A pattern language

Where can the need for concealment be expressed; the need to hide; the need for something precious to be lost, and then revealed ?

We believe that there is a need in people to live with a secret place in their homes: a place that is used in special ways, and revealed only at very special moments.

To live in a home where there is such a place alters your experience. It invites you to put something precious there, to conceal, to let only some in on the secret and not others. It allows you to keep something that is precious in an entirely personal way, so that no one may ever find it, until the moment you say to your friend, “Now I am going to show you something special”—and tell the story behind it.

There is strong support for the reality of this need in Gaston Bachelard’s The Poetics of Sface (New York: The Omen Press, 1964). We quote from Chapter 3;

With the theme of drawers, chests, locks and wardrobes, we shall resume contact with the unfathomable store of daydreams of intimacy.

Wardrobes with their shelves, desks with their drawers, and chests with their false bottoms are veritable organs of the secret psychological life. Indeed, without these “objects” and a few others in equally high favor, our intimate life would lack a model of intimacy. They are hybrid objects, subject objects. Like us, through us and for us, they have a quality of intimacy. . . .

If we give objects the friendship they should have, we do not open a wardrobe without a slight start. Beneath its russet wood, a wardrobe is a very white almond. To open it, is to experience an event of whiteness.

930

An anthology devoted to small boxes, such as chests and caskets, would constitute an important chapter in psychology. These complex pieces that a craftsman creates are very evident witnesses of the need for secrecy, of an intuitive sense of hiding places. It is not merely a matter of keeping a possession well guarded. The lock doesn’t exist that could resist absolute violence, and all locks are an invitation to thieves. A lock is a psychological threshold. . . .

Therefore:

Make a place in the house, perhaps only a few feet square, which is kept locked and secret; a place which is virtually impossible to discover—until you have been shown where it is; a place where the archives of the house, or other more potent secrets, might be kept.

Classic types of secret places are the panel that slides back, revealing the cavity in the wall, the loose board beneath the rug, the trap door—closets between rooms (198), thickening THE OUTER WALLS (21 I) , FLOOR-CEILING VAULTS

(219). . . .

931

CONSTRUCTION

At this stage, you have a complete design for an individual building. If you have followed the patterns given, you have a scheme of spaces, either marked on the ground, with stakes, or on a piece of paper, accurate to the nearest foot or so. You know the height of rooms, the rough size and position of windows and doors, and you know roughly how the roofs of the building, and the gardens are laid out.

The next, and last part of the language, tells you how to make a buildable building directly from this rough scheme of spaces, and tells you how to build it, in detail.

* *

The patterns in this last section present a physical attitude to construction that works together with the kinds of buildings which the second part of the pattern language generates. These construction patterns are intended for builders—whether professional builders, or amateur owner-builders.

Each pattern states a principle about structure and materials. These principles can be implemented in any number of ways when it comes time for actual building. We have tried to state various ways in which the principles can be built. But, partly because these patterns are the least developed, and partly because of the nature of building patterns, the reader will very likely have much to add to these patterns. For example, the actual materials used to implement them will vary greatly from region to region . . .

Perhaps the main thing to bear in mind, as you look over this material, is this: Our intention in this section

935

places can be in the boundaries of neighborhoods, communities and subcultures—see subculture boundary (13); others, not noisy or noxious, can be built right into homes and neighborhoods. In both cases, the crucial fact is this: every home is within a jew minutes of dozens of workplaces. Then each household would have the chance to create for itself an intimate ecology of home and work: all its members have the option of arranging a workplace for themselves close to each other and their friends. People can meet for lunch, children can drop in, workers can run home. And under the prompting of such connections the workplaces themselves will inevitably become nicer places, more like homes, where life is carried on, not banished for eight hours.

This pattern is natural in traditional societies, where workplaces are relatively small and households comparatively self-sufficient. But is it compatible with the facts of high technology and the concentration of workers in factories? How strong is the need for workplaces to be near each other?

The main argument behind the centralization of plants, and their gradual increase in size, is an economic one. It has been demonstrated over and again that there are economies of scale in production, advantages which accrue from producing a huge number of goods or services in one place.

