Wear prediction modelled in 3D

Advances in modelling software allow accurate prediction of the effects of erosion in pipes and fittings in really inaccessible places, wear in mining and construction equipment, and the behaviour of granular solids. Used to predict the effects of entrained sand on components used by the Oil and Gas industries beneath the sea, advances in technologies to predict erosion and other kinds of particle induced wear are of interest to anyone concerned with wear activities that result in a shape change with the potential to lead to unexpectedly early failure. While the demonstrations look impressive, even those who develop and sell the software warn that it is essential to validate these predictions with experimental measurements. The latest developments were announced at a recent user conference organised by Ansys Fluent for the oil and gas industries. Neil Barton, of TUV NEL, told delegates: “People use CFD [Computational Fluid Dynamics] to a point that is almost scary.” He cited the example of using CFD to predict when an undersea choke would need changing. The problem here was not wear caused by entrained sand, but by cavitation – because what had originally been a producing oil well was being converted to water injection, to try to force out residual oil that would no longer come out on its own. Barton said it was always wise to validate any modelling with above-ground experiments, but added: “People often decline to do even $15,000 worth of testing.” The importance of verification was supported by Fluent software developers. But knowing that some users put their faith in modelling alone, the developers strive to make their predictions accurate – and offer alternative scientific models within the software, so that users can compare results based on different assumptions and get a feel for possible errors and uncertainties. Franz Zdravistch from Ansys Fluent in the US explained that it is possible to model shape changes caused by fluid-induced erosion and wear – using a dynamic mesh model in a new Fluent module. “In some cases, the changing shape does not make much difference, but in other cases it does,” he said. Many factors affect erosion, including the angle of particle impingement, particle velocity, diameter, mass, shape, collision frequency and material type. The new software module offers Fluent’s own erosion model plus four others. Zdravistch explained: “Users have to do some calibration from test rig measurements.” Using the module, it was possible to reproduce some of the strange wear patterns encountered in real components – including instances where parts not only wear, but also distort under load as a result of localised thinning. Another way of attacking such problems is to use the kind of method developed by DEM Solutions, based in Edinburgh. DEM’s John Favier explained that its EDEM software is used to “model the dynamics of each representative particle”. It considers the mechanical and inertial properties of individual particles in terms of momentum, mass and heat transfer between particles. As well as oil and gas applications, it has been used to study wear in excavator and dragline buckets, conveyors, hoppers, rock crushers and grinding mills. In many applications, EDEM can be coupled with Finite Element Analysis, in order to estimate structural deflections caused by interacting with the particles, ranging from fine sand up to rock size. Alternatively, it is possible to couple it with CFD to model particle fluid flows. “If the suspension is dilute, fluid dynamics dominate. If it is dense, particle dynamics dominate,” he says. Favier said that EDM-Fluent coupling is suitable for systems with large particle size distribution, non-spherical particles, dense packed particles in a fluid, particle agglomeration or segregation, build-up of particles on surfaces, particles sticking to equipment surfaces, and coalescence or break-of particles and droplets. The technique is also versatile. It is being used by DeBeers to improve marine diamond mining, to improve the design of packed bed and fluidised bed reactors and to improve rock drilling. The rock is modelled using particles bonded together with brittle bonds that fracture under load. The fluid pressure acting on the rock and cuttings can also be modelled. Fluent too, we learned is also working on what it terms ‘Macroscopic Particle Modelling’ (MPM) – which it makes available to select customers through its consulting services. However, it says there are no immediate plans to include it in the standard version of its software. The problems MPM addresses, like most real world problems, are extremely complicated to analyse. Particles can collide, interact with each other, grow larger through coalescence or agglomeration or become smaller through attrition. They can also stick to each other and to walls or other solids. One application is the manufacture of pills for the pharmaceutical industry. Because of the different models offered and the different assumptions that can be made, it is definitely not software to be used by the non-expert. Ansys Fluent DEM Solutions Pointers * Shape changes caused by erosion and wear can be modelled as they develop, which may lead to enhanced wear effects and consequent early failure * The behaviour of particles in fluids and solids – including sand, mud, rock and pill manufacture – can also be modelled * The modelling of heterogeneous mixtures is still at an early stage and should always be backed by experimental validation