Search Documents

By Yvette Heritage

"The risk of fatalities from rib failure is still prevalent in the coal mining industry. This risk has prompted further research to be conducted on rib deformation in order to understand the mechanisms of rib failure, with the long-term objective being to improve rib support design. This paper presents the results of ACARP research project C25057, which investigated the mechanics and drivers of rib failure. The results of rib deformation monitoring at three different mines in Australia provides rib deformation characteristics for overburden depths ranging from 160 m to 530 m. Monitoring includes deformation during development drivage conditions and during the longwall retreat abutment stress environment. The rib deformation monitoring covered three different seams: the Goonyella Middle Seam, Ulan Seam, and Bulli Seam in the Bowen Basin, Western Coalfield, and Southern Coalfield, respectively. The observed mechanisms driving the rib deformation ranged from bedding shear failure along weak claystone bands to vertical shear fractures to kinematic failures driven by shear failure dilation. The variation in mechanisms of rib failure, together with the seemingly consistent method of rib support design, prompts the question: What exactly is the role of rib support?INTRODUCTIONThe current risk of rib fatalities in the coal mining industry has prompted further research into understanding the mechanisms of rib failure. Three Australian mines assisted further research by providing sites for rib deformation studies as part of the Australian Coal Association Research Program (ACARP) project C25057. The mines included in this study were Moranbah North Mine, Ulan West Mine, and Appin Mine. This paper provides key results of the rib deformation monitoring relevant to assessing the mechanisms of rib deformation.The approach of monitoring rib deformation and rib support loads was used to measure and characterise the dynamic rib deformation present in both development and longwall retreat stress environments. The characterisation of rib deformation, together with rock failure modelling, was used to determine the mechanisms for failure within the rib and to assess the interaction of rib support with the failure process. The mechanisms of rock failure are considered an important factor in understanding the drivers for the failure. Consideration of the site-specific failure mechanisms can provide for a more tailored support design."

Jan 1, 2018

By Gary R. Corbett

The federally owned Cape Breton Development Corporation (CBDC) mines approximately 2.5-3.0 Mt of coal per annum from its Phalen Colliery. As part of an ongoing process to become more commercially viable the Corporation has implemented changes in its support method, namely, the transition from traditional steel set inseam supports to roof bolts as primary support in the single entry longwall retreat drivages. During 1991 and 1992, a monitoring program was implemented by the Cape Breton Coal Research Laboratory (CBCRL) to determine the loading on support systems and ground reaction in the gateroads in front of the retreating longwall faces of the 4 East and 5 East panels at Phalen Mine. The 4 East gateroad had been supported by steel sets alone. The 5 East gateroad had been supported by steel sets as well as roof bolting and strapping of the gateroad roof. Steel sets in both gateroads were instrumented with strain gauges and load cells to record support reaction to the retreating face abutment pressure. Multipoint borehole extensometers and multipoint strain gauged roof bolts were installed in the gateroad roof to monitor strata displacement and bolt loading above the gateroad as the longwalls retreated. This paper describes the details of the instrumentation and monitoring techniques utilized during this program and discusses the results, trends and differences observed in the two differently supported access roadways. Work on future data acquisition instrumentation is also briefly described.

Jan 1, 1993

By Simon H. Prassetyo

Violent failures of coal pillars, known in practice as coal mine bumps, have long been a subject of investigation. Many field investigations have considered geological conditions that create high stress in the pillar to be the main causative factor leading to bumps. In recent years, constraint at interfaces has been observed to be associated with the occurrence of coal bursts. This paper shows how interface friction and width-to-height (W/H) ratio affect the degree of violence of coal specimen failure. A series of UCS tests on 178 coal samples from bump-prone (Utah, called UT coal) and non-bump-prone (West Virginia, called WV coal) mines were conducted. Three violent failure parameters? peak sound pressure level, core zone failure, and ultimate stress? were used to assess the violence of failure. The degree of violence were investigated at varying interface frictions (high: µ = 0.40, medium: µ = 0.22, and low: µ = 0.13) and W/H ratios (W/H = 2, 3, 4, 5, 6, 7, 8, 10, and 12). A simulation series of modeled coal specimen under uniaxial loading was also conducted using the finite element method to observe the equivalent plastic strain distribution experienced by the modeled coal specimens. It was found that the violence of coal specimen failure depends on W/H ratio and interface friction. Interface friction has more significant influence on the violence of failure at high W/H ratios than it does at low W/H ratios.

