Search Documents

By L. Forbes, G. Forbes, J. Wang

Bubble size measurement is a vital part of flotation system analysis and diagnostics. Over the years, a great many different bubble size analysis tools have been developed. However, all of them rely on ex-situ measurement and in-pulp sampling techniques. This paper evaluates the Mettler Toledo ParticleView™ Measurement probe as a novel method for in-situ bubble size measurement. This device is then compared to the conventional Anglo Platinum Bubble Sizer (StoneThree). It was found that the in-situ bubble size analysis technology appears to be a more user-friendly and reliable option for determining bubble size in flotation. However, the current instrument design is not suited to industrial applications. Keywords: Bubble size measurement, bubble coalescence, PVM probe

Jan 1, 2020

By R. G. Votteler, A. C. Brent

"The first in this series of two papers, on the potential of renewable electricity sources for mining operations in South Africa, investigated the internal structure of mining corporations to evaluate feasible alternative electricity sources that were identified as through earlier research. The purpose of this paper is to combine current knowledge about the external macroeconomic and the internal environments to produce a clear picture of how renewable sources of electricity could perform from the perspective of mining corporations in South Africa. The multi-attribute value theory (MAVT) approach was adapted to structure the research and results. The model was fed with real-time data provided from five different energy companies and four mining corporations operating in South Africa. The results show that the performance of hybrid versions of the currently used electricity sources (diesel generators and Eskom grid connection) with solar PV and onshore wind is favourable compared with the current sources alone. The advantage of diesel generators is significantly greater than that of the Eskom grid connection. By combining the macroeconomic influences with the MAVT results, hybrid solar PV versions are identified as having the greatest potential. In second place are hybrid wind solutions, which have the shortcoming that good wind conditions occur only in coastal regions where there are fewer mining activities. Geothermal hybrid versions are the least favourable owing to the lack of service infrastructure and high initial investment costs.IntroductionRecent developments in the macroeconomic environment necessitate an investigation of the internal structure of mining corporations to evaluate possible alternative electricitygenerating sources (Roehrl and Riahi, 2000). To be able to combine current knowledge of macroeconomic factors and the internal structure of mining corporations, the research process is presented in two papers. The first paper (Votteler and Brent, 2016) investigated the internal structure and argued that a strategic tool should be used to evaluate electricity sources from the perspective of mining corporations. The multi-criteria decision analysis (MCDA) approach was selected as the most appropriate strategic tool. Based on a literature review of previous similar MCDA approaches in energy planning, it was concluded that no adaptation of MCDA methods could be found from a corporate, mining, or South African perspective. Finally, the paper investigated the internal evaluation structure in collaboration with mining corporations, by identifying the criteria they use to evaluate electricity sources."

Jan 1, 2017

By D. S. Loubser

The technique of backfilling in the gold mining industry is dependent on the dewatering of slurry to form a solid support system. In many cases flocculants are added to the slurry to increase the rate of dewatering. To ensure that a stable support is obtained it is important that the chemistry, application and limitations of these flocculant additives are understood. Flocculant additives have to be added at a point where minimum shear will occur. (Shear occurs when a flocculant is exposed to excess energy resulting in breakdown, of the floc.) To ensure that optimum results are obtained it is important that personnel involved are trained to use the flocculant additives and dosing equipment correctly.

Jan 1, 1988

By C. J. Alberts

The perception exists in many organizations that, once an energy management system is employed, there need not be any further adjustments to such a system. This is not true, as the field of energy management is a dynamic field, and constant feedback and adjustments are needed to enhance the efficiency of a system to the point that substantial pay-back can be obtained. It is stressed that for any energy management system to be successful, careful planning should permeate through the organization, from the top level decision-makers to the workplace. This paper describes a dynamic technique by which energy management strategies can be evaluated by viewing the effects of the strategies on the organization as a whole.

