WOOD Research @ UPM

50 %
50 %
Information about WOOD Research @ UPM
Education

Published on March 11, 2014

Author: drsidekaziz

Source: slideshare.net

Description

Collection of working papers and journals on wood research, conducted at Universiti Putra Malaysia

Physical | Elasticity | Dielectric S i d e k A b A z i z

Pertanika. J. Sci. & Techno!. 3(2): 389-399 (1995) ISSN:0128-7680 © Penerbit Universiti Pertanian Malaysia COMMUNICATION II Penggunaan Kaedah Dinamik Ultrasonik bagi Menentukan Pemalar Kenya! Kayu Tropika Diterima 7 May 1994 ABSTRAK Kertas ini melaporkan penentuan pemalar kenyal 56 spesis kayu-kayan tropika menggunakan teknik dinamik ultrasonik. Gelombang membujur (45 kHz) digunakan untuk mengukur halaju pada arah ketakisotropan jejarian (R), tangen (T) dan longitud (L) bagi setiap sampel kayu. Pemalar kenyal bagi setiap spesis kayu ditentukan dengan menggunakan nilai halaju ultrasonik dan ketumpatan menerusi persamaart Christoffel. Hasil k~ian menunjukkan terdapat satu korelasi yang linear diantara pemalar kenyal dan ketumpatan kayu bagi arah L, R dan T. Perbandingan hasil pengukuran ulrasonik kepada hasil pengukuran statik menunjukkan teknik ultrasonik juga mampu digunakan untuk menilai kualiti sampel-sampel kayu tropika jenis sederhana dan keras. ABSTRACT This paper reports the determination of elastic constants for 56 tropical wood species using the dynamic ultrasonic technique. Longitudinal waves (45 kHz) were used to measure the velocity in the radial (R), tangential (T) and longitudinal (L) anisotropic directions of each sample of wood. The diagonal elastic constants of each species of wood studied were estimated using the values of ultrasonic wave velocities and their mean densities via the Christoffel equations. The results show that there is a linear correlation between the elastic constants in the L, Rand T directions and density of wood. Comparison between the ultrasonic measurement and static measurement indicate that the ultrasonic measurement technique is also capable of assessing the quality of tropical medium and heavy hardwood. PENGENAlAN Teknik ujian memusnah biasa diguna bagi mendapatkan sifat fizikal bahan yang boleh dikaitkan dengan kualiti kayu (Szymani and McDonald 1981; Bucur 1985). Kaedah yang berkesan, kos yang rendah dan kepentingan kajian merupakan faktor-faktor yang harus dipertimbangkan bagi mendapatkan hasil pengukuran yang memuaskan. Teknik ujian memusnah dikatakan kurang efektif kerana memerlukan banyak sampel kajian di samping kos penyelenggaraan yang tinggi dan hanya dapat dilakukan di makmal sahaja. Teknik ujian tak memusnah ultrasonik merupakan satu alternatif bagi tujuan pengujian, pemiawaian dan kawalan mutu kepada bahan kayu (Bucur 1983; Bucur 1985; Bucur and Rocaboy 1988). Teknik

Sidek Hj. Abdul Aziz, Abdul Halim Shaari dan Chow Sai Pew ini tidak memerlukan banyak sampel kerana sampel kajian tidak mengalami kemusnahan dan sampel yang sarna dapat diuji semula bagi mendapatkan parameter fizik yang lain. Di samping itu ia berupaya untuk memberikan data-data fizik dalam masa yang singkat di makmal atau ujian seeara in situ. TEDRI DAN TEKNIK PENGUKURAN Kayu adalah bahan ortotropik kenyal dengan eiri kimia dan struktur binaannya yang kompleks serta berbeza pada arah longitud (L), tangen (T) dan jejarian (R) seperti yang ditunjukkan seeara skematik pada Rajah 1. Kayu menunjukkan sifat kenya! dan berkemampuan mengatasi sebarang tegasan luar seperti ketumpatan dan ketegangan sehingga ke had kenyalnya; ia mematuhi hukum Hooke iaitu dengan Cijkl adalah pemalar kenyal bahan uj dan ek1 masing-masing mewakili komponen tegasan dan terikan (Bueur 1983; Bueur 1985). Model ortotropik Cartesian digunakan bagi memudahkan kajian perambatan gelombang ultrasonik bagi meneirikan aspek-aspek kekenyalan bahan kayu. Data-data pemalar kenyal kayu ini sangat penting bagi menganggarkan atau menentukan pemalar-pemalar teknikal yang digunakan olehjurutera bahan. Arah L Arah T Arah R 1 390 Rajah 1. Arah rujukan bagi sampel kayu Tatanda L mewakili longitud (arah pemanjangan), R mewakili jejarian (arah pembesaran cecincin) dan T merujuk kepada tangen terhadap arah pembesaran cecincin). Pertanika J. Sci. & Technol. Vol. 3 No.2, 1995

Penggunaan Kaedah Dinamik Ultrasonik bagi Menentukan Pemalar Kenya! Kayu Tropika Pemalar kenyal bagi komponen normal (diagonal) iaitu Cll , CRR dan Crr bagi bahan kayu dapat diperolehi menerusi data masa perambatan halaju gelombang ultrasonik yang merambat dalam arah L, R dan T. Indeks pertama dan kedua masing-masing merujuk kepada arah perambatan gelombang ultasonik dan arah sesaran zarah-zarah. Pemalar kenyal Cj . dapat dikaitkan dengan sebutan halaju gelombang utrasonik V melalui persamaan berikut: Persamaan ini sebenarnya merupakan hasil penyelesaian persamaan Christoffel (Bucur 1985; Sidek et at. 1990) iaitu C.·kl n·n· - pV 2 8'k = 0lJ ! } 1 dengan npj merujuk kepada arah perambatan gelombang, p adalah ketumpatan sampel kayu dan 8ik adalah delta Kronecker yang bernilai 1 apabila nilai i bersamaan dengan nilai k. Matlamat utama kajian ini adalah untuk menentukan pemalar-pemalar kenyal kayu tropika menggunakan kaedah dinamik ultrasonik. Konsep asas pengukuran adalah begitu mudah dengan gelombang ultrasonik 45 kHz yang dipancarkan menerusi tranduser piezolektrik ke dalam bahan kajian. Gelombang ini akan dikesan oleh alat pengesan gelombang ultrasonik yang juga diperbuat dari bahan piezoelektrik. Bagi bahan pepejal seperi kayu, di samping gelombang longitud (membujur), terdapat juga gelombang ricih (melintang) dan gelombang Rayleigh iaitu gelombang kenyal yang merambat pada permukaan bahan. Dalam kajian ini kaedah pemancaran terus digunakan bagi pengujian bahan kayu yang mempunyai struktur butiran kasar. Frekuensi 45 kHz digunakan supaya gelombang ultrasonik dengan jarak gelombang yang panjang mampu dirambatkan di dalam sampel dan dapat melalui halangan-halangan kecil. Alat ultrasonik BPV Steinkamp (buatan German) dengan operasinya berdasarkan kaedah pemancaran mampu mengukur masa perambatan gelombang ultrasonik sehingga 999.9 f.Ls telah digunakan bagi mencirikan sifat kenyal kayu tropika. Rajah blok bagi keseluruhan set eksperimen ditunjukkan pada Rajah 2. Dua buah prob berbentuk kon dipilih disebabkan keadaan permukaan kayu yang kasar dan tidak sarna rata jika ditinjau secara mikroskopik. Dengan mengetahui jarak dan masa perambatan gelombang yang merambat dalam kayu pada arah-arah tertentu, maka nilai halaju gelombang ultrasonik diperolehi melalui sebutan V (=jarak/masa). Bagi setiap arah L, R dan T, lebih dari dua puluh nilai pengukuran halaju Pertanika J. Sci. & Techno!. Vol. 3 No.2, 1995 391

Sidek Hj. Abdul Aziz, Abdul Halim Shaari dan Chow Sai Pew dilakukan untuk setiap sampel kayu dan nilai purata halaju digunakan bagi mendapatkan nilai pemalar-pemalar kenyal. Penjana Denyut lITtrasonik Dellyuf 600V Sampel kayu Set Pengukur Masa Rambatan I 261.7J1sl ...... ..... Transduser Penerima Rajah 2. Rajah skema bagi peralatan ultrasonik BPV Steinkamp yang digunakan untuk mengukur masa perambatan gelombang ultrasonik dalam Sfl;({!J sampel kayu tropika. BAHAN KAJIAN Sampel kayu tropika yang dikaji diperolehi dari Institut Penyelidikan Perhutanan Malaysia (FRIM) Kl'pong yang dikeringkan seeara pendedahan pada udara. Kandungan kelembapan bagi sampel-sampel kayu ini dianggarkan antara 12-15%. Jadual 1 menyenaraikan sampel kayu tropika yang dikaji berdasarkan nama tempatan dan saintifik serta nilai ketumpatan purata setiap satunya. Sampel-sampel kayu tropika yang kering mempunyai dimensi 11 em (panjang) x 7 em (lebar) x 1 em (tebal) (ralat pengukuran ± 0.01 em) dan setiap satunya mempunyai arah longitud L, tangen T dan jejarian R yang merujuk kepada arah utama ketakisotropan kayu (lihat Rajah 1). Nilai pemalar kenyal normal kayu dapat ditentukan dari nilai ketumpatan dan halaju gelombang (V) ultrasonik melalui persamaan berikut dengan indeks L, R dan T mewakili arah-arah paksi utama kayu iaitu longitud, jejarian dan tangen (Rajah 1). 392 Pertanika J. Sci. & Techno!. Va!. 3 No.2, 1995

