T T r PROPERTIES OF CHIPBOARD. IC/92/433 INTERNAL REPORT (Limited Distribution) ABSTRACT

Transcription

1 V Internatinal Atmic Energy Agency and / United Natins Educatinal Scientific and Cultural rganizatin INTERNATINAL CENTRE FR THERETICAL PHYSICS PRPERTIES F CHIPBARD IC/92/433 INTERNAL REPRT (Limited Distributin) ABSTRACT Density, permeability, prsity, water absrptin and swelling characteristics have been assessed fr chipbard made ut f saw dust and an adhesive. Results indicated that permeability increases as prsity increases and bth decrease with increase in bard density. Permeability, and prsity, p were fund t relate t the density, p as fllws: and - = At af> a p = e tp while prsity and permeability were fund t fllw the relatin: P = C(-) d Edward A. Baryeh Internatinal Centre fr Theretical Physics, Trieste, Italy and ESIE, BP 311, Bingerville, Ivry Cast. ' MIRAMARE - TRIESTE March 1993 It was als fund that the lwer the bard density, the higher its water absrptin and swelling. Thermal cnductivity, specific heat and calrific value were als assessed. Results revealed that thermal cnductivity increases nnlinearly with increasing density, while it decreases linearly with increasing prsity. Specific heat als increases linearly with increasing density and temperature. Hwever, calrific value was fund t be independent f density. Mechanical prperties increase as the density increases except fr the penetratin depth. It was fund that the tensile strength, mdulus f rupture, shear strength, hardness, resistance t cutting and impact strength vary linearly with the density, while the degree f elasticity has a nnlinear relatin with the density. The bard is generally strnger in bending than in cmpressin and tensin in descending rder f strength. Permanent address T T r

2 use f particle bards in exterir applicatins in the U.S.A. and Eurpe fr example, range frm marine applicatins t sheathing and pallet decks (2). The studies here are designed t evaluate the physical, thermal and mechanical prperties f chipbard. The bards used fr the studies were made ut f dum tree saw dust. dum is a famus timber species in Ghana. 2. Experimentals: 1. Intrductin: chipbard is made frm sawdust which is a byprduct f saw mills in Ghana. It is made by mixing saw dust f 8% misture cntent (wb) with 6% (slid basis) liquid phenl frmaldehyde (1) and 1% (slid basis) wax and pressed at a temperature f 170 c C at a pressure depending n the density desired. Different types f bards with different engineering prperties can be frmed by varying the type and quantity f resin and wax, the manufacturing temperature and pressure and the particle size. It is used extensively in Ghana in the building industry fr making the ceilings f buildings and fr panneling walls f rms. The physical, thermal and mechanical prperties f the bard are unknwn. As a result, ther uses have nt been explited, when the prperties are knwn, it may be pssible t explit it fr furniture manufacture, making cabinets and clsets, thermal insulatin applicatins, use as a heat surce, filtering air in industries and making crks fr bttles. The 2.1 Cnditining f specimens: In determining the varius prperties, the misture cntent f the chipbard samples were kept lw and cnstant by hlding them in a chamber which was aerated at a rm temperature f 28 t 30 C, which is the average temperature at the lcatin in Ghana where the investigatins were cnducted. Mrever, the test specimens were ven dried fr 10 hurs at 90 C befre each test t ensure cnstant misture cntent. Part 1: Physical Prperties Density: Since chipbard is a prus bdy, it is likely that its physical, thermal and mechanical prperties depend n its density. Fr prus and adsrbent materials, density values are usually reprted as bulk density (3). Cnditined chipbard

