Western Juniper Drying Project

Contents:

• Abstract
• Purpose
• Need
• Background
• Prior Work
• Current Practices
• Methodology
• Results
• Discussion/ Recommendations
• Literature Cited
• Moisture Meter Correction Factors
• Suggested Kiln Schedule
• Draft Grading Rules

Abstract

Purpose

Return to Contents

Need

Return to Contents

Background

The focus group expressed an interest in better utilization of and markets for western juniper (Juniperus occidentalis). While it is highly unlikely that juniper will be the "salvation" of the Northwest timber industry, there is real potential for the creation and retention of jobs in harvesting, transporting, manufacturing, and marketing western juniper. Projects such as this one are designed to help manufacturers overcome production challenges so that they may effectively utilize western juniper.

Return to Contents

Prior Work

Herbst (1977) discusses juniper's reputation for warping and twisting when drying and states this reputation is not deserved. He refers to Kozlik's (1976) article and states that, "The wood has been air dried by entrepreneurs for making furniture and novelty products for many years. The wood, especially if cut into fairly thin boards, kiln dries very well." In 1973 and 1974, Herbst air dried 1, 2, and 4 inch thick juniper lumber using several methods and reported varying results. He reports that drying resulted in little warp or checking in the 1 inch boards, however the cracking and splitting was substantial in the 2 and 4 inch thick boards with the exception of the material wrapped in a tarp to retard drying. The Dry Kiln Operator's Manual (1991) supports Herbst's comments in stating, "Thick, wide, flatsawn lumber is more susceptible to surface checking than thin, narrow lumber".

This current project has built upon the prior work by expanding and refining the kiln schedules developed by Kozlik for material in addition to "distressed" paneling , by exploring other drying methods including dehumidification kilns and vacuum kilns, by developing moisture meter correction factors (see Table 1, Appendix A), and by testing methods such as high-temperature steaming to reduce warping due to inherent stresses in the wood. Flexibility was maintained throughout the project to allow results of initial work to provide the direction for the work to follow.

Return to Contents

Current Practices

In addition to air drying, several juniper manufacturers have contracted their drying with larger mills owning conventional steam kilns. Many of these mills have relatively large kilns (50 to 200 MBF, where MBF = thousand board feet) and either require customers to completely fill a kiln with juniper or ask that small volumes be dried with other species (primarily pine or fir in eastern Oregon), using the other species' kiln schedule. Filling even a 50 MBF kiln would require many, many months of production for most western juniper manufacturers. While the manufacturer was accumulating enough juniper material to fill even a relatively small kiln, the material first sawn would be nearly dry unless it were kept submerged in water.

Partially filling a kiln and drying juniper with other species may have its drawbacks also. In trials where juniper was dried with pine to a target moisture content of 10 percent, the juniper came out closer to 12 percent. Pulling the juniper out earlier in the conditioning cycle (end of the drying cycle where moisture is added back into the wood to reduce dryingstresses) may result in failure to adequately relieve the drying stresses in the juniper and, in addition, would interrupt the air flow within the kiln. Mike Milota, wood drying specialist for Oregon State University, mentioned other potential problems with mixing juniper and ponderosa pine in a dry kiln. If dried with pine heartwood, juniper would likely be dried too quickly resulting in excessive checking and warp. If dried with pine sapwood, juniper may become too dry, again resulting in increased checking. Dr. Milota mentions these things merely to encourage caution and states that only experience will indicate if juniper can be kiln dried properly if mixed with pine.

Return to Contents

Methodology

In the Sycan Project, a total of 354 logs weighing just over 160 tons were scaled and processed. Logs were scaled by US Forest Service, Bureau of Land Management, and Southern Oregon Log Scaling Bureau scalers. Logs were segregated into two types designated as "intermediate" and "mountain" juniper. Recovery was significantly higher for the intermediate juniper than for the mountain juniper. Average size for intermediate juniper logs was 25 to 30 feet in length and 14 to 16 inches in diameter at breast height (DBH). Average size for mountain juniper logs was 50 to 60 feet in length and 18 to 20 inches DBH. Gross Eastside Scribner scale for the 208 intermediate juniper logs was 6890 BF (BF = board feet) and net scale was 6100 BF. Gross Eastside Scribner scale for the 146 mountain juniper logs was 6890 BF and net scale was 4620 BF. Boards were graded for #1, #2, and #3 fencing by a West Coast Lumber Inspection Bureau grader.

