: simple analytical models and
should be able to incorporate the following elements of the physical
physical testing. Analytical models, however, were never used as a
problem [7]:
stand-alone methodology and were generally employed to explain
the behavior seen in the tests. The only exception to this was
a) Variable sheet width and length
Coates [3] who used some simple equations to estimate the axial
b) Variable wire diameter and spacing
force in the mesh wire without any laboratory testing to support the
c) Wire’s non-linear load-deformation properties
theory.
d) Weld’s load-deformation properties
e) Large wire deformations
In the physical testing, a square sheet of mesh is directly loaded
f) Variable bolt spacing, bolt tensions and bearing plate
either in lab or field using a steel plate or frame to generate the
sizes
load-deformation curves and to obtain the peak load capacity of the
g) Slippage behavior at the bearing plates
mesh. This methodology is the most reliable way to choose a
h) Variable rock loads and their distribution patterns
proper wire mesh but has the following practical limitations:
Numerical modeling can incorporate all these factors in the analysis
the sheet size used is a shorter version of the actual one
and has some other advantages which include, cheaper analysis
and thus the results may not be directly applicable to the
cost, no need of complicated test set up and the ease of conducting
full-size sheets
parametric studies over a broad range of input values.
testing requires elaborate laboratory or field set up
parametric studies are hard to conduct thus limiting the With modeling it is also easy to investigate if non-standard screen
number of variables that can be studied configurations are needed in special applications. For example, if
difficult to gain insights on the way loads and high rock loads are anticipated, then instead of increasing the wire
deformations are distributed among individual wires and diameter over the entire sheet, it may be possible to achieve
require subjective interpretation about critically loaded satisfactory design by incorporating large diameter wires at a few
areas strategic locations within the mesh or by changing wire spacing
extremely difficult to vary how the rock load is only in certain areas. Such flexibility is almost impossible with any
distributed on the screen other design methods. Since the readily available screens in market
can not accurately simulate the conditions at the bearing are made to certain standard sizes in popular use, any non-standard
plates, unless tested in situ design needs to be custom manufactured for testing in the
interpretation of the load-deformation results is laboratory. If the initial design is unsatisfactory, then another mesh
complicated beyond the first peak as several factors can must be built before its adequacy is checked. With modeling,
produce the same result and it is difficult to pinpoint the however, there is no need to fabricate the screen for testing. On
actual cause for the observed behavior (discussed more in contrary, based on several numerical models, one can specify what
section 6.0) type of mesh may be manufactured to achieve the design goals.
requires special fabrication of the mesh if a non-standard
design is to be checked Another factor in favor of modeling is that in recent years it has
needs capital investment and the entire process is very become a routine tool for ground control studies and hence is not
time consuming totally unfamiliar to the mining community. Further, a majority of
the commercially available computer programs include the
‘structural elements’ needed to build wire mesh models. Thus the
144
25th International Conference on Ground Control in Mining
mesh design by numerical modeling will not be a significant
deviation from routine ground control modeling and with only a
little more effort it can be included as a part of the regular analysis.
145
25th International Conference on Ground Control in Mining
(a) (a)
1600
20
1400
1200 16
1000
Load, lb
Tested 12
800
Load, KN
Modeling
Model
600 Tested
8
400
200
4
0
0 5 10 15 20 25
Deformation, in
0
(b) 0 50 100 150 200 250 300 350
3500
Deformation, mm
3000 (b)
15
2500
2000
Load, lb
Tested 12
Model
1500
9
Load, KN
1000
Model
Tested
500
6
0
0 5 10 15 20 25
Deformation, in 3
146
25th International Conference on Ground Control in Mining
In a typical physical test, a small screen is bolted on 4ft x 4ft square
1
pattern and is pulled using a 12 inch plate to failure. As a result of
the smaller size, the boundary conditions along two sides of the
sheet are not correctly implemented. This has two consequences.
