Woodwool Slabs – Manufacture, Properties and Use by Erik Johansson Contents

Acknowledgements 4 1 Introduction 4 Problem 4 Method 4 Organization of the report 4 2 General considerations 5 History of woodwool slabs 5 Production 5 Properties 6 Products and uses 6 3 Recommendations for production 7 Manufacturing process 7 Quality control 10 Plant 10 Equipment 10 Staff 11 Setting up the plant 12 Consumption of raw materials and energy 12 Production costs 12

4 Choice of wood 14 Technical requirements for the wood 14 Availability of the wood 14 Improving the compatibility of the wood with cement 14 Tests of suitability 14 5 Properties 15 Thermal properties 15 Strength 15 Acoustic properties 16 Fire performance 16 Moisture properties 17 Erik Johansson was born in Sweden in 1963. Durability 17 He holds a master’s degr ee in Civil Engi neer ing from Emissions of harmful compounds 17 Lund Univer sity. Since gradu a ti on he has done resear ch 6 Quality control 18 on build ing ma te ri als, and has par tic i pated in de vel op- Size and density 18 ment resear ch projec ts on building technique and build- Strength 18 ing ma te ri als in Tu ni sia, Al ge ria and Ethi o pia. He is engaged as a resear cher at Lund Centr e for Habi tat Other tests 18 Studies and at the Divi sion of Building Mate rial s, the 7 Applications 19 In sti tute of Tech nol ogy, Lund Uni ver sity. Exterior walls 19 His mas ter’s the sis dealt with wa ter leak age in flat Roofs and ceilings 22 roofs in Tuni sia, and he wrote a Building Issue on the Surface finishes 23 topic in 1989. He is the house expert on foot ball and spends all his free time out in the Swedish rain cheering References 24 on his home team Malmö FF. Appendix 26

3 Volume 6 • Number 3 Building Issues 1994

Acknowledgements 1 Introduction

I would like to acknowl edge some of the help I recei ved Problem during this study. Ther mal in su la tion is im por tant to im prove in door cli - First I would like to thank Mr Bengt Rääf (Produc ti on mate and save ener gy in a building. The use of therm al Man ager) and Mr Lennart Rääf of Tepro Byggmaterial, in su la tion ma te ri als are, how ever, of ten lim ited in de vel - Österbymo, Sweden, for their gener ous hospi tal ity on the oping countri es. There are several reasons, for exam ple: many occa sions they invit ed me, and my col leagues, to their factory and for the time they took to · The ad van tages of ther mal in su la tion ma te ri als are not discuss the produc ti on and use of woodwool slabs. known. In valu able in for ma tion on the pro duc tion and use of · Lo cal ther mal in su la tion ma te ri als are not avail able. woodwool slabs was provided by Elten Sys tems, · Im ported ther mal in su la tion ma te ri als are ex pen sive Barne-veld, the Nether lands , who also com mented on the and require foreign currency . draft report. I am very grateful to Mr John de Wit (Divi - · Use of ther mal in su la tion ma te ri als re quires modifi- sion Manager) and Mr Arnold Plak (Export Manager). cat ion in current building methods. My visit to Climatex Indústria de Madeir a Minerali-zada, Porto Alegre, Brazil, greatly in creased This reports deals with woodwool slabs, a well tested my knowl-edge about woodwool slabs and their use. I ther mal insu la tion mate rial that can be made lo cally in would like to thank Mr Carlos Roberto Simm (Oper a - most countri es. Becaus e of their versa ti lit y, woodwool tions Direc-tor), Mr Luiz Sergio Bocchese, Mr Werner slabs are easy to in te grate with most con struc tion tech - Dopheide, and Mr Romildo Feijó da Rosa for their warm niques. re cep tion, hos pi tal ity and will ing ness to pro vide in for- Many of the ma te ri als tra di tion ally used in con tem po - ma tion. rary buildings – bricks, concret e, stone, soil block, etc. – Last but not least, thanks to Ms Rosane Bauer, Archi- have poor therm al insu la ti on capac it y. To reach the same tect, who par tic i pated in the Ar chi tec ture & De vel op - ther mal insu la ti on as 100 mm of woodwool slab require s ment course at LCHS in 1993. She acted as guide and about 4 m natu ral stone, 1.8 m concret e, 700 mm soil transla tor during my visit to Porto Alegre. block or concret e hollow block and 500 mm hollow brick. Erik Johansson Method This report is prim aril y based on the expe ri ence gained from a re search co op er a tion be tween Lund Uni ver sity, Di vi sion of Build ing Ma te rials and LCHS, and the Nati onal Centre for Building Resea rch and Studies (CNERIB) in Alge ria during the period 1991–1993. This project , which include d test produc ti on of wood-wool slabs with Alge ria n woods and full scale tests of the slabs in building com ponents , laid the founda ti on for in- troduc ing woodwool slabs on the Al geria n market. Visits were made to mod ern, com pletely au to mated fac to ries in Swe den and the Neth er lands. A semi-auto - mated, la bour inten sive factory was studied in Brazil. Detai led studies were also made on the use of woodwool slabs in these countri es. The re port is also based on a study of the liter a ture , includ ing a large part of all the mate ria l on woodwool slabs written over the years.

Organization of the report The re port is divided into two parts. The first part in- cludes Chapter 1, the background to the study, Chapter 2, a short descri pti on of the produc ti on, proper ti es and use of woodwool slabs, and Chapter 3, a descri pti on of how woodwool slabs can be produced indus tri ally in devel op - ing countri es. The second part include s Chapter 4, choosing an ap - propri ate wood, Chapter 5, the proper ti es of woodwool slabs, Chapter 6, qual ity con trol, and Chapter 7 exam ples of use.

4 Building Issues 1994 Volume 6 • Number 3

2 General considerations

History of woodwool slabs Slabs of woodwool (excel sior), gypsum and water were patente d in Germ any already in 1880. During the 1910s produc ti on of woodwool slabs with magnesite as the binding agent (patente d 1908) started in Austri a. Magne- site1 gave better dura bil ity than gypsum . Portland cem ent was intro duced at the end of the 1920s, and is the most com mon binder today , which is why they are com monly called woodwool ce ment slabs Fig. 1 Making woodwool in a vertical wood shredding machine (Tepro, Sweden). Shredders sold today are horizontal. and, in North America, ce ment ex celsior boards (CEB). See Figs 9 and 10. For a long time woodwool slabs were made in Germ any with either gypsum , magnesite or ce ment as binder. Gypsum slabs are no longer made. duces the amount of sugar and other com pounds in the The technique to produce woodwool slabs – mainly wood that inhibi t setti ng of the slabs, and lowers the cem ent-bonded, but even magnesite-bonded – spread moisture content (shredding is more dif ficult with green quickly from Austri a and Germ any to other Euro pean wood). coun tries and North Amer ica. A great in crease in pro- To make woodwool, a half metre long log is placed in duction occurred dur ing the years before and after the a shredding machine (Fig. 1) fit ted with scoring knives sec ond world war, and woodwool slabs were spread even perpen dic u lar to the planing knives. The cross sec tion of far ther geo graph i cally. the woodwool is deter mined by adjust ing the speed with At first cem ent-bonded woodwool slabs were pro- which the log is fed toward the planing knife, and the duced by hand in small plants. The equipm ent was lim - distanc e betwee n the scoring knives. The thickness of the ited to wood shredding machine s, to make the wood- woodwool can vary betwee n 0.2 – 0.5 mm, and the width wool, and a mixer. Man u fac ture be came more and more betwee n 1.5 – 5 mm depend ing on how the slab will be mecha nize d over the years, with signif i cant ly high pro- used. The amount of woodwool in a slab varies betwee n duc tion ca pac ity. Mod ern fac to ries are nor mally fully about 75 – 200 kg de pending on the densit y of the slab. mecha nize d, and about 15 persons can produce up to 150 m3 slabs a day. Binder The most com monly used wood for woodwool slabs The most com monly used binder in woodwool slabs is com es from coni fers, mainly pines and firs. During the Portland cem ent, but magnesite can also be used. The 1960s a great num ber of other spe cies were tested, in- Portland cem ent is norm ally of ordi nary type (OPC), cluding tropi cal woods, to see if they could be used for although rapid-hardening cem ent can be used to make woodwool slabs. A num ber of spe cies were suit able, the setting faster. Sometimes white ce ment is used for which led to produc ti on of woodwool slabs on other con- aes thetic rea sons. ti nents. There is current ly produc ti on in Afric a (Ghana, For magnesite-bonded slabs the binder can be said to Malawi, Namibia, Zam bia), Asia (Burma, In dia, In do ne - be magne sium oxide (MgO), one of the com ponents of sia, Ja pan, Ma lay sia, the Phil ip pines, Tai wan, Thai land) mag ne sia ce ment. and Latin Amer ica (Brazil, Mex ico, Pan ama, Peru, Ven e - The amount of binder depends on the densit y of the zuela). Produc ti on in these countri es varies in the degree woodwool slab, and varies betwee n about 150 and of mech a ni za tion – ev ery thing from man ual to com - 400 kg/m3. pletely au to mated. Binder additives Normally some kind of binder addi ti ve is neces sary . The Production set ting of the Portland cem ent and water mixture can be The com ponents needed for woodwool slabs are wood- inhib it ed to greater or lesser extent by sugars and other wool, binder (Port land cem ent or magnesite) and wa ter. chem i cals in the wood wool. An ac cel er a tor is there fore Normally a small amount of binder addi ti ve is added to often added so that the slabs set within 24 hours. The speed up set ting. most com mon ac cel er a tor is cal cium chlo ride (CaCl2). The amount of accel era tor de pends on the kind of wood, Making woodwool but is usuall y about 2% by weight of the water . Woodwool can be made from a num ber of woods. To For magnesite-bonded slabs the binder addi ti ve is ei- make produc ti on eas ier, the wood should allow easy ther mag ne sium chlo ride (MgCl2) or magne sium sul- shredding (have a low densit y) and not contai n com - phate (MgSO4). pounds that se riously inhibi t the setti ng of the slab. Usually the tree trunks are air-dried (sea soned) be fore cutti ng into logs and shredding to woodwool. This re -

1 The binder is actually magnesia cement, produced by mixing magnesium oxide (MgO) with a solution of magnesium chloride (MgCl2) or sulphate (MgSO4). The magnesium oxide is produced by heating minerals such as magnesite (MgCO3) or dolomite (Ca(MgCO3)2).