Penggunaan Kaedah Dinamik Ultrasonik bagi Menentukan Pemalar Kenyal Kayu Tropika JADUAL 1 Nama tempatan, spesis, nilai ketumpatan purata dan pemalar kenyallongitud bagi sampel kayu tropikayang dikaji menggunakan teknik dinarnik ultrasonik (dalam unit 1010 N/m2) Sampel Nama Nama Ketum- Tempatan SaintilIk patan eLL Crr CRR CLL (kgm·3) Kayu keras Al Tembusu Fagraea fragrans 800 1.47 0.20 0.23 1.40 A2 Merbau Intsia palembanica 800 1.57 0.19 0.20 1.54 A3 Giam Hopeaspp. 975 1.67 0.21 0.24 1.65 A4 Balau merah Shorea spp. 880 1.74 0.19 0.23 1.70 A5 Resak Vatica spp. 945 1.88 0.18 0.24 1.81 A6 Kekatong Cynometra spp. 975 1.87 0.22 0.24 1.84 A7 Chengal Neobalanocarpus heimii 945 1.96 0.23 0.24 1.96 A8 Keranki Dialium spp. 960 1.56 0.22 0.23 2.01 A9 Balau Shoreaspp. 975 1.73 0.21 0.23 2.01 AI0 Bitis Madhuca utilis 1105 2.19 0.24 0.26 2.38 All Bakau Rhizophora spp. 1040 1.99 0.23 0.25 2.07 Kayu sederhami Bl Meransi Carallia spp. 800 1.41 0.17 0.22 1.34 B2 Simpoh Dillenia spp. 735 1.49 0.16 0.19 1.43 B3 Rengas Glutaspp. 835 1.52 0.20 0.23 1.49 B4 Kulim Scorodocarpus 835 1.51 0.19 0.22 1.49 B5 Punah Tetramesrista glabra 720 1.54 0.17 0.19 1.54 B6 Merawan Hopeaspp. 690 1.62 0.19 0.23 1.55 B7 Keledang Artocarpus spp. 800 1.57 0.16 0.21 1.55 B8 Mengkulang Heritiera spp. 755 1.65 0.18 0.20 1.60 B9 Mata ulat Kokoona littoralis 880 1.76 0.19 0.24 1.63 BI0 Kasai Pometia spp. 800 1.84 0.15 0.23 1.70 B11 Kelat Eugenia spp. 800 1.77 0.17 0.23 1.76 B12 Tualang Koompassia excelsa 835 1.81 0.17 0.20 1.78 B13 Merpauh Swietonia spp. 755 1.87 0.15 0.19 1.81 Bl4 Kempas Koompassia 880 1.90 0.19 0.24 1.86 B15 Kapur Dryobalanops spp. 755 1.98 0.18 0.21 1.80 B16 Keruing Dipterocarpus 880 2.01 0.19 0.23 2.23 Kayulembut Cl Terentang Campnosperma spp 435 1.09 0.17 0.18 0.70 C2 Jelutong Dyera costulata 465 1.19 0.13 0.18 0.80 C3 Pulai Alstonia spp. 465 1.39 0.14 0.20 0.71 C4 Sesendok Endospermum 530 1.40 0.16 0.19 0.85 malaccensis C5 Melantai Hopea macroptera 530 1.21 0.17 0.21 0.79 C6 Geronggang Cratoxylon 545 1.25 0.20 0.22 0.80 Pertanika J. Sci. & Techno!. Vol. 3 No.2, 1995 393

Sidek Hj. Abdul Aziz, Abdul Halim Shaari dan Chow Sai Pew Uadual 1) sambungan Sampel Nama Nama Ketum- Tempatan SaintiIIk patan CLL CIT CRR CLL (kgm·3) C7 Meranti Shoreaspp. 545 1.40 0.21 0.24 1.94 merah muda C8 Petai Parkiaspp. 545 1.13 0.12 0.23 1.07 C9 Perupok Lophopetalum spp. 560 1.38 0.20 0.22 1.26 C10 Machang Mangifera spp. 560 1.45 0.18 0.21 1.43 C11 Kedondong Burseraceae 575 1.45 0.13 0.24 1.29 C12 Terap Paratocarpus spp. 575 1.34 0.16 0.22 1.20 C13 Panarahan Myristicaceae 595 1.45 0.16 0.18 0.94 C14 Medang Lauraceae 610 1.53 0.16 0.22 1.26 C15 Ramin Gonystylus bancanus 625 1.46 0.17 0.23 1.64 C16 Kayu getah Hevea brasilliensis 640 1.32 0.17 0.20 0.92 C17 Mersawa Anisoptera spp. 640 1.37 0.18 0.21 1.26 CI8 Melunak Pentace spp. 655 1.31 0.18 0.20 1.20 C19 Meranti Shoreaspp. 655 1.38 0.20 0.22 1.21 kuning C20 Meranti Shoreaspp. 675 1.46 0.17 0.20 1.94 puteh C21 Kungkur Pithecellobium spp. 675 1.26 0.17 0.22 1.07 C22 Meranti Shorea uliginosa 675 1.50 0.16 0.18 1.47 bakau C23 Sepetir Sindora spp. 675 1.49 0.16 0.20 1.36 C24 Merawan Hopeaspp. 690 1.59 0.11 0.19 1.55 C25 Cerutu Parashorea lucida 690 2.06 0.17 0.20 2.06 C26 Bintangor Calophyllum spp. 690 1.84 0.19 0.22 1.43 C27 Durian Durio spp. 690 1.62 0.18 0.21 1.58 C28 Kembang Scaphium spp 705 1.64 0.20 0.22 1.70 semangkok C29 Meranti Shoreaspp. 705 1.54 0.21 0.24 1.39 Data pemalar kenyal CLL *dilaporkan oleh (MTIB) Malaysia Timber Industry Board (1986) . HASIL DAN PERBINCANGAN Perubahan nilai-nilai pemalar kenyal terhadap ketumpatan pada arah L, R dan T bagi sampel kayu tropika yang diukur pada suhu bilik ditunjukkan pada Rajah 3 (a)-(c) dan nilai puratanya dinyatakan dalam Jadual 1. Perbezaan nilai pemalar kenyal bagi satu spesis kayu dengan spesis yang lain adalah kerana kayu merupakan bahan yang bersifat tak homogen samaada pada jenis yang sama atau jenis yang berbeza. Struktur asas binaan kayu akan menentukan nilai ketumpatan dan sifat fizik kayu yang lain. Secara amnya struktur kayu keras terdiri dari bahan molekul makro 394 Pertanika J. Sci. & Techno!. Va!. 3 No.2, 1995

Penggunaan Kaedah Dinarnik Ultrasonik bagi Menentukan Pemalar Kenyal Kayu Tropika (40-45% hablur polimer alfa selulos, 15-35% hemiselulos dan 17-25% amorfus lignin) dan bahan bermolekul rendah (bahan organan - ekstraktif dan bahan tak organan). Manakala kayu lembut pula mengandungi 40- 45% selulos, 20% hemiselulos dan 25-35% lignin. Peratusan ini berbeza dari satu jenis dengan jenis kayu yang lain. Kandungan selulos tidak banyak berbeza bagi kayu jenis keras berbanding dengan kayu lembut (Desch 1980). Molekul selulos diorientasikan sepanjang paksi serabut (atau sepanjang butiran kayu) bagi menghasilkan kekuatan serta kekenyalan yang maksimum pada arah L (Patton 1986). Hal ini jelas diperhatikan dari Rajah 3(a)-(c), corak pemalar kenyal bagi kayu-kayan tropika adalah CLL >CRR>Crr sepertimana yang diperolehi oleh Bucur (1983) yang mengkaji kayu beech. Perbezaan inilai CLL' CRR dan Crr bagi setiap spesis kayu kajian (lihat Jadual 1) adalah di antara lain bergantung kepada struktur binaan kayu, ketumpatannya dan juga kelembapan bandingan persekitaran. Data- data ini adalah unik bagi setiap sampel kayu kajian dan amat berkait rapat dengan kualiti kayu berkenaan. Analisis korelasi linear ke atas nilai pemalar kenyal dan ketumpatan bagi sampel-sampel kayu tropika setiap satunya pada arah L, R dan T menggunakan komputer menunjukkan terdapatnya perkaitan yang berikut; i) CLL =1.2758 X 1O-3p + 0.643228 (r= 0.59), arah L ii) CRR = 7.1621 X 1O-5p + 0.16402 (r = 0.31), arah R iii) Crr = 9.43053 x 1O-5p + 0.112512 (r = 0.31), arah T dengan CLL'CRR'~ diukur dalam unit (x 1010 Nm2 ) dan p dalam unit kgm-3. 2.50 N E ........ Z ~2.00 x '--" o >-c Q) ~ '- , .50 o o EQ) 0... 1.00 400 Arch L 600 BOO 1000 1200 Ketumpatan Kayu (kgjm 3 ) Pertanika J. Sci. & Techno!. Vol. 3 No.2, 1995 395