Jan 1, 2011

By Francis S. Kendorski

Rock reinforcement has been in widespread use and generally has been accepted in underground mining and tunneling since the 1950s. The first rock reinforcement technologies employed were mechanical anchors such as split wedges or expansion shells with 5/8-inch-diameter steel bolts. Failures occurred in weak strata which provided poor anchorage or in ground with corrosive waters. Friction rock fixtures consisting of relatively thin-walled tubes have been in use for about 15 to 20 years. While generally performing adequately, longevity problems have developed from corrosion in water bearing ground Longevity of rock reinforcement is much enhanced with grouted bolts. Portland-cement-grouted rock reinforcement has been in use since the mid-1950s, primarily in funneling and other civil engineering underground construction. Tests of decades-old installations have revealed few problems except in ground with aggressive waters. Polyester-resin-grouted rock reinforcement was introduced in the United states to musing in the late 1960s and to tunneling in the early 1970s. Experience from 30 years of resin-grouted bolt installations and field tests have identified longevity problems associated with degradation of steel reinforcing bars, but in generally unusual situations. Improvements continue in resin chemistries, packaging, corrosion protection, grout quantities, and mixing and distribution in the drilled hole to achieve long-term performance. Specific case histories cited with resin- or Portland-cement-grouted rock reinforcement longevity or performance problems, upon close examination, reveal that the causes of the problems were qualify control procedures being inadequate and not in accordance with good practice. Manufacturers' recommendations and engineering specifications, if followed in the field, and with competent inspection and supervision, would have prevented must, if not all, reported longevity or performance difficulties.

Jan 1, 2000

By D. Reddish

The combination of free standing steel supports, such as arches, and rock bolts (mixed support) has Frequently been in used in UK coal mine roadways when it is considered that the strata and/or environmental conditions would not be suitable for rock bolts solely. However, how such mixed support systems operate is not fully understood. Due to the differences in the support method provided by rock bolts and free standing steel, the total support generated is not simply a case of adding the contribution to the support of the roadway given by the two systems. This paper describes a numerical investigation into mixed support systems when used in coal mine roadways. A series of 2D and 3D geomechanical models are described and the method of modelling the rock bolts and steel supports in coal mine tunnels is discussed. The numerical models described include rock boltd only roadways, steel supported roadways and mixed supported roadways in a variety of geological and in-situ stress conditions. Conclusions are formed that allow an insight into mechanisms of strata reinforcement provided by the support systems modelled. Further to this, a review of the numerical modelling of pre-tensioned rock bolted systems is described identifying the contribution to support efficiency given by the adoption of tensioning in full column resin anchors. The results obtained highlight that the effectiveness of the pre-tensioned load is dependant on the rock mass quality and in- situ stress level at the site where it is to be used.

Jan 1, 2004

By S. J. Mitchell

An analysis of bolting reinforcement of several large [approximately 18 m (60 ft) cube] pillars was performed. The many overcoring stress profiles in pillars at the mine were used to produce generalized stress distributions at various stages in pre- and post-failure. A complete stress-deformation curve was estimated from this data. The effect of pillar bolting on the load-deformation behavior was extrapolated from the observed unbolted behavior. Bolting can increase pillar strength by up to 10 percent. The effect of bolting on general mine stability was examined with a shoW computer analysis. The computer model was calibrated against measurements performed at the mine, and produced good agreement between measured and modeled stresses. The analysis indicated that bolting increases the bolted pillar safety factors in the event of additional pillar failure by a small amount (approximately 5 percent). Progressive pillar failure is unlikely, because arching transfers most load from yielding pillars to large unmined abutments rather than to the smaller panel pillars.

