Mid-century Modern Concrete and Glass Facades by The Architect's Newspaper

Mid‐century Modern Concrete and Glass Facades New Design and Restoration for Today’s University Richard A. and Susan F. Smith Campus Center Harvard University

Hopkins Architects ‐ Design Architect Bruner/Cott Architects ‐ Executive Architect & Façade Restoration Architect Simpson, Gumpertz + Heger ‐ Concrete Restoration Engineer Arup – Building Envelope Engineer for New Construction

Josep Lluis Sert introduced modernism derived from Le Corbusier into university settings starting in the mid‐1950’s. Holyoke Center was an ambitious, ten‐story exposed concrete structure that covered an entire city block in the heart of Harvard Square. It was basically an office tower with a section devoted to university health services, street level retail including a bank. It was not a popular building.

Concrete structure was exposed with downturned board‐marked spandrel beams , precast mullions, and story‐height store front glazing. Windows were typically single‐pane without treatment for safety in case of breakage. Glazing preceded introduction of Low‐emissivity coatings. Facades were composed with clear vision panels and translucent Plyglass panels of glass fiber mats between two edge‐sealed sheets of ¼” glass.

The project was completed in two phases between 1960 and 1965. This image shows Sert’s innovative elevations with figure‐ground patterns of clear and translucent glazing, cast‐in‐place and precast concrete, and the use of glass at lower floors to levitate the building’s massing. Note the large pane sizes and simple steel mullions at street level for comparison with glazing in our new additions. The building exterior is under jurisdiction of the Cambridge Historical Commission, who takes its architecture significance very seriously while recognizing the necessity of change for today’s uses.

In April 1970, street rioters smashed windows at street level. We bent over backwards to save their replacements as large sheets of tempered glass for insulated units are extremely difficult to obtain. We added safety films to these and to all windows within 25’ above sidewalks or occupied roofs. We combined this film with solar films on existing windows except on north‐facing elevations. Removal of metallic solar films inherited from the OPEC days of 1979 re‐established Sert’s figure‐ground patterning of facades as today’s films are far more transparent.

The design approach at Smith Campus Center interweaves contemporary facades and interior spaces with Sert’s 55‐year old massing and elevations. New glazed facades are by Roschmann Steel and Glass Construction in Germany. The approach to transparency is less about massing and more about revealing interior spaces and activity to the pedestrian world surrounding the building. Harvard programmed the building to welcome the public, existing students, faculty, and staff to a series of gathering places, meeting rooms, and food venues. Hopkins wanted the glazing to disappear. A series of story‐height sliding doors a plaza level will accomplish just that.

Germany has produced several firms like Roschmann and Gartner that manufacture both engineered glazing and support steel along with fastening systems. This certainly helps with coordination and Ken Amano will talk about their highly effective design/assist involvement. These images show a new multi‐story gathering space with transparent roof, elevated walkways, and interior garden of woodland plants open to the sky.

In terms of thermal control and energy conservation the new structures are overglazed, but the many layers of transparency and overlapping sources of natural light are key to integrating new with old while overcoming urban design failings of the original design. People sit everywhere and every corner is visible from several points‐of‐view and therefore feels safe.

The woodland garden in the vitrine is open to the sky and can be seen across from one street to another through the building. This image shows how the base plates for “castellated” mullions are integrated into the enclosures for mechanical systems. It also shows the invisibility of the toggle system and the sealant joint between panes instead of physical mullion penetration.

Glazing panels are typically about 7’x11’ with some taller units. IGUs are 25 mm thick and made up of paired laminated panes of low‐iron separated by an Argon‐filled void. There are eight surfaces, one with low‐emissivity coating. Glass is by St. Gobain. This image shows the new exterior at its best. Mostly during daytime the glass is reflective, especially when viewed at an angle. The lack of exterior mullions emphasizes the panes’ surfaces with any distortion due to tiny differences in planar alignment and reduces the sense of transparency compared to Sert’s steel mullions that project to separate large single panes.

