Free Redacted Document - District Court of Delaware - Delaware

Case 1:06-cv-00751-MPT

Document 52

Filed 07/22/2008

Page 1 of 3

IN THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF DELAWARE PARKER-HANNIFIN CORPORATION, Plaintiff,
V. ZIPPERTUBING (JAPAN), LTD.,

Civil Action No. 06-751-MPT

REDACTED
Defendant,

DECLARATION OF WILLIAM J. MARSDEN, JR. IN SUPPORT OF ZIPPERTUBING'S ANSWERING CLAIM CONSTRUCTION BRIEF WILLIAM J. MARSDEN, JR. declares: 1. I am a member of the bar of the State of Delaware and have been admitted

to practice by this court. I have personal knowledge of the facts stated herein and would testify to them under oath if called upon to do so. 2. I am an attorney in the law firm of Fish & Richardson, P.C, counsel for

Defendant, Zippertubing (Japan) Ltd. ("Zippertubing") in this matter. 3. I make this declaration to identify documents submitted with

Zippertubing's Answering Claims Construction Brief.

4.

Exhibit 1 is a true and correct copy of W. Denney Freeston, Jr., Knife

Coating of Porous Substrates, in Coated Fabrics Technology: A Collection of Papers from the Journal of Coated Fabrics 27 (Technomics 1971). 5. Exhibit 2 is a true and correct copy of John C. Zemlin, Development of a

100% Solids Urethane Fabric Coating Process, Coated Fabrics Technology, Newton, MA , March 28-29,1973 (AATC 1.973).

Case 1:06-cv-00751-MPT

Document 52

Filed 07/22/2008

Page 2 of 3

REDACTED
6. Exhibit 3 is a true and correct copy of N.A. Favstritsky and J.-L. Wang,

Flame-Retardant Brominated Styrene -Based Polymers. X. Dibromostyrene Grafted Latexes, 69 J. Coating Tech. 39 (May 1997). 7. Deposition of Action No. 1:07-cv-266-MPT. 8. Dated: July 15, 2008 I declare under the penalty of perjury that the foregoing is true and correct. FISH & RICHARDSON P.C. Exhibit 4 is a true and correct copy of a confidential excerpt from the in Parker-Hannifin v. Schlegel Electronic Materials, Inc., Civil

By: William J. Marsden, Jr. #2247) FISH & RI SO P.C. 919 N. Market Str , ite 1100 P.O. Box 1114 Wilmington, DE 19899-1114 (302) 652-5070 [email protected] Attorneys for Defendant ZIPPERT UBING (JAPAN), LTD.

Case 1:06-cv-00751-MPT

Document 52

Filed 07/22/2008

Page 3 of 3

CERTIFICATE OF SERVICE I hereby certify that on July 15, 2008, I electronically filed with the Clerk of Court this Declaration of William J. Marsden , Jr. in Support of Zippertubing ' s Answering Claim Construction Brief using CM/ECF which will send electronic notification of such filing(s) to the following Delaware counsel. In addition, the document was caused to be served on the attorneys of record, at the following addresses and in the manner indicated::

VIA EMAIL Rudolf E. Hutz, Esquire Francis DiGiovanni, Esquire Steven A. Nash, Esquire (pro hac vice) Connolly Bove Lodge & Hutz LLP
The Nemours Building 1007 N. Orange Street Wilmington, DE 19801 rhutz@cblh. corn

Attorneys for Plaintiffs PARKER-HANNIFIN CORPORATION and PARKER INTANGIBLES, LLC

[email protected] snash@cblh. com

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 1 of 18

EXHIBIT 1

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 2 of 18

COATED FABRICS TECHNOLOGY
A Collection of Papers from the Journal of Coated Fibrous Materials

TECI

publication

TECHN'OMIC Publishing Co., Inc. 265 W. State St., Westport, Conn. 06880

iir n IIRRARY

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 3 of 18

COATED FABRICS TECHNOLOGY
A Collection of Papers from the Journal of Coated Fibrous Materials

©TECHNOMIC Publishing Co., Inc. 1973 265 W. State St ., Westport, Conn. 06880

a ,1., B CP LV (DMIC publication

Printed in U.S.A.