Of the material that was not sent to Fresno Pallet to be remanufactured into fence pickets, approximately 7200 BF of 1 5/8 inch thick western juniper lumber was dried to 8-10 percent moisture content in a dehumidification kiln at North Douglas Wood Products in Drain, Oregon. Material was 2 to 8 inches in width and 6 to 12 feet in length. An additional 1600 BF of 4/4 material was air dried at the Sycan mill. The material was 4 to 8 inches in width and 8 to 16 feet in length. The lumber was stacked and stickered using a 2 foot sticker spacing. The material was then banded and stored in a covered shed.

Projects to follow the Sycan Project have included: Projects conducted by Mike Milota, Wood Drying Specialist for Oregon State University's Department of Forest Products:

• Drying Western Juniper, July 1995 (Milota and Swan, 1995)- The objectives of this research were to expand upon Kozlik's (1976) work and to assess the effect of kiln schedule and finishing on product quality. Four different kiln schedules were used to dry 4/4 and 7/4 juniper lumber in a lab-scale (2 MBF) steam kiln. One kiln schedule involved first air drying material, to slowly reduce the moisture content, then kiln drying. Drying stresses were examined during the kiln cycle and defects after drying were recorded. All material was then remanufactured into finger-jointed/ edge-glued panels. To evaluate the effects of finishing, the panels were coated on half their length with lacquer, and the other half was left unfinished. The panels were cycled through varying moisture regimes and defects were compared for the finished and unfinished ends of the panels. Lastly, moisture meter correction factors were developed for western juniper (see Table1, Appendix A).
• Steaming of Juniper, September 1995 (Milota, 1995a)- Juniper boards were subjected to increasing temperatures from 160-240°F (in 20 increments) in the presence of steam over a period of 20 hours. The goal was to attempt to determine if there was a temperature at which the growth stresses in juniper would "relax" and therefore reduce the potential for warp during drying and remanufacture. Boards were subjected to a 3-point bending configuration and deflection was measured at each temperature increment with the intent of finding a temperature at which juniper would deflect dramatically, indicating the wood had become "plasticized".
• Pre-Steaming of Juniper, November 1995 (Milota, 1995b)- Approximately 1 MBF of juniper lumber was steamed for 22 hours, then kiln dried. Another load of approximately 1 MBF was kiln dried without the steaming period. A group of air dried boards formed a third group for comparison. Boards selected from each treatment group (steamed/kiln dried, kiln dried, and air-dried) were ripped in half. Crook ("warp") was then measured for boards in each group and compared. The objective of this project was to attempt to simulate a procedure known as "plasticization" used with other species, such as plantation-grown radiata pine (Radiata Pine Association of Australia, 1992), prone to warp during remanufacture. As in the "Steaming of Juniper" project mentioned above, the goal was to determine if inherent stresses in the wood could be relieved by the use of steam, and therefore warp during remanufacturing might be minimized.
• Saw-Dry-Rip for Juniper, January 1996 (Milota, 1996)- A load of juniper lumber was kiln dried and a sample of 30 boards was ripped to 2½ inches in width (i.e, the material was sawn, dried, then ripped). Warp for these boards was compared to boards that had been ripped to 2½ inches prior to drying (i.e., the material was sawn, ripped, then dried). The objective of this project was to simulate work that has been done for warp-prone species, particularly hardwoods, by the USDA's Forest Products Lab (FPL) in Madison, Wisconsin. The FPL's program is known as the SDR (saw-dry-rip) program and has been shown to effectively reduce warp (Hallock and Bulgrin, 1978; Maeglin and Boone, 1986). Boards are sawn and lightly edged, dried in wide widths, then ripped to the desired dimensions. The theory is that where narrow boards have only the restraint provided by the stickers, wider boards have both the sticker restraint as well as the restraint provided by being part of a larger board. Again, the goal was to find a technique manufacturers could use to minimize warp during drying and remanufacture.
• Vacuum Drying of Juniper #1, February 1996- This project was conducted by Gary Coelyn at All Native Hardwoods in Roseburg, Oregon. All Native Hardwoods uses a Wood-Mizer Vacu-Kiln to dry native Oregon hardwoods. Vacuum kilns take advantage of the principle that water boils at a lower temperature in a vacuum. Therefore wood can be dried more quickly without the associated defects such as strength loss and checking that often occur when wood is subjected to temperatures above 212°F. Logs were end-coated within a day after arriving at the mill, were sawn, and the lumber was loaded into the kiln within 10 days of sawing. Lumber was dried in the vacuum kiln and defects were recorded.
• Thomas Lumber Kiln Drying Project, May 1996- This project was conducted in cooperation with the Western Juniper Debarking Project. Approximately 20 tons of logs were debarked at Klamath Veneer in Klamath Falls, and sawn into 7/4 and 4/4 x random width and length lumber by High Desert Wood Products in Dairy, Oregon. The approximately 4 MBF of 7/4 lumber was then sent to Thomas Lumber Co. in Klamath Falls to be dried. The material was dried mixed with ponderosa pine lumber. The material was planed and observations were made regarding the quality of the end product. Recovery figures were recorded.
• Fremont Sawmill Kiln Drying Project, July 1996- Juniper logs for the project were sawn by High Desert Wood Products in Dairy, Oregon. All logs had been end-coated and were dried within 2 weeks of sawing. The approximately 40 MBF of lumber was dried in a commercial kiln owned by Fremont Sawmill, Lakeview, Oregon. The material was dried by itself, that is, not mixed with other species. The lumber was then planed, graded according to draft grading rules developed by Bill Breedlove, Western Juniper Industry Facilitator (see Appendix B), and observations were made regarding the quality of the end product. Recovery figures were recorded.
• Vacuum Drying of Juniper #2, July 1996- Fourteen juniper logs were sawn at All Native Hardwoods in Roseburg, dried in a vacuum kiln, planed, and quality of the end product was recorded. Recovery figures were recorded.