Jan 1, 1984

By Daniel W. Su

This paper presents the implementation and evaluation of the hydraulic fracturing technique and Longwall Visual Analysis (LVA) software to mitigate the impact of a 1,000-ft (305-meter) widemassive sandstone channel on a 1,500-ft-wide (457-m-wide) longwall face. Based on a underground roof geology reconnaissance program, four frac holes were drilled and fraced along the center axis of the sandstone channel. To further provide detailed monitoring of the longwall face, the Longwall Visual Analysis (LVA) software was installed to track the face pressure and cavity formation index. In mid-December 2011, the longwall face approached and entered into the western edge of the sandstone channel. The longwall face mined through the center axis of the sandstone channel in mid-January 2012 and advanced outby the eastern edge of the sandstone channel in early February 2012. Results from daily underground observations and production delay analysis confirmed the effectiveness of the hydraulic fracturing program and the validity of the LVA Software.

Jan 1, 2012

By Joe Zelanko

In order to prevent roof fall accidents and injuries, U.S. underground coal mine operators must develop and implement effective roof control strategies. Various tools are available to assist engineers in developing designs. For example, criteria to address global stability issues (e.g., pillar stability) are well established in the U.S.; effective tools (e.g., empirical and numerical) are available to diagnose potential problems and avoid catastrophic failures. However, most miner injuries occur as a result of local instability (e.g., a roof fall in a single heading or a rib roll from a pillar corner). Hazard recognition and the availability of support plans suitable to those conditions are key to maintaining stability and safety on the working section.

Jan 1, 2011

By Charles D. Albright

FMC built its first angle bolter in 1969. This was a single mast unit which could drill holes for conventional bolts or angle holes for installing roof trusses. From this drill we learned very quickie that the mast design was an important factor to the success or failure of any angle bolter. The first mast installed on this machine (EXHIBIT 1) was fixed in length with endless chain and hydraulic motor feed. It took less than a shift underground to determine that the fixed length was not going to work. The undulating top created conditions where the mast was too long to be positioned at a 450 angle. Several mast designs were tried and discarded before we felt we had the answer to our problems. (EXHIBIT 2} We developed a mast that would extend or retract without moving the position of the drill unit. This allowed the operator to "wedge" the mast between the floor and roof before starting his drill feed. This arrangement worked exceptionally well, but unfortunately the demand, as perceived by our Sales Department, for angle bolters did not materialize. This unit became an "orphan" and was used by many coal mines until it died a natural death. FMC did not decide to re-enter this market again until 1979. At this time many coal producers were contacted to determine what features should he incorporated into a dual mast angle bolter. From this information we were able to arrive at three main areas of concern that we should consider when designing a new machine. 1. Eliminate overturning moment on drill unit. 2. Provide some sort of steel support near roof line to aid in starting a 15° hole. 3. Dual mast units need a method of raising center of mast rotation to facilitate various mining heights and still maintain a 45° angle and proper distance from rib. I would now like to discuss each of these individually. EXHIBIT 3 indicates the overturning moment resulting from the feed cylinder thrust and the reaction of the drill hit against the roof. To eliminate this moment a mast was designed using two feed cylinders mounted in the plane of the drill center line, thus removing all overturning moments on the drill unit carriage. This mast has been used extensively in vertical drilling with no downtime attributed to wear on mast guides and drill unit supports. The cylinders are designed to provide uniform thrust and feed speed throughout the total stroke. For manufacturing and economic reasons, the final mast design on the first drill was not used for this new drill. A drill steel support mounted on the mast, and positioned by a hydraulic cylinder provides the needed guidance near the roof for starting a hole at a 45° angle. (EXHIBIT 4) This support does not surround the drill steel but simple cradles the drill steel for "collaring" the hole. Once the hole is started, the support can he lowered. The support plate is hinged to allow the drill mast to pass. The advantage of having vertical motion of the center of rotation on a dual mast drill is shown in EXHIBIT 5. Following the criteria recommended by Birmingham Bolt Company, the truss bolts should he at a 45° angle and one fifth of the width of the entry from the rib. Therefore, in an entry 20 ft. wide, the hole should be drilled 4 ft. from the rib and at the 45° angle previously mentioned. With the limitations inherent to dual boom drills, that is, the masts cannot pass each other nor can one mast pass the center line of the machine even if the other is positioned out of the way, it is impossible to accommodate much variation in seam height and maintain the recommended criteria. As shown in EXHIBIT 5, a center of rotation 17 inches above the floor will start a hole 4 ft. and 45°angle at a height of 67 inches. Any height above 67 inches