Library of Congress No. 73-86380 Standard Book No. 0-87762-118-7

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 4 of 18

W. DENNEY FREESTON, JR.

Fabric Research Laboratories, Inc. Dedham, Mass.

Knife Coating of Porous Substrates
(Received March 17, 1971)

ABSTRACT: A theoretical analysis of the knife coating of porous substrates is derived. The purpose of this analysis is to obtain a ,better understanding of the effects on the penetration of the coating material of the knife geometry, clearance between the knife and substrate, rheological properties of the coating material, coating speed, and porous nature of the substrate. KEY WORDS; Knife coating, theory, analysis, fluid dynamics, rheology, viscosity, dynamic pressure, porosity, coating speed, penetration.

INTRODUCTION
S UliFACE COATINGS are applied to porous fibrous substrates for ' any of a number of reasons, most of which are functional in nature, although

some are purely aesthetic. Often it is desired to reduce the permeability of the structure to gases or liquids, increase durability under certain use conditions, or improve compatibility of the surface with its surroundings. A coated fabric may also be easier to care for than uncoated fabric, or it may just look nice. No matter what the reason for tbQ coating, and no matter how it is to be applied, control of the penetration of` the viscous liquid into the porous substrate is absolutely essential. Too little penetration can result in an inadequate bond; too much in excessive stiffening, reduced tearing strength, and an unacceptable appearance on the underside of the coated material. At the present time, penetration control depends largely upon skill and experience of the coating manufacturer and the coater, and the general excellence of the products attests to that skill. Nevertheless, it is unsettling to realize that we understand very little about this complicated but commercially extremely important process. It was to contribute to an improved understanding of the basic mechanics of knife coating that the work described herein was carried out. The application of coatings to fibrous substrates is a rather complex problem in fluid dynamics. In some of its aspects, it has much in common with
27

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 5 of 18

W. D, Freeston, Jr.

the hydrodynamic theory of lubrication to which a good deal of study has been devoted [1]. Several examples of the application of this theory to the Po Pa blade coating of paper exist in the N) ^^ r N - --^ _ _ _ published literature, of which only iE- -^ a two will be mentioned. Windle and Beazley [2] have studied the hydroFigure 1 . Knife Coating Model dynamic pressure developed under the blade of a paper coating machine from a theoretical point of view, while Follette and Fowells [3] have provided experimental verification of some of these principles. The dynamic interaction of the coating and the substrate has also been discussed by many authors. For example, penetration of viscous liquids into a: porous system was studied by Bohmer and Lute [4], and Lazor [5] has discussed the influence of the rheology of the coating paste. These examples of -published work, along with others which can be found in the literature dealing with paper coating, represent an important contribution to our understanding of the processes. Unfortunately, similarly detailed studies of the fluid dynamics of coating more porous substrates such as woven or knitted fabrics, or non-woven fibrous webs, do not exist. These substrates and the techniques used for coating them, give rise to considerations which differ in significant ways from those applicable to paper processing.

MODEL A pictorial representation of a coating knife-fabric interface is shown in Figure 1. The knife is at rest and the fabric moves at the constant velocity U with respect to it; the coating material flow takes place in the wedgeshaped channel between the knife-and -the fabric: -Sine -this- model is analogous to the lubrication of slipper bearings, the hydrodynamic theory of lubrication can be applied to the analysis of the knife coating of fabric. It is reco nized that not all coating knives are of the shape shown in Figure 1. Many used in commercial coating have a square lower edge and are held vertically, tilted forward or tilted back depending on the coating material and the product to be produced. When the square edged blade is tilted forward or is vertical, the fabric blade interface is of the form shown in Figure 1; when the blade is tilted back, other considerations may be introduced which would alter the analysis, and possibly the implications of its results, from those discussed herein. A thick blade with a rounded edge is also often used. An analysis for this blade geometry was not undertaken, since mathematical complications did not appear warranted at this time.
28