Results

Project participants decided to test market green juniper lumber to established incense-cedar fence markets. This decision was based on similarities between western juniper and incense-cedar (e.g., color, texture, knot structure, aroma, and defects) in addition to rapid taper (which necessitated shorter length finished products in order to maximize recovery), current and historic market values, and the mill's lack of an on-site dry kiln. The green fence boards were remanufactured into fence pickets at Fresno Pallet in Fresno, California. Test fence panels were constructed, and after 2 days displayed an unacceptable amount of distortion. As mentioned above, results such as this were responsible for encouraging western juniper drying research.

For the 7200 BF dried in the dehumidification kiln, boards were visually inspected and no severe shrinkage or warp were noticed. Any warping that occurred appeared to be reduced for the shorter, narrower boards. Chris Cornett, OEDD consultant for the project, mentioned a potential drawback to the use of dehumidification kilns. These kilns lack the ability to condition lumber at the end of drying in order to reduce drying stresses. Conditioning helps to reduce the warp that often accompanies the stresses during resawing.

Conditions were reported to be ideal for the air dried material; winds were light to moderate and dry. No excessive shrinkage, cracking, or warp was noticed. One item of note was the speed with which the material dried. Green moisture content was approximately 40 percent for the material. After just 4 days, sapwood moisture content was approximately 30 percent and the heartwood was 22 percent. After 3 weeks, both heartwood and sapwood had equalized to around 12 percent moisture content.

Drying Western Juniper, July 1995 (Milota and Swan, 1995)- To summarize the findings:

• Drying stresses were as expected (similar to other species) during the kiln cycle. Warp and splitting in juniper lumber seem to be more a function of the way the trees grow than from stresses induced during drying. The grain patterns caused by highly tapered stems and large limbs cause more pronounced shrinking and swelling-related problems for juniper than for species with fewer and smaller limbs, and straighter stems.
• Drying had little effect on panel performance. Panels produced from material that had been air-dried, then kiln dried at a very mild schedule allowed for fewer through splits, but had as many end splits and surface checks as more severe drying methods. Knots, grain deviation (spiral, cross-grain, or deviations caused by large knots), pith (center of the tree), and what happens to the lumber before it is kiln dried (length of storage and how it's stored for example) seem to be more important than drying technique.
• Causes of defects during moisture cycling were due to (in decreasing order of occurrence): knots, grain deviation, weakness along growth ring, weakness along ray cells, manufacturing snipe, and weakness near pith.
• Finishing appeared to reduce defects.
• Knots greater than ½ inch in diameter should be avoided if possible. Cutting out knots and finger-jointing is one way to avoid the splitting and warping associated with larger knots.
• Moisture meter correction varied from +1.3% to +2.2% for resistance ("pin") meters and +1.8% to +5.2% for capacitance meters.
• Splitting problems appear to be associated with end checking of logs. This problem may be reduced by sawing logs as soon as possible after felling. If logs must be stored, they should be end-coated.
• The book, Dry Kiln Schedules for Commercial Woods (Boone, et al., 1993) provides a kiln schedule for drying western juniper (see Table 2, Appendix A). Dr. Milota feels this schedule will work well with juniper if adjusted for kiln type.

Steaming of Juniper, September 1995 (Milota, 1995a)- The high temperature study did not reveal a temperature (normally used in dry kilns) that is capable of reducing warp during remanufacturing due to stresses inherent in juniper. Deflection of loaded members did increase as temperature increased, however there was no point at which deflection increased dramatically. Juniper has been reported to bend very easily without damage; in this study, at 200 F juniper deflected 0.26 inches compared to 0.01 inches for Douglas-fir at the same temperature. Finger-jointing, and the use of shorter, narrower and thinner boards than commonly produced from western woods will be more successful in reducing the effect of growth stresses than will the use of steaming.

Pre-Steaming of Juniper, November 1995 (Milota, 1995b)- As in the steaming project discussed above, pre-steaming juniper lumber before kiln drying did not appear to significantly reduce warp during remanufacturing. One interesting point discovered in the project was that air dried lumber was no more prone to warp than was kiln dried material.

Saw-Dry-Rip for Juniper, January 1996 (Milota, 1996)- Like many other species tested using the SDR process, juniper will warp less if ripped after drying rather than before. In addition, for this project, there was much less variability in the amount of warp measured for boards that were ripped after drying rather than before drying. Based on Dr. Milota's report, Scott Leavengood, OSU Extension Agent, mentioned some things to consider before using SDR with juniper:

1. SDR involves live sawing logs rather than conventional cant sawing. Live sawn lumber is sawn through and through on the same plane whereas cant sawn logs are flipped continuously to remove the higher quality boards around the outside of the log, leaving a center cant. In SDR, the live sawn flitches are lightly edged in order to better fit into the kiln. After drying, the flitches are then ripped to the desired final width. Each manufacturer will need to decide if live sawing is the best method to use to get the highest quality lumber from juniper logs.

2. If SDR is used, the material is likely to be ripped just before some other operation such as edge-gluing. If the material warps after ripping, this may cause splitting in the final product, or at the very least, downstream material handling problems. On the other hand, if the material is sawn, ripped, then dried and it warps during drying, the warp can be handled (cut out, scrapped, etc.) before the initial remanufacturing processes.

Vacuum Drying of Juniper #1, February 1996- The operator commented that "the material appeared to dry o.k. without any noticeable degrade". One mystery still remaining from this project was the fact that the rubber seal on the kiln had just been replaced and it was deteriorated and full of cuts after drying juniper. This problem did not occur on the second run of the vacuum kiln with juniper.

Thomas Lumber Kiln Drying Project, May 1996- The schedule used at Thomas Lumber was for 5/4 pine shop lumber. Although target moisture content for the schedule was 8 percent, the pine exited the kiln at closer to 10 percent and the juniper came out of the kiln at closer to 12 percent. This raises some questions regarding the ability to dry juniper with other species. Dried lumber was tested and found to be relatively free of drying stresses, indicating the conditioning period in the kiln was successful. The lumber was planed at 450 feet per minute using a Woods 407 planer with 14 knives and no problems were reported.

