Utilization of structural wood in housing

J. WESTOBY

Chief, Forest Economics Branch, Forestry Division, FAO

This article is Chapter 4 of a study entitled Trends in Utilization of Wood and its Products in Housing undertaken by the joint secretariat of FAO and the Economic Commission for Europe (ECE). The study has a limited aim and has been undertaken at the request of the ECE Timber Committee, in the hope that a review of this kind might throw light on future trends in sawnwood consumption in Europe. Since it deals with only one sector, new housing, which accounts for only one fourth of all sawnwood consumed in Europe, it can afford only certain indications of that trend. A comprehensive assessment of prospects for sawnwood would require the preparation of similar studies covering the other principal sectors of consumption. It is intended to prepare parallel studies of this kind, starting with packaging, as and when resources permit.

The wooden house or wood-walled house is not characteristic of European housing. It is highly localized, and accounts for but a small fraction of the European housing stock and for an even smaller fraction of the annual additions to that stock. The overwhelming proportion of sawnwood consumed in European housing enters into one- and two-family houses and in apartment blocks built of bricks, stone or concrete; it is consumed either in the house itself - as a building element (in the roof structure, joists and beams covering the basement and supporting the floors and ceiling, and various framework) or in joinery (flooring, doors, windows, stairs, cupboards, etc.) - or as part of the equipment used in the erection of the house (scaffolding, and molds for concrete constructions). Of the wood that becomes a permanent part of the house, that used for the building elements is required, generally speaking, in larger dimensions and with higher strength qualities than that used for joinery.

Reports from a number of countries point to the fact that it is as a structural element that the use of wood has declined most, and that this decline is closely bound up with changed methods of construction. On this point some informative figures are available for various regions of Austria (Table 1).

These figures show that, in Austria, wood consumed in joinery work per dwelling unit has fallen, as compared with before the war, by 10 percent to 15 percent whereas that used for the building elements has in general declined by 30 percent to 45 percent. The fall in the use of wood for building elements has been more marked in blocks of apartment houses than in one- and two-family houses (which in any case use, on the whole, more structural timber) and it has accelerated since the war.

Thus the impact on total structural wood consumption for housing in Austria (and elsewhere) has been greater than is suggested by the figures just quoted, since, from 1949 onwards, a greater proportion of the dwelling units provided have been in the form of blocks of apartments.

Trends similar to those recorded in Table 1 have also been noted for Greece. In Sweden, while structural wood has shown a marked decline, the amount consumed per dwelling unit in joinery work has actually increased, the rise being associated with an increase in the size of the average dwelling unit.

TABLE 1. - AUSTRIA CONSUMPTION OF SAWNWOOD PER DWELLING UNIT, 1936-1937 TO 1950-1964

			1936-1937	1946-1949	1950-1954	Percentage fall from 1936-1937 to 1960-1954
Blocks of apartment houses			in cubic meter
	Kärnten	A	6.7	6.4	3.6	46
		B	3.6	3.3	3.1	14
	Steiermark	A	10.5	9.5	6.1	42
		B	3.6	3.3	3.2	11
	Vorarlberg	A	6.9	6.3	4.2	39
		B	2.1	1.9	1.8	14
	Tyrol	A	9. 24	0.80	0.80	92
		B	2.55	2.55	2.55	-
	Oberösterreich	A	5.10	3.95	2.66	48
		B	2.50	2.20	1.60	40
Suburban one- and two-family houses
	Kärnten	A	9. 5	8.5	6. 8	28
		B	3.6	3.4	3.2	11
	Steiermark	A	10.5	9.5	7.3	30
		B	3.6	3.3	3.2	11
	Vorarlberg	A	7.0	6.2	4. 8	31
		B	2.2	2.0	1.9	14

A: for building elements
B: for joinery work

Piecing together the scattered indications available, we may hazard the guess ¹ that building elements today account 50 percent of all the sawnwood used in new residential construction in Europe, as against perhaps 60 percent before the war. If this guess is near the mark, it would mean that about 8 ½ million cubic meters of Europe's current sawnwood consumption are used structurally in new housing.

¹ This estimate is based on reports of the trends in various components in different countries, as well as on the data already cited. The following additional figures for Austria (relating this time to the whole country) may be of interest:

Proportion of Wood Consumed for Building Elements

	1937	1950	1964
	percent
Blocks of apartment houses	60	50	52
Suburban one- and two-family houses	66	69	62
Rural houses	74	71	68

The roof

Much of the structural wood goes into the framework of the roof, and it is here that important reductions in per-unit use have occurred. Yet reports suggest that the direct substitution for timber of other materials has been a relatively minor factor; where this has occurred it has been associated with changed housing patterns and changed methods of construction.

The lower consumption of roofing timber per dwelling unit may be ascribed to a number of factors:

(a) the increasing proportion of multi-family constructions, which automatically reduces the roofing per dwelling unit;

(b) the increased proportion of flat, as distinct from triangular roofs, and the adoption of lower-pitched roofs;

(c) the trend towards lighter roof structures; making use of smaller dimensions;

(d) the replacement of timber by other materials, notably reinforced and more latterly, pre-stressed concrete.