However, large centralized organizations are not intrinsic to mass production. There are many excellent examples which demonstrate the fact that where work is substantially scattered, people can still produce goods and services of enormous complexity. One of the best historical examples is the Jura Federation of watchmakers, formed in the mountain villages of Switzerland in the early 1870*5. These workers produced watches in their home workshops, each preserving his independence while coordinating his efforts with other craftsmen from the surrounding villages. (For an account of this federation, see, for example, George Woodcock, Anarchism: A History of Libertarian Ideas and Movements, Cleveland: Meridian Books, 1962, pp. 168—69.)

In our own time, Raymond Vernon has shown that small, scattered workplaces in the New York metropolitan economy, respond much faster to changing demands and supplies, and that the degree of creativity in agglomerations of small businesses is vastly greater than that of the more cumbersome and centralized

CONSTRUCTION

has been to provide an alternative to the technocratic and rigid ways of building that have become the legacy of the machine age and modern architecture.

The way of building described here leads to buildings that are unique and tailored to their sites. It depends on builders taking responsibility for their work; and working out the details of the building as they go—mocking up entrances and windows and the dimensions of spaces, making experiments, and building directly according to the results.

The patterns in this section are unique in several ways.

First, the sequence of the patterns is more concrete than in any of the earlier portions of the language. It not only corresponds to the order in which a design matures conceptually, in the user’s mind, but also corresponds to the actual physical order of construction. That is, except for the first four patterns, which deal with structural philosophy, the remaining patterns can actually be used, in the sequence given, to build a building. The sequence of the language corresponds almost exactly, to the actual sequence of operations on the building site. In addition, the patterns themselves in this section are both more concrete, and more abstract, than any other patterns in the language.

They are more concrete because, with each pattern, we have always given at least one interpretation which can be built directly. For instance, with the pattern root foundation, we have given one particular interpretation, to show that it can be done, and also to give the reader an immediate, and practical, buildable approach to construction.

CONSTRUCTION

Yet at the same time, they are also more abstract. The particular concrete formulation which we have given for each pattern, can also be interpreted, and remade in a thousand ways. Thus, it is also possible to take the general idea of the pattern, the idea that the foundation functions like a tree root, in the way that it anchors the building in the ground—and invent a dozen entirely different physical systems, which all work in this fundamental way. In this sense, these patterns are more abstract than any others in the book, since they have a wider range of possible interpretations.

To illustrate the fact that a great variety of actual building systems can be developed, based on these patterns, we present three versions that we have developed, in response to different contexts.

In Mexico: Concrete block foundations with re-bar connectors; hollow self-aligning molded earth blocks reinforced with bamboo for walls and columns; burlap formed concrete beams; steep barrel vaults with earth and asphalt covering—everything whitewashed.

In Peru: Slab floors poured integrally with wall foundations j finished with soft baked tiles j hard wood (diablo fuerte) columns and beams ; plaster on bamboo lath acting as shear walls between columns j diagonal wood plank ceiling/floors; bamboo lattice partitions.

In Berkeley: Concrete slab finished with colored wax; walls of exterior skin of i x boards and interior skin of gypboard filled with light weight concrete; box columns made of i x boards, filled with lightweight concrete j 2-inch concrete ceiling/floor vaults formed with wood lattice and burlap forms.

937

CONSTRUCTION

As you can see from these examples, we have formulated these patterns with very careful attention to cost. We have tried to give examples of these patterns which use the cheapest, and most easily available, materials $ we have designed them in such a way that such buildings can be built by lay people (who can therefore avoid the cost of labor altogether); and we have designed it so that the cost of labor, if done professionally, is also low.

Of the three parts of the language, this third part is the least developed. Both the part on Towns and the part on Buildings have been tested, one partially, the other very thoroughly, in practice. This third part has so far only been tested in a small number of relatively minor buildings. That means, obviously, that this material needs a good deal of improvement.

However, we intend, as soon as possible, to test all these patterns thoroughly in various different buildings •—houses, public buildings, details, and additions. Once again, as soon as we have enough examples to make it worth reporting on them, we shall publish another volume which describes them, and our findings.

In many ways, rough though it is, this is the most exciting part of the language, because it is here, in these few patterns, that we can most vividly see a building literally grow before our eyes, under the impact of the patterns.

The actual process of construction, in which the sequence of their patterns creates a building, is described in chapter 23 of The Timeless Way.

938

Before you lay out construction details, establish a fhilosofhy of structure which will let the structure grow directly from your flans and your conceftion °f the buildings.