Jan 1, 1982

By C. C. White

Much benefit can be achieved by a mining company and a manufacturer if both parties can communicate effectively with each other and assist in product development; mutually beneficial ideas and products will develop. This communication can best be achieved if both parties have the opportunity of having on their respective staffs, specialists whose function is to improve the use of products used in mine development. Although a great wealth of roof control knowledge has long existed among mine operators, until recently there was not an abundance of joint development of roof control material by manufacturers and mine operators. This was perhaps due, in past lean years, to the preoccupation of mine management with other overwhelming problems that arise in mining. Manage¬ment had little time for the lengthy process of product development. Until recently, only a few of the largest mining companies had fulltime personnel devoted to roof control. A bolt manufacturer who is invited to work with an operator's roof control people in developing new material is lucky indeed in that he learns what is wanted and he acquires the benefit of the operator's knowledge and experience. He has only to come up with a satisfactory product at an acceptable price - which is not always easy: The development of new products is usually costly and time consuming, and many new products fall by the wayside. In the case of point anchors, however, this product is here to stay. The advantage of point anchors over mechanical anchors is now a proven fact. This advantage is due to less bleed off and higher ultimate load capacity of the resin point anchors. Point anchored bolts have proven to be superior to fully grouted bolts in many areas. We at Birmingham Bolt Company are fortunate in that we have been involved in joint efforts to develop point anchors with several coal companies from whom excellent advice and direction was received. It is appropriate to mention before continuing that the use of point anchors began in the 1960's in Europe and became widespread only recently in the United States. They fall in two categories, point anchors and combination point anchors. These categories have been defined by the Bureau of Mines as follows: (1) point anchors having 24" of grouted length and (2) combination anchors having 24" or more of grouted length. In the present nomenclature, all are anchored with resin or some sort of cementing compound between the rebar or tensioned element and the wall of the hole. The latest development includes an expansion shell as well as rebar, the purpose of the expansion shell being to facilitate installation without the delay necessary in use of types that require a waiting periond for setting of the resin before tightening. (See figures #1 and #2.

Jan 1, 1982

By Steve P. Signer

Researchers from the Spokane Research Laboratory of the National Institute for Occupational Safety and Health installed 39 instrumented, fully grouted bolts at six test sites in a trona mine retreat panel to study mine roof stability for the improvement of workplace safety. Variables at each test site included bolt spacing, bolt length, roof span, and location in the panel layout. At most test sites, two rows of instrumented bolts were installed, one in or near the intersection created during development and the other in or near the intersection created during second-pass mining. The most significant factor affecting bolt load was roof span. The highest loads were on the bolts installed in the intersections of the 6 m-wide entry. Mining-induced stress resulting from panel layout was the next most significant factor affecting bolt loading. Minor varia¬tions in bolt loading could be attributed to changes in bolt spacing and bolt length. Gas pressure release in the immediate roof contributed to some bolts showing compressional loading during gas bleed-off.