5 Volume 6 • Number 3 Building Issues 1994

Density Thermal conductivity Material (kg/m3) (W/mK) Woodwool slab 400 0.085 Cellular plastic, mineral wool 20 – 50 0.036 Lightweight concrete 400 0.10 Softwood 500 0.14 Hollow brick 800 0.47 Table 1 Comparison of density and thermal conductivity (values in practice) for woodwool slabs and some other materials.

Other properties Woodwool slabs have very good acous tic proper ti es and are often used to absorb sound in, for exam ple, facto rie s, publi c gather ing places, sports and concert halls. Fig. 2 Mould filled to overflowing with a mixture of wood- The mate ria l is known for its good dura bil ity. It has wool, cement and water before compressing very good fire resis tance, toler ate s damp and is not at - (Tepro, Sweden). tacked by mould or rot. Woodwool slabs have good re- sis tance to in sect pests, as ter mites. Compared to many other in su la tion ma te ri als, wood- Water wool slabs have good bending and com pression strength. The wa ter should not contai n any thing that would They are easy to saw, drill and nail. They have good inhibi t the setti ng of the slabs. The amount of wa ter ad he sion to ren der ing/plas ter ing mor tars and con crete. require d is about 50% of the cem ent by weight. The mate rial is con sid ered to be healthy, since it has very Mixing low emissions of harm ful com pounds. Before the woodwool is mixed with the binder, it is soaked in a water bath contai ning the binder addi ti ve. Products and uses The wet woodwool is transferred to a mixer, where dry binder is added.2 Insulation slabs The ra tio by weight of binder to woodwool is about In su lation slabs are nor mally made of rel atively coarse 2:1. There are small vari a tions from this ra tio. Since the woodwool (3 – 5 mm wide). They are used for therm al binder norm ally is the most expen sive part of the wood- insu la ti on, and are norm ally not visi ble but rendered/ wool slab, at tempts are made to use as lit tle as pos si ble. plas tered. The slabs can be 2 – 3 m long, 500 – 900 mm Too littl e binder, however , means that not all the wood- wide and 15 – 150 mm thick. Their densit y ranges from 3 wool is coated, which result s in poorer binding and 250 – 700 kg/m depend ing on use. If therm al insu la ti on strength. ca pac ity is impo r tant, they are made with low den sity; if strength is im portant , they are made with high densit y. Making the slabs The mixture of woodwool, binder and wa ter is put into Acoustic slabs moulds which are filled with the require d amount of Woodwool slabs meant for sound absorp ti on are often mixture by weight (Fig. 2). The moulds are then stacked made of finer woodwool (1.5 – 3 mm). These slabs are on top of each other and put under pres sure so that the vis ible, and they are of ten painted for aes thetic rea sons. mixture in each mould is com pressed.3 They might also be made with white cem ent which gives After the slabs have hardened, usuall y in 24 hours, them a whitewood colour . Acousti c slabs are usuall y 15 they are demoulded and the edges trimmed with a saw. – 50 mm thick. They cure for two to three weeks be fore they are de liv - Special products ered. There are also a num ber of spe cial prod ucts. To in crease the loadbearing capac it y for use in roof struc tures, the Properties slabs might be rein forced with wooden poles or bars, or the sides of the slabs can be strengthened with gal va- Thermal insulation nized steel chan nels (Fig. 4). These re in forced slabs can Woodwool slabs give good therm al insu la ti on. Therm al even be used as standing, loadbearing wall ele m ents. con duc tiv ity is, how ever, rel a tive to their den sity (see In many countri es the slabs are sold with a finishe d 3 Fig. 23). At a densit y of 400 kg/m the therm al conduc - sur face. These might be a layer of ce ment-based mor tar tivit y is about 0.085 W/mK in practi ce. A com pari son of or gypsum plas ter. The surfac ing can be done so thinly the therm al conduc ti vit y of woodwool slabs and some that the tex ture of the slab still shows (Fig. 5). other mate ri als is shown in Table 1.

2 Less commonly the binder and water are mixed before the woodwool is added. A small-scale process using this method of mixing is described in Hawkes and Cox (1992). 3 Magnesite-bonded woodwool slabs can also be produced industrially in a continuous process during which the slabs are compressed and heated to 400°C to make them cure faster (see Kollmann 1955).

6 Building Issues 1994 Volume 6 • Number 3

3 Recommendations for production

The man u fac tur ing pro cess rec om mended is in prin ci ple the same as for more auto m ated produc ti on. This means the proposed plant can be gradu all y mecha nize d with more equipm ent without al ter ations in the plant layout. This descri pti on assum es the binder is Portland cem ent. Magnesite can also be used ac cord ing to Chap ter 2.

Manufacturing process Debarking and air-drying (seasoning) Imm edi ately after fell ing, or at least be fore stack ing in the wood yard, the bark is rem oved from the trunks with a de bark ing tool, such as a steel scraper. The tree trunks should then be air dried unti l the moisture content of the wood drops to the right level, usuall y about 20 – 30%. This norm ally takes three to six months and has to be done in open air. Making woodwool The trunks are cut into logs of about half a metre long Fig. 3 Example of an insulation slab (above) and moved to the wood shredding machine on carts. and an acoustic slab (below), life-size photographs. During shredding, the planing and scoring knives are ad- (Tepro, Sweden). justed to produce the desire d width and thickness of wood wool strands. Weighing and soaking woodwool Dry woodwool is taken to the scales in a wheel barrow and the right amount for a batch is weighed. It is then carrie d by hand and fed into an im mersion tank. The im - mersion tank is also filled with ac cel er a tor so lu tion from an at tached con tainer. (The amount and type of accel er a - tor in the solu ti on depends on the type of wood, and it can hap pen that no ac celer ator is needed). Before the wet woodwool leaves the im mersion tank, it passes be tween rub ber roll ers to re move the ex cess liquid. The woodwool is then carrie d by conveye r belt to the mixer. Adding the cement Fig. 4 Cross section of slabs reinforced with wood (above) Cem ent is deli vered to the plant in bags. The bags are and galvanized steel channels (below). emp tied into a con tainer and the ce ment is trans ferred to the mixer through screws. The correct amount of cem ent for a batch is control led by a ce ment dos ing unit next to the mixer. Mixing and moulding Wet woodwool and dry cem ent are mixed conti nu ously in a hor izon tal mixer. The hom oge nous mix is then spread in oiled moulds that are pushed into place under the mixer on a line of rollers. The amount of mate ria l in the moulds de pends on the densit y of slab to be pro- duced. The mixture must be spread evenly in the mould and pressed down along the edges. Note that gloves must be used to avoid skin contac t with the mixture , since ce - Fig. 5 Slab coated with gypsum plaster, resulting in a smooth ment is corro sive . (Skin contac t with ce ment might even surface that retains the texture of the slab lead to chrome aller gy.) (Climatex, Brazil).

7 Volume 6 • Number 3 Building Issues 1994

Fig. 6 Fig. 7 Fig. 8 Debarking the tree trunks before they are air dried. Cutting the trunks into De barked half a metre long logs. and cut logs stacked on carts.

Fig. 11 Fig. 12 Fig. 13 Woodwool Dry woodwool is soaked in an immersion Con tinu ous mixing of wet wood wool with ce ment, is weighed manually tank and transported to the mixer. and filling the moulds. The mixture must be spread before soaking. Before the wet woodwool reaches the conveyer belt, uniformly in the moulds and pressed down along it goes between rubber rollers to press out excess the edges. liquid.

Fig. 17 Fig. 18 Stripping the 24 hour old slabs. Stripping 24 hour old slabs (Climatex, Brazil).

8 Building Issues 1994 Volume 6 • Number 3

Fig. 9 Fig. 10 Producing woodwool in a horizontal Shredding woodwool. As the log is fed down, wood shredding machine. The logs are the planing and scoring knives go back and forth fastened and pressed down against the planing horizontally at high speed. and scoring knives by toothed rollers. The woodwool falls down under the shredder.

Fig. 14 Fig. 15 Fig. 16 Filling the cement container Compression The slabs set un der that is connected to the mixer of filled moulds pres sure for 24 hours. by cement screws. in a hydraulic press. The re quired pres sure is maintained by a con- crete slab weigh ing about one ton.

Fig. 19 Fig. 20 After a weeks curing, the slabs are sawn to trim the edges Storing of slabs stacked on top of each other. and ensure the correct length and width.

9 Volume 6 • Number 3 Building Issues 1994

Compressing and setting fices, etc. Some of the machines require cast con crete The filled moulds are stacked on top of each other and foun da tions. The plant should have san i ta tion fa cil i ties com pressed in a hydrau li c press. The bottom of one and fire equipm ent. mould becom es a lid on the mould under . Then the enti re The plant yard should be large enough to allow stack of filled moulds is moved and allowed to set for 24 air-drying of tree trunks. hours. Pressure is maintai ned by, for exam ple, plac ing a concret e slab (weighing about one ton) on top of the Equipment stack. The equipm ent needed for the suggest ed produc ti on pro- Stripping cess is shown in Tables 2 – 4. Some of it must be im - When the slabs have set they are stripped from the ported, and some should be avail able local ly. The prices moulds, and the moulds are cleaned and oiled for reuse. given are es ti mates for new im ported equipm ent. The Curing and trimming the edges placem ent of the equipm ent in the plant is shown in The stripped slabs must be cured indoors or under cover, Fig. 21. protec ted from direct sunli ght. The best cur ing occurs if the surround ing air is somewhat moist, so the slabs are Note that the equipm ent to make woodwool slabs is not al lowed to dry out. It is recom mended that each stack very robust, and it is recom mended that used equip - of boards be covered with a plasti c sheet dur ing the first ment be bought if available . week. To avoid high tem pera ture s due to hydration, spac - ers could be put betwee n the boards. Item Number Code Cost (US$) When the slabs are suffi ciently hard ened, say af ter Wood shredding machines 2 2 130,000 one week, they can be sawn along both long and short Immersion tank 1 5 56,000 sides to the cor rect size. Cement bag tipping unit 1 7 14,000 Af ter a to tal of two to three weeks cur ing the slabs are Cement screws 2 8 19,000 ready for deli very . Cement dosing unit 1 9 16,000 Continuous mixer 1 10 54,000 Distribution station (mould supplier, Quality control spreading machine, stripping table) 1 11 32,000 Qualit y is mon i tored by in spect ing ran domly se lected Switch box 1 12 28,000 sam ples and should be routi ne at the plant (propri eta ry Hydraulic press 1 14 54,000 inspec tion) and, if possi ble, at an accred it ed testing cen - Grinding machine for wood shredding machine planing knives 1 18 20,000 tre. Chapter 6 gives recom menda ti ons on how to de ter - Set of spare parts and tools 1 – 27,000 mine the qual ity of woodwool slabs. Table 2 Equipment to be imported. Approximate cost 1 Dec Proprietary inspection 1994 according to Elten Systems, the Netherlands. The size and weight of one sam ple slab, of each product The code number shows the placement of the equipment in the plant (Fig. 21). made, should be checked each produc ti on day, and the findings enter ed in a perm anent record. Item Number Code Monitoring by a testing centre Bark removers (steel scrapers) 5 – The plant’s own inspec tion should, if possi ble, be cor- Saw to cut logs 1 1 rob o rated by an in de pend ent test ing cen tre, per haps Manual fork-lifts 3 15 twice a year. Saw blade grinding machine 1 17 An overall inspec tion of each product should be done Air compressor 1 19 reg u larly, per haps once a year, by an ac cred ited testing Concrete blocks cen tre. The most im por tant checks are size, den sity, (for compressing slabs in the moulds) 20 – 30 – bending strength and com pression strength. Delivery pallets (for storage of slabs and delivery to the customer) – – Oil for the hydraulic system 600 litres – Plant Oil and grease for machine maintenance – – For the produc ti on proces s descri bed, it is rec om mended Main transformer station with distribution station 1 – that the plant be about 55 ´ 40 m with a height of at least Distribution system for process water – – 5 m. See the de sign in Fig. 21. This building is large Tools for the workshop – – enough to per mit in crease au to ma tion with out ma jor al- Laboratory equipment for testing slabs – – ter ation. There will be unused space at first, but this Basic electric installation material – – space might well be used to pro duce pre fab ri cated el e - Equipment for plant cleaning – – ments. Containers for transport of waste – – The plant should be service d by roads suitable for lor- Table 3 Equipment normally available locally. ries. It needs to be con nected to water and elec tricity . Lighting is nec essary , and possi bly even heating. Some parti tion walls are needed to sepa rat e the workshop, of -