Sidek Hi .hdul A~iz, Abdul Halim Shaari dan Chow Sai Pew 0.30 N E '-.. z ~o 0.25 o >. c Q) .Y L 0.20 o o EQ) r Arch R j -1 1 J i....--'" ._~.~. _~,._~'U'__"-J.'~·_L'~· ~,'_',~;_'_L, ....., '-,Wi'U'__"-1.'-.L.J,i 0.30 40(1 600 800 1000 Ketumpatan Kayu (kg/m 3) 1200 E '-..0.25 Z o -0 00.20 >. c Q) .Y L o E0.15 ~ I 0.10 400 Arch T 600 800 1000 Ketumpatan Kayu (kg/m 3) 1200 396 Rajah 3. Perubahan pemalar kenyal terhadap ketumpatan pada (a) arah L, (b) arah R dan (c) arah T bagi 56 spesis kayu tropika Pertanika J. Sci. & Techno!. Vo!. 3 No.2, 1995

Penggunaan Kaedah Dinamik Ultrasonik bagi Menentukan Pemalar Kenyal Kayu Tropika Nilai pemalar kenyal longitud kayu yang didapati dari eksperimen ultrasonik lebih tinggi berbanding dengan nilai pemalar kenyal kayu yang dilaporkan oleh MTIB (1986) menggunakan teknik statik. Rajah 4 menunjukkan perkaitan di antara pemalar kenyal CLL yang diukur menggunakan kaedah ultrasonik dan kaedah statik bagi menentukan sejauh mana penyisihan kedua-dua teknik bagi sampel-sampel kayu tropika. Dari eksperimen ultrasonik didapati nilai pemalar kenyal CLL bagi kayu tropika yang berketumpatan kurang dari 840 kgm3 lebih tinggi berbanding dengan nilai pemalar kenyal kayu yang dilaporkan oleh MTIB (1986). Analisis seterusnya terhadap data-data ini menunjukkan terdapat penyisihan yang agak tinggi bagi kayu tropika yang mempunyai ketumpatan yang rendah, tetapi bagi kayu jenis sederhana dan keras, nilai purta perbezaan antara kedua-dua teknik ini tidak melebihi 5%. Bagi kelompok kayu sederhana dan keras, teknik ultrasonik berkeupayaan memberikan data-data pemalar kenyal sepertimana dari teknik pengukuran statik. Perbezaan yang ketara ini disebabkan oleh isyarat gelombang ultasonik yang meranlbat di dalam sampel kajian diterima dalam masa yang amat singkat, menyebabkan tegasan yang amat kecil terhasil dari gelombang ultrasonik yang berfrekuensi antara 40 hingga 45 kHz ini. Hasil yang diperolehi bersesuaian dengan hasil yang diperolehi oleh Bucur (1985) iaitu nilai pemalar kenyal bagi kayu beech yang dikaji dengan menggunakan kaedah dinamik ultrasonik didapati lebih tinggi berbanding dari kaedah statik (lihatJaduall dan Rajah 4). Bodig and Jayne (1982) juga melaporkan bahawa nilai pemalar kenyal kayu yang diperolehi dari kaedah dinamik agak tinggi (sekitar 10-15%) berbanding dari kaedah lenturan statik bagi sejalur kayu; proses pelenturan kayu tersebut akan melibatkan canggaan ricih. Faktor ini disebabkan oleh perbezaan keadaan termodinamik bagi ujian statik dan ultrasonik. Secara teorinya pemalar kenyal yang terlibat dalam pengukuran statik adalah bersifat isoterma sedangkan bagi pengukuran dinamik ultrasonik ia lebih mematuhi hukum adiabatik terutama apabila frekuensi gelombang ultrasonik meningkat. Corak taburan data-data pengukuran dari kedua-dua teknik adalah seragam dengan sisihan taburan tidak melebihi 10% dan ini menunjukkan bahawa teknik ultrasonik berupaya menyediakan data-data pemalar kenyal sebagai suatu faktor pencirian untuk menentukan kualiti kayu. KESIMPUIAN Secara keseluruhan dapat disimpulkan kaedah dinamik ultasonik dapat digunakan bagi kayu tropika. Teknik ini dapat digunakan bagi tujuan pengujian dan pempiawaian bahan berkayu kerana nilai pemalar kenyal yang didapati melalui kaedah ini walaupun lebih tinggi tetapi peratusan ralatnya kecil iaitu kurang dari 10 peratus dari nilai yang diperolehi menerusi kaedah statik. Bagi kayu tropika corak pemalar kenyal adalah CLL>CRR>CTI dan terdapat satu korelasi yang linear di antara nilai pemalar-pemalar kenyal dan ketumpatan bagi 56 sampel kayu tropika yang dikaji. Pertanika J. Sci. & Techno!. Vol. 3 No.2, 1995 397

Sidek Hj. Abdul Aziz, Abdul Halim Shaari dan Chow Sai Pew 2.50 Aroh LN E ~2.00 5?O • o • : 2 • IV ./" o /" /" /" . ...... o. • 0 ......s o. • "01.50 >, c Q) .Yo L ..2 E1.00 Q) 0.. .. .., .. 9 'U' 'U' Kcedch Ultrcsonik ~ Kaedah Statlk 1200 0.50 400 600 BOO 1000 Ketumpatan Kayu (kg/m 3) Rajah 4. Perbandingan hasil pengukuran pemalar kenyal CLL menggunakan kaedah dinamik ultrasonik dan kaedah statik (MTIB 1986) PENGHARGAAN Projek penyelidikan ini dibiayai sepenuhnya oleh Kementerian Sains, Teknologi dan Alam Sekitar Malaysia menerusi geran penyelidikan jangka panjang IRPA/RME, Kod no. 1-07-05-062. Ucapan terima kasih ditujukan kepada Encik Zaidi Hassan, Cik Tijah Pardi dan stafJabatan Fizik di atas perbincangan dan bantuan teknikal. SIDEK HJ. ABDUL AZIZ, ABDUL HALIM SHAARI dan CHOW SAl PEW Jabatan Fizik, Fakulti Sains dan Pengajian Alam Sekitar Universiti Pertanian Malaysia 43400 UPM Serdang, Selangor, Malaysia. RUJUKAN BODIG, J. and B.A. JAYNE. 1982. Mechanics of Wood and Wood Composites. New York: Van Nostrand Reinhold. BUCUR, V. 1983. Ultrasonic method for measuring the elastic constants ofwood increment cores bored from living trees. Ultrasonics May: 116-126. BUCUR, V. 1985. Ultrasonic, hardness and x-ray densitometric analysis ofwood. Ultrasonics November: 269-275. 398 Pertanika J. Sci. & Techno!. Vo!. 3 No.2, 1995

Penggunaan Kaedah Dinamik Ultrasonik bagi Menentukan Pemalar Kenyal Kayu Tropika BUCUR, V. and F. FOCABOY. 1988. Surface wave propagation in wood: Prospective method for determination of wood off-diagonal terms of stiffness matrix. Ultmsonics November: 344-347. DESCH, H.E. 1980. Timber Its Structure, Properties and Utilisation. London: McMillan. MAL4,.YSIAN TIMBER INDUSTRY BOARD (MTIB). 1986. 100 Malaysian Timbers. Kuala Lumpur: MalaysianTimber Industry Board. PATTON, WJ. 1986. Materials in Industry. NewJersey: Prentice Hall. SIDEK H]. AB. AzIZ, H]. SALLEH HARON, ABDUL HALIM SHAARI, CHOW SAl PEW, GAZALI AHMAD dan MOHD. SALLEH MOHD DEN!. 1990. Pendekatan ujian tak memusnah bagi penganalisaan ciri kenyal bahan kayu. In: Seminar Kebangsaan Fizik Dalam Industri, 19-20 September (1990) Universiti Teknologi Malaysia, Sekudai,Johor. SZYMANI, R. and K.A. McDoNALD. 1981. Defect detection in lumber; state of the art. Forest ProductsJournal31 (11): 34-43. PertanikaJ. Sci. & Technol. Vol. 3 No.2, 1995 399

