USING BAR ELEMENTS Bar elements in CRISP carry axial forces only. Bar elements may be used to represent ties, pin-ended struts (ie no bending moment) and geotextile memberane. How do I create bar elements? Exacltly the same way as all other elements. Click on the first node and then click back on the first. The element will then turn green to show that the program has closed the element successfully. Define material properties for the bar by selecting BAR in the list of material models. Then select the bar elements and assign them this property using the Element Type feature. When do I use 2-noded bar elements as opposed to 3-noded elements? If the element representing the bar is meant to be embedded between the two adjacent 2D elements, then it would have to be a 3-noded bar. This would insure that the mid-side node of the bar is the same node as the mid-side nodes of the two adjacent 2D elements. Geotextile memberane may be represented by a 3-noded bar element as can be seen in the image below: If the bar is to represent a strut (or tie) connecting two points, then a 2-noded bar would be appropriate. For example you might wish to fix a wall against a point on the boundary of the mesh using pin-joints in which case a single 2-noded bar would be sufficient. You must avoid using a 3-noded bar if it is not embedded in soil (or between two 2D elements) as the unrestrained mid-side node of the bar would create ill-conditioning in the stiffness causing the program to stop. Why do I get axial force for bars in KN/m2 per meter run in the stress output section? For bars, the output stresses are in fact forces in KN per meter run. The stress array VARINT caculates the axial forces as follows: VARINT=epsilon(x).E.A where E is Young's mudulus and A is cross sectional area and therefore output in the stress table is force and not stress How do I model prestresed force which is in tension in the anchor but pushing (compression) towards the wall? Let's say you are installing a pre-stressed anchor, modelled very crudely by a 2-noded bar. The diagram below shows a bit of a wall, and a bit of the retained soil in which the fixed length of the anchor terminates. '*' indicates which nodes the bar will be attached to (e.g. nodes 31 and 264 respectively). Procedure: In the Nth increment block of you analysis, you apply prestress loads at nodes 31 and 264. Suppose you had a load of 500 kN in the anchor, which was inclined at 30 deg to the horizontal. Then given the orientation of the example (and noting the +ve directions for x and y) at node 31 you would have to apply DFX = 500*cos(30) and DFY = -500*sin(30) ... whereas at node 264 you would apply DFX = -500*cos(30) and DFY = 500*sin(30). This has the effect of "pulling together" nodes 31 and 264 with a resultant force of 500 kN. In the (N+1)th block, you then install the 2-node bar element. Subsequent axial loads picked up by the bar (either +ve or -ve) will be the CHANGES in load w.r.t. the datum at installation (i.e. 500 kN in this example). A PROP with a COMPRESSIVE prestress would be handled just the same, but with the loads applied in such away to as to be pushing apart the two nodes before bar installation. Amir