The first of these is discussed elsewhere in the study. The trend there noted seems certain to continue, and it will lead to a further reduction of the timber required on average per dwelling unit. Normally, multi-family construction is flat-roofed, though not invariably so. Some countries (e.g., France) report a reaction against flat concrete roofs in multifamily construction as defects appear in blocks built 10 to 20 years ago. However, the reaction is slight, and certainly only temporary, since concrete building methods have made great advances in the last two decades. In one- and two-family houses, flat roofs do not appear to be gaining ground in western and northern Europe. Indeed, they may be less common, among recent constructions, than in the immediate postwar years, when timber shortages in a number of countries prompted exceptional measures to reduce timber consumption. Of new residential constructions authorized in Belgium over a recent period, only 19 percent provided for flat roofs. In many countries the pitched roof is still preferred, both on aesthetic grounds and because of the additional space afforded, and on the whole there seems to have been a reversion to this traditional pattern for the one-and two-family house in countries where controls established in the early postwar years have been relaxed. The shift towards lower-pitched roofs, however, although in the first instance deliberately undertaken by architects to meet specific quantitative restrictions on the use of timber, seems to be permanent.

The utilization of smaller dimensions has undoubtedly been a factor of great significance; it reflects the application of engineering principles to the rational use of wood which started before the war and was given a great impetus in the postwar years both by specific timber shortages and by the rise in the price of wood in relation to those of other materials. This feature of postwar development which is not, of course, confined to the use of wood in housing, is discussed elsewhere in this report. Here we may content ourselves with noting that by assessing carefully the strength properties which will be required of each wood element, it has been possible to have recourse to shorter lengths, smaller - often different - cross-sections and wider spacing, while yet preserving adequate safety margins. Further progress certainly will be made in this direction, though the difficulties in applying engineering principles to wood - which is a natural product and therefore subject to considerable variations - are many. So far as roof structures are concerned, the technician's task has been facilitated by the adoption of new and lighter insulation materials. Significant wood savings, as well as labor savings, have also been realized by the adoption of nailed roof trusses, nailing replacing this carpenter's joint.

This trend towards lighter framing may be expected to continue. In many western European countries, the typical one- or two-family house of brick or artificial stone still retains its double-pitched, timber-framed roof. Undoubtedly, the failure of steel and prestressed concrete to make considerable inroads in smaller houses has been due to the fact that architects and builders have found the way to considerable wood economies. The emphasis on lighter framing in the Scandinavian countries has already been mentioned. In the United Kingdom, roof trusses developed by the Timber Development Association are reported to have won some acceptance. In Belgium, it is reported that a new design developed by the study and research service of the Société nationale des habitations et logements a bon marché requires only 2 cubic meters, as against 4 cubic meters before the war and less than 3 cubic meters in 1954. ² It is estimated that in France the amount of roofing timber required per square meter of area covered fell from 0.10 cubic meter in 1935 to 0.05 cubic meter in 1950 and 0.03 cubic meter in 1955; a further decline to 0.025 cubic meter by 1960 is envisaged. In Greece, it is estimated that the per-dwelling consumption of wood in roofs declined from 3.6 cubic meters in 1938 to 1.6 cubic meters in 1955, a reduction of 2.0 cubic meters, or 56 percent. Lighter wood structures not only reduce cost, but emphasize one of the special advantages of timber in relation to its competitors. Steel is heavier, and in addition requires more maintenance. Prestressed concrete is much heavier, requiring stronger walls and thus adding to total costs.

² Thus roof timber accounted for au important part of the total decline in unit consumption in Belgium, from 9.5 cubic meters in 1938 to 6 cubic meters in 1964.

Thus, in the traditional one- and two-family house in western Europe, timber's place in the framework seems assured. Consumption per dwelling unit in several countries has probably reached its minimum, at somewhere between 2 and 3 cubic meters, since there is no indication that lighter roof-covering materials will displace the traditional clay tile. Certain countries, however, still consume more than this minimum; a German report, for example, sets the roofing timber required per unit at from 2.3 to 4.8 cubic meters, according to the design of the framework and the roof covering. This suggests that there is still scope for a reduction in timber consumption. In northern Europe, wood is still used more lavishly, in both framework and roof covering, but a steady shift towards timber-saving designs will undoubtedly bring about further reductions in wood use. Information available for Sweden indicates a per-dwelling reduction in roof and attic joists from 6.6 cubic meters in 1938 to 3.2 cubic meters in 1965, or a decline of 52 percent. During 1955 it is estimated that a small Swedish wood or brick or stone house required 8.4 cubic meters and 9.0 cubic meters of wood respectively for roof and attic joists, while similar apartment dwellings utilized 3.8 cubic meters and 1.1 cubic meters. In Sweden, a small brick or stone house has on the average used more wood for roof elements than a wooden house.

But while timber maintains, and is likely to continue to maintain its position as roof structural material in the traditional one- or two-family house in western Europe, thanks largely to wood savings which have been effected, in southern and eastern Europe, the typical small urban house is already built in artificial stone and is usually flat-roofed. In Poland, for example, the roofs are generally prefabricated in concrete. Moreover, in these areas one- and two-family houses represent an ever-diminishing proportion of urban residential buildings. In multi-family blocks, throughout Europe, the quantity of timber used for roofing is very small.

Finally, if we look at the situation for Europe as a whole, we note a tendency for housing programs in western Europe to stabilize, while those in eastern Europe, where housing activity has admittedly lagged during the last decade, are expanding. This development, in association with the trends which have previously been mentioned, will tend progressively to reduce European consumption of roofing timber per dwelling.

How much timber is currently consumed in roofs for new residential construction in Europe? The total volume may be regarded as a function of the total housing program, the proportion provided in multi-family blocks, and the unit timber consumption per roof. In round figures, we may regard the European housing program as consisting of, say, 1.1 million one-family houses and 1.1 million in multifamily blocks, the latter corresponding (assuming an average of 4 stories) to 0.3 million roof structures. As to the proportion of each timber-roofed, some French estimates are available which afford a partial clue According to these estimates, the percentage timber-roofed is said to have developed as follows in Table 2.