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 6 of 18

Knife Coating of Porous Substrates

In principle, however, the approach would not be greatly different, and the results of the analysis would probably be quite similar to those obtained below for a blade with a beveled edge, since the critical region of the knifefabric interface would be essentially the same as that shown in Figure 1. It is assumed that the distance h (x) between the knife and the fabric is very small compared to the thickness l of the knife. Also, the fabric is approximated initially by a plane, nonporous surface. Since in reality the surface of a fabric is rough and the coating penetrates the fabric, b (x) represents the distance from the knife to the surface of the fabric plus some fraction of -the`fabric`thickrress,...,._.... ,^r, -....,.w......-..__ - ,__..,.._..__... _,...-.. ..,^..-..,.......a,,....,.__., ,w^......_-__- -. DYNAMIC PRESSURE UNDER THE COATING KNIFE ., r.A'complete description of the motion'of a compressible fluid requires not only the fundamental equations of fluid motion (i.e., the Navier-Stokes equations) [1], but also the equation of state which gives the relationship between pressure, density and temperature in the fluid, and the energy equation which specifies a balance between heat and mechanical work. However, since the knife is assumed stationary and the fabric moving with constant velocity U, the knife coating of a fabric can be considered a steady-state process. Additionally, the coating material can be treated as an incompressible fluid (p = constant). Temperature variations in the fluid can be assumed small although this temperature may be different from that of the surroundings. Also, since the shearing rate can be assumed constant, the viscosity also may be assumed constant. (However, the appropriate value for the fluid temperature and the shearing rate is still required.) Therefore the equation of state and the energy equation become superfluous in calculations of the flow field. Furthermore, since fabric coating can be treated as a steady-state process, all terms in the Navier-Stokes equations containing derivatives with respect to time can be equated to zero. Because the coating knife is also assumed very long and the fabric wide, the problem involves only two dimensions. Letting x, y, z denote the three orthogonal directions of the Cartesian coordinate system and u, v, w the three respective orthogonal components of the velocity vector, all terms containing derivatives with respect to z can be equated to zero and w = 0. With the foregoing simplifications the equations of motion reduce to
au au ap a2u ax2 a2u

p

u- v-

ax +

at

- ax +

+

at 2

,
(1)

av

av

u- +V---

1

ap

a2v

a2 v

ax

at

at

axe + ay2
29

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 7 of 18

W. D. Freeston, Jr.

where p is the fluid density, p. the viscosity and p denotes the difference between the total pressure and the hydrostatic pressure (pressure at rest). This definition of p results in cancellation of the body forces since they are in equilibrium with the hydrostatic pressure.
ou It is further assumed that although the convective acceleration u - is ax

different from zero, the viscous forces are considerably greater than the inertia forces and thus the inertia terms can be omitted from the equations of motion. An estimate of the ratio of the inertia forces to the viscous forces is Inertia Forces Viscous Forces pu au/dx tL a2u/aye P U2/1 p, U/h2 p Uhl p,l (2)

Since the inertia forces are proportional to the square of the velocity whereas the viscous forces are proportional to the first power of the velocity, a flow for which inertia forces are neglected is obtained when the velocity and thus Reynolds number is very small. When the inertia terms are omitted from the equations of motion, the resulting solutions are approximately valid for Re < < 1. The reduced Reynolds number is given by [ 1) U1 ( h ) 2 t (3)

The following are typical ranges of values for the parameters in the reduced Reynolds number for laboratory and commercial knife coating of fabric: U = 4 to 10 yards/minute (commercial) 4 to 10 feet/minute (laboratory) f^ = 10,000 to 80,000 centipoises p = 1 to 1.5 gms/cm3 .a 0' to 10 mils

0 = 30° (typical)
1 = 0.125 to 0.250 inch

0 < Re* <0.03. Thus for all practical cases.Re* << 1 when knife coating fabric.
The second of Equations (1), the equation for the y direction, can be omitted altogether because the component v of the coating material flow velocity is very small with respect to u. The pressure gradient in the y direction is also small and can be neglected. In the equation for the x direction 2 2

a2
30

can be neglected with respect to a a because the former is smaller

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 8 of 18

Knife Coating of Porous Substrates

than the latter by a factor of the order (h/1)', With these simplifications, the differential Equations (1) reduce to dp dx a2u a µy2 (4)

and the equation of continuity in differential form can be replaced by the condition that the volume rate of flow in every vertical plane under the knife must be constant:
h(x)

Q= o

udy= const.

