Good morning friends, faculty, and jurors. My name is Chris Barlow. I am in the structural option of AE, and I am here to present my thesis on the Tree House. The Tree House is a residence hall owned by the Massachusetts State College Building Authority and occupied by students from several of the many colleges and universities in the nearby area of Mission Hill in Boston. The building is 21 stories and 260 feet tall. It is important to note that the Tree House is not a mere college dorm building and indeed offers a variety of spaces more often seen in high end residential buildings. On the first floor, there is an open lobby and cafe. The second level is a student medical facility. The third floor is a variety of communal spaces including a community kitchen, fitness center, laundry room, and more. The remaining floors, excluding a mechanical penthouse, are student residences.
I have provided you with a slide outline to help guide you through this presentation. As you can see it is rather simple. First, we will discuss the existing structure. During my investigation, I saw that there were many design challenges that the architect and structural engineer had to resolve together. In other words, the existing structural solution is married to the existing architecture. Rather than set an arbitrary goal of intruding minimally on the existing architecture, I decided to fully reconsider both systems during my redesign. In this presentation I will briefly cover two of the major challenges with the design of this building and show you my alternative solutions. Lastly, we will compare cost and schedule of the original design with my own.
The existing structure is a steel frame superstructure with composite metal decking. The metal decking is long spanning, eliminating the need for in-fill beams in most bays. A combination of moment frames and braced frames provide lateral support. The foundation system utilizes cast-in-place concrete elements and steel friction piles.
A typical floor plan shows that there are 7 major column lines in the long direction and 4 major column lines in the short direction for the third floor to the roof. However, at the base and second level, there are only 3 column lines, and one of them is now curved. This curvature is due to a large sewer easement beneath the site. This easement creates a cantilever, the severity of which can be seen in this image.
Looking at the gravity system, the first design challenge I would like to address is height constraints. The typical floor to floor height is 10’ - 2”. The typical ceiling height in a residence is 8’ - 11”. This architectural provision leaves 1’ - 3” of space for both structure and MEP. Using steel framing, the only means of fitting both systems in this tight space is to use beam penetrations. A typical floor has 15 penetrations of various types. Specialty design such as this is costly and difficult to construct. Keep in mind that this doesn’t occur on one floor but all of them.
So, the issue here is height constraints. I believe a better solution is a reinforced concrete flat plate, which is typically slimmer than steel framing.
Moving on to the lateral system, this graphic shows the placement of lateral elements: moment frames are in red, and braced frames in green. As I mentioned earlier, the sewer easement causes one of the four major column lines to recede at the base. Additionally, the architecture restricts another of the four major column lines from reaching the base. As you can see there is a moment frame along Line A, which is the line that recedes at the base for the easement. A moment frame like this is of course costly and its effectiveness is somewhat questionable. In order to determine its effectiveness I investigated the building with and without this frame. By removing the frame, the Center of Rigidity moves about 8.5 feet, creating more eccentricity. This of course causes significantly more torsional displacement at the top of the structure. Without the moment frame along Line A, I saw that the building had issues with both extreme torsional irregularity and torsional irregularity. With the moment frame, I saw that the issue of extreme torsional irregularity was resolved, eliminating one irregularity. However, keep in mind that this frame is offset, which is in itself an irregularity. So we've eliminated one irregularity but gained another.
Ultimately, the design challenge here is limited space for lateral elements. With my concrete redesign I decided to implement reinforced concrete shear walls. These shear walls can be placed around the stairwell and elevator openings, which typically require thicker material around them anyway for fire rating requirements.
At this point we will shift to my redesign. As you can see I used Revit in order to create a full 3D model of my redesign. This model was created to help you understand the design holistically. Also, the BIM capabilities of Revit were utilized for accurate take-offs in my detailed estimate. A link to this model will be provided at the end of the presentation.
Starting with the gravity system, you can see that the base of the structure now looks a bit different. Mainly, I have brought Line B to the base of the structure and introduced three circular columns to the open lobby space. This decision was made for both structural and architectural reason. During my visit to the Tree House, I saw that there were many opportunities within the building for the students to express themselves creatively; however, this wasn't true in the lobby space, which was very open and vacant. I did notice that there were some students hanging posters on the columns along line B.1 and along the glazing after they had run out of room. So it seemed rather clear to me that the students wanted to make this space their own but couldn't. These columns offer students that ability to express themselves creatively. This change also gives a greater sense of purpose to the furniture arrangement in the lobby as can be seen here at The Museum of Modern Art. Detailing of one of these circular columns as well as a square column are shown here.
