University of Mississippi  Department of Civil Engineering 
CE407: Civil Engineering Laboratory II 

Experiment S2a: Stress Analysis of a Circular Arch-Numerical Experiments
Dr. Chris L. Mullen
Last Updated: 10-24-01



1.0 INTRODUCTION
1.1 The SABLE or SAP200 computer programs installed in the Civil Engineering Computer Graphics Laboratory, Room 213 Carrier Hall, will be used to find the internal forces in the circular arch portion of the Three-Hinged Arch studied in Experiment S1. From these forces, stress and strain may then be computed using the elementary beam bending theory of the mechanics of materials.

1.2 The objective is to use the finite element method to estimate the horizontal thrust reactions for comparison with those measured in Experiment S1, and to further esticmate the internal forces, stresses, and strains for comparison with the strains to be measured in Experiment S2b) .

2.0 THEORY
2.1 Definitions.

2.2 Analysis

Compute the internal forces using the method of sections taken at an arbitrary location along the arch. The system is statically determinate when the arch acts as a single unit (without the deck and spandrels).

Compute the normal stress as a function of distance from the neutral axis using N and M and elementary beam bending theory where plane sections remain plane during deformation. Use Hooke's law and the assumption of plane stress to obtain the normal strain.

3.0 SOFTWARE
The SABLE or SAP2000 finite element programs are to be used to construct a model of the arch including the pinned supports and the hinge at the crown. The beam elements provided in the program are to be used to define the stiffness of the arch with respect to vertical loads applied to the deck.

Prepare an Excel spreadsheet that computes  the coordinates of points on a circular arch in a Cartesian system with origin where the horizontal line passing throught the points A and B intersects a vertical line at midspan.

4.0 NUMERICAL PROCEDURE
For the symmetrical arch configuration in Experiments S1, consider the three load cases. Use the finite element software to define a planar model of the arch without the deck or spandrel elements.

Key modeling features are:

    In order to obtain printed results for all three internal forces at the gage position, it is necessary to place a joint at that location. Otherwise, it will be necessary to plot the individual diagrams and graphically determine the values.
    The loads in the vertical direction are most conveniently applied using joint loads instead of beam loads which are defined in a local beam coordinate system (inclined in this case).
    The pinned support joints should be restrained in all three DOF (x-translation, y-translation, and z-rotation).
    The hinge joint should be restrained only in the third DOF (z-rotation).
    The beam elements that attach to the support and hinge joints should have a pin member end condition at that end so that no bending moment will be permitted.
    All other member end conditions should be fixed so that a bending moment will be developed as required.
     
Perform a mesh sensitivity analysis to determine the best accuracy for the least number of elements.  Use elements of equal arclength and a criteria based on the effect on changes in computed bending moment.  Summarize your results in an Excel spreadsheet.

5.0 RESULTS AND CALCULATIONS
5.1 Computed Results Table

Summarize the numerical results in a simple table of the style developed for the recorded data. Calculate theoretical predictions from the elementary theoretical analysis, and compute percent differences (error) between the numerical and theoretical stresses for  the case of a central load of 50 N, taking the theoretical stresses to be the "correct" ones.

6.0 DISCUSSION
Comment on sources of error in the computed data.


7.0 CONCLUSIONS
State what you accomplished through the experiment. Were your numerical experiments reasonable? Was the software well suited for this experiment? Do you have any suggestions to improve the procedure?