The final construction time lapses have been assembled. Short but sweet. We hope it goes so quickly for our 5 day construction schedule at the Cooper-Hewitt!
Especially for the die-hard structural engineers, here is Part I of the demolition: the 10 minutes of springing displacement and intermediary failure before the final collapse (Part II in previous post). It still has its dramatic moments and demonstrates the hinging mechanisms which form in the vault, leading to its collapse at 6 inches of displacement. Keep in mind that there were also several significant sledgehammer blows directed from the inside of the vault around the quarter-span region – precisely at the level where we will see the final failure hinges.
Play as we did, the real end-game – as JAO mentioned – was the inevitable dominance of gravity as Silman gamingly abused our vault. The final event – to destruction – would test our design intention to accommodate all horizontal thrust through friction along the base of the vault springings. The Silman engineers marked out half-inch increments and then displaced one springing with sledgehammers to see how long it would hold up, observing at each major behavioral alteration the cracks and mechanisms which were developing in the vault shell.
“So it is down to you, and it is down to me! If you wish her dead, by all means, keep moving forward.” – TPB
But since she did hold up rather long indeed – for an awesome horizontal displacement of approximately 6 inches, “off the charts” – I will include now the sudden-death moment of demolition, and add the drawn-out version for the structural masonry fanatics when I have 4 or more hours of uninterrupted upload time…
What remained was intriguing, really. That undulating vault is no more; but – even in ruins – she still had her curves. Our great thanks to the engineers from Silman, Smithsonian reporter Logan Ward, MIT wood-shop director Chris Dewart, JAO, and the vault team!
Signing off for now –
Lara
And so it has come to a close. The end-game for this January’s prototype vault was a project for the Masonry Research Group and Robert Silman Associates, the structural engineering firm charged with the verification of our vault’s structural stability and safety. The task was to de-center the vault under their observation, and then to test the many ways in which it might be made to fail. A coup for the Masonry Research Group would be a successful de-centering – and then the longest endurance against the beatings of our counterparts in construction: Silman the demolishers!
Here are just a few clips of the various “tests” concocted by Silman and John Ochsendorf:
But what playful and sensitive demolishers they were. Have you ever heard of a destroyer who listened with his hands? During the second video, Derek Trelstad sounds the frequencies induced by the hammer with his hands, measuring the differential between those frequencies conveyed through the end arch of the vault and those conveyed through a section that he had smashed-in with a rubber mallet. More such tests occurred… indeed, our vault became the structural jungle-gym for our engineers as they sought out the weaknesses and strengths of our vault – our advisor John O among them of course. It was really quite amazing to watch them all in action. Despite our team’s initial reluctance to destroy our first undulating, unobstructed view of the de-centered vault, we soon caught their spirit and followed suit:
– Lara
De-centering was a little hair-raising, especially with an audience, but watch her fly!
Here is a video clip of the last of the de-centering process:
We haven’t posted in a few days, an effect no doubt of the crunchy lack of sensation in one’s fingertips and the moderately enfeebling effect of daily overexertion – perhaps from carrying the extra weight of mortar in cuffs, pockets, and hair, to be later deposited on the floor of one’s apartment. So it feels as though I must go back to briefly recap the closing. Indeed, very briefly.
(Coming soon… Stop-motion construction.)
In fact it was done almost as soon as it had begun; in six days, all masonry had been laid. The most challenging – and yet gratifying – part was the closing-in, each course in succession as its last bricks were laid to complete an arch. These sections had constructional complexity that involved: 1. custom-cutting of bricks to fit within the arches, 2. a condition of full cantilever in which, against gravity, the mason must support bricks until mortar has been sufficiently set, and 3. the most extreme curvature of the vault which required steep joint hinging and difficult mortaring between each course for it to turn into the curvature.
Though I might add that the greatest difficulty by far was squeezing into or over the deep and narrow grid-shell compartments of our formwork. Since the brick itself was designed by the measure of the body, I think we will probably reconsider the usefulness of such simple measures as shoulder-spans in the next version.
Through these days, my greatest thanks go to Mallory, my mortar mixing and brick-laying teammate, who worked patiently alongside me to spot me in the hairiest of moments, to read, anticipate and direct the myriad of tasks necessary to set a good brick – and occasionally, to catch a falling glob of wet mortar. None of this work would be possible without the patience, timing and teamwork between the members of our vault crew – in all facets of vault design, formwork design, materials sourcing, fabrication and construction.
The satisfaction of the final brick was meant to go to John O’s toddler, whose timbrel-flavored, multi-syllabic vocabulary includes such words as “Guastavino” and “Barcelona”. We considered how perfect it would be for John to claim to his future students that his daughter completed her first vault before age two. Not this time, I’m afraid – though she did practice with warmish globs of play-doh Hydrocal, holding it up with a brick as though to indicate she knew they went together. Soon enough, young vaulter.
– Lara
Consider for a moment that a masonry structure is composed of details, many details, which in their whole organization must come to work as a system. So now, when we look at the detail – our brick – we must always see two things at once: a unit (which has its ideal position) and a system (composed of these units). In a 3d model, these units always fall in their proper place – but a brick, well, it is an idiosyncratic thing! Each brick has its own shrinkage proportions based on its position in the kiln and exposure to heat, it has cracks which distort it, it has uneven edges on the surface side of sand-molded bricks, such as the ones we are currently using. Each brick has a different water content (since they must be soaked in sequence in preparation for laying). Each brick is laid in a sequence of plaster mortar batches, during which period plaster changes its character from wet and runny to dry and thick, thus slightly altering the thickness of a mortar joint. These are some of the many myriad of things which make a brick an idiosyncratic detail of a highly organized system. So the devil is here in the details… in sorting them out to best position a wonky brick (this is a technical term), and to see the whole system in such a light that it may respond to the alterations that each brick induces into the often cascading distortions of a system. But to see these dramatic distortions, one must anticipate them as they aggregate, indeed, one must anticipate them in the very inclination of each brick. And the difficulty of this task increases as the geometry of the overall system is complexified.
For this reason, most especially for complicated geometries, registration or guide lines become very important to ensure that the masonry system does not distort beyond an acceptable range. In our case, with a one inch thick shell, the most critical distortion which cannot be accepted is that which falls significantly outside of the range of our catenary thrust lines.
For any pattern that is established as a masonry system, there are counter-patterns which the bricks may take. Personally, I find these patterns extremely interesting, so I study them. They are often as beautiful as a regular masonry pattern, yet they have a tendency towards entropy – a pattern that destroys itself. The mason directs negative entropy in the system, the entropy which must be exported for the system to correct itself.
I will take up Scott’s last thread, in the hopes to spur a blog-debate to keep our minds active while we work:
It is true, new technologies such as CNC fabrication do enable new aesthetic and structural potentials, which were unavailable to previous architects, engineers, and builders of masonry. However, I believe – more importantly perhaps – new technologies also create new kinds of errors, which require innovative responses. And it is the process of discovering error through models and full prototypes, as this one, which allows a design to evolve. What I find extremely applicable here about the greatest structures in the history of masonry innovation, is that their daring created problems which had to be resolved by their own architects and their inheritors. Just as di Cambio constructed the 42′ span octagonal drum for the dome of the Florence Cathedral, which could not be built for more than one hundred years, until Brunelleschi’s double-shell innovation proposed to span it without formwork. Just as Isodore of Miletus and Anthemius of Tralles built a dome which collapsed repeatedly, and had to be repaired and renovated by Sinan and other Ottoman architects. Just as the fear of collapse of the dome of St. Peter’s caused Poleni to so greatly advance contemporary theory of masonry structure through his analysis of Hooke’s ‘hanging chain’. My colleague Peter Christiansen once queried whether it was the history of innovation in engineering which led to such impressive structures, or whether it was the failure of impressive and daring structures, which required – post-hoc – more significant advances in structural engineering to solve the impossible problems they first presented.
I do not believe that we can stand upon the shoulders of the masonry greats, simply because they have been built and engineering has advanced – because we have not yet had the opportunity to learn first-hand from their failures. Perhaps we will. Several weeks ago, some generous project managers of the MIT dome renovation took me for a tour: The view from the very top of the dome is one to which only an MIT hacker so aspires, ah, but more exciting by far was to see what Walsh Brothers excavated from the experimental original thin-shell dome skylight, which has been leaking since before WWII. And so much more by far, the view from between the two domes: the indexes of ship-building technology in 8″ formwork board marks on the inside of the outer structural concrete shell, arrayed willy-nilly with no apparent geometrical coherency – yet still describing a perfect surface for the eye of these early 20th century masons. How does the mason see this surface? How does the computer see it?
This grid-shell of formwork profiles and thin grid of mason’s line approximates a surface, which our team must now begin to ’see’ with brick. It will not be an easy task to develop the skills of the eye in ‘lofting’ such a surface, though these are skills to which the human eye (and in particular the eye of the mason) is much better adapted than the automated robot – as Gramazio and Kohler of the ETH in Switzerland have discovered through their robotic masonry constructions. This surface curves in two directions, and each bricks must be placed with respect to its relationship to its neighbors and its role in correcting, adjusting, “splitting the difference” in the errors of its neighbors in describing the whole surface. This is the fun part.
– Lara
Gentle readers,
You may have noticed that there are many faces behind this project, reflected in the varied voices heard here in this blog. There are seven of us on the ground here at MIT (as evidenced by the above picture from last September), and over the course of the project, four have put fingers to keys and contributed thoughts to cyberspace (newest voice being Scott’s from yesterday). Come Spring, I hope that everyone will have their say here – after all, we’re a pretty democratic crew.
Someone asked me recently what my role on the project is. Somehow this question puzzled me: role? I managed to mention a few of the tasks I happen to take lead in, but in reality, we each do our fair share of lifting and have a say in all things. Yes, there are some who are gurus in certain areas more than others, but hey – as a team, I would say we act pretty communally (sometimes too much so). In our project update emails to our advisor, we even sign them as “MRG,” standing for “Masonry Research Group,” despite the fact that they are clearly sent from one account. We aren’t a perfect Brady Bunch, but at least we’re not as dysfunctional as Al Bundy’s household. (O imperfect analogies.) And to think, half of us barely knew each other just last summer!
- – - back to regular programming – - -
Anyways, on to the actual project update:
Current status:
Formwork assembled, mechanical fasteners installed, guide strings strung, bricks cleaned and custom cut as needed.
Details:
Today was one for taking care of some unanticipated tasks that were unexpectedly tedious yet necessary. Yesterday we completed the formwork and installed all mechanical fasteners in each joint, which is needed to help the wood counteract any lateral stresses. This was a great milestone to achieve, but necessitated today’s check, going through each and every one of the 440 bolts, washers, and nuts to make sure they were present and tightened. After kneeling, peering, and straining to reach all of them for the second time in two days, Lara’s comment was: “This is ‘Crouching Tiger, Pain in the _____.” (We have a sense of humor.)
Installing the strings that would act as guides for each brick course also proved to be a headache – literally for Mallory, who had to endure the bolt of string hitting her on the head several times as it was thrown from one side of the vault to the other (see photo for why).
And finally, we were at the point of laying bricks – ! – when we realized they were covered in sand and needed to be cleaned, in order for the plaster mortar to adhere correctly. This lead to a brick scrubbing party outside with buckets of warm water and bare hands. Thankfully, God granted us with 45 degree weather – truly a miracle at this time of year in Boston.
Cutting the bricks was also an endeavor in itself. It inspired a new fashion statement on the job site, but I have to get the model’s permission first before unveiling it here.
Sorry we promised bricks today, but they will be going up tomorrow for sure. Stay tuned.
- eLo
The Pantheon
Florence Duomo
Hagia Sophia
Rafael Guastavino - file him under 'Takes Care of Business'
The Baths of Caracalla. The Pantheon. Florence Duomo. Hagia Sophia. Decent efforts all, but consider that every such masonry vault that has existed previously has led us to this moment, the culmination of some thousands of years of building tradition. When we trace the lineage of this project, we count among our forebears giants of the ages, designers such as Rafael Guastavino and Eladio Dieste. Out of the most common of building materials, they produced the original and the resplendent. They remain the standard by which masonry vaults must be compared, and to participate in this project is to stand on their shoulders.
Yet we have something none of those previous generations had. Something with the potential to revolutionize the design and construction process – a robot. Just what is a robot, you ask? Ours is a Computer Numerical Control Router, commonly referred to as a CNC machine. The CNC has produced the panels of plywood formwork for the latest iteration of our vault, completing in a matter of hours the same steps we measured in days and weeks of construction time for our first iteration. Afforded this new power, however, one must remain vigilant. Unfeeling, unflinching, and capable of extraordinary cutting speed and precision, robots would no doubt prove formidable adversaries should they one day rise up against us and attempt to destroy us; for now, they remain our allies and friends, and the CNC machine, under constant watch, has proved remarkably adept at carrying out our bidding.
Already, word of mouth about our project has begun to generate considerable press buzz*. The first brick will be laid tomorrow, from which point the project will proceed quickly. Regular progress updates to follow.
STF
*Please inform any friends who may be interested in this project, but especially any journalist friends. You don’t have to know them that well, or at all, really.
