Shake Table Model
This study aims to examine the response of a structure to seismic activity and its ultimate failure. For the purpose of this exercise, a quarter inch scale model twelve stories tall is used and a shake table simulates the seismic activity.
The main goals of this structural study were to reduce the mass at the top of the structure to decrease the load, to implement a lateral system that ties the load into the foundation, and create a joinery system to raise the strength and stability at the connection points at the columns and beams. To achieve this we created an initial massing design of four connected towers with each tower decreasing by one fourth in height. The four towers share a central core around a continuous column that runs from the base to the top of the tallest section. Continuous columns were also placed in each corner and beams were tied to columns with one half interlocking joints. We then added an angled lateral system to help tie the structure to together and to the base.
Framing Plan and Detail
This structure is made of quarter inch bass wood for the columns and beams and the lateral system is made of eighth inch by quarter inch bass wood attached by wood glue. The structure is then applied to a half inch thick piece of plywood measuring twelve inches by twelve inches. The overall massing of the structure measures forty feet by forty feet and one hundred eighty feet tall at quarter inch scale. This model uses a basic grid divided into four quadrants. A continuous vertical column is placed in each corner and the center of the square. The structure varies in height in each quadrant; quadrant one is one hundred eighty feet tall, quadrant two is one hundred thirty five feet tall, quadrant three is ninety feet tall, and quadrant four is forty five feet tall. Each quadrant is twenty and a half feet wide. When possible a forty foot wide continuous beam is used to attach two quadrants together. Each connection is joined by an eighth inch notch in each column/ beam and glued together for added
Joint Detail
stability and strength. This is created by cutting out half of each of the column and beams at the connection joint and putting the two pieces together. Once put together, the joint of the two pieces fits flush and does not require either element to be offset. The resulting structure has 12,000 square feet of total rentable area and a central core that connects the four towers and provides access to each level.
Once the Columns and beams are placed, a lateral system is added to the structure using a radial distribution. The center of a circle is placed on each base corner of the building and a line is drawn every twelve degrees from the base. This creates an angled overlapping pattern across the facade of the structure that helps pull the load of the upper levels back into the base of the structure.
Lateral System Design
Given this design, we expect the structure to perform well on the shake table. We do not expect any issues with the interlocking column and beam connections. Since the load has been reduced at the top of the structure, we do not expect any failure at the upper levels. The lateral system is designed to tie the load from the upper levels into the base and we expect this keep the structure stable. The only issue we expect might occur is torsion at the reentrant corners created by the irregular heights of the four towers, but we are expecting the exterior lateral system to help control the twisting.
During the shake table experiment, the structure was able to hold up against the simulated seismic activity under its own weight. For the second round we added weights at the center and upper levels of the structure randomly. At this point, we began to see some vibration at the top of the structure, but the rest of the structure remained stable. For the third round, we added more weights in the middle section of the structure at the exterior beams. With this added weight, the movement of the upper levels became more pronounced, but the center and lower levels remained unaffected. For the fourth round, we moved most of the weights to the upper floors and exterior beams. We also rotated the structure on the shake table since it only moves in one direction to see if seismic load in the opposite direction would have more effect on the irregular shape of the structure. During this test, we began to notice some torsion in structure around the irregular connections of the varied tower heights; however, the movement of the upper floors seemed to remain the same as the previous test.
Torsion at Reentrant Corners
For the fifth round, all weights were move to the exterior beams and a large ten pound weight was added at the one half height tower level. Torsion continued to be observed during this test and the movement of the structure seemed to remain the same as observed in previous tests. In the sixth round we added a brick at three quarter height tower, but movement of the structure remained the same. For the seventh round we added another brick at the tallest tower and removed the weight from the one half height tower. We also increased the speed of the shake table to a very high level. The added weight and speed began to cause some of the lateral system braces to break. Removing the weight from the lower tower caused the load to become uneven and concentrated to one side of the structure. This combined with the added weight and speed began to cause some of the lateral system braces to break. Once the lateral system became compromised, the added weight and increased speed ultimate caused the connection of the structure to the base to fail and the structure overturned in the direction of the unbalanced load, or bricks.
The failure of the structure was the base connection facilitated by an unbalanced load applied to the structure. As expected, the reentrant corners did cause torsion at the irregular connections. The interlocking connections of the columns and beams did provide a strong connection that held up throughout the tests. The lateral system also helped to keep the structure stable and carry the load to the base. However, we discover that the lateral system had some redundancies and could have been reduced as a cost cutting option moving forward with the design. With the lateral system tracking the load back to the foundation, the load became too great at the base connection and this is where the structure finally failed. To counteract the forces acting on the base connections we could have used base isolators to separate the base from the columns and allow the movement to occur without damaging the connection. Another options would have been to extend the vertical columns at the base and embed them into the foundation. The increased depth of the base connection could have helped hold the structure against the seismic forces.
Base Separation
In conclusion, we started with the design concept of a multiple height tower and used continuous columns and beams where possible, an angled lateral system, and interlocking joint connections to hold the structure and create strength and stability. While some things we expected to happen during the shake table test did occur, the ultimate failure of the base connection was not expected. The experience of designing, building, testing, and analyzing our model has helped us understand the effects of seismic activity on tower typology buildings, what causes failure in these structure, and strategies to counteract these effects or even fully prevent them from occurring.
Design: Joanna Jankowska, Renee Langley Salai, Rodrigo Valesquez Gonzalez
Model Construction: Rabi Aldaylami, Joanna Jankowska, Rodrigo Valesquez Gonzalez
Construction Drawings: Rodrigo Velasquez Gonzalez
Written Analysis & Images: Renee Langley Salai