3 r specimens were therefre weighed and their dimensins measured p = internal resistance cefficient t assess their vlumes and t calculate their densities Permeability: Chipbard may be apprximated t cnslidated prus media which cmprises a highly cmplex netwrk f channels. Sme f the tiny pres in the bard are intercnnected (accessible t air flw frm bth ends f the pres), sme have dead end (cnnected t the utside f the bard nly frm ne end) and sme are islated (inaccessible t external air flw). neither straight nr f cnstant diameter. The pres are If a difference in abslute pressure is established acrss a chipbard plate, a hydrdynamic flw f air thrugh the bard will ccur (4). in such a case, transitin frm laminar flw t turbulent flw will be gradual allwing fr the use f the Frchheimer general flw equatin (5) given belw: Where Pi - P; L 9 PP.vf 3 p ; = abslute upstream pressure p 2 - abslute dwnstream pressure L = thickness f chipbard V a - superficial velcity f air (based n ttal crsssectin) p. = density f air p g a - viscsity f air = dimensinal cnstant = viscus resistance cefficient The term invlving V. in equatin (1-1) is negligible fr viscus flw and hence can be neglected. Equatin (1-1) then results in the Darcy equatin f flw thrugh prus media: Pi - (1-2) The quantity ' is referred t as the permeability cefficient defined by Muskat (6) as the vlume f fluid f unit viscsity passing thrugh a unit crss-sectin f a medium in unit time under unit pressure gradient. The apparatus in figure 1-1 was cnstructed fr the permeability tests. Air was blwn by the fan at different pressures, filtered and passed thrugh the dip tube t stabilise the pressure and keep it cnstant. Air frm the dip tube was dried by the silica gel and passed int the air bx, thrugh the chipbard specimen and ut f the ther end f the air bx. manmeter measured the pressure difference, while the pitt static tube measured the velcity f the air (7). The specimens were 100 mm diameter and 5 mm thick. The The cpper-cnstantan thermcuples measured the air temperatures. Air flw was cntrlled by valves. Readings were taken under steady state cnditins t ensure unifrm pressure differential. The air temperatures facilitated the readings f the viscsities f air. The air flw was viscus, laminar and isthermal. The chipbard acted as an undefrmable, hmgeneus, prus medium with cnstant misture cntent during the tests. The tests were

4 cnducted fr different chipbard densities and fr five times at each density Prsity: The dimensins f cnditined test specimens were measured t evaluate their uncmpressed vlumes. The specimens were then cmpressed withut changing their lateral dimensins until they were practically incmpressible. The cmpressed dimensins were measured t evaluate the cmpressed vlume. The prsity, 7 was then evaluated using the fllwing frmula: at a depth f 50 mm. The specimens were then taken ut and suspended fr five minutes t drain them f excess water. They were weighed and their thicknesses were measured at the same five different, pints. The thickness measurements were averaged. The prcedure was repeated fr immersin times f 10 hurs and 20 hurs. The quantity f water absrbed, Q was expressed as a percentage r fractin f the specimen dry weight, while the increase in thickness r swelling, S was expressed as a percentage r fractin f the riginal thickness as fllws: (V - V )100% (1-3) Where V u = uncmpressed specimen vlume V s = cmpressed specimen vlume Thus the difference in the tw vlumes expressed as a percentage r fractin f the riginal r uncmpressed vlume was used as a measure f prsity Water absrptin and swelling: Absrptin is the penetratin f a substance int the bdy f anther. The absrptin capacity f a substance relates t the amunt f liquid it can absrb, while the swelling has t d with dimensinal changes due t liquid absrptin. The thickness f cnditined specimens were measured at five different pints with a mcrmeter and averaged. The specimens were weighed and immersed in water at a rm temperature f 28 C fr ne hur. The specimens were held hrizntally in the water Where and Q = (W ; - W 1 )100% w, S = (X, - X 1 )1% x, V 1 = specimen weight befre immersin W 2 = specimen weight after immersin X x specimen thickness befre immersin X 2 = specimen thickness after immersin 1-3. Results and discussin: ( 1-4; ( 1-5) The chipbards with high densities were fund t have a better surface finish r were smther n the surface than thse with lw densities. Frm equatin 1-2, it is seen that the slpe f the graph f (p 1 - p 2 ) against V a yields the viscus resistance cefficient, a. Three such graphs fr different densities are shwn in figure 1-2. The figure indicates that fr a given density, the pressure

