II M l' 2 \ll 2 1 2 2|! 2 1 l I I “I 2|”! 12 2} 11' W1 22 1. 2 5 1|” 2:; .22 2‘22 222 222:2 “2.? 222 22.222 22 222222 2222 2212‘ Thesis fo2t12e Degree of M. S. 22 “2’32"?” SMTE bNt‘éERSWY H-EL MEIR ‘22515223523, Jr. 1969 “mm? 2222222222 EMEEWI ABSTRACT THE EFFECTS OF SCREW DESIGN ON SCREW HOLDING POWER IN PARTICLEBOARD By Helmer N. Jensen, Jr. The purpose of this thesis was to study the relation— ship between screw design and screw holding power in the edges of a three layer, commercially produced, particleboard. Six special screws were used throughout the testing. Three screws had 12 threads per inch, the other three, 20. Shallow, medium and deep threads were used in each of the two threads per inch groupings, and three pilot hole sizes were tested with each of the six screws. All board samples were brought to the same moisture content prior to testing. It was concluded that screw design had no effect on the holding power of the screws tested. The average withdrawal values for the screws fell around a regression line described by Y = 85.29(X) + 390.H3 where Y was the withdrawal force per lineal inch of thread required to extract the screw and X was the surface area of the screw per lineal inch of thread. Pilot hole size had an inconclusive effect on the holding power of the screws. Helmer N. Jensen, Jr. The outside diameter of the screws tested was, for all practical purposes, equal. The shear area was therefore essentially unchanged as shear area is determined by the dia- meter of a screw if all thread lengths are the same. In the screws tested, there was an obvious shear fail- ure of the board material at the outside diameter of the screws. It is felt that a change in shear area, brought about by either a change in screw diameter or screw length, will cause a proportional change in the holding power of a screw. Density variations in the samples tested were small. No conclusions could be drawn from this except that small density differences have no noticeable effect on holding power and can be ignored. THE EFFECTS OF SCREW DESIGN ON SCREW HOLDING POWER IN PARTICLEBOARD By Helmer Niels Jensen, Jr. A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Forestry 1969 ACKNOWLEDGEMENTS The author wishes to express his most sincere appreciation to Dr. Otto Suchsland, of the Department of Forestry, for his guidance and counsel in the preparation of this thesis. ii TABLE ACKNOWLEDGEMENTS . . . . . LIST OF TABLES . . . . . . LIST OF FIGURES. . . . . . Chapter I. INTRODUCTION. . . . OF CONTENTS II. MECHANICS OF A SCREW. . . . . . Defining a Screw Screw Mechanics. Screw Designs Utilized . . III. DESIGN OF EXPERIMENT. . . . . . Variables Studied and Preliminary Tests. Experimental Design. . . . . IV. EXPERIMENTAL PROCEDURE. . . . . Samples. . . . . Conditioning . . Screw Insertion. Screw Withdrawal Density Borings. V. RESULTS AND DISCUSSION. . . . . VI. CONCLUSIONS . . . . LITERATURE CITED . . . . . iii Page ii ‘iv O\CDO\ O\ l-’ Table U'I LIST OF TABLES Page Listing of the dimensions of the six screws tested. Dimensions are illustrated in Figure l. . . . . . . . . . . . . . . . . . . . 18 Listing of the values of certain percentages of the root diameters of the screws tested . . . 21 Summary of design of experiment showing variables tested and replications. . . . . . . . 3“ Results of withdrawal tests for 360 samples tested, showing withdrawal force and standard deviation for each of the 18 screw—pilot hole combinations, and for the six screws tested. . . . . . . . . . . . . . . . H9 Results of analysis of variance showing the variables compared . . . . . . . . . . . . . . . 50 iv Figure l O 10. ll. 12. LIST OF FIGURES Diagram of a screw thread showing angles and diameters. . . .. . . . . . . . . . . . Representation of resultant forces in a screw with a load, without friction . . . Representation of resultant forces in a screw with a load, with friction. . . . Torque meter for controlling insertion torque of screws, showing screw and sample in place. Insertion torque plotted against time . . . . . Seating torque as determined by preliminary testing . . . . . . . . . . . . . . . . . . . Diagram of typical sample tested showing dimensions of sample and location of density boring holes. . . . . . . . . . . Screws used for testing . . . . . . . . . . . . Jig used for withdrawing screws from samples shown mounted in Instron testing machine. Withdrawal force as recorded by testing machine Particles remaining on screws following complete withdrawal, 7/6“ inch pilot hole. . . . . . . Particles remaining on screws following complete withdrawal, 9/6" inch pilot hole. . . . . . . Surface area of threads plotted against with- drawal force, showing the regression line and