Printer Friendly

In-forest assessment of veneer grade Douglas-fir logs based on acoustic measurement of wood stiffness.

Abstract

Acoustic technology has proven to be a well-established nondestructive technique for assessing potential product performance by identifying logs with high stiffness. In an ongoing endeavor to optimize merchandizing and enhance timber value recovery, 7 second growth Douglas-fir stands of similar age class in western Oregon were sampled, totaling 1,400 trees and more than 3,000 logs. This research investigated the effects of spatial as well as internal and external log characteristics on Douglas-fir wood stiffness. Stand-level in-forest log acoustic measurements, as well as dynamic modulus of elasticity values correlated well with the actual G1/G2 veneer grade recovery ([R.sup.2] of 0.91 and 0.82, respectively) once bark removal adjustments were made. External log characteristics such as diameter and length were found to have limited predictive capability in terms of acoustic velocity and hence wood stiffness. The presence and size of branches were found to be negatively correlated to acoustic velocity readings, and the addition of the tree-length difference improved the regression model. Logs produced from the lowest part of the tree had the largest acoustic velocity, and velocity decreased in each subsequent log along the length of a tree stem.

**********

Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) is of great economic importance for the forest products industries of the United States, Canada, New Zealand, and parts of Europe (Gartner et al. 2002). The area of nonreserved timberland presently stocked with Douglas-fir ranges from 134,000 and 330,000 ha in Germany and France, respectively, to about 4.5 million ha in Canada, and to more than 14 million ha in the United States (Hermann and Lavender 1999). It is expected that international and U.S. wood product markets, especially high-quality structural lumber markets, will continue to demand Douglas-fir logs (Schuler and Craig 2003).

Timber resources in the Pacific Northwest have gradually shifted from unmanaged old growth to intensively managed young growth stands (Adams et al. 2002). Younger stands yield lower quality timber in comparison to old growth stands because of the higher proportion of juvenile wood, which in turn affects properties such as strength and dimensional stability (Gartner 2005). Moreover, harvesting younger trees increases the variability in product performance (Carter et al. 2004) and limits log producers in their ability to meet market demand for wood products (Murphy et al. 2003). As world log markets are becoming increasingly competitive and complex, forest products companies face increasing global competition not only from other wood producers but also from other industries, such as steel, aluminum, plastics, and composites (Marshall and Murphy 2004, Eastin 2005). To be competitive, forest products companies need to control costs, sort and allocate logs to the most appropriate markets, and maximize the value of the stands at the time of harvest. Beyond the silvicultural aspects of production forestry, harvesting and transportation of timber constitutes a large portion of the cost of the wood at the time it is delivered to the manufacturing facility (Marshall and Murphy 2004). Identifying stem quality in the stand (Acuna and Murphy 2006), determining its most appropriate use and processing operation, and consequently shipping the product to the right location are steps to achieving reduced costs and increased product values (Murphy et al. 2005). Optimally matching wood quality to markets can mean cutting logs for very specific end uses and classifying them into several categories or "sorts" to improve product uniformity, productivity, and profitability along the seedling-to-customer supply chain. Research into technologies for measuring stem quality attributes is progressing on a number of fronts with varying levels of success; e.g., acoustics, optical and laser scanning, x-ray, microwave, ultrasound, and near infrared (NIR) spectroscopy (Carter et al. 2004).

Wood modulus of elasticity (MOE), also known as stiffness, is one of the most important mechanical properties and is the most frequently used indicator of the ability of wood to resist bending and support loads. Wood stiffness and strength have long been recognized as crucial product variables in both solid wood and pulp and paper processing (Eastin 2005). Raw timber material is highly variable in these properties, dependent upon site, genetics, silviculture, and location within the tree and stand. Nondestructive testing (NDT) and evaluation of wood products for stiffness and strength has been proven and commercialized for years (Wang et al. 2004). Initial research trials indicate high correlation between yield of structural grades of lumber and acoustic velocity of logs processed as measured using acoustic techniques (Ross et al. 1997, Wang et al. 2001, Carter et al. 2004, Wang et al. 2007a). Segregation of logs based on tools that measure stiffness is already being used by some forest companies to improve the value of lumber recovery (Dickson et al. 2004, Wang et al. 2007b).