Ultrasonic detection of knots, cross grain and bark pockets in wooden pallet parts Kabir, M. F..1 , Schmoldt, Daniel L.2 and Schafer, Mark E.3 ABSTRACT This study investigates defect detection in wooden pallet parts using ultrasonic scanning. Yellow-poplar (Liriodendron tulipifera, L.) deckboards were scanned using two rolling transducers in a pitch-catch arrangement to detect unsound and sound knots, bark pockets and cross grain. Data were collected, stored, and processed using LabView™ software. Six ultrasonic parameters—three involving time-of-flight, two involving ultrasound pulse energy, and one using ultrasound pulse duration—were measured for each defect type. Four of the six parameters were affected by transmission through unsound knot regions. Sound knots also showed decreased values forthe energy-related parameters. All ultrasonic parameters changed sharply for bark pockets. Cross grain also affected ultrasound energy transmission. Small coefficients of variation for repeated measurements indicates that this scanning arrangement is stable and scanning rate has little effect on the measurements. Results indicate that on-line detection of these defects is possible by ultrasonic scanning. INTRODUCTION Wooden pallets are the largest single use of sawn hardwood logs, consuming around 40% of all US hardwood lumber produced. Every year, over 400 million wooden pallets are manufactured using 4.5 billion board feet of hardwood lumber. Pallets are integral to the US transportation infrastructure, and wood is the primary raw material used in pallets. Most wooden pallets consist of two parts, stringers—the structural center members that carry the product load—and deckboards—the top and the bottom members that provide dimensional stability and product placement. There are many types of pallet designs—depending on the size, number, and position of stringers and deckboards—but most are produced from solid wood, lumber, or from the center cant material of logs. Most solid wood material, that is manufactured into pallet parts, is low quality (having a high percentage of defects) and therefore has less market value for other solid wood products. The most common defects in pallet parts are knots, cross grain, reaction wood, bark pockets, insect holes, splits, decay, shake, and wane. For quality pallet production, it is necessary to detect defects during manufacturing and then grade and sort parts prior to pallet assembly. An economic analysis by Schmoldt et al. (1993) indicated that improved pallet durability and performance imparts much greater value to carefully manufactured pallets. Current pallet manufacturing operations, however, do lend themselves to manual grading and sorting of parts. Therefore, this research program aims to develop automated techniques that include scanning, defect detection, and grading. Previous work has investigated a variety of ultrasonic waveform parameters to detect defects in wood (McDonald 1980, Lemaster and Dornfeld 1987, Patton-Mallory and DeGroot 1990, Ross et al. 1992, Schmoldt et al. 1994, Fuller et al. 1995, Kabir et al. 1997). Most of these studies were conducted using laboratory samples or surfaced lumber, whereas in practice, conditions may be quite different. This is especially the case in the pallet industry where low quality, unsurfaced wood must be scanned. Furthermore, simple ultrasonic propagation velocity alone may not be sufficient to detect most defects. Other ultrasonic parameters, e.g. peak amplitude, time to peak amplitude, centroid time, root mean square of the time domain, pulse length, energy, frequency domain modes, frequency domain energy, etc., may be required. Recently, Halabe et al. (1993, 1994, 1996) conducted a study using ultrasonic frequency analysis for decay detection in wooden 1 Postdoctoral Fellow, Dept. of Wood Science and Forest Products, Virginia Tech, Blacksburg VA 24061-0503 2 Research Forest Products Technologist, USDA Forest Service, Biological Systems Engineering Dept., 460 Henry Mall, Madison WI 53706-1561 3 Vice-President, Ultrasound Technology Group, Forest Products Division, Perceptron Inc., 2935 Byberry Road, Hatboro PA 19040

timbers. They reported that frequency domain analysis can significantly increase prediction sensitivity for modulus of elasticity and strength of clear and defective wood under controlled laboratory conditions. More research needs to be done comparing different ultrasonic parameters’ sensitivities to defects. The final goal of this study is to develop an automated ultrasonic scanning system for defect detection in pallet parts. This initial experiment was carried out to determine which ultrasonic parameters respond well to particular defects and also to observe the reliability and repeatability of data collection using pressure-contact rolling transducers. MATERIALS AND METHODS Scanning equipment A materials handling system was designed by the Forest Products Division of Perceptron and purchased by the USDA Forest Service. It consists of in-feed and out-feed roll-beds and an ultrasonic scanning ring where rolling transducers are mounted. Perceptron provided the necessary electronics and software to control material movement, signal generation, and waveform capture and analysis. Pallet parts move through the system lying on a face and pitch-catch ultrasonic transmission propagates through the part’s thickness. The transducers can be operated using a range of frequencies in 90- 180 kHz. A single scan line of data is collected during each pass of the part through the scanner. The desire resolution of this scan line (number of waveforms per inch) can be achieved by controlling roller speed and number of pulses/sec. Data were processed to create six ultrasonic parameters—time of flight-centroid (TOF-centroid), time of flight-energy (TOF- energy), time of flight-amplitude (TOF-amplitude), pulse length (PL), energy value (EV), and energy/pulse value (EPV), to be discussed further below. Definition of the ultrasonic parameters4 The most important parameters relate to the energy in the received signal. Wave energy is expressed as the time integral of the voltage squared: E = v2 (t)dtò (1) Because of the wide variation in transmitted energy levels between sound wood and defects, it is more convenient to express the energy on a logarithmic basis. The energy value (EV) is derived from the energy E, and is expressed in decibels (dB). By convention, this is a negative number, with lower signals (containing less energy) being more negative. The pulse length parameter (in units of microseconds) is simply the time for which the pulse is “on”, and depends upon the transmitted ultrasound frequency. These two parameters, energy value and pulse length, can be combined to provide a single parameter, which is known as energy/pulse value (EPV). Again, because of the wide range of energy levels, EPV is also expressed on a logarithmic scale (in dB). TOF-energy is calculated as the time at which the energy integral (Equation 1) crosses a threshold value—as a percentage of the final (maximum) value. If the threshold value is, for instance, 40%, then TOF-energy is simply the time at which the integral value reaches 40% of the final value. Similarly, TOF-amplitude is the time at which the amplitude of the signal first reaches, for instance, 40% of the maximum amplitude. TOF-centroid is the time to the centroid of the time waveform, which is based on the ratio of the first- and zero-th order moments. No frequency domain parameters were calculated in this study. Data collection Twelve, fresh-cut yellow-poplar boards 40 inches in length and approximately 1/2 inch thickness were collected from a pallet manufacturer. The boards were kept in cold storage to reduce their drying rate. A line was drawn on each board through a defect of interest and scanning was performed along this line. All parameters were calculated for each scan. Three boards were scanned for each defect type. The boards were scanned with two scanning rates—10 waveforms/inch (70 ft/m roller speed) and 4 waveforms/inch (220 ft/m roller speed). Each board scan line was repeated ten times for each defect type and scanning rate. All measurements were made at 120 kHz transmitting frequency and 500 kHz sampling frequency. 4 For proprietary reasons, full details regarding the ultrasonic parameters and their measurement cannot be released at this time.

RESULTS AND DISCUSSION Fig. 1 depicts graphs of ultrasonic parameters plotted against board length for a scan line that includes an unsound knot. These data are taken from one of the sample boards; other boards behaved in a similar manner. Data collected at 10 waveforms/inch and 4 waveforms/inch are also shown. The graphs show that PL and TOF-centroid increase sharply with the unsound knot. TOF-amplitude and TOF-energy seem relatively unaffected, however. On the other hand, the presence of an unsound knot causes a dramatic decrease in EV and EPV. Scanning rate does not appear to affect parameter measurements (Figs. 1c & 1d); both, EV and EPV have nearly identical values at scanning rates of 10 waveforms/inch and 4 waveforms/inch. In addition to a dramatic loss in transmitted energy, unsound knots also tend to spread out the received waveform, increasing the pulse length and centroid value. 0 50 100 150 200 0 10 20 30 40 50 µµµµs Pulse length TOF-centroid (a) 0 20 40 60 80 0 10 20 30 40 50 µµµµs TOF-amplitude TOF-energy (b) -120 -100 -80 -60 -40 -20 0 0 10 20 30 40 50 EV (dB) EPV (dB) (c) -120 -100 -80 -60 -40 -20 0 0 10 20 30 40 50 EV (dB) EPV (dB) (d) Figure 1. Ultrasonic measurements taken at 10 waveforms/inch along the length of a scan line (x-axis) are shown for all six parameters (a, b, c). An unsound knot is present between the 5- and 10-inch locations. For comparison, the last graph (d) shows measurements made at 4 waveforms/inch for EV and EPV. The effects of sound knots on ultrasonic parameters appear in Fig. 2. PL, TOF-centroid, TOF-amplitude, and TOF- energy seem unaffected by the sound knots, but EV and EPV exhibit a sharp decrease around the regions of sound knots. While unsound knots contain incipient or advanced decay or have a complete separation of knot from surrounding wood (loose), sound knots influence wood properties by interrupting the longitudinal direction of wood fibers (Anon 1987). Wood fibers around a sound knot are distorted, developing localized cross grain which may have substantial impact on ultrasonic measurements. While it is well established that TOF measurements can detect the presence of sound knots (e.g., McDonald 1980, Kodama and Akishika 1993, Schmoldt et al. 1996; Kabir et al. 1997), it appears, from this study, that energy losses (Figs. 2c & 2d) are more sensitive than TOF measurements (Figs. 2a & 2b). Like unsound knots, sound knots do not have any noticeable effect on measurements at different scanning rates (Figs. 2c & 2d).