205. STRUCTURE FOLLOWS SOCIAL SPACES

206. EFFICIENT STRUCTURE

207. GOOD MATERIALS

208. GRADUAL STIFFENING

205 STRUCTURE FOLLOWS SOCIAL SPACES**

940

. . . if you have used the- earlier patterns in the language, your plans are based on subtle arrangements of social spaces. But the beauty and subtlety of all these social spaces will be destroyed, when you start building, unless you find a way of building which is able to follow the social spaces without distorting or rearranging them for engineering reasons.

This pattern gives you the beginning of such a way of building. It is the first of the 49 patterns which deal specifically with structure and construction; it is the bottleneck through which all languages pass from the larger patterns for rooms and building layout to the smaller ones which specify the process of construction. It not only has its own intrinsic arguments about the relation between social spaces and load-bearing structure—it also contains, at the end, a list of all the connections which you need for patterns on structure, columns, walls, floors, roofs, and all the details of construction.

v v •:*

No building ever feels right to the people in it unless the physical spaces (defined by columns, walls, and ceilings) are congruent with the social spaces (defined by activities and human groups).

And yet this congruence is hardly ever present in modern construction. Most often the physical and social spaces are incon-gruent. Modern construction—that is, the form of construction most commonly practiced in the mid-twentieth century— usually forces social spaces into the framework of a building whose shape is given by engineering considerations.

There are two different versions of this incongruence.

On the one hand, there are those buildings whose structural form is very demanding indeed and actually forces the social space to follow the shape of the construction—Buckminster Fuller domes, hyperbolic paraboloids, tension structures are examples.

On the other hand, there are those buildings in which there are very few structural elements—a few giant columns and no

more. In these buildings the social spaces are defined by lightweight nonstructural partitions floating free within the “neutral” physical structure given by the engineering. The buildings of Mies van der Rohe and Skidmore Owings and Merrill are examples.

We shall now argue that both these kinds of incongruence do fundamental damage—for entirely different reasons.

In the first case the structure does damage simply because it constrains the social space and makes it different from what it naturally wants to be. To be specific: we know from our experiments that people are able to use this pattern language to design buildings for themselves; and that the plans they create, unhampered by other considerations, have an astonishing range of free arrangements, always finely tuned to the details of their lives and habits.

Any form of construction which makes it impossible to implement these plans and forces them into the strait jacket of an alien geometry, simply for structural reasons, is doing social damage.

Of course, it could be argued that the structural needs of a building are as much a part of its nature as the social and psychological needs of its inhabitants. This argument might perhaps, perhaps, hold water if there were indeed no way of building buildings which conform more exactly to the loose plans based on activities alone.

But the next fete 'patterns in this hook make it very clear that there do exist ways of building which are structurally sound and yet perfectly congruent with social space, without any compromise whatever. I't is therefore clear that we may legitimately reject any form of construction which cannot adapt itself perfectly to the forms of space required by social action.

What of the second kind of incongruence between social space and building form—the kind where the structure creates huge areas of almost uninterrupted “flexible” space, punctuated by occasional columns, and the social spaces are created inside this framework by nonstructural partitions.

Once again, many important patterns cannot be incorporated into the design—fight on two sides of every room (159), for example simply cannot be included in a giant rectangle. But in this type of building, there is an additional kind of incongruence between social space and engineering structure which comes from the fact that the two are virtually independent of each other. The engineering follows its own laws, the social space follows its laws—and they do not match.

This mismatch is perceived and felt not merely as a mismatch, but as a fundamental and disturbing incoherence in the fabric of the building, which makes people feel uneasy and unsure of themselves and their relation to the world. We offer four possible explanations.

First: the spaces called for by the patterns dealing with social and psychological needs are critical. If the spaces are not right, the needs are not met and problems are not solved. Since these spaces are so critical, it stands to reason that they must be felt as real spaces, not flimsily or haphazardly partitioned spaces, which

943

CONSTRUCTION

only pay lip-service to the needs people experience. For instance, if an entrance room is created with flimsy partitions, it will not take hold; people won’t take it seriously. Only when the most solid elements of the building form the spaces will the spaces be fully felt and the needs which call for the space then fully be satisfied.