10 Building Issues 1994 Volume 6 • Number 3

Fig. 21 Design for a semi-automated plant. (Detailed proposals for plants with different degrees of automation are available from Elten Systems, the Netherlands.)

Item Number Code Cost (US$) Staff Scales for woodwool 1 3 10,000 For produc ti on 22 labour ers (see Table 5), a super vi sor Salt solution preparation unit 1 4 20,000 and a mainte nance en gi neer are needed, which makes a Woodwool conveyor 1 6 28,000 total of 24 produc ti on staff. There are also adm inis tra - Roller conveyor 1 13 6,000 tors, sales and market ing staff, etc. Edge trimming station 1 16 47,000 No. of Table 4 Equipment that might be purchased locally. Tasks workers Approximate cost 1 Dec. 1994 according to Elten Systems if the equipment is imported. Slab production Debarking tree trunks, sawing into logs, shredding Moulds and transport of woodwool to scales. 3 It is rec om mended that the moulds have bottom s of wa- Weighing woodwool and loading the immersion tank 1 ter-resistant plywood and sides of wood strips, see Fig. Preparation of accelerator solution, filling cement, 22. The plywood should be about 20 mm thick and pref- cleaning, oiling, etc. 2 er ably be sur face treated with phe no lic resin. The edge Filling and spreading the mixture in the moulds 4 strips should be about 50 mm wide. The recom mended Transport of filled moulds with a manual fork-lift 1 inside dim ensions for the moulds is 2 – 3 m long, 500 – Stacking the moulds, compression in the hydraulic press 2 600 mm wide, and 15 – 150 mm high. Stripping the set slabs from the moulds 2 The num ber of moulds needed de pends on the pro- Transport of the stripped slabs and trimming the edges 3 ducti on volum e and the thickness( es). With a daily pro- Stacking the slabs on pallets 2 ducti on of 32 m3, if the slabs are 50 mm thick and 1 m2 Delivery in area (for ex am ple 500 ´ 2,000 mm), the the o ret i cal Transport with manual forklift 1 num ber of moulds needed is 640. However , there should Maintenance alway s be some spare moulds, and one should have Maintenance of machines, sharpening the shredder knives and saw blades 1 about twice as many, that is 1,300. A mould should last about five years with good care and mainte nance . Total 22 Table 5 Labour needed for production.

11 Volume 6 • Number 3 Building Issues 1994

470 kg/m3. An annual produc ti on of 7,700 m3 require s the fol low ing raw ma te ri als.

Wood 1,000 metric tons (including wastage) Cement 1,875 metric tons Water 1,100 metric tons (including general use) Accelerator1 18 metric tons

1 The accelerator can be calcium chloride or any other accelerating salt. A solution of 2% is assumed. However, the amount of accelerator depends on the type of wood. Fig. 22 The moulds are normally made with a plywood bottom and wooden strips on the sides. Note there are sev eral ad van tages in making slabs with a densit y lower than 400 kg/m3: ther mal con - Setting up the plant duc tiv ity decreases (see Fig. 23), the slabs are eas ier Before start ing produc ti on, the equipm ent must be set up to handle (lighter) and the produc ti on volum e is and tested. Me chan i cal and elec tri cal in stal la tions are su - greater for the same amount of raw mate ri als. pervise d by quali fie d profes sion als with help of the fu - A lower densit y however means that strength de- ture labour ers. Insta lla ti ons can take 1.5 – 2 months. creases. This makes no dif fer ence in most ap pli ca - Each machine must be tested af ter in stalla tion. When tions. Lower densit y however re quires good qual ity all machin es work well, the plant starts pilot pro duc tion, woodwool, which must be elasti c and strong. while the staff is trained by ex perts. This train ing takes 2 If slabs are made with a densit y of 300 inste ad of – 3 weeks. 400 kg/m3, ther mal con duc tiv ity im proves from about 0.085 to about 0.072 W/mK, and the volum e of produc ti on increa ses about 30% for the same Consumption of raw materials and energy amount of raw mate ri als. Raw materials Produc ti on capac it y is usuall y governed by the num ber of shredders. With a norm al thickness of woodwool Energy consumption (about 0.3 mm) each shredder has a ca pacit y of 300 – The equipm ent recom mended re quires a total of about 75 400 kg of woodwool per hour. kW. The most de manding machine is the shredder which The produc ti on method de scribed here makes slabs draws about 20 – 30 kW depend ing on the brand. Ener gy with a densit y of 250 kg/m3 or more. With an air-dry consum pti on for the plant with one shift per day is about densit y of 400 kg/m3, about 32 m3 of woodwool slabs 75,000 kWh per year (240 work days and a load factor of can be made during an 8 hour shift (7 hours ef fecti ve about 60%). work). This gives a year’s produc ti on of 7,700 m3 (reck- oning 240 work days per year). Production costs Estimated production cost 3 To produce 8,000 m of woodwool slabs per year An exam ple is given of the cal cula ti on of produc ti on 3 require s about 2,000 m wood. An annual fell ing of costs for 1 m3 of woodwool slab with an air-dry densit y this volum e of wood require s a planta ti on of an of 400 kg/m3. The calcu la ti on is done for the plant rec- esti mated 200 – 500 hect ares. om mended in this chapter and as sumes one work shift It is im portant that produc ti on of woodwool slabs per day, equiva lent to an annual produc ti on volum e of does not lead to defor es ta ti on. The wood must be 7,700 m3. The current prices for im ported equipm ent taken from well-managed forest s and planta ti ons, were used. Other prices and costs, which vary greatly with con tin u ous re plant ing. be tween coun tries, were es tim ated in this ex am ple. The Indus tri es using wood do not neces sar il y lead to a calcu lation is set up so that the cor rect val ues can eas ily de crease in for ests. On the con trary, comm er cial for - be put in. estry can prom ote wood planta ti on, if re plant ing is The calcu la ti on is done for the first year of produc - seen to pay. tion. Over time bank inter est will go down, but the cost for mainte nance will go up. Therefore it is not expect ed The maxi m um capac it y of a plant is thus 23,000 m3 that the total produc ti on costs will change very much. (with 3-shift produc ti on). To increa se produc ti on capac - Possi ble im port taxes on the equipm ent are not in-cluded ity per shift require s more shredders and moulds. (Pro- in the calcu la ti on. It is assum ed that there is conti nu ous posals for larger plants than that de scribed here can be produc ti on, that is the dem and is rather consta nt and obtai ned from Elten System s, the Neth erlands .) there are no big techni cal problem s. The ra tio betwee n woodwool, cem ent and water (in- The calcu lat ed produc ti on cost for 1 m3 woodwool cluding ac cel era tor) is about 1:2:1. For slabs with a slab is 52 US dollar s (see Table 6). If 50 mm thick slabs air-dry densit y of 400 kg/m3, the wet densit y is about are produced, the cost per m2 is 2.6 US dollar s.

12 Building Issues 1994 Volume 6 • Number 3

Type of cost Cost/m3 (US$) Calculations Raw materials 16 Cost of raw materials1 (in US dollars) Consumables 2 Total Salaries 13 Quantity Unit price annual cost Energy 1 Cement 1,875 tonnes 50 94,000 Capital costs for equipment 14 Wood 1,000 tonnes 25 25,000 Capital costs for plant and site 6 Water 1,100 tonnes 1 1,000 Total production costs 52 Calcium chloride 18 tonnes 100 2,000 Sum 122,000 Table 6 Estimated total production costs during the first year (in US dollars) for 1 m3 woodwool slab with an air-dry 1 Including transport density of 400 kg/m3 with an annual production of 7,700 m3. Cost of consumables (in US dollars) Quantity Unit price Total annual cost Note that if the slabs are made with a densit y of 300 Moulds 1,300 pcs 50 13,0001 inste ad of 400 kg/m3 (which means produc ti on in- Other 2,0001 creases from 7,700 till 10,000 m3/year), the cal cu - Sum 15,0001 lated produc ti on cost decrea ses from US dollar s 52 1 With an expected lifetime of 5 years to 40 per m3 (a de crease of 23%). Note also that if the in ter est rate is 5% in stead of Salaries (in US dollars) 10% the cal culat ed produc ti on cost drops from US Number Cost/person1 Total annual cost dollar s 52 to 46 (a decrea se of 12%). Engineer 1 8,000 8,000 Supervisor 1 6,000 6,000 Labourers 22 4,000 88,000 Comparison with other materials Sum 102,000 The ac tual cost for the mate rial will be sig nif icantly 1 Annual salary including insurance, employers tax, etc. greater than the esti mated produc ti on cost. Overhead Energy cost (in US dollars) costs in clude costs for man age ment, ad min is tra tion, sales and market ing, taxes, profit, etc. When the produc ti on Quantity Unit price Total annual cost starts for the first time, there are also costs for in stal la - Machines 75,000 kWh 0.04 3,000 tion, training and the pilot produc ti on. Other 25,000 kWh 0.04 1,000 It is dif ficult to com pare prices of woodwool slabs Sum 4,000 and other building mate ri als since the prices vary greatly Capital costs for equipment (in US dollars) betwee n countri es. In general woodwool slabs are some- what more expen sive per m2 than ther mal in su la tion ma- Investment cost te ri als such as mineral wool and cel lu lar plas tic to give Imported equipment 450,000 equal ther mal in su la tion. Transport (6%) 30,000 Compared to tra diti onal construc tion using, for exam - Other equipment 160,000 ple, hol low blocks, build ing com po nents con tain ing Total 640,000 woodwool slabs are often cheaper. This is becaus e con- struc tion is faster with the mate rial, and it combin es good Total annual cost ther mal in su la tion ca pac ity with high strength. Amortization 640,000 / 15 years 43,000 Interest 640,000 ´ 10% 64,000 Sum 107,000

Note that if second hand or local ly produced equip- ment are used as much as possi ble, invest ment costs will be sig nif i cantly lower.

