but they are shielded to a considerable degree in comparison with the alternative plan of common piling where the end stickers are placed at the ends of the shorter lumber and the long ends are allowed to project by themselves. It must be remembered that in both endpiled and cross-piled loads there is a sweep of air past the ends of the load. In an end-piled compartment kiln the hot air rises from the pipes between the loads as well as in their length, and when a progressive kiln is end piled this is true, also, of that portion of the track under which the heating pipes extend. In a cross-piled progressive kiln the longitudinal circulation is stronger past the ends of the lumber than through the middle of the loads. Even in a crosspiled compartment kiln, with the heating pipes directly under the middle of the load, projecting ends are subjected to faster drying, though the differences may not be so great here as in the other types mentioned, because the air at the sides of the kiln is apt to be cooler air, about to descend and repeat its cycle.

Another advantage of the box-piled load is that it offers a longer effective air barrier than the load of similar lumber with projecting ends. If two such loads are butted on an endwise track, there is less air space between the ends for idle air movement; and if cross piled, the end reaches closer to the side of the kiln and here, too, leaves a narrower passage for idle air.

In box piling two lengths of lumber together it is well to have the second sticker line back from the end just the difference in length, so that it will take the end of the shorter lumber. The same plan can be carried out with three or four different lengths to be piled together, by so disposing the stickers that each board end will reach a sticker.

Another point to remember is that the stickers get kiln-dried as well as the fumber and more thoroughly after repeated trips through the kiln. A sticker just out of the kiln often has a bad effect on green lumber and sometimes will pull enough moisture out at the sticker crossings to check the lumber, even before it is moved into the kiln. This can be avoided in most cases by letting the stickers lie in open pile for a day or two before using them again, giving them a chance to reabsorb a little moisture from the air. It has been noticed in this connection that the checking of a board near a sticker takes place particularly where wide stickers are used, which retard the drying rate in the area in contact with the sticker, with the result that the adjacent area dries more rapidly and checks. Subsequently these checks may extend under the sticker. Incidentally this drying influence of stickers on lumber is a good thing if it does not go too far, because it tends to counteract the usual tendency to slower drying at sticker crossings.

The use of lumber for pile crossers for itself is fortunately unknown in dry-kiln piling, as the width of piles does not accord with usual lumber lengths. Short-dimension stock and squares are the only material that can be self-crossed for kiln-drying, and this is usually carried out in such fashion as not to be objectionable.

It is unfortunate that such studies as have been made of kilndrying do not give any results of comparative tests as between box piling and plain piling of mixed lengths, of the use of chimneys and no chimneys, or of placing stickers at the ends or back from the

ends on any of the species of lumber usually kiln-dried. It is usually possible to decide which method will give the best results, but whether the better results are worth the added effort should be definitely determined.

COMMON FAULTS IN COMPARTMENT-KILN OPERATION

Compartment kilns are usually operated at temperatures and humidities which make it highly uncomfortable to enter the interior of the kiln and are largely operated on what can be learned near the doors. Kiln samples are usually inserted where they can be readily reached, and hand instrument readings are taken a few feet in from the doors. There are some heat losses even through good doors, well maintained, and a certain amount of steam leakage around them; and it is probable that both heat and humidity are higher in the kiln interior. Recording instruments have an advantage over hand thermometers, in that their bulbs can be installed at a greater distance in from the doors. An effort should be made to determine what differences exist in heat and humidity between the kiln interior and the easily reached space near the doors, so that due allowance may be made in kiln control. This difference is favorable as far as kiln samples are concerned, because when stock near the doors is adeqautely seasoned the general contents are apt to be down to the same moisture content or a little lower. If the air where the readings are taken is materially cooler than in the interior, it will probably also show a higher humidity reading than the average for the kiln and thus lead unwittingly to too dry operation.

Another difficulty occurs when kilns are operated without night supervision and without automatic control, and particularly when the conditions of night steam supply differ from daytime conditions. Usually the hand controls are set for a little higher humidity over night in order to be on the safe side, and yet a morning examination of the lumber would show that surface drying and perhaps some surface checking had occurred.

