U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000


Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

FHWA Home / Highways for LIFE / Technology Partnerships / Bridge Technology / Precast Bent System for High Seismic Regions: Laboratory Tests of Column-to-Footing Socket Connections

Precast Bent System for High Seismic Regions: Laboratory Tests of Column-to-Footing Socket Connections

APPENDIX E: DESIGN CONCEPTS

This appendix describes four concepts for connecting a precast concrete bridge column to a cast-in-place foundation. These concepts arose while the research team was preparing the proposal for this Highways for LIFE project. The purpose was to permit participants in the WSDOT ABC group to see the concepts prior to a meeting on July 30, 2009. It was hoped that, at that meeting, one concept could be selected for use in the Grand Mound Bridge, and that preparations for testing it at the University of Washington could begin. The Highways for LIFE schedule required testing to start that summer.

The goal of these four design concepts was to develop a footing-to-column connection that combines good constructability and seismic performance. The concepts use a small number of large bars in the column, so that they will be compatible with the cap-beam connection shown on the plans.

These concepts constituted a basis for discussion and were not finished designs. The discussion that follows explains the rationale behind their development. The concepts were divided into two groups. In group 1, the bars project below the bottom of the precast concrete and are used for supporting the column while the footing is cast. In group 2, the precast concrete extends below the bottom of the bars and is used to support the column's weight.

The drawings show circular columns, but the same concepts could be used with rectangular columns.

Columns with Projecting Bars

Figure 163 shows a detail similar to the one presented on the WSDOT preliminary plans in early summer 2009 that includes a steel ring attached to the bottom of the main column bars. However, the construction sequence for the detail on the plans was not clear. The detail shown here was intended to be used in a construction sequence in which the ring is used both as a temporary locating device during construction and as end anchorage for the main column bars under seismic loading.

The main bars have threaded ends and project from the bottom of the column. A light ring template is attached (with two nuts per bar) to hold the bars in exactly the right location when the column is (pre-)cast. The ring is removed after casting, but one nut remains on each bar. On-site, a similar (but maybe heavier) ring is set and secured in position, and it acts as a location device for the column. The bottom mat is placed on top of the ring. The column is brought in and set with the bars projecting through the holes in the ring. The ring ensures that the column base is located correctly. By careful pre-placement of the nuts on the bars (jam-nutted, if desired, to prevent movement), the column alignment and plumb should also be very close to true. Adjustments can be made as needed by turning the nuts. When the column is plumb, a second nut is screwed onto the end of the bar below the ring, thereby securing the ring to the bar for anchorage.

This drawing shows a column with a connection to the footing using projecting bars and a steel anchor ring. Step 1 shows the ring detail, while step 2 shows how the column projecting bars are connected to the ring in the base. The drawing shows a column section (marked A) and a detail of the nut configuration in the ring detail (marked B).

Figure 163. Diagram. Column detail with projecting bars.

This drawing shows a column with a connection to the footing using projecting bars and a steel anchor ring. Step 1 shows the ring detail, while step 2 shows how the column projecting bars are connected to the ring in the base. The drawing shows a column section (marked A) and a detail of the nut configuration in the ring detail (marked B).

Figure 163. Diagram. Column detail with projecting bars (continued).

The proposed construction sequence was:

  1. Fix the light ring template to the threaded bars. Place it the form.
  2. Precast the column.
  3. Remove the template. Leave one nut on each bar.
  4. On-site, place and secure the heavy ring.
  5. Place the bottom mat of the footing steel.
  6. Set the precast column on the heavy ring. Adjust for level and plumb as needed using nuts on bars.
  7. Place and tighten a nut on each bar beneath the ring.
  8. Install and tie the top mat. Pass top bars through the spiral and column bars as needed.
  9. Cast the footing.

Columns without Projecting Bars

Three details were shown in this group—socket columns, long struts, and short struts. In each case, the goal was to have no bars projecting below the bottom of the concrete, thereby making the system more robust for transportation and handling.

Socket Columns

The socket column was the simplest connection to erect. This concept formed the basis of the socket connection that was tested in the program. Its advantages were:

  • Robust for handling and transportation.
  • Connection is simple. No steel to connect on site.

Disadvantages were that the footing may have to be deeper, to handle the joint shear forces.

The proposed construction sequence was:

  1. Precast the column.
  2. Place the bottom mat of the footing steel.
  3. Set the precast column, level (using shims or other devices), and brace it.
  4. Install and tie the top mat.
  5. Cast the footing.

Notes on this connection:

  • Figure 164 shows a notch in the bottom of the column in each direction. This was to permit some bottom bars of the footing to pass under the column, where they are most useful. Note that these bars will not engage directly with the vertical column steel. The couple that resists the column moment will therefore consist of horizontal forces at the top and bottom of the footing.
  • From a construction viewpoint, it would be slightly easier to place no bars under the column, and to distribute them all uniformly either side of the column. However, this might be less satisfactory from a structural viewpoint.
  • The top mat will either have to be placed with no bars passing through the column, or (straight) bars will have to be placed in ducts through the column. Under seismic load, the negative cantilever moment on the uplift side of the footing is much smaller than the positive moment on the high soil pressure side, because it is caused only by the self-weight of the footing. Therefore, the top steel plays a much less significant role than the bottom steel, and a top mat that requires no steel to pass through the column may be feasible.
  • The column must resist vertical punching shear. Two critical perimeters should be checked: the usual one, forming a cone through the footing, and a second, cylindrical one at the interface between the precast column and the cast-in-place footing. However, the shear stress on the latter is on the order of 100 psi in the Grand Mound Bridge, and that stress can easily be handled by roughening the surface of the precast column directly or by casting a corrugated steel tube around the part of it that is to be embedded in the footing.
This drawing shows a horizontal section through the column (marked A) and an elevation of the column (marked B).