5 T difference increases linearly with the air velcity. Als fr reduces the size and number f air pres thereby reducing the a given air velcity, the pressure difference increases with the permeability and prsity. The graphs crss each ther at a density. n the cntrary, fr a given pressure difference, the density f 287 kg/m 3. Frm the crssing pint, the gap between air velcity increases as the density decreases. the curves gets wider as the density increases. The freging Frm graphs such as thse in figure 1-2, the permeability cefficient. I were calculated and the values btained pltted ( against the prsity f the bard. The resulting graph is shwn in figure 1-3. The figure indicates that the prsity increases as the permeability increases but the relatin is nnlinear. zer prsity, the permeability is practically zer. At Frm the trend f the graph, hwever, it seems the curve will be assympttic t the 100% prsity line as the permeability increases. suggests that in applicatins in which air flw thrugh the bard is undesirable, it is advisable t use high density bards, while if air flw is desirable, as in a filtratin prcess, lwer density bards may be mre desirable. Bth the permeability and prsity yield straight line graphs when pltted against the density n linear-lg r semi-lg graph sheet as displayed in figure 1-5. This reveals that the permeability, * and prsity, p have the fllwing mathematical relatinship with the density, p: Since the viscus resistance cefficient, a varies with the density (see figure 1-2), the permeability and prsity were bth pltted against the density as shwn in figure 1-4. The figure Ae-* 1 a = e *> A ( 1-6) reveals that bth permeability and prsity decrease as the and p = Be""' (1-7) density increases. The relatin is nnlinear in bth cases. Shukla et al (8) als fund a similar trend with rice husk bards. The decrease is mre rapid fr permeability than prsity. Bth prperties decrease mre rapidly fr densities between 200 and 500 kg/m s cmpared t the rate f decrease between 500 and 1000 kg/nv\ The results reveal that bth permeability and prsity depend n the size and number f air pres in the bard rather than the saw dust particle size. Hence when the density is lw, the accmpanying large number and size f air pres increase bth the permeability and prsity. High fabricatin pressure, in the case f the high density bards, where A, B, a and b are cnstants which can be determined experimentally r frm the graphs r by the technique f regressin analysis. The relatin between prsity and permeability als plts linearly n lg-lg sheet as shwn in figure 1-6. This indicates that the variatin f prsity with permeability is characterised by an equatin f the frm: (1-8)

6 where C and d are cnstants which can be determined experimentally r frm the graph r by regressin analysis. Graphs f water absrptin capacity and increase in thickness versus density are displayed in figures 1-7 and 1-8 fr ne hur, 10 and 20 hurs f immersin in water. Bth the water absrptin capacity and increase in thickness decrease as the density increases. When the density is lw, the bard has mre pres and the adhesive bnd is less effective. Hence when it is immersed, these pres admit and hld water as the air in them is displaced. Alternately, the mre prus the bard is, the mre water it hlds and the mre it swells. This suggests the use f high density bards where there is likely t be misture. Bth the water absrptin capacity and the increase in thickness als increase with increase in immersin time. At a density f 400 kg/m 3 fr example, the water absrptin capacity fr 20 hurs, 10 hurs and ne hur immersin are 63, 46 and 27% respectively, while at 800 kg/m 3, the crrespnding values are 35, 23 and 15% respectively. The abve results shw that the physical prperties evaluated depend n the bard density. Gatchell et al (9) have revealed that the density f such particle bards are affected by the manufacturing temperature and pressure, and the amunt and type f adhesive, resin and wax. Hence, further studies may be cnducted t find hw such cmpnent variables affect these physical prperties fr chipbard. If the bard is waxed r painted, the water absrptin capacity and the increase in thickness reduce as Shukla et al (8) fund when they waxed rice husk bards Cnclusin: The apparatus devised has been used successfully t determine the physical prperties f chipbard. It was fund that all the physical prperties are a functin f the chipbard density. The higher the bard density the less permeable and prus it is, the less water it absrbs and the lwer the swelling r increase in thickness. Part 2: Thermal Prperties Thermal cnductivity: Thermal cnductivity f a substance is the rate f heat flw per unit area per unit temperature rise fr a unit thickness f the substance (10), and may be fund in mst treatises n heat transfer as fllws: k = (2-1) where k = thermal cnductivity f the material Q = rate f heat flw acrss the material x = thickness f the material A = surface area f heat flw Ae = temperature difference between the ht and cld material faces In these experiments, the lngitudinal and lateral dimensins f the specimens were much larger than their thickness. The