These figures reflect timber's relatively firm hold in one-family dwellings and the-renewed progress which concrete is expected to make, following earlier setbacks, in multifamily construction. Reports suggest, however, that for multi-family construction in Europe as a whole the proportion is nearer one half. As for one-family building, the proportion is undoubtedly as high or higher in other western European countries and in Scandinavia, but lower in southern and eastern Europe. Having regard to the size of the individual housing programs, 85 percent may not he far off the mark. This suggests that fractionally over 1 million out of the 1.4 million roof structures in Europe today are of timber. Unit timber consumption, as we have seen, varies enormously: about 2 cubic meters in the United Kingdom; rather less in Greece; slightly more in France; 3 cubic meters in Belgium; 2.5 to 4.5 cubic meters in Germany; rather more in Scandinavia.

TABLE. 2. - FRANCE: ESTIMATED PERCENTAGE OF TIMBER ROOFED HOUSES, 1920 1960

Year	Multi-family	One-family
	percent
1920	95	98
1936	60	80
1950	70	85
1955	75	85
1960 (expected)	60	80

These (except for Germany) are averages for all dwelling units, including nontimber as well as timber. They suggest that it may be reasonable to assume a European average (all dwellings) of about 2.5 cubic meters, corresponding to average consumption in timber roofs only of 3.25 cubic meters. This leads to the conclusion that roughly 3.5 million cubic meters of timber are being consumed annually in the roofs of new dwellings in Europe - i.e., about 40 percent of all timber used structurally in new residential construction goes into roofs.

The interstory structure, ceilings and floors

The building element which supports the floor may be of wood and the actual floor surface of some other material, and vice versa. This suggests the advisability of separating a discussion of structural timber in floors from that of the flooring surface. The following paragraphs are primarily concerned with the building elements, but it will be necessary to make some reference to the flooring because, in some of the data available, a rigid distinction cannot be drawn. :

The function of the interstory element is to provide a support for the floor above and the ceiling below, and to give rigidity to the whole structure. It is, therefore, necessary to distinguish between ground floors and first floors in two-story structures, and between two-story and multi-story constructions. While the fate of wood in each has differed some what from country to country, broadly speaking it can be said that, for Europe as a whole, wood has been routed in the interstory elements of multi-story constructions, has lost considerable ground in ground floors and is meeting with varying success in resisting the encroachments of other materials in first floors.

As a ground-floor support, concrete has met with much more success in displacing wood than in roof structures. Indeed, in the days of intense wood short ages immediately after the war, there was active discouragement, and in some cases total prohibition, of wooden ground floors. For suspended floors, there was a switch to prefabricated concrete joists of various types, some times to composite steel and timber joists. But frequently the suspended floor gave way either to solid concrete floors cast on the site, or to prefabricated concrete floors. In the United Kingdom, for example, where controls for many years ruled out wood floors, the return to wood floors since the ending of building controls and the delicensing of softwood has not been particularly marked. Where the market has been recaptured by wood, it, has been either because site conditions enhanced the cost of concreting, rendering wood more economical, or, more usually, because the consumer preferred the wood floor for reasons of taste even though it might be more costly. One reason why concrete floors have proved more economical is that they eliminate the brick sleeper walls which formerly carried the plates and joists.

Elsewhere in the study it is noted that there were very great differences in timber consumption per dwelling unit among the various European countries, and that these differences by no means corresponded exactly with the availability of domestic timber supplies. Traditional usage is a powerful factor, and that may stem from accessible indigenous materials; but also it may relate to a period when imported supplies were plentiful and cheap. Thus while it is clear that the prevalence of all-wood houses is closely connected with local timber availabilities - such constructions remaining typical of rural areas in wood-surplus countries and surviving almost as curiosities in many wood-deficit countries - the fate of the wood floor in nonwooden one- and two-family houses in the various European countries on the whole reflects the general timber supply situation, modified in certain instances by the weight of tradition. In Italy, wood floors were supplanted many decades ago; in Scandinavia, the decline is relatively recent. In the United Kingdom, but for war and postwar shortages, the wooden ground floor might have survived, in spite of its falling competitive power. In fact, it was driven out under controls, and its limited comeback after decontrol has largely been due to traditional factors.

Thus, whereas timber ground floors were undoubtedly the most important in Europe for one- and two family houses 20 years ago, today they have been largely replaced by solid concrete floors. This trend has proceeded at different rates from country to country, but there is little likelihood that it will be permanently reversed anywhere. Unfortunately, representative statistics on ground-floor constructions are totally lacking, ³ and it is difficult to sum up the situation from the descriptive accounts which are available However, if we assume that the dwelling units annually provided today in Europe represent 1.4 million ground-floor constructions, the probability is that no more than half a million, and perhaps as few as a quarter of a million, involve structural wood.

³ Though it is reported that, for one- and two-family houses in Sweden, 80 percent of ground-floor joists were of wood in 1960; by 1955 the proportion had fallen to 32 percent.

As the interstory element in two-story, one-family houses, timber has resisted the challenge of concrete (reinforced to a greater or lesser extent, often combined with ceramic elements) much more successfully. The United Kingdom, for example, reports no evidence of any general substitution for timber in this end-use. The situation seems to be similar in Germany. In Belgium and in Sweden ⁴ the competition of concrete is making itself felt, while in France timber has lost ground rapidly in the last two decades, accounting today for only 30 percent of the ground-floor and interstory building elements in one-family houses: ⁵

TABLE 3. - FRANCE: DIMINISHING USE OF TIMBER IN HOUSING, 1920-1955

Year	Percentage in wood
	Multi-family constructions	One-family houses
1920	15	95
1936	2	85
1950	6	40
1955	8	30

⁴ The percentage of wooden first-floor joists in one- and two-family houses in Sweden fell from 77 in 1960 to 60 in 1955.