Capital costs for plant and site (in US dollars) Investment cost Plant building 300,000 Site (8,000 m2) 50,000 Total 350,000

Total annual cost Amortization 350,000 / 30 years 12,000 Interest 350,000 ´ 10% 35,000 Sum 47,000

13 Volume 6 • Number 3 Building Issues 1994

Improving the compatibility 4 Choice of wood of the wood with cement

This chapter deals only with cem ent-bonded slabs, but Air-drying (seasoning) Air dry ing lowers the content of sugars and other com - the proce dures to choose a suitable wood are the same pounds that inhibi t setti ng, and norm ally takes 3 – 6 for magnesite-bonded slabs. months. Technical requirements for the wood Addition of an accelerator The inhi bi ti on of the cem ent setti ng by the wood can be Ease of shredding compen sated by the use of an accel er ator . Dis solved in The wood must allow shredding at a rea sonable cost. wa ter, these are of ten called “min er al iz ing flu ids.” They This means it should not be too hard and/or contai n work by making the slab set before the “cem ent poi sons” much sil ica, which causes ex ces sive wear on the knives. in the wood have time to dif fuse into the ce ment paste. Woods with an air-dry densit y over 750 kg/m3 are not The most com mon ac cel er a tor is cal cium chlo ride nor mally suitable. The most suit able spe cies have few (CaCl2). Other pos si ble ac cel er a tors are mag ne sium branches and trunks that grow rela ti vely straight. chlo ride (MgCl2), wa ter-glass (so dium sil i cate, Na2SiO3, or po tas sium sil i cate, K2SiO3), alum inium sulphate Wood specie s that give long, strong and elasti c (Al2(SO4)3) and lime wa ter. The ac celer ator is usu ally strands are best to achieve good ther mal insu la ti on added to the water bath that the woodwool is soaked in capac it y. This type of woodwool allows produc ti on before it is mixed with the cem ent. The norm al concen - of slabs with low densit y and suf ficie nt bending tra tion is be tween 1 – 5% by weight of the water in the strength. On the other hand if the woodwool easil y mix. breaks into short strands, it is dif fi cult main tain both There are sev eral rea sons to keep the amount of ac cel- low densit y and good bending strength in the slabs. era tor low. One is produc ti on costs, es pecia lly if the ac- cel era tor is expen sive or must be im ported. If the slab The di am eter of the tree trunk should norm ally be con tains chlo ride, which it will if cal cium chlo ride or greater than 100 mm and less than 400 mm. (There are mag ne sium chloride was used as ac celer ator , there is a also spe cial shred ding ma chines that han dle di am e ters clear risk of corro sion and galva nize d nails and screws under 100 mm). must be used. The wood’s effect on setting Rapid-hardening cement Cem ent-bonded woodwool slabs consis t of or ganic The amount of accel er ator can be re duced – per haps woodwool enclos ed in inor ganic cem ent paste (Portland none is needed at all – by using rapid-hardening (quick cem ent and water). Like all cel lu lose mate ri als, wood to set ting) Portland cem ent. It is more finely ground than greater or lesser ex tent in hib its the set ting of the ce ment. ordi nary cem ent and thus more expen sive . This is caused by wood sugars and other com pounds that leach out of the wood in contac t with the ce ment paste. If Leaching the specie s of wood prevents set ting, or delay s it too An other, of ten eco nom i cal, way to im prove the com pat i - much, the wood is not suit able. bil ity of the wood with ce ment is leaching. This is done by pre-soaking the woodwool in water for 24 hours (pos- sibly 48 hours with a change of water after the first day). Availability of the wood Someti mes the woodwool is soaked for some period in If a wood spe cies is to be consid ered suit able for produc - hot wa ter or in solu ti on of cal cium chloride , water -glass tion of woodwool slabs, it is not enough that the techni - or so dium hy drox ide (NaOH). cal re quire ments are met. It is also impo r tant that the wood is avail able in ade quate quanti ty and at a rea son- Tests of suitability able price. The proxim ity to the source of wood is im por- tant, since transport costs would other wise be high. Full-scale tests Quick grow ing species are of ten pref er able, since these The only way to be sure that a wood spe cies is suit able woods are norm ally the cheapest . for woodwool slabs is to produce test slabs on full scale. An annual produc ti on of 8,000 m3 woodwool slabs, Such a test re quires ac cess to a woodwool shredder . as assum ed in Chapter 3, require s a planta ti on of about If this is not avail able, logs can be sent away for shred- 200 – 500 hect ares, assum ing it is well maintai ned and ding, and the woodwool shipped back. It might be possi - re planted af ter fell ing. The pre cise area de pends on how ble to conduct tests at the closest woodwool slab fac tory, quickly the spe cies grows. which gives one access to the fac tory’s prac tical ex pe ri ence, which can be very valu able. Full-scale tests Tree spe cies with lit tle commer cial value might be can also be conducte d by Elten System s, the Neth er- used in woodwool slabs. Quick growing spe cies, lands. such as pulpwood, that are not appro pri ate for furni - Observe how the wood shreds. The wood should be ture, etc. are of ten suit able. easy to shred so that the knives do not wear out too fast.

14 Building Issues 1994 Volume 6 • Number 3

(Normally the knives should last for eight hours produc - tion before they need sharpen ing.) 5 Properties The slabs them selves can be made by hand. The ra tio of woodwool : cem ent : water should be 1:2:1. First soak Thermal properties the woodwool in water (perhaps with an accel era tor), Thermal conductivity then sprinkle on the ce ment. (Alter nati vely the dry The therm al conduc ti vit y of woodwool slabs is mainly woodwool can be put into a ce ment slurry.) Mixing can a functi on of their densit y and moisture content ; it in - be done with a pitchfork in a trough or in a ce ment creases as den sity and/or mois ture con tent in creases. Fig. mixer. The mix is packed into a mould of appro pri ate 23 shows therm al conduc ti vit y in practi ce as a func tion size, such as 50 ´ 500 ´ 1,000 mm. of air-dry densit y. (Ther mal conduc ti vit y for oven-dry The main aim of making test slabs is to see if the slabs ma te rial is sig nif i cantly lower.) Ta ble 7 com pares the hold togethe r and how long time they take to set before ther mal conduc ti vit y of a woodwool slab, with they can be stripped from the mould. If the test slab sets 400 kg/m3 den sity, with other mate ri als. sat is fac to rily, in a rea son able time, one can ex am ine its bending strength and com pression strength. For these, and other tests, see Chap ter 6 Qual ity con trol. A large num ber of wood spe cies have already been tested by produc ti on of full-scale slabs. They are listed in Ap pen dix (Ta ble A1). Screening tests There are sev eral small-scale tests to de ter mine the suit - abil ity of a wood. Note that it is not possi ble to estab - lish suit abil ity with these tests; they only help elim i - nate com pletely un suit able woods. A rather good method is to make small test slabs with wood wool, ce ment and wa ter (per haps con tain ing ac cel- era tor). This gives an im pression of the shredding qual ity of the wood, and if the slabs set satis fac to rily. If these Fig. 23 Approximate values for the thermal conductivity (l) as a function of air-dry density for woodwool slabs tests are prom ising, then full-scale tests are done and the (moisture content 8 – 10%). slabs are tested for strength. Woods tested by this method (Sources: Information from different manufacturers and are shown in Appen dix (Table A2). Cammerer (1962): Der Wärme- und Kälteschutz in der There are other more or less sim ple screen ing tests, Industrie). but since they are not reli able they are not de scribed here. Woodwool slabs allow air to pass easil y, which could in crease the ther mal con duc tiv ity by forced con vec tion when the mate rial is left un plas tered (acous tic ap pli ca - tions as in Fig. 44). Studies show however that if one side of the slab is sealed, norm al values for therm al con - ducti vit y apply as well to this kind of construc tion. Thermal capacity Woodwool slabs have a high specif ic therm al capac it y (about 1,600 J/kgK at 10% moisture content ). The rela - tively high therm al capac it y of roofs and walls con - structed with woodwool slabs can signif i cant ly im prove in door com fort, since in door tem per a ture changes are atten u ate d, when there are large diur nal varia ti ons in out - door tem per a ture. The therm al capac it y of woodwool slabs is com pared with other mate ri als in Ta ble 7.

Strength The standards for strength ac cording to DIN4 1101 for cem ent and magnesite-bonded woodwool slabs are given in Table 8. The mini m um values accord ing to DIN are low; mod ern in dus tri ally pro duced wood wool slabs might have several times the strengths shown.

