This is part 4 of 6
Working with historic materials
One of the largest challenges in the restoration was to effectively clean and repair the wide palette of materials Kimball had used in the decorative façade to express his client’s taste for opulence (Table 1). The restoration required all these elements to be cleaned, or replaced/repaired/repainted as appropriate. Each presented specific challenges to the team, whether through the selection of appropriate lime putty mortar to match existing; replication of the terra cotta bricks for spot replacements; sourcing new stone to match existing; cleaning the stone and terra cotta; or patching the stone with Jahn repair mortars.
The project was funded with money from the American Recovery and Reinvestment Act 2009, so this introduced additional challenges to comply with the Act’s “Buy America” clause. This imposed a general requirement that any public building or works project funded by the stimulus package must use only iron, steel and other manufactured goods produced in the US. As the contract documentation was mostly prepared before this Act came into existence,
it presented the team with a huge challenge to achieve compliance, effectively requiring redesign of many key components while the works were under way.
Inside the building the team faced similar issues, sourcing three kinds of wood for the window framing repairs, and matching marble for the decorative wainscot panelling and floor tiles and slate for repairs to the historic stair core.
Cast iron façade repairs
The original design called for wholesale removal of cast iron elements in the façade, so that individual pieces could be documented, cleaned down to bare metal, and repainted. Damaged pieces were to be replicated. This approach was based on the assumption that the cast iron was erected bottom-up after the main masonry façade, as was common in many similar buildings of this period.
Work commenced with the careful removal of the decorative cast iron leaf-shaped tracery elements that stood proud of the window framing (Fig 15). Unfortunately, as removal of the internal wood window framing began, it was discovered that the original cast iron armatures were built into and behind the decorative terracotta window surrounds. It would be impossible to remove these elements without wholesale damage to the terracotta, and extending the project schedule by several months. Fortunately the back side of the cast iron windows was found to be generally in excellent condition.
Faced with this, the team recommended a change of approach. While the multiple small decorative elements fixed onto the face of the windows would still be removed and either cleaned off-site or replicated if too damaged, the cast iron windows and frames would be left in place and repaired there.
This gave the team new challenges: to clean and repaint the cast iron in situ, repair non-structural cracking in infill panels and window sills, and the in situ structural repairs to load-bearing armatures.
Solutions had to be rapidly developed while engaging the contractor to perform necessary field testing and mock-ups to ensure that both designer and client were happy with the final solutions.
Cleaning and repainting cast iron in situ
The team conducted shop and field testing of several cleaning options, including Vacu-Blast and needle-guns, to determine which would be most effective. Traditional blasting was considered but dismissed, due to the likely excessive cost of site containment and blast media collection. Although Vacu-Blast performed well in the shop, it did not translate effectively to the field due to the quantity of decoration on the existing metalwork forming a high relief and preventing a good seal between equipment and working face.
Fortunately the needle-gun (a drill-like device with multiple metal needles driven percussively by pneumatic action) proved effective in cleaning but still avoiding damage to the base metal decoration. Meanwhile, PACA examined the paint on the existing ironwork and came to a startling conclusion: it had not always been black as originally thought, but rather a bright red color (somewhat ironically named “shy cherry”). This was further backed up when PACA found a contemporary citation to Corbin as “the red building”.
Repair of non-structural cracking in infill panels and window sills
Cast iron cannot be easily welded, and material tests on the cast iron at Corbin confirmed high levels of carbon in the base metal that would precipitate cracking if welding was attempted. An alternative approach was needed, and research revealed “cold stitching”, a common means of repair to cast iron engine blocks that could potentially be applied here.
Cold stitching consists of carefully drilling a series of interconnecting holes through templates that allow shear keys to be installed, providing tensile resistance across the crack. The crack line is then drilled out with contiguous threaded holes which accept inserts of slightly larger diameter so as to achieve a compression fit. The system is then polished down to be flush with the surrounding metal (Figs 16–18).
The infill metal is a high nickel steel alloy with the same coefficient of expansion as cast iron, to avoid any stress cracking of the repairs when subjected to the norms of temperature cycles.
The method, although effective, is somewhat empirical, so the design team worked with the sub-contractor on a series of field tests to establish maximum allowable spacings and edge distances between the shear keys. Ultimately this solution proved very successful for the areas of plain cast iron, which could be polished down afterwards to form a smooth finish without any signs of the intervention. The system is difficult to apply to decorative areas, so it was fortunate that these were almost entirely intact, and thus left as found.