Bark pockets contain some bark in place of wood, so one would expect them to have a significant effect on ultrasonic measurements. Fig. 3 illustrates how the ultrasonic parameters vary with bark pockets. PL, TOF-centroid, and TOF- energy increase sharply with bark pockets whereas EV and EPV decrease. The tremendous increase or decrease of the parameters may be associated with the presence of a small split and decay in the bark pocket of this sample board. 0 20 40 60 80 100 120 140 0 10 20 30 40 50 µµµµs Pulse length TOF-centroid (a) 0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 µµµµs TOF-amplitude TOF-energy (b) -80 -70 -60 -50 -40 -30 -20 -10 0 0 10 20 30 40 50 EV (dB) EPV (dB) (c) -80 -70 -60 -50 -40 -30 -20 -10 0 0 10 20 30 40 50 EV (dB) EPV (dB) (d) Figure 2. Ultrasonic measurements taken at 10 waveforms/inch along the length of a scan line (x-axis) are shown for all six parameters (a, b, c). Sound knots are present between the 12- and 18-inch locations and also between the 32- and 38- inch locations. For comparison, the last graph (d) shows measurements made at 4 waveforms/inch for EV and EPV. Cross grain represents a generalized slope of grain; it is often measured as a ratio of perpendicular deviation versus longitudinal reach, e.g. 1:10 or 1:6. Deviations from longitudinal (X direction) are measured in both the Y (dy) and Z (dz) directions, with the resultant cross grain d calculated as in (2). Ultrasound propagation is known to differ with grain direction (anisotropy), so one would expect several measured parameters to be affected. PL shows high variability in the cross grain region (Fig. 4a). There are also some decreases in EV and EPV measurements, but not as dramatic as bark pockets or knots. Again scanning rate showed no effect on data collection. d = dy 2 + dz 2 (2) To test the repeatability of measurements and the reliability of the data, boards were scanned ten times with coefficients of variation calculated. These are presented in Fig. 5 for a bark pocket scan line using the board appearing in Fig. 3. The low CV% values suggest that the repeatability of data collection using pressure-contact rolling transducers is very good. Some inflation of CV values likely occurs because of errors in data point registration between repeated scans.

0 100 200 300 400 500 600 0 10 20 30 40 50 µµµµs Pulse length TOF-centroid (a) 0 50 100 150 200 0 10 20 30 40 50 s TOF-amplitude TOF-energy (b) -140 -120 -100 -80 -60 -40 -20 0 0 10 20 30 40 50 EV (dB) EPV (dB) (c) -140 -120 -100 -80 -60 -40 -20 0 0 10 20 30 40 50 EV (dB) EPV (dB) (d) Figure 3. Ultrasonic measurements taken at 10 waveforms/inch along the length of a scan line (x-axis) are shown for all six parameters (a, b, c). A bark pocket is present at approximately the 10-inch location. For comparison, the last graph (d) shows measurements made at 4 waveforms/inch for EV and EPV. CONCLUSIONS There appear to be significant, defect-specific differences in several ultrasonic parameters for yellow-poplar deckboards. Most of the ultrasonic parameters examined here change rapidly in the region of defects, which can be used for on-line inspection of defects. Because species and individual pallet parts will vary in the magnitude of various ultrasonic parameters, relative changes may prove to be the most informative and diagnostic characteristic of ultrasonic signal propagation. Energy value and energy/pulse value were found to be the most sensitive ultrasonic parameters for the defects examined in this study. Bark pockets and unsound knots are more easily detected compared to sound knots and cross grain. Small values for coefficients of variation indicate that repeatability and reliability are acceptable. Scanning rate has little effect on data collection, which means that it should be possible to scan at relatively high industrial speeds. Continued work in this project will expand data collection further. Other wood species will be tested to determine if significant inter-specific differences exist with respect to the ability of ultrasonic parameters to distinguish defect types. We also plan to examine still other ultrasonic parameters, e.g., in the frequency domain, that have been shown to discriminate between certain defect types. Eventually, multiple scan lines will be collected for each board to enable us to create ultrasonic 2-D maps (images). Such maps will actually be 3-D (multi-dimensional), as the 2-D maps will have values at each scan point for a variety of useful ultrasonic parameters. The contribution of several parameters should produce fairly accurate defect characterization at each scan point on the pallet part.

0 20 40 60 80 100 120 140 0 10 20 30 40 50 µµµµs Pulse length TOF-centroid (a) 0 20 40 60 80 100 0 10 20 30 40 50 µµµµs TOF-amplitude TOF-energy (b) -100 -80 -60 -40 -20 0 0 10 20 30 40 50 EV (dB) EPV (dB) (c) -100 -80 -60 -40 -20 0 0 10 20 30 40 50 EV (dB) EPV (dB) (d) Figure 4. Ultrasonic measurements taken at 10 waveforms/inch along the length of a scan line (x-axis) are shown for all six parameters (a, b, c). A region of cross grain is delineated by the tilted oval. For comparison, the last graph (d) shows measurements made at 4 waveforms/inch for EV and EPV.

-140 -120 -100 -80 -60 -40 -20 0 0 10 20 30 40 50 EPV(dB) Average CV% (a) -120 -100 -80 -60 -40 -20 0 0 10 20 30 40 50 EV(dB) Average CV% (b) 0 100 200 300 400 500 600 0 10 20 30 40 50 Pulselength Average CV% (c) Figure 5. Average values and coefficients of variation (%) are shown for 10 repeated measurements of three ultrasonic parameters taken at 10 waveforms/inch along the length of a scan line (x-axis). The board and scan line appearing in Fig. 3 are the source of the data. REFRENCES Anon. 1987. Wood Handbook: Wood as an Engineering Material. Agriculture Handbook 72, USDA Forest Service, Madison. 460pp. Fuller, J.J., Ross, R.J., and Dramm, J.R. 1995. Non destructive evaluation of Honeycomb and surface check in Red Oak lumber. Forest Products Journal 45(5): 42-44. Halabe, H.B., GangaRao, H.V.S., and Hota V.R. 1993. Nondestructive evaluation of wood using ultrasonic frequency analysis. Pages 2155-2160 in D.O. Thompson and D.E. Chimenti, (Eds.) Review of Orogress in Quantitative Nondestructive Evaluation Vol. 13. New York, Plenum Press. Halabe, U.B., GangaRao, H.V.S., and Solomon, C.E. 1994. Non destructive evaluation of wood using ultrasonic dr- coupled trasducers. Pages 2251-2256 in D.O. Thompson and D.E. Chimenti, (Eds.) Review of Progress in Quantitative Nondestructive Evaluation Vol. 12. New York, Plenum Press. Halabe, H.B., GangaRao, H.V.S., Petro, S.H., and Hota V.R. 1996. Assessment of defects and mechanical properties of wood members using ultrasonic frequency analysis. Materials Evaluation 54(2): 314-352. Kabir, M.F., Sidek, H.A.A., Daud, W.M., and Khali, K. 1997. Detection of knot and split of rubber wood by non- destructive ultrasonic method. Journal of Tropical Forest Products 3(1): 88-96. Kodama, Y. and Akishika, T. 1993. Non-destructive inspection of of defects in wood by use of pulse-echo technic of ultrasonic waves. I. Measurements of enclosed knots. Mokuzai Gakkaishi 39(1): 7-12.

Lemaster, R.L and Dornfield, D.A. 1987. Prelimanary investigation of the feasibility of using acoustic-ultrasonics to measure defects in lumber. Journal of Acoustics Emission 6(3): 157-167. McDonald, K.A. 1980. Lumber defect detection by ultrasonics. Res. Pap. FPL-311, Madison WI: USDA Forest Service. Forest Products Lab. 20p Patton-Mallory, M. and DeGroot, R.C. 1990. Detecting brown-rot decay in southern yellow pine by acousto-ultrasonics. Pages 29-44 in Proceedings of the 7th International Nondestructive Testing of Wood Symposium, September 27- 29, 1989, Madison WI, Conference and Institute, Washington State University. Ross, R.J., Ward, J.C., and Tenwolde, A. 1992. Identifying bacterially infected oak by stress wave non-destructuve evaluation. FPL-RP-512, USDA Forest Service, Madison Schmoldt, D.L., McCleod III, J.A., and Araman, P.A. 1993. Economics of grading and sorting pallets parts. Forest Products Journal 43(11/12): 19-23. Schmoldt, D.L., Morrone, M., and Duke Jr., J.C. 1994. Ultrasonic inspection of wooden pallets for grading and sorting. Pages 2161-2166 in D.O. Thompson and D.E. Chimenti, (Eds.) Review of Progress in Quantitative Nondestructive Evaluation. Vol. 12. New York, Plenum Press. Schmoldt, D.L., Nelson, R.M., and Ross, R.J. 1996. Ultrasonic defect detection in wooden pallet parts for quality sorting. In S. Doctor, C. A. Lebowitz, and G. Y. Baaklini (Eds.) Nondestructive Evaluation of Materials and Composites, SPIE 2944: 285-295.