⁵ These figures relate to solivages i.e., joists and beams employed in both ground-floor and interstory structures. However low the survival of the wooden ground floor, the figures imply a major incursion of concrete in the first-floor elements.

Lower cost, better insulation and greater structural rigidity are the advantages claimed for various forms of concrete construction. Yet it may be doubted whether these provide the whole answer; certainly they do not explain the disparity in the trends in different countries just noted. Indeed, it is by no means certain that, on cost alone, the advantage lies with concrete.

In 1953, M. A. Martinoff carried out a study on the cost per square meter of realizing the interstory construction in various ways - e.g., the bearing element in timber, concrete or hollow bricks, the ceiling in lath and plaster, plasterboard, fibreboard or chipboard, the flooring in softwood boards or concrete, and the final floor covering in balatum, tiles, linoleum, parquet, etc. (Table 4).

TABLE 4. - BELGIUM: PRICE PER SQUARE METER OF DIFFERENT TYPES OF INTERSTORY CONSTRUCTION

(in Belgian francs: estimates based on prices ruling in the first quarter of 1953)

Type	Ceiling	Bearing element	Floor surface	Floor covering	Total
Joists in native pine 2 ½ x 7 in.		156
Covered with plaster boards	65
Flooring boards in ¾ in. northern red pine			119		340
The same, but with balatum covering				48	388
The same, but with joists in. northern red pine	65	173	119	48	405
Concrete or hollow bricks faced with concrete	34	199
Surfaced with lightweight concrete			62
Covered with balatum				48	343
The same, but covered with:
2-3 mm. vinyl	34	199	62	103	398
Linoleum on felt				188	485
Parquet				238	533
Concrete or hollow bricks faced with concrete	34	199
With floor boards in ¾ in. northern red pine			140
with balatum				48	421
The same, but with:
Cement squares	34	199	144		377
Granite squares			184		417
Ceramic squares			245		478

Thus, at the prices then ruling, the traditional interstory construction with wood flooring did not compare unfavorably, from the cost standpoint, with most of the alternatives. Nor had this competitive advantage disappeared in Belgium in 1955. Comparative costs per square meter adduced by an architect consulted at the end of 1955 were as follows in Table 5.

These figures suggest that the margin of cost advantage may have narrowed, but that it had not disappeared at the end of 1955.

In the United Kingdom, too, the traditional softwood structure remains competitive. An official publication, Houses that Save Softwood (H.M.S.O., 1953), gives some interesting data drawn from demonstration houses built on four sites to illustrate the effect of using materials other than softwood. The softwood savings and extra cost on a standard three-bedroomed house as compared with the traditional softwood joists and softwood flooring are shown in Table 6.

TABLE: 5. - BELGIUM: COMPARATIVE COSTS

Material		Belgian francs per m²
Wood
	Three 2 ½ in. x 7 in. joists	144
	¾ in. floorboards	115
	Ceiling work on 16-mm. laths	70 329
Hollow clay blocks
	Reinforced hollow blocks	230
	Ceiling work	30
	Covering in 35-mm. cork	78 338
Reinforced concrete
	9-cm. concrete at B fr. 2,800 per cubic meter	262
	Plaster ceiling work	30
	Covering in 36-mm. cork	76 358

Here, too, it is unlikely that the cost advantage of the traditional softwood floor has seriously diminished since 1953. These figures make it easy to understand how timber has retained its place for the interstory element of traditional one-family house-building in the United Kingdom. One would expect the cost comparison to be about the same in wood-deficit western Europe, more advantageous in Scandinavia and less advantageous in the wood-deficit countries of southern and eastern Europe. In Italy, for example, timber floors and ceilings, like timber roofs, are virtually unknown in the southern half of the country, in the plains and in urban areas generally. Even in the mountainous and rural areas, where wood is the traditional material, cost is said to be the main factor bringing about a steady recourse to concrete.

TABLE 6. - UNITED KINGDOM: SOFTWOOD SAVINGS AND EXTRA COST INVOLVED

Type	Softwood saved		Extra cost £
	Standards	m²
Prestressed concrete joists and particle-board flooring	0.394	1.84	34
Prestressed planks and hollow clay blocks with thermo-plastic tiles	0.443	2.08	39

Timber's greater success in retaining its place in first floor construction than in ground floors does not rest on direct cost alone. Concrete constructions are very much heavier for a given load (Table 7).

Thus, a square meter of this construction weighs (taking the 4-meter length as an example), 175 kilograms. A timber construction giving the same useful load would require, per square meter, 3 meters of 2 ½ inches x 7 inches joists and 1 square meter of 4 feet 4 inches flooring - a total of 0.061 cubic meter of softwood, weighing about. 30 kilograms. The weight of a timber construction therefore works out at about one fifth of that of its concrete rival; hence the higher weight of concrete is a fact which has to be taken into account in planning the foundations.

