US2325055A - Thermal insulation - Google Patents

Description

THERMAL INSULAT C. C. HERITAGE ION Filed May 22, 1940 2 Sheets-Sheet l Way? - 20 DEF/ BER .f

.sA/.E WATER A40/.s7- 23 38 7J/P5672 F/L TE?? afer 44 4f wer l Dl? Y 27 73 Jg LA PJ a FF F2735]- URY www F/.z/FF 46 am FL UFF FL UFF my F/.UFF

z a@ 24 2 Y 49 WE-74 WE T WET WE T WET WE T FEll'J FELT FELT FELT FFLT FELT 4 f 5f e2 54 58 URY 7K? RV 01W r DRV FELT FEL T FELT FELT FELT FELT 29 37 g5 is /i 0 am mr my my [my wer fieyefeaw" 67am? C. /rz'zaye if 0722@ v July 27, 1943. c, C, HERWAGE 2,325,055

THERMAL INSULATION Filed May 22,- 1940 2 Sheets-Sheet 2 arveg Patented July 27, 1943 v THERMAL msULA'rIoN Clark C. Heritaga-Cloquet, Minn., assignor to Wood Conversion Company, Cloquet, Minn., a

corporation of Delaware Application May 22, 1940, Serial No. 336,495

8 Claims. (Cl. 154-44) The present invention relates to thermal insumasses and to insulation formed from such fibermasses into loosely felted masses such as are advantageous for thermal insulation and sound deadening.

The primary use of fibrous insulation is for thermal purposes, but it is also useful acoustically. The major characteristics of the fiber herein described are with reference to thermal uses, but some of the characteristics apply to the assembly of bers into mats or masses, and more particularly to handling and to felting the fiber into spaces, independently of the use.

Heretofore, the applicants assignee has provided a chemically-cooked vegetable fiber especially prepared for packaging into sealed slabs for thermal insulation of refrigerator cabinets. A special machine was provided Afor packing the fiber. Spaiford U. S. Patent No. 2,073,655 describes the machine, its use, and to a degree,.the ber referred to. It is therein explained that theloosened fiber is dropped intol a hopper to fall in front of a reciprocating'ram. The ram forcesit forward into a spout or nozzle of rectangular cross-section. This discharges a continuous rectangular bat which is somewhat elastic. 'I 'he bat is fed endwise into a. receiving bag which is thereafter to besealed or closed to provide.an insulating slab. The bat is a loosely feltedstructure subject to compression and to a spring back" after release from compression. A length of freshly made bat which is longer than the slab when not compressed, is pushed into the bag endwise thereof, and the bag sealed over it. The immediately available elasticity causes the bat to press endwise on the bagY and its other faces with a view'to assuring a full bag and avoiding creation of void space by settling. However, it has now been discovered that this elasticity is notl permanent, and that the customary initial degree of felting is not sufficient l which was for a long time incomprehensible, was

the loss of elasticity and eventual settling. This is now understood and the knowledge of it has lead to the present invention. Practice of the invention employs a system to evaluate the fiber, permitting the provision of fiber to meet desired objectives when used in accordancewith the evaluation.

The above mentioned prior use is a good example of some requirements of bulk fiber for insulation and other uses. It must felt well and have high insulating value for a minimum of material. It must felt well to retain a packed formation. It must have elasticity and resistance to compression. For a pneumatic felting process it must have a moderately low resistance to felting, or in other words, felt easily. 'I'hese properties are vital -to packing bulk fiber into a permanent insulating form of predetermined density. These qualities of the mass call for fibers which per se are iiexible and resilient, and which in amass have suitable properties to form a permanent and thermally efficient felted mass.

It is to be understood that many properties of bulk ber are referred to a mass of the fibers, rather than to individual fibers, in spite of the fact that the properties of the mass arise in the individual fibers. A massof fibers exhibits properties .which depend on all the fibers of the mass and also upon the manner in which the fibers arev associated. When the density of a' heterogeneous mass of fibers is changed, the properties of the mass are changed, even though the properties of individual fibers are unchanged. It follows, therefore, that any one property taken for consideration is not absolute. For example, the thermal insulation value of one batch of fiber is dependent upon how it is assembled and at what density it is assembled.

It is also known as a result of experience in developing the present invention, that the properties of a fiber per se vary according to the kind of natural lignocellulose used, for example whether it be aspen or jack pine;.vary withlthe treatment of the lignocellulose material before, during or after the deiibering process, for example the exposure to one or more of the following conditions: heat, water, moisture, chemicals; and vary with mechanical conditions encountered in debering. In short, every variable which affects the bers per se, has an effect on the mass properties of the ber.

It is the general object of this invention to produce from lignocellulose as a raw material, a fibrous product which has properties inherent' in the bers per se, so combined and of such value, that-they may be employed in controlled felted aggregates to give thermally eicient insulation, which is non-settling when no binder is used and the union in the mat is dependent upon the qualities of the mass as to felting and elasticity. v

It is an object of the present invention to convert lignocellulose materials into a brous product having a predetermined combination. of desirable qualities.

It is a particular object of the invention to regulatethe berizing process to control-the discharge thereof so as to give initially a high de'- gree of the desired ber size'distribution. i

It is an object of the invention to produce a low cost brous material suitable for eicient thermal insulation when placed in random arrangement at a predetermined density without the use of binders.

It is another object of the invention to produce ber with a controlled resistance to felting t A particular object of the invention vis to con-` trol some properties'of the ber-by ,controlling particle size distribution of the ber'in forming Variousl other Vand ,ancillary objects land advantages' of the :invention will become apparent fromthe following description and explanation of the invention.' The invention is accompanying drawings in which:

Aexplained by reference to the Fig. 1 is a plot of "free-footage? and thermal conductivity of a series of bers.

Fig. 2 is a plot of free-footage and 'thermal conductivity of a more general relationship with the curve of Fig. 1 imposed upon it.

no extraction of constituents. It is not indicated that the heat and steam effect no chemical change, but the absence of liquid water and of active'cooking chemicals, and the short time. mitigate against chemical changes comparable to chemical cooking, or even water-softening steps. Use of water in a process extracts some material and alters others. In particular it gelatinizes cellulose material where frictional Work is done on it in the presence of water. These and other changes in the substance of the ber greatly alter the properties exerted by a felted aggregate of the bers. Therefore, one important step in controlling properties is to minimize change in substance. Suitable processes are to be found in British Patent No. 15,105, of 1911, Asplund U. S. Patent No. 2,008,892, MacMillan U. S.' Patents No. 1,344,180, No. 1,476,032 and No. 1,515,062, the Banbury apparatus (see Robinson No.'2,142,334, referring to Serial No. 101,014, -led September 16, 1936), the Respats apparatus of Respess No. 1,976,297, the explosion processes of the Masonite Corporation, Laurel, Mississippi, and others. Noting that the Asplund machine must bespecially operated and controlled in order to produce ber within the scope of the present 'invention,.it is to be understood that this is like? wise true of other machines and processes. The

Vquicker'the process, the less the change.