(5)

.,...,... ,..... The^·boundary.oonditions are-:-....._...... w..^.._.:.a,

w_, ... _a.._...._._x^...._....M........_..,...._...^..

y=O : u=U; x=a . p=po

y=h : u=0; x=1 : p=po

(6)

i.e., the pressure distribution must satisfy the condition that p = po on both sides of the knife. The solution of Equation (4) which satisfies the boundary conditions is

V
u =U`1__..f )

d p dx 2µ

y h h

(7)

where the pressure gradient dp/dx must be determined in such a way as to satisfy the continuity Equation ( 5) and the boundary conditions for pressure. Inserting ( 7) into (5) Uh Q= 2 or, so lving f or dpl dx =12µ ( 2 h - hs } l^ Hence, by integration (9) h3 (dp 12µ ` dx (8)

d
rx dx fdx - l2 Q O o h3

p (x) = po - f- 6µ U

(10)

Inserting the condition p = po at x = 1
1

dx (11)

-2 U

1

f hz o
L dx 0 h3

Q

31

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 9 of 18

W. D, Freeston, Jr.

Thus the mass flow is known when the shape of the flow channel is given as the function h (x) . Equation (9) gives the pressure gradient, and Equation (10) gives the pressure distribution over the knife edge. In the case of a knife with a flat face for which h (x) = tan 6 (a-x ) 0 (a-x) (assuming 0 < 30°) where a and b are constants ( see Figure 1) Q= U and for the pressure distribution P (X) = B(a - l)a 2a - l (12)

p°

+ 6p

x(l - x)

U h2( 2a-l)

(13)

The pressure p4 at the center of the knife, x = 1/2, is given by
3 2

p#

p° + 2 w

U

hm2 ( 2a-l)

(14)

where hm = h at x = 1/2, i.e., the distance between the knife and the fabric at the midplane of the knife. For small angles of inclination between the knife and the fabric the pressure distribution from Equation (13) is nearly parabolic. Hence the mean pressure difference becomes:

pm - UU
From Figure I

l2 (2a - l) hm2

(15)

hm=b+ l tanO^8+ 0 2 2

(16)

and
a= 6 + l tanO tanB 8+10 0

(17)
u (18)

I.- Therefore-.Equation.-(15-Y,can be rewritten xn the Poll©wing ft^rm pU120 Pm = 2(o x- 1 O ) s

For cases where b < < hm, Equation (18) can be further simplified 4µ U

PM = 102
32

(19)

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 10 of 18

Knife Coating of Porous Substrates

As this expression shows, the greater the viscosity µ of the coating material, the faster the fabric is moving, the thinner the knife and the smaller the angle 0, the greater is the mean pressure generated under the knife. Of these four variables, the one that has the greatest effect on the pressure and the one that can be most readily varied is the angle 0 of the blade to the fabric. Fabrics are usually coated at the fastest possible speeds because of economic considerations. The thickness of the knife is dictated by structural considerations. The viscosity of the coating material is a complex variable whose -_..... influence.....can.,.be-.vaned-.. vithi.r ...sever-a.1.,.orclers^af...niagruituder How.ever,...as---_-discussed below, this influence is neither simple nor usually predictable. The foregoing analysis also indicates that the angle of the sides of the coating knife with the fabric surface should have only a small secondaryeffect do the pressure generated under the knife edge. Additionally, since very high pressures are generated under the knife, the height of the reservoir of coating material behind the knife does not have a significant effect on the pressure.

