A Century of Ice:
the architecture of phase change

Or: 10 lessons learned leading up to the construction of the
Centennial Ice Pantheon.

by Pieter Sijpkes

"Mon pays ce n'est pas un pays, c'est l'hiver..." rings true for
all Northerners, but it is particularly evocative for children and
builders, who at the mere thought of winter conjure up visions of
construction materials falling from the sky and water turning to
rock. Snow forts and igloos have been the result of this insight by
kids and the native Innu since time immemorial, while ice palaces
have been built in Canada since 1883, as related in the excellent
book "Ice Palaces":
"At the 1882 banquet of the Montreal Snowshoe Club, R. D. McGibbon
suggested holding a regular winter sports festival, in order to
enjoy and show off the glories of Canadian Winter.... Ideal weather
helped to make the 1883 Winter Carnival a resounding success. An
estimated 15000 foreign visitors flocked to the city..McGibbon's
inauguration of the ice palace was one of the highlights of the
fete"

Designed by local architect A.C. Hutchison, this first Montreal Ice
Palace was a grand affair erected on what is now called Canada
Square. Built of ice blocks sawn from the frozen St. Lawrence
river, -a technique long used to harvest ice intended for
refrigeration-, this first palace started a boom of palace building
in northern cities. Ever bigger palaces were built in Montreal,
Quebec, Ottawa, St. Paul, Minneapolis and Leadville, Colorado until
the end of the century, when the fervour to construct ice palaces
started top die down. Even though ice structures continued to be
built throughout the 20th century, never was the monumentality of
the early era matched.


Let's jump forward to the 1970's, at the McGill campus.
Studying at the School of Architecture in the early seventies, my
student friends and I were quite intrigued by an ice structure
going up one winter. Built of manufactured ice blocks, the
structure`s size seemed disappointing in the context of the
surrounding buildings. Word had it that the sponsor of the winter
carnival donated an annual, fixed amount for construction, and that
inflation had gradually reduced the size of structures over time.
We asked ourselves how we might reduce the cost of ice
construction. Frozen laundry on a clothesline led us to design a
fabric-formed ice structure, supported by a wooden triangle
suspended from three sturdy campus trees.  A few frosty nights of
hosing this assembly down transformed it from a flapping in-the-
wind sheet of tensile fabric into a rigid, compressive structure.
I will never forget the excitement of cutting the three steel wires
that initially held up the structure.(picture of 1972 ice structure
from slide.)

Professor Peter Collins, had lectured us on the inconvertibility of
tensile forces into compressive ones and proposed 'a wager of 10
dollars' that the structure would not stand. Rarely did beer  taste
sweeter than the supply bought with that ten dollar bill.
 
Lesson no. 1: It can be quite delicious to disagree with your
professors, particularly when you're right

Standing up for several days, the structure collapsed suddenly
after a short thaw, giving rise to another lesson.

Lesson no. 2. Light-weight ice construction has little resistance
against even short-term melting

When a cold winter struck in 1981,the frozen-fabric idea was taken
up in a new context; this time I was a teacher and the students 
designers in competition with each other. Ed Hurkin's concept was
chosen, and what a striking structure he produced!   Three graceful
hyperbolic paraboloids assembled from 20 feet-long scaffolding
pipes covered by a sown together nylon membrane were erected on the
lower campus and sprayed with water. These three `hypars` formed a
complex double-curved nylon-reinforced ice space, admired by the
master of the hyperbolic paraboloid, Felix Candela, when I showed
him photographs once (Photo of structure : Colour Xerox large
format.) The surface of the structure was barely an inch thick, and
to make up for the ice sublimating ('gassing off'), we had to hose
it down every other day. Frances Bronet, then a student, now the
head of the School of Architecture at Rensselaer Polytechnic
Institute in Troy, NY, did a snappy computer analysis of this
structure's behaviour under wind loading. Red neon tubes were
fastened to the edges of the structure.The most beautiful pictures
were taken at night, when the ice conducted light along its surface
and glowed in a red hue. The McGill flag that we had proudly
fastened to the top, almost thirty feet in the air, was gone the
next morning.

Lesson no. 3.: Never underestimate the energy and ingenuity thieves
and vandals will expend on an unusual structure.