In one phase of the'Britishv patent process wood chips and steam are fed'to rotary grinders, such asl stones with spaced faces, between which the wood is reduced to ber. In the Asplund process, the ber Ais similarly ground, but with metal plates, 'ina heated gaseous atmosphere, such as of steam,v above 212 F. and at a temperature where the wood or other lignocellulose is softened. The'higher the temperature the softer the wood,

and hence the process is more quickly carried out,

and the wood more thoroughly debered.

the vproduct may be rigid slivers, such as ragged forms resembling in size a tooth-pick or a match. These? are referred to as sliversfandthe range [of size may be large. As the spacing is narrowed,

Fig. 3 is a chart of courses which ber may take from a debering process to a nished mat, which courses inuence the properties of the ber for felt formation, and also inuence the properties of the mat.

Fig. 4 is a front view of apparatus for use in determining the coarseness modulus of ber.

Fig. 5 is a side view of the apparatus of Fig. 4.

Figs. 6, 7 and 8 are respectively cross-sections of Fig. 4 on the lines B-B, 1-1, and 8 8, showing the ber-filtering means.

The processes of reducing wood vor like lignocellulosic material to ber are many and dierent. For -thepresent invention a process is preferred which reduces the raw material to ber with little Waste as non-brous material, with a minimum of chemical change, and with substantially the slivers are-reduced in number, and exible ber bundles increase and mayfpredominate. Bundles are natura1 aggregates of ultimate bers which are not completely separated. They may be secured together at one end, at the middle, or overall, or have mixed relations. Many bundles are found which are frayed at an end like a broom, presentingsingle or a few joined ultimate bers. From twoy to several hundred ultimate bers may constitute a' ber bundle as it is here 'dened for the purposes of the present invention. vIt is also to be understood that a sliver which is rigid maybe fractured or bruised so that it is exible to be considered herein as a bundle.

When the spacing of the grinding elements is made still closer, the bundles become less, slivers maydisappear entirely, and ultimate bers increase. As the setting may be made still closer particles.

non-brous material which lies naturally between bers.

' bination of such properties for selected uses.

To practice the present invention, it is first desir able to analyze the iiber by dividing it into fractions of predetermined sizes to determine a definitive coarseness modulus. From informa- 'tion thus obtained it may be predetermined what adjustment of the machine will deliver a product within a selected range oi' coarseness modulus, the'whole of which product may be most emciently used as bulk fiber useful for insulation. By such means a bulk `iiber may be obtained meeting predetermined specifications.

The infomation obtained from use of one kind oi raw material does not necessarily dictate the same operating practice and results for another kind or another species of the same kind. Properties of the mass derive in part from properties of the iiber, and these vary with the source of raw material. Nevertheless, the general procedure for practice of the invention is the same. It is, therefore, not considered necessary to give duplicative disclosures for several materials. To explain generally, it is stated that suitable bers derived from aspen have a range of coarseness modulus from 60 to 160. while for jack pine the range is from 130 to 250. Nevertheless, certain other properties may be identical. In order to illustrate the invention, a discussion of the procedure is'given with reference to jack pine, reduced to chips, and treated to the action of an Asplund machine in steam at a-temperature of around 365 F. (150 lbs. gauge pressure).

'I'he chips, particularly of old wood, are-wet slightly with water so that they are more moist than air-dry, thus to prevent burning in the machine. Green wood is sufficiently moist for use without adding moisture. It may be mentioned that control of moisture at this point is one variable for eilecting property-value control. The chips in the machine are heated, ground and may be discharged from the machine in about 45 to '7 0 seconds, more or less. The rate oi' production is another control variable. 'I'he discharge may have from about 50% to 150% moisture on the dry wood basis. The material is preferably dried directly without immersion in water. Water treatment is another control variable.

A result of direct drying is the inclusion of nes and dust. These may be eliminated by taking up in water, filtering, and preserving the large sizes. This adds to the cost by the extra drying which must follow. It also removes watersoluble material. Also under some conditions the;

drying after contact with water tends to lay down ends of fibers in bundles, or bundles or fibers cling together as agglomerate masses, and alter the insulation value. Therefore, for the tests herein reported, the material has not been immersed and then dried, but dried directly, unless otherwise specied.

The nes and dust are undesired in bulk iiber for insulation, for reasons discussed later. In handling, for example, these escape and may fill the air. By adding a bonding agent to the chips or other form of raw material entering the machine, the dust and ilnes may be anchored to large particles. Rubber latex or other treating agents may be so used, as` described in my copending 'application U. B. Serial N0. 334,761, led May 13,

1940, and Serial No. 336,577, filed May 22, 1940. The use of a treating agent must be considered as a factor changing the character of the fiber.

It has been determined that where the same wood wax, asphalt, montan wax, and rosin Also there may be used lfire-proofing agents, mold-preventatives and other agents. Of these rubber latex is preferred to minimize dustiness. It gives the fiber other desirable properties for felting into insulation mats.' Fire-proong agents, moldpreventing agents and other materials have been used with latex, as well as combination of material.

In proceeding to test the properties of the iiber discharged from the machine under the specific conditions above described, the machine was diierently controlled in a series of runs, producing a relatively iine product in run No. 1, and

a relatively coarse product inv run No. 8, with gradation in between. The particle size distribution of each run was determined and expressed numerically on some selected but arbitrary standard. The one specifically referred to at this time results in a value reported herein as coarseness modulus.

DETERMINATION OF COARSENESS MODULUS.

The principle employed involves a standardized screening of the material through a series of graded perforated plates, or screens, or both, of successive neness, whereby a fraction No. 1 of coarse material is obtained, followed by frac'- tions of successively iiner material. There are numerous classification apparatuses available for such operations using fixed procedures. -Some operate wet, some dry, and some by air. The present results were obtained on a wet classifier herein referred toas the Laurel classifier (Shop Order No. 4336) provided by the Machine and Foundry Company, Laurel, Mississippi. Itl provides iirst and second plates ywith drilled holes, and then two wire screens.

l The classification results are expressed as folows:

Fraction No.-

Retained by Ms inch drilled holes. u inch drilled holes. 32 mesh screen. mesh screen. Passes 80 mesh.

4 P4 (past No. 4) -I I 4 X percent of fraction No. 1 3 X percent of fraction No. 2 2 X percent of fraction No. 3 1 X percent of fraction No. 4

giving total of (a number) which is the coarseness modulus, frequently referred to as C. M. This is a standard method of weighting results.

