Captive Column . Com

arrow1.gif (893 bytes) Timeline - 1985: UND HPR State Study (1)-82(A)...
Winder The winder UND used to wrap the Captive Column test bridge beams. Seated at the machine is Gary (son in law), with Lawrence Bosch on the right. Various Captive Column test beams can be seen in the background. (24k)

HPR State Study (1)-82(A): "Design Construction and Testing of Captive Column Bridge Girders" was sponsored and funded by a grant from the Materials And Research Division of the North Dakota State Highway Department, in cooperation with the U.S. Department Of Transportation, Federal Highway Administration.

Some of the topics covered in Study (1)-82(A) are: Girder Design and Construction, Load Tests, Equations, Computer Models and Computer Model Verifications, and so on. This UND study may be obtained from the National Technical Information Service, Springfield, Virginia 22161.

Here is an excerpt taken directly from this 195-page study. This study is unclassified and is not copyrighted. Furthermore, there are no restrictions placed on its distribution, which is why it is reproduced here:

"For each of the static tests, load-deflection curves were plotted for increasing and decreasing loads. This explains the two curves shown in Figure 2-21. Figure 2-22 demonstrates the repeatability of the beam's performance as the two curves (two lines for increasing load and two lines for decreasing load) virtually plot on top of each other.

The following observations summarize an analysis of all of the static load test data:

  1. The load-deflection curve is essentially linear.
  2. For each load case, the load-deflection behavior is repeatable from cycle to cycle through the test.
  3. There was no appreciable change in the load-deflection response for different load application rates.
  4. The girder stiffness as determined from the experimental data is approximately 38,500 lbs./in. Using this value, the midspan deflections at loads of 19,500 lbs. and 22,800 lbs., were computed to be 0.51 in. and 0.59 in., respectively. The allowable deflection is span / 800, which equals 0.6 in. Therefore, the girder satisfies the deflection requirement under HS-20 loading.

The girder was subjected to cyclic loading between the limits of 1,000 and 19,000 lbs. at a rate of three cycles per second for well over one million cycles. The purpose of this test was to investigate possible fatigue effects on the structure. Analysis of the results showed that there was no detectable change in the load carrying ability of the girder resulting from these load cycles. The load-deflection curves for before and after the test were essentially identical. In addition, no noticeable deterioration of the girder components was observed.

The girder was subjected to an ultimate static load at midspan using a hydraulic ram and load cell arrangement having a maximum load capability of 120,000 lbs. The load-deflection curve for this test is shown in Figure 2-23. The curve becomes significantly nonlinear at a load of approximately 52,000 lbs. The ultimate load capacity of the girder was approximately 70,000 lbs., which produced noticeable plastic deformation of the top two caps and breaking of a number of wrap strands. It should be noted, however, that this same girder has since been rotated 180 degrees to place the plastically deformed caps in tension, with the result that the load-deflection characteristics are only slightly different than before the ultimate load test.

2.10 Computer Model Verification
Table 2-11 shows a comparison of experimentally determined midspan deflections with the results of a number of analytical model's for three different load levels. The following analytical models were considered.

  1. Beam theory model, neglecting shear deformation.
  2. Beam theory model, including shear deformation, for the case of ideal wrap pretension (IWPT).
  3. Beam theory model, including shear deformation, for the case of no wrap pretension (NWPT).
  4. SAP computer model with ideal wrap pretension (IWPT).
  5. SAP computer model with no wrap pretension (NWPT,).

Table 2-12 is a similar comparison for the case of cap stresses at midspan. Results are presented for strain gages at two locations on each cap, along with their average value, plus values obtained from beam theory, the computer model, and an internal couple analysis, wherein the bending moment is assumed to be carried entirely by the caps.

Analysis of the results shows that the computer model provides the best prediction of deflection, while all of the models produce reasonable predictions of cap stress. These observations are consistent with conclusions drawn from previous captive column research.

2.11 Summary of Girder
The captive column bridge girder, as designed, constructed and tested, demonstrated the validity of the captive column concept and provided verification of the computer model. Although the project originally called for construction of a second girder, performance of the first girder was judged so positive as to suggest that additional effort could be more efficiently directed in other areas such as sign and luminaire support structures. As reported, the girder was subjected to an equivalent HS-20 loading applied through more than 1,000,000 cycles. It was also subjected to an ultimate load test. The girder is currently stored in the structures laboratory for final disposition as directed by the sponsoring agencies. "

UND Study Figure 2-19 UND Study Figure 2-20 UND Study Figure 2-21
Figure 2-19
Figure 2-20
Figure 2-21
UND Study Figure 2-22 UND Study Figure 2-23
Figure 2-22
Figure 2-23

The term "cap" (highlighted in red throughout the excerpt) refers to the column elements of the Captive Column. For some reason the engineering community insists on using its own terms instead of the ones given by the inventor. This particular renaming was done even though the inventor specifically requested that the term "column" be used. This behavior was interesting considering Mr. Bosch was on the UND payroll for 3 years at $3,000.00 a month. In exchange for his paycheck, Mr. Bosch acted as a consultant to UND and signed over a Captive Column license for Utility Transmission Towers. However, the university never made use of their license -- they just sat on it. Why they did this remains a mystery to this day.

The following question is for the skeptics in the audience: If the Captive Column doesn't work, why would a university engineering department go to the trouble of doing so many studies? By the way, this wasn't just an ordinary study; it was an Application study. You don't do an application study until you already know it works.

For problems or comments about this site, contact

Since the information contained in this web site is being provided free of charge, please read this Disclaimer carefully before using any of the data in this web site (click here).

Copyright 1998-2004 by Lawrence R. Bosch.