COATING VISCOSITY
Coating materials rarely have simple, Newtonian viscosity characteristics. Usually they display some degree of thixotropy or pseudoplasticity, and occasionally they are dilatant [5]. This means that their viscosity is dependent on shear rate and time. It is also strongly temperature dependent. The general magnitude of the viscosity can be altered by means of additives, but at the same time these are likely to alter the thixotropy of the system. Thus, while one can vary the viscosity under the knife, and therefore the value pertinent to Equation (19) this cannot be done without regard to the effect on the viscosity at other shear rates or temperatures which may be encountered after the spreading operation. For the remainder of this discussion, which deals with penetration of the coating into pores, it will be assumed that the viscosity is p', which is different from µ in Equation (19). In an actual case,. a pore size distribution will imply a range of shear rates, and therefore a viscosity distribution as well, so assigning a single value for the viscosity is an over-simplified representation of reality. Nevertheless, it will serve as a good approximation to indicate the effect of coating rheology, with the realization that the derived expressions cannot be applied quantitatively without dealing more precisely with the rheological problems. PENETRATION INTO A POROUS SUBSTRATE The actual penetration of the coating material under the knife into the substrate being coated can be treated as viscous flow through an aggregation
33

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 11 of 18

W. D. Freeston, Jr.

of very small channels. The channels in a woven or nonwoven fibrous structure are similar in shape and arrangement to the spaces within a bundle of randomly oriented short cylinders. As the fluid penetrates the structure it will flow successively through channels of varying size. A stream flowing through a large pore space will be divided into several streams which flow through smaller spaces and then unite with each other or with other streams at the next large pore space, etc. In a dense woven or nonwoven structure the coating material flows between fibers and consequently all the channels are relatively small in crosssection. However, when coating a moderately open woven structure, in addition to. penetration of the coating material between fibers, it flows between the yarns. For most structures these latter channels are substantially larger than the channels between the fibers in the yarns. However, the crosssectional area of the channels between the yarns also varies along the channel length. Consequently' the channels in a fibrous structure can be visualized as sets of short capillaries connected in series with the cross-sectional area of the capillaries in each set varying along their length. An exact analysis of such a structure appears sufficiently complex to question the wisdom of attempting it for the knife coating of fabric. Instead, the following simplified model of channels in a fibrous structure is used. An aggregation of parallel capillaries, circular in cross-section, going from one fabric surface to the other is assumed [6]. Each capillary consists of numerous uniform capillaries (capillaries with a constant cross-sectional area) having various lengths and cross-sectional areas connected in series. In this model each capillary of varying cross-sectional area in the fibrous structure is replaced by a uniform capillary having the same resistance to viscous flow. Similarly, each group of capillaries that feed or are fed by larger capillaries are replaced by single capillaries having the same resistance to viscous flow. The fraction of the total number n of capillaries per unit volume of the fibrous structure having the radius r and length- l is denoted herein by the,.. function (r,l) where N(r,l) (20a) (r,l) dr is a small increment of capillary radius, Al is a small increment of capillary length, N (r,l) is the total number of capillaries per unit volume which have a radius within ± Ar/2 of any prescribed radius r (0 < r < co) and length within ± 0l/2 of any prescribed length 1(0 < l < oo ), and ).^N (r,l) = NT is the sum of all such total numbers of capillaries. In the limit, i.e., as dr and of both approach zero, Equation (20a) can be used to define 4' (r,l) as a continuous function:

(r,l} =
34

1 d2N NT drdl

(20b)

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 12 of 18

Knife Coating of Porous· Stthsfra(e.s

From this definition of the distribution function 4 (r,l ) n
k=1

n

1^

4:(rk °, lk

Ark Alk =

k=1 Ark Alk NT

I

N(rk°,lk

°)
Ark Alk = 1

(21)

where
rkGrk l'k

+1,OGr kGCO ,

lkGlk

1k +1, 01k ' cc.