Good measurements and predictions of the external and internal properties of the wood in each stem are crucial to optimally match logs to markets (Clarke et al. 2002). In 2006, a study was launched to determine whether log stiffness could be successfully measured in-forest using acoustic technology and to evaluate the factors influencing its accuracy. This paper summarizes the results of the investigation into modeling the effects of spatial as well as internal and external log characteristics on Douglas-fir wood stiffness from a range of sites in western Oregon.

Materials and methods

Study sites

In summer 2006, six Roseburg Forest Products company (RFP) stands, located in the Coastal (A--near Bellfountain, D and E--near Elkton, and F--near Lorane,) and Cascade (B--near Sutherlin, and C--near Tiller) Ranges of Oregon, were harvested as part of two studies evaluating novel technologies for in-forest measurement of wood properties. In summer 2007, a seventh stand (G--near Corvallis), located within Oregon State University's McDonald-Dunn College Forest, was also harvested as part of these studies. All sites were second growth Douglas-fir stands of similar age class (50 to 70 yr) chosen to cover a range of elevations and tree sizes (Table 1). Site G had been commercially thinned on three occasions. Sites A to F had no commercial thinning but may have received a precommercial thinning. Two hundred trees from each stand were sampled, totaling 1,400 trees and more than 3,000 logs. Only veneer grade lengths were cut ( 18, 27, and 35 ft or 5.5, 8.2, and 10.7 m, respectively); no sawlogs or pulp logs were produced. Measurements included, but were not limited to, tree-length, merchantable length, diameter at breast height (DBH), biggest branch diameter on each 20 ft (6.1 m) segment of the tree, acoustic velocity of the standing tree (using the Director ST300[R] tool), acoustic velocity of the whole stem with and without the branches (using the Director HM200[R] tool), and acoustic velocity of each log made out of the stem. All acoustic measurements (on the standing tree, the whole stem and the logs produced) were performed within a 5-day window for each stand to ensure that no differences in "sample" moisture content (MC) existed; whole stem measurements and log measurements were performed within 20 minutes of felling. Approximately 100-mm thick disks were collected from a subsample (40 trees per stand) of the trees. These samples were taken at different heights from the tree, one from the base and one from the top of each log. More than 800 disks were collected. The disks were labeled, placed in dark plastic bags to prevent moisture loss, and later stored in a cold refrigerated room. Green densities, as well as sapwood/heartwood ratios, were calculated for each disk. Dynamic MOE was obtained by using the green density and the velocity of the acoustic wave through the material, expressed by the following formula:

[MOE.sub.d] = [rho]/g x [V.sup.2]

where

[MOE.sub.d] = dynamic modulus of elasticity (psi (Pa))

[rho] = density of the material (pcf (kg/[m.sup.3]))

g = acceleration due to gravity (386 in/[s.sup.2] (9.8 m/[s.sup.2]))

V = velocity of the wave through the material (ft/s (m/s))

After the in-forest measurements on the logs were completed, they were transported to a veneer mill, debarked, cut into 8 ft (2.4 m) bolts, kiln-heated, shape scanned, and peeled into veneer sheets. They were then scanned for defects and moisture, sorted into moisture classes, dried and then sorted into several veneer grades (G1, G2, G3, AB, C+, C, D, X, and XX)based on in-line acoustic measurement of wood stiffness using the mill conveyor system's Metriguard[R] model 2800 DME Ultrasonic/RF/IR veneer tester. Percent veneer recovery in all grades was calculated for each stand; because of operational limitations we were not able to track recovery from individual logs.

Acoustic velocity measurement tools

The acoustic velocity of the standing trees was measured using the Director ST300[R] (CHH Fiber-gen, New Zealand) tool, which is a time-of-flight (TOF) acoustic measurement system, developed by a research team for measuring acoustic velocity in standing trees (Wang et al. 2004a). The system and the working protocol are described in detail by Wang et al. (2007a). Findings, based on standing tree measurements in stands A to G, however, will be presented in a separate paper. The same authors also describe the resonance based acoustic tool (Director HM200[R], CHH Fiber-gen, New Zealand) used to measure longitudinal wave velocity in the logs. They point out that the latter method is a well-established NDT technique "for measuring long, slender wood members".