Second: a building will also seem alien unless it gives to its users a direct and intuitive sense of its structure—how it is put together. Buildings where the structure is hidden leave yet another gap in people’s understanding of the environment around them. We know this is important to children and suspect it must be important to adults too.

Third: when the social space has, as its own surrounding, the fabric of the load-bearing structure which supports that space, then the forces of gravity are integrated with the social forces, and one feels the resolution of all the forces which are acting in this one space. The experience of being in a place where the forces are resolved together at once is completely restful and whole. It is like sitting under an oak tree: things in nature resolve all the forces acting on them together: they are, in this sense, whole and balanced.

Fourth: it is a psychological fact that a space is defined by its corners. Just as four dots define a rectangle to your eye, so four posts (or more) define an imaginary space between them.

* %

Four foints make a rectangle.

This is the most fundamental way in which solids define space. Unless the actual solids which make up the building lie at the corners of its social spaces, they must, instead, be creating other virtual spaces at odds with the intended ones. The building will only be at rest psychologically if the corners of its rooms are clearly marked and coincide, at least in the majority of cases, with its most solid elements.

205 STRUCTURE FOLLOWS SOCIAL SPACES There fore:

A first principle of construction: on no account allow the engineering to dictate the building’s form. Place the load bearing elements—the columns and the walls and floors—according to the social spaces of the building; never modify the social spaces to conform to the engineering structure of the building.

You will be able to guarantee that structure follows social spaces by placing columns at the corner of every social space— columns at the corners (212); and by building a distinct and separate vault over each room and social space—floorceiling vaults (219).

For the principles of structure which will make it possible to build your building according to this pattern, begin with efficient structure (206) ; for the class of compatible materials, see good materials (207) ; for the fundamentals of the process of construction, see gradual stiffening (208). . . .

industrial giants. (See Raymond Vernon, Metrofolis 1985, Chapter 7: External Economics.)

To understand these facts, we must first realize that the city itself is a vast centralized workspace and that all the benefits of this centralization are potentially available to every work group that is a part of the city’s vast work community. In effect, the urban region as a whole acts to produce economies of scale by bringing thousands of work groups within range of each other. If this kind of “centralization” is properly developed, it can support an endless number of combinations between small, scattered workgroups; and it can lend great flexibility to the modes of production. “Once we understand that modern industry does not necessarily bring with it financial and physical concentration, the growth of smaller centers and a more widespread distribution of genuine benefits of technology will, I think, take place” (Lewis Mumford, Sticks and Slones, New York, 1924, p. 216).

Remember that even such projects as complicated and seemingly centralized as the building of a bridge or a moon rocket, can be organized this way. Contracting and subcontracting procedures make it possible to produce complicated industrial goods and services by combining the efforts of hundreds of small firms. The Apollo project drew together more than 30,000 independent firms to produce the complicated spaceships to the moon.

Furthermore, there is evidence that the agencies which set up such multiple contracts look for small, semi-autonomous firms. They know instinctively that the smaller, more self-governing the group, the better the product and the service (Small Sellers and Large Buyers in American Industry, Business Research Center, College of Business Administration, Syracuse University, New York, 1961).

Let us emphasize: we are not suggesting that the decentralization of work should take precedence over a sophisticated technology. We believe that the two are compatible: it is possible to fuse the human requirements for interesting and creative work with the exquisite technology of modern times. It is possible to make television sets, xerox machines and IBM typewriters, automobiles, stereo sets and washing machines under human working conditions. We mention in particular the xerox and IBM typewriters because they have played a vital role for us, the authors of

946

. . . this pattern complements the pattern structure follows social spaces (205). Where that pattern defines the relationship between the social spaces and the structure, this pattern lays down the kind of structure which is dictated by pure engineering. As you will see, it is compatible with structure follows social spaces, and will help to create it.

V

Some buildings have column and beam structures; others have load-bearing walls with slab floors; others are vaulted structures, or domes, or tents. But which of these, or what mixture of them, is actually the most efficient? What is the best way to distribute materials throughout a building, so as to enclose the space, strongly and well, with the least amount of material?

Engineers usually say that there is no answer to this question. According to current engineering practice it is first necessary to make an arbitrary choice among the basic possible systems—and only then possible to use theory and calculation to fix the size of members within the chosen system. But, the basic choice itself— at least according to prevailing dogma—cannot be made by theory.