In situ structural repairs to loadbearing armatures
The structural repairs to the vertical mullions (Fig 20) were probably the most technically challenging impact of the decision to repair the cast iron façade in place. The existing armature comprised a hollow section built up from flat and U-shaped cast iron pieces with staggered joints.
As these vertical elements were very long, it was impractical to remove them without wholesale deconstruction of the façade, but typical corrosion patterns were only evident at the base of the armatures coincident with window sills (in many cases due to drains from 1970s air-conditioning units).
Arup worked with the specialist subcontractor to develop an internal “splint” detail — a series of stainless steel struts that could be inserted after the lower corroded part of the cast iron mullion was cut away (Fig 21). The repair could then be reclad with new cast iron pieces formed to the original profile. The internal splints were bolted to the armature above and below using flat head bolts in countersunk holes and the join between old and new repaired by cold stitching.
Terracotta replication and repair
Individually mapping each piece of terracotta in the façade for cleaning, repair or replacement was an exhaustive task undertaken by PACA with the specialist façade repair sub-contractor. Out of the total of over 5000, around 500 were beyond repair and needed to be replaced (including over 225 in the high-level parapet zone) (Fig 22).
Replicating terracotta is a complex process, and selection of the specialist supplier (Boston Valley Terra Cotta) was probably the most important decision in getting the right result on site. Replication starts with matching the clay body to give a close colour match after firing. Similar clay can give a range of colour depending on the heat of firing, so a series of firing tests had to be carried out by the supplier. Even then the natural variability of temperature in any kiln means that within a single firing some colour variation is to be expected.
Clay shrinks about 10% in firing, and the implications were significant. Instead of being formed from original pieces from the building, new moulds exactly 10% larger than the proposed finished article had to be made. This is no mean feat of artistic skill, as the sculptor creates the clay master working by eye from an original piece (Fig 23). Even when the moulds were formed, the inability to press out any kind of re-entrant detail required that the individual pieces be hand-finished and stippled/marked to match the originals prior to placing and firing in the kiln (Fig 24). Given the level of artisan skill needed, the typical price was around $500/piece, uninstalled.
The pieces not replicated needed to be cleaned. During design it had been envisioned that a dry system called “sponge jet” — bombarding the façade with thousands of micro-sponges that are then collected and recycled — would be an acceptably mild approach, but preliminary testing with this revealed the existing terracotta to have a very fragile fireskin, susceptible to mechanical damage.
The team looked for alternatives.
Trials using the wet Prosoco alkali-based cleaning agent were carried out (Fig 26), with different dwell times and various degrees of agitation, until an acceptable result was achieved. This had some advantages in that a similar system had been specified for nearby areas of stone cleaning, so the contractor could readily adapt his means and methods of protection to extend this approach.
The final results were generally very good, although micro-analysis of the terracotta surface showed that, despite the design team and contractor’s best efforts, significant areas of the fragile fireskin were lost. In fact it was highly probable that beneath the grime this had always been the case.
To extend the life of the newly cleaned terracotta, a specialist KEIM coating was used. Being a stone-based product, this was relatively inert, and had the benefit of consolidating the terracotta surface asreplacement for the fireskin, giving a more uniform surface appearance without acting as a cheaper sealant would, trapping salts and moisture within the terracotta and causing potential long-term damage.
As the different cleaning and coating systems were considered a significant change to the original details submitted to SHPO, the design team had to detail everything in a technical memo and seek SHPO approval, which was duly granted.
14. Restored room interior.
15. Detail of restored cast iron façade element.
16. Stages of cold stitching process.
17. Drilling holes for crack repair.
18. Completed repair to crack.
19. Damaged cast iron window before repair.
20. Detail of armature repair.
21. Armature repair: first stage was local removal of corroded elements.
22. Colour matching tests of the terracotta: clay body and firing times were varied to achieve near-perfect results.
23. Recreating the detailed design using originals as a guide.
24. Finished pieces ready to be fired.
25. Close examination was needed to ensure the fireskin was left intact.
26. Cleaning tests on the terracotta.
27–28. Details of restored terracotta façade with wood windows.
Originally published in the Arup Journal, November 11, 2014. Authors include: Ian Buckley, Craig Covil & Ricardo Pittella.
Coming up next: The Corbin Building, Fulton Center: Rediscovering and Renewing An Architectural Gem – Part 5