Data analysis

Statistical analyses of the data were undertaken following either a simple linear least squares regression analysis or a stepwise multiple regression methodology described by Ramsey and Shafer (2002). It included the following steps: graphical analysis of the data, examination of the correlation matrix, fitting of the linear model, exploration of the residuals, significance test of the variables, and improvement of the final regression model. Both SAS[R] 9.1 statistical software (SAS 2004) and the Data Analysis Tool Pak of MS Excel were used for the analysis, and ap-value of 0.05 was used as the threshold for determining significance of explanatory variables.

Results and discussion

Stand A produced the largest number of logs totaling 572 while Stand G yielded the least with 353 logs; the average log length was 9.2 m, ranging from 8.5 m for Site F to 9.5 m in Site B; HM200 acoustic velocity averaged 3.77 km/sec throughout the 3077 total logs and ranged from 2.73 to 4.69 km/sec (Table 2). The variation and distributions of the log lengths and the HM200 acoustic velocities across all sites are shown in Figures 1 and 2, respectively.

Although no particular trend was observed in terms of the spatial location of the stands, not all sites yielded the same quantity and/or quality of veneer (Fig. 3). While the overall G1 and G2 (the highest quality grades) veneer grade recovery percentage for sites A, B, D, and E was about the same (around 50%), the other three sites (C, F, and G) were considerably lower (32, 37, and 37%, respectively). This highlights the variation in internal wood properties between stands and emphasizes the need for preceding log quality information in order to make informed management decisions.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

Both average log acoustic velocity and dynamic MOE were good predictors of the subsequent veneer grade recovery from the processed logs. Initial investigation of the relationship between G1/G2 veneer grade recovery and average acoustic velocity for the delimbed logs measured on site yielded a correlation coefficient [R.sup.2] of 0.52 and the linear model was not significant with a p-value of 0.07. It was observed, however, that the two "outlier" points (Fig. 4) coincided with the stands (F and G) where trees were processed using a mechanized Waratah processor head which resulted in most of the log bark being removed; trees in stands A to E were felled, bucked and delimbed manually with a chain saw, leaving most of the bark on the logs. A study in a radiata pine stand in New Zealand (Lasserre 2005) found that removal of bark significantly increased acoustic velocity by an average 4.1 percent. The author commented that the acoustic resonance method measures a volume-weighted MOE and that bark adds mass to the stem without contributing much MOE. In order to evaluate the effect of bark removal on Douglas-fir logs acoustic measurement, the authors of the current paper conducted a short experiment in a mill yard on 81 Douglas-fir logs and found an average increase of 4.6 percent for debarked logs (Murphy and Amishev 2008). Possible explanations about the demonstrated effect of debarking on acoustic velocity readings, other than the explanation given by Lassere (2005), might be the difference in MC between outerwood and bark, or the bark/ wood interface influencing the sound wave. Once the average acoustic velocity values for sites A to E were adjusted by 4.6 percent, the regression model was significant with an [R.sup.2] of 0.91 (Fig. 5).

[FIGURE 4 OMITTED]

Green density measurements were combined with the acoustic velocity to explore the relationship between dynamic modulus of elasticity (MOE) and veneer recovery (Fig. 6). Yielding a significant linear model with an [R.sup.2] of 0.82, dynamic MOE was found to be strongly correlated with veneer grade recovery. The "outlier" (stand F), designated with a triangle instead of a diamond on Figure 6, had logs with green density values that were 5 percent greater on average than the rest of the sites.