To anyone with an enquiring mind, this seems quite unlikely. That such a fundamental choice, as the choice between column and beams systems and load-bearing wall systems and vaulted systems, should lie purely in the realm of whim—and that the possible myriad of mixed systems, which lie between these archetypes, cannot even be considered—all this has more to do with the status of available theory than with any fundamental insight.

Indeed, as we shall now try to show, the archetypal, best solution to the problem of efficient structure in a building is one which does lie in between the three most famous archetypes. It is a system of load-bearing walls, supported at frequent intervals by thickened stiffeners like columns, and floored and roofed by a system of vaults.

We shall derive the character of the most efficient structure in

three steps. First, we shall define the three-dimensional character of a typical system of rooms and spaces in a building. We shall then define an efficient structure as the smallest cheapest amount of stable material, placed only in the interstices between the rooms, which can support itself and the loads which the rooms generate. Finally, we shall obtain the details of an efficient structure. For a similar discussion, see Christopher Alexander, “An attempt to derive the nature of a human building system from first principles,” in Edward Allen, The Responsive House, M.I.T. Press, 1974.

I. The three-dimensional character of a typical building based purely on the social spaces and the character of rooms.

In order to obtain this from fundamental considerations, let us first review the typical shape of rooms—see the shape of indoor space (191)—and then go on to derive the most efficient structure for a building made up of these kinds of rooms:

1. The boundary of any space, seen in plan, is formed by segments which are essentially straight lines—though they need not be perfectly straight.

2. The ceiling heights of spaces vary according to their social functions. Roughly speaking, the ceiling heights vary with floor areas—large spaces have higher ceilings, small ones lower—

CEILING HEIGHT VARIETY (190).

3. The edges of the space are essentially vertical up to head height—that is, about 6 feet. Above head height, the boundaries of the space may come in toward the space. The upper corners between wall and ceiling of a normal room serve no function, and it is therefore not useful to consider them as an essential part of the space.

4. Each space has a horizontal floor.

5. A building then is a packing of polygonal spaces in which each polygon has a beehive cross section, and a height which varies according to its size.

If we follow the principle of structure follows social spaces (205), we may assume that this three-dimensional array of spaces must remain intact, and not be interrupted by structural

elements. This means that an efficient structure must be one of the arrangements of material which occupies only the interstices between the spaces.

We may visualize the crudest of these possible structures by means of a simple imaginary process. Make a lump of wax for each of the spaces which appears in the building, and construct a three-dimensional array of these lumps of wax, leaving gaps between all adjacent lumps. Now, take a generalized “structure fluid,” and pour it all over this arrangements of lumps, so that it completely covers the whole thing, and fills all the gaps. Let this fluid harden. Now dissolve out the wax lumps that represent spaces. The stuff which remains is the most generalized building structure.

II. The most efficient structure for a given system of spaces.

Obviously, the imaginary structure made from the structure fluid is not real. And besides, it is rather inefficient: it would, if actually carried out, use a great deal of material. We must now ask how to make a structure, similar to this imaginary one, but one which uses the smallest amount of material. As we shall see, this most efficient structure will be a compression structure, in which bending and tension are reduced to a minimum and a continuous structure, in which all members are rigidly connected in such a way that each member carries at least some part of the stresses caused by any pattern of loading.

I. A compression structure. In an efficient structure, we want every ounce of material to be working to its capacity. In more precise terms, we want the stress distributed throughout the materials in such a way that every cubic inch is stressed to the same degree. This is not happening, for example, in a simple wooden beam. The material is most stressed at the top and bottom

CONSTRUCTION

of the beam; the middle of the beam has only very low stresses, because there is too much material there relative to the stress distribution.

As a general rule, we may say that members which are in bending always have uneven stress distributions and that we can therefore only distribute stresses evenly throughout the materials if the structure is entirely free of bending. In short, then, a perfectly efficient structure must be free of bending.

There are two possible structures which avoid bending altogether: pure tension structures and pure compression structures. Although pure tension structures are theoretically interesting and suitable for occasional special purposes, the considerations described in good materials (207) rule them out overwhelmingly on the grounds that tension materials are hard to obtain, and expensive, while almost all materials can resist compression. Note especially that wood and steel, the two principle tension materials in buildings, are both scarce, and can—on ecological grounds—no longer be used in bulk—again, see good MATERIALS (2O7) .