Jan 1, 2001

This paper describes the experiences with the application of the Coal Mining Roof Rating (CMRR) System to South African strata conditions. Roof classification studies using the CMRR system were undertaken as part of a SIMRAC investigation into falls of ground in South Mean collieries. Colliery roof rating exercises were undertaking in a range of geotechnical conditions from shallow level coal seams under flat topography to workings with cover depths of greater than 300m A comparison is made between the CMRR and other Rock Mass Rating systems used to classify different roof geotechnical conditions. The need for geological discontinuity orientation, blasting and rockmass stress adjustments to the CMRR is discussed. The effects of acid mine water on strata weatherability and the need to incorporate immersion tests using acid mine water are also detailed. In addition limited correlation is given between CMRR values and other geotechnical factors such as joint trace length, joint shear strength, joint orientation, weatherability and roof damage due to horizontal stress. Taking cognizance of these observations, a concept of stress induced roof damage triggering is discussed.

Jan 1, 2001

By P. Palmer

Using 3D numerical modeling as a tool to predict rock behavior during mining extraction is still an uncommon practice at many underground operations. This paper will discuss how 3D numerical models were used to identify potential underground instability problems in the later stages of mining. Detailed modeling was then used to make modifications to the mine design in order to minimize the impact of the potential instability problems. This paper will also discusses how the back analysis of a previous underground collapse event at the same mine in a room and pillar area was used to identify the instability problems for a new orebody being developed deeper in the mine. A 3D numerical model was first created in the MAP3D software to determine the conditions of the pillars and the magnitude of the vertical displacement above the mining zone prior to the underground collapse. The results from the ground collapse model were then used as a threshold limit for predicting potential instabilities and developing a model for the new mining zone. sing the results from this analysis, mine management has been able to implement a much higher percentage recovery of the reserve in a new, high-grade area of the orebody while minimizing potential instabilities in the future.

Jan 1, 2005

By Jun Lu, Mark Van Dyke, Daniel W. H. Su, Greg Hasenfus

"This paper presents the geologic and ground control challenges that were encountered by CONSOL Energy’s mining operations in southwestern Pennsylvania. A sandstone-to-limestone geology transition was encountered by Mine A, where the primary 8-foot bolts were anchored in claystone. Upon experiencing roof sags and a roof fall in the return entry, 6-foot high-tension, fully-grouted primary bolts were employed along with 16-foot center cable bolts at every other strap to improve beam building and to ensure proper anchorage into the competent limestone. No longwall delay was reported within this geologic transition zone, although there was some difficulty in maintaining an open #2 entry airway behind the longwall face.The geologic encounter in Mine B was more complex, where fluvial deposits of massive sandstone channels, shale channels and pyritic-rich, green claystones were encountered on both sides of the mains, and which impacted both development and longwall ground control. In particular, poor transitional roof areas near the edges of the sandstone and shale channels often posed significant ground control challenges. In addition, hydraulic fracturing was sometimes utilized to enhance caving of massive sandstone behind the shields to relieve pressure at the face. Longwall face conditions and advance rates were documented to illustrate the effects of sandstone channels, shale channels, and green claystones on safety and productivity and to demonstrate the effectiveness of the hydraulic fracturing.Shale channels, slickensided zones, laminated roof zones, soft floor, and a regional syncline were encountered in Mine C, the combination of which posed unique ground control challenges during both development and longwall mining. The presence of soft floor, coupled with the presence of thick floor coal and deep cover, resulted in headgate convergence during retreat of the first right-hand longwall panel. Two-piece 8-foot bolts were sometimes observed to fail at the coupler within the laminated roof zone, and 6-foot, 1-piece, fully-grouted point-anchor bolts were employed as the primary bolts thereafter."