Some forest managers in the Pacific Northwest, as well as researchers overseas (e.g., Leith Knowles, New Zealand Scion Research, personal communication) believe that there is a correlation between diameter and log acoustic velocity. Investigating the relationship between tree DBH and the acoustic velocity of the felled trees revealed that, overall, there was a very weak, statistically nonsignificant relationship with an [R.sup.2] of 0.10. There was a weak trend for the measured acoustic velocity to decrease with increasing tree DBH. The relationship between diameter and acoustic velocity measurements for fixed-length butt logs was investigated as well, thus removing the diameter-length interaction factor. In this case, the measured DBH was found to be very weakly related to the acoustic velocity readings with [R.sup.2] ranging from 0.15 for the 5.5 m butt logs to 0.19 for the 10.7 m butt logs. Also, an investigation of the relationship between tree-length and acoustic velocity showed no relationship between these two variables with an [R.sup.2] of 0.006. In other words, both DBH and log length have limited predictive capability in terms of acoustic velocity and hence wood stiffness.

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

Two models were developed to assess the impact of branches on acoustic readings once the cross-sectional areas of the measured tree branches were calculated and then summed for each tree. In the first model, simply by regressing HM200 acoustic velocity measurements with the limbs on against those taken with the limbs removed, the relationship was fairly strong (Fig. 7) with an [R.sup.2] of 0.67. Potentially influential points were identified using both the Cook's Distance diagnostic and the studentized residual statistic test (cutoff value of 3) in SAS. After examining them 28 points of the total 1,375 observations were identified as outliers and removed from the sample based on additional indications (sampling errors) regarding the validity of those measurements.

[FIGURE 7 OMITTED]

The resultant model yielded an [R.sup.2] of 0.76 meaning that the velocity measurements with the limbs on correlated fairly well with those with the limbs removed. Lasserre (2005) reported that acoustic velocity for radiata pine logs with the branches still attached was 2.7 percent lower compared to that after they had been removed. In our model, for Douglas-fir logs, we found the same change (2.7%) in velocity. This model, however, only accounts for the "presence" of branches and non-merchantable tops but not for the effect of their size (basal area difference and tree-length difference). The second model included total branch basal area for each tree as well as the length difference between total and merchantable tree lengths as explanatory variables. The resultant [R.sup.2] was 0.83 with the model, intercept and slope coefficients all being highly significant (Table 3). These results highlight the need for consistency in acoustic measurement procedures, particularly for logs with large limbs.

Depending on the site and log lengths, an average tree yielded 1 to 4 logs with most trees yielding 2 or 3 logs in our trials. The correlation between the acoustic velocity measurement taken on the felled whole stem with the limbs and tops present and the acoustic velocity reading on the logs produced from that stem was investigated. The analysis showed that logs produced from the lowest part of the tree are different from those produced from higher sections. On average, the butt log had the largest acoustic velocity (Table 4), and it decreased in each subsequent log along the length of each tree stem.

The correlations between whole tree acoustic velocity readings with limbs and tops present and those taken on the logs produced were statistically significant and quite strong ([R.sup.2] ranged from 0.60 to 0.72) for all the logs along the stem, except for the fourth log whenever such was produced. However, one might ask whether the length of those logs matters. Investigating this relationship only for the butt log it was found that longer logs had higher acoustic velocities than shorter ones (Table 5) as compared to the acoustic velocity reading on the whole tree. Those regression models were all significant as well and had coefficients of determination [R.sup.2] ranging between 0.57 and 0.61, meaning that a whole tree measurement can be a fairly good predictor of the wood stiffness for each log produced based on its length and position up the stem. These findings are consistent with what other researchers have reported for other species; Xu and Walker (2004) found that a low-stiffness wood zone forms from the base to about 2.7 m stem height in a sample of 62 radiata pine trees, resulting in limited structural usability of that portion of the tree; Xu et al. (2004) concluded that the cell ultrastructure and more specifically microfibril angle is responsible for the "rapid increase of wood stiffness in vertical direction, especially in corewood zone," reporting that in the S3 layer it dropped from over 80[degrees] at the bottom to about 51[degrees] at the top of a radiata pine butt log.

It is important to emphasize that this research is associated with several limitations influencing its power and scope of inference: due to the commercial nature of the operation and the resulting time and logistics constraints, analysis was possible and performed only on a stand average level in terms of the green veneer recovery (log level veneer recovery data were not collected); only one tree species--Douglas-fir--was sampled in the study; only one end product--green veneer--was considered and produced regardless of other external and internal tree attributes.