mean values of screws i one standard deviation . . . . . . . . . . . . . . . . . . Page 11 ll 23 25 28 30 35 39 Al AA A6 51 CHAPTER I INTRODUCTION Particleboard is manufactured from wood chips and/or flakes by pressing the particles between heated platens after applying an adhesive to the particles. The particles are randomly arranged in the plane of the board. The properties of particleboard are the result of the combination of component variables such as wood, particle geometry, resin, and wax, with the processing variables such as temperature, and rate and degree of compression (8). Particleboard can be classed as interior or exterior depending on the type of binder used. Exterior board is made with a waterproof resin, usually phenol formaldahyde, while interior particleboard is normally manufactured with urea formaldahyde, a resin that is only water resistant. An interior board was used in the procedures that follow. In the twenty years since particleboard was introduced commercially in the United States, its production has grown at a very rapid rate until, in 1967, over one billion square feet (3/4 inch basis) were produced by the platen process (A). This represents 95 per cent of the particleboard produced, the other 5 per cent being manufactured by the extrusion process. Furniture core grades accounted for 57 per cent of the platen process production (A) thus making it the most important single use. As early as 1956 (13) it was reported that the major impetus for establishing a wood particleboard industry in this country came from the furniture producers desiring to use it as core stock for high pressure laminates and hard- wood veneers. Particleboard has proven itself to be highly satisfactory when used in dining room furniture, cedar chests, bedroom furniture, television cabinets and sewing machine cabinets. It is also being used in desk tops, kitchen cabinets, counter tops in stores, truck and trailer bodies, or nearly any place that plywood can or has been used (18). In spite of its growing popularity in the furniture and building construction fields, particleboard has some serious drawbacks. The prediction of dimensional stability is more easily accomplished in a homogeneous mass such as particleboard than in any but the highest quality wood (12). Thickness swelling is excessive, however, and there can be permanent deformation (9) if the particleboard becomes too wet. Most particleboards machine less smoothly than lumber because of their structure, but in typical applications, the machined surfaces are concealed, and somewhat less exacting standards of machining often suffice (6). Large particles, especially when combined with light pressure, produce a board of coarse texture with numerous voids or openings between particles. These voids cannot be sanded out because they occur throughout the entire thickness of the sheet. Large voids lead to chipping of the particles and produce rougher cutting. In addition, these coarse—textured boards must be cross banded before applying a face veneer. If cross-banding is not used, the phenomenon of telegraphing will occur. Telegraphing may be described as the surface distortions of the board showing through the face veneer. Large holes or voids in the board will show as depressions, while the particles, upon becoming wet, will swell perma— nently due to a loosening of glue bonds, and appear as raised places in the face veneer. Another problem, and one of great concern to the furni— ture industry, is that of the screw holding power of particle- board, particularly in the board edges. It has been found that the screw holding ability is usually less in particle- board than in lumber (12). Another point of note is that screws inserted into the face of a piece of particleboard invariably hold better than those inserted into the edge (1), with ratios of two to one not being uncommon (12). In order to alleviate this problem, several means have been studied and recommended. Larger and longer screws or the application of resin to the pilot hole before screw insertion have been offered (12). Urling (13) suggested using a minimum size pilot hole. Yale, in 195A, felt that the use of wider edge bands or the insertion of solid wooden blocks would provide the desired results (17). In 1956, Yale (18) found that the screw holding power could be in— creased by increasing the density of the particleboard or by using a sheet metal screw in place of a wood screw. Wyckoff agreed (l6), and found that sheet metal screws, with their more closely spaced and more deeply cut threads, when com— pared to wood screws, usually offer more holding power than either a wood screw or a common wire nail. The screw holding power of particleboard