7 T T r apparatus used fr the determinatin f the thermal cnductivities f pr cnductrs (3) was adapted t determine the thermal cnductivities f the samples as fllws: 1. the mercury thermmeters were replaced by cppercnstantan thermcuples; 2. the apparatus was lagged with ccnut fruit fibre r cir (3) t reduce the heat lss. Specimen diameters f 100 mm and thickness f 5 mm were used. The thermal cnductivities were determined as a functin f the bulk density fllwing the prcedure utlined by Baryeh (3). Accrding t Brady (11), the thermal cnductivity f a prus material may increase if water is absrbed int the spaces within it. The cnditining f the samples, as utlined abve, hwever, eliminated this drawback. In the experimental prcedure, the heat lss frm the apparatus cnsidered negligible, the heat flw S = M^te - e t ) + M;S;(9-9 t ) (2-2) where S = specific heat f chipbard Sj - specific heat f calrimeter and stirrer S 2 = specific heat f paraffin Mj = mass f calrimeter and stirrer M 2 i mass f paraffin M = mass f chipbard specimen 9j = initial temperature f paraffin and calrimeter initial temperature f heated chipbard specimen 6 = equilibrium temperature f mixture The specific heat was evaluated as a functin f the bard density and als fr different ambient temperatures. was assumed t be linear due t the small specimen thickness Calrific value: The calrific value f a substance is the heat generated by Specific heat: The specific heat f a substance is the heat required t raise the temperature f unit mass f it thrugh ne degree(12). It was determined using the methd f mixtures and the apparatus used by Shukla et al ( 8), replacing the glass thermmeter with a cpper-cnstantan thermcuple. When the specimen is heated and put int paraffin and stirred, the equilibrium temperature is btained. At the equilibrium temperature, the heat released by the specimen is equal t the heat absrbed by the lagged calrimeter and paraffin resulting in the fllwing equatin: burning unit mass f the substance. This was determined by burning a weighed piece f cnditined chipbard in a bmb calrimeter (13). The calrific value, C was calculated frm the relatin: mc = IV^S. + M C S C + M b S b (9 2-9,} (2-3) where m = mass f chipbard M,, = mass f water M c = mass f calrimeter M b = mass f bmb

8 S u = specific heat f water The specific heat variatin with density and temperature are S c = specific heat f calrimeter displayed in figures 2-3 and 2-4. Bth figures indicate that the S h = specific heat f bmb specific heat increases linearly as bth density and temperature Si = initial temperature f bmb increase, and are a little higher cmpared with sme slid wds 6;, = final temperature f bmb. due t the adhesive and resins used t bind their particles (14). The specific heats are in general high als because chipbard is 2-3 Results and cmments: Figures 2-1 and 2-2 shw the thermal cnductivity variatin a pr cnductr f heat. The specific heat values range frm 1.5 J/g C fr a density f 400 kg/m 3 and a temperature f 30 C t with density and prsity. The figures shw that thermal cnductivity increases as density increases. Similar results 3.5 J/g C fr a density f 1000 kg/m 3 and a temperature f 90 C cmpared t 0.4 J/g C fr cpper, fr example, which is a gd have been btained by Shukla et al (8) fr rice husk bards and cnductr f heat. This further suggests the suitability f Baryeh (3) fr plantain stem fibre, plantain fruit stalk fibre and ccnut fruit fibre. The relatin fund in the present studies is, hwever nt linear and therefre des nt cnfrm cmpletely with the linear relatinship Shukla et al (B) fund fr rice husk bards. The variatin is, hwever very similar t chipbard fr thermal Insulatin purpses. The heat f cmbustin was fund t be cnstant at kj/kg as the density varies frm 400 t 1100 kg/m 3. Hence the calrific value des nt depend n the pres f air in the bard- This is because the apparatus fr the determinatin f the ne btained by Baryeh (3) fr the increasing prtin f the the calrific value is itself filled with a cpius supply f air graphs fr the thermal cnductivities f plantain fruit stalk which is used up in the chipbard cmbustin prcess. The fibre, plantain stem fibre and ccnut fruit fibre. The figures calrific value culd hwever depend n the misture cntent f als indicate that thermal cnductivity increases linearly as prsity decreases. When the bard is prus and has mre air in its pres thermal cnductivity is lwer because air has a lw the bard but this was nt investigated. This is thus a subject fr further investigatin. The calrific value f the bard is high enugh fr it t be used as fuel fr dmestic cking and thermal cnductivity. Accrding t Brady (11), the efficiencies f fibrus insulatin is partly due t their air spaces, and frm figures 2-1 and 2-2, it is bvius that Brady's statement is als fr sme small and medium size industrial 2-4. Cnclusin: plants. true fr chipbard. Due t the generally lw thermal The studies have revealed that the thermal cnductivity and cnductivities, chipbard can be recmmended fr use as a thermal insulatr t replace imprted insulating materials. specific heat f chipbard depend n the bard density, while the calrific value is independent f it. The thermal cnductivity

