Semi-destructive method for in-situ evaluation of compressive strength of wood structural members.
A method of testing the compressive strength of wood using small-diameter cores was developed. A special core drill was used to cut 5-mm-diameter cores from wood members. Cores were drilled perpendicular to the fibers and were tested in concave jaws that applied the compressive force parallel to the fibers. Correlation between the strength of cores and the strength of American Society for Testing and Materials (ASTM) specimens in compression and tension was established. While the correlation between compressive strengths of the cores and ASTM specimens was relatively good, the correlation between core strength and tensile strength was weak.
Historic buildings contain a considerable number of timber structural members. The timber elements can be either load bearing or may only have architectural significance. When deciding on preservation, repair, reconstruction, or replacement of defective components of historic buildings, preservation is always preferred. To select an optimal solution, basic parameters such as strength, modulus of elasticity (MOE), density, wood species, moisture content, and extent of possible damage must be determined.
Structurally, the most important parameters are strength, MOE, and extent of damage. Strength parameters cannot be obtained by direct measurements due to the destructive nature of strength tests. Over the years, various nondestructive methods have been developed to evaluate strength and stiffness of in-situ wood structural members. These include ultrasound, transverse or longitudinal stress waves, low-energy x-ray, and other methods utilizing various physical phenomena. The advantages of indirect methods are that they are relatively fast, easy to use, and do not require extraction of a specimen from the member.
Indirect methods provide a good estimate of properties over the length or depth of the member, which is especially valuable where there is no direct access to the member. In indirect methods, some physical (nondestructive) quantities, such as ultrasound velocity or stress wave velocity, are measured and related to elasticity and strength (destructive) properties through correlation. The drawback of the indirect methods is that the correlation between nondestructive and destructive parameters may be weak and affected by other factors such as moisture content and density. This drawback weakens the reliability of the indirect methods and favors direct measurements.
Semi-destructive methods include increment cores, pins driven into the wood, a resistance drilling technique, and a core drilling technique described in this paper. The increment core method can be used for direct measurements of density and moisture content. The resistance drilling gives only relative density information. Neither of the two methods can be used for direct strength measurements. Their disadvantage over the indirect methods that measure the global member characteristics (such as stress wave methods) is that the information extracted from a member pertains to local areas.
Given the spatial variability of the material, the sampling technique is critical and may affect results. Also, procedures must be established to correlate small, clear specimen values to properties of full-size members. The advantage of semi-destructive methods over the indirect methods is that they provide very specific information about the location of interest, such as the base of a column or bearing area of a beam.
The major drawback of all nondestructive methods is the relatively inaccurate information generated about the strength parameters of a structural member. Such parameters are essential in structural design (or redesign) and evaluation of existing wood buildings. The economic and cultural importance of historic buildings and monuments justifies the use of more sophisticated methods that can provide more reliable information about the structural integrity of the object.
[FIGURE 1 OMITTED]
Core drilling techniques have been used in wood structures for some time and the purposes vary from assessment of wood preservation (1) to measuring the wood density, or shear strength of the glueline (2,3). The technique presented in this paper uses small-diameter cores to evaluate the compressive strength of the wood along the fibers. The idea behind using the small-diameter cores is that the hole created by the tool is smaller that most knots present in a timber and so will not compromise the strength of the member. The hole can be easily plugged, which will restore some of the compressive strength. Clearly, the method will not be suitable for elements of small cross sections. Experiments were conducted to find the correlation between small cores subjected to radial pressure along fibers and the American Society for Testing and Materials (ASTM) specimens traditionally used to establish the wood strength in compression and tension along fibers.