Jan 1, 1997

By J. N. Van der Merwe

"The probability of survival of coal pillars is a more rational indicator of stability than the safety factor. Over time, pillars undergo reduction in size by progressive scaling of the sidewalls. The probability of survival is therefore not constant but reduces over time. The rate of scaling can be quantified, and the reduction of the probability of survival as a function of time can therefore also be quantified. The paper describes the link between the factor of safety, probability of survival, and reduction of the probability of survival. Using the safety factor, probability of survival, and the time at which the probability of survival will reduce to 50% is proposed as a more rational method of evaluating pillar stability than the safety factor alone.IntroductionCoal pillar strength has long been the subject of research in South Africa and several other coal-producing countries. However, it began receiving serious attention only after the Coalbrook disaster in 1960. The most important outcomes of the initial investigations were the groundbreaking research by Salamon and Munro (1967) that resulted in the well-known power formula for pillar strength, and the work by Bieniawski (1969) that resulted in a linear formula for strength. The former approach was based on empirical back-analysis of failed and stable pillar cases, and the latter on in situ strength tests on large specimens.The Salamon and Munro (1967) formula has been widely used in South Africa and even adapted for application to hard rock pillars by researchers such as Hedley and Grant (1972).At the time of the initial work by Salamon and Munro, there were only 27 failed pillar cases suitable for analysis. It was not possible at the time to determine site- or seam-specific strength, and the result was a general formula for use in all seams in all mining districts. Since that time pillars have continued to fail and the database continued to grow – as of 2013, there were 85 cases in the database. Using the additional data, there have been a number of re-evaluations of the strength formula and it has also become possible to differentiate between the different mining seams and areas."

Jan 1, 2016

By Q. Zhang, X. C. Wang, S. Mao

Phosphate ore is an important strategic mineral resource in which apatite is the main valuable mineral. Flotation is the most commonly used method to separate apatite from middle-low grade phosphate rock. Selective flotation separation is based on the difference of physical and chemical properties of minerals occurring at the three interfaces of mineral/bubble/water. Surface and interface chemistry of mineral have been playing an important role in interpreting and governing its flotation phenomenon. In-situ, visualized and high-precision analytical techniques, such as Time-of-Flight Secondary Ion Mass Spectrometry, Atomic Force Microscopy and Molecular Simulation, have been applied to explore and uncover microscopic nature of apatite flotation. Consequently, a great progress at molecular or atomic level on surface and interface chemistry of apatite flotation has achieved in the past decades. In this paper, the research progress of surface and interface chemistry of apatite flotation such as surface wettability, surface-charging characteristics, surface dissolution behavior, adsorption mechanisms of fatty acid collector and depressor on apatite surface were systematically reviewed. This review will be helpful for the development of more selective reagents and efficient technologies for apatite flotation. Keywords: Apatite, Flotation, Surface and interface chemistry, Surface analysis technology

Jan 1, 2020

By H. W. Read, L. G. D. Napier

Change is endemic in our society. Future change will be accelerated by an explosion in technology, communication, and political thought. To meet these challenges, rapid structural changes have been taking place in mining organizations both here and abroad. The management of shaft-sinking projects is a case in point. Shaft sinking is a seminal element in the sustained well-being of a deep level mining venture. Its accomplishment is fraught with risk, involves a high level of capital expenditure, and is a precursor to profitable mining investment. The purpose of this paper is to present a concise and practical picture of the concepts, methods, and techniques of professional project management as applied to shaft-sinking projects.

Jan 1, 1994

By V. G. Pienaar, B. M. Mbiya, P. T. Slatter

[a constant b constant cc onstant c' constant D pipe diameter (m) Dshear shear diameter (m) E error function fth theoretical friction factor F function F' function K fluid consistency index (Pa sn) kv valve loss coefficient L pipe length (m) n flow behaviour index N number of data points ?P valve pressure drop (Pa) Q flow rate (m3/s) Re Reynolds number Re3 Reynolds number (Slatter) V average velocity (m/s) Vann annulus velocity (m/s) x variable X number of independent variables y variable Y number of fundamental dimensions z variable µeff effective viscosity (Pa s) ? valve degree of opening d partial derivative ? fluid density (kg/m3) [ ] function]

Jan 1, 2007

By J. A. C. Diering

The determination of stresses and displacements close to tabular excavations is of considerable importance in the planning of tunnels and of remnant and pillar layouts. This paper gives a description of a simple yet effective method for the achievement of improved values from an analogue or MINSIM type of analysis. Essentially, closures and rides (the movement of hanging-wall relative to footwall) in mined-out areas are assumed to vary in a piecewise linear fashion as opposed to a piecewise constant fashion. As the former fashion more closely resembles the true variation, it is possible for the stresses and displacements at points close to the excavations to be calculated with much better accuracy.