First, due to the lack of restraint that would otherwise be available
in a full-size sheet, the measured deformations in lab tests would be
higher. Second, when the load carrying wires start to yield, they 2
shed a part of their load to adjacent wires that are not fully loaded
and this load shedding process can be different in a smaller sheet as 3
compared to a full-size one. These limitations are demonstrated by
the modeling results in Figure 7 and Figure 8. The axial force
distribution is shown using pipe diagrams and the displacements
are plotted by vectors. A feel for the magnitudes of the force or
displacement could be gained by comparing the relative size of the
pipes or vectors. To aid in the visualization, the values of maximum
central displacement and wire axial force are also given in the
figures. Comparison of results in Figure 7(a) and Figure 7(b) shows Figure 8. Axial force variation at select wires shown in Figure 7(b)
that the measured displacements in lab tests are higher. The load in the full-size sheet as a function of the model solution step.
distribution patterns in the immediate neighborhood of rock load,
however, are similar. The modeling also shows that a majority of under the bearing plate at those locations. Before the peak was
rock load is carried by the wires passing under the bearing plates reached in wire 1, the load in wires 2 and 3 was negligible and they
with very little axial load seen in the rest of the wires. started picking up load as wire 1 started yielding. Subsequently as
additional wires yielded and shed more load, the load in wires 1,2
and 3 fell a little before stabilizing at their final equilibrium values
shown in Figure 8. A vertical dotted line is added to Figure 8 to
clearly show the load-shed mechanism. This shedding can not
(a) happen in a small-size sheet as wires 2 and 3 do not exist.
Max. displacement = 7.11 in The difference in the displacements shown in Figure 7 and Figure
2 Max. axial force = 1064 lb 9(a) is not only because of the differences in the boundary
1 conditions, but also due to the smaller spacing between bolts in the
loaded region in Figure 9(a). This shows the importance of the span
on mesh deformations. The smaller the bolt spacing, the higher the
Figure 7. Load-displacement distributions in (a) small- and (b) mesh stiffness and vice versa. Therefore, if no other option exists,
full-size sheet for identical rock loads. higher mesh stiffness could also be achieved by reducing the bolt
spacing.
The variation of the axial load in a few critically loaded wires was
monitored as the solution was progressing as shown in Figure 8.
The wires for which the loads were monitored are indicated in
Figure 7(b). Once the load in wire 1 reached its yield value, it
started shedding load to wires 2 and 3 as those two wires passed
147
25th International Conference on Ground Control in Mining
the three. But the worst possible loading occurs when the broken
(a) rock falls on the entire sheet. Therefore, the modeling results are
presented only for this condition. Similarly, it was pointed out in
section 5.0 that in numerical models the fixity conditions at the
bearing plates were either all-nodes or one-node. Since the actual
number of wires passing below a bearing plate in the field lies
somewhere in between the simulated extremes, the true
displacements of the mesh will probably be closer to an average of
these two limiting conditions. These results are shown in Figure 10
and labeled as average-fixity conditions. Additionally, in this figure
Max. displacement =6.287 in results are also presented separately for the two bracketing fixity
Max. axial force = 1041 lb conditions. Three different unstable rock heights, 3”, 6” and 12”
were considered in the modeling and the results were given
separately for each of them. For convenience, a line is added in
Figure 10 at twelve inch displacement to indicate the limit of
bagging failure. Further, if there is a significant deviation from a
(b) near-straight line trend in displacements with a change in the wire
gauge, it means for those conditions the screen has suffered
strength failure at a few points but could still take the load. If no
results are shown for any condition, it indicates that several wires
failed in the mesh and the model could not reach its equilibrium
state.
Max. displacement =6.912 in
Comparison of the results in Figure 10 shows the importance of the
Max. axial force = 1035 lb
boundary conditions at the bearing plates. These results show that
the performance of a mesh could be improved by a good quality
control during installation. With everything else being the same, a
higher mesh stiffness and load capacity could be obtained by
ensuring that the maximum possible number of wires are held tight
by the plates. These results also indicate that the mesh stiffness and
(c) strength could also be improved by increasing the size of the
bearing plate as a larger plate will have more wires passing under it
than a smaller one.
Max. displacement =6.763 in For fully grouted resin bolts with very little installed bolt
Max. axial force = 1143 lb tension, one-node fixity is perhaps appropriate
For mechanical or combination bolts, if the initial tension
is within the range that is normally used, the average
fixity conditions are applicable
If the installed tension is very high and the roof
Figure 9. Load-deformation changes due to the difference in the conditions do not allow tension bleed off, then all-node
loading locations, (a) left only, (b) left and central and (c) all three. fixity may be used
With mechanical and combination bolts, if experience at
8.0 PARAMETRIC STUDIES the mine shows possible tension bleed-off with time, then
the one-node fixity conditions must be used
Having shown the validity and the kind of details that can be Finally, installing a trial screen at the mine will provide a
obtained with numerical modeling, the next step is to extend the better idea of the fixity conditions pertinent to that mine.