has been shown to be a function of board density and to a lesser extent, a function of the resin content, chip orientation, and particle size and shape (12). Another study (3) showed that holding power generally increases with the size of the screw. Very little conclusive work has been undertaken to determine what effects are obtained by comparing various screw designs for their holding power in particleboard, especially in the edges. It is for this reason that this study was initiated. Specifically, the intent of this thesis was to inves- tigate the effects of screw design on the holding power of screws inserted into the edges of particleboard. The char- acteristics to be compared were the slope (threads per inch) and thread depth within the screw itself, and the size of pilot hole into which the screw was inserted. Special screws were selected to be manufactured because there could be much more flexibility and control of design than would be found in commercially produced screws. This also per- mitted control over thread depth to allow the screws to be compared with one another on a meaningful basis. Taper was eliminated in order to have screw-to-particleboard contact throughout the entire thread length. The screws had double heads. The head nearest the threads allowed the screw to be tightened and seated in the normal manner while the other head permitted the attachment of the device used for with— drawing the screws from the samples. The special screws also eliminated the variables and tolerances inherent in the use of commercial screws. CHAPTER II MECHANICS OF A SCREW Defining a Screw Screw type fasteners are made in many different designs for different purposes. They can be made of several differ- ent materials such as wood or plastic, but most commonly are made from a metal or an alloy. One common element groups all screw fasteners together, and that is the helically descending thread. Screw fasteners are, in general, of two types. The first type encompasses those that have a threaded receptacle to receive the screw, such as a nut receiving a bolt. The other type, without a threaded receptacle, is characterized by the common wood screw that must form its own threads in the material being fastened. In the first type, the strength of the fastener is a function of the characteristics of the external threads of the bolt and the internal threads of the nut. Since the threads are preformed and there are certain tolerances, there is no appreciable friction involved in assembling the Joint 7 except for the final tightening. The tensile strength is that of the bolt at its root diameter in the plane perpendi- cular to the axis of tensile force while its shear strength is that force which can be resisted by the bolt or nut at the base (root) of their threads in the plane parallel to the direction of pull. In both instances, the strengths of this type fastener are relatively high as they normally deal with metal to metal contact. The second type, and that with which this thesis will be dealing, must form its own threads in the wood particles as it is being fastened. In order to do this, the material receiving the fastener must be considerably softer than the screw. The amount of friction to be overcome for insertion is much larger, too, as threads must be formed in the board by the metal threads of the screw displacing the material to be fastened. As might be expected in this type of fastener, the strength values, especially in shear, are lower since the fastened material is considerably weaker than the screw. A problem that must be contended with, especially where power driven equipment is used to insert screws, is that of over- tightening. This can, and frequently does, lead to a stripping of the threads in the board. When this occurs, the screw becomes loose and, for all practical purposes, has no holding power. Essentially the same thing can happen when screws are subjected to withdrawal loads in use. The threads in the fastened material will eventually give way and the holding power will be lost. Screw Mechanics A screw may be compared to an inclined plane (10) with its resultant forces. The threads are the inclined plane, merely spiralling up a shaft, while the load is the force resisting insertion. The force resisting insertion of a wood screw is friction. This friction is a function of the coefficients of friction between the screw and, in this case, the particleboard, and is the result of the surface characteristics of the two materials. (The distance through which the force is moved in one revolution of the thread is equal to the lead, as shown in Figure 1. In a single-threaded screw the lead is equal to the pitch. In a double—threaded screw the lead is twice the pitch, etc. The friction is overcome by a force acting at right angles to the axis of the screw. If there is no friction, a load W is supported by the reaction force N which acts normal to the plane of the helix. Therefore, without fric- tion, the force required to initiate movement of a load or to Just hold it is QO as seen in Figure 2—A. QO = W tana (Eq 1) When friction is present the force Q required to move the load along the plane is increased. With 8 the friction angle and u the coefficient of friction, ie p = tanB, (Figure 2-B) then, Figure l.