The following year we continued to work on the nylon ice method.
Randy Cohen's attempt to create a thirty foot crystal 'needle'
protruding from the frozen campus, Howard Davies' inflatables,
Stefan Wisniofski's tents and, most dramatic of all, Mark Pimlott's
attempt to create an ambitious structure supported by a cable slung
from the roof of the McLennan library, all were valiant experiments
that year.(pictures in slide format and black and white in book.)
The 'limits of growth' were stretched too far, however in the cable
suspended structure: the exposed site and a stiff wind made it
impossible to perform the magic of turning the flapping suspended
structure into a rigid, self supporting one by simply spraying it
with water. The material kept folding when the structure deformed
under the weight of the water or when wind deflected it. Instead 
of a graceful free-form structure, a frozen lump of nylon was the
result of this exercise jokingly called the "Edsel".

Lesson no.4: New techniques bring new opportunities as well as new
problems


Back to "Pure Ice"

The ice-on-nylon method was deemed not 'pure' by some critics, so 
our thoughts turned to methods of casting light-weight low-cost
blocks. The ubiquitous two-liter milk carton was chosen as the most
suitable form. We tried to recycle used cartons, but were slow at
acquiring the 2000 units needed. Fortunately one student had
connections to a milk processing plant, and soon a full truck with
empty milk cartons pulled up to the school. Unloading 2000 cartons,
filling them with water, submerging  them in hot water to shake the
ice blocks from the cartons and laying the blocks on the catenary
formwork in a slush mortar turned out to be quite a task.

Lesson no. 5: "Doing even the smallest thing two thousand times
takes a lot of time" 

Sometimes I wonder: did architecture student Robert Libman plot his 
future as the founder of the Equality party while he was drudging
away in the bitter cold of 1983?

Lesson no. 6: Cause and effect are not always easily linked. ("the
butterfly effect").

The design that year called for a catenary arch, twenty feet high
spanning twenty feet. A formwork , eight feet wide was constructed
in a catenary shape by simply following the curve traced out by a
free hanging rope, sagging twenty feet between two nails, driven
twenty feet apart into the wall of the engineering lab.
The plywood form was, of course, reversed to turn the downward
tensile curve into an upward compressive one.

(picture)

When we finished laying the blocks on the form it was almost dark,
and it was decided to take the form out only after the deep frost
expected overnight had done its work on the slushy array of
iceblocks. The next morning a small void had formed between the top
of the form and the ice: the slush joints had expanded when they
froze overnight .. and lifted the whole twenty-ton shell off the
form; having made that happy observastion it was easy to be
confident that 'we can take the form out now'.
It was sweet to see the civil engineers scratch their head once
more; lesson 1 was paying off. In addition to getting praise for
simply standing up, this particular structure was appreciated for
the effortless elegance that comes so naturally to the catenary
arch, as the work of Antonio Gaudi  shows over and over again.


Lesson no 7  Don't worry about emulating concepts- just make sure
to pick good ones.

Reusing the wooden catenary formwork ribs a year later, and
rearranging them to form a dome rather then a linear vault was a
simple experiment, meant to prove how much stronger a double curved
shell is when compared to a single curved arch. The idea was to use
an inflated plastic sheet as formwork between the radially arranged
ribs; but hard-to-repair tears in the nylon made that plan
impossible, and the plastic sheet was instead stretched between the
ribs. The final form of the catenary dome was compromised by this
change in technique: the level of grace of the previous year's
arch, was equalled  only by the 'lumpiness' of its successor.

Lesson no  8 The aesthetic success or failure of a structure may
depend on seemingly small variables

The value of this experiment was revealed through its demise. When
the warm weather came, the dome melted on the sunny side, but of
the six 'slices' forming the original dome, two stood up for almost
a week. This taught us how much stronger the double curved dome is
when compared to its single curved predecessor which had collapsed
quite suddenly.

Lesson no.9: Double curved surfaces are much stronger than flat or
single curved surfaces.

The method of subtraction.

Every child discovers how easy it is to dig into a mound of snow or
dirt, creating a cave. The 'architecture of subtraction' is as old
as architecture itself, and can be found in places as diverse in
time and space as China, where today millions of people live in
houses dug from the ground, to Roman temples carved out from the
sandstone cliffs at Petra in Jordania, to excavated churches of
Ethiopia and Capadocia. 
Making an "ice pub" for fifty people using snow and water was the
assignment students received in the winter of 1994. (picture)
After blowing a  pile of snow twenty feet high with a snowblower,
a dozen students dug to their heart's content. On  a cold Friday
night in January of 1994 we counted 52 people comfortably sipping
drinks protected by the vaulted roof and walls of the ice pub.