Fremont Sawmill Kiln Drying Project, July 1996- The kiln schedule used at Fremont Sawmill for 7/4 juniper lumber and was adapted and altered slightly from a schedule developed by Mike Milota at OSU for the "Drying Western Juniper" project. Target moisture content was 8 to 10 percent, however the lumber exited the kiln at 10 to 12 percent moisture. The kiln operator stated he thought the material should have remained in the kiln an additional 24 hours, which would agree more closely with the time frame of the schedule developed by Mike Milota. An additional 24 hours in the kiln may have resulted in more closely meeting the target moisture content. Dried lumber was tested and found to be relatively free of drying stresses, indicating the conditioning period in the kiln was successful. Warp was common in the 14 to 16 feet long material and surface checking was common, although the checking was not apparently related to board moisture content (that is, checking was present in boards both at relatively high and low moisture contents).

Lumber was surfaced at 350 feet per minute (150 feet per minute slower than the mill would surface pine or fir) with a 14-head planer. Due to excessive "thick and thin" in the material, the planer was set to surface to 1½" instead of the target size of 1 9/16". Some tearout was noticed around knots. The boards were graded as they exited the planer using draft rules developed by Bill Breedlove (see Appendix B). Total yield was 29,813 BF with 984 BF (3 percent) in the tight-knot grade, 19,981 BF (67 percent) in the cut-stock grade, and 8,848 BF (30 percent) in the rustic grade.

Vacuum Drying of Juniper #2, July 1996- Gross scale (Scribner Dec. C) for the 14 logs was 400 BF. Logs were sawn to 1 inch x random width x 8 foot lumber. Total lumber recovery was 614 BF, or approximately 1.5 as a fraction of gross log scale. The lumber was dried to 9 to 10 percent moisture content in the vacuum kiln and then planed. Jacket boards (lumber produced from the outer, rounded-off portion of a log) contained surface cracks after drying. Some bow and slight crook were evident also. No changes were reported in quality of the lumber after planing.

Return to Contents

Discussion/ Recommendations

1) Checking/ splitting- Stresses develop in wood as it dries due to the outer "shell" of a board attempting to shrink accompanying loss of moisture while the moisture saturated inner "core" resists this shrinkage. Checks and splits occur when these stresses exceed the tensile strength of the wood perpendicular to the grain (across the width or thickness of a board). The larger the material is in width and thickness, and the more severe the drying conditions, the greater the stresses, and therefore the potential for checking and splitting. Tests have shown the tensile strength perpendicular to the grain of western juniper to be less than other common commercial softwoods (Burke, 1994) and therefore there is evidence to support the tendency of juniper to check and split during drying. Several techniques may be used to minimize these problems:

• Logs should be end-coated within hours of felling and processed as soon as possible. Moisture loss occurs immediately as the log is cut from the stump and therefore the use of end-coat will help reduce the rate of (but not stop) water loss. Wood loses moisture 10 to 15 times faster from the end-grain as from the surface of a board or log. As logs are stored, moisture loss continues, and therefore both the number of size of checks increases. These checks will then be present in the ends of finished lumber and will continue to propagate as the wood is dried and remanufactured.
• Drying conditions that are "slow and gentle", particularly in the initial stages of drying, will minimize checking and splitting. The initial settings for a kiln schedule should provide for equilibrium moisture content levels (EMC) above 11 percent (Kozlik, 1976). For air drying, Reeb and Brown (1995) recommend several practices to reduce checking (a balance must be found, however between drying slow enough to minimize checking and fast enough that mold and staining do not occur):
  
  use relatively wide stacks
  minimize space between stacks to 2 feet or less
  butt lumber edge to edge as closely as possible
  use thin stickers (½ or 3/4 inch) and
  use sheds and/or shade cloth to protect lumber from direct sunshine
  
  
• For air drying, in addition to the above practices to minimize checking, the use of an insecticide to control wood borers may be necessary.
• Where practical, manufacturing western juniper to relatively small dimensions (widths and thicknesses) will help to minimize checking and splitting. In the case of large timbers, such as fireplace mantels, the slow drying practices described above should be followed.
• Cutting out knots greater than ½ inch diameter, avoiding the pith, finger-jointing, edge-gluing, and finishing will help to minimize the impact of checks and splits on finished product quality.

a) Stresses due to growth are caused by anatomical changes controlled by the tree in response to its environment. "Abnormal" growth such as juvenile wood, compression wood, and spiral grain are all related to tree growth and often lead to warp.