4 Deutsches Institut für Normung (“German standards institution”).

15 Volume 6 • Number 3 Building Issues 1994

The strength of woodwool slabs, espe cia lly if they are Acoustic properties ce ment-bonded, is in sig nif i cantly af fected by high rel a- Sound absorption tive hu mid ity. Woodwool slabs have good sound absorp ti on, which Bending strength makes them suit able for all kinds of publi c gather ing The bending strength of a slab is high rela ti ve to its places, in dus tries, etc. weight, be cause the two com ponents – woodwool and The sound absorp ti on norm ally increa ses somewhat binder – com plem ent each other: the wood strands take with in creased thick ness, es pe cially for low fre quen cies. the ten sile stress, while the hard ened ce ment paste (or Sound absorp ti on is also af fected by proxim ity to other mag ne sia cem ent) takes most of the compres sive stress. mate ri als, while paint ing the slabs has only slight ef fect. The tensil e strength of the wood strands is crucia l to the Fig. 24 shows sound absorp ti on at dif ferent frequen - slab’s bending strength. cies both for a free-standing slab, such as a sus pended Bending strength increa ses with densit y. Normally slab, and for a slab next to a hard mate ria l, such as a thin slabs are made with higher densit y than thick slabs woodwool slab against cast concret e. to give ade quate bending strength. Load bear ing ca pac ity can be in creased sig nif i cantly by re inforc ing the slabs. Steel channels or wooden poles or bars can be used, see Fig. 4. These slabs can have a bending mom ent capac it y of over 2 kNm and are about 4 times stronger than an ordi nary slab of the same thick- ness. Compressive strength Woodwool slabs have a rela ti vely high com pressive strength that, as bending strength, increa ses with densit y. The com pressive strength per pendic u lar to the plane of the slab is ex pressed as compres sive stress at 10% com- pression (see Table 8). Com pressive strength of wood- wool slabs is com pared with other mate ri als in Table 7. Fig. 24 Sound absorption at different frequencies for a 50 mm thick woodwool slab. Curve A is the slab with at least Volumetric 50 mm clear space over. Curve B is the slab lying Thermal thermal Compressive against a hard material. Density conductivity1 capacity strength (Source: Anon. 1985). Material (kg/m3) (W/mK) (J/m3K) (MPa)

3 2 Woodwool slab 400 0.085 640 ´ 10 0.2 – 1.0 Autoclaved Sound insulation aerated A woodwool slab it self gives very moder ate sound insu - 3 concrete 400 0.10 420 ´ 10 1.7 lati on, but if it is plaster ed on one side, the sound insu la - 3 Glass wool 50 0.036 44 ´ 10 0.01 tion proper ti es are good (sound re ducti on of about Expanded 30 dB). If a plas tered woodwool slab is part of a heavy polystyrene 30 0.036 36 103 0.1 ´ wall, made of brick or concret e for exam ple, sound insu - Hollow brick 800 0.47 700 103 – ´ lati on is very good; (sound reduc ti on is 35 – 55 dB de- 1 Values in practice. 2 At 10% compres sion. pending on the weight of the wall). Good sound insu la - tion can also be achieved with a wall made of two plas- Table 7 Comparison of properties for some insulation materials and hollow brick. Note the high thermal conductivity for tered woodwool slabs with an air cavit y (sound reduction hollow brick and the low thermal capacity for mineral of about 50 dB). wool and expanded polystyrene.

Bending Compressive Fire performance Thickness Weight1 Density1 strength2 strength2 3 In spite of the wood content , woodwool slabs have good (mm) (kg/m2) (kg/m3) (MPa) (MPa) re sis tance to fire. The mate rial is classed as hard to ig - 15 8.5 570 1.7 0.20 nite, and is therefore approved for indoor surface s ac - 25 11.5 460 1.0 0.20 cording to inter nati onal standards. A 50 mm thick slab 50 19.5 390 0.5 0.15 resis ts fire for 1 hour and a 100 mm thick slab for 75 28.0 370 0.4 0.15 2 hours. 100 36.0 360 0.4 0.15 The good fire per for mance of the mate rial is re lated to 1 Maximum value. the fact that the wood strands are protec ted by the binder, 2 Minimum value. 3 At 10% compression. as well as its ther mal in su lating ca pac ity and coarse struc ture. If the mate rial is cov ered with a layer of ce - Table 8 Standards for weight and strength of woodwool slabs for different thicknesses according to DIN 1101. ment or gyp sum plaster , fire resis tance in creases fur ther.

16 Building Issues 1994 Volume 6 • Number 3

Moisture properties The mate ria l has been buried in the ground for 30 years and kept under water for 10 years without destruc tion. Woodwool slabs have the abil ity to absorb large amounts Mould re sis tance is an im por tant qual ity for healthy of moistur e. If the rel a tive hu midity in the air ex ceeds buildings. Where moisture is high, mould is com mon on 95%, the moisture content of the slab is more than 20%. untre ated wood and on wooden boards that are not ce - (When dipped in water unti l satu ra ti on, the moisture con- ment-based. tent is about 30%.) Because of their capac-ity to ab sorb moisture , woodwool slabs are suitable where the rel ati ve Resistance to insects and termites hum idit y is occa sion all y very high, for exam ple in sports Becaus e the wood strands are covered by binder, resis - halls. The slabs at tenu ate the varia ti ons in the indoor air tance to in sects and ter mites increases sig nif icantly . hum idit y, by absorb ing moisture rapidl y when there is a Some studies show however that both cem ent and moisture input (when the rela ti ve hum idit y rises) and re- magnesite-bonded woodwool slabs might be at tacked by leas ing this mois ture when the rel a tive hu midity de - ter mites (see Kumar 1980). creases. Most ther mal in su la tion ma te ri als lack this abil - The re sults are very good in practi ce. Woodwool slabs ity. Fig. 25 com pares the abil ity of woodwool slab with have been used at least since the 1960s in countri es with wood (pine), brick and mineral wool to absorb moisture . severe term ite problem s, without any reports of term ite When wet woodwool slabs dry to air-dry (about 50% attack on the slabs. rel ati ve hum idit y), shrinkage in length is about 3‰. For safety reasons the risk of term ites must however The vapour diffusivity of woodwool slabs (at 20°C) is be consid ered, if the slabs have an acti ve functi on in the about 10 ´ 10–6 m2/s, whereas the vapour resistivity is loadbearing con struc tion. about 20 MNs/gm. Plas tering the slabs will fur ther re duce the risk of ter mite at tack. Resistance to an aggressive environment Woodwool slabs have good resis tance to aggres sive air. The mate ria l re sists sulphur in the air and has been very succes sfull y used in swim ming halls where the air often contai ns chlorine and chrome. Woodwool slabs in facto - ries where ag gres sive chemi cals are used have not de teri - o rated.

Emission of harmful compounds Woodwool slabs have littl e emissions . Unli ke glued wood-based boards, such as ply wood and chipboard, the ma te rial does not re lease form al de hyde. The to tal amount of vola ti le or ganic com pounds (TVOC) relea sed by a ce - ment-bonded slab was measured at less than 11 mg/m2h, which is very low (Tests by Swed ish Na tional Testing Fig. 25 The ability to absorb moisture. The moisture content and Re search In sti tute). (kg moisture per m3 material) as a function of relative humidity for woodwool slab (density 400 kg/m3), pine (500 kg/m3), brick (1,700 kg/m3) and mineral wool (20 kg/m3). (These absorption isotherms are valid at equilibrium at 20°C) (Source: Own measurements and Elmarsson and Nevander, 1981: Fukthandbok)

Durability Resistance to rot and mould Cem ent and magnesite-bonded woodwool slabs have surpris ingl y good re sis tance to rot and mould. This is be- cause the binder creates a chemi cally basic en vi ron ment that protec ts the wood strands (pH-value ³9). Woodwool slabs have been in use for over 80 years, and the ex pe ri ences of all areas of ap pli cation are uni - formly good. Rendered slabs have sat on facade s ex- posed to heavy rains for over 50 years without rot ting or moulding. Expe ri ence from swim ming halls, where the rel ati ve hum idit y can be over 80%, is also very good.

17 Volume 6 • Number 3 Building Issues 1994

Method. Five slabs of the same type are tested in a hy- 6 Quality control drauli c press (or the equiva lent ). Each slab is placed on two sup ports (pref er ably roll ers) that are at least as wide The sugges ti ons in this chapter are largely based on the as the slab (Fig. 26). The slab is placed so that the sup- Germ an standards for woodwool slabs, DIN 1101, but ports are placed one quarter in from each short end (so lo cal stan dards may be es tab lished. that the weight of the slab itself does not affect the test). The slab is loaded with a lin ear load across its cen tre Conditions of testing (paral lel to the supports) as shown in Fig. 26. The load is The tests should be carrie d out on five random ly selec ted gradu all y increa sed unti l the slab breaks. The bending slabs of each product produced. Before test ing, the slabs should be condi ti oned for at least two weeks under as strength (the modulus of rupture ), sb (MPa), is cal cu - con stant room tem per a ture and hu mid ity as pos si ble. lated as follow s: 2 sb = 0.75F ´ L / (w ´ t ) Size and density where F is the breaking load in Newtons (N); L, w and t are the length, width and thickness of the slab re- Size Re quire ment. The slab is measured and the follow ing spec tively (mm). varia tions are al lowed: length ±10 mm, width ± 5 mm, Compressive strength thick ness ± 3 mm. Re quire ment. A standard for the low est com pressive Method. Use a steel measur ing tape. Measure the strength (de fined as the com pressive stress at 10% com - length of each slab in three places and the width in four pression) should be estab li shed for each type of slab pro- places. Mea sure the thick ness of the slab with callipers in duced. See for exam ple Table 8, which gives the DIN 10 places. The av er age of the measure ments for each di- standard. The aver age com pressive strength of five spec- mension should fall within the range spec ifie d. im ens should not fall below the chosen standard. The com pressive strength of an indi vid ual sam ple should not Density be more than 10% be low this standard. Re quire ment. A standard for the high est densit y should Method. From five slabs of the same type, saw square be es tab lished for each type of slab made. See for ex am - spec im ens with an edge L of, say, 200 mm. The speci - ple Table 8, which gives the DIN standards. The aver age mens are tested in a hy drau lic press (or the equiv alent). of the five slabs must be equal to or lower than the cho- Each spec im en is placed between two rigid metal plates sen standard. The densit y of an indi vid ual slab may not with the shorter side at least as long as the edge of the excee d this standard by more than 15%. spec im en. Each spec im en is sub jected to a load over its Method. Five slabs of the same type, whose dim en- centre (Fig. 27). The speci m en is first loaded with a force sions are de ter mined as described above, are each of 100 N to de ter mine its initi al thickness. The pressure weighed on scales with an ac cu racy of ± 0.5 kg. The is then in creased steadily un til the spec imen is com- den sity of each slab is cal cu lated. pressed to 90% of its ini tial thickness. Com pressive strength, sc (MPa), is then cal cu lated as fol lows: Strength 2 sc = F / L Bending strength where F is the load (N) at 10% com pression. Re quire ment. A standard for the low est ac cept able bend - ing strength should be estab li shed for each type of slab made. See for ex am ple Table 8, which gives the DIN standards. The aver age value for the bending strength of the five slabs must not fall below the chosen standard. The bending strength of an indi vid ual slab may not be more than 10% be low this standard.