TABLE: 7. - BELGIUM DATA FOR REINFORCED HOLLOW CLAY BLOCK CONSTRUCTIONS

Length of slab (m.)	Thickness (cm)	Weight (kg/m²)
	Useful load of 150 kg. per m²
2.63-3.12	7	90
3.13-3.50	9	145
3.51-3.87	9	145
3.88-4.37	9	146
4.38-4.76	11	175
	Useful load of 250-300 kg. Per m²
2.38-3.60	9	145
3.51-4.50	11	175
4.51-5.75	13	195

In multi-story constructions (i.e., three or more stories) the interstory structure, like the ground floor, is today almost universally in some form of concrete. Steel - i.e., profile steel, as distinct from reinforcing steel - has never been very important save in one or two countries and in a few multi-story blocks. It had already lost much ground before the war and has continued to do so since. Wood is used on a very limited scale. Architects seem to agree generally that concrete elements give greater rigidity and stability to the whole structure. For this reason, and also because of fire risk, building regulations frequently preclude the use of wood structurally in multi-story buildings.

The current European housing program corresponds roughly to just under 2 million interstory structures (1.1 in one-family construction, 0.8 in multi-story blocks), of which perhaps half are realized in wood. Adding to these an estimate for timber ground floors, and reckoning 3 cubic meters on average for a ground floor or floor and ceiling unit, we can tentatively estimate European consumption in this end-use at 4 million -cubic meters.

Roofs, floors and ceilings together account for the vast bulk of timber consumed structurally in housing; the other end-uses are diverse and difficult to classify. Moreover, the general considerations which apply to the use of timber in roofs, floors and ceilings equally govern other structural housing timber, so that separate treatment is scarcely necessary. However, before attempting to sum up the prospects for timber in the building elements in housing, it is necessary to offer one or two general observations on the technical merits and demerits of wood as a structural material, and on price trends.

Timber's limitations as a structural material

How far has the decline in the use of wood as a structural material in housing been the result of the manifest technical superiority of its rivals? This springs to mind as a central question in the present context, yet it may be doubted whether, even if a complete and satisfactory answer could be made, it provides the key to the developments which have so far taken place and which are continuing.

The technical merits and demerits of wood as a structural material in relation to its rivals have been well summed up by M. Campredon ⁶ in Table 8.

⁶ See, for example, his article "Le matériau bois." in the Nov.-Dec. 1954 issue of Bois et forêts des tropiques.

A variety of measures can be taken to eliminate or reduce the technical drawbacks of each of these three materials. Wood is a natural material, and the preservation of its form depends on its moisture content in relation to its surroundings. Complaints of contortion have been encountered with increasing frequency since the war, largely because circumstances have compelled the use of wood which had been inadequately or incorrectly seasoned. In recent years, these complaints have diminished, but the prejudice they aroused has survived in the minds of those called upon to make the choice of materials. One factor which has focused attention on the propensity of wood to shrink and swell has been the increasing proportion of houses and flats built since the war and equipped with central heating, involving marked seasonal changes in the humidity of the atmosphere. This is a disadvantage which can easily be countered by appropriate safeguards; in northern Europe, structural wood has been used in centrally heated dwellings for many decades without ill effects. Similarly, though various kinds of wood treatment can reduce proneness to attack by insects and fungi, in Europe, these measures are seldom necessary. Most architects agree that the exercise of care in construction and necessary maintenance can render this risk negligible.

Wood's inflammability is a more serious matter. It not only leads to enhanced insurance rates, but in a number of countries has brought about absolute prohibitions on the use of wood for certain purposes in constructing residential premises in congested areas.

The following premium figures, which apply to Norway, are believed to be representative:

	Mills per annum
Frame house	1.3
Brick house with timber joists	0.75
Fire-resistant buildings with fireproof floors	0.3

The fire risk can be considerably reduced by the prior application of fire retardants to the timber, and this does not raise the cost significantly. Unfortunately, few timber traders are equipped to provide the necessary service, and few consumers are aware of the possibility of reducing fire risk by appropriate treatment. Nor, in general, is any incentive provided by insurance companies in the shape of reduced premiums. In any case, it is not the builder, anxious to keep his costs as low as possible, who is primarily concerned with the degree of fire risk and the cost of carrying that risk.

TABLE 8. - COMPARATIVE: TECHNICAL MERITS AND DEMERITS ON WOOD AS STRUCTURE MATERIAL

Wood	Steel	Concrete
Technical advantages
Low density	Low cross-sections, taking less space	Permits strong and decorative solutions
High resistance in relation to density, permitting, by new techniques, firm and decorative solutions	Light aspect	Long spans and heavy loads
Low maintenance costs	Factory preparation	Modern architectural forms
Easy to repair, join and strengthen	Rapid installation	Prefabrication
Easy to install, with cheap and simple tools		Prestressing
Resists fumes and chemical agents
Disadvantages
May distort	Heavier pieces. Higher weight of finished construction	Much heavier, calling for strengthened foundations
Subject to insect and fungus attack	Maintenance more difficult; mechanical means needed for moving and lifting	Setting time delays construction
Fire danger; higher insurance premiums	Special tools required for installation	Useless after serious fire
Not acceptable for party walls	Difficult to reconstruct after fires
Difficult to obtain in long lengths	Subject to corrosion unless regularly painted

As a natural material, wood not only has natural enemies: its suitability for structural purposes is subject to limitations by virtue of its heterogeneity and the lengths in which it is available. For centuries the craftsman has applied his arts to overcome these limitations, but only in recent decades - indeed only since the war in many European countries - has this problem been tackled in a scientific way and a determined effort made to place timber construction on a sound, up-to-date technological foundation. The timber engineer made his appearance in North America during the thirties. In Europe, although his advent is more recent, he has already many successes to his credit. One has already been mentioned - the reduction of cross-sections and the adoption of wider spacing. Of course wood has always been sorted and selected for particular uses in accordance with the strength properties required. But the selection has been on the basis of a superficial inspection, not always a sure guide to inherent strength properties, and the known variability of wood has made for tolerances many times those required for factory-made structural materials. Experiments carried out by Professor Siimes in Helsinki have shown that one deal or batten may be 12 times as strong as another of the same size and species. It is thus clear that very substantial wood savings could be made if it were possible to carry out strain tests generally, selecting for given use the pieces conforming to predetermined strength requirements, with an adequate safety margin. This solution, however, is not practicable. It is possible, however, to grade sawnwood before using it in accordance with its strength properties, and thus effect considerable savings. But stress grading raises a number of problems.