It is unknown, at this point, whether juvenile wood is a significant problem in western juniper. In many other species, juvenile wood is known to cause warp, strength reductions, and other problems. Juvenile wood is limited to the first 5 to 20 years of growth. Due to very slow growth rate, it is unlikely that juvenile wood comprises a significant volume or poses a serious problem in western juniper.

Compression wood, on the other hand, is likely to be an important factor in the warp manufacturers have experienced with western juniper. Compression wood forms in leaning stems (due to wind, snow, landslides, or other factors) and under limbs. The often large and numerous limbs characteristic of western juniper, as well as the climatic conditions (high winds and snow) in juniper's native range make juniper a likely candidate for compression wood.

Spiral grain is often evident as well in western juniper logs. The shrink and swell behavior of wood is often relatively predictable, provided the grain runs parallel to the long dimension of the board (i.e., the board is relatively "straight-grained"). Spiral grain and the grain deviation caused by large limbs often result in the shrink and swell behavior being far from predictable.

Most experienced manufacturers are already accustomed to dealing with juvenile wood, compression wood and spiral grain. Since it is likely that western juniper manufacturers won't have the luxury to simply reject logs that appear to contain such growth abnormalities", steps must be taken to minimize their influence on the quality of the finished product:

• using weights (top-loads) for air or kiln dried stacks of lumber to restrain the load during drying reduces warp in many other species
• cutting out very large knots will help to minimize the influence of both the grain deviation caused by large knots as well as the compression wood often associated with the knots
• research has shown that, where practical, ripping after drying (Saw-Dry-Rip), rather than before drying will lead to less warp, and minimizing piece size will help to minimize the influence of growth stresses in general

b) As mentioned above, drying can cause stresses in wood that lead to warp. One very common such stress is known as casehardening. If not properly relieved by conditioning in the dry kiln, casehardened lumber will cup ("warp") when it is resawn.

c) Differential shrinking and swelling are also know to cause warp in wood. Wood shrinks approximately twice as much in the tangential direction as in the radial direction. In "normal" wood (that is, excluding juvenile wood, compression wood, wood with spiral grain, or other "abnormalities"), shrinkage along the grain (i.e., in length) is negligible. This difference in shrink/ swell characteristics depending on grain orientation is, in part, responsible for cupping of wide boards. Again, if the grain orientation is not parallel to the long axis of the board, as if often the case with juniper, differential shrinking and swelling can cause the board to warp. Cutting out large knots (to minimize the effect of grain deviation due to knots) and limiting piece size (such as in finger-jointing) will minimize the effect of differential shrinking and swelling.

Return to Contents

Literature Cited

1. Tight knot, no hole- Best Face
Knots not to exceed 1½" diameter. Small, tight, black knots well spaced over 70% of length are acceptable. No bark, no wane , no voids.
PRODUCTS: Flooring, cabinetry, paneling, high-end furniture, decking, novelties, #1 and #2 fencing.

2. Cut stock (Para 99) Knotty and clear
Must contain multiple cuttings over 70% of length. Minimum cutting of 2½ x 6" desired. Bark and/or wane not to exceed ½ of thickness or ¼ of width over 1/3 of length of board. No voids in cuttings.
KNOTTY: Tight knots not to exceed ½ the width of cutting. No spike knots in cutting. Small, tight, black knots not to exceed 2 per 6" cutting.
CLEAR: No knots, no wane.
PRODUCTS: Flooring, knotty paneling, clear paneling, high-end furniture, picture frames, cottage industry products; suitable for bent-wood or weave products.

3. Rustic
Well spaced bark encased knots or voids acceptable. Knots or voids should not extend to edge of board. No spore rot over 70% of any board.
PRODUCTS: Rustic paneling, rustic furniture, novelty products, #3 fencing.

NOTE: This material will contain clear sapwood cuttings, however the "Fall Down" would exceed 50%.