Jan 1, 1980

By W. A. Hodgson

"In order to be able to map the variation in the concentration of elements in soils, it is necessary to sample at intervals that will give adequate coverage of the area and will allow determination of the rate of fluctuation of concentration with distance in any environment. This rate of fluctuation with distance can be referred to as the continuity of concentration.The continuity of concentration in soils varies according to the chemical characteristics of the element being considered, the underlying rock type or orebody and in response to the dispersion processes that operated during soil formation. Within any environment there is a maximum distance between sampling points within which adjacent sample values are related. This distance, the critical sampling interval, can be determined by means of a correlograrm.Where the sampling grid has been chosen to provide an adequate coverage over unmineralized rock, adjacent samples being related, anomalies may be missed if the element continuity in the anomaly is less than in the background. Results show that a knowledge of the critical sampling interval of the environment being examined may used to select a more efficient sampling pattern for a geochemical survey."

Jan 1, 2014

A General Meeting and Colloquium was held in Kelvin House, Johannesburg on Wednesday 29th September, 1971. The theme of the Colloquium was "The Ferro-alloy Industry". Prof D. D. Howat (President) was in the Chair. The meeting was attended by 106 delegates and was opened by the President at 10.05 a.m. OBITUARY The President: "It is my sad duty to announce the death of G. C. Needham, Fellow, who joined the Institute in 1966 and passed away on 27th September, 1971. As a mark of respect to the memory of the deceased and in sympathy with the bereaved I would ask you all to rise and observe a few moments silence." MEMBERSHIP The President: "I have much pleasure in announcing that the names of the following candidates, having been published in accordance with By-Law 5.2.2, Council has elected them to membership of the Institute. Fellows: Richard Lloyd Blandy, Brian Roger Borton, Brian Teversham, Peter Donald Dickson. Members: Frederick Brian Peacock, Peter John Stacey, Kenneth Niles Thanstrom, Byron J. Hayley, Ian Flemming Buchan, Abraham Hercules le Roux. Graduate: Ian William Coy. Students: Peter Boesley J ames, Kenneth Brian Perel. TRANSFERS From Member to Fellow: Cecil Eugene Fivaz, Peter Edwards. From Student to Graduate: Roy Stavely Traviss. From Student to associate: William Mitchell. I welcome the new members to the Institute and congratulate those members who have been transferred to a higher grade. COLLOQUIUM The first paper, presented by Mr S. Selmer-Olsen, Consulting Metallurgist, Amcor Limited, on 'Trends in Ferro Alloy Production' was published in the May issue of the Journal and comprised information and data relating to new developments in ferro alloy production in various parts of the world. This was followed by a paper-'Conductivity and Viscosity Properties of Slags produced in ferro-alloy manufacture' by D. 1. Ossin et al. This was a record of laboratory test work on slags containing both lime and magnesia, corresponding with the composition of slags encountered in the production of ferro chrome alloys when a dolomite flux is employed. The third paper-'Review of the work of the Pyrometallurgical Research Group of N.I.M. and Metallurgy Department, Wits.' presented by Dr P. R. Jochens, leader of the Research Group, outlined the progress made on ten different research projects presently under way in the Department of Metallurgy. These projects range from studies of the kinetics and mechanism of the reduction of chromite ores through the determination of the activities of MnO in slags to phase studies and melting point determinations in the alloy systems Cr-Fe-Si-C and Cr-Fe-Si. A new feature of this Colloquium was the introduction of the discussion panel composed of four or five individuals known to have special knowledge or interest in the subject under duscussion. Mr J. J. Coetzee was chairman of the first panel, Professor D. D. Howat of the second and Dr R. E. Robinson of the third. During the discussion of Mr Selmer-Olsen's paper an interesting contribution was made by Mr Jack Hayley of Transalloys (Pty.) Limited who pointed out that the lowest power cost in the U.S.A. was that from the TV A at 6 mils per unit (0,43 cents S.A.) while costs from other sources were 7 to 8 mils per unit (0,50 to 0,57 cents S.A.) He also referred to the use of 'pile sintering' of ore fines plus coke at certain ferro alloy plants in the V.S. and to the recent practice of 'graphitising' the paste used for Soderberg electrodes. The last session in the afternoon was devoted to a group discussion on 'Practical problems in Arc Furnace Operation', Mr H. R. Reid of Middelburg Steel Alloys (Pty) Limited occupying the chair, members of the panel being Mr J. Meintjies (Rand Carbide Limited), Mr J. Benn (Amcor) and Mr E. J. Jackson (Transalloys (Pty.) Limited). Mr Reid initiated the discussion by talking of the design and composition of electrode shoes, mentioning the use of the 85/15 Cu/Zn alloy and the comparison of hollow shoes and shoes cooled by tubes. Mr Meintjies discussed electrode management with particular reference to breakages, and dealt with casing and fin design and the welding of fins. Mr Benn dealt with aspects of closed top furnace operation with particular reference to the control of eruptions in the furnace and indicated that preheating of the raw materials might be well worth considering together with the development of a system for the detection of the water vapour present. Mr J ackson was concerned with shells, linings, refractories and tap holes. He deprecated the use of oxygen for opening the tap hole emphasising how uncontrolled use of oxygen led to loss of shape of the tap hole and quite severe burning out of the front hole resulting in time and labour in repairs. A notable feature of the Colloquium was the free discussion 'from the floor' of the hall. Many valuable and highly pertinent contributions were made spontaneously and usually in reply to some controversial topic raised by the discussion group.