Dielectric and ultrasonic properties of rubber wood. content, grain direction and frequency M.F. Kabir, W.M. Daud, K. Khalid, H.A.A. Sidek Holzals Roh-und Werkstoff56 (1998)223-2279 Springer-Verlag1998 Effect of moisture Dielectric properties of rubber wood have been studied at low and microwave frequencies with different moisture content and grain direction. The ultrasonic properties were studied with pulsed longitudinal waves of frequency 45 kHz. Two anisotropic directions have been considered for this study - parallel and perpendicular to grain. The low frequencies were of 0.01, 0.1, 1.0, 10 and 100 Hz and microwave frequencies were of 1, 2.45, 6, 8, 10, 14 and 17 GHz. The moisture content affected the dielectric constant and dielectric loss factor both at low and mi- crowave frequencies. The moisture content above 30% showed little influence on dielectric properties whereas it increases linearly from 0 to 30% in both the grain direc- tions at low frequencies. A continuous increase of dielec- tric properties was obtained with the increase of moisture content at microwave frequencies and the trend becomes concave upward. Dielectric properties increase as the fre- quencies increase except dielectric loss factor at micro- wave frequencies where reverse trends were observed. Little change of dielectric loss factor was obtained at fre- quencies above 6 GHz. The parallel to grain direction showed higher dielectric constant and dielectric loss factor compared to perpendicular to grain direction. This di- electric anisotropy of wood may be attributed due to the microscopic, macroscopic molecular as well as chemical constituents of wood. Ultrasonic properties were also af- fected considerably by the moisture content and grain direction. The dried wood showed higher ultrasonic ve- locity and elastic stiffness constant compared to green wood. The parallel to grain direction exhibits higher ul- trasonic velocity and elastic stiffness constant than per- pendicular to grain. Oielektrische und UltraschalI-Eigenschaften von Hevea brasiliensis: EinfluB yon Feuchte, Faserrichtung und Frequenz Die dielektrischen Eigenschaften yon Hevea brasiliensis wurden bei niedriger und Mikrowellenfrequenz sowie unterschiedlichen Feuchten und Faserrrichtungen unter- sucht. Die Ultraschalleigenschaften wurden mit gepulsten Longitudinalwellen von 45 kHz bestimmt. Beide Bestim- mungen erfolgten parallel und senkrecht zur Faser. Als niedrige Frequenzen wurden 0,01, 0,1, 1,0, 10 und 100 Hz eingesetzt, im MikroweUenbereich 1, 2, 4, 5, 6, 8, 10, 14 und 17 GHz. Die Feuchte beeinflu~t die Dielektriziffits- konstante und den Verlustfaktor in beiden Frequenzbe- reichen. Zwischen 0 und 30% Feuchte steigt die Dielektrizit~itskonstante bei niedrigen Frequenzen linear M.F. Kabir, W.M. Daud, K. Khalid, H.A.A. Sidek Department of Physics, Universiti Putra Malaysia, 43400, Serdang, Selangor, Darul Ehsan, Malaysia mit der Feuchte an. Ober 30% ergeben sich nur geringe Anderungen. Im Mikrowellenbereich steigt die Dielektri- zit~itskonstante fiber den gesamten Feuchtebereich expo- nentiell an, wogegen der Verlustfaktor bei hohen Frequenzen bis 6 GHz abnimmt; oberhalb davon wurden nur geringe Anderungen beobachtet. Auch die Ultraschall- Eigenschaften werden deutlich yon der Feuchte und Fa- serrrichtung beeinflut~t. In trockenem Holz sowie parallel zur Faserrichtung sind Geschwindigkeit und Elastizitiits- modul hSher als in feuchtem Holz und senkrecht zur Fa- ser. 1 Introduction The dielectric properties of rubber wood are important for understanding the structure of wood and cellulose at molecular level as well as for measuring the density, moisture content by nondestructive method. It was re- ported that the detection of knot, defects, spiral grain, etc. are also possible by measuring dielectric properties (Martin et al. 1987). The dielectric properties of wood are essential for its efficient use in engineering application where it is subjected to alternating fields, such as in large power transformers. It plays a significant role in heating, drying and gluing and thus, improving the quality of the wood and wood based materials. The dielectric properties of wood vary with physical parameters such as moisture content, density, grain direction and temperature. They also vary in an extremely complicated fashion with frequency. The overall effects of these parameters on dielectric properties of wood interact with each other and add to the complexities of the die- lectric properties. Though there are some reports on the variation of dielectric properties with these parameters, most of them deal with low moisture content and with few frequencies (James 1975, 1977; Kroner and Pungis 1952; Nanassy 1972; Norimoto and Yamada 1976; Rafalski 1967; Skaar 1948;Vermas 1974, 1976; Venkateswaran and Tiwari 1964). The present work deals with the variation of die- lectric properties, such as dielectric constant and dielectric loss factor of rubber wood with moisture content, grain direction and frequency. The conventional static test (destructive) for evaluating the wood properties is quite an expensive, time consuming process and it would take decades of work to accomplish the test for various species. As an alternative to the static test, several methods such as vibration, X-ray radio graphics, pilodyn wood testers have been employed for evaluating wood properties (Parker and Kennedy 1973). The ultrasonic technique may be suitable for the rapid determination of the mechanical properties. It is also re- ported that the detection of defects such as honeycomb, check, split, knot, etc., are also possible by ultrasonic method (James et al. 1995; Kabir et al. 1997). The effect of 223

224 moisture content and grain direction on the ultrasonic properties have also been discussed in this paper. 2 Materials and methods Rubber wood (Hevea brasiliensis) was supplied by the Farm Department of Universiti Putra Malaysia. Specimens were prepared in the form of discs of 35-40 mm in dia- meter and 3.0-3.5 mm in thickness for the measurement of dielectric properties at low frequencies; and 22-30 mm in diameter and 3.5-5.0 mm in thickness at microwave frequencies. Two main anisotropic directions have been considered for this study - parallel to grain and perpen- dicular to grain. Both surfaces of the specimen were smoothed with sand paper so that it made good contact with the electrodes. The low frequency measurements were carried out with a parallel plate electrode at frequencies from 10-2 to 102 Hz by using Dielectric Spectrometer consisting of Chelsea Dielectric Interface (CDI 4c/L-4, Dielectric Instrumentation, UK) and Frequency Response Analyzer (SI 1255, Schumberger, UK). The microwave experiments were done with a 4 mm open ended coaxial sensor (HP 8507 M) and computer controlled Network Analyzer (HP 8720B). Frequenices used for this study were of 1, 2.45, 6, 8, 14 and 17 GHz. Following the proper ca- libration method, the accuracy of the measurement is about +5% for dielectric constant and +3% for dielectric loss factor. Ultrasonic measurements were carried out with a commerical ultrasonic tester (BP V - Steinkamp, Ger- many) of 45 kHz pulsed longitudinal waves. Two conical transducers were used for transmitting and receiving the pulses. Transmission times were digitally displayed and recorded manually. The ultrasonic velocity was calculated by dividing the specimen length with transmitting time. The elastic stiffness constant was determined using the following equation: eli = p V 2 whereC/j is the elastic stiffness constant, p is the density of the wood specimen and V is the ultrasonic velocity. To measure the dielectric and ultrasonic properties at different moisture content, initially the specimens were fully soaked in water for a sufficiently long time to achieve full saturation. After that the weight of the specimen was taken and measurement was carried out. It was then dried in air to reduce the moisture. This cycle of measuring, drying and weighing was repeated until the specimen showed no change of weight by drying. Finally the oven dried weight of the specimen was taken by drying in an electronic oven at 100 + 3 ~ for 24 hours. 3 Results and discussions The dielectric constant of rubber wood at frequencies 0.01, 0.1, 1.0, 10 and 100 Hz in parallel and perpendicular to grain directions are presented in log scale in Figs. 1 and 2 respectively. Figures 3 and 4 show the variation of die- lectric loss factor with moisture content in both directions. Regardless of grain direction, two distinct regions of moisture content have been found for the variation of dielectric properties at low frequencies- one below 30% and the other above the 30% moisture content. The die- lectric constant and dielectric loss factor increases as the moisture content increases form 0 to 30% and thereafter a slight increase of the dielectric properties are observed 8 7 ............... CI + = I E 6 ............ ~'.i .... /................................. 5 ...... ...... .iiiiiii iiiii iiii iiii iiiiiiii il-.t2 .... i i i i i i ii i 0 10 20 30 40 50 60 70 80 90 1O0 Moisture content (%) Fig. 1. Dielectric constant vs moisture content in parallel to grain direction at low frequency [] o.oi Hz, + o.1 Hz, * 1.o Hz, 9 zo.o Hz, x lOO Hz Bild 1. Dielektrizitiitskonstante parallel zur Faser in Abh~ingig- keit yon der Feuchte bei niedriger Frequenz 8 7 . . . . . . . . . . . . . . . . . . . . . . . . r . . . . . . . . . . . . . . . . . = . . . . . . . . . . . . . ~4 ........... 0 3 ...... ~5 1'o 2' . . . . . . .0 0 0 40 50 60 70 80 90 100 110 Moisture content (%) Fig. 2. Dielectric constant vs moisture content in perpendicular to grain direction at low frequency [] o.ol Hz, + o.I Hz, * x.o Hz, 9 IO.O HZ, X 1OO HZ Bild 2. Dielektrizit~itskonstante senkrecht zur Faser in Abh~in- gigkeit yon der Feuchte bei niedriger Frequenz -2 ! , ,i i i ~ i i i I0 20 30 40 50 60 70 80 90 1O0 Moisturecontent(%) Fig. 3. Dielectric loss factor vs moisture content in parallel to grain direction at low frequency [] o.ol Hz, + o.1 Hz, * x.o Hz, i lo.o Hz, x lOO Hz Bild 3. Verlustfaktor parallel zur Faser in Abhiingigkeit yon der Feuchte bei niedriger Frequenz