This research has primarily focused on the use of acoustic technology for evaluating internal properties on Douglas-fir veneer grade logs. Other researchers have found the use of acoustic tools to be viable in assessing stiffness in radiata pine sawlogs and green lumber (Grabianowski et al. 2006), structural lumber logs and boards (Carter et al. 2006) and young seedling clones (Lindstrom et al. 2004); Eucalyptus dunnii veneer and LVL logs and structural lumber boards (Joe et al. 2004). Acoustic techniques have been successfully used and implemented for nondestructive evaluation of mechanical properties of other wood products (structural lumber, poles, pulp logs, decay detection, etc.) and species as well as in tree selection and breeding based on stiffness (Huang et al. 2003).

Conclusions

The objective of this study was to determine whether acoustic technology could be used for in-forest assessment of veneer grade log stiffness and whether spatial and within tree characteristics could influence the accuracy of those measurements in second-growth Douglas-fir stands in Oregon. Acoustic velocity and dynamic MOE were found to correlate very well with the mill veneer recovery. Hence, despite the inherent variability between and within trees, acoustic technology could be a promising and valuable tool in assessing log dynamic MOE early in the supply chain even on a whole-tree basis. These measurements could be incorporated into optimal bucking decision making based on market requirements for wood stiffness by taking into account the effect of log position and length, branch size, bark presence/absence, green density. All these variables were found to influence acoustic velocity measurements, and their influence could be predicted based on linear models developed in this study.

Segregation of logs based on acoustic tools that measure stiffness is already being used by some forest companies to improve the value of lumber recovery (Dickson et al. 2004). Preliminary results from our acoustics trials show that in-forest sorting of logs is likely to lead to improvements in recovery of higher value Douglas-fir veneer grades. Our initial studies strive to address an array of questions related to the technical feasibility of using acoustic technology in forest environments. Much more work, however, needs to be undertaken to examine the costs, benefits and economic viability of this technology.

Literature cited

Acuna, M.A. and G.E. Murphy. 2006. Geospatial and within tree variation of wood density and spiral grain in Douglas-fir. Forest Prod. J. 56(4):81-85.

Adams, D.M., R.R. Schillinger, G. Latta, and A. Van Nalts. 2002. Timber harvest projections for private land in Western Oregon. Res. Contribution 37. Oregon State Univ., Forest Res. Lab. 44 pp.

Carter, P., D. Briggs, R.J. Ross, and X. Wang. 2004. Acoustic testing to enhance Western forest values and meet customer wood quality needs. In: Productivity of Western Forests: A Forest Products Focus, Harrington, C.H. and S. Schoenholtz, Eds. GTR-PNW-642. USDA Forest Serv., Pacific Northwest Res. Sta. pp. 121-129.

--, S.S. Chauhan, and J.C.F. Walker. 2006. Sorting logs and lumber for stiffness using Director HM200. Wood and Fiber Sci. 38(1): 49-54.

Clarke, C.R., R.D. Barnes, and A.R. Morris. 2002. Effect of environment on wood density and pulping of five pine species grown in Southern Africa. Pap. presented at the 2002 Tech. Assoc. of the Pulp and Pap. Industry of Southern Africa Conf., Durban, October 2002. http:// tappsa.co.za/archive/APPW2002/Title/ Effect_of_environment_ on_wood_/effect_of_environment_on_wood_.html (accessed March 2006).

Dickson, R.L., A.C. Matheson, B. Joe, J. Ilic, and J.V. Owen. 2004. Acoustic segregation of Pinus radiata logs for sawmilling. New Zealand J. of Forestry Sci. 34(2):175-189.

Eastin, I. 2005. Does lumber quality really matter to builders? In: Productivity of Western Forests: A Forest Products Focus, Harrington, C.H. and S. Schoenholtz, Eds. GTR-PNW-642. USDA Forest Serv., Pacific Northwest Res. Sta. pp. 131-139.

Gartner, B.L., M.N. North, G.R. Johnson, and R. Singleton. 2002. Effects of live crown on vertical patterns of wood density and growth in Douglas-fir. Can. J. Forest Res. 32:439-447.