2. A continuous structure. In an efficient structure, it is not only true that individual elements have even stress distributions in them when they are loaded. It is also true that the structure acts as a whole.

Consider, for example, the case of a basket. The individual strands of the basket are weak. By itself no one strand can resist much load. But the basket is so cunningly made, that all the strands work together to resist even the smallest load. If you press on one part of the basket with your finger, all the strands in the basket—even those in the part furthest from your finger—work together to resist the load. And of course, since the whole structure works as one, to resist the load, no one part has, individually, to be very strong.

This principle is particularly important in a structure like a building, which faces a vast range of different loading conditions. At one minute, the wind is blowing very strong in one direction; at another moment an earthquake shakes the building; in later years, uneven settlement redistributes dead loads because some foundations sink lower than others; and, of course, throughout its life the people and furniture in the building are moving

all the time. If each element is to be strong enough, by itself, to resist the maximum load it can be subjected to, it will have to be enormous.

But when the building is continuous, like a basket, so that each part of the building helps to carry the smallest load, then, of course, the unpredictable nature of the loads creates no difficulties at all. Members can be quite small, because no matter what the loads are, the continuity of the building will distribute them among the members as a whole, and the building will act as a whole against them.

The continuity of a building depends on its connections: actual continuity of material and shape. It is very hard, almost impossible, to make continuous connections between different materials, which transfer load as efficiently as a continuous material; and it is therefore essential that the building be made of one material, which is actually continuous from member to member. And the shape of the connections between elements is vital too. Right angles tend to create discontinuities: forces can be distributed throughout the building only if there are diagonal fillets wherever walls meet ceilings, walls meet walls, and columns meet beams.

111. The details oj an efficient structure.

If we assume now that an efficient building will be both compressive and continuous, we can obtain the main morphological features of its structure by direct inference.

i. Its ceilings, floors, and rooms must all be vaulted. This follows directly. The dome or vault shape is the only shape which works in pure compression. Floors and roofs can only be continuous with walls, if they curve downward at their edges. And the shape of social spaces also invites it directly—since the triangle of space between the wall and ceiling serves no useful purpose, it is a natural place for structural material.

951

CONSTRUCTION

3. Walls must all be. load-bearing. Any non load-bearing partition evidently contradicts the principle of continuity which says that every particle of the building is helping to resist loads. Furthermore, columns with non load-bearing partitions between them need shear support. The wall provides it naturally; and the continuity of the walls, floor, and ceiling can only be created by the action of a wall that ties them together.

3. Walls must be stiffened at intervals along their length by columnar ribs. If a wall is to contain a given amount of material, then the wall acts most efficiently when its material is redistributed, nonhomogeneously, to form vertical ribs. This wall is most efficient in resisting buckling—indeed, at most thicknesses this kind of stiffening is actually required to let the wall act at its full compressive capacity—see final column distribution (213). And it helps to resist horizontal loads, because the stiffeners act as beams against the horizontal forces.

4. Connections between walls and floors, and between walls and walls, must all be thickened by extra material that forms a fillet along the seam. Connections are the weakest points for continuity, and right-angled connections are the worst. However, we know from the shape of indoor space (191) that we cannot avoid rough right angles where walls meet walls; and of course, there must be rough right angles where walls meet floors. To counteract the effect of the right angle, it is necessary to “fill” the angle with material. This principle is discussed under column CONNECTIONS (227).

Therefore:

Conceive the building as a building made from one continuous body of compressive material. In its geometry, conceive it as a three-dimensional system of individually vaulted spaces, most of them roughly rectangular; with thin load-bearing walls, each stiffened by columns at intervals along its length, thickened where walls meet walls and where walls meet vaults and stiffened around the openings.

continuity of material

*£«

The layout of the inner vaults is given in floor and ceiling layout (210) and floor-ceiling vaults (219); the layout of the outer vaults which form the roof is given in roof layout (209) and roof vaults (220). The layout of the stiffeners which make the walls is given in final column distribution (213); the layout of the thickening where walls meet walls is given by columns at the corners (212); the thickening where walls meet vaults is given by perimeter beams (217)-, the construction of the columns and the walls is given by box columns (216) and wall membranes (218) ; the thickening of doors and window frames is given by frames as thickened edges (225); and the non-right-angled connection between columns and beams by column connection (227). . . .

954

955