with the increase of moisture content. The increase of dielectric constant with the increase of moisture content upto 30% was also observed by Mkhaiovskaya (1972) and ]ames (1975). Lin (1976) stated that the polar groups in the cell wall and cellulose have increased freedom of rotation when moisture content increases and thus increasing the dielectric constant. The free water, i.e. the moisture cont- ent above 30% does not have much effect on the dielectric properties. Dielectric constant of rubber wood at microwave fe- quencies of 1.0, 2.45, 6.0, 8.0, 10.0, 14.0 and 17.0 GHz in parallel and perpendicular to grain directions are shown in Figs. 5 and 6 respectively. Dielectric loss factors in both the directions are shown in Figs. 7 and 8. The dielectic constant and dielectric loss factor increase with the in- crease of moisture content throughout the whole ranges. The abrupt change of the dielectric properties are observed at very high moisture content and the curve becomes concave upward. The dielectric constant and dielectric loss factor vary almost linearly at low moisture content. Free water molecules interact with the microwave field inde- pendently of the cell wall substances and bound water. Therefore, the change of dielectric constant and dielectric loss factor is determined mainly by the dielectric proper- ties of free moisture and its volume. It is observed from Figs. 1-4 that the dielectric constant and dielectric loss factor in parallel to grain direction is greater than in perpendicular to grain direction. The va- riation of dielectric properties between parallel and per- pendicular to grain direction is due to the difference in the arrangement of cell wall and lumen in addition to the anisotropy of cell wall substances (Norimoto et al. 1978). The greater dielectric constant in parallel to grain direc- tion may be explained in terms of the transition proba- bility of dipole jump to an adjacent site when the field applied to parallel to grain direction is considerably higher than that when the electric fields were applied in perpen- dicular to grain (Norimoto and Yamada 1970). The che- 225 35 .. 67~ ...................................... ,................. 9....................... ,. 30 ..................................................... 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . g} . . . . . . . . . . o 3 ............................................... o ......................................................... -15 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~~ ........................................................... "J .................................................................. o, . . . . . -4 n ' ' ' ' ' ' ' ' ' ' 0 20 40 60 80 1O0 0 10 20 30 40 50 60 70 80 90 100 110 Moisture content (%) Moisture content (%) Fig. 4. Dielectric loss factor vs moisture content in perpendicular to grain direction at low frequency [] o.o1 Hz, + o.1 Hz, * 1.o Hz, 9 10.0 Hz, x zoo Hz Bild 4. Verlustfaktor senkrecht zur Faser in Abh~ingigkeit yon der Feuchte bei niedriger Frequenz Fig. 6. Dielectric constant vs moisture content in perpendicular to grain direction at microwave frequency [] 1.o GHz, + z.45 GHz, 6.o GHz, 9 8.o GHz, x zo.o GHz, 9 14.o GHz, A 17.o GHz Bild 6. Dielektrizit~itskonstante senkrecht zur Faser in Abh/in- gigkeit yon der Feuchte bei Mikrowellenfrequenz 35 3O 25t- 2o ._ alO 5 0 I I o 2'0 4'0 6'0 ao loo Moisture content (%) Fig. 5. Dielectric constant vs moisture content in parallel to grain direction at microwave frequency [] 1.o GHz, + z.45 GHz, * 6.o GHz, 9 8.o GHz, x lO.OGHz, 9 i4.o GHz, A 17.o GHz Bild 5. Dielektrizit~itskonstante parallel zur Faser in Abhtingig- keit yon der Feuchte bei Mikrowellenfrequenz o (.] ~5 14 12- 10- 8- 6- 4- 2- O~ 0 10 20 30 40 50 60 70 80 90 100 Moisture content (%) Fig. 7. Dielectric loss factor vs moisture content in parallel to grain direction at microwave frequency [] x.o GHz, + 2.45 GHz, * 6.0 GHz, 9 8.o GHz, x lO.OGHz, 9 14.o GHz, A 17.o GHz Bild 7. Verlustfaktor parallel zur Faser in Abh/ingigkeit yon der Feucht bei Mikrowellenfrequenz

226 mical constituent of wood may also be responsible for the dielectric anisotropy. According to Norimoto and Yamada (1972), the dielectric properties of wood are strongly in- fluenced by cellulose and mannan in parallel to grain direction whereas in perpendicular to grain direction the dielectric properties are influenced by lignin. The measurement frequency also affected the dielectric properties considerably. At low frequencies, the lower the frequenices, the higher the dielectric constant and dielec- tric loss factor (Figs. 1-4). This is also true for the die- lectric constant in microwave frequencies. But for dielectric loss factor at microwave frequencies, higher values were obtained at higher frequencies. A little varia- tion of dielectric loss factor was observed above 6 GHz. The interaction of the electromagnetic field with the molecules of the wood substance at higher frequencies differs from those at lower frequencies, as the period of field oscillation at microwave frequencies can be compared with relaxation time of the molecules (Torgovnikov 1993). A phase shift, therefore arises between the field strength 140 12o- % ~xlOO- E 80- 8 60- E ,-n 40 ................................................................ o i t i ~ i i t 0 10 20 30 40 50 60 70 Moisture content (%) 80 Fig. 10. Elastic stiffness constant vs moisture content in different grain direction: [] Longitudinal, + Radial, * Tangential Bild. 10. Elastizit/itskonstante bei verschiedenen Faserrichtun- gen in Abh~ingigkeit yon der Feuchte: [] longitudinal, + radial, 9 tangential 12 10 ............................................................. o 6 =o ........................... ~* ~. .~.... -~ 4 t5 2 0 0 2'0 4'o 6'0 8'o t60 Moisture content (%) Fig. 8. Dielectric loss factor vs moisture content in perpendicular to grain direction at microwave frequency 1.o GHz, + 2.45 GHz, * 6.0 GHz, [] 8.0 GHz, x lO.O GHz, 9 14.o GHz, A 17.o GHz Bild 8. Verlustfaktor senkrecht zur Faser in Abh~ingigkeit yon der Feuchte bei Mikrowellenfrequenzen 5000 ii ii iiii ii 3500 ........................................................ = 3000 ................................................................o o 2 5 0 0 . . . . . . . . . . . . . . . ". . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2000- ~ - - - - .............................. 15oo. . . . . . . . . . .~ : ~......... L § 1000 . . . . 5'O io0 10 20 30 40 60 80 Moisture content (%) Fig. 9. Ultrasonic velocity vs moisture content in different grain direction: [] Longitudinal, + Radial, * Tangential Bild 9. Uhraschallgeschwindigkeit bei verschiedenen Faserrich- tungen in Abh/ingigkeit yon der Feuchte: [] longitudinal, + radial, 9 tangential vector and the polarization vector resulting in a reduction of dielectric constant and an increasing of loss factor with the increase of frequency were also reported by James and Hamil (1965), Torgovnikov (1993) and Tiuri et al. (1980). Empirical equations of the type Y = ax b have been determined for calculating dielectric constant and dielec- tric loss factor from the moisture content at low frequency. The value calculated from the equations are shown by the solid lines in Figs. 1-4. Fourth order polynomial equations of the type Y = ax 4 + bx 3 + cx 2 + dx + e were found sui- table for the curve fitting at microwave frequencies. The value calculated from the equations of curve fitting were shown by solid lines (Figs. 5-8). The ultrasonic velocities and elastic stiffness constant in parallel and perpendicular to grain directions at different moisture content are shown in Figs. 9 and 10 respectively. As the moisture content increases, the ultrasonic velocities were found to decrease for all grain directions, and cons- equently elastic stiffness constant. A sharp rise of the ul- trasonic velocities and elastic stiffness constant was observed as moisture content decreases from 30% in- dicating fiber saturation point. Similar results were also obtained by Sakai et al. (1990). But Nakamura and Nanami (1993) obtained continuous decreasing of velocity instead of getting any sharp division line in the region of fiber saturation point. The ultrasonic velocities and elastic stiffness constant in parallel to grain were found to be greater than the perpendicular to grain directions. This result agrees well with findings of the previous workers (Kamioka 1988; Bucur 1983, 1988; Bucur and Feeny 1992). The higher acoustic velocity in parallel to grain direction may be due to the longitudinal orientation of cell along the axial directions since cell walls provide a continuous wave path. Polge (1984), on the other hand, found a strong correlation between the fiber length and ultrasonic velocity that permits the possibility to develop a non destructive methodology for fiber length. The radial and tangential directions both of which are perpendicular to grain also have substantial effect on the ultrasonic velocity and el- astic stiffness constant. The radial direction showed higher ultrasonic velocity and elastic stiffness constant compared to tangential direction.