Jan 1, 2015

By Al Campoli

One of the critical variables affecting resin bolt performance is the size of the resin annulus, defined as the difference between the hole and bolt radius. The vast majority of the millions of resin roof bolts installed each year in the United States employ an annulus between 0.125 and 0.188 inch. As the annulus size decreases the resin flow around the bolt becomes more difficult, increasing the resin pressure induced by bolt insertion. As the resin pressure increases the loss of resin into bedding planes and vertical fractures in the rock increases. As the annulus size increases the mixing of the resin paste and mastic becomes less efficient. Thus, the lower and upper limits of the optimum annulus size is determined by insertion pressure and mixing considerations, respectively. Previous U.S. Bureau of Mines researchers have concluded that a 0 125 inch annulus is optimal. However, others have recently reported that a 0.095 inch annulus with special rolled deformation (SRO) bolts provide superior anchorage to standard rebar with a very large 0.250 inch annulus, attributing SRD resin loss in weak ground to oversized holes. The effect of annulus size on unit length pull strength and resin loss to fractured rock was investigated. Short encapsulation pull tests with resin annuluses of 0.125 and 0.095 inch resisted an average of 2.25 tons per grouted inch, with no significant variation. However, annulus size has shown a dramatic effect on the magnitude of resin lost to simulated rock fractures. No. 5 standard . No. 6 standard, and No 7 J-Bar rebar were inserted into two different viscosity resin cartridges designed for 4 foot boll encapsulation into a one inch inside diameter metal tube, To simulate a rock fracture, a 0.250 inch diameter orifice was located near the capped end of the tube Resin loses increased as annulus size decreased, with a minimum loss of 35% and a maximum loss of 60% associated with 0.125 and 0.095 inch annuluses, respectively.

Jan 1, 1999

By Daniel W. H. Su

This paper presents a retrospective assessment of ground control research and its successful application by CONSOL Energy over the past 25 years. Recognizing the need to improve the understanding of coal pillar design mechanics under shallow, medium, and deep covers, an extensive underground coal pillar instrumentation program was initiated in the mid 1980s. Combined with sophisticated numerical modeling technique and laboratory tests, results from this instrumentation program were used to advance the concept of coal pillar design and to demonstrate that numerical modeling can be used to predict in situ coal pillar strength, especially under non-ideal conditions where interface friction and roof and floor deformation are the primary controlling factors. To evaluate the impact of longwall mining, a 10-year subsidence research program was initiated in 1986 that encompassed surface, subsurface, and underground measurements, as well as a sophisticated numerical modeling technique. Results from the research revealed that interface friction and shearing between rock layers within the overburden had a profound effect on surface subsidence development and the resulting strain. Recognizing the impact of horizontal stresses on CONSOL?s mining operations, extensive horizontal stress measurements were conducted using the USBM Borehole Deformation Gage and the CSIRO Mininfrac system in the mid to late 80s. Roof extensometers and CSIRO HI Cells were used to measure roof displacement and roof stress concentration in longwall headgates and tailgates, which resulted in the definition of right-hand and left-hand longwall panels. Pillar design modifications, combined with cable bolt application and innovative standing support installation, were then successfully employed to mitigate horizontal stress impact on the right-hand longwall headgates and left-hand longwall tailgates. An in-house floor bearing capacity test apparatus was fabricated and employed underground to determine mine-specific floor bearing capacities which, along with other geotechnical parameters, were then used in the mine-specific shield design. To promote caving of thick, massive sandstone immediately behind longwall face supports and to improve face conditions, the hydraulic fracturing technique was successfully employed by CONSOL to pre-fracture the sandstone prior to longwall mining. Face monitoring and numerical modeling confirmed the effectiveness of the hydraulic fracturing technique. A roof geology reconnaissance program was initiated by CONSOL in the early 1990s to improve roof control, safety, and production. Information obtained from the reconnaissance program was used successfully to guide mine planning decisions and daily roof support practices.