——Diagram of a screw thread showing angles and diameters. do = outside diameter dr = root diameter d = mean diameter or E2_i_3£ 10 11 Figure 2.——A.—-Representatlon of resultant forces in a screw with a load, without friction. B.—-Representation of resultant forces in a screw with a load, with friction. 2 ll load 20 ll force when friction is present Q0 = force without friction 12 Lead *-*- 2nr \Normal Figure 2. 13 Q = W tan(a + 8) (Eq 2) In equation (2), Q is the force that must be exerted at the midpoint of the thread, and the torque at radius rm is Q(rm). The operating torque of the screw is applied through a lever at a distance D/2 from the center line of the screw. Thus F(D/2) = Qrm = Wrm [tan (a + 8)] (Eq 3) This is the equilibrium equation for the operation of a screw when the load is being moved up the screw (10). The shear area of a screw may be at the root diameter or the outside diameter (see Figure l) of a screw depending on the type of fastening involved. It is the place at which the threads will be stripped if a sufficient tensile force is applied parallel to the axis of a screw. If an aluminum bolt is threaded into a nut made of hardened steel, for example, one would expect the shear to occur at the root diameter of the bolt. In wood screws, the opposite is found, in that the shear failure occurs at the outside diameter of the screw when the wood fibers fail. Any screw will rupture at its root diameter when a sufficient amount of tensile force, parallel to its axis is applied. The tensile stress to which a screw may be sub- mitted is found by the equation: IN S = F/A (Eq 4) where F is the force in pounds and A is the cross-sectional area absorbing the load in square inches (10). The bearing pressure per thread is supported by the area of the annulus (10) having an outside diameter of (10 and an inside diameter (root diameter of the screw) of dr. This area is described by: 2 2) A square inches. When n number of threads are in contact, this area becomes: 2 2 n1r(dO - dr ) A = 2 ' (Eq 5) The unit bearing pressure, P,is equal to the total load W in psi, divided by the area A from equation (5). Thus: W 2 w n =4: u nfl(d0 2) (Eq 6) '—_—___7T--—_—_ The screw mechanics discussed up to this point are those used primarily for an analysis of bolts and power screws Operating with a load, both with and without friction. The preceding paragraphs also pertain to the general mechanics of a wood screw which, during insertion, operates without a load, 15 but with friction. The actual mechanics of wood screws have not been studied as fully as those of power screws, but some limited information is available. The resistance of screws to withdrawal from the side grain of seasoned wood varies directly with the square of the specific gravity of the wood. Within limits, the with- drawal load (that required to pull the screw from the wood) varies directly with the depth of penetration and the dia- meter of the screw (15). The effective length of a screw is limited by the length at which the screw will fail in tension. This limiting length decreases as the density of the wood increases. For wood screws inserted into the side grain of sea— soned wood, the allowable load may be expressed: 2 p = 2370 G (do) (Eq 7) in which p is the allowable load (one—sixth of the ultimate load) per lineal inch of penetration of the threaded portion of the screw; G is the specific gravity of the wood based on oven dry weight; and d0 is the outside diameter in inches (15). From an analysis of the above discussion, it would appear that the optimum screw design to be used in a given situation would be one where the values of shear failure, S (tensile stress in equation A), P (bearing pressure in equation 6), and p (withdrawal load in equation 7) all 16 reached their maximum values simultaneously. The design of any screw for a specific purpose should therefore be made with this object in mind. Frictional forces must be overcome in order to insert a screw. In this experiment these forces are the result of the surface area and texture of the screw, the density of the board, and the particle geometry. Torque is used to overcome