Newsletter - Spring 2010
Welcome to the LimitState newsletter where you will find news about LimitState software and short notes on technical features. This month we have an example of the use of 'cutoff' material models in LimitState:GEO, and a guide to the use of support movements in LimitState:RING.
LimitState to Present at ICE Innovation Event on 16th March
An 'Innovation in Geotechnics' event has been organized by the Institution of Civil Engineers Yorkshire Geotechnical Group, and Drs Gilbert and Smith of LimitState will be making a presentation on 'Advances in Geotechnical Limit Analysis'.
If you wish to register for the seminar, please visit the ICE Yorkshire Group website.
Presentation: 'Advances in geotechnical limit analysis'
Limit analysis has for many decades played an important role in geotechnical engineering practice. To extend the usefulness of traditional hand-based limit analysis methods, a variety of computational limit analysis methods have recently been successfully developed. In particular, the recently developed Discontinuity Layout Optimization (DLO) analysis procedure now provides geotechnical engineers with a very powerful means of assessing the stability of slopes, retaining walls, footings and virtually any other geotechnical feature, including those incorporating soil reinforcement.
In the presentation, details of the DLO procedure will be described and the DLO-based LimitState:GEO software will be applied to various design problems to illustrate how effectively the new tool bridges the gap between existing single application hand-type analysis methods and those based on much more complex incremental analysis methods (e.g. non-linear finite element analysis). The emphasis will be on practical assessment of the Ultimate Limit State (ULS), covering appropriate problem definition for a range of ULS design problems and application in accordance with Eurocode 7.
A New-look limitstate.com
The re-launched LimitState website brings together the old limitstate.com and masonryarch.com sites, and is designed to make it as easy as possible for visitors to quickly browse our products and services, and to find all the information that they want.
Users who have previously registered on masonryarch.com will need to re-register their details to be able to use limitstate.com, but will hopefully find the new site a much improved experience.
LimitState clients also get their own dedicated area of the site, with access to an all-new Support Portal, as well as a growing collection of literature, training materials and software archive. We will be contacting each of you in the coming weeks with further details.
Your opinions are very important to us. If you find something wrong, or have any suggestions as to ways in which we may further improve the site, please feel free to get in touch with us using the online contact form.
We hope you like our new look - visit the home page.
LimitState Client Support Portal Online
In addition to the new look, the re-launched limitstate.com also introduces a dedicated client support portal, which provides:
- Access to an expanding knowledge base, which can be searched to find solutions to your problems.
- A webform allowing you to log support requests.
- The ability to view all previous support requests from your organization logged via the Support Portal, or via email to support@limitstate.com.
LimitState clients can visit the support portal here.
GeoFlorida 2010 - Advances in Analysis, Modeling and Design
At this years GEO Institute conference, GeoFlorida 2010, during the session on "Modeling of Geotechnical Engineering Systems", LimitState director, Matthew Gilbert is presenting the paper "Advances in Computational Limit State Analysis and Design". He is also be giving a demonstration of the software on the Tuesday (23rd January) at 3:15pm.
Registered users may download the paper from the Client Area. For more details of the event, please visit the Events pages or download the pdf below:
 |
Full details of all the technical sessions at the GeoFlorida 2010 conference. |
LimitState:GEO featured in 'The Structural Engineer'
LimitState:GEO was recently the subject of a 3-page article in The Structural Engineer - the international journal of the Institution of Structural Engineers.
Entitled "Soil structure revolution", the article explains the theory behind discontinuity layout optimization (DLO), the powerful analysis technique used in LimitState:GEO, and how this dramatically simplifies ULS analysis of geotechnical and soil-structure interaction problems.
The paper has been made available to us by permission from the Institution of Structural Engineers and can be downloaded below. LimitState clients may also access the paper from the Client Area.
|
Download the Structural Engineer article 'Soil structure revolution' |
Imperial Units and US Distributor
When we launched LimitState:GEO in early 2008, we prioritised SI Units. With the latest version, the LimitState:GEO interface can display either Metric (kN, m) or Imperial (lb, ft) units. Users can switch units for the entire application by selecting Tools > Preferences > Units
Selecting the option “Use defaults tailored for Imperial Units” means that default material properties are rounded appropriate to the units, rather than being equivalent to the SI values. This means that, whichever units are selected, users are presented with intuitive values, but keep the ability to analyse models created anywhere in the world and get identical results.