9 T increases with increase in density r decrease in prsity, while the specific heat increases with increase in bth density and ambient temperature. = X Ḣ (3-4) The direct r nrmal stress and strain result in the mdulus f Part 3 Mechanical Prperties elasticity, E while the shear stress and strain give the mdulus f rigidity, G as fllws: Theretical cnsideratins: The defrmatin f an engineering material due t applied frces depends n the applied frces, the area, A ver which the frce is applied and the length, L r height, H f the material (see figure 3-1). The applied frce, F results in nrmal r direct stress, a and shear stress, x given by: 9* (3-5) The breaking stress in tensin, a, is a = F (3-7) A (3-1) where F, is the breaking frce. X =_. A (3-2) The degree f elasticity in cmpressin, D. c may be defined frm frce-defrmatin measurements as (15): Fifure 3-l(a) illustrates nrmal r direct stress, while figure 3-l(b) shws shear stress. The applied frce defrms the material as shwn in figure 3-1. Figures 3-l(a) and 3-l(b) illustrate direct r nrmal defrmatin, e and shear defrmatin, x respectively. The defrmatin cmbine with the length, L r height, H f the material t give direct r nrmal strain, s and shear strain, <f> as fllws: (3-8) where D r = distance recvered by sample after the lad is remved D c = distance thrugh which the sample is cmpressed The degree f elasticity in tensin, D et may als be defined as (16): t (3-3) (3-9)

10 where D r is as stated abve and D e is the distance the sample is extended. Shukla et al (6) have defined the cmpressibility and retentin f cmpressin which qualitatively give similar cmparisn because the mre elastic a material is, the higher its degree f elasticity and the lwer its cmpressibility and retentin f cmpressin. Hence degree f elasticity deals with hw elastic a material is, while cmpressibility and retentin f cmpressin deal with hw inelastic a material is. When a beam with the dimensins shwn in figure 3-2 is simply supprted and is laded at mid-span, the mdulus f rupture r breaking bending stress, a b is given by: where My (3-10) I M = maximum bending mment n the beam y = distance frm the neutral axis t the uter fibre I = sectin mdulus Methdlgy: a b =.PL. _t_ 12t' 3 = 3PL 4-2 b? Tensin and cmpressin tests: (3-11) Cnditined chipbard specimens measuring 2 5 mm by 5 mm in sectin and 100 mm lng were used fr the tensile tests using the Instrn universal testing machine. The test was perfrmed n 20 samples and the samples were tested t the fracture r breaking pint. The lading rate was 0.1 mm/s. Cmpressin tests were als cnducted n 20 cnditined samples using the Instrn universal testing machine. Fr each test, fur specimens measuring 2 5 mm by 25 mm by 5 mm were put tgether t frm a cmpsite specimen measuring 25 mm by 2 5 mm in sectin and 20 mm high. The lading rate was 0.1 mm/s. Sme f the tensile test specimens were extended t 105% f their initial length using the Instrn machine in rder t assess the degree f elasticity in tensin, while sme cmpressive test specimens were cmpressed t 95% f their initial height t assess the degree f elasticity in cmpressin Bending test: Fr the bending tests, cnditined specimens measuring 25 mm by 5 mm in sectin by 150 mm lng were used. They were placed n knife edges with 25 mm verhanging at each knife edge supprt. They were laded in the middle as shwn in figure 3-2 using the Instrn machine until the specimens brke. Twenty specimens were tested fr each bard density cnsidered and the lading rate used was 0.1 mm/s Shear test: The shear tests were cnducted n samples measuring 25 mm by 25 mm by 5 mm using the equipment and prcedure described by Baryeh (16) and a lading rate f 0.1 mm/s Hardness and resistance t cutting tests: The hardness was evaluated as resistance t indentatin in the frm f pressure fr indentatin and depth f indentatin. A 10 mm diameter steel rd with a hemispherical indenting end and