these forces during insertion. Friction also plays an important role in the resistance of screws to withdrawal (tensile) stresses applied parallel to the axis of the screw. A screw that has been driven into wood derives most of its holding power from those fibers which contact it at their ends. When a withdrawal force is applied, the fibers are bent upward and fail in horizontal shear. These fibers do not fail locally, but may be moved up to 3/8 inches along the grain with the screw as the load is applied (11). Screw Designs Utilized The six screws selected to be tested were special made by turning them down from a hardened steel rod on a lathe. The hardened steel permitted the reuse of all screws through— out the experiment with no apparent wear. They were designed with no taper in order that the withdrawal force could be spread over the entire length of the threads, thus eliminat- ing any effects caused by tapering where some threads would be more deeply imbedded in the particles than others. 17 Table l, in conjunction with Figure 1, describes and gives the dimensions for each of the six screws. The screws will be referred to by number as indicated in Table 1 through- out the remainder of this thesis. The portions of a screw that could be varied in a meaningful way were the thread slope (threads per inch) and the thread depth. Changes in these factors will result in proportional changes in the surface area of the screw. An increase in the threads per inch will, at a given thread depth, increase the surface area by a proportional amount. Likewise, an increase in thread depth (decrease in root dia- meter) will increase the surface area. Thus the effects of different slopes and thread depths could be measured in a controlled experiment. In this experiment, three thread depths were used in each of two slope designs. .m®£OCH CH mCOHmCmEHQ HH< 18 omam.o o:ae.o omos.o eewe.o mmme.o same.s Amado eeea ted mete teachem mme.a mom.H mm:.a mam.a mm:.a 0:3.H hammers do hemeea m smo.o mHo.o mao.o Heo.o Hmo.o omo.o heuoe heats eeehee \ o (H mma.o mefi.o mma.o moH.o mmH.o mmfi o e heheeeaa poem 0 o \ o o o \ o O mmfi o mag 0 Han o Has 0 Hmfi 0 was a e emphases teameao a m a m m H aeheez steam deed eased: eoaamhm deed eases: aoaamem heads esteem om om om NH NH NH some had heathen .H as empeaemaaaa mum mCOHmCmfiwH .flm mmp m m Mam m0 mun“ map %0 MCHpmHHII.H MHMdU CHAPTER III DESIGN OF EXPERIMENT Variables Studied and Preliminary Tests The first variable introduced was that of the number of threads per inch on the screws. It was felt that by having two very different slopes, ie 12 and 20 threads per inch, any effects caused by a change in slope would be apparent between these two choices. Three thread depths, ie shallow, medium, and deep, were considered necessary for each slope to test for differences since the change in total surface area would not be great, and several tests might have been required to detect the small differences involved. The thread depth, although controlled, could not be varied in a manner entirely independent of the threads per inch. All screws were made with a 60° thread angle (see Figure l), the angle normally used for screws. The screws with 20 threads per inch had a smaller pitch than those with only 12 threads per inch. Thus, since the thread angle was a constant, those screws with 12 threads per inch could have greater thread depths than those with 20. 19 20 The screws did not have the same length of threaded distance. The varying lengths were not used as a controlling device for the experiment, and were corrected for in the cal— culation of all results. All results are therefore given as withdrawal force or surface area per lineal inch of screw thread unless otherwise stated. Prior to selecting the pilot hole sizes to be used, tests were performed to determine the effects of pilot hole size upon the different screws within the range of 70 to 90 per cent of the root diameters of the screws. The 70 per cent figure is recommended for use in softwoods (7) while a pilot hole of 90 per cent of the root diameter is recommended for screws used in hardwoods (2). This range varied from 0.076 to 0.151 inches, or from between 4/64 and 5/64 inches to between 9/64 and 10/64 inches (see Table 2). In nearly all trials, the samples utilizing the 5/64 and 6/64 inch pilot holes were split. These two were thus rejected, as was the 10/64 pilot hole. The rejection or the 10/62 pilot hole was due to its being too large to provide a sufficient bear- ing area for good screw-to-particle contact. Pilot holes of 7/64, 8/64, and 9/64 inch diameters were selected since they were best suited to be used with all six screws. These values range from approximately 95 per cent of the smallest root diameter to approximately 85 per cent of the largest root diameter. By utilizing these three pilot holes, splitting of samples was eliminated entirely. 