If the steam sprays have been left partly on, a drop in steam supply will reduce the amount of spray, and less directly of heat if a reducing valve is used; and the humidity in the air of the kiln drops quickly, while the heat in the lumber will carry evaporation forward for a considerable period. A drop in boiler pressure will directly affect the high-pressure steam spray but may not affect the reduced pressure of the heating. The air in contact with the lumber is warmed by contact and its evaporative effect is increased, hastening surface evaporation, and this effect will continue as long as the wood remains hotter than the air. A fall in steam supply may produce the same effect if there are no steam sprays on, if for any reason ventilation to the outside is increased and air humidity reduced inside the kiln; and the ventilating stacks sometimes develop a stronger draft at night because the outside air is cooler than during the daytime. Cutting the ventilating dampers well down at night will tend to hold both heat and humidity in the kiln in the face of lowered steam supply and secure all the seasoning without undesirable surface effects that the heat in the lumber and the amount of steam supply will produce. If, however, the night supply of steam can be

depended on, all that is necessary is to allow a safely high humidity that will be needed when the air temperatures in the kiln lead instead of lag behind the temperature inside the wood.

Cross-circulation compartment kilns should have an equal supply of heat at all points throughout their length, but in many of them there is a gradation of heat from one end to the other. This is because the heating pipes run the entire length and are supplied with steam at one end, the other being the discharge end. The entering steam is a little the hottest, particularly with low-pressure steam. Such a kiln has to be operated with regard to the cool end, leaving the lumber in until adequate drying is secured there, and it is often possible to grade the charge somewhat, putting the lumber that will take longest to dry in the hot end of the kiln; or, if the stock runs uniform in that respect, loads intended for the slow end can be piled more openly so they will dry more quickly and thus not hold up the entire charge. This will slightly reduce the total footage in the charge, but total production from the kiln will be increased by the faster drying per charge. In kilns with a return-pipe system the discharge header is at the same end as the supply header and the pipes run through and back again, thus equalizing the heating effect throughout the kiln. In long kilns the length is sometimes divided into separately valved sections for control and equalization of the heat.

This decrease of temperature with increasing distance from the steam supply is due to a decrease in steam pressure inside the pipe, and steam flows only toward points of lower pressure. At 10 pounds pressure saturated steam has a temperature of nearly 240° F.; at 5 pounds, a little over 225° F.; and atmospheric pressure, 212° F. If the supply of steam is throttled down, the difference gradient along the pipe becomes steeper, because a larger proportion of the steam becomes condensed in the nearer length of pipe, causing a greater difference in heat in the kiln. A more uniform moderate heat can therefore be provided by dividing the pipe into sections, each with a separate valve, and shutting off some of these sections to reduce heat, supplying the remaining sections with more nearly a full throttle of steam.

A kiln-heating system, whether compartment or progressive, has a steam supply but no steam discharge; all the steam supplied is condensed to water, which must be trapped off with an efficient steam trap. If the trap fails to work, the water backs up into the pipes, reducing their temperature; and accumulation of water in the pipes because of insufficient drainage slope has the same effect. The system may become airbound instead of waterbound if there are not adequate air valves to allow the steam to drive the air out, or if these valves do not work. Both steam traps and air valves usually work best on a rather narrow range of steam pressures; a system adjusted for exhaust steam would not work well on live steam unless the pressure is reduced to an equivalent of the regular supply. There may be no air at all in the pipes when the steam is shut off, but condensation of the steam remaining in the pipes would produce a vacuum, inducing air leakage around joints and valves; and this air must be driven out again when steam is turned on. Most of the steam traps used in dry-kiln work discharge the water only against atmospheric

pressure; if it is desired to return this hot water to the boiler for boiler feed, additional special appliances are usually employed.

Air seasoning tends toward uniformity of seasoning if the stock remains sufficiently long in pile, because the drier stock stops drying when it has reached balance with the air, and the heavier stock goes on seasoning until it has reached that point. The heavier stock dries somewhat more rapidly, because it attracts better air circulation, cooling the air more rapidly by evaporation from the wood and causing it to sink away, while fresh air is drawn in to replace it. This effect occurs also in kiln-drying but has less result because of the shorter period. There is therefore less uniformity of seasoning in kiln-dried product than in thoroughly air-died product. This difference can be reduced during the early stages of drying down to the fiber saturation point. Stock having an initial moisture difference of 20 per cent moisture content or more can be brought to a difference of 5 per cent or less at 25 per cent moisture content by holding heat and moisture during the early stages, in relation to the heavier stock, and thus holding back the drying of the drier portions of the charge. Occasional steaming treatments seem also to have a leveling effect, increasing the drying rate most, on the greenest stock.