Lesson no.10 Subtraction can lead to the same result as addition in
architecture, given the proper materials.


CENTENARY ICE
The lessons applied.

To celebrate the School's centennial (later combined with the
University's 175's) the idea of a celebratory ice structure came
up. We wanted to evoke the grandeur of nineteenth century ice
structures with the experience we had garnered building with snow
and ice since the seventies.

Lesson no 2 had taught us that "small mass=short life", and ruled
out a lightweight approach for a structure that had to serve
various functions for two weeks. Lots of mass was the only defense
against evaporation as well a possible mid winter thaw and rain.
Depending on conditions, mass can be obtained by ice blocks,
massive amounts of snow, or by snow injected and sprayed with water
The search was on for a massive structure that would not only be
interesting to build, it had to be architecturally striking as
well. It may well have been the presence of Orson Wheeler's cutaway
model of the Pantheon in the glass case on the third floor of the
school that gave the impetus for the idea of using the Pantheon in
Rome as the inspiration for the Centennial ice structure.

The Pantheon fitted our bill perfectly: It was a massive structure,
and it was novel because it would span a sizeable space with ice.
We intended to change the character of Montreal's historical look-
at ice structures to a new live-in quality, analogous to the change
in antiquity from Greek look-at architecture to Roman live-in
architecture.
The Pantheon offered a two level challenge: spanning the space with
a massive dome was the main challenge, while decorating the
structure and adorning it with a monumental entrance was another.

The most difficult decision was the one of picking the size of the
structure. In the context of the campus anything under three
storeys had proven to be to too small to "hold its own", as we
observed in the ice structures of the sixties. The biggest clear
span we had reached so far in ice was 20 feet with the 4" thick
catenary ice arch and dome of 1983 and '84. Doubling that span
seemed too much in one step.. so a span and height of about 32 feet
(or 10 meters) was choosen. That gave the Ice Pantheon a scale of
a bit less than one to five of the original Pantheon. A scale of pi
(1 over 3.14) would have been the most elegant one, but that scale
would have pushed the span to a scary 45 feet.
After determining the size of the project, the method of
construction was next. One idea, following lesson 10 was to have
the snow pushed up to a 35 foot high mountain, and carve the
Pantheon out like the '94 snow pub. It was rejected because of
worries whether the snow would be sufficiently stable to hold up
before it could be reinforced by injected water. Precast snow-ice
blocks, moved into place by a crane were looked at but rejected as
to expensive in crane time. ( Lesson 5. it takes time doing many
things over and over again.) Finally a technique used for centuries
in adobe construction was adopted. Lightweight, eight foot long
curved plywood elements, laterally clamped together were the basis
of the system. Strung along to form two parallel circular walls,
kept apart by notched, two by four spreaders. a  four foot high
"donut"  was  formed, which was filled with snow, The forms were
removed and reinstalled on top of the newly cast snow below etc.
Since the two by four spreaders holding the bottom  of the form in
place had to be pulled out of the wall, tell-tale holes were left,
very much identical to the holes left by adobe construction.  This
system was very successful, (following part one of lesson 4): the
forms went up easily and the snow was dumped effortlessly by the
frontloader of the physical facilities department. There was no
shortage of snow in the winter of 95/96: like biblical manna, our
building material fell from the sky in prodigious quantities.
Our pace of construction slowed down considerably when we reached
level 4, 16 feet up. Handing up the forms to that height was
becoming tricky in the wind, and the frontloader could no longer
reach over the edge of the forms: we now had to shovel the snow
from the frontloader bucket into the forms.
Even that became impossible for the next level; for a while we used
buckets and pulleys to get the snow in place, but 1600 cubic feet
of snow were required per 4 foot layer of the structure: it became
clear that by hand we would run out of time. A roofer's motor-
driven hoist proved to be to difficult to operate efficiently.
After discussions with the helpful people of McGill's facilities
management an experiment was done using the university's snow
blower. The machine was certainly able to blow the snow high
enough, the problem was that it was hard to deposit the snow
exactly in the forms. Filling the first 4 foot layer of the
structure resulted in almost as much snow falling inside the walls
as inside the forms, and a massive cleanup of the inside space was
needed. To avoid having to do this several more times, it was
decided to try to roof the dome in 'one shot', rather than layer by
layer. This decision also was brought about by a desire to speed
the process up: the opening date was only ten days away, and warm
weather was announced for the days coming. The last twenty five
feet of the thirty-two foot span was domed over in a hurry, using
the inside as well as the outside wall forms, improvising patching
the pie shaped spaces between the panels, propping everything up
from a central point on top of the scaffolding in the center of the
structure. The beauty of the original design of the Pantheon 
became evident here: as the radius of the plan is the same as the
radius of the dome, the curved plywood wall panels could be used as
form work for the dome as well.
On Tuesday afternoon January 16 the snow was blown over the dome,
and disaster almost struck, (in accordance with the second half of
lesson 4): Even though the blower operator had become a real artist
controlling the machine and dropping tons of snow with pin-point
accuracy, having to drive around the structure to blow the snow on
top, it was inevitable that the dome would be loaded asymmetrically
for a few minutes at a time, and we all had the scare of our lives
when the wooden props inside started popping and buckling. Fearing
that the whole form might give way under the weight of tons of snow
we stayed close to the exit. But even though the form work settled
a foot or so, the dome consolidated and held. As with the catenary
ice dome, the beauty of the dome was marred by the unevenness
introduced by these unforeseen events; lesson 8 was learned again.