Looking at the typical floor slab, you can see that a minimum column size of 12” x 12” is needed to eliminate the need for shear reinforcement in the slab. Of course, these columns can be made wider if the contractor needs more space for concrete consolidation in the reinforcement cage. This minimum is overwritten in five locations in order to limit the clear span of the slab. Doing so allowed a 10” slab to be chosen. The reinforcement details for this slab are shown here.
In order to transfer gravity loads from column line A across the cantilever and into the shear wall cores, a transfer slab and outrigger system was designed. Level 4 was chosen to be the transfer level. This is because, as you may recall, Level 3 is the most mixed use floor. Already in the original design there are thick partitions between these spaces. My redesign actually gives back some square footage of usable space to the occupant while still maintaining a solid acoustical barrier between spaces. The thickness of the transfer level is 2 feet, and the thickness of the outrigger walls is 16”.
At this point we will move on to the lateral system. While designing the shear walls, I noticed that ETABS did not properly analyze the complex shapes in my model. As you can see here, ETABS utilized the full section in bending. For example, when bent about the neutral axis in this direction, the full lengths of the transverse pieces of the wall were considered in compression and tension. From ACI 318 we know that concrete does not behave in this manner. Rather, there are dimensional limitations on the effective flange widths of box-shaped sections. In other words, the walls ought to look like this. So, strength design was based on hand calculations, and detailing is shown here. ETABS was utilized for displacement calculations by using a stiffness modifier of 0.5 for all concrete elements, which is allowable by code. Allowable drifts were met for both wind and seismic loading which limit damage to the facade.
At this time I would like to discuss an interesting feature of my lateral system. After completion of trial sizing for the shear walls, I noticed that allowable drift was not met for sustained dead and live loading. Which may seem like an odd thing to say, but I’ll explain. Due to the cantilever mentioned previously, the building leaned several inches toward the cantilevered edge. This leaning caused differential rates of creep to occur across the main superstructure. Additionally, the shear walls along line D had high tensile stresses under certain ultimate load combinations. Of course, concrete shear walls do not perform well in tension.
So they say when you are faced with a difficult challenge that you should approach it from a fresh perspective. And as I looked at my building from this angle I saw a solution. Lying on its side, the building becomes a cantilevered beam with a uniform load and moments due to the eccentric floor loads. Typically, if the deflection of a beam is excessive we provide camber to the section. This camber is often achieved by use of post tensioning. So after that I quickly Googled "vertical post-tensioning," and learned that it is actually a possible design solution, though it is rather new. In fact, the concrete code doesn't talk about it all that much. My research into this solution primarily centered on one particular project, which is the River House Project. From this research I came up with the following solution.
Of course, vertical post tensioning is very different from horizontal. Mainly, you cannot shore up the building as it is being built. Therefore, I analyzed the building floor by floor for construction loads to determine the critical location where post-tensioning is needed. In this case, post tensioning is installed after completion of the 7th floor.
Now, we will quickly compare the cost and schedule for these two systems. The final budgeted amount for the steel frame superstructure, which includes allowances for fireproofing and the beam penetrations, was $7million. According to my detailed estimate, which included 47 different line items, considerations for winter conditions, city cost index factors, worker’s comp., etc. I estimated a cost for the concrete redesign to be about $6.4million. However, this is mostly due to the extremely high city cost index factors for concrete at this time. Apparently, in 2010 concrete was not very popular in Boston. Using national average cost data, the cost of the redesign becomes $5.4million. Now we will move onto schedule. In just 2 years, the original design went from a conformed set of documents to full occupancy. Meanwhile, I estimated that my redesign would lag behind the steel design by about 3-4 months.
In conclusion, the decision to use structural steel for this project is certainly justifiable from a scheduling perspective since steel construction was more popular in Boston at this time. However, I feel that not all the design decisions made by the structural engineer and architect are equally justifiable. From the very beginning of my academic career in AE I have sought to break down the barrier between structural engineering and architecture. In my opinion, the two go hand-in-hand. In this presentation, I have shown how considering both systems equally can be beneficial to the overall design.
Thank you.