Fig. 27 Set up for the test of compressive strength.

Other tests Fig. 26 Set up for the bending strength test. Thermal conductivity Re quire ment. A standard for the high est per mit ted ther - mal conduc ti vit y should be es tabli shed with re spect to the densit y of the slab. The same standard can be used

18 Building Issues 1994 Volume 6 • Number 3 for slabs of dif ferent thicknesse s but with about the same densit y. (The DIN standard for slabs with thicknesse s 7 Applications ³ 25 mm is 0.090 W/mK. Method. Ther mal con duc tiv ity is closely tied to den - Exterior walls sity (compar e with Fig. 23), that is a test of den sity is an Thermal insulation requirements in di rect test of ther mal con duc tiv ity. How ever, if pos si - It is dif ficult to give a general standard for the therm al ble, at least one therm al conduc ti vit y test should be done trans mittance (U-value) of an ex terior wall, since it per product in a labo ra tory . How to measure therm al con - depends on the local cli mate and on econom ic factors. ducti vit y is descri bed in, among others, DIN 52 612 part Adam son and Åberg (1993) rec om mend a maxi m um 1 and 2, Brit ish Standard (BS) 874 and Ameri can Stan- U-value of 1.0 W/m2K for walls of air condi ti oned dard (ASTM) C 518. Since therm al conduc ti vit y is also houses in warm and hum id clim ates. The types of walls re lated to the mois ture in the mate rial, the slabs should de scribed here fol low this rec om men da tion. be condi ti oned to a define d moisture content before test - ing. Walls built on site H Acoustic properties The wall in Fig. 28 is made of ele m ents contai ning Slabs intende d for sound absorp ti on should, if possi ble, a cavit y. Each ele m ent consis ts of two woodwool slabs be tested to produce a profil e of sound absorp ti on at dif - sandwiching spacer strips of woodwool slab. The spacer fer ent fre quen cies (see Fig. 24). strips should be at least 50 mm thick. The parts are as- The sound insu la ti on proper ti es of the slabs might be sembled with ce ment slurry. The slabs in the el em ents tested for the types of walls and roofs they are norm ally might stand ver tically or hor izon tally. In the lat ter case used for. Such tests are descri bed in BS 2750 (sound re - the dis tance can be in creased between the con crete col - duc tion in dex). umns, and each col umn can be cast in ver tical stages. Fire performance Fire per for mance is re lated to the den sity of the slabs, that is if the slabs meet the den sity re quire ments, their fire perfor m ance is also known. If possi ble the fire per- form ance of each product should be tested at least once. Tests of fire per for mance are described in, for ex am ple, Fig. 28 Wall of vertical (or horizontal) elements consisting of two DIN 4102 part 1, BS 476 part 1, 5, 6 and 7. woodwool slabs sandwiching spacer strips of woodwool Chloride content slab. The elements are joined by in-situ cast columns of reinforced concrete. (Using 30 mm thick woodwool slabs Re quire ment. Accord ing to DIN 1101 a woodwool slab gives a U-value of 0.9 W/m2K). may not contai n more than 0.35% wa ter solu ble chlo- rides (measured in percent weight of the slab’s oven-dry weight). Ad van tages. The walls are light and fast to build. The Method. The prin ci ple for the test is that chlo rides are cav ity can be used for electri cal cables. The wall has leached out by disti lla ti on and the content is de ter mined good ver tical loadbearing ca pac ity and is sta ble be cause by potentiometric titra tion. The method is descri bed in of the good adhe sion of the concret e col umns to the el e- DIN 1101. ments. It has no signif i cant ther mal bridges and is safe in case of ter mites, since the load is car ried by the con crete. Dis ad van tages. Wall ele m ents must be pre fab ri cated, requir ing an extra step in the work proces s.

Calculation of U-value for the wall in Fig. 28 t (m) (W/mK) R (m2K/W) = t / l l External rendering 0.02 1.0 0.02 Woodwool slab 0.03 0.085 0.35 Some national standards Cavity 0.05 — 0.17 for woodwool slabs Woodwool slab 0.03 0.085 0.35 Plastering 0.01 1.0 0.01 Aus trian ÖNORM B6021 External + internal Brit ish BS 1105 surfaces — 0.17 Total 0.14 1.07 Ger man DIN 1101; re quire ments, testing U = 1/R = 0.93 W/m2K 1102; ap pli ca tion total The small decrease in the insulation capacity caused by the In dian IS:3308 concrete columns is ignored in the calculation. Dutch NBN 638 Swed ish SIS 238101

19 Volume 6 • Number 3 Building Issues 1994

H The wall in Fig. 29 consis ts of hom oge neous slabs with the ends sawed into a V-shape. A square cavit y is cre ated when the slabs are placed next to each other, and re in forcem ent is placed in the cav ity. Con crete of fluid consis tency is poured in. The slabs can be placed verti - Fig. 31 Wall of vertical slabs reinforced by galvanized steel cally or hor izon tally. In the latter case the dis tance be - channels. (Using 70 mm thick woodwool slabs gives tween the con crete col umns is greater, and they can be a U-value of 1.0 W/m2K). cast in stages. kind of wall is used in Mexico, Vene zuel a and Panam a, among other places. Ad van tages. The wall is thin and in spite of that has a low U-value (1.0 W/m2K for a thickness of 100 mm). It has good ver tical loadbearing ca pac ity, is ex tremely Fig. 29 Wall of vertical (or horizontal) woodwool slabs with quick and easy to build and safe in case of term ites, since V-shaped ends. The slabs are joined by columns of the load is car ried by the steel chan nels. in-situ cast concrete. (Using 70 mm thick woodwool Dis ad van tages. The steel chan nels form therm al slabs gives a U-value of 1.0 W/m2K). bridges which raise the U-value of the wall. H Fig. 32 shows a wall of slabs rein forced by wooden Ad van tages. The wall is thin and in spite of that has a poles to carry the ver tical load. low U-value (the U-value is 1.0 W/m2K for a total wall thickness of 100 mm). It has good verti cal loadbearing capac it y and is stable be cause of the good adhe sion of the concret e to the slab. The wall is light and quick to build, and it is safe in case of ter mites since the load is car ried by the con crete. Dis ad van tages. The concret e col umns go through the Fig. 32 Wall of vertical woodwool slabs reinforced by wooden wall and functi on as therm al bridges, which raises the poles. (Using 70 mm thick woodwool slabs gives a U-value of 1.0 W/m2K). U-value somewhat .

Ma te rials with very high ther mal con duc tiv ity, such Ad van tages. The wall is thin and in spite of that has a as concret e and steel, should not be al lowed to go low U-value (1.0 W/m2K for a thickness of 100 mm). It through an outer wall or the roof if pos sible. They has good verti cal loadbearing capac it y. The wall has no func tion as ther mal bridges pro mot ing ther mal flow. ther mal bridges and is very easy and quick to build. Also, when it is colder out doors than indoors, con- Dis ad van tages. The wall is not safe in case of ter - densa ti on can oc cur on these ther mal bridges caus - mites, since the load is carrie d by wood studs. ing dirty ing. H Fig. 33 shows a wall with woodwool slabs nailed on H The wall in Fig. 30 consis ts of pre fabri cat ed, both sides of a wooden loadbearing structure . I-shaped concret e col umns, with woodwool slabs placed hor i zon tally be tween. The dis tance be tween the col umns is the full length of the slab (2.4 m). This kind of wall is used in Zam bia (see Hawkes and Cox 1992). Ad van tages. The wall is thin and in spite of that has a low U-value (1.0 W/m2K for a thickness of 100 mm). It Fig. 33 Wall with a wooden loadbearing structure against which has good ver tical loadbearing ca pac ity, is ex tremely woodwool slabs are nailed on both sides. (Using 30 mm thick woodwool slabs gives a U-value of 0.9 W/m2K). quick to build and is com pletely ter mite safe. Dis ad van tages. The concret e col umns functi on as ther mal bridges and raise the U-value somewhat. Ad van tages. The wall has good verti cal loadbearing capac it y and is quick and easy to build. The cavit y can H Fig. 31 shows a wall of slabs rein forced with gal va- be used for electri cal cables. nized steel chan nels, which carry the ver tical load. This Dis ad van tages. The wall is not safe in case of ter - Fig. 30 Wall of prefabricated, I-shaped concrete columns, with mites, since the load is carrie d by wood studs. woodwool slabs placed horizontally between. The slabs are joined by mortar. (Using 70 mm thick woodwool slabs gives a U-value of 1.0 W/m2K).

20 Building Issues 1994 Volume 6 • Number 3

If the accel era tor used in making the woodwool slab con tains chlo rides, all nails and other metal fit tings that come into contac t with the slab should be hot-dip gal va nized.

H Fig. 34 shows a wall made of woodwool blocks. (The blocks are sawn from as thick a slab as possi ble, at least 70 mm). The blocks are laid to bond with lime ce ment mor tar, with thin joints.

Fig. 36 Wall element in the “Climatex system” consisting of 50 mm thick woodwool slabs with columns and beams Fig. 34 Woodwool block masonry. (Using 70 mm wide of reinforced concrete. The elements are rendered and woodwool blocks gives a U-value of 1.0 W/m2K). plastered (U-value = 1.3 W/m2K).

Ad van tages. A very quick and easy technique with great flex i bil ity. Dis ad van tages. The wall has limited ver tical load - bearing ca pac ity. H Fig. 35 shows a wall for a multi-storey building. The wall is made of re in forced con crete, cast in-situ be tween woodwool slabs which functi on as perm anent shutter ing.