Jan 1, 2011

By Thomas P. Mucho

To document the performance of mine roof and roof support systems the U.S. Bureau of Mines (IJSBM) has been installing instrumentation at selected sites in U.S. coal mines. Much of this support and roof instrumentation has been installed in high stress conditions to maximize differences in performance. Summarized in this paper are the results of four such site investigations. The roof geology at the four sites is quite different and is quantified using the USBM's Coal Mine &f Rating (CMRR). The studies included detailed measurements of roof movement using multi-point extensometers, as well as measurements and monitoring of bolt loading. The investigations include developmental loading and abutment loading from longwall and room-and-pillar mining operations. Some of the issues examined are the effect of the horizontal stress field, the effect of installed tension (torque-tension versus resin bolts), the effect of reduced annulus bolt holes, and the differences between similar bolts supplied by different manufacturers.

Jan 1, 1995

The U.S. Bureau of Mines (USBM) has developed the coal Mine Roof Rating (CMRR) to aid in ground control. This paper describes the application of the CMRR at four coal mines operated by a major U.S. coal company. The mines are widely separated geographically, and cover a wide variety of geologic conditions and mining methods. The study found that the CMRR was a useful predictor of roof performance. In mines where the CMRR varied, and other parameters held constant, roof falls were associated with low CMRR values. Important similarities in roof performance were found in mines with similar CMRR values, despite radically different geologic environments. The analyses also indicated that statistics can be used to evaluate ground control performance. For these mines, a "roof fall probability" can be estimated from the CMRR and the intersection span. A spreadsheet program has been developed to facilitate treatment of CMRR data, and it is briefly describe

Jan 1, 1994

By Christopher Mark

The demographics and safety record of the pillar retreat segment of the U.S. coal industry was analyzed using statistics collected by the Mine Safety and Health Administration. Pillar recovery is practiced primarily by mines in Appalachia and the Midwest. Using 1993 data, the accident rates and productivity of a large sample of pillar re- mines were found to be similar to other room-and-pillar mines in the same geographic areas. Pillar recovery apparently accounts for about 10% of all U.S. underground production, but has been associated with about 25% of the mf and rib fatalities during 1989-96. However, of the 28 fatalities that were analyzed, only 4 occurred for which no citations were issued for violations of mining law. Nearly one- half of the fatal incidents occurred during the mining of the last lift or pushout. All four no-citation incidents occurred during the removal of the last lift during a "Christmas tree" extraction sequence.

Jan 1, 1997

By Andrew Starkey

Ground anchorage systems in the form of rock bolts are used extensively throughout the world as supporting devices for civil engineering structures such as bridges and tunnels. It is widely recognized that non-destructive testing methods for ground anchorages need to be developed as a high priority [1 ], with only a small proportion of anchorages currently being monitored in service [2]. The condition monitoring of anchorages is a relatively new area of research, with the objective being a wholly automated or semi-automated condition monitoring system capable of repeatable and accurate diagnosis of faults and anchorage post tension levels. The GRANIT tm system developed by the University of Aberdeen is capable of measuring the integrity of ground anchorages and has won two major UK Awards - the National John Logic Baird Award for Innovation in 1997 and a Design Council Millennium Product Award in 1999. The normal procedure with the GRANIT tm system [3,4] is to train a neural network on dynamic response data that has been sampled from an anchorage over a range of post-tension levels. Further data is needed in order to test the neural network. The paper presents neural networks that have been trained on data taken from a number of anchorages on different sites, and the results of the diagnosis of these neural networks on new test data is presented. The results show that the cross-diagnosis of anchorages is possible, where a neural network trained on one anchorage is capable of successfully diagnosing the post-tension level for an anchorage of similar construction, and the high level of accuracy that can be attained with the GRANIT tm technique is shown. The paper also introduces a mathematical model of the GRANIT tm process.

Jan 1, 2001