By Jean Michel Marino, Jean-Pierre Slama

"The computer methods used by the French Commissariat a l'Energie Atomique, which are applicable to any type of ore deposits, are:(i) Ore reserve estimation. The data are either the core analyses or gamma leggings. To determine the spatial distribution of the ore, three different methods of interpolation are used. Two of these are based on the least-squares theory and the third one is an application of the regionalized variable theory. The interpretations of the orebody by the geologists are introduced in the computer program as often as possible.(ii) Optimal open pit design. From the spatial distribution of ore, the three-dimensional outlines of the optimal pit are determined by a geometrical algorithm which allows different slopes varying with the azimuth.Finally, the optimization of annual successive plans arc discussed."

Jan 1, 2014

By D. F. Malan, J. A. L. Napier

A number of shallow coal or hard rock mines employ pillar mining systems as a strategy for roof failure control. In certain platinum mine layouts, pillars are designed to ‘crush’ in a stable manner as they become loaded in the panel back area. The correct sizing of pillars demands some knowledge of the pillar strength and the overall layout stress distribution. It is particularly important to understand the impact of the layout geometry on the effective regional ‘stiffness’ of the rock mass around each pillar. An important design strategy is to model relatively detailed layout configurations which include a precise representation of the local pillar layout geometry and to analyse multiple mining scenarios and extraction sequences to select optimal pillar sizes and barrier pillar spacing. Although computational solution techniques are now impressive in terms of run time efficiency, a major difficulty is often encountered in assigning suitable material properties to the pillars and in devising an effective material description of the layered rock strata overlying the mine excavations. This paper outlines an efficient numerical strategy that can be used to assess large-scale pillar layout performance while retaining the ability to modify individual pillar constitutive behaviour. The proposed method is applied to selected layouts to compare estimated average pillar stress values against values determined by detailed modelling and against observed behaviour.

May 3, 2023

By S. M. H. Ahmadi, K. Barani, F. Moradpouri, R. Ghaedrahmati

As exploration is time-consuming, costly, and risky, determination of the erosion surface of a metalliferous deposit before geophysical surveying and exploration drilling might be very helpful. Geochemical haloes can be used to determine whether the erosion surface is supra-ore or sub-ore and thus reduce the risk of exploration operations. The aim of this investigation is to determine the erosion surface of the North ROK porphyry deposit (NRPD) in northwestern British Columbia in Canada using linear productivity (LP), which is the content of an element defining the halo multiplied by the width of the halo. A total of 2045 soil samples from the B horizon were analysed using ICP-MS for 36 elements, including Cu, Mo, Pb, Zn, Au, As, Ag, Ni, Co, Fe, and Mn. The data-set was snalysed to obtain the statistical parameters and the elements Cu, Mo, Pb, and Zn were chosen to calculate the linear productivity and the total linear productivity. These four elements were modelled using probability plots to identify and separate subpopulations in terms of anomalous haloes and background, including the threshold values of each subpopulation. The results of the probability plot modelling and thresholds values were then used to map the distribution of each element in a GIS to calculate the linear productivity. The total linear productivity indicated that the erosion surface is supra-ore. Finally, a 3D orebody model of the Cu, Mo, Pb, and Zn distributions was constructed using borehole data and used to validate the results.

Feb 6, 2023

By H. A. D. Kirsten

Written contribution by L.J. Besaanst I am a Mechanical Engineering Technician (T3), and wish to further my education to at least NHD (T4) level. I read your paper in the Journal of The South African Institute of Mining and Metallurgy with great interest, and must compliment you on its brilliance. However, there are some questions I should like to ask, and certain points I should like to highlight. I should appreciate your comments. Written contribution by J.F. Curtis The comprehensive and finely detailed survey of, and proposals for, engineering education in South Africa reflects the author's wide experience in the needs for, and the provision of, such education. This paper wil make a major contribution to the enhancement of the profession of engineering and the improvement in quality of engineering activities in South Africa. Contribution by G.W. Annandalet I have studied your paper in some detail, and am of the opinion that it makes an important contribution to the topic of the development of technological manpower, which is currently being debated in South Africa at al levels. Debate on this topic has risen above the level of academic discussion; it has been discussed by the Economic Advisory Council of the President, and is being considered at Ministerial level. Author's reply to L.J. Besaans Thank you for your contribution and complimentary remarks. As you have pointed out, there are certainly many aspects that will have to be addressed in the resolution of the crisis in engineering education. I trust that the following comments satisfactorily answer the queries you have raised. Author's General Comments Since publication of the paper, attention has become focused on the provision of employment opportunities an aspect not dealt with specifically in the paper. The following comments serve to indicate some of the directions in which job opportunities will have to be developed in the numbers that will be required.

Jan 1, 1991

By J. Cilliers, R. Seitz, D. Drinkwater

Mineral processing produces a saleable product from high-cost ore. The Mineral Processing Engineer must constantly improve the separation to create the most value from the ore and reduce the associated costs. Graduate Metallurgists (or Chemical Engineers) have had the university theoretical training required, but often are illprepared to perform their plant roles. The in-work Metcelerate programme was designed specifically to develop the practical skills that Mineral Process Engineers require, by combining three key elements: 1. Knowledge transfer through self-paced, on-line training; 2. Practical skills acquisition through assessed, practical plant exercises; 3. Skills development through online tutorials. A final process improvement project adds significant value to the operation that far exceeds the training cost. Experience from Metcelerate clients has indicated improved staff retention, greater confidence, and pro-active process intervention.

Jan 1, 2023

By J. H. Zietsman, S. Maharajh, J. Muller

In the chlorination process for TiO2 pigment production, blends of titania feedstocks such as ilmenite, synthetic rutile (SR), natural rutile, upgraded slag, and chloride-grade slag are reacted with coke and chlorine at a temperature of around 1000°C to form TiCl4, which is the main product, and other waste metal chlorides. TiCl4 is the main feed material for the TiO2 pigment-making process. Feeding different titania materials to the chlorinator affects the amounts of coke and chlorine required for the process, the amount of waste generated, waste disposal costs, the amount of TiCl4 produced, and bed build-up rates. These factors influence the value of the feedstock. Generally, a higher TiO2 feedstock is more valued since less waste is generated and less reagents are consumed. To quantify the impact of different feedstocks on the chlorinator, a techno-economic model was developed to describe the chlorination process and estimate process variables at steady state. This paper describes the development of the model and studies in which the model has been used to quantify the effects of using different feedstocks.