Figure 164. Diagram. Socket column concept.

Long Struts

The goal of this detail was to enable the column to be supported on concrete components, rather than resting on projecting bars, while the footing is cast. Details are shown in figure 165. The column bars are not fully encased in the precast concrete, and their end anchors are embedded in the cast-in-place concrete of the footing. This achieves better engagement between the footing steel and the column steel than would be the case with a socket column.

This drawing shows a horizontal section through the long struts at the bottom of the column (marked A) and an elevation of the long struts (marked B).

Figure 165. Diagram. Long struts concept.

The column is cast with a cruciform blockout at the bottom. The blockout depth is approximately the same as the footing depth and has a slight taper on it to facilitate removal of the blockout formwork. The four concrete struts, or legs, projecting from the bottom of the solid part of the column support the weight of the column while the footing is cast in place round it. Within the footing depth, the column spiral is shown at the outer face of the column, with no cover. This arrangement has several consequences:

  • It allows the blockout to be formed and removed more easily than if the spiral were placed 3 inches in from the curved surface.
  • It creates a rough surface at the outer face of the precast concrete, to improve bond.
  • It provides some support for the struts during handling.
  • It provides a mechanism for joint shear resistance within the column core.

The proposed construction sequence was:

  1. Precast the column.
  2. Place the bottom mat of the footing. Leave the center four bars each way under the column loose.
  3. Set the precast column, level (using shims or other devices), and brace it.
  4. Move the four center bars in each direction of the footing mat so they lie on top of the column bar terminators. Tie the bars.
  5. Install and tie the top mat. This requires passing some of the bars through the cruciform blockout space in the column. (Those bars will have to be straight.)
  6. Cast the footing.

Notes on this connection:

  • The top of the blockout has a slope to facilitate compaction of the cast-in-place concrete and to avoid air pockets.
  • Leveling may be done using shims or other leveling devices.
  • A "rat slab" may be cast under the column (or precast) if the ground will be unable to support the local contact pressure caused by the leveling devices. (The column segment in the Grand Mound bridge weighs less than 8 kips.)
  • The bottom mat bars will probably have hooks. Therefore, they cannot be installed after the column is set.
  • The plan is to lift the loose bars of the bottom mat over the T-heads of the main column bars, once the precast column is in place. This is straightforward for the top layer of steel in the bottom mat, but the bottom layer creates some challenges. That issue needs to be studied and resolved.
  • The column spiral is cast with the precast column and is to be exposed to provide a rough surface. The intent is to avoid placing a corrugated tube there, because the tube would have to be cut extensively at the blockouts, thereby causing additional work and reducing the tube's effectiveness for shear resistance and confinement.
  • Some vertical steel is needed in the long struts to prevent bending failure during handling. If it extends past the top of the footing, it provides additional flexural strength at the column-footing interface. There is nothing inherently bad about this, but it would be necessary to check that the increased flexural strength does not cause problems elsewhere (such as at the partial-height column splice).
  • The concrete surfaces at the blockout should be roughened to optimize transfer of shear across the interface between the precast column and the cast-in-place footing. Roughening could be achieved by use of a retarder on the blockout form.

Short Struts

This detail, illustrated in figure 166, was intended to overcome one of the potential weaknesses of the "long struts" system, namely the fact that the struts are quite slender and therefore vulnerable to damage during handling. The detail was in most ways identical to the "long struts" system, but the struts are made shorter. This means that the top mat bars cannot be passed through the blockout and will have to be placed in the footing outside the column. For the socket column detail, it was argued that this may be acceptable. The same argument holds here. If not, horizontal ducts could be placed in the column to accommodate these top mat bars, but they represent additional complexity and congestion, and they create a weak plane for horizontal shear, so they were undesirable.

The main potential disadvantage of the short struts system lies in the configuration of the strut-and-tie system in the footing. The primary struts (of the strut-and-tie model) may have to pass through the cast-in-place footing and precast concrete, with the associated interfaces between. The precast concrete will be roughened at the interfaces, but if the shear capacity there is inadequate it may compromise the ability of the struts to carry their full load. As with the long strut detail, the spiral should be placed at the outer cylindrical surface to simplify the blockout formwork and to create a rough surface there.

The conservative version tested in this research (specimen SF-1) was a modified version of the short strut concept.

This drawing shows a horizontal section through the short struts at the bottom of the column (marked A) and an elevation of the short struts (marked B).

Figure 166. Diagram. Short struts concept.

Page last modified on June 28, 2016
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000