The façade restoration project addressed concrete cracks and spalling, accumulated stains, PCB impregnation of concrete by original sealants, failed glazing units‐ especially the translucent Plyglass panes, and solar control and safety films. The building was fully scaffolded. The extent of disruption was only partially foreseen‐ mostly vibration and noise associated with concrete cutting and chipping. More disturbing to occupants was the noise of sealant removal and pumps for cleaning concrete with water‐borne garnet sand. Harvard decided not to replace the single‐glazed windows in order to keep existing offices running throughout the construction period. This decision may have been influenced by inadequate life‐cycle costing that would include occupany’s effect on logistics, and energy savings. For buildings of this scale and construction type the logistical implications of continued occupancy are more significant than with most other campus buildings and warrant very thorough study to quantify and manage the costs of disruption. This is Sert’s south‐facing façade with precast brise soleils, precast mullions, and different window types.

Inadequate cover to reinforcing steel, cracks that admit water, advancing carbonation, and loss of alkalinity combine to cause spalling. This photo illustrates rectilinear cutting for a Dutchman‐like patch, chipping away of concrete behind the rebars, stainless steel anchors for repair concrete, zinc‐rich rebar coating, and cathodic protection puck. Simpson Gumpertz + Heger hand‐sounded the entire façade after scaffolding was in place. The CDs located cracks, spalls, and efflorescence from sample areas accessed by lift along with ground‐level binocular surveys. Board marking obscured fine cracks and the final number of repairs was more than double the locations shown on drawings along with contingency.

It takes time to establish mix designs that can practically address the many different textures and tonalities in cast‐in‐ place concrete within the same building. These in situ sample patches demonstrate differences among cement type, aggregates including sand. Comparison requires a number of likely mix types, almost a month of cure time for each successive trial, and a single surface treatment. See the difference between earlier mixes at the top and the lower eight.

At Harvard, we started with our most successful mix from another Sert tower at Boston University, but along with 37 different cleaning trials, mix selection at Harvard took about three months during pre‐construction.

Working with the Cambridge Historical Commission, we tried to leave one stair tower of board marked cast‐in‐place concrete without cleaning. The contrast between repairs and soiled concrete convinced Harvard to pay for cleaning the remaining three towers where patches are now far less visible. When Harvard built the second phase of Sert’s building, they economized by using ungalvanized reinforcement in the precast mullions. Within 12 years, they had spalled so badly that they were replaced with white aluminum. We specified new silicone sealants in white to match the aluminum and within two months many of the sealants were stained to a rust color by contact with concealed hydrocarbons, fragments of neoprene, or other active contaminants. We changed the sealant color to gray and the staining was invisible. The glazing on this façade has new solar film. In 1979, the metallic films made the clear glass vision panels look as gray as the translucent panels destroying Sert’s animated pattern of unpredictable window alignments.

The 3” thick (or THIN!) precast brise soleils, mullions, and sills failed rarely, but almost always because of inadequate structural anchorage and later repairs. The stain on the right is from sealant oils impregnating the concrete and trapping dirt. Low pressure garnet sand blasting required that water be trapped for ticketed disposal due to PCBs from the 1960s sealants. Matching precast was easier than the cast‐in‐place because the aggregate was more dominant than paste and relatively easy to match. The cleaned precast re‐established its lighter tone and contrast with tan cast‐in‐place as Sert intended.

We discovered that in situ precast repairs were faster and easier than removal and replication. This image shows the most distressed panel of a brise soleils with new anchorage to the building.

Sert used a much smoother concrete a street level. This retail pavilion was chemically cleaned to preserve the original plywood formwork markings on its smooth concrete. The windows are original (post 1970 riot). Concrete repair patches are beginning to weather and become less noticeable. Actually, everyone on this street is looking at their phone‐ not the architecture…

This completion photo of the south façade shows the Hopkins insertion between concrete masses, the cleaned and repaired concrete, and the relatively infrequent wide vision panels interrupting the repetitive mullion spacing. It is now a very popular destination and the mid‐century modernism is becoming fashionable once again. When it was completed in 1965, the most frequent question about it was when will it be finished? When will the brick be applied?

Facades+ Conference: Boston 2019 Henry Moss

Bruner/Cott Architects end