Fig. 37 Above: House in Porto Alegre, Brazil, built with the “Climatex Fig. 35 Concrete wall for a multi-storey building where the system” (Fig. 36) with woodwool slabs are used as permanent shuttering. prefabricated wall (Using 25 and 50 mm thick woodwool slabs gives elements of wood - a U-value of 1.0 W/m2K). wool slabs and concrete. Ad van tages. Compared to an ordi nary concret e wall, this saves the work of taking down the formwork and Left: Interior showing plastered walls and a fin ish ing the sur face of the con crete. Elec tric cables and ceiling of gypsum other in stalla tions can be cast in the con crete. The wall coated woodwool has no ther mal bridges and is safe in case of ter mites. slabs (see Fig. 5). Prefabricated wall elements In Porto Alegre, Brazil, a building system is used for sin - gle-storey houses – the “Climatex sys tem” – with both ex te rior and partition walls of pre fab ri cated wall el e - ments made of woodwool slabs and concret e. During produc ti on in the fac tory, slabs are placed in a mould and

21 Volume 6 • Number 3 Building Issues 1994

Fig. 38 Supplementary insulation on a masonry wall. The slabs are fixed by mortar. re in forced con crete cast around them, leav ing space for windows and doors (Fig. 36). Dur ing as sembly the el em ents are tied to gether with rein force m ent bars sticking out from the ele m ents (the joints be tween the el em ents are later cov ered with con - crete). When all walls are up, they are ren dered and plas- tered. The wall el em ents are cur rently made of 50 mm thick Fig. 39 Woodwool slabs used in pitched roofs. woodwool slabs and have a U-value of about 1.3 W/m2K. A. Slabs nailed to the rafters as a base for the roofing material. (Using 100 mm thick woodwool slabs gives 2 To reduce the U-value to a maxi m um of 1.0 W/m K, the a U-value of 0.6 W/m2K). thickness of the slabs must be increa sed to 70 mm. B. Slabs nailed to the ceiling joists. A thinner slab is used Corru gate d asbes tos sheets have been the main roof- as a ceiling. (Using 100 and 25 mm thick woodwool ing mate ria l togethe r with a ceil ing of 25 mm woodwool slabs gives a U-value of 0.5 W/m2K). slab. Over 7,000 houses have been build with this system in southern Brazil since the 1970s. The houses are largely main tenance free and the res idents are very pleased with them (Fig. 37). See also van Elten (1982). Supplementary insulation Woodwool slabs are very suit able for both exter nal and in ter nal sup ple men tary in su la tion of walls. Against ma- sonry (with or without render ing) or concret e walls, the mate rial can be fixed with ce ment slurry, lime cem ent mortar or gypsum plas ter (Fig. 38). The mate ria l can be nailed or screwed to wood walls.

Roofs and ceilings Thermal insulation requirements A roof should have at least as good insu la ti on as the Fig. 40 Nailing woodwool slabs to the rafters on a pitched roof. walls. In trop ical cli mates, with high so lar el e va tion and strong solar radi a ti on, much of the heat transm ission is through the roof, and here the roof should be better insu - Rein forced slabs are often used in roof appli cati ons, lated than the walls. In air condi ti oned houses in tropi cal becaus e they are stronger than ordi nary slabs and allow a cli mates, Adam son and Åberg (1993) rec om mend a max- greater distanc e betwee n the support ing beams. im um U-value for the roof of 0.5 W/m2K. Most of the roofs descri bed here meet this rec om menda ti on. Flat roofs The roof shown in Fig. 41 is a vari ant of a type of flat Pitched roofs roof that is very com mon in many countri es. It norm ally A com mon use for woodwool slabs in pitched roofs is to con sists of pre fab ri cated (of ten pre stressed) con crete nail them to the raf ters as a base for the roof ing mate rial, beams with hollow blocks (of either concret e or burnt such as roof ing tiles, metal sheets, or simi lar (Figs 39A clay) in be tween and is cov ered with a re in forced con - and 40). Before apply ing the roofing mate ria l the slabs crete slab cast in situ. By re plac ing the hol low blocks are nor mally coated with a ce ment-sand screed. with woodwool slabs, and by increa sing the dis tance The woodwool slabs can also be nailed to the ceil ing betwee n the beams, the U-value of the construc tion can joists (Fig. 39B). be im proved sig nif i cantly.

22 Building Issues 1994 Volume 6 • Number 3

Fig. 41 Flat roof with woodwool slabs laid between prefabricated concrete beams. A slab of reinforced concrete is cast over the woodwool slabs and beams. The ceiling can either be left as it is, which give good sound absorption, or plastered. (Using a 150 mm thick layer of woodwool slabs gives a U-value of 0.5 W/m2K).

H The roof in Fig. 41 was used in two buildings in Tu - Both roofs met the strict Al geria n and Tuni sia n re- nisia (Fig. 44). The cost for the roofs was less than for a quire ments for loadbearing capac it y and defle cti on in tra diti onal roof using hol low blocks, even though the full-scale lab o ra tory tests (see Åstrand et al., 1994). woodwool slabs were im ported. The reason for the lower Ceilings cost was mainly that it was quicker to build. One of the most com mon uses for woodwool slabs is as Ad van tages. The roof is very quick to build and needs acous tic ceil ing panels in publi c gather ing places, corri - no formwork, only posts. It has good therm al insu la ti on dors, etc. The slabs can either be fixed to the roof (cast (U = 0.5 W/m2K). against concret e or screwed in as in Fig. 39B) or sus- Dis ad van tages. The beams must be pre fab ri cated and pended (Fig. 43). The air space betwee n the slabs and the they functi on as therm al bridges. roof influ ence s the sound absorp ti on somewhat (Fig. 24). H Fig. 42 shows a varia ti on of the roof in Fig. 41 using two layers of woodwool slabs. In this roof, enti rely cast in-situ, the therm al bridges create d by the concret e Surface finishes beams are broken by the under ly ing woodwool slab. Woodwool slabs provide and excel lent base for render - Ad van tages. The roof is quick to build. It require s no ing and plaster ing becaus e of their coarse texture . formwork when cast, only support consis ting of beams Reinforcement and posts. It has no therm al bridges and excel lent ther - Before apply ing render or plaster , all joints be tween mal insu la ti on (U = 0.5 W/m2K). slabs, or be tween slabs and an other mate rial (for ex am ple Dis ad van tages. This roof require s more support when a concret e col umn), should be rein forced. This is done to casting than the roof in Fig. 41. avoid cracks in the mate ria l caused by movem ents in the Note that for the roofs in Figs 41 and 42 it is very im - woodwool slabs due to tem pera ture and moisture por tant to cal cu late the shear rein forcem ent in the beams. changes. Re inforce m ent can be done with galva nize d This verti cal rein force m ent, which should be anchored in steel wire net ting, pref er ably welded netting (chicken- the con crete slab, in creases with the dis tance be tween the wire can be used, but be cause it is elas tic, it does not pre - beams.

Fig. 42 Flat roof consisting of two layers of woodwool slabs. Space is left in the top layer for reinforced concrete beams. These beams and the concrete slab are cast at the same time. The under layer of woodwool slab breaks the thermal bridges created by the concrete beams. The ceiling can either be left as it is, which gives good sound absorption, or plastered. (Using 30 and 120 mm thick woodwool slabs gives a U-value of 0.5 W/m2K).

Fig. 43 Ceiling of suspended woodwool slabs.

23 Volume 6 • Number 3 Building Issues 1994

References

Adam son, B and O Åberg 1993 Design for Climatization: Houses in warmhumid ar eas. Building Issues Vol. 5, No. 1. LCHS, Lund Uni ver sity, Swe den. Anon. 1985 Leichtbauplatten-Fibel. Bundesverband der Leichtbaupla ttenindustrie e.V., München, Ger - many. (In Germ an) Anon. Fig. 44 The roof in Fig. 41 as built in the youth centre of 1990 Woodcemair Wood Wool Cement Building Slabs Tameghza, Southern Tunisia. Woodwool slabs provide – Handbook . Torvale Building Products , both thermal insulation and sound absorption. Leominster, Herefordshire, United Kingdom . Åstrand, J, L Bessadi, E Johansson, S Laïd, H Teggour vent small cracks). The net, which should cover at least and N Toumi 100 mm on each side of the joint, is fixed to the slab 1994 Matériaux thermiquement isolants – béton with hot-dip galva nize d nails. mousse et panneaux en laine de bois. Re port 2. LCHS, Lund Uni versit y, Sweden. (In French) External rendering The slabs should be neither too damp nor too dry at fin- DIN ishing. A dry slab, for exam ple exposed to direct sun- 1989 Woodwool Slabs and Multi layere d Slabs as light, can be dampened a littl e before finish ing. A damp In su lat ing Ma te rials in Build ing: Re quire ments, slab, for exam ple one that has been in the rain, must dry Test ing. DIN 1101. Deutsches Institut für be fore fin ish ing. Normung, Berlin, Germ any. (In Germ an) The exter nal render ing should be in two or three van Elten, G J coats: spat ter dash coat (a prep ara tion of the base, max i- 1975 “Wood wool ce ment boards used for low cost mum 2 mm, opti onal), under coat (10 – 15 mm) and a fin - houses and other appli cati ons.” World Consul- ishing coat (paint or render 0 – 5 mm). The mortar of a tation on Wood Based Panels, Feb. 6–16, FAO, previ ous coat should be stronger (contai n more ce ment) United Nati ons, New Delhi, India. or as strong as the mortar for the next coat. A suit able 1977 “Pre fab El em ents from Wood Wool Cem ent for spat ter dash coat is lime-cement mor tar with a small Econom ic and Low Cost Housing in Argen ti na, amount of lime. A suit able mate ria l for the under coat and Brazil, Hon du ras, Ma lay sia, Mex ico, Pan ama, the fin ish ing coat is lime-cement mor tar with equal parts Spain, Yu go sla via.” In ter na tional Con fer ence (by weight) lime and cem ent, or with more lime than ce- on the Use of Prefab ri cate d Building El ements , ment. (See also Anon. 1985 and Anon. 1990). Con struc tions in De veloping Coun tries. World Internal plastering As so ci a tion for El e ment-building and Pre fab ri ca- Inter nal plaster ing should be done on air dry slabs. The tion, Sep.19–22, Ham burg, Germ any. plas ter ing ma te rial can be ce ment mor tar, lime-cement 1982 “Climatex Sys tem for econom ic hous ing pro- mor tar or gyp sum mor tar. If a lime-cement mor tar is jects.” 3rd In ter na tional Con fer ence on the Use used, a thin prepa ra ti on of the base is prefer a ble be fore of Pre fab ri cated Build ing El e ments. World the under coat (10 – 15 mm) is appli ed. Plaster ing with As so ci a tion for El e ment-building and Pre fab ri ca- gypsum mortar should be done in two or more coats tion, March 8–10, Orlando, Florida, USA. starting with a thin under coat . Hawkes, A J and D R S Cox 1992 A Small-scale Pro cess for Manu fac turi ng Wood- Vapour barri ers are often recom mended in air condi - wool/ce ment Slabs in Devel oping Countri es. tioned buildings to im prove tight ness and avoid Bul le tin 49. Nat u ral Re sources In sti tute, moisture transport that could dam age or ganic mate - Chat ham, United Kingdom . rials . Woodwool slabs are moisture resis tant, so there is no need for a vapour barrie r in these appli - Kohler, R cati ons. Finishe d woodwool slabs meet the require - 1966 “La Fabricación de la Tab la de Pajilla de Madera ment for air tight ness. y Cemento.” Boletín – Instituto Forestal La tino- americano de Investigación y Capacitación. No. 20–21: pp. 5–19. Mérida, Vene zuel a. (In Spanish)