[FIGURE 2 OMITTED]
Material and methods
A special core drill was designed to cut cores 5 mm (0.197 in.) in diameter. The outside diameter of the drill was 9.5 mm (0.375 in.). The inner diameter of the hollow drill gradually increases from the tip to the shank end to prevent the frictional forces that can disintegrate the extracted sample. The tool creates a 10-mm- (0.394-in.-) diameter hole in the wood member. High strength tool steel was used for the drill to achieve high quality cores. The cores must be consistent to eliminate the effect of the variation in core geometry on the core strength. The drill with a typical specimen is shown in Figure 1.
The testing device consists of a pair of cylindrical jaws that are used to apply compressive force along the concave surface of the core in the direction along fibers. The length of the core should be sufficient to include a reasonable number of annual rings to avoid possible bias. The ASTM standards (4) require a 50- by 50-mm (2- by 2-in.) or 25- by 25-mm (1- by 1-in.) cross section of the specimen. European standards use a 20- by 20-mm (0.79- by 0.79-in.) cross section. Therefore, the length of the cylindrical specimen should not be less than 20 mm to minimize the bias that can be introduced by differences between mechanical properties of latewood and earlywood. The schematic of the testing apparatus is shown in Figure 2. The cylindrical specimen is inserted into the jaws so that the fibers are parallel to the direction of load. The orientation of the specimen in the device is crucial and will significantly affect the results. The concave surfaces of the device distribute the load over the surface of the specimen. The surface is only partially loaded due to the gap between the surfaces of the compression device. The gap is necessary to allow the specimen to deform. Two miniature linear variable differential transducers are used to measure deformation between the concave jaws and this information can be directly correlated with the MOE in compression. A load-deformation curve is also needed to establish the yield point of the material.
Five species were tested to establish a correlation between the strength of a cylindrical specimen and the ASTM specimen: white ash, sugar maple, redcedar, southern yellow pine, and red oak. Ideally, the relationship between the core strength and ASTM-specimen strength should be only a function of the specimen geometry and test method, given that all other parameters (density, orientation, moisture content, etc.) are equal for both methods. However, due to the variability between specimens (within board), the data will always be scattered. The species were selected to cover a wide range of densities and strengths to increase the span of the data. The location of individual specimens in the 50- by 50-mm (2- by 2-in.) sample is shown in Figure 3. Cylindrical specimens were drilled in the vicinity of the ASTM specimens to reduce the effect of variability between specimens. All specimens were equilibrated at 22[degrees]C and 65 percent relative humidity environment. ASTM tension and compression tests were performed and correlated with the cylindrical specimens. The strength of the cylindrical specimen in compression was calculated as
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
[FIGURE 5 OMITTED]
[sigma] = [F.sub.max]/[l X d]
[sigma] = compressive strength of the specimen
[F.sub.max] = failure load
l = specimen length
d = diameter of the core
Note that this is an apparent strength because the testing device will not distribute the pressure equally over the core surface and the end effects cannot be suppressed.
MOE cannot be directly calculated from the load-deformation curve because of the nonuniform strain field. However, the slope of the load-deformation curve can be correlated with the value of MOE in compression.
Results and discussion
Figure 4 shows a typical load-deformation curve. The lack of a distinct yield point is due to the surface restraint, Poisson's effect, and the shape of the specimen, which create a multiaxial stress situation. The value of the [F.sub.max] was established as an intersection of the straight line extension of two quasilinear portions of the load-deformation path and this is shown in Figure 4.
Figure 5 shows the correlation between the ASTM compressive strength and strength of the cylindrical specimen along the fibers. As discussed above, no perfect correlation can be achieved. However, the correlation coefficient ([r.sup.2] = 0.89) indicates that the core drill method may be feasible for establishing the compressive strength of the clear wood. Figure 6 shows the correlation between the slope of the load-deformation curve shown in Figure 4 and MOE in compression along fibers obtained from the ASTM tests. The correlation coefficient ([r.sup.2] = 0.76) reflects a relatively good relationship between the two quantities. The deviation of experimental values from the straight line can be attributed to the variability between the specimens and to possible deviation between the angle of the force and fibers in cylindrical specimen tests. Theoretically, the slope of the load-deformation curve for the cylinders should be perfectly correlated with MOE despite the non-uniform strain field.