The first is that of grading with adequate precision on the basis of external characteristics. Studies carried out in Finland, Sweden, and the United States and elsewhere have shown that this problem is soluble, the quality grading being based on the following properties: knots, wane, slope of grain, density, rate of growth, shakes, bow, spring, compression wood, blue stain, rot, ingrown bark, worm holes, top defects, pitch pocket, cross grain, spiral grain, twist, concavity. Of these, density is the only property which cannot be readily determined on the basis of external signs. But even density can be measured with adequate precision by focusing attention on the spacing of the growth rings, though this is certainly the most difficult, part of stress grading. Maximum limits for all these defects can be established for various strength classes. Stress grading, however, requires a more thorough training than current commercial, or appearance, grading.

Another problem arises from the fact that in many applications of sawnwood the external appearance is at least as important as, and sometimes more important than the inherent strength properties. Current commercial grading lays emphasis on appearance, with such indirect indication as appearance may give of strength properties. Already the sawnwood trade is complicated by the proliferation of grades and sizes, and the parallel existence of two separate grading systems would lead to further confusion. This problem would be soluble if there were a close correlation between quality and appearance grading, but researches to date suggest that, while a correlation undoubtedly exists, it is too loose to allow of a reconciliation of the two systems. Moreover, it has to be borne in mind that the use of wood structurally, where quality counts, has for some time been declining in relation to its consumption for purposes where appearance is the major consideration.

The need for stress grading, however, centers on a limited number of sizes used in building components. This simplifies its introduction in domestic markets. In Finland, for example, it has been suggested that stress grading be introduced for use in seven size classes, mainly deals and battens, leaving appearance grading for timber used otherwise.

Its introduction in the export trade, however, is complicated by the fact that the standard structural sizes differ from country to country. It may be doubted whether, in the absence of a measure of international agreement on standards, stress grading can ever be practicable for internationally traded timber.

Finally, there is the problem of the attitude of producers, consumers and dealers towards stress grading, which are by no means in harmony. Consumers, in so far as they are aware of the possibilities afforded by stress grading, realize that it would effect wood savings and reduce costs. For both producers and dealers the adoption of stress grading could bring advantages in the long term, inasmuch as it would strengthen the competitive position of timber, though in the short run wood savings might mean lower sales. It would undoubtedly lead to greater price differentials between various qualities - a fact which militates against unanimity on the part of producers. Greatest resistance comes from the dealer, who is accustomed to buying appearance graded timber and effecting what is, in essence, a kind of stress grading in his own yard to meet the individual needs of his customers. Part of his profit stems from this service, and he is naturally loath to forgo it. This is why dealers have on many occasions, while paying tribute to the possibilities offered by stress grading, at the same time maintained that it should be carried out in the importing country. However, apart from the fact that this procedure would complicate the formation of international prices, it has to be borne in mind that the structure of the trade in most importing countries today, with a multiplicity of small firms, is such that the overwhelming majority of firms would be unable to find, train and utilize the necessary skills. There is no reason why the desired end, however, should not be reached through cooperative efforts on the part of smaller firms.

Finally, existing building codes have some bearing on this problem. Were these codes to be altered so as to presuppose stress grading, the adoption of stress grading would undoubtedly be accelerated.

Stress grading is the logical extension of the application of technological thinking to problems of wood construction. If it has made little progress to date, that is not to say that the ideas on which it is based have not been applied. Tolerances have been reduced, wood savings effected, and timber's competitive power strengthened - and this trend will undoubtedly continue however the debate on stress grading may ebb and flow.

But the technologist's contribution has not been limited to focusing attention on required strengths. In co-operation with the engineer, he has developed new, ways of using wood. New assemblies, often prefabricated, have been designed, giving greater rigidity for less weight. Clear roof spans of 8 to 10 meters am now practicable, and industrial trusses with spans of up to 30 meters have been realized. Research on timber connectors and nailed structures has yet to find its full application, while glued lamination offers infinite possibilities. It is true that many of the recent advances must be regarded as more relevant to other kinds of building than to residential construction. To realize their possibilities, and perhaps to find new applications in housing, it is necessary, first, that architects and consumers should be aware of the reduction in timber's limitations afforded by current research; secondly, that the architect should be able to enlist the professional advice of competent timber engineers in the same way as he can in solving problems of steel and concrete; and, finally, that timber used in novel forms should be competitive with its rivals - i.e., if it is more expensive, the extra cost should not outweigh the compensating advantages.

Both steel and concrete have certain technical disadvantages, but in both these materials determined efforts are being made to remedy them. In steel, the trend is towards lighter sections, and various means have been found for checking corrosion. Concrete has increased its competitive position within the last decade through the development of prestressed concrete. Though demanding higher-quality cement and high-tensile steel wire, it provides considerable economies in both cement and steel. It is waterproof and does not crack to the same extent as ordinary reinforced concrete. The lower weight reduces transport costs and promotes prefabrication. Besides the advantage in many cases of lower initial costs, it carries important savings in maintenance. The use of prestressed concrete has steadily expanded since the war, but its full impact in most countries is yet to be felt.

