Basics, potential uses and limitations of the resistance-recording drilling method for tree inspection and risk management can best be understood by reviewing its developmental history and technical specifications. These matters are reflected in the different properties of the currently available models of the instrument. In addition to technical specifications, an understanding of wood anatomy and mechanical material properties is required for proper application of this method and reliable interpretation of results. Manual resistance drilling used by carpenters since the 1920s is not described here because it does not provide recordable data, and thus is not used by professional tree risk assessors.
Origin and purpose of resistance drilling
Today, it seems obvious that the penetration resistance of drilling with thin needles (drill bits) can correlate to wood condition. The evolution from concept to development of the first working-prototypes, and finally the current models, however, took quite a bit of time, research, and effort.
Based on the idea of Prof. Gersonde, working at German Federal Material Testing Institute (BAM, Berlin), during the 1970s, the German company WESERHÜTTE AG developed a machine for improving the penetration of preservatives into wooden utility poles by pushing thin needles into the wood to create channels for the chemicals. This method has since become established worldwide, and is known as “In-Sizing”. It is especially useful for species like spruce (Picea) where wood preservatives do not penetrate well, even under pressure, because the connections between the wood cells close when they die. The engineer in charge at WESERHÜTTE, Thowald Kipp, observed that some needles broke, while others penetrated quite easily. He concluded that the recorded penetration resistance could tell something about wood condition and strength. WESERHÜTTE, which specialised in large machinery, realised that it was virtually impossible to develop a wood diagnostic method for their application.
Years later, two retired leading engineers of WESERHÜTTE (W. Kamm, S. Voss), got permission from the company to apply for a patent describing the idea of needle resistance-drilling. Although this 1985 patent was later declared invalid by the German Patent Supreme Court, it triggered a flurry of development that ultimately led to the first operational resistance drilling machine (Rinn 1986), and the first series of portable drills (Rinn & FEIN 1987). In 1984, KAMM and VOSS developed a drill that recorded the penetration resistance of a thin needle using a spring-loaded gear box connected to a scratch pin. But, resonance effects of the recording spring mechanism (triggered, for example, by tree-ring density variations) led to inaccurate readings and profiles—too high in late-wood, too low or even zero in softer earlywood. Such misleading spring-resonance effects can be partially reduced by compensation springs. But this introduces stimulus thresholds into the recording system, causing another systematic and inherent error—drill resistance values below the threshold are suppressed and not displayed, or plateaus appear in the profile where the curve does not change although the local wood properties are changing.
If the spring-recorded profile drops down to zero in soft (but intact) earlywood, for example, this is likely misinterpreted as insect damage or decay. As a consequence, profiles from intact, but soft wood (commonly found in the centres of most conifers or in the sapwood of many other species) are often misinterpreted as being decayed. These mechanically recorded profiles thus had been shown as non-linear, non-reproducible, imprecise, incorrect and unreliable. And, even more importantly, such systematically wrong profiles can never be the base for a correct and reliable evaluation of wood condition and stability. Because of that, KAMM and VOSS developed a new drill in 1985 with an electrical recording mechanism, starting with a loud speaker and headphone connected to the drilling motor. They applied for a patent and asked German electric tool companies if they would be interested in buying the intellectual property rights.
FEIN of Stuttgart/Germany, the company that invented the first electric drilling machine in the 1880s, was interested in their offer, but wanted an independent opinion from a University regarding whether the needle resistance method could work. In a joint project of the tree-ring lab at Hohenheim University and the Environmental Physics Institute of Heidelberg University, the resistance drilling idea was then further developed with the aim to measure the intra-annual density variations of tree-rings for climate reconstruction (Rinn 1986-1988). Early measurements clearly showed that electronic regulation and electronic recording of motor power consumption are required to obtain reproducible profiles and reliable results that can be linearly correlated to wood density, as opposed to spring-loaded recordings which were shown to be unreliable and systematically inaccurate.
The power consumption of both electrical motors, one responsible for feed and the other for rotation of the needle, was measured individually and recorded while the needle was moving forward and backward, producing four curves per measurement. Detailed comparative analysis showed that the variations of the power consumption of the feed-motor at constant speed, and of both motors, while pulling the needle backwards did not contain significant additional information (Rinn 1989). Consequently, resistance drills from then on usually measured and recorded the electrical power consumption of a direct-current, needle-rotation motor while penetrating wood. This value is proportional to the mechanical torque at the needle, assuming the motor acts linearly. This requires a correspondingly linear type of electric direct-current motor. If the needle’s tip is flat and twice the diameter of the shaft, the torque mainly depends on density at the major point of contact at the needle’s ‘front line’ while penetrating the wood at a high rotational speed (Rinn et.al. 1991). The ratio between rotational speed and thrust was shown to be critical to achieve a high linear correlation to wood density. Only this allows a reliable interpretation of the obtained profiles.
Since the goal was to determine radial density profiles at the highest possible resolution, drill resistance must be measured at one point of the drilling path at a time. Therefore the end of the needle had to be flat. A thin centring tip was added to the end of the needle to guide it along a straight path. Because tree-ring borders are not linear-laminar (arranged in a linear or flat manner, but are concentric or even undulating, the width of the needles tip determines the radial resolution by tangential averaging, and therefore should be as small as possible. A wider needle would not allow the method to detect thin tree rings because the tip would rotate in earlywood and latewood of two or even more rings at the same time. As a result, the measured resistance cannot differentiate between the density of individual earlywood and latewood zones, and it would be impossible to identify thin tree rings. Therefore, for tree ring identification, the needles should be as thin as possible.