The correlation between compressive and tensile strength may be somewhat questionable due to the different nature of load and failure modes (Fig. 7). As expected, the correlation coefficient is relatively low ([r.sup.2] = 0.67). One of the reasons for low correlation is the fact that the ASTM tensile test requires the cross section of the specimen to be 4.8 by 9.5 mm (3/16 by 3/8 in.). This leads to variable results, especially for species with wide annual rings where the bias due to sampling variable amounts of earlywood and latewood may be significant. The bias perhaps can be eliminated by using specimens of larger cross sections or by a significant increase in the number of tests. Several core values are associated with one value of strength in tension as documented in Figure 3. Although the cores were cut in the vicinity of the "necked" section of the ASTM specimen, the variability between individual cores due to the location in the board is clearly visible in Figure 7.
[FIGURE 6 OMITTED]
[FIGURE 7 OMITTED]
The method can give a good estimate of compressive strength and MOE of small, clear specimens when cutting ASTM specimens is impossible. The relatively small outside diameter of the core drill will not cause structurally significant damage to in-situ members. The hole created by the drill can be easily plugged, thus somewhat preserving the integrity of the member. Possible drawbacks of the method are: 1) the relatively large number of specimens needed to establish a reliable estimate of the clear wood strength; and 2) determination of allowable stress based on clear wood values. The number of cores drilled per member should be sufficient to calculate the statistical parameters needed to establish a basis for allowable stress calculation. This may not always be possible. To estimate allowable stress values for in-situ members, one must know the extent and size of various defects. Therefore, a visual inspection of each member is necessary.
It appears that the core-drill method can complement other nondestructive techniques such as the stress wave method, which determines the global characteristics but may be weak in prediction of actual strength, moisture content, or density.
The proposed method can be used in conjunction with other nondestructive techniques to establish compressive strength and MOE of in-situ wood structural members. While no isolated method can provide a complete picture of the wood elements in question, the combination of various techniques can increase our confidence in established values. The positive attribute of the method is that it establishes the strength of the wood in a destructive way while not reducing the strength of the member. The major drawback is the local character of obtained data. If the tested specimens do not adequately represent the population (properties of the member) erroneous conclusions can result. This drawback can be somewhat eliminated by combining the method with other "global" methods.
The core method can be potentially valuable in establishing compressive strength, bearing capacities, direct density and moisture measurements, microscopic evaluation, and other measurements that require small samples from an in-situ member.
[c]Forest Products Society 2003.
Forest Prod. J. 53(11/12):55-58
(1) Sievert, R. 1996. Anwendung der DIN 68 8000 Teil 3. Praxisorientierte Beispiele und Umsetzungsemphehlungen fur die innerbetriebliche Arbeit. Bauen mit Holz. 2.
(2) Gaudert, P., and M. N. Carroll. 1973. A plug test for plywood lumber composites. Forest Prod. J. 23(1): 31-36.
(3) Selbo, M.L. 1962. A new method for testing glue joints of laminated timbers in service. Forest Prod. J. 12(2): 65-67
(4) American Society for Testing and Materials (ASTM). 2000. Construction wood. Annual book of ASTM Standards, Sec. 4, Vol. 04.10. ASTM, Philadelphia, PA.
The author is a Professor, Dept. of Wood and Paper Science, Box 8005, North Carolina State Univ., Raleigh, NC 29695-8005. This paper was received for publication in April 2002. Article No. 9475.
*Forest Products Society Member.
|Printer friendly Cite/link Email Feedback|
|Publication:||Forest Products Journal|
|Date:||Nov 1, 2003|
|Previous Article:||The incorporation of paper deinking sludge into fiberboard.|
|Next Article:||Estimated stress and friction distributions on tool rake face in the medium density fiberboard cutting process.|