analysis and develop design guidelines for the welded wire mesh By noticing the number of wires passing under the
selection. For this purpose parametric studies were conducted by bearing plates with regular quality of installation
varying a few common input design variables. The bolting pattern expected at the mine, the correct number of wires to be
shown in Figure 6 with 4.5ft row spacing was used to estimate the fixed in the numerical models could be determined.
expected rock load on the mesh and the sheet size. It was assumed
that the bearing plate’s influence extends to 6” on either side of the Based on the results in Figure 10, it appears it is not efficient and
bolt center and the load on the screen is estimated using the total perhaps not economical to use welded wire mesh to handle rock
remaining area of rock between adjacent bolts. Additionally, for all loads greater than 1ft thick. If the roof instability extends beyond
the parametric studies the most commonly used aperture size 4” x this height then probably the roof bolting pattern itself may need to
4” is taken. The factors varied were the wire gauge, the height of be evaluated for its effectiveness. Figure 10 also shows the wire
rock load, the distribution of rock load and the number of wires gauge should be 6 or less to support higher loads. Of course, the
fixed under the bearing plates. stiffness and capacity of the mesh could also be improved by
changing the wire spacing. In any case, if the screen is planned for
As mentioned in section 7.0, the rock load can act in the left, large rock loads, then some non-standard mesh designs may be
middle or right regions of the mesh or any other combinations of
148
25th International Conference on Ground Control in Mining
(a) considered and it is better to conduct custom numerical modeling
studies to arrive at the best design.
Wire gauge
2 4 6 8 10 12 14
0
For the most common mining conditions, the modeling results in
Figure 10 indicate that a 4” x 4” wire mesh with 6- to 8-gauge
5 wires is sufficient. This is consistent with what is seen in the
practice. While it is preferable to conduct numerical modeling
10 using the site specific conditions, the results given in this section
may be used as an initial guide. Based on the subsequent
Max. displacement, in
(b) 1. Evaluate the need for a skin control measure. This could
be done based on past experience at the mine or nearby
Wire gauge
2 4 6 8 10 12 14
mines. As an initial guide if the roof CMRR value is
0 below 35, consideration may be given for adopting some
surface control measure, especially, for openings that
5 require long-term stability. Similarly, estimate the
probable height of unstable roof that may be considered
10 as a skin failure. Again, no fixed rules are available at
this juncture and educated guesses based on experience
Max. displacement, in
15
3"
9. If multiple wire gauges satisfy the design criteria, then
20 6" choose the one that is most economical and convenient to
12"
use.
25 10. If site conditions are significantly different from those
used for Figure 10, then build a numerical model
30
representing all the pertinent site-specific parameters to
choose the wire spacing and gauge that satisfies the
35
design criteria.
40
11. If non-uniform wire spacing, variable gauges or non-
standard yield strengths are considered, again construct a
Figure 10. Maximum displacement in the wire mesh as a function numerical model incorporating those details to check if
of wire gauge for different rock load heights with (a) all-node (b) the design criteria are satisfied or make changes until a
one-node and (c) average-fixity at the bearing plates. satisfactory design is achieved.
149
25th International Conference on Ground Control in Mining
10.0 DISCUSSION AND CONCLUSIONS testing. Specifically, it was found that results from such tests form
the upper-bound on displacements for the same loads acting on a
Welded wire mesh is the most commonly used skin control full-size sheet used in practice.
measure in the US underground coal mines. Despite it popularity,
rigorous design methods for a proper selection of the mesh are Based on parametric studies, design guidelines were developed for
lacking. The industry practice so far has been to pick a mesh based the mesh selection for the most common mining and other input
on past experience or what the engineer ‘thinks’ will work from a variables. These studies confirm the adequacy of the popular 8-
small pool of standard designs supplied by vendors. Even though gauge, 4” x 4” aperture wire mesh for the regular ground
scanty, literature review showed that most of the available conditions. It was also found that the limiting rock load for the
information on the mechanical behavior of the welded mesh was screen’s application may be around 1ft high. If it is necessary to
derived from physical testing in laboratory or in field. As pointed support more skin rock than this, the choice of the final gauge and
out in this paper, due to several practical constraints, the past aperture size may be made based on an economic analysis. The cost
testing was not comprehensive or totally informative. Similarly, of using a sheet with a lower gauge (higher diameter) and lower
because of the complicated interactions of several factors, efforts to wire spacing over the entire mine or section must be compared with
develop a closed-form solution or a ‘formula’ to guide wire mesh the costs that would be incurred if a lighter mesh is used in general,
design were also inadequate. Therefore, an alternative approach is and rehabilitation is done in the areas where it might fail
warranted which is easy to use, economical and yet can derive the functionally. Alternatively, non-standard screens may be designed
load-deformation characteristics of the mesh with an acceptable to handle the higher loads using the modeling methodology.
accuracy. Research in this paper showed 3D numerical modeling
can fulfill these requirements. While numerical modeling has several desirable features, it has
some limitations too. As mentioned before, at this stage, modeling
During the course of this study it was found that currently there are can not accurately incorporate two factors, weld failure and
no available methods to estimate the extent of unstable skin rock or slippage at the bolts, in the analysis. The impact of ignoring these
even to decide under what conditions some form of surface two could be negligible to considerable depending on whether the
coverage is necessary. As an initial guide, it is proposed to use quality of welding is good or bad and whether the bolt plates are
CMRR as the baseline parameter and consideration may be given to installed properly or not. Despite this, since the limitations of the
apply some skin control if its value is below 35. The most effective modeling are no worse than the laboratory testing and considering
selection of the mesh could be achieved by generating CMRR the other significant benefits described in this paper, numerical
contours for the entire mine, section or entries under consideration. modeling may be used as a routine design tool for the welded wire
By looking at these contour maps a qualitative ‘feel’ could be mesh selection.
developed about the probable mesh load heights in different parts
of the mine. Based on this information, different mesh ACKNOWLEDGEMENTS AND DISCLAIMER
specifications appropriate for the different loading conditions could
be chosen. The authors would like to thank Dennis Dolinar of the National
Institute for Occupational Safety and Health (NIOSH) for sharing
Two different criteria, strength and bagging failure, were identified some of his tested data and for the many stimulating discussions on
to guide wire mesh selection. While the strength failure is a clearly the engineering behavior of the welded wire mesh.
identifiable phenomenon, the bagging failure is dependent on
several site-specific conditions. For the lack of any guideline, it is Reference to some products in the paper does not constitute
proposed to use 12 inch as the limiting deformation to classify the endorsement of the product by Peabody Energy or its affiliates.
failure as bagging. This number may need to be modified based on
a systematic study from underground observations.
REFERENCES
The non-linear 3D modeling methodology adopted in this paper
showed a very good match with the data derived from some lab 1. Manual of Standard Practice –Structural Welded Reinforcement
tests. The match, however, was limited to the monotonically (WWR-500). Wire Reinforcement Institute, Inc., Hartford, CT,
increasing part of the load-deformation curves. From a design point 1999, 35p.
of view, this limitation of modeling is not of serious consequence
as the mesh has suffered bagging failure by then and was also at the 2. Villaescusa, E. Laboratory testing of weld mesh for rock support.
verge of strength failure. Past laboratory studies showed that in the Paper in the Proceedings of International Symposium on Rock
post smooth-curve region, several wires and welds failed causing Reinforcement Practice in Mining’, Kalgoorlie, Balkema,
the observed saw-tooth behavior. From practical standpoint, such Rotterdam, 1999, pp. 155-159.
failures will demand rehabilitation of the mesh even if the peak
screen capacity exists in the post-smooth region. Therefore, the 3. Coates, D. F. Rock Mechanics Principles. Energy Mine
simulation of the complete load-deformation curve is only of Resources Canada, Monograph 874, 1981, 410 p.
academic interest. As noted before, if the weld stress-strain curves 4. Molinda, G., and Mark, C. The coal mine roof rating (CMRR)-A
and plate-wire-rock interface properties are known, modeling can practical rock mass classification for coal mines. USBM IC 9387,
simulate the mesh behavior over full range of displacements. 1994, 83 p.
For the square bolting pattern used in the US coal mines, modeling 5. Pakalnis, V., and Ames, D. Load tests on mine screening.
showed that the mesh loads were totally carried by the wires Ontario Ministry of Labour Report, 1983, 16 p.
passing below the bearing plates. In addition to the wire gauge and
spacing, distance between the bolts and the number of wire-plate 6. Tannant, D. D. Load capacity and stiffness of welded-wire mesh.
contacts determine the mesh capacity and stiffness. The modeling Paper in the Proceedings of 48th Canadian Geotechnical
also highlighted the limitations of using small screens for physical conference, CIMM, Vancouver, 1995, pp. 729-736.
150
25th International Conference on Ground Control in Mining
7. Thompson, A. G., Windsor, C. R., and Cadby, G. W.
Performance assessment of mesh for ground control applications.
Paper in the Proceedings of International Symposium on Rock
Reinforcement Practice in Mining, Kalgoorlie, Balkema,
Rotterdam, 1999, pp. 119-130.
151