Thesis (PhD (Forest and Wood Science))--Stellenbosch University, 2007. / ENGLISH ABSTRACT: There has been a reported reduction in the durability of creosote treated, wooden utility poles in South Africa in recent years. Several factors could have been responsible for this. In this study, Eucalyptus grandis pole drying schedules currently used, methods of measurement of moisture content (MC) after drying and drying defects were investigated. Relationships involving drying defects and MC gradient before treatment which may help in the development of simple nondestructive methods of assessing defects such as surface checking, honeycomb and collapse, and MC gradient were also explored. The long term goal of this study was to increase the durability of treated poles by avoiding unacceptable drying defects and MC values before treatment. The results of this investigation are presented in the following chapters: an introduction motivating the aims of the investigations (Chapter 1); a review of literature relevant to MC and drying defects in sawn or round wood (Chapter 2); ready for press manuscripts on MC and drying defects in kiln-dried E. grandis poles (Chapters 3 to 7) and a general conclusion that links up chapters 3 to 7, and recommendations (Chapter 8). The auger drill method gave reliable MC values when samples in increments of 25 mm depths into the pole were taken. However, when single samples of 50 mm and 75 mm depths were considered, the auger drill MC measurements were unreliable. It was concluded that SABS SM 983 (2000), which specifies taking single radial auger drill samples of depth 70±5 mm at pole mid length, is not a reliable method of measuring MC in a kiln-dried E. grandis pole. Correlation results indicated that tree growth factors such as sapwood depth, green MC and heartwood percentage were related to final drying defects and may be used as criteria for pole sorting before kiln drying. Also, the number of valleys per unit length (VPUL) of the circumference at the theoretical ground line (TGL) was positively correlated with honeycomb and closed surface checks, implying that VPUL as a parameter of the pole circumference profile can be used to assess invisible, internal defects. The dry bulb temperature (Tdb) of 80oC, used to dry poles in industry, was too high and resulted into unacceptable levels of drying defects in kiln dried E. grandis poles. In addition, the drying period of 8 days was too short to attain an acceptable MC gradient in poles. It is, therefore, recommended that: To use the auger drill method to reliably determine the moisture content of a pole, samples in increments of 25 mm should be taken. MC measurements should be made at the most critical zones of a pole such as the TGL and not higher up since there is normally considerable MC variation in the longitudinal direction. It is also important to consider sampling more than one position on the pole circumference to cater for the MC variation in the tangential direction. For poles with large sapwood depths i.e. >15mm, mild drying conditions at the beginning of a drying run should be used, since such poles may be more susceptible to surface checking. Poles with large heartwood percentages should also be dried with suitable kiln schedules, i.e. with low Tdb to minimise honeycomb and collapse. In general, Tdb lower than 80oC should be considered in order to reduce the defects to acceptable levels. Poles should be dried for longer than 8 days even at Tdb as high as 80oC to reduce the MC to acceptable values and gradients. Since only three schedules were tested, more drying schedules should be investigated to make reliable conclusions about the effect of schedule on drying defects. In addition, more poles per schedule should be considered in order to obtain statistically reliable results. The relationship of pole diameter and drying defects also requires further investigation. Since a limited sample was used to test for the effect of source of poles and drying schedules on defect correlations, further studies should be done on sufficient samples to come up with more reliable conclusions. It is also necessary to further investigate the possible variation in correlation of VPUL and internal defects on a large sample of poles from specific sources and