A full description of the classier and of its operation is 4given later in the specification.

In Table I, the data and the corresponding coarseness modulus of the Jack. pine fiber are given for said runs No. 1 to 8, for which the machine setting was recorded, and is arbitrarily represented herein by the Run No.

ATable I TABULATING PERCNT BY WEIGHT AGAINST FRACTIONS AND CORRESPONDING PROPERTIES Run. No.-

Fraction No.-

1 1.0 4.4 8.2 19.0 32.4 39.6 49.6 53.8 4.2 ,6.8 10.6 13.4 11.8 10.8 8.8 7.6 42.6 40.2 33.2 27.8 21.0 21.4 17.8 16.8 29.8 27.8 29.2 24.0 21.6 16,212.8 11.0 22.4 20.8 18.8 15.8 13.2 12.2 11.0 10.8 131.6 146. 2 160. 2 196. 8 228. 6 249. 8 273. 2 292. 6 "K" value .231 .239 .235 .239 .238 .249 .257 .257 Freefootage 8.64 8.16 8.9 8.95 8.15 6.76 6.5 5.8 SpccicelastcityKa.. 118 75 82 841.5 67 '53 38 24 Absolute elastctyME .373 .399 .331 .341 .363 .478 .478 .517 Specific fulting Kir..." 82 65 78 29.5 18 1 1 8 .0 Abslutefelting Mr... .347.358 .327 .26 250 .281 271 .0

,V'Table I also gives the insulation value.(K

value) fof. thefiber measured at 3.8 pounds per cu. ft. density, which corresponds to the` foot- 'age" of 3.15 board `feet per pound. This footage islthe one chosen in the testing procedure forv the"specific values KE and KF used herein. The given K" value represents heat transmission in B. t. u., per sq. ft., per 1inch of thickness, per 1 F. temperature differential, per hour, when the test is made on a sample of l-inch thickness ywith the fiber in direct contact with the hotand cold plates, and is a standard unit well known in the art. The table also gives free-footage, which represents an extrapolated hypothetical state where the fiber is unfelted :except by the force of gravity. In plainer words, lit represents the. flufilness of'thefiber, and is mathematically expressed as a volume per unittions. But practically, it is signicant when the fibers of different coarseness modulus are being compared, if the fibers are produced by the same conditions, as for example by the same machine. The distribution is an average result of all the conditions. Considered mathematically, the said average is the average of the probabilities. The same machine, or the same process tends to present the same probabilities when the conditions are the same, and to vary from sameness in a regular way related to changing conditions. Accordingly, the coarseness modulus is an indication of both changed conditions and changed result, and isa comparable value.

The higher the coarseness modulus, the more.

coarse the fiber mass will be.

From wood, the mass may have slivers, bundles,

vend indicates lowered thermal eiilciency for insulation.

On the other hand,ultimate fibers favor insulating efficiency, land decreasing coarseness modulus indicates increasing content of ultimate fibers, and high thermal efdciency. It ,follows that ultimate fibers is the optimum to be desired. However, other properties and other considerations enter the picture. A mechanical defibering action cannot readily be carried out to produce 100% ultimate fibers. As-the process is adjusted to eliminate slivers and bundles and increase ultimate fibers, some ultimate fibers are broken to make dust and fines. The process may be adjusted to produce mixtures exhibiting a desirable combination of fiber sizes, and the coarseness modulus is a simple key for production control.V

The matter of handling fiber masses, in baling, shipping, forming mats, felting, and the like, is related to the so-called Idustiness of the fiber mass. For the-purposes of description dustiness or flying dust of a mass of fibers, is the content which can leave a felted matrix ofthe -fibers the screen-e'ect of the matrix would be finer and ultimate fibers would not appear as' dust.

It follows,.therefore, that in converting wood'or j the like to bulkinsulating fiber, without loss of a portion, or addition thereto to control the properties, it is very important to avoid a high percentage of coarse, and also to avoid a high percentage of fine Where there is considerable of coarser material. The present invention permits the defibering of wood or the like to be controlled to yield a total'product falling within a given* range of particle size distribution, in orderto secure a high degree of practical utility commensurate with a highAdegree of'insulation eiliciency. i

With respect to the Asplund machine operating on lignocellulose materials, such as Wood, the invention contemplates adjusting the machine to yield a total product falling with a range of coarseness modulus. Therefore, the thermal conductivity is not the sole criterion on which to base the desirability of the fiber. However, it is significant to check the fiber chosen for other desirable characteristics.

The coarseness modulus referred to in this description is a convenient index for particle size distribution. It is specic for the classification system employed, and that is arbitrary and in practice not a universally accepted standard. However, one system may beconverted fairly accurately to another, using the known laws governing particle size distribution. Fiber masses being 12 x 12 x 1- inches.

with the same coarseness modulus may vary in important mass properties. l

There is another property of fiber masses which appears to be an aggregative property, comprehending other properties, such as the elasticity of the fibers per se, and the felting characteristics of the ber. Reference is made to the application of Anway U. S. Serial No. 313,920, filed January 15, 1940, which discusses means and methods for determining the so-called compressive properties of fiber masses. Free footage is one of the compressive properties and is the preferred term for the said aggregative property. It is dened in terms of a density-factor of a ber mass at zero compression. Actually, the free footage is the reciprocal of density, and is given in units of board feet per pound of fiber, a board foot The value is one obtained by extrapolation by the the procedure of the below-quoted paragraphs'. These paragraphs are quoted from said Anway U. S. Serial No. 313,920, led January 15, 1940, and are sufficient herein to give the full procedure for determining the compressive properties, including free-foot- PRACTICAL PROCEDURE inder of 8-inches diameter and 22 grams of the same loosened fiber in a 3-inch diameter cylinder presenting the same character of interior surface. Place the 'cylinder on a platform scale adjusted to `read zero (or else the Weight of the liber, the diiference being negligible). With a at plate compress the large column of fiber at the rate of 1 inch in 32.4 seconds until it is 2 inches high (a reciprocal density of 2 board-feet per pound). Record the pressure on the scale as reading P pounds, and the corresponding height of the column, at approximately each half-inch of advance, or a suflicient number of observations to locate a mean straight line. Record the final pressure Q at column height of 2 inches. Upon attaining said last mentioned height, immediately remove theplate to release the column. Measure the height (board-feet per pound) attained by the column after such spring-back, and record as H inches (or board-feet per pound).

Then substitute the smaller cylinder, and compress at the same rate until the column height is 2 inches (board-feet per pound), and record the pressure as R pounds.