Jan 1, 2014

By K. Forsberg, Ş. Kaya, C. Dittrich, M. Svärd, E. M. Peters

Scandium can be recovered from waste streams of other metallurgical processes such as red mud from the Bayer process and titanium dioxide acid waste. This is accomplished by an integrated process comprising leaching using acidic media, solvent extraction using organic solvents and stripping of Sc from the organic phase using NH4F solutions. The precipitation of scandium as (NH4)3ScF6 from NH4F strip liquors has been demonstrated at lab-scale with recoveries and purities greater than 98% and 98.3%, respectively. It is desired to assess the possibility of recovering the solvents for re-use in the process to improve its economy and environmental friendliness. Investigations have been conducted to recover alcohol by simple distillation from strip liquor-alcohol mixtures containing about 50 v/v% alcohol. The mixtures were obtained after anti-solvent precipitation of (NH4)3ScF6 using an alcohol to strip liquor volumetric ratio of 1:1. The starting strip liquor contained about 2000 mg/L scandium in 3 mol/L NH4F solution. In the recovery of methanol, the first distillate product collected had a purity of about 87 v/v% and this decreased slightly with time over two hours. The methanol distillate collected in the first 30 minutes with a purity of ca. 85 v/v% was re-used in anti-solvent crystallisation of (NH4)3ScF6 from the initial strip liquor, and the scandium recovery attained was about 98.6 w/w%, almost similar to that of 99.2 w/w% obtained using 100 v/v% methanol. For ethanol recovery, the first distillate product collected was approx. 84 v/v% pure and the distillate collected in the first 30 minutes with a purity of about 75 v/v% achieved a scandium recovery of about 97.2 w/w% in comparison to 98.5 w/w% obtained using 100 v/v% ethanol during anti-solvent crystallisation. Keywords: Alcohol recovery, methanol, ethanol, anti-solvent crystallisation, (NH4)3ScF6

Jan 1, 2020

By S. W. Jacobsz, E. P. Kearsley, N. K. Schoeman

A physical modelling approach was adopted to simulate the process of cave mining propagation at various horizontal to vertical stress state ratios, and image data was captured and used to characterize the resulting failure mechanisms. Image processing was integrated into the study to determine minor principal strains in models to identify where critical extensional strains had been exceeded relative to crack propagation. The study exhibited a scale factor magnitude of 119 using a calculated critical extensional strain value of 0.014% for the modelling material used. It was perceived that cave propagation occurred through successive parallel extensional ‘fracture bands’ in all cases, as noticed in models of prior studies, deviating from that described by the Duplancic model. Cracks tended to bisect regions of minor principal strain equal to the critical extensional strain, which suggests ‘fracture banding’ is an acceptable failure mode for brittle, rock-like materials. Moreover, it was found that models with lower horizontal to vertical stress ratios (K ratios) showed a greater degree of vertical development of the caving mechanism (at similar vertical stresses), while the vertical extent of the cave was suppressed in models with higher K ratios. Models subjected to larger vertical stresses experienced slower cave formation in latter time-steps. These models exhibited larger two-dimensional caved perimeters and areas, respectively.

Nov 30, 2022

By M. Pierce, E. Karampinos

"One of the main challenges in the numerical modelling of ground support in underground excavations is to reproduce the performance of the sequential installation of the reinforcement while capturing the rock mass behaviour of an advancing face in 3D. The 3D modelling approaches used to simulate the progressive advance of excavations are mostly continuum and often cannot reproduce the rock mass failure mechanisms. The 3D discrete element method (DEM) can better reproduce structurally controlled rock mass failure mechanisms and can explicitly represent the reinforcement elements. This paper builds on previous work that reproduced the structurally controlled squeezing conditions in an underground hard-rock mine using 3D DEM and the in situ behaviour of reinforcement under pull conditions. It addresses important issues on the way reinforcement is explicitly introduced in discrete element models. A pseudo-3D model is employed to overcome the computational restrictions and time limitations of a 3D modelling approach. The work focuses on the scaling of the material properties of the ground support elements when the thickness of the model is not equal to the out-of-plane spacing of reinforcement. It demonstrates the scaling methodology based on the type of the modelled reinforcement elements used and investigates the significance of this approach. The advantages and disadvantages of this approach with respect to 2D and 3D methods are discussed. The results are compared with field data and previous modelling work done at the mine. IntroductionNumerical models can provide significant insight into the anticipated behaviour of underground excavations. The modelling of ground support, however, is not a trivial exercise. An effective representation of ground support in underground excavations should capture the behaviour of both the rock mass and the support elements. When numerical models are used for design purposes, the complexity of the problem requires significant time and expertise (Sweby, Dight, and Potvin, 2016).At a laboratory scale, numerical models have successfully reproduced the results of laboratory tests on rockbolts by explicit modelling of the bolt, the borehole, and the grout. Chen and Li (2015) used a continuum FLAC3D (Itasca Consulting Group Inc., 2012) model to simulate the mechanical behaviour of D-bolts as observed experimentally. Grasselli (2005) used a 3D finite element code to model the behaviour of fully grouted rods and expandable bolts under shear."

Dec 1, 2018