These facts are known and often made use of in factory kilndrying of air-dried lumber, which may have heavy variations in moisture content at the beginning of the kiln-drying process. They have a bearing also in sawmill kiln-drying green from the saw, because of the differences in moisture content of green boards in the same species. It is true particularly in the softwoods that the heavier boards will usually be of sapwood, which kiln-dries more rapidly than heartwood, so that this difference tends to equalize in kiln-drying. This result will be made more certain if, in operating a compartment kiln, samples of the heavier as well as the lighter contents of the charge are included and if the schedule of operation is guided chiefly by the heavier pieces until they are brought in harmony with the rest.

It will sometimes be noticed that kiln units with side walls exposed to outside air, or the outside kilns of a battery with one side so exposed, are affected more by humidity conditions of the outside air than the units in the interior of the battery. This is because of the loss of moisture through the outside walls, which occurs more rapidly in fine dry weather than in damp or rainy weather. A great deal of moisture often escapes in this way, working through the walls and evaporating from the outside, and humidity inside the kilns must be compensated for this loss. It can be reduced by keeping the inside of the kiln well painted with a waterproof asphalt paint. Condensation on the inside of the walls will then run down to the bottom and should be drained off.

OPERATION OF COMPARTMENT KILNS

Compartment kilns are usually operated on a definite schedule; in many instances the schedule used is that which has been worked out by the Forest Products Laboratory for that particular species, with perhaps slight local modifications. A compartment kiln has to be kept under proper control in order to operate successfully. In many cases kilns are unintentionally not operated on a definite schedule,

because of inaccurate recording or controlling instruments, improper placing of instrument bulbs and of moisture samples, and inadequate circulation. If a kiln has shortcomings, it is apt to be too sluggish air circulation, though this is usually better in compartment kilns than in progressive kilns. Unless there is a good movement of air through the lumber, the drying results are not uniform throughout the load.

Probably the chief fault in sawmill operation of compartment kilns is overcrowding them. This may mean piling stock too closely, or it may mean operating on too severe and speedy a schedule. A schedule may be speeded up somewhat, especially on softwood stock thicker than 6/4, if occasional steaming treatments are given to relieve surplus surface-drying effects; but sometimes the steaming treatment is not continued long enough to have proper effect and the stock too soon gets in condition to need another steaming. Whatever is gained in speed of drying by such methods is usually lost in increased degrades in the kiln-dried product, which develop in the planing mill, even if not apparent in the rough kiln-dried stock.

STEAMING TREATMENTS, PARTICULARLY IN COMPARTMENT KILNS

A preliminary steaming is generally recognized as being a good way to start the kiln-drying process in both hardwoods and softwoods, though more essential for some hardwoods, and is used with both progressive and compartment kilns, and perfectly practical with either if the progressive kiln is provided with steam-spray pipes in the green end. Some of the steam may work back toward the dry end if no cut-off curtain is used, unless the dampers are left open to continue the longitudinal air current, but thorough steaming usually requires closing the dampers. The lumber toward the dry end will not be likely to be injured by such steam as reaches it; more probably the contrary. Steaming is especially indicated to start kiln-drying of stock which has previously been air-dried, in order to soften the outer portion of the lumber as well as to heat the interior. Steaming is also desirable for lumber green from the saw as the quickest way to heat the lumber through thoroughly and start off the kiln-drying process.

In order to be commercially successful, kiln-drying must be fairly speedy, and operation on this basis often produces some excess of surface dryness; this is the beginning of casehardening, and a steam treatment at that stage is desirable to moisten the dry surface. These intermediate steam treatments are impossible in a progressive kiln but entirely feasible in a compartment kiln, and their use on stock, as mentioned before, thicker than 6/4 will result in speedier drying with less degrade from checking, loosening of knots, casehardening, and warping.

Both initial and preliminary steaming should be sufficient to accomplish the desired result and no more. In both one desired result is to heat the lumber through, and by the time this is accomplished in the intermediate steam treatments the other desired effect of moistening the outer wood will probably have been accomplished. Recommended steamings seem to vary from one to four hours for each inch in thickness of the stock. A. E. Krick, of