FAILURE OF PINUS RADIATA VENEER IN TENSION ACROSS THE GRAIN

Transcription

1 120 NOTE FAILURE OF PINUS RADIATA VENEER IN TENSION ACROSS THE GRAIN A. MICHELLE CARRINGTON, ROGER B. KEEY, Deprtment of Chemicl nd Process Engineering, University of Cnterbury, Christchurch, New Zelnd MARTIN D. PUGH, Deprtment of Mechnicl Engineering, University of Cnterbury, Christchurch, New Zelnd nd JOHN C. F. WALKER School of Forestry, University of Cnterbury, Christchurch, New Zelnd (Received for publiction 2 June 1994; revision 29 August 1994) INTRODUCTION Preliminry tests on smll pieces of veneer under room conditions were used to determine the tensile strength of Pinus rdit D.Don perpendiculr to the grin in the rdil direction. These tests were undertken to provide filure informtion for devising loding tests t elevted tempertures with veneer sheets. The use of veneer hs ttrctions in tht, firstly, repliction of smples is simple s the properties of flt-sliced veneer chnge reltively slowly long veneer sheet or between djcent veneers nd, secondly, the lods t filure re reltively smll, so mking ny test pprtus comprtively simple in construction. MATERIAL AND METHOD The mteril used ws nominlly 0.6-mm-thick veneer of P. rdit from 28-yer-old pruned mnged stnd in Southlnd, New Zelnd, nd ws produced by slicing in tngentil direction to the rings. Smples were cut from sheets to give visully homogeneous test pieces of erly wood from 15 to 200 mm long nd 50 mm wide. Thickness of ech smple ws mesured with vernier guge nd rnged between nd 0.6 mm. The test pprtus ws designed to llow vrible lod to be pplied to piece of veneer in tension. The veneer ws securely held in plce between pir of clmps, with the bottom New Zelnd Journl of Forestry Science 24(1): (1994)

2 Crrington et l. Filure of Pinus rdit veneer in tension 121 clmp ttched to the lower edge of n I-shped frme. This frme supported system of two pulleys over which single cord connected the top clmp to pn on the other side of the frme through spring blnce. The lod ws incresed by progressively plcing smll weights (steel shot) in the pn nd ws mesured by the spring blnce. A smple of veneer ws plced between the clmps in the testing rig nd the lod ws incresed until the veneer filed. The corresponding stress ws clculted ccording to the pprent cross-sectionl re of the veneer, nd not the re occupied by cell wll tissue. Tensile strengths perpendiculr to the grin were determined for both green veneer nd ir-dried veneer t 12% moisture content. RESULTS Initil tests on 100-mm-long veneer smples showed some vrition in the clculted stress (Fig. 1). The coefficient of vrition ws Filure loctions were sttisticlly rndom long the "guge length" of the smples nd ny pieces on which filure occurred within the clmps were rejected. There ws no indiction of preferentil filure djcent to the clmps, which would indicte stress concentrtions due to the clmps themselves. However, the filure stresses mesured were much smller thn those reported in tests on similr species, Pinus sylvestris L., in which lrge smples 10 mm thick hd been used (Siimes 1967). These results my lso hve been ffected by the different loding method used which, being lod-controlled rther thn deflection-controlled, possibly influenced the nture of crck growth in the specimen. The work done by Siimes (1967) ws with wood loded t constnt rte of bout 50 kp/minute with tensiometer redings being tken every 5 or 10 kp. Bier (1986) lso gve much higher vlues for filure stress for P. rdit in bending tests perpendiculr to the grin. Smples 100 mm long x 180 mm wide x 45 mm thick hd filure stress of 3.60 MP. Initilly it ws thought tht the low vlues for filure stress were due to pprent defects in the veneer, possibly microchecks which hd developed during slicing, but when smples 1.20 T 1.00 CM ^ z w 0.60 in o o o ^ o ~ o Green Air-dried Smple length/ (mm) FIG. 1-Filure stress of ir-dried nd green Pinus rdit veneer smples.

3 122 New Zelnd Journl of Forestry Science 24(1) of different lengths were tested, the shorter the smple the greter ws the stress t filure (Fig. 2). When the pprent defects on the surfce of the veneer were exmined under scnning electron microscope, they were not checks s nticipted. The rounded depressions on the surfce of the veneer (Fig. 3) indicted tht they were rndomly dispersed xil resin cnls, between 0.3 nd 0.4 mm dimeter, with some on the surfce nd others embedded in the veneer. It ws lso possible to see some smll rdil resin cnls, 0.04 mm dimeter, 3.00 T ST 2.50 E E s i.5o fo N <$ Green (this work) Siimes, Air-dried (this work) Siimes, Smple length / (mm) FIG. 2-Vrition of filure stress of ir-dried nd green Pinus rdit smples with smple length. FIG. 3-An xil resin cnl on the surfce of the veneer.