It is not easy to draw firm conclusions from this short review of wood and its rivals as structural materials in housing. The trend against wood has clearly been due largely to the change in the type of dwelling unit built and to changes in methods of construction. This trend has frequently been accelerated by physical shortages of timber; at the same time it has been restrained by the application of new techniques which have permitted wood savings and by what may roughly be described as traditional factors. It would seem, therefore, that if wood is to retain its role as a structural material, emphasis should be laid on those particular advantages which wood possesses from the standpoint of the consumer, and on speedy application of the findings of the wood technologist and timber engineer, notably by taking advantage of any opportunities afforded by prefabrication techniques.

The significance of this last point is underlined if one reflects on the circumstances under which methods of construction underwent rapid changes in the years immediately following the war. At that time, most European countries turned to new ways of building houses under the pressure of shortages of various materials. But an equally important factor, which operated against wood even in countries with adequate domestic resources, was the scarcity of skilled craftsmen and the need to keep labor costs down to the minimum. Hence methods of construction which permitted a higher degree of mechanization on the site, which permitted the transfer of labor costs from the site to the factory, through prefabrication, and which required either more readily available or more quickly acquired skills, were adopted, even where this meant an increase in crude material costs. Woodworking skills are not easily acquired, and over most of Europe the generation that should have been acquiring them was otherwise engaged for nearly a decade. The kind of constructions required during the war years called for different skills and different materials. This meant that in the postwar years, wood found itself under an additional disadvantage, which has survived to the present day. In western and northern Europe, the building industry is traditionally a small-firm industry, where the apprentice of yesterday is the skilled craftsman of today and the entrepreneur of tomorrow. Thus many of the builders in the postwar decade were men who had become accustomed to handling nontraditional materials in novel ways. The lost decade can never be recovered, but a wider understanding of wood's new possibilities might well check, though not reverse, the trend away from wood.

The influence of price

In the phases of residential construction discussed in the preceding pages, wherever timber has been displaced it has been replaced by concrete in one form or another, reinforced to a greater or lesser extent, frequently combined with ceramic elements such as hollow clay blocks. The competing materials are thus cement, reinforcing steel and clay ware; profile steel has nowhere been a serious rival.

The price of timber, in relation to these other materials, is very much higher than before the war, as shown by the examples in Table 9 for the United Kingdom.

TABLE: 9. - UNITED KINGDOM: BUILDING MATERIALS PRICE RATIOS (1938 = 1)

Year	Timber/cement	Timber/bricks	Timber/steel
1945	1.8	2.0	1.9
1946	1.8	1.7	1.8
1947	2.2	2.0	2.3
1948	2.3	2.2	2.1
1949	2.3	2.2	2.1
1950	2.2	2.0	2.0
1951	2.7	2.8	2.7
1952	2.8	2.8	2.4
1953	2.6	2.5	2.1
1954	2.6	2.4	2.1
1955	2.7	2.6	2.1
1956	2.6	2.6	2.0

Thus at the end of the war timber had doubled, or nearly doubled, in price as compared with alternative materials. In 1951, its competitive position suddenly worsened, and the improvement recorded since then, though marked in relation to steel, is negligible in relation to cement and clay components.

There is another aspect of price movements, besides the relative price trends, on which attention must be focused. If we examine the magnitude of year-to-year variations in the price of each material over the periods 1930-1938 and 1945-1955, we see that these fluctuations have been much more marked for timber than for other materials (Table 10).

TABLE 10. - UNITED KINGDOM: YEAR-TO-YEAR PERCENTAGE CHANGES IN THE PRICES OF BUILDING MATERIALS (1930-1938 AND 1945-1956)

Changes, percent	Timber	Cement	Steel	Bricks (postwar only)
Unchanged	1	1	3	1
Not more than 5%	8	14	9	8
5 %, but not more than 10 %	3	5	7	3
10 %, but not more than 20 %	5	1	1	1
Over 20 %	2		1

Thus a year-to-year change in price of over 10 percent occurred once only for cement, twice for steel, but no less than seven times for timber. In many respects, stability of price is no less important than the level of prices to those concerned with constructional work. Though timber prices in Europe have remained fairly steady since 1953, there is no doubt that recollections of the unhappy consequences of past violent fluctuations, and fears of future sharp changes, have created in the minds of those responsible for the choice of building material a prejudice which will be hard to dispel.

The price data cited in the foregoing Tables relate to the United Kingdom. Similar data could be quoted for nearly every European country. It is evident that the increase in timber's relative price has been responsible for some of the decline in timber consumption per dwelling unit, though it may be doubted whether this factor has been as potent as is commonly supposed. So far as the structural elements are concerned, the change in the size and type of dwelling unit provided would have brought about a considerable reduction in timber usage per unit independently of relative price considerations, though undoubtedly the higher price of timber, and specific timber shortages, have helped to hasten those changes. In particular end-uses, there are examples of timber being preferred in spite of higher cost; there are equally examples of shifts towards concrete even where the traditional structure appears to remain competitive.

The prospects for structural timber in housing

About 8.5 million cubic meters of sawnwood are being used annually as structural elements in new residential construction in Europe (excluding the U.S.S.R.). This is about half of all sawnwood used in housing, and represents a sharp fall from the estimated pro-war proportion of 60 percent. Of this 8.5 million cubic meters, about 4 million cubic meters go into floors, ceilings and interstory structures, slightly less into roofs, and the rest into a variety of miscellaneous uses. Thus it is in the building elements that the use of sawnwood has declined most.