Erik Funegard of our US distributor GEMSoft says,
LimitState:GEO has had a very good reception among our geotechnical customers. Everybody, without exception, is impressed by the combination of simplicity of use and sophisticated analysis, allowing the engineer to analyze far more alternatives than previously possible. We have also been very impressed with LimitState's responsiveness to questions about the software and to suggestions for improvements.
LimitState:GEO allows free switching between imperial and SI units. Whichever is chosen, all displayed values use the selected units. In this image, units of different dimensions are displayed in the property editor.
Effective Use of the 'Cutoff' Material
Introduction
The 'Cutoff' model in LimitState:GEO has many uses in modelling geotechnical problems. This article introduces the user to this often overlooked material type by using the built-in 'Cutoff' material to help with two applications:
- the modelling of a tension crack behind a retaining wall and
- modelling a propped retaining wall.
Modelling Tension Cracks with the 'Cutoff' material model in LimitState:GEO
Consider the 8m high cantilever wall depicted in Fig. 1. The wall is embedded 2m into a cohesive soil of undrained shear strength 35 kN/m2, unit weight 20 kN/m3 and a surcharge load of 10kN/m2 is applied on the surface of the retained soil. For simplicity the soil/wall interfaces are modelled as smooth, and all analyses will be short term undrained. In this example the focus is on the external failure of the wall, and so the wall itself is simply modelled as a 'Rigid' material of unit weight 24kN/m3. In the results that follow all analyses were undertaken using a medium nodal resolution.
The problem was set up in LimitState:GEO with the 'adequacy factor' applied to the surcharge load. The question being asked is therefore how much larger does the surcharge load have to be to cause collapse (i.e. drive the system to its Ultimate Limit State). If the problem is modelled exactly as in the scenario described above, the analysis predicts an adequacy factor of 1.25 (i.e. collapse when the surface load reaches 12.5 kN/m2) with the collapse mechanism predominantly involving sliding of the wall.
Figure 1: Retaining wall collapse mechanism showing formation of tension crack at the upper wall/soil interface
However since the soil is cohesive there are likely to be near surface tensile forces appearing in the retained soil. To eliminate this effect at the soil/wall interface, a tension cutoff material can be 'added' to the wall/soil boundary. (This can be done in various ways e.g. drag and drop the system defined 'No-Tension Cutoff' material, available in the Materials Explorer, onto the right hand soil/wall boundary, and click 'Add'). This allows this boundary to be both smooth and not allow any tensile forces to be transmitted across it. Solution of this problem produces a lower adequacy factor of 1.16 and a modified collapse mechanism involving sliding and rotation as depicted in Fig. 1, where the tension crack can clearly be seen. (Note that by adding the 'No-Tension Cutoff' material to the whole retained soil mass would also allow modelling of tension cracks within the soil and a further reduction of adequacy factor).
Modelling Wall Rotations about a Prop in LimitState:GEO
The margin of safety for the above wall is clearly very low. To increase this, a prop can be introduced located e.g. centred 1.0m from the top of the wall as depicted in Fig 2. In this case a 'Rigid' material is used and is joined to the wall along a height of 0.2m. If the prop is connected directly to the wall, a 'fixed' joint will be modelled. In this case the failure mode that LimitState:GEO will find is of a base heave type failure with soil flowing around the rigid wall. To allow free rotation about the prop it is necessary to again utilise the 'No-Tension Cutoff' material. In this case it should be added to the prop/wall interface. (As before drag and drop the built in 'No-Tension Cutoff' material onto the 0.2m length of prop-wall interface). On solving, the wall is now allowed to rotate about the prop and the failure mechanism in Fig 2. is obtained together with a much increased adequacy factor of 5.28.