The following data are thus obtained:

Table IV Board-feet per Foota e on 8inch 3-inch i pound or footage" sprng-back 841mb column column 2 H Q R. (The series) (The series P) Select semi-log graph paper having a log scale vertically beginning at log of 1, (which is zero), and mark equal divisions of footage units horizontally. On the chart, plot the series of obpoint tothe free-footage and label the line Csline,' or spring-back line. Draw a vertical line at H footage and label sprung-back footage. On the latter line, subtract the ordinate values of the intercepts with the C-line and Cs-line, and record the difference. Mark this difference on the log scale as a point on the H-footage line, draw a straight line from the point tothe free footage point, and label Km-line. Take any two points on the Ka-line. Subtract the ordinate values (the log units, not the pounds) and divide the difference by the difference of the free-footage values. Neglect the minus sign and record the solve for y. Using the formula (wherein K=co efllcient of friction) solve for the value K.

On the vertical line for footage of 2, plot the value of 11, and draw a straight line to the free footage, marking the line friction line. On the friction line at H-footage record the force value as a deduction value. On the C-line at H-footage read the force value, subtract from it the said deduction value, and record'the difference as felting at 2 board-feet. Plot this f elting value on the-verticalline for 2 board-feet, connect the point by a straight line with the free-footage point, and label the line Kr-line or felting line. Determine the slope of the KF-line by dividing the difference in ordinate values (the log units, not the pounds) of two selected pointsl it by the difference in footage values of said two points, and label the ligure so obtained with or without the minus sign as absolute felting or Mia At the specific-footage on the KF-line record the force value as specific felting. Thus the values determined are compressive properties, listed in Table V.

Table V Absolute values Specictggat m5 Free footage Absolutaelasticity (MA) Absolute felting (M Coefficient of friction (K) Specific elasticity (Kn). Specific ielting (Kr).

On the lines: C'-line, Cs-line by projecting horizontally to the C-line, Kir-line. friction line, KF- line, may be read for any selected footage the following values. respectively: pounds per sq. ft. in compressing, the spring-back height obtain- -able at any time on releasing during the compression, pounds per sq. ft. exerted by elasticity Serial No. 313,919, filed January 15, 1940. Briefly, explained, a vertically axial container, carrying a column of fiber to be tested is dropped repeatedly to give an impact of 1 foot-pound on the column. The fiber thus acquires a constant density or volume. The density, expressed as its reciprocal in board feet per pound of ber at 1 foot-pound impact is the impact resistance adopted for comparisons of fiber.

The impact resistance in foot-pounds is but another manifestation of elasticity of a. felted fiber mass. Whereas the specific elasticity is measured at 3.15 board feet per pound, the impact resistance is related to that density when expanded which just resists settling at 1 footpound impact. A mathematical relation between specific elasticity and "impact resistance is more complicated than saying they are the same thing, recognizing, however, that they arise from the same characteristics of the fiber. It has been determined that where a mass is self-felted for insulation mats to a density at least equal to the tested impact resistance as defined, the mat is sufficiently non-settling for use in refrigerators and like places where substantial shock in daily use,.as in closing a refrigerator door, is encountered.

Hundreds of determinationsjof-the properties of fiberA by the above compressive and impact procedures evidence their real values in evaluating fibers. Hundreds of determinations of coarsenessl modulusrior` the evaluated fibers show its ducing it, to give the 'desired evaluations.V These evaluationspermit the setting of limits for the coarseness modulus for the present invention.

` A ber mass may be characterized in part by its coarseness modulus, which gives very definite l Free-footage, on the other hand, is not such an independentbsingle property. It is determined in part by intrinsic qualities of individual fibers, and

, also in part'by the coarseness modulus. However, it eliminates the state of aggregation. Free` footage must be determined` on fibers in a mass which is substantially unfelted. A mass of fibers which has been felted must be unfelted to determine its free-footage, and both acts may be so severe as to break fibers, thus changing both the coarseness modulus and the free footage. Therefore, in unfelting fibers to determine free-footage, care should be exercised not to break fibers.

For a given kind of fiber having the same general history but differing in coarseness modulus, there is a general correlation between coarseness modulus and free-footage. However, for different types of fiber, the same correlation does not hold. Accordingly, for fibers of the same type, comparisons may be made by reference to coarseness modulus. For fibers varying in type, comparisons may be made by reference to free-footage or to both free-footage and coarseness modulus. For example, it has been found that the thermal conductivity of many iiberauntreated, and treated, fractions, Vand. mixtures of fractions, are correlatable with free footage.

When lignocellulose is deiibered in an Asplund machine, treating agents may be added as set forth'in Asplund No. 2,047,170. Rubber, such as latex, is a treating agent described in my copending application, U. S. Serial No. 336,577, filed May 22, 1940, as a particularly valuable one to minimize dust in bulk fiber, and to improve the properties for certain insulation applications. Bituminous substances, such as petroleum asphalt, is another type of treating agent, described in my copending application, U. S. Serial No. 334,761, filed May 13, 1940. These agents weight the fibers, and two masses which are respectively treated and untreated, may have the same coarseness modulus, but different freefootage values.

Plotting the values given in Table I of free footage and coarseness modulus, or the lc-values and coarseness modulus, gives correlations less perfect than that obtained by plotting lt-value against free-footage. This is in agreement with much other experience, namely that free-footage correlates "k-values very Well for miscellaneous varieties of fibers, as

reliability for control of ber, especially in pre-y shown in the said Anway application U. S. Serial No. 313,920,file` d.January 15, 1940.

In Fig. 1. hereof, the series of Table I is plotted with 1c-value of interfelted fiber mats lacking adhesive on' the vertical axis I0, and free-footage on the axis II, giving the straight line correlation I2. -This is in agreement with Fig. 2 hereof which is Fig. 17, of Anway application U. S. Serial No. 313,920, filed January 15, 1940.

Fig. 2 has thermal conductivity (measured at 3.8 pounds per cu. ft.) of numerous interfelted mats of fiber lacking adhesive, on the axis I3, and free footage of the samel fiber on the axis Il. The band'area I5 defined by the straight lines I6 and I1 was produced by plotting on this graph 48 samples of wood fiber having irregular variations in particle size distribution, with and Without treating agents, and drawing lines bounding the average area determined by the specific plots. 75% of the points plotted lie within the band.

The values of line I2 in Fig. 1 have been transferred toV Fig. 2 as line I8 thereon, in confirmation of the fact that the specific runs of the Table I, fall within a more general relation covering greatly different conditions. Neither the band of Fig. 2, nor the line I2 of Fig. 1 are given herein as a criterion for other bers, but only in illustrations of the general tendencies of the relationships, and of the value of the control and evaluation procedures.