The fall that has taken place in per unit use may be attributed to:

1. changes in the size and type of the average dwelling unit (including changed methods of construction and the shift towards multi-family blocks);

2. straightforward wood economies (smaller dimensions, improved design);

3. replacement of wood by other materials.

Though not statistically demonstrable, it is likely that the order in which these factors are cited reflects their relative importance. Price has, of course, entered into each, though to a varying extent.

Changes in the dwelling unit have come about partly as a result of social influences (demographic trends, urbanization, etc.) and partly as a result of changed methods of construction designed to economize in labor costs, to meet shortages of specific labor skills, or to make use of materials which were cheaper or more readily avail able than timber.

The effect of more rational use of wood, through better design and less extravagant tolerances, is most clearly seen in roofs, where timber retains a firm hold on the market in one-family houses in many countries, though its role in multi-construction has considerably diminished. It may be expected that most one-family house roofs will continue to require timber though inter-country comparisons suggest that in several countries there is still scope for further reductions in use per unit.

Wood is also used more rationally today in ground-floors and interstory constructions, and though in one-family houses wood has fared better in first than in ground floors, further losses in both are likely. In multi-construction, timber's share is likely to diminish further.

How will current trends, if they continue, affect the future demand for sawnwood for structural elements in new housing? The foregoing discussion will have made it abundantly clear that the statistical material available does not admit of an intelligent forecast.

Nevertheless, it may be useful to consider certain assumptions which appear to accord with current trends and to examine their possible impact on requirements of structural timber in new housing. Suppose, for example, that the total European housing program (excluding the U.S.S.R.) were to expand by about half a million dwelling units by 1970, and that the proportion provided in multifamily blocks slightly increased. Then the total number of roof, ground-floor and interstory units would increase as follows in Table 11 (assuming an average of four stories for multi-family blocks):

TABLE 11. - EUROPE: POSSIBLE: TRENDS IN HOUSING 1955 AND 1970

Number of units	1955			1970
	One- and two-family	Multi family	Total	One- and two-family	Multi-family	Total
	in millions
Dwelling units	1.13	1.05	2.18	1.30	1.40	2.70
Roof units.	1.13	0.26	1.39	1.30	0.35	1.65
Ground-floor units	1.13	0.26	1.39	1.30	0.35	1.65
Interstory units	1.13	0.79	1.92	1.30	1.05	2.36

Making certain assumptions as to the proportion of each unit constructed of timber (percentages shown in parentheses), we arrive at the following estimates of the number of timber roof, floor and interstory units respectively required in 1955 and 1970 (Table 12).

If we now assume that the average timber roof in 1956 required 3.25 cubic meters and the average timber ground-floor or interstory unit 3 cubic meters in 1955, ⁷ and that by 1970 these figures will have declined slightly to 3 cubic meters and 2.75 cubic meters, respectively, we arrive at the aggregates given in Table 13 for the consumption of structural timber in new housing.

⁷ These averages are, of course, higher than certain of those quoted earlier, which were averages for components of all dwelling units.

TABLE 12. - EUROPE: ESTIMATED NUMBERS OF TIMBER ROOF, FLOOR AND INTERSTORY UNITS, 1955 AND 1970

Unit	1955			1970
	One and two-family	Multi family	Total	One and two-family	Multi family	Total
	in millions
Roof	0.96 (85)	0.13 (50)	1.09	0.98 (75)	0.11 (30)	1.09
Ground-floor	0.34 (30)	0.03 (10)	0.37	0.26 (20)	0.02 (5)	0.28
Interstory	0.79 (70)	0.16 (20)	0.95	0.65 (50)	0.11 (10)	0.76

Though, as already stressed, little credence can be attached to the absolute figures set out in this statistical exercise, their implications are of some interest. They suggest that a continuation of present trends, in so far as we have been able to decipher them, will lead to a further decline in the per-unit consumption of sawnwood for structural purposes in new residential construction. If the assumptions made in the foregoing paragraphs are realized, this decline may be of the order of 30 percent over the next 16 years. Should this happen, then even a substantial expansion in the European housing program (by half a million units) would still mean a lower aggregate need for structural sawnwood. On the assumptions which have been made here, this decline would amount to 1.6 million cubic meters.

TABLE 13. - STRUCTURAL WOOD NEEDS IN NEW HOUSING, 1955 AND 1970, ON CERTAIN ASSUMPTIONS

Need	1955	1956
	in million cubic meters
Roofs	3.5	3.3
Ground-floors and interstories	4.0	2.9
Other structural timber	0.9	0.7
TOTAL	8.4	6.9
Per dwelling unit	3.8 m³	2.6 m³

THE SOCIETY OF AMERICAN FORESTERS AWARDS THE GIFFORD PINCHOT MEDAL

A unique event place in the course of the last Conference of FAO. Simultaneously with the annual dinner of the Society in Syracuse, N.Y., members of the Society of American Foresters present in Rome gave a luncheon party for all the forestry delegates attending the Conference. In the course of the luncheon, on behalf of the Society, Dr. R. A. McArdle and Dr. V. L. Harper, Chief and Assistant Chief respectively of the U.S. Forest Service, presented the Gifford Pinchot medal for outstanding services in the cause of forestry to Mr. Henry lilt Clepper, who has been Secretary of the Society for the past 30 years, Only four such awards have been made during the 50-year history of the Society. Our photograph shows (center) Mr. Clepper after having received the award. On hi, left is Mr. M. Leloup, Director of FAO. Forestry Division, on the right Dr. McArdle, extreme right and left two of the outstanding Honorary Members of the Society of American Foresters of whom there are 12 in all, Professor Eino Saari of Finland and Professor Aldo Pavari of Italy.