21 .mam>Hpomammn .:m\m cam :m\> now mosam> on» .mo:a.o cum :moa.o cmozpmn wcHHHmm modam> mmpmoaocHx smo.o :oe.o oaa.o* mHH.o* emfl.o* HmH.o* mma.o* e :HH.o* mmH.o* omH.o* mma.o* s:a.o mmH.o mma.o m mHH.o* mmH.o* :mH.o* m:a.o Hma.o oma.o mma.o : meo.o mmo.o mwo.o mmo.o mmo.o 30H.o moa.o m omo.o smo.o moa.o oaa.o* mHH.o* mma.o* mNH.o* m moa.o :HH.o* mmfl.o* mmH.o* smH.o* zea.o mmH.o H 1 ponesz os ms om mm om mm OOH 1111111 steam / pamo pom .ompmop mzmnom map no mpmpmemao poop on» mo momwpcmopmo :Hmppmo mo mozam> on» mo mcfipmfiqnl.m mqmao39 coeH nod momoane .mcofimeHHan was ompmmp mmaowfimm> wcfizonm pcmsfipooxm mo cwfimmo mo mam5§3m11.m mqm¢e 35 Figure 7.--Screws used for testing. From tOp to bottom, screw numbers are: 6, 4, 5, 3, 2, l. 36 CHAPTER IV EXPERIMENTAL PROCEDURES Samples The samples used in this experiment were 3" x 6" blocks cut from two 4' x 8' x 3/4" sheets of standard commercial three layer particleboard. One half of the samples were cut with their long dimension parallel to the 8 foot length of the board. The remainder were cut with this dimension per- pendicular to the 8 foot length. The samples were randomly placed in 18 groups of 20 samples each for testing. Conditioning All samples were placed in a conditioning room with temperature and relative humidity controlled at 70°F and 65 per cent respectively. The average moisture content, based on oven dry weight was 9.83 per cent. The standard devia— tion was $0.31 per cent. The highest and lowest moisture con— tents were 10.89 per cent and 7.28 per cent respectively. Screw Insertion The screws were inserted to within 1/8 inch of seating with a spring loaded electric screwdriver as seen in Figure 3. This insured that the screws would be inserted vertically 37 38 and an equal amount of downward force would be used in all cases. The final seating of the screws was accomplished by hand tightening them with a screwdriver until the proper torque was reached. The torques utilized for screw insertion were measured for each sample and were limited to prevent stripping of the threads within the particleboard. A torque of 22.7 inch pounds was used for all tests with the 8/64 and 9/64 inch pilot holes. The 7/64 inch pilot holes required 28.3 inch pounds to properly seat the screws. This higher torque requirement was due to the increased friction between the screws and the sides of the smaller pilot hole. Screw Withdrawal The screws were withdrawn using the special Jig seen in Figure 8. An Instron Universal Testing Machine was used to extract the screws from the samples at a speed of 0.5 inches per minute. Figure 9 shows an example of the maximum withdrawal force required for one sample and the manner in which it was recorded by the testing machine. In seventeen of the twenty replications, the withdrawal was continued until the load dropped to 200 pounds after go- ing through a maximum. In the other three replications, the screws were completely withdrawn in order that an analysis of the amount and type of particles remaining on the screws could be made to determine what type of failure occurred 39 Figure 8.—-Jig used for withdrawing screws from samples shown mounted in Instron testing machine. ' . . C a“. I Q. 1 o .3 o. a. 9 z. 9 o '4. " .,."’5“>'1 ‘._ 41 Figure 9.--Withdrawal force as recorded by testing machine. Maximum withdrawal force shown is 600 pounds, occurring at 4.8 seconds and 0.04 inches of downward movement of head after test began. _ *‘r—v Withdrawal Force (lbs) 600 500 400 300 200 100 42 12 24 Time (sec) 0.1 0.2 Head Movement (in) Figure 9. 36 O. 43 within the samples. Figures 10 and 11 are typical examples of the results of complete withdrawal. Density Borings Immediately following the withdrawal tests, density borings were made on each sample. To obtain the density, the samples were weighed individually. Two holes were then drilled in the samples (see Figure 6) followed by another weighing. The change in sample weight divided by the vol- ume of the holes (in cm3) gave the density of the particle- board at the area of the test. Values obtained during actual testing were: Highest Density 0.75 gm/cm3 Lowest Density 0.64 gm/cm3 Average Density 0.69 gm/cm3 Standard Deviation 0.02 gm/cm3 These figures agreed substantially with those obtained during the pilot test mentioned earlier and, since there was a relatively small deviation from the mean, they were all considered to be equal for the purposes of comparison and calculation. 44 Figure 10.