The needle had to be kept as narrow as possible to minimise damage to the tree being tested. Unfortunately, thin needles (<1mm) are often deflected by wood rays, knots or other wood anatomical features, and do not maintain a straight path. In addition, to be able to penetrate hardwoods, needles had to be thicker and stronger. After thousands of tests, a shaft diameter of 1.5mm with a 3mm tip was found to be a good compromise between minimising damage and maximising information in the profiles (Rinn 1989a; Rinn et.al. 1990).
All subsequent profiles used in this document have been obtained using a resistance-drill that records and is regulated electronically. It was also equipped with a flat tipped 1.5/3mm steel needle as described above.
Although the resistance-drilling method was developed for tree-ring analysis, its ability to detect decay now drives the market. Several thousand drills have been sold worldwide since 1987 by different manufacturers, but they differ dramatically in size, weight, resolution, precision, applicability and price. A linear correlation to wood density (as a mandatory prerequisite for reliable profile interpretation) was demonstrated only for electronically regulated and electronically recording device types (Winnistorfer and Wimmer 1995).
Ability to detect tree rings and decay
If the needle’s geometry follows the guidelines as described here, local wood density at the point of the needle’s tip is the main factor influencing mechanical penetration resistance. Consequently, due to density variations between earlywood and latewood, tree-ring structure and penetration angle determine the shape of the obtained profiles. Therefore, it is necessary to understand basic wood anatomical properties in order to interpret resistance profiles correctly.
In addition to needle geometry and electronic regulation, the drilling angle determines the ability to detect tree rings. Maximum resolution of tree-ring structures is provided by radial drillings so that the needle penetrates the tree-ring borders radially inward (perpendicular to the stem). The more the drilling angle deviates from 90°, the less clear tree rings appear in the profiles.
Regarding wood material properties and the cutting mechanism at the needle’s tip, local shear strength may be even more closely related to the measured drill resistance value. But, comparison with high-resolution radial x-ray density profiles showed that electronically regulated and recorded resistance-drills produce profiles linearly correlated to local wood density, revealing tree-ring density variations down to a width of about 1/10mm or even less. That is assuming the needle is equipped with a flat tip that penetrates the tree-ring borders radially (Rinn et. al. 1996).
The combination of the following wood anatomical properties determines the appearance of typical radial resistance profiles, which are correlated with density profiles:
There are many other consequences of the combination of these wood anatomical properties for tree inspection: conifer stems, for example, are commonly soft in the centre and stronger in the outer areas of the cross section, and the opposite applies to oak and all other ring-porous species. Most importantly for resistance drilling is that radial profiles derived from conifers naturally tend to drop down in the centre. This is not a sign of decay and can only be distinguished reliably from a profile drop caused by internal decay if the profile is linearly correlated to density, and the resolution is high enough to clearly differentiate between earlywood and latewood zones. If the profile drops down below the lowest earlywood resistance level, this indicates decay, a crack, or the pith as shown above.
In ring-porous wood, such as oak (Quercus), the profiles are commonly very low in the wet and soft sapwood with narrow rings, containing primarily soft earlywood. They rise up in the centre of the cross sections. Again, a high resolution and linear correlation to wood density is required to differentiate between intact but soft sapwood and decay.
Resistance profiles derived from tropical species without distinct tree rings are similar to most diffuse-porous trees from moderate climate zones, but tend to rise up slightly in the centre. Thus, in general, three types of wood have to be distinguished in terms of typical profile shape:
These typical trends have to be taken into account while interpreting resistance-drilling profiles. At the same time, the influence of the angle of the needle’s path in relation to the tree- ring borders has to be considered. If a drilling profile cannot be interpreted, a reference drilling further up the stem may help, although the number of drillings should be limited in order to minimise damage to the tree.
Needle penetration angle in relation to tree-ring borders and wood anatomical properties of the tree species determines the typical shape of the resulting resistance drilling profiles. Only profiles of electronically regulated and electronically recording resistance-drilling machines have been shown to linearly correlate to wood density, allowing the user to interpret correctly and evaluate reliably. Thus, technical properties of resistance drills have to be taken into account before purchasing and using a resistance drill.
Proper evaluation requires knowledge about the species-specific typical density trends within the tree rings and along the drilling path. Wood anatomical properties differ strongly between the three wood-type groups (conifers, ring-porous and diffuse-porous trees). These properties influence the resistance-drilling profiles, as well as the mechanical behaviour of trees in general. Ring porous trees have the highest strength values in the centre, and are much weaker outside. Conifers are just the opposite. This aspect influences all technical measurements of mechanical properties and behaviour of trees, not only by resistance drilling but also stress-wave-timing (sonic tomography) and static-load tests. It has to be taken into account while estimating strength and eventual strength loss due to decay (while moisture content plays an important role, too). Thus, high-resolution resistance drilling profiles reveal valuable information about intact and decayed areas, increment growth rates and mechanical properties of wood.