The next day a week long thaw started, with temperatures as high as
10 degrees, and the whole McGill Community seemed to be concerned
about the well being of the Pantheon. Despite the 'tropical'
temperatures, the thermal mass of the structure prevented disaster,
loosing only a few inches of its bulk in a week  of thaw. Lesson no
2 proved its worth.

The form work could be removed the day the temperature returned to
normal arctic levels, and it was a proud moment to be able to show
Mayor Bourque the structure just when the dome could be seen
without props, and it was another proud moment to be able to
welcome Principal Shapiro and Cancellor Chambers as well as many
members of the McGill community into the structure after it was
opened on January 26. That evening a concert by a McGill music
student Jazz combo was a magical experience: the cool music, the
candle-lighting, the hot mulled wine all contributed to the unique
event. The massive snow walls and roof created an outstanding
acoustic environment, while hardly a note could be heard outside in
the crisp minus 15 C air. 

The Ice Pantheon was primarily built to celebrate the Centenary of
the School of Architecture, and Friday February 2 was the night
chosen to celebrate the beginning of the centenary festivities.
For the event Professor David Covo had carved into the portico
pediment a marvelous scene portraying the Macdonald-Harrington
building, the Engineering library, the Ice Pantheon and several
reclining figures. (Photo.)    
A large group of members of The School community, over 125 strong,
gathered inside at after dark, for a historic picture taking
session. The slow exposure time required, combined with the
suitably cold minus 18 C degree temperature froze that event in 
our memory as well as in our limbs, as the suitably serious faces
in the picture shows. (Photo.)
The Centenary Committee chairman, Professor Bruce Anderson made a
much appreciated rousing speech welcoming eveyone and highlighting
the events to come during the centennial year. 
The willingness of humans to endure the cold for this historic
event was not matched by performance of the equipment belonging to
the rock band "the Snitches". Invited to perform as part of the
inaugural activities, the amplifiers of the group refused to work
in the cold, despite efforts to coddle them with blankets and
heaters, (a case of matter over mind,) and the concert was moved
into the Exhibition room of the school.

The Pantheon's formal role was now over, and it entered a stage
which was, in some ways, it's most interesting. The slow melt of
the structure was followed by the university community with great
interest, and  as the south wall and the dome slowly thinned, the
structure took on a romantic and ruinous quality that recalled its
progenitors in Rome. It became a late night place to observe the
moon through the oculus, and to be generally romantic. The entrance
was finally was boarded up and the perimeter fenced in to prevent
accidents and vandalism (lesson 3)  
The structure stood until the day before Good Friday, when it was
finally levelled by the same frontloader that filled the forms,
three months before.
As all other ice structures, the Centennial Pantheon disappeared
effortlessly as water into the ground and as vapour into the sky.
"Notre Pays" does have a summer, after all.

























References

http:132.206.23.82/archdocs/events/ice/index.htm

http://softeng.cs.mcgill.ca:80/ice/

"Ice Palaces" Fred Anderes and Ann Agranoff, Macmillan of Canada
1983.

"Teaching the properties of thin shells to undergraduta architects-
An empirical method using ice as a construction material".
Preoceedings of the First International Conference On Lightweight
structures in Architecture, Sydney. Australia. 1986

Credits.
To all the students who have slaved over the years to prove that
"Mon pays ce n'est pas un pays, c'est l'hiver.