--Particles remaining on screws following complete withdrawal, 7/64 inch pilot hole. Screw numbers from top to bottom are: 6, 4, 5, 3, 2, l. 45 46 Figure 11.—-Particles remaining on screws following complete withdrawal, 9/64 inch pilot hole. Screw numbers from top to bottom are: 6, 4, 5, 3, 2, l. 47 CHAPTER V RESULTS AND DISCUSSION The results of the withdrawal tests are shown in Table 4. The average withdrawal forces shown for each pilot hole size under the various screw designs are based on twenty individual tests. It appears that the effects of screw design, if any, are very limited over the entire range of the variables with the exception of screw number 1, which shows a decidedly lower value. An analysis of variance was made and the results are shown in Table 5. These results seem to indicate that the variable combination of slope and thread depth has a highly significant effect on withdrawal power. Slope and depth of the thread determine the entire contact area of the screw. The result of the analysis of variance might therefore be expected to be supported by a strong relationship between withdrawal power and total con- tact area. This relationship is shown in Figure 12. A statistical analysis of the mean withdrawal values for each screw showed that the significant difference pointed up by the analysis of variance is limited to the difference 48 49 0 6 2 8 2 O . . . . . . zepom pom monom Hmzmnonufiz 3 M Mn W .64 5 M no coapwa>om camocmum 763 208 289 696 028 267 QOQOW HQ3Q¢HU£PH3 350 915 831 “.07 113 “18 53.“. “uh". U. 35 .U. 32 U. 55 an.” .HO COHPmfi>0Q Uhmvcmpm W M, mm % mod N ACHEEC 398m pom u. u. u. .4 u. u. mohom Hmzmnvnpfiz mwwno>< was she. awe ohm, 1% my» 223: .2... 7 4 4 4 444 444 444 444 444 444 HM3MQU£OH3 mwmhm>< 9/64 8/64 7/64 9/64 8/64 7/64 9/64 8/64 7/64 9/64 8/64 7/64 9/64 8/64 7/64 9/64 8/64 7/64 Aces .smfia «Hem eoafim m m z m m H ponesz 3mhom doom Esaooz soaamnm doom adage: zoaamzm apnea ommpne om ma cocH pom momonne .Umpmmp mzmzom xam map you one .mcoapmcfin IEoo maon poafidlzmpom ma one no some now coapmfi>mm onmocmum one wopom Hmzmnonpaz mcfizozm .Umpmmp mmademm 0mm pom mpmmp memmoepfiz mo mpHSmmmII.: mqm<9 50 monom Am Vx ma manmfimm> pcmocoaoa mHom poafim A: Vx memHnm> mnowopwo ma 9 spoon unease Am vx mHanpm> anowmpMo ma 0 macaw whence Am vx magmanm> whowmuwo ma m nosemeeaemm AH vx manwanm> showcase ma < mmm mamm.wm:mom Hapoe mmom.mmom mmm Hamm.mmomsm poupm wcacamsmm mow.o mmmo:.o Hmoo.mmm : :mmo.:m=m mom :s:.o ommmm.o mmwa.mzma : mmmm.owm> no moo.o mmmmo.m :mzm.mmmma m n:mo.mmmmm mm omm.o :mmmm.a woma.mamm m uawm.mmwm a mooo.ov :smmm.m mmmm.mmasa m mmms.mmmzm om mHH.o mmo:a.m on:m.m~z: m mmmw.mzmw o wmo.o mmmmm.: mwom.momoa H mmom.momoa m msm.o mmmom.o ommm.mmma ma Hmma.ssmmm < Eoommpm meanwpm> .emem m we seHHHQmQOhm oepmfiempm so mmhmsum so mocmofimficwfim .xopdd< m cheddm new: mmmnmmo mo 53m condom unmeaeficmflm .m use .om .om .Uondeoo mmapmfipm> esp wcfizonm mocmfinm> mo mammawcm mo measmmmln.m mqm¢9 umhm TOQMHQN> .HO wmohdom 51 Figure l2.—-Surface area of threads plotted against withdrawal force, showing the regression line and mean values of screws 1 one standard deviation. 52 .NH themes Amcflv mmz< oommpsm mm. om. ms. ow. +1 . . II sum .eem H1 1/1/(\\\\\w1 me.omm + xmm.mm u m ecfid eoemmeemmm . .emo .eem H+ e... . 11 088 009 ogn 09n ohn 08h ooh (SQI) eoaos temeapqatm 089 53 in values between screw number 1 and screws 2, 4, and 6. In any event, the practical variance of withdrawal force over the range of the surface areas studied is of very little significance. The variable of slope, either alone or in combination with other variables, is seen in all of the significant relationships in the analysis of variance. Statistically, the mean withdrawal values of screws 1, 2, and 3 (12 threads per inch) are significantly less than those of screws 4, 5, and 6 (20 threads per inch). This difference, again, may be attributed to the decidedly lower values of screw number 1. Figure 12 shows the relationship between withdrawal force and the surface area of each screw. The standard deviation of the withdrawal values for each screw was plotted to show the ranges of dispersion. It was found that the values varied around the linear regression curve in the form of Y = mX + b where: Y = withdrawal force (lbs) X = surface area of the screw (in2) m = slope of line = 85.29 b = y—intercept = 390.43 The relationship is stated as: Y = 85.29(X) + 390.43. This line is horizontal for all practical purposes and shows that the design of a screw, in itself, has no effect on its holding power. 