drying schedules. / AFRIKAANSE OPSOMMING: Die afgelope paar jaar is waargeneem dat kreosootbehandelde transmissiepale 'n afname in duursaamheid toon. Verskeie faktore kan hiervoor verantwoordelik wees. In hierdie studie is huidige Eucalyptus grandis oonddrogingskedules en voggehalte- (VG) bepalingsmetodes vóór en na droging asook drogingsdefekte ondersoek. Verwantskappe tussen drogingsdefekte en VG-gradiënt vóór behandeling wat kan meehelp om nie-destruktiewe metodes te ontwikkel om defekte soos oppervlaktekrake, interne barste en instorting en VG-gradiënt te bepaal, is ook bestudeer. Die langtermyn doelwit van hierdie studie is om die duursaamheid van behandelde pale te verbeter deur onaanvaarbare drogingsdefekte en voggehaltes vóór behandeling te vermy. Die resultate van hierdie ondersoek word in die volgende hoofstukke bespreek: 'n inleiding waarin die doelwitte van die onderskeie ondersoeke motiveer word (Hoofstuk 1); 'n oorsig van die literatuur relevant tot VG en drogingsdefekte in gesaagde en rondehout (Hoofstuk 2); persklaar manuskripte oor VG en drogingsdefekte in oondgedroogde E. grandis pale (Hoofstukke 3 tot 7) en 'n algemene gevolgtrekking wat bevindings uit hoofstukke 3 tot 7 saamvat gevolg deur aanbevelings (Hoofstuk 8). Die augerboor-metode kon betroubare VG-waardes lewer indien monsters van 25mm diepteinkremente in die paal geneem is. Augerboor-voggehaltemetings met enkel 50mm en 75mm monsters was egter onbetroubaar. Daar is bevind dat SABS SM 983 (2000), wat 'n enkel radiale augerboormonster tot op 'n diepte van 70±5 mm in die middel van die lengte van die paal spesifiseer, nie 'n betroubare VG-bepalingsmetode vir oondgedroogde E. grandis pale is nie. Die resultate het aangetoon dat boomgroeifaktore soos spinthoutdiepte, groen voggehalte en die persentasie kernhout aan drogingsdefekte verwant is en as basis vir sortering van pale vóór oonddroging gebruik kan word. Ook kon die aantal valleie per eenheidslengte (e. VPUL) van die omtrek by die teoretiese grondlyn (TGL) positief met interne barste en geslote oppervlaktekrake korreleer word wat impliseer dat VPUL as parameter van die omtrekprofiel van die paal gebruik kan word om onsigbare, interne defekte mee vas te stel. Die droëbaltemperatuur (Tdb) van 80oC wat in die industrie gebruik word, was te hoog omdat dit onaanvaarbare vlakke van drogingsdefekte in oondgedroogde E. grandis veroorsaak het. Daarbenewens was die drogingsperiode van 8 dae by 'n Tdb van 80oC te kort om 'n aanvaarbare VG-gradiënt te lewer. Die volgende word derhalwe aanbeveel: Om betroubare VG-resultate d.m.v. die augerboormetode te verkry, behoort monsters in inkremente van 25mm geneem te word. VG-metings behoort in die mees kritiese dele van 'n paal gedoen te word byv. by TGL en nie hoër nie omdat daar normaalweg beduidende VG-variasie in die longitudinale rigting is. Dis ook belangrik om op meer as een posisie op die omtrek te meet om vir die variasie in die tangensiale rigting voorsiening te maak. Ligte drogingstoestande behoort aan die begin van die drogingsproses by pale met groot spinthoutdieptes, bv. > 15mm, gebruik te word omdat sulke pale meer geneigd tot oppervlaktekrake is. Pale met hoë kernhoutpersentasies moet ook met geskikte drogingskedules, bv. met lae Tdb, gedroog word om interne barste en ineenstorting te verminder. Oor die algemeen behoort Tdb laer as 80oC oorweeg te word om die omvang van defekte by aanvaarbare vlakke te hou. Selfs by Tdb so hoog as 80oC behoort pale langer as 8 dae te droog om die VG tot by aanvaarbare waardes en 'n gradiënt te bring. Omdat slegs drie skedules evalueer is, behoort meer skedules ondersoek word om betroubare gevolgtrekkings oor die effek van skedule op drogingsdefekte te verskaf. Addisioneel behoort meer pale per skedule oorweeg word om statisties betroubare resultate te verkry. Die verwantskap tussen paaldeursnee en drogingsdefekte benodig ook verdere ondersoek. Aangesien 'n beperkte aantal pale gebruik is om die herkoms en drogingskedule met defekte te korreleer, behoort verdere studie op genoeg pale gedoen te word om meer betroubare gevolgtrekkings te kry. Dit is ook noodsaaklik om die moontlike variasie in die korrelasie tussen VPUL en interne defekte op 'n groot aantal pale van spesifieke herkoms en drogingskedules vas te stel.