--. 2005. Assessing wood characteristics and wood quality in intensively managed plantations. J. of Forestry 100(2):75-77.

Grabianowski, M., B. Manley, and J. Walker. 2006. Acoustic measurements on standing trees, logs and green lumber. Wood Sci. and Tech. 40:205-216.

Hermann, R.K. and D.P. Lavender. 1999. Douglas-fir planted forests. New Forests 17:53-70.

Huang, C.-L., H. Lindstrom, R. Nakada, and J. Ralston. 2003. Cell wall structure and wood properties determined by acoustics--a selective review. Holz als Roh-und Werkstoff 61:321-335.

Joe, B., R. Dickson, C. Raymond, J. Ilic, and C. Matheson. 2004. Prediction of Eucalyptus dunnii and Pinus radiata timber stiffness using acoustics. RIRDC Publication No. 04/013. RIRDC, Kingston, Australia. 121 pp.

Lindstrom, H., P. Harris, C.T. Sorensson, and R. Evans. 2004. Stiffness and wood variation of 3-year-old Pinus radiata clones. Wood Sci. and Tech. 38:579-597.

Lasserre, J.P. 2005. Influence of initial stand spacing and genotype on Pinus radiata corewood properties. M.S. thesis, Univ. of Canterbury, New Zealand. 107 pp.

Marshal, H.D. and G.E. Murphy. 2004. Economic evaluation of implementing improved stem scanning systems on mechanical harvesters/ processors. New Zealand J. of Forestry Sci. 34(2): 158-174.

Murphy, G.E. and D.Y. Amishev. 2008. Effects of bark removal on acoustic velocity of Douglas-fir logs. New Zealand J. Forest Sci. (accepted September 2008).

--, H.D. Marshall, and I. Conradie. 2003. Market complexity and its effect on variables that gauge the economics of harvesting production. New Zealand J. of Forestry Sci. 32(2):281-292.

--, --, and A.W. Evanson. 2005. Production speed effects on log-making error rates and value recovery for a mechanized processing operation in radiata pine in New Zealand. Southern African Forestry J. 204:23-35.

Ramsey, F.L. and D.W. Shafer. 2002. The statistical sleuth: A course in methods of data analysis. 2nd ed. Duxbury Press. Florence, Kentucky. 742 pp.

Ross, R.J., K.A. McDonald, D.W. Green, and K.C. Schad. 1997. Relationship between log and lumber modulus of elasticity. Forest Prod. J. 47(2):89-92.

SAS Inst. Inc. 2004. SAS/STAT User's Guide: Version 9.1.3. Cary, North Carolina.

Schuler, A. and A. Craig. 2003. Demographics, the housing market, and demand for building materials. Forest Prod. J. 53(5):8-17.

Wang, X., R.J. Ross, and M. McClellan. 2001. Nondestructive evaluation of standing trees with a stress wave method. Wood and Fiber Sci. 33:522-533.

--, N. Sharplin, P. Carter, and R.J. Ross. 2005. Method and apparatus for evaluation of standing timber. US Patent 20050160819.

--, R.J. Ross, and P. Carter. 2004. Assessment of standing tree quality--from baseline research to field equipment. In: Forest Products Soc. 2004 Annual Meeting--Tech. Forum, Grand Rapids, Michigan.

--, --, and --. 2007a. Acoustic evaluation of wood quality in standing trees. Part I. Acoustic wave behavior. Wood and Fiber Sci. 39(1):28-38.

--, P. Carter, R.J. Ross, and B.K. Brashaw. 2007b. Acoustic assessment of wood quality of raw forest materials--a path to increased profitability. Forest Prod. J. 57(5):6-14.

Xu, P., L. Donaldson, J. Walker, R. Evans, and G. Downes. 2004. Effects of density and microfibril orientation of low-stiffness wood in radiata pine butt logs. Holzforschung 58:673-677.

-- and J. Walker. 2004. Stiffness gradients in radiata pine trees. Wood Sci. and Tech. 38(1):1-9.