11 r T a hexagnal head t fix int a hexagnal cavity in the lading head f the Instrn machine, served as the indenter (see figure 3-3). Test specimens were bnded tgehter t frm cmpsite specimens measuring 100 mm by 100 mm by 30 mm thick. The specimens were indented using a frce f 25 N and a lading rate f 0.1 mm/s fr 10 s. The depth f indentatin was nted as the hardness (16). ther tests were cnducted in which the indenter was made t penetrate 5 mm int the specimen. The maximum frce divided by the prjected area caused by the hemispherical indenting end was evaluated t be the hardness. The resistance t cutting was evaluated as utlined by Baryeh (16) using samples measuring 25 mm by 25 mm by 120 mm and cutting acrss the 25 mm width Impact test: Charpy impact tests were perfrmed as utlined by Warnck and Benham (17). The specimens measured 10 mm by 10 mm by 80 mm. A Charpy testing machine type CPSA (range: 0 t 4 J) manufactured by Amsler tt Wlpert-Werke f Germany was used fr the tests. 3-3 Results and discussin: Stress-strain curves frm the tests are shwn in figures 3-4(a) and 3-4(b) fr different densities. It is clear frm the figure that the higher the density the higher the stress fr a given strain, and the higher the density the lwer the strain fr a given stress. The stress-strain relatin is nnlinear. Hence it is impssible t use the graph t find the mdulus f elasticity f the bard withut a knwledge f the equatin f the curve. The graphs shw that the mdulus f elasticity depends n the density. Apprximating the curves t straight lines, the apprximate mdulus f elasticity fr densities f 400, 600 and 800 kg/m 3 are 26.67xl 5, 60xl 5 and Bxl 6 respectively. This variatin is due t the air pres and the resin and wax which bind the particles tgether. Thus, the air pres and resin make the different density bards behave as different materials. At the breaking pint, the bnd between the bard particles breaks up making the graphs in figure 3-4(a) axhibit a remarkable drp in stress fr little r n inr rease in strain. It can be seen frm the figures that the cmpressive Strength is abut 1.25 times the tensile strength fr a given density. Hence the bard is strnger in cmpressin than in tensin. The test specimens were cnsidered t be at the rupture r breaking pint when cracks started appearing n them. At this pint, the resin bnd between the particles gave in. This behaviur is indicated by a sudden drp in stress fr little r n increase in strain. Figure 3-5 shws the tensile strength variatin with density. The figure indicates that the tensile strength increases linearly with increasing density. This trend is due t the larger number f pres and less adhesive and wax in the lwer density bards. These pres ffer abslutely n strength t the bard. Hence, as the pres decrease, the bard strength increases. Accrding t Marra (18) and Gatchell et al (9), the resin r edhesive in particle bards gives superir bnds between particles at higher densities than at lw densities, resulting In imprved bard strength

12 The degree f elasticity is displayed as a functin f density in figure 3-6. The definitin given in equatins 3-8 and 3-9 are bth pltted. The graphs reveal that the degree f elasticity increases nnlinearly with density. Hence as the bard gets denser, it becmes mre elastic thugh the degree f elasticity is very lw cmpared t that f metal- Shukla et al (8) hwever, wrking with rice husk bards, fund a linear relatinship fr bth cmpressibility and retentin f cmpressin variatin with density. Equatin 3-8 gives higher degree f elasticity values than equatin 3-9. The gap between the graphs fr the tw equatins increases as the density increases. Fr example, at 400, 800 and 1000 kg/m 3, the difference in the degree f elasticity fr the tw equatins are 0.005, and respectively. This culd be due t the difference in the bard strength in cmpressin and tensin. This is nt in agreement with the results fund fr ccyam by Baryeh (16) in which the degree f elasticity was fund t be higher in the tensile situatin than in the cmpressive situatin. Figure 3-7 depicts the mdulus f rapture in bending as a functin f density. The mdulus f rupture increases linearly with density fr the reasns given already fr the tensile and cmpressive strengths. Cmparisn f the mdulus f rupture and tensile strength results reveals that the mdulus f rupture fr a given density is abut twice the tensile strength. This indicates that the bard is strnger in bending than in tensin. A plt f the shear strength against density appears in figure 3-8- Here als, shear strength increases linearly with density mainly fr the same reasns given fr tensile strength, cmpressive strength and mdulus f rigidity. The shear strength values are in general lwer than the tensile and cmpressive strengths and the mdulus f rupture in bending. This indicates that the bard is strnger in tensin, cmpressin and bending than in shear. Hence the bard may nt be used in applicatins in which shear is the prevailing mde f stress. The shear strength variatin with density indicates further that the mdulus f rigidity f the bard varies with density. The relatinship between the hardness and density is displayed in figure 3-9. The hardness expressed as depth f penetratin decreases linearly as the density increases. The reasn fr this trend is that at high densities, the bard has less vids and therefre ffers mre resistance t penetratin than at lw densities. This cnsequently results in lwer depths f penetratin and vice versa. The hardness expressed as frce divided by the prjected area caused by the hemispherical indenter, hwever increases linearly with increasing density as shwn in figure 3-9. This is because at high densities, higher frces are needed t make the indenter penetrate t the same depth cmpared t lw densities. Hence the higher the density, the higher the required frce and vice versa. This later statement has als been made by Shukla et al (8) fr rice husk bards. Resistance t cutting is, in a way, a measure f hardness t because, in general, the harder a material, the lnger it takes t cut it. The variatin f the resistance t cutting with density is shwn in figure The figure depicts that the

13 r T T higher the density, the higher the time needed t cut it and vice versa. The reasns fr this trend are the same as thse given earlier fr the tensile and cmpressive strengths trends. The resistance t cutting is als smewhat similar t the screw and nail pulling resistance test cnducted by Shukla et al (B) n rice husk bards in which they cncluded that the higher the density, the higher the frce needed t pull ut a nail r a screw frm them. The impact strength f the bard als increases linearly with density fr reasns already given abve (see figure 3-1), The impact strengths are generally lw because the resin and wax binding the particles cannt stand impact lads. Therefre chipbard is nt recmmendable fr impact and shck lad applicatins. applicatins in which shear and impact frces are high. These mechanical prperties are likely t depend n the type and quantity f resin and wax, misture cntent f the bard and pressing pressure and temperature during manufacture. These were nt, hwever investigated and therefre can frm a tpic fr further investigatin, ACKNWLEDGMENTS The authr wuld like t thank Prfessr Abdus Salam, the Internatinal Atmic Energy Agency and UNESC fr hspitality at the Internatinal Centre fr Theretical Physics, Trieste. The mechanical prperties studied are likely t vary with the resin and wax cntent and the pressing pressure and temperature used in the bard manufacture as was fund by Gatchell et al (9) fr particle bards made ut f Duglas-fir flakes. These aspects were nt cvered in this study. This is therefre an area fr further study Cnclusin The tests perfrmed have revealed that all the mechanical prperties studied are linear functins f the bard density except the degree f elasticity. All the prperties increase as the density increases except the depth f penetratin. Fr applicatins in which strength is imprtant, it is advisable t use bards f high density. The bard is nt recmmendable fr

14 11. Brady, G.S. Materials Handbk, 10th Editin, McGraw Hill References: Bk Cmpany, New Yrk, Abbt, A.P. rdinary Level Physics, 4th Editin, Heinemann 1. Shields, J. Adhesives Handbk, 3rd Editin, Butterwrths, Lndn, Heebink, B.G. and Hann, R.A. Hw wax and particle shape affect stability and strength f ak particle bards. Frest Prduct Jurnal, Vl. 9, N. 7, pp 197, Baryeh, E.A. Thermal cnductivities f sme West African fibres. The Agric. Engineer, Vl. 40, N. 1, pp 10-14, Treyball, R.E. Mass Transfer peratins, 3rd Editin. McGraw Hill Bk Cmpany, New Yrk, Perry, R.H. and Green, D. Perry's Chemical Engineers' Handbk, 6th Editin, McGraw Hill Bk Cmpany, New Yrk, Muskat, M. Physical Principles f il Prductin. McGraw Educatinal Bks, Lndn, Rgers. G.F.C. and Mayhew, Y.R. Engineering Thermdynamics- Wrk and Heat Transfer, Lngmans, Lndn, Marks, L.S. Mechanical Engineering Handbk, 4th Editin, McGraw Hill Bk Cmpany, New Yrk, Mhsenin, N.N. Physical Prperties f Plant and Animal Material, Vl. 1, Grdn and Breach, New Yrk, Baryeh, E.A. Rhelgical prperties f ccyam. The Agric. Engineer, Vl. 45, N. 4, pp 92-98, Warnck, F.V. and Benham, P.P. Mechanics f Slids and Strength f Materials. Pitman Paperbacks, Lndn, Marra, G.G. Binders fr particlebards. US Frest Prduct Lab. Reprt N. 2183, Madisn, Hill Bk Cmpany, New Yrk, Streeter, V.L. and Wylie, E.B. Fluid Mechanics, 6th Editin, McGraw Hill Bk Cmpany, New Yrk, Shukla, B.D., jha, T.P. and Gupta, C.P. Measurement f prperties f rice husk bards. AMA, Vl. 16, N. 2, pp 53-60, Gatchell, C.J., Heebink, B.G. and Hefty, F.V. Influence f cmpnent variables n prperties f particle bard fr exterir use. Frest Prduct Jurnal, Vl. 16, N. 4, pp 46-59, Reynlds, W.C. and Perkins, H.C. Engineering Thermdynamics. McGraw Hill Bk Cmpany, New Yrk,

15 CD «6 9=800 CD a e.a a QJ u d 0) f CD w- «^ dj r Z) tn (XI =600 = Velcity (m/s) FIGURE i-2. Pressure difference against air velcity r

16 m- I If U I V WmmBmm^SuSm m urn m.-< i i i i H Permeability* 10" l3 (m 2 ) jr 60 CL v\. prsity 40 permeability 20k Density (kg/m 3 ) FIGURE,'RE i-3. Prsity against permeability FIGUEE 1-4. Permeability and prsity density againsf

17 Permeability x10" l3 ( CL fd 13 V) *~.t" 10 rt> 3 i ' XI r 3 r D CT ^r ic: D CL (X) ZJ TD V) Ui </) prsity (%) Prsity (%)

18 ! i i Water absrptin (%) K) Q- p TD n P n 13 (/I fd in N 3 C Increase in thickness (%) I h 8 T" UJ f LTI 8 ^-' P n 3 10 p S 8

19 .20 y t.12 > u f.08 a.04 u a e Density (kg/m ) Prsity (%) Fig 2-1 Variatin f thermal cnductivity with density Fig 2-2 Variatin f thermal cnductivity with prsity T r~ T~

20 r t CT 1 a QJ U CD Q_ U QJ Q J L Density (kg/m 3 ) Temperature ( C) Fig 2-3 Specific heat variatin with densi ttj Fig 2-4 Specific heat variatin with temperature

21 riqinal psitin... psitin Final psitin J ^ r * ': A-' ' >--J' (b) Figure 2>~\ Nrmal and shear stress and stratn 27 H c IP ( Q_ i I I. if si: 3 VI c re «n p 3

22 ?= 800 kg/m 3 7=600 kg/m kg/m kg/m 3 9= 4Q0 kg/ m 3 x V) v> 0/ a 0?= 400 kg/m Strain x 10" Figure 3~4-( a ) Tensile stress-strain curves (9 =density) Strain x 10 " Figure 3»~4-(b) Cmpressive stress strain curves (9--densityJ

23 Tensile strength x 10 (kn/m z ) r C I 5 in CD t m CP I Degree f elasticity a 2L ID 3 Ch a a a. hi

24 Mdulus f rupture x 10 (kn/m 2 ) CD c ID fd ^ CD "1 C3 3 CD ex. r Z3 VI 200 c fd (J 1 C 200 Shear strength x 10 (kn/m 2 ) Shear Q 009 a fversus a. m nsit Densiti <a 800 a IS]

25 15 r 30 X VI 10 (LI I 5 (kn/m z ) -i 10 6 i in r a 25.E f LJ I 10 I/I \n a> Density (kg/m 3 ) &00 Density {kg/m 3! 1200 Figure 3>~9 Hardness against density Figure 3-t Resistance t cutting against density T r" T

26 J CL &0& Density (kq/m 3 ) 1200 Figure ^-ft Impact strength variatin with density