24 Building Issues 1994 Volume 6 • Number 3

Kollmann, F Addresses 1955 Technologie des Holzes und der Holzwerkstoffe, 2. Aufl. II Bd., pp. 468–489, Springer-Verlag, Manufacturers of equipment Germ any. (In Germ an) for woodwool slab production Kumar, S Elten Sys tems 1980 Devel opment of Wood-wool Boards from P.O. Box 15 In dig e nous For est Ma te rials. Fi nal Re search NL-3770 AA BARNEVELD, The Nether lands Report, PL-480 Project . Forest Resea rch Insti - tute, Dehra Dun (U.P), India. Manufacturers of wood shredding machines Lee, A W C MECCAT 1991 “The Lat est Devel op m ents in the Ce ment-bonded Zona Industriale Wood Ex cel sior (Wood wool) Board Indus try .” Via Braille 5 Proc. Inor ganic Bonded Wood and Fiber Com- I-391 00 BOLZANO, It aly pos ite Ma te rials. For est Prod ucts Re search So ci ety, Mad i son, Wis con sin, USA. Some Manufacturers of woodwool slabs Simatupang, M H, G H Schwarz and F W Bröker Österreichische Heraklith AG 1979 “Small Scale Plants for the Man u fac ture of A-9702 FERNDORF, Austria Min eral-bonded Wood Com pos ites.” Spe cial (Man u fac turer of magnesite-bonded wood wool slabs. pa per FID-II/21-3, Eighth World Forest ry Factori es in Austri a, Germany and Greece) Con gress, Oc to ber 1978, Ja karta, In do ne sia. S.A. DHENACLITE 35, rue Emile Claus B-8798 SINT-ELOOIS-VIJVE, Belgium Climatex Indústria de Ma deira Mineralizada Ltda. Caixa Postal 7054 91130-040 PORTO ALEGRE – RS, Brazil FIBRALITH GIE Zone Industrielle Conversion factors F-68190 UNGERSHEIM, France from SI units to British and US units E. Schwenk Dämmtechnik GmbH & Co. KG 1 m = 3.281 ft Isotexstraße 1 1 mm = 3.937 ´ 10 –2 in D-86899 LANDSBERG/LECH, Germ any 2 2 1 m = 10.76 ft Nederlandse Bouwplaten 1 hect are en Isolatiematerialen Industrie BV = 10,000 m2 = 2.471 acres P.O. Box 375 1 m3 = 35.31 ft3 NL-4900 AJ OOSTERHOUT, The Neth erlands 1 kg = 2.205 lb 1 tonnes Träolit AB = 1,000 kg = 1.102 short tons = 0.984 long tons Box 20 1 kg/m3 = 6.243 ´ 10 –2 pcf (lb/ft3) S-570 60 ÖSTERBYMO, Sweden 1 N = 0.2248 lb f Torvale Building Products 1 Nm = 8.850 in lb f Pembridge 1 MPa LEOMINSTER = 1 N/mm2 = 145.0 psi (lb f/in2) Herefordshire HR6 9LA, United Kingdom 1 J = 9.485 ´ 10 –4 Btu 1 kWh = 3.414 ´ 10 3 Btu Organizations with information 1 W = 1 J/s = 3.414 Btu/hr about woodwool slabs 1 kW = 1.341 hp Bundesverband der Leichtbauplattenindustrie e.V. TK = T°C + 273 = 5/9 (T°F – 32) + 273 1 W/mK = 6.938 Btu in/ft2 hr °F Beethovenstr. 8 1 W/m2K = 0.1762 Btu/ft2 hr °F D-80336 MÜNCHEN, Germ any 1 J/kgK = 2.390 ´ 10 –4 Btu/lb °F (Trade asso ci a ti on for the 14 woodwool slab man u fac tur ers in Ger many) from SI units to CGS units For est Re search In sti tute 1 N = 0.102 kp P.O. New Forest 1 MPa = 10.2 kp/cm2 DEHRA DUN 248 006 1 W/mK = 0.860 kcal/m h°C Uttar Pradesh, India 1 J/kgK = 2.39 ´ 10 –4 kcal/kg°C

25 Volume 6 • Number 3 Building Issues 1994

Appendix

Bo tan i cal name Suit abil ity Bo tan i cal name Suit abil ity Bo tan i cal name Suit abil ity Abies pindrow* s Garcinia sp. (“manggis hutan”) n Pinus kesiya s Agathis borneensis Warb n Gonystilus brunescens A. Shaw. n Pinus khasya s Ai lan thus malabarica DC. n Gossweilerodendron balsamiferum Harms. s Pinus mercusii s Albizia falcataria s Grevillea robusta s Pinus nigra* s Albizia lebbek s Haplolobus celebicus H.J.L. s Pinus patula s Anisoptera costata Korth. s Holoptelia integrifolia n Pinus roxburghii* s Anisoptera marginata Korth. n Hopea dryobalanoides Miq. s Pinus sylvestris* s Anogeissus latifolia n Hymenodictyon excelsum s Pinus taiwanensis Hay. s Arau ca ria arau ca ria* s Irwingia malayana Olive n Pinus wallichiana* s Azadirachta indica n Koompasia excelsa Tamb. n Polyathia hypoleuca HK. & TH n Bombax cieba s Koompasia malacencis Maing n Populus deltoides s Bridelia retusa s Koordersiodendron pinnatum Meer s Populus tremula s Calophyllum inophyllum s Lannea coromandelica n Pterocarpus indica n Calophyllum soulatri Burn. f. s Larix leptolepis s Pterospermum celebicum Miq. s Cassia siamea r Licania laxiflora n Quercus alba L. n Cedrela toona s Liquid ambar styraciflua L. n Quercus falcata Michx. n Cedrus deodara s Lirio den dron tulipifera L. s Salmalia malabarica s Cinnamomum seylanicum n Maesopsis eminii n Sandoricum indicum n Cordia myxa s Mangifera foetida Lour n Santiria laevigata Bl. n Cratoxylon sp. (“geronggang”)* s Mangifera indica s Shorea bracteolata Dyer n Cunninghamia lanceolata Hook. s Mangifera minor Bl. n Shorea elliptica Burck. n Dacryodes excelsa s Melanorhoea wallichii Hook. n Shorea gibbosa Brandis n Dalbergia sisoo s Miristica lowiana King n Shorea gysbertsiana Burck. s Dehassia caesia Bl. s Mora excelsa n Shorea hopeifolia Sym n Delonix regia s Morus sp. (“shehtoot”) n Shorea koordersii Brandis n Dialium platysepalum Baker. n Octomeles sumatrana Miq. n Shorea leprosula Miq. n Diospyros macrophyla Bl. s Ougeinia oojeinensis n Shorea ovalis Bl. s Dipterocarpus sp. (“gurjan”) s Palaquium ferox H.J.L. n Shorea palembanica Miq. s Dipterocarpus appendiculatus Scheff. n Palaquium hexandrum Baill n Shorea pauciflora King. s Dipterocarpus condiferus Merr. n Palaquium obovatum s Shorea pinanga Scheff. n Dipterocarpus costulatus V. Sl. n Palaquium obtusifolium Burck s Shorea sp. (“Meranti”) r Dracontomelon dao Meer & Rolf. n Palaquium rostratum Burck n Simaruba amara r Dracontomelon mangiferum Bl. n Parastemon versteeghii Meer & Perry n Spondias cytherea Sonn. n Drypetes longifolia Pax & Hoffm. n Parinari corymbosa Miq. n Swietenia macrophylla n Durio zibethinus Merr. r Payena leerii Kurz n Syzygium cumini s Emblica officinalis n Pericopsis elata n Tarrietia javanica Bl. n Eperus falcata n Picea abies* s Tectona grandis r Eschweilera sp. n Picea smithiana* s Terminalia spread n Eu ca lyp tus globulus* s Pinaceae sp.* s Tetrameles nudiflora* s Eu ca lyp tus grandis r Pinus sp. (“southern pine”)* s Toona ciliata s Ficus sp. (“gular”) s Pinus caribaea s Tsuga chinensis Pritz. r Ficus sp. (“bar”) s Pinus densiflora s Xylopia malayana HK. f. & TH n Ganua motleyana Pi erre n Pinus elliottii* s Table A1. Woods tested by production of full-scale slabs. s suit able A wood is considered suitable if the slab meets the requirements n not suitable for bending strength of DIN 1101 or an equivalent standard. r re stric tive suit abil ity (Information from different sources. Information about other woods can be obtained from Elten Systems, the Netherlands.) * This wood spe cies is used com mercially for woodwool slab produc tion

Bo tan i cal name Suit abil ity Bo tan i cal name Suit abil ity Bo tan i cal name Suit abil ity Afzelia bipindensis n Daniellia ogea r Nauclea diderrichii n Antiaris africana n Distemonanthus benthamianus n Nesogordonia papaverifera r Antrocaryon micraster n Entandrophragma angolensis s Ongokea gore n Berlinia grandiflora s Entandrophragma cylindricum r Piptadeniastrum africanum n Canarium schweinfurthii r Entandrophragma utile s Pterygota macro carpa n Cedrela odorata s Euca l yptus camaldulensis r Tarrietia utilis r Ceiba pentandra n Euca l yptus gomphocephala s Tectona grandis r Celtis zenkeri n Guarea cedrata n Terminalia ivorensis r Chlorophora excelsa n Khaya sp. (“khaya, mahog any”) n Terminalia superba s Chrysophyllum africanum s Lovoa trichilioides s Tieghemella heckelii n Chrysophyllum albidum s Mansonia altissima r Triplochiton scleroxylon r Cola gigantea r Mitragyna stipulosa r Cylicodiscus gabunensis s Musanga cecropioides s Table A2. Woods tested by production of small test slabs. s suitable The wood is considered suitable if it can be shredded easily n not suit able and if the slabs set satisfactorily. If these tests are promising, r re stric tive suitability then full-scale tests are done and the slabs are tested for strength. (Information from different sources.)

26