With respect -to the value of free-footage, it is to` be understood that it is a value dependent on the fibers being completely free and 'unfelted by an applied force of compression. If the fibers are felted'byfcompression, or by immersion in water, filtering and drying, the so-called freefootage determined on such partially felted fibers is not the true free-footage. Thus, for example, dry unfelted fiber which is baled, then broken, must be completely loosened from the felted relation produced by baling, before the true free-footage may be determined. This may be done for test purposes by suction apparatus, or adequate picking means, so used as not to introduce felting, and to avoid breaking fibers.

One of the characteristics of the Asplund machine is, that it discharges moist fiber in a loose substantially unfelted condition, which may be easily and directly dried. In fact the steam pressure at discharge from the machine may be used to convey the fiber in a conduit to a conveyer belt of a drier, without thereby subjecting it to a felting compression which destroys its substantially unfelted form vin which the freefootage may be immediately determined. Avoidance of wetting the liber to the point of matting, and avoidance of suspending it in Water, are both economically important where a freefootage value is sought, for two major reasons. The wetting produces clotting, which requires unfelting, with possible injury or breaking of the liber. Immersion in water and filtering, in addition to the above, dissolves solid matter from the fiber with economic loss. It has other effects which change the physical properties of the fiber, such as the felting and elastic characteristics, thermal conductivity and freefootage. Thus, it may be resorted to where a desirable change may be effected.

However, where controlled coarseness modulus is sought, the clotting of the ber by water is not a serious interfering matter. The clots may be unfelted in water in which the fiber may be present for determining the coarseness modulus as later described. Accordingly, when bers'are to be characterized by coarsenes modulusjthe i ber, as discharged from an Asplund machine or the like, may be wet and clotted, or merely moist, and either form may be dried directly, or further wetted or suspended in water, recovered and dried.

Where the fiber is to be characterized by freefootage, it is preferred that the fiber be formed in the absence of suspending or clotting water. and that it be dried directly from its original condition. These make unnecessary any effort to un-felt it for the purpose of evaluation. Such preferred procedure preserves all the solid substance of the vegetable matterl from which it is derived.

Not all the fiber masses made by direct drying or otherwise as herein referred to, are efficient for insulation purposes. A large amount of slivers gives high coarseness modulus, lower free-footage, and poorer insulation efficiency. To avoid too much of the fines and dust, the product must be made with a permissible coarsenes, and a small amount of slivers may be tolerated. fresh directly dried fiber has a free-footage in excess of 6.0, it is satisfactorily efficient for handling, baling, loosening, and packing into insulating spaces, with the proviso that where natural felting without adhesive is used, it has also other required properties. It is likewise satisfactory for building up flexible blankets, with adhesive or without it. .Where a wateradhesive is used, the fiber need not be dried, except to determine free-footage. The moist fiber may be directly conveyed to forming machines where it is built up in the presence of adhesive to provide a mat. Where adhesive is used as a binder for the fibers Ain the mat, certain qualities may be neglected.

For the purposes merelyof insulation the fibers may be made without slivers with lower coarseness modulus. tend to increase, giving dustiness, and also W- ered insulation value. Practically, therefore there is a range for coarseness modulus wherein the fibers as a whole product of lignocellulose, are valuable as bulk fiber for insulation. The various kinds of lignocellulose may have individual ranges within the broader general range. Aspen is typical of the short- `ilbered Woods, and jack pine is typical of the It has been found that when thev However, nes and broken bers aspen more or less lies at one end of the general range, While the range for jack pine lies at the other end. From experience with many samples of ber made in the Asplund machine, a coarseness modulus from 60 to 250 is the general range which is practically useful and efficient. Aspen has the range 60 to 160,while Jack pine has the range from 130 to 250. y

The free-footage value determined after' direct drying runs generally from 7 to 12, but sometimes higher and lower values are encountered. This includes fiber with and without chemical treatment.

. The -process is preferably carried out in the Asplund'machine by introducing moist or slightly wetted Wood lchips into a chamber within which is a steam pressure of 150 pounds per sq. inch, or a temperature of 365 F. This is not a critical pressure, nor is the steam temperature an indication Vof the actual temperature of the wood while vbeing reduced to fiber. 'I'he 4important points in the process, are the absence of water to suspend the fibers, the action of heat in softening the wood, and the quick action. The free-footage of the resulting fiber is dependent upon the setting and operation of the machine, in a complex relationship, which can be best determined by empirical tests, comparing results to operating conditions. The series in Table I is representative of changes in the product resulting from adjustment of grinding elements. In many months of controlling Asplund machines, the coarseness modulus has been used to regulate the machine to control the product.

At the same time the compressive properties (including free-footage) and the impact resistance test have beenused to evaluate the fibers.

With guiding experience, a selected range of coarseness modulus practically assures that the fiber falls within the useful range of desired properties, and lthe evaluation tests point out what the specific properties actually are. In several hundreds of determinations, the coarseness modulus has thus been shown to be a reliable index to the suitability of the fiber for thermal insulation .and the handling thereof to produce it.

Although the foregoing has referred to raw wood as the original material, this is largely because of more experience with it. In practice it is an economic raw material. However, the same process has been applied to chips which have been softened by chemical cooking and washed. It has been practiced on chips fed to the Asplund machine'with cooking chemicals in solution, such as caustic soda, monosodium sulfite and others. It has been operated upon mixtures of raw chips and cooked chips, and mixtures of raw chips and both wet and dried pulp-mill screenings, from both sulte, soda, kraft sulfate 'and monosulfite cooking processes. Accordingly, the term lignocellulose" as used herein, contemplates such mixtures and modified forms.

Where the fiber is to be dry-felted, using the inherent felting qualities of the mass of ber, the felting and elasticityv properties, as determined by the compressive procedure, are very 'important controls. These are related to the .formation of eflicient mats which do not settle and which may exhibit an elasticity in the mat. But where the fiber in the mat is bound with an adhesive, the mat is more stabilized against settling, and the felting and elastic properties are not essential controls. The free-footage howlike nails in a barrel. 'I'hen consider the same sizes of material which is like the wood fiber here described with frayed ends, as in bundles, with elasticity, with flexibility and with bent, curved or other non-linear particle forms. These will pack with greater volume than the rod-like forms. This difference is indicated by the freefootage value.

Examination of the compression-line of the above described compression test procedure, will show clearly that the free-footage value is not dependent upon the action of the fiber under compression (even though it is so determined), as are the elastic and felting properties. It follows, therefore that the elimination of the characteristics of felting end elasticity from the broader definition of the fiber of the present invention is justified by experience, by hypothetical considerations, and by a test procedure and the mathematical analysis of the results of such procedure.