Figure 2: Propped wall collapse mechanism with free rotation about the prop
If the wall/prop joint is zoomed into as shown in Fig. 2, it will be seen that the wall is rotating about the lowest point of the wall/prop interface. It is able to do this because the interface now has zero tensile strength, but effectively an infinite compressive strength allowing it to rotate about an almost infinitesimal contact area. More sophisticated use of the cutoff model can be used to model e.g. a plastic hinge in the prop or wall and compressive yield of the prop (e.g. representative of a buckling load). These topics will be covered in future articles.
Download the source .geo files used for the unpropped or propped model and try them yourself.
Using the Support Movement Functionality
Introduction
LimitState:RING is most commonly used as a tool to rapidly assess bridge ultimate load carrying capacity, but the inbuilt ‘support movement wizard’ can provide valuable insights on pre-collapse bridge behaviour. Using this mode, users can explore the likely causes of crack patterns seen in bridges in the field, and also examine potential load paths in a given bridge. This article shows the user how to set up and solve a simple single-span bridge problem using support movement mode. Once familiar with the concept, significantly more complex scenarios can also be considered using support movement mode, e.g. involving multi-ring and / or multi-span arch bridges.
Support Movement Wizard
When the centering supporting the arch barrel of a single-span arch is lowered, the thrust from the arch is transmitted to the supporting abutments. In practice these are liable to move apart slightly and this example explores this case.
Try this:
- First set up the software so that problems are solved immediately after a vehicle is dragged with the mouse. This is achieved by selecting Tools>Preferences>General>Solve after dragging vehicle
- Begin a new project by selecting File>New
- Click Finish to complete the New Bridge Wizard using the default parameters.
- Add a heavy vehicle to the project. Select Tools>Loads>Vehicle Database>European Union Vehicle>EU Triple Axle (2.60m) and click Import.... Click OK then select EU Triple Axle (2.60m) from the Vehicle dropdown list. Click OK.
- Drag the vehicle off the left-hand side of the bridge using the mouse. You will find the load factor is reported as ‘locked’ as, even if the load factor is increased to infinity, the arch will not collapse with the load positioned here.
- Now move the right-hand support outwards by 10mm. Select Tools>Support Movement Wizard then click Pick support block to move. Select the right-hand support block (green) and in the x (mm) column, type 10mm. Finally, click OK.
- Now solve the problem. Click the green solve button, press F5 or select Analysis>Solve.
You will obtain the result shown in Figure. 1. LimitState:RING has identified a suitable load path, together with implied hinge / crack positions, caused by the movement of the support. If a real bridge under consideration has cracks that are different to those predicted by the software, this is likely to indicate a different underlying cause (e.g. differential vertical settlements of the supports). By exploring different support movements, the potential underlying causes of cracks can be investigated.
Figure 1: Line of thrust and hinges in unloaded arch bridge subject to outward support movement
Try this:
- Magnify the displacement using the slider (highlighted in Figure 1) to show the mode of response more clearly. Note that the reported ‘Total work done by moving supports’ is in this case likely to be of much less interest than the observed mode of response (the ‘Total work done by moving supports’ is the horizontal component of the thrust exerted on the support multiplied by the specified displacement).
- Now drag the vehicle with the mouse to various locations across the bridge, as shown in Fig. 2:
Figure 2: Lines of thrust and hinges in loaded arch bridge subject to outward support movement
It is evident that the presence of the vehicle loads leads to migration of the hinges from their original locations. When the applied load is greater the degree of migration becomes more marked, and there is even the potential for complete reversal of the intrados / extrados location of the hinges at a given cross section (the axle loads can be modified by setting the axle load factor via Tools>Partial factors). Clearly the implied repeated opening and closing of hinges under the action of a moving vehicle is likely to be detrimental, potentially leading to rapid deterioration of the structure.
Finally it should be borne in mind that since the original, undeformed, geometry is used in all calculations the support displacements are instantaneous, and it is not currently possible to e.g. determine the amount of support movement required to cause collapse.