54 The effect upon the values due to the pilot holes was varied, and no significant relationship was found to be caused by this variable. The type of failure occurring is seen in Figures 10 and 11. These figures clearly indicate that there is an obvious shear failure in the particles at the outside dia- meter of the screws. Figure 11, showing extraction from a 9/64 inch pilot hole, had fewer particles in those screws with 20 threads per inch. This, it was felt, was caused because the root diameter was slightly larger than the pilot hole and the threads were closer together, giving only a small volume in which the particles could be retained. The combination of these factors would not allow the particles to cling to the threads, thus leaving them in the sample. CHAPTER VI CONCLUSIONS From the results obtained, it may be concluded that the design of the screw had no effect on the holding power of the screws tested. The values for the resistance to withdrawal fell around a regression line described by Y = 85.29(X) + 390.43 where Y is the withdrawal force and X is the surface area of the screw. The effect of pilot hole size was varied and no con— clusions could be drawn concerning this variable. The board material failed in shear at the outside dia- meter of the screw threads. All outside diameters were essentially equal as was the holding power per lineal inch of thread. The holding power must therefore be a function of the shear area of the screw. The difference in density variation encountered within the samples tested was slight and no conclusions could be reached except that small density changes have no significant effect on the holding power of a screw. Further research in this area could be carried out by studying the differences caused by using a non-tapered 55 56 screw, as in this experiment, as compared with the normal tapered screw since the values obtained in this experiment were much higher than preliminary tests with tapered screws had indicated. Another project could be centered around the effects on holding power of a screw under varying moisture content conditions in the particleboard. In this thesis, the mois- ture content was held constant because of the unknown effects that would be caused by varying it. It is further suggested that the standards for testing screw holding power in particleboard be modified to such an extent that the screws be completely seated before being extracted. This should give a more accurate indication of how the screws would perform under actual use conditions. LITERATURE CITED 57 4. 10. LITERATURE CITED Akers, L. E. 1966. Particleboard and Hardboard. London: Pergamon Press. American Standards for Testing Materials. 1964. ASTM D1037-64, Sections 89 Anonymous. 1968. "Screw Holding Power of Particle- board." National Particleboard Association, Technical Bulletin No. 3. “ Anonymous. 1968. Wood_and Wood Products. Reported by U. 8. Bureau of Census. VoITfi73, No. 10, p. 37. Commercial Standards. 1966. CS 236-66, Mat Formed Wood Particleboard, p. 5. Davis, E. M. 1962. "Machining Particleboard. " Furniture Design and Manufacturing, Vol. 34, No. 5, p. 26. Fairchild, I. J. 1926. "Holding Power of Wood Screws." U. S. Dept. of Commerce, Bureau of Standards. Tech. Paper 319. Gatchell, C. J., Heebink, B. G., and Hefty, F. V. 1966. "Influence of Component Variables on Properties of Particleboard for Exterior Use." Forest Products Journal, Vol. 16, No. 4, p. 46. Heebink, B. G. 1967. "A Look at Degradation in Particleboards for Exterior Use." Society of Wood Science and Technology_. SWST Research Paper No. 10, p. 59. Jefferson, T. B. and Brooking, Walter J. 1951. Introduction to Mechanical Design New York: The Ronald Press. 58 11. 12. 13. 14. 15. 16. 17. 18. 59 Jones, Burce R. 1954. "The Screw Holding Power of Veneered and Laminated Wood Panels." Forest Products Journal. Vol. 4, No. 3, p. 1I9. McGee, L. B., McLean, R. A., and Carlyle, A. A. 1957. "Properties of Particleboard Related to Its Use in Furniture Manufacture." Forest Products Journal, Vol. 7, No. 3, p. 9l.I Urling, Gerald P. 1956. "Wood Particleboard-—A Giant in the Making." Forest Products Journal. Vol. 6, No. 10, p. 363. I Whittington, J. A. and Walters, C. S. 1969. "With- drawal Loads for Screws in Soft Maple and Particleboard." Forest Products Journal, Vol. 19, No. 3, p. 39. 1 Wood Handbook. 1955. U. S. Department of Agriculture Handbook No. 72. Washington, D. C.: UT S. Government Prihting Office. Wyckoff, William R. 1962. "Application of Screw Fasteners to Wood Construction." Mechanical Fasteners for Wood. Building Research InStitute Puincation No. 1003. Yale, Rollin H. 1954. "Wood Versus Substitutes in the Furniture and Cabinet Industry." Forest Products Journal. Vol. 4, No. 2, p. 24—A. 1956. "Composition Board Has Found Its Place in the Furniture Industry." Forest Products Journal. Vol. 6, No. 10, p. 365. I ICHIGAN STQTE UNIV. LIBRRRIES HIM Ill HHII! lllllll llll l 3 9 1 12 30 5938784