For the working of bulk fiber to produce selfielted (without binder) insulation mats of the desired character, individual properties of the fiber mass have significance. The specific felting value KF must not be too high, or a well felted self-supporting mat, as in a refrigerator cabinet, cannot be obtained. On the other hand, the specific elasticity value KE cannot be too low, or a felt resistant to settling cannot be obtained. The desired thermal effect or K-value imposes an upper limit on C. M. Settling imposes a lower limit on KE, which in turn imposes an upper limit on C. M. Ease of felting imposes an upper limit on KE, which in turn imposes a lower limit on C. M. Adequacy of felting imposs a lower limit to KE, which in turn imposes an upper limit on C. M.

As a result, the suitable self-felting fiber may be characterized as having a range for coarseness modulus, a range for specific felting, a minimum limit for free-footageand a minimum limit for specific elasticity. Collateral and subsidiary. properties may be added, which flow from those specied. Thus, for example, the absolute elasticity may have a lower limit. A self-felting fiber mass within the present invention must therefore have:

Additionally, for eflicient insulation a felt which is resistant to settling, and which has a density of 3.15 board feet per pound, must have a K-value not over .25. The numerous limits above relate to the working qualities of the fiber to produce a mat of such K-value'. Where the term resistant to settling is herein employed, it means with respect to repeated impacts of not more than one foot-pound.

As explained these are interrelated and dependent properties; no one of which predeterbe operated in an ordinary manner at one time to have quite different properties.

mines one other. The debering machine may produce fibers within the above, and at another time to produce fibers outside the above. Treatments of such fibers may change properties within the above to properties outside the above, and vice versa.

Many conditions not referred to above affect the results. For example, Table I was the result of operating one size of Asplund machine. A different size of such machine gives a different distribution of the C. M. determinant fractions, and hence different properties. Thus fibers from said two machines, having` the same C. M; may Table II shows Jack pine f'lber made on a machine of size different from the machine used to make the fibers of Table I. i

- Table II Sample S-218- C. M. fraction N o.-

4 28. 2 3l. 0 27. 0A 30. 2 Z3. 2

P4 22. 8 17. 2 17.0 15. 7 18. 6 Coarseness modulus... 159 160 175 176 193 Free .footage-. l0. 2 9. 2 9. 5 9. 7 8. 9 Specific elasticity KE. 147 115 145 90 110 Absolute elasticity ME .31 34 34 31 34 Specific telting KE... 62 33 43 40 27 Absolute felting ME... 26 25 26 24 .24 Kvalue 227 220 232 In the above table, sample S-218-28 is only a i fairly satisfactory self-felting bulk fiber, since its fiber may not evaluate as satisfactory. It may be observed that'as the C. M. goes up, the KF goes down. However, such fiber is more satisfactory for making adhesivelybound felted mats, Where self-felting is not so important.

In Table I, an examination of the fibers in the light of the limitations placed on them by the present invention for self-felting purposes, indicates that only runs Nos. 2, 3 and 4 provided the proper fiber.. The KE values of the others are outside the range 20 to 80. Run No. 8 is also not a suitable self-felting fiber because of low freefootage, and runs No.' 7 and 8 are also out on account of low KE. Run No. 1 is out on account of high KE. It is to be noted that although excluded run No. 1 of Table I has a K-value below .25, the fiber does not produce the physical properties for the most desirable mat.

Run No. 4 of Table I is comparable to S-218-26 o1' Table 1I on coarseness modulus, and shows a wide variation in KE. Conditions within the different sizes of machines and their operations,. are thus reflected in difference in properties for f Runs Nos. 7 and 8 are excluded evaluations, and indicates treatments.

i frabiem (roughiaclcfpme)k The following table gives a number of fiber '.covery is exemplified at 38 by the term filten The recovered lfiber "may be dried in any form at .l v f vNone (1% on ber) Asphalt ical -"1.Treatment- ,f l i Item'i `item2 Items item4 item5 Items Item? Items Ite'me Itemio o.' tir 131.6 196.8 292.6 [132 294 254.4 242 252 311 229 Kve1ue 1231 -.239 .257 :238 .237 .249 .268 M .373 .341 .517 .422 .324 .441 490 .495 .624 .40 MF ".347 .263 .0 '.326 .25 .273 .255 .326 .420 .29 KE.- 11s y134.5 V24 98.5 114 133.1 99.3 65 15.3 196 KF-" 82 29;5 .0 34.6 38.5 .265 10.6 15.6 6.3 49 F. F 8. 64 s, 95 is 7.87 9. 49 8. 31 7. 17 6. 78 5.125 43.9 Item In" or Out" on ro rties shown. as aself- Kr Kr 'K :eltmgnher p pe KF In C.M. In. In C' KF {0.M. QM In..

Item In or ut"ior adhesive'mats. i. lIn In In In C. M. In C. M. In

r:In Table III, the item 19 ischemieauy treated ber videntied as sample S2l8 11.

The invention contemplates theproductIon .of

ber; with control offparticle size distribution, and subsequent treatment anduse of the resulting fiber forthe ultimate purpose vof providing loosely felted masses having' insulation efliciency, wherein properties" of the ber to form the felts,

` and properties'of the felt are predeterminedby the coarseness modulus and other properties.`

Fig.4 3 shows 'diagrammatically the relation of the originaliiber to the subsequent uses. l Gen` erally, 'and practically the fiber may be provided in dry or moist bulk, dry compacts lsuch as bales and dry laps, and moist compacts, such as bales and wet lap. "These are forms of convenience from which ultimate felts may be made. Bulk iiber, either dry or moist may be assembled to form ultimate mats or other masses at controlled densities. Compacts, moist or dry, may be fiuied or disintegrated, to reproduce a. bulk form.

Dry bulk ber may be assembled into a selffelted insulating body by mechanical packing, pneumatic conveyance and deposition from air currents as described in Heritage U. S. Serial No. 294,212, led September 9, 1939, or other methods.

Wet felting comprises assembling wet or moist bers with or without adhesive, to' form mats. Felting from water suspension may be done under conditions to avoid forming dense mats. The foaming process of Byrant U. S. No. 1,821,856 is one such method.

Fig. 3 is a diagram showing the courses fibers may take in the formation ultimately of loosely felted bodies, such as insulation mats. The

block 20 represents a process of producing :liber of coarseness modulus desired. Lines 2|, 22, and 23, indicate respectively the paths to direct felting, direct drying. and additioni of water. Along path 2l, there are represented felting 24 with or without adhesive while moist from the grinding, and drying 25.

Along path 22, the fiber may be baled at 26 'directly from the grinding, followed by uing 21, then felting moist at 28 and drying 29. Also the iluied ber 21 may be first dried at 30 and dry felted at 3| with or without adhesive.

0n line 22 (direct drying) the dried fiber 23 may be felted atv 32 with or Without adhesive, or be baled dry at 33. After such baling it may be luifed at 34, and then either dry` felted at 35 with or without adhesive, or moist felted at 36l with or without adhesive, and dried at 31.

On path 23, water may be added, or the bers suspended in water, in which case a fiber re- 20 39,1k then iluffedat 40. The dry fluffed fiber may v bedryffeltedat .4l with or without adhesive, or wetfelted at 42 with or without adhesive and dried at 43. Instead of drying the recoveredber from ltration V38, it may in the so-calledflter- '25 ing be obtained as a wet-lap, like paper pulp, as at 44. This may be dried at 45, oruied wet as at 46, or dried at and rewetted and luied wet (46). The dry lap 45 may -be dry flued at 41, and dry feltedv at 48 with or without adhesive.

30 or wet` felted at 49 anddried at 50. The wet fiuied ber 46 may be wet felted at 5| with or Without adhesive and dried at 52, or ,the wet Ilui orwithout adhesive.

Table IV 50 Sample 202-'4174 202-46 Treatment .9%latex solids None Coarseness modulu 198 198 Free-footage- 9. 2 9. 5 ME .333 .327 KE. 100 125 Mr .257 .266 Ks 31 49 Impact reslstanoe- 3. 7 4. 1 value .232 .228 Dust percent 1. 0 2. 8

From many determinations it has been ascertained that for insulating self-felted fibrous bodies the bers used should have properties within the following limits.

6 as as Free-footage.. Speciiic felting Kr 62g Nog?) peciic elasticityv Kg 50 None oarseness modulus 60 250 For adhesivelybound mats, the elastic and {elting qualities need not meet the above limitaons.

Given fibers within these limits the K-va1ue It is` to be understood that various treating This illustrates fibers ofl the sai-ne'k methods, and the stability of the insulating mat particularly against settling when the mat lacks.

adhesive, win be assured, if it is formed in acf cordance. with the evaluation of thev liber. 'Ihe formation mustobserve the impact resistance in order to employ a non-settling density, ywhich will vary according to the conditions to which the will "be satisfactory, thev mechanical working l properties will be satisfactory forone .or more In addition, aboveeach piston head there are two short nozzlesl |02 for water-Jets. These are eter. Water pressure of 25 to 30 pounds at gauge swirl to iibers'on the perforated plates or screens.

82 yforces VJets through' the nozzles to give a rotary The top-plate operates at a level of 81,4' to 11% inches from the top rim of cylinder 61;

` and thetwo lower perforated members are revalue rises." For the libero! the present -invenl tion whenV used 'in self-felted mats, it has 'been' found that asa general rule theK-value'isnot above .25, when the density of the mat .iswithin ,therange 4from 2 -to-6 pounds per cu. ft. Where V-the C..M. isfrom 100l to 250,` the said density may spectively 6 and 12 inches lbelowlplate 98.

In operation to determine'coarseness modulus,

the cylinder is filled by operating valve 84, until 4 thewater-level is 1 inch from the top. Then' the motor is started, and valves 86 and '8| opened, whereby water enters and drains away,the adjustments being such as to maintain the original be anywhere in the-range from 2 to 6,.but for to assure the desiredK-value of'no't over .25.

Frama Cnassrr'rrm fflberhaving a C; M. of below'. aboutlOO, the density vmust 4be nearer the middle'of the range 4 In order -thatthe coarseness modulusfmay be 1 determinedas it' is'herein dened and used,- the A base `60' has uprights 8i supportinga table V j plates or screens arev collected, oven-dried, or4

62, ,which'table presentsv twoupright guides 83 and64 for a sliding cross-head 66.v The table also supports a vertical cylinder 61, of 5% inches inside diametenropen at the top 68 and capped at the bottom 69 where an outlet pipe 1| is connected. *In the cylinder, a plunger operates. It comprises a plunger rodin the form of a waterconduit 12, carrying three piston-heads $13,' 14

and- 18 which providef'perforations 'for filtering; fiber.: The plunger-pipe12 hasa telescopic water-connection with a iixed water-pipe :located at'the top of the guides 63-84, and said xed pipe connects by line I8y to a pressure controlled water supply line 19. valve 8| and pressure gauge 82 indicate the controls of the water supply -to the plunger. Also a water-supply connection 83 with valve 84 leads to the lower capped end 69 of the cylinder, whereby the cylinder may belled with water. A valve 86'in the cylinder drain-line Y1iV controls the outflow in line 81 whicndischarges yThe cross-head 66 f isbalanced lby counterweights .92 over pulleys 93,.and the partsare so l arranged that the crosseheadmay bev raised to bring the, plunger and its lower. perforated head water level in the cylinder.I Then a weighed' separated according to size, as thekmachine is operated i'or'15 minutes; '.'Ihen .the water is drained from the cylinder, and the fibers on the dried in'a steam press, and weighed, giving 4 vfractions for use in determining the coarseness The diilerence passing screen 9| is modulus. merely recorded as the past 4th fraction.

FRAcrroNAran FIBER A mass of jack pine fiber A (see Table VI) evaluating with suitable characteristics for a selffelting fiber has been treated by a suitable process of dryfractionation whereby it is split into fa coarse fraction C and a nefraction'F. 'I'hen the coarse fraction is again similarly divided producing a coarse fraction CC and a fine fraction "I6 olitsideofthe cylinder 61.' In` normal.opera tion the plunger rides in the` Ycylinder-in the positionshown, with a stroke of 3inches fromv a rotary drive operating at ,65 R; P; M. v '.I'hisdrive comprises a motor 92, a; gear box'93,` and tWo crank pins `94. From the crankpins 94 there areV connecting rods 96 to :the vcross-head 166, where a; detachable union 91 is provided, forremovalVv A of the driving rods forraising the cross-head.

The `pistons 13, 14 and 18 have top areas of 41/2 inches diameter. lInpiston 13, this area has holes 98 (Fig. 6) which areinch in diameter-and arranged atla uniform spacing of 100%per sq; in.

In piston 14 (Fig. 7) there are holes 99 which are :ya inch in diameter and arrangedat a uniform spacing of 144 per sq. in. In piston 16 (Fig. 8)

there is a wire screen i0l of 32 meshes per inch;

C F. These masses'oi' fiber evaluate as follows:

Table VI A C F CC GF Property Com- Coarse Fine Coarse Fine posite iromA iromA from C from C Free-footage-F. F 9.75 9.80 9.45 9.22 9.00

KE .l 141 150 133 '110 115 `38.2 2. 8 53. 6 Il. 8 V 14. 4 19. 2 14. 0 25. 2 f 20.2 35.0 16.0 33.2 l 14:4 25.0 6.2 is. 2

250 163.8 A294.6 205.4 K-value .227 .222 .224 .241 .232 In or Out on prop-l erty shown, for seli-felt- 1ngtlber v In (l). In CM In 1 nw but at 11min on CM.

When ilbers are fractionated a different set of laws appears tovhold with respest to the resulting K-value as related to the CM from the ilbers andthe densityjof the mat.

'I'he above table confirms the fact that high i coarseness modulus produces high K, yet it shows thatthe presence of the coarser particles mixed with .fines as ln the composite is permitted. It shows also that with very little change in freeft. It follows that no one property may be used as a direct indication oi the thermal eiilciency. These particular results show that any `judgment evaluation of any one ber on the basis of uiness (free-footage) is not a completely reliable indication of its utility as a thermal insulating ber.

Anmisrvnty BoUNn Mar The bulk ber may be adhesively bound to form exible insulating blankets of high thermal elciency. Thus; rough green jack pine is reduced to ber usinga re-retarding agent with the chips entering an Asplund machine, in accordance with the applications of Heritage and Walter, Serial No. 332,682, and Serial No. 332,683, both led May 1, 1940. Thus, for 100 pounds of dry wood are used 8 pounds oi' ammonium sulfate and .08 pound of calcium carbonate. The machine is adjusted and operated to give ber of the desired characteristics. The discharged ber is exemplied by reference to one having i the following characteristics:

Free-footage 8.3 Specic elasticity (Ka) 97 Absolute elasticity (Ma) .400 Specic felting (Kr) Absolute felting (Mr) .291 K-value .241 Impact resistance 13.1

1 (In board-feet per pcund, or 3.88 in pounds per cu. ft.)

Particle sizedistribution: a

1st fraction 4.2 2nd fraction vv177.4 3rd fraction 22.0 4th fraction 28.4 4+ fraction 28.0 Coarseness modulus 141.4

The discharged .ber has about 100 pounds of water per 100 pounds of dry wood, the mass being moist bers not felted. These may be felted moist by suitable process, but since aqueous binder is added to effect union of the telt, the bers are partially dried so that the formed mat will not be too wet. They are dried to have 20 to 30 parts of water per 100 parts of dry ber example a water-proong size, and mould-proofing agents, insecticides and the like. An aqueous starch solution at about 2.5% has been successfully used. Such a solution having also about .5% of other materials, such as wax in emulsion form, and insecticide.

Such a mat made at 1 inch thickness, may have by weight about 120 to 140 parts of ber, 20 parts of starch, about 15 parts of re-proong agent, about 2 to 4 parts of wax, and about 6 to 8 parts of preservative, per 1000 sq. ft. It has a K-value of about .235. Such a mat may be placed between paper liners, asphalted, plain, at or creped, in varying combinations to provide exiduces Ielting, and such mats are hereinabove indicated as having been so made, where the K-value is given. It has been determined that frames with mats of the same density. packed either by machine, by hand, or by a pneumatic process as above referred to, have identical K-values, where the ber is the same, and where the arrangement is heterogeneous, and the mat substantially uniform in density.

It is of course understood that some bers within the scope of the present invention are so far from the border lines oi.' eicien'cy that they are capable of dilution with other bers. without bringing the mixture outside of the bounds of the presentinvention.

The foregoing description and explanation shows that the invention may be practiced in many ways diering from' the specic illustrations given, without departing from the spirit and scope of the invention as expressedin the appended claims.

I claim:

l. Thermal insulation material comprising an aggregate mass of bers the composition of which is essentially lignocellulose, said bers being characterized by a coarseness modulus within the range from 60 to 250, and further characterized when dry and in substantially unielted mass formation by a free-footage not less than 6 boardfeet per pound; by a specic felting resistance within the range of 20 to 80 pounds per sq. ft. at a density corresponding to 3.15 board-feet Iper pound: and by'a "spcic elasticity of n'ot less than 50 .pounds :per: s`q.1:f-ft.' at. a density corresponding to..3.15. -board-feet per pound.

2.l A ber mass according to claim 1 in which the bers are derived from long-bered wood, characterized bya coarseness modulus within the range from to 250.

3. A ber mass according to claim 1 in which the bers are derived from short-bered wood, characterized by a coarseness modulus within the range from 60 to 160.

4. Thermal insulation comprising felted ber mat having bers of lignocellulose material entangled and bound together by the natural felting property of the ber, said mat having a density not in excess of 4 pounds of ber per cu. it., said bers having the characteristics of the ber of claim 1.

5. Thermal insulation comprising felted ber mat having bers of lignocellulose material entangled in felt formation and adhesively united,

said mat having a density not in excess of 4 agent being selected from the group consisting of bituminous asphalt and of solids of rubber latex, said bers being characterized by a coarseness modulus within the range from 60 to 250, and further characterized when dry and in substantially unfelted mass formation by a `freefootage not less than 6 board-feet per pound;` by a specic felting resistance within the range of 20 to 80 pounds per sq. ft. at a density corresponding to 3.15 board-feet per pound; and by a speclc elasticity of not less than 50 pounds pes sq. tt. at a censity corresponding to 3.15 board-feet per poun 7. Thermal insulation material comprising an aggregate mass of fibers the composition of which is essentially lignocellulose, said fibers having at the surface a small quantity of the solids of rub.-

ber latex, said bers being characterized by a 'j coarseness modulus within the range from 60 to 250, and further characterized when dry and in substantially unfelted mass formation by a freefootage not less than 6 board-feet per pound: by a specific felting resistance within the range of 20 to 80 pounds per sq. ft. at a density corresponding to 3.15 board-feet per pound; and by a specific elasticity of not less than 50 pounds per sq. ft. at a density corresponding to 3.15 boardfeet per pound.

8. Thermal insulation material comprising an aggregate mass of fibers the composition o! which is essentially lignocellulose, said ilbers having at .the surface a small quantity oi bituminousr footage not less than 6 board-feet per pound: by

a speciiicv f'eiting vresistance within the range of 20 to 80 pounds per sq. ft. at a density corresponding to 3.15 board-feet per pound: and by a specic elasticity of not less than 50 pounds per sq.- it. at a density corresponding to 3.15 board-'feet per CLARK C. HERITAGE.

cEm'iinzcATE oF' comuzcTIoNf.v Patent No. 2,525,055.' l Y July 27, 191,5.'

CLARK' c. HERITAGE."

' It is hereby certified that` error appears inthe printed specification of the above numbered patent requiring correction 'as follows: lPage 5', sac-4 a thesaid Letters Patent' should be read 'with this correction therein that the same may.conform to the record of the case in the Patent Off-ice.

Signed' and sealed this lhth day of September, A. D. 1914.5.

` Henry van Arsdelle,` n (Seal) Acting-Commissioner of Patents.'

Images