4 Conclusions Moisture content of rubber wood affected dielectric pro- perties considerably at both low and microwave frequen- cies. The dielectric constant and dielectric loss factor increases as the moisture content increases from 0 to around 30% and thereafter increases slightly for both parallel and perpendicular to grain directions at low fre- quency. In microwave frequencies, the dielectric constant increases continuously with the increase of moisture content with concave upward. In low frequencies, the dielectric constant and dielectric loss factor increases as the frequency increases. The dielectric constant also inc- reases with the increase of frequency but dielectric loss factor decreases. The dielectric loss factor almost remains constant above 6 GHz. Fourth order polynomial equation can be used for estimating dielectric properties of rubber wood from moisture content at microwave frequencies whereas exponential equations were found suitable at low frequencies. The grain direction plays an important role for measuring ultrasonic properties. The parallel to grain direction showed higher ultrasonic velocity and elastic stiffness constant compared to perpendicular to grain. This result can be used for detecting grain defects of wood. References Bucur V (1983) An ultrasonic method for measuring the elastic constants of wood increment cores bored from living trees. Ul- trasonic 21(3): 116-126 Bucur V (1988) Wood structural anisotropy estimated by acustic invariants. IAWA. Bulletin 9(1): 67-74 Bucur V, Feeny F (1992) Attenuation of ultrasound in solid wood. Ultrasonics. 30(2): 76-81 lames WL, Hamil DW (1965) Dielectric properties of Douglus Fir measured at microwave frequencies. Forest Prod. I. 15(2): 51-56 lames WL (1975) Dielectric properties of wood and hardboard. Variation with temperature, frequency, moisture content and grain direction. Res. Pap. USDA Forest Service. Forest Prod. Lab. Madison, Wisconsin lames WL (1977) Dielectric behaviour of Douglus-fir at various combination of temperature, frequency and moisture content. Forest Prod. J. 27(6): 44-48 lames IF, Robert IR, Drum JR (1995) Nondestructive evaluation of honeycomb and surface checks in Red Oak lumber. Forest Prod. I. 45(5): 42-45 Kabir MF, Sidek HAA, Daud WM, Khalid K (1997) Detection of knot and split of rubber wood by non destructive ultrasonic method. Journal of Tropical Forest Product 3(1): 88-96 Kamioka H (1988) Effect of ultrasonic bonding materials on ve- locity and attenuation of sound in Red Lauan wood lpn. J. Appl. Phys. 27(2): 188-191 Kroner K, Pungis L (1952) Zur dielektrischen anisotropie des natureholzes im frequenzbereich. Holzforschung 6(1): 13-16 Lin RT (1967) Review of the dielectric properties of wood and cellulose. Forest Prod. I. 17(7): 61-66 Mikhailovskaja KP (1972) Investigation of moisture characteris- tics of wood electric parameters. Author's paper on Thesis for Candidate of Science degree, LTI, CBP, Leningrad. (in Russian) Nanassy AJ (1972) Dielectric measurement of moist wood in a sealed system. Wood Sci. Technol. 6:67-77 Nakamura N, Nanami N (1993) The sound velocity and moduli of elastic in the moisture desorption process of Sugi wood. Mokuzai Gakkaishi. 39(2): 1341-1348 Norimoto M, Yamada T (197o) The dielectric properties of wood IV. On the dielectric anisotropy of wood. Wood Res. 50:36-49 Norimoto M, Yamada T (1972) The dielectric properties of wood VI. On the dielectric properties of the chemical constituent of wood and the dielectric anisotropy of wood. Wood Res. 52:31-43 Norimoto M, Yamada T (1976) Dielectric properties of wood. Wood Res. 59/60:106-152 Norimoto M, Hayashi S, Yamada T (1978) Anisotopy of dielectric constant in coniferous wood. Holzforschung. 32(5): 167-172 Parker ML, Kennedy RW (1973) The status of radiation densi- timetry for measurement of wood specific gravity. Proc. IUFRO, Pretoria. 5(2): 882-893 Polge H (1984) Essais de caracterisation de la veine verte du merisier. Ann. Sci. For. 41:45-58 Rafalski J (1967) Dielectric properties of compressed Beech wood. Forest Prod. J. 17(8): 64-65 Sakai H, Minamisawa A, Takagi K (199o) Effect of moisture content on ultrasonic velocity and attenuation in woods. Ultra- sonics. 28:382-385 Skaar C (1948) The dielectric properties of wood at several radio frequencies. NY State Coll. For. Syracuse NY, Tech. Pub. 69, 36 pp Torgovnikov GI (1993) Dielectric properties of wood and wood based materials. Springer Verlag, New Work. Tiuri MK, Jokela K, Heikkila S (1980)Microwave instrument for accurate moisture and density measurement of timber. Journal of Microwave Power. 15(4): 251-254 Vermas HF (1974) Dielectric properties of Pinus pinaster as a function of its Alcohle-Benzen-soluble content. Wood Sci. 6(4): 363-367 Vermas HF (1976) The dielectric constant of solid wood sub- stances calculated with two different methods. Holzforschung. 30(3): 97-98 Venkateswaran A, Tiwari SY (1964) Dielectric properties of moist wood. Tappi. 47(1):25-28 227

COMPARATIVE STUDIES OF ELASTIC PROPERTIES OF COMMERCIAL-TYPE WOODS BUREAU OF RESEARCH AND CONSULTANCY UNIVERSITITEKNOLOGI MARA 40450 SHAH ALAM, SELANGOR MALAYSIA By AZMAN KASIM AMRAN SHAFIE AZHAN HASHIM @ ISMAIL January 2006

i) u R i: A i; O P R E . S C A « C H A CONSULTANCY Biro Penyelidikan dan Perundingan Universiti Teknologi MARA 40450 Shah Alam. Malaysia Tel : 03-55442094 / 5 / 3 / 2 Website : www.uitm.edu.my/brc Fax : 03-55442096 -i UNIVERSITI TEKNOLOGI MARA Penolong Naib Canselor (.Penyelidikan) 03-5544 2094/5 aznizl32@salam.itm.edu.my {Coordinator Penyelidikan (Sains dan Teknologi) 03-5544 2091 'zainonm@salam.itm.edu.my [Coordinator Penyelidikan (Sains Kemasyarakatan & Kemanusiaan) 03-5544 2097 rosmimah@salam.icm.edu.mv [Coordinator Perundingan (Kewangan) 03-5544^2090 shidah@salam.itm.edu. my (Coordinator Perundingan '3-5543 5120 ato@salam.itm.edu.my Penolong Pendaftar 03-5544 2092 dapeah794@salam.itm.edu.my Pegawai Eksekutif 03-5544 2098 rohani734@ salam. itm.edu. my Pentadbiran 03-5544 2093 Surat Kami Tarikh 600 - BRC/ ST. 5/3/491 11 September 2002 Encik Mohd Halil Marsuki Penolong Akauntan Unit Kewangan Zon 17 Universiti Teknologi MARA Shah Alam Tuan GERAN PENYELIDIKAN - Merujuk kepada perkara di atas, bersama-sama ini dimajukan salinan surat kelulusan menjalan

Add a comment

Related presentations

Related pages

UPM | The Biofore Company

As the frontrunner of the new forest industry UPM leads the integration of bio and forest industries into a new, sustainable and innovation-driven future.
Read more

Innovation | UPM

UPM's Biofore strategy is based on the versatile use of renewable wood biomass, combined with innovation, resource efficiency and sustainability.
Read more

UPM (company) - Wikipedia, the free encyclopedia

The material was a result of UPM's own research and product development. UPM ProFi ... The pavilion provided an opportunity to demonstrate how UPM´s wood ...
Read more

Wisa Plywood

UPM's WISA plywood products have quite a reputation in China - a reputation quite literally build on the foundations of some of the country's most iconic ...
Read more

Upm-Kymmene Wood Oy: Private Company Information ...

Upm-Kymmene Wood Oy company research & investing information. Find executives and the latest company news.
Read more

Universiti Putra Malaysia - Wikipedia, the free encyclopedia

Universiti Putra Malaysia (formally known as Universiti Pertanian Malaysia or College of Agriculture Malaya or School of Agriculture), or UPM, is ...
Read more

Timber Construction Research Group -UPM

Timber Construction Research Group -UPM Timber Construction Research Group g i . c o n s m a d e r a @ u p m . e s ... Wood Research, 56(3):2011, 297-310.
Read more

Universidad Politécnica de Madrid (UPM) - News

New woods to make new wines. Researchers from the UPM and the Forest Research Centre (CIFOR-INIA) have found a new technique of wine aging based on the ...
Read more

Advanced biofuel production | Biofuel production | UPM ...

The creation of UPM BioVerno renewable diesel is an advanced biofuel production process that has been developed in the UPM Biorefinery R&D Centre.
Read more