Dzhamal Amishev *

Glen E. Murphy *

The authors are, respectively, Research Scientist, Scion, Rotorua, New Zealand (Dzhamal.Amishev@scionresearch.com); and Professor, Forest Engineering Dept., Oregon State Univ., Corvallis, Oregon (glen.murphy@oregonstate.edu). Thanks are extended to Roseburg Forest Products, and especially the Oregon Logging Manager Donald Persyn, for providing the stands, equipment, tools, and personnel to assist with this study. We are grateful to the OSU College of Forestry for providing a stand in the College Research Forests, as well as funding for equipment and personnel. Funding for this study came from a USDA CSREES Center for Wood Utilization grant CO338B FIFL (in-forest log segregation technologies). This paper was received for publication in February 2008. Article No. 10459.

* Forest Products Society Member.
Table 1.--Characteristics of the seven study sites.

 Elevation
 of the DBH range of
 site Stand age trees selected
Site (m) (years) (cm) *

A 180 62 19.3 to 96.8 (52.2)
B 900 66 16.5 to 69.6 (36.3)
C 1040 56 17.5 to 79.0 (50.6)
D 220 54 14.2 to 66.8 (39.5)
E 120 51 15.5 to 59.4 (32.0)
F 290 53 16.3 to 77.2 (38.9)
G 280 72 15.0 to 78.5 (41.6)

Site Site location latitude/longitude

A 44[degrees]24.04'N/123[degrees]23.24'W
B 43[degrees]22.58'N/123[degrees]03.54'W
C 42[degrees]58.56'N/122[degrees]48.52'W
D 43[degrees]40.09'N/123[degrees]43.19'W
E 43[degrees]40.16'N/123[degrees]44.58'W
F 43[degrees]48.40'N/123[degrees]18.34'W
G 44[degrees]42.55'N/123[degrees]19.58'W

* Average DBH in parentheses.

Table 2.--Log summary statistics for the seven research
sites.
 Log count Log length HM200 acoustic velocity
 total average
 Average Min Max
Study site (number) (m) (km/sec)

Site A 572 9.4 3.92 3.03 4.58
Site B 399 9.5 3.77 2.80 4.69
Site C 458 9.2 3.46 2.73 4.23
Site D 447 9.2 3.76 2.98 4.63
Site E 395 9.3 3.84 2.88 4.48
Site F 453 8.5 3.82 2.96 4.47
Site G 353 9.3 3.77 2.78 4.32
Overall 3077 9.2 3.77 2.73 4.69

Table 3.--Summary for the regression between HM200
acoustic velocity (km/s) measurements with limbs and tops
removed and the explanatory variables: HM200 acoustic
velocity (km/s) measurements with limbs and tops on, branches
basal area ([cm.sup.2]) and length difference (m).

 Coefficients Value Standard error t stat p-value

Intercept 0.925 0.04992 18.532 2.39E-67
Limbs on velocity 0.762 0.01287 59.237 0
Branches BA -0.00048 6.72E-05 -7.1605 1.43E-12
Length difference -0.01629 8.71E-04 -18.691 2.44E-68

Table 4.--Difference between whole tree acoustic velocity
with limbs on and each subsequent log produced from that
tree (km/sec).

 Log 1 Log 2 Log 3 Log 4

Average -0.201 -0.051 0.205 0.381
Minimum -0.940 -0.950 -0.560 -0.210
Maximum 0.690 0.980 0.800 0.680

Table 5.--Some statistics for the difference between whole
tree acoustic velocity with limbs on and different length butt
logs (Log 1) produced (km/sec).

 Log 1 length (m)

 5.5 8.2 10.7

Average -0.117 -0.197 -0.207
Minimum -0.600 -0.870 -0.940
Maximum 0.578 0.130 0.690
COPYRIGHT 2008 Forest Products Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2008 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Amishev, Dzhamal; Murphy, Glen E.
Publication:Forest Products Journal
Geographic Code:4EUUK
Date:Nov 1, 2008
Words:5008
Previous Article:Grading and properties of small-diameter Douglas-fir and ponderosa pine tapered logs.
Next Article:Effects of bark content and particle geometry on the physical and mechanical properties of particleboard made from black spruce and trembling aspen...
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters