Presenter Information

Randy R. Frank

Research Advisor

Hornsey, Edward

Advisor's Department

Mechanical and Aerospace Engineering

Meeting Name

1st Annual UMR Undergraduate Research Symposium (1991: Apr., Rolla, MO)


Joint flexibility plays an important role in the distribution of strains and displacements in the frame structures of buildings, towers, automobiles, etc.. The objective of this project is to study the effect of flexibility (or rigidity) of two simple mechanical joints in frame structures on the stress and displacement distributions, using both experimental and analytical methods. Three welded and three bolted samples were fabricated using A-36 low carbon steel in a two column and horizontal crossmember (table) arrangement. This arrangement allows for both shear forces and axial forces in the joint. From a three element rectangular rosette strain gage attached to the bottom of the cross- member, strain readings were collected by a computer based data acquisition and processing system, during point and line loading (elastic range) of the specimens by a load frame. These results were compared to each other and to values found by analytical methods for the same arrangement with theoretically rigid and theoretically completely flexible joints.

The data for the line loading of the samples were plotted and compared with the analytical results. Also a quantitative flexibility index was developed and used to compare the bolted and welded joints. The flexibility index is defined as the ratio of the difference between the actual and theoretical rigid joint flexibility to the difference between the theoretical flexible joint and the theoretical rigid joint flexibility in percentage. These comparisons were made for each sample at load increments over 50 lbs and within the elastic limit of the specimen. This was chosen because of the linear relation of the data received in this range and the slight error of the experimental apparatus at low loads. The test was conducted for each of the six specimens of the bolted and welded joints (3 specimens per joint type) and then repeated. This allowed for error in the measurement and in the manufacture of the specimens. The average flexibility index for the bolted joint samples (line loading) were 52.23%, 50.46%, and 46.6% resulting in a total average flexibility index of 49.73% with standard deviation of 3.21. The average flexibility index for the welded joint samples (line loading) were 46.38%, 42.95%, and 42.27% resulting in a total average flexibility index of 43.87% with standard deviation of 2.78. The data from the point loading runs were used to see if anything unusual happened in this case. Because of averaging by the strain gage of the values of the strain at the center due to the stress concentration of a point loading the values were not used to compute a flexibility index. There were no unusual findings from the point loading data.

The results show that the bolted joints used in this experiment were more flexible than the welded joints for both line and point load. Furthermore, both joints did not act as completely flexible or completely rigid joints. This fact confirms the importance of the joint flexibility consideration in the design of steel frame structures.

Research Category

Improving Structures

Document Type



3rd Place

Presentation Date

April 1991