4 Crrington et l. Filure of Pinus rdit veneer in tension 123 completely penetrting through the veneer (Fig. 4). The embedded xil resin cnls would probbly be the dominnt weknesses in the veneer; their size suggests tht the true vlue for the filure stress is greter thn the pprent filure vlues for 100-mm-long smples consequence of the cnls wekening the test smple. If the veneer ws 0.6 mm thick nd resin cnl ws 0.3 mm in dimeter, there would be 0.3-mm skin of veneer t the thinnest plce. Such mteril would exhibit two-fold decrese in the pprent tensile strength mesured over the whole cross-section. FIG. 4-Rdil resin cnls. Occuption of some of the volume by resin cnls cuses decrese in the effective crosssectionl re, giving rise to stress concentrtions in the mteril. As the elstic stress concentrtion fctor is pproximtely 2.2 (Benhm & Crwford 1987), the mximum stress djcent to the resin cnls is likely to be pproximtely 4.4 times lrger thn the pprent stress (lod divided by pprent veneer cross-section). Elstic nd plstic deformtions will led to reduction in the vlue of the stress concentrtion fctor. However, the true filure stress will still be lrger thn the pprent stress, s indicted by the dt. Scnning electron microgrphs were tken of the filure surfces (Fig. 5). Both xil nd rdil resin cnls ppered on the filure surfces s rounded grooves. A piece of veneer tht hd filed t very high lod hd no trce of n xil resin cnl (Fig. 6), but some rdil resin cnls were visible on the filure surfce. This suggests tht the smll rdil cnls rther thn the lrger xil cnls were the strength-limiting defect in this cse, resulting in the higher filure stress. Brrett (1974) showed tht size effects in smples in tension perpendiculr to the grin cn be explined by the filure behviour being governed by the "wekest link". The pprent stress is function of smple volume for geometriclly similr specimens. In

5 124 New Zelnd Journl of Forestry Science 24(1) FIG. 5-The frcture surfce with remnnts of n xil resin cnl. I^-^^*,!^^^WS^: :;^!^- ^''~5fe^ FIG. 6-The frcture surfce without remnnts of n xil resin cnl. our tests, the xil resin cnls formed the wekest link, nd the pprent stress exhibited liner reltionship with smple volume on log-log plot (Fig. 7). CONCLUSION The differences in filure stress with vrition in smple length cn be explined by the nturl distribution of xil resin cnls within the veneer. Filure stress might be expected to be influenced by the probbility of finding one or more xil resin cnls within the veneer (which increses with smple length), by the nturl distribution nd the dimensions of xil

6 Crrington et l. Filure of Pinus rdit veneer in tension i I I I I I I i i i i i i I I I I I I Volume of smple / (mm 3 ) FIG. 7-Vrition of filure stress of ir-dried nd green Pinus rdit smples with smple volume. resin cnls, nd by the loction of the xil resin cnl reltive to the veneer surfce (Gordon 1978). Short veneer smples (<20 mm) my be free of xil resin cnls, nd their strength should pproximte tht for "cler" smples. In thicker smples resin cnls re still present but their effect on the strength of the smple hs been reduced s the rtio of cnl dimeter to smple thickness is reduced. The limiting vlues of the filure stress correspond to reported vlues for P. sylvestris reported by Siimes (1967). This conclusion is demonstrted by the filure surfce of the veneer, since ll filure surfces except those with high stress t filure showed trces of xil resin cnls (Fig. 5). Therefore, the cuse of most filures ws the presence of the embedded xil resin cnls. The originl premise for this work ws the suggestion tht tests of thin veneers would obvite some of the difficulties in smpling nd nturl vrition tht occur when working with lrger specimens. However, the distribution of xil resin cnls mens tht unless the specimen is very short (<20 mm) this pproch would be imprcticl. ACKNOWLEDGMENTS This work ws supported by reserch contrct UOC302 from the New Zelnd Foundtion for Reserch, Science nd Technology. The Deprtments of Mechnicl Engineering nd Plnt nd Microbil Sciences t the University of Cnterbury llowed used of their scnning electron microscopes, nd the smples were cut from sliced veneer sheets provided by New Zelnd Veneers Ltd, Christchurch. REFERENCES BARRETT, J.D. 1974: Effect of size on tension perpendiculr-to-grin strength of Dougls-fir. Wood nd Fiber 6(2): 126.

7 126 New Zelnd Journl of Forestry Science 24(1) BENHAM, P.P.; CRAWFORD, R.J. 1987: "Mechnics of Engineering Mterils". Longmn Scientific & Technicl, Wiley, New York. BIER,H. 1986: Rdit pine in cross-grin bending. The New Zelnd Journl of Timber Construction 2(1): 16. GORDON, J.E. 1978: "Structures or Why Things Don't Fll Down". Plenum Press, New York nd London. SIIMES, F.E. 1967: The effect of specific grvity, moisture content, temperture nd heting time on the tension nd compression strength nd elstic properties perpendiculr to the grin of Finnish pine spruce nd birch wood nd the significnce of these fctors on the checking of timber t kiln drying. The Stte Institute for Technicl Reserch, Finlnd, Julkisu 84 Publiction.