If the function D(r)) is continuous throughout the range of values of r and l increments -of length _Arr, ,and,, olk, approacli zero,-and the -number,_ of -_111-111 1.11 1 increments n approach infinity, the above relation can be expressed as an integral
Ar--*0 At-.) O

lim_
k=1

rk

.°, l kO

^,1ark

Alk =. f

Co 1=0

fr=0 "

4) (r))drdl

(22)

and from Equation (20b),

f= 0 f.r 4,(r,l) drdl = l '0 =0

f 1- 1 NT f=000

°° dON

r=0 drdl

drdl = 1

(23)

This conversion of the summation to the integral presumes a sufficiently large number of capillaries per unit volume of the fibrous structure to exist so that discrete capillaries can be replaced by a continuum. The actual internal pore structure of a real material can be determined by Mercury Penetration.* In this procedure the size and quantity of the pores are measured by determining the quantity of mercury which can be forced into the material at various pressures [7]. Mercury wets few materials, thus it will not penetrate pores unless forced. Under pressure mercury will penetrate circular pores, in accordance with the expression pD = - 4o cos t i (24)

where p is the applied pressure, D the diameter of the pore, u the surface tension of mercury, and 0 the contact angle between mercury and the material forming the opening. Additionally the volume of the mercury forced into the pore is a'direct function of the pore volume distribution. Thus, a penetrated volume-pressure curve can be converted into an effective cross sectionpore number distribution curve. The analysis of the flow through a straight tube of circular cross-section is the classical case of the Hagen-Poiscuille theory of flow through a pipe. An outline of the analysis follows [1] : Let the x-axis be the axis of the pipe and r denote the radial coordinate
°e.g., Mercury Penetration Porosimeter manufactured by Micromeritics Instrument Corp., Norcross, Georgia.

35

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 13 of 18

W. D. Freeston, Jr.

measured from the axis outwards . The velocity components in the tangential and radial directions are zero ; the velocity component parallel to the x-axis, denoted by u, depends on r alone, and the pressure is constant in every cross-section. Of the three Navier-Stokes equations in cylindrical coordinates, only the one for the axial direction remains , and it simplifies to d2u 1 du dp = dx 25 ()

dr2 + r dr

The boundary condition is u = 0 at r = R, where R is the radius of the pipe. The solution of Equation (25) is u(r) = - k dp ( 4µdx R2 _ r2) (26)

where -dp/dx = ( Pi - P2) /I = constant is the pressure gradient along the pipe or capillary. This expression is an exact solution of the Navier -Stokes equations. The mean velocity u of the flowing fluid is u= 82 T
µ dx

(27)
(28)

the volume rate of flow is Q = dV/dt = irR2u= irR4 (- dp /dx) /8M' and the pressure drop through a uniform capillary of length x is

AP = fo x 8µ4 dV dx wRdt

(29)

The laminar flow described by the foregoing expressions occurs in practice only when the . Reynolds number Re = puR/µ' is less. than appr iUa ely 2300. The flow becomes turbulent for values greater than 2300 . However, it can be shown that the Reynolds number of the flow of coating into the fabric pores would indeed be small. ^'I'lie disTaiice^x tl"ie°$u^d°fl'ows iii'a given°"periocl"'of'tim^ t" ^^" obta'in' Equation (27) as follows
U __-

dx

R2 op
8M'
E

dt
z

x

R2Ap

f xdx = °

8
µ

dt

(30)

/ R2Ap t 11
36

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 14 of 18

Knife Coating of' Porous Substrates

Neglecting entrance effects , to a first approximation , the pressure drop through capillaries connected in series can be assumed additive. Therefore, the pressure differential from the surface of the fibrous structure into which the fluid is flowing to the depth x, per unit surface area of the structure is given by

rz
AP = ''x=

n

8t^ N(rk,k)
70 k dV 8µ,

dV

k=1

dt

A (31)

NT

( r k lk)

A'rk Al, dx

k=1

If Ark ---> 0, Alk -4 0 and n --> oo, the above expression for the pressure drop through the fibrous structure can be expressed in integral form AP = 81tdV , 7r dt 1X r ,^ ,1c=o r=o N (r'l) drdl dx r4 (32)

Performing the integration with respect to x

AP

8p'x dV
7r at

7
r=o

(' a
r=o NT

4)(r l)
r4

drdl

Rewriting Equation ( 33) the total volume rate of flow of fluid through a unit area of the fibrous structure is
dV 7r Ap 1 1
,,t =O

(34) r

fr= O

00

NT

4

drdl

At any distance x through the fibrous structure the rate of flow dx/dt will not be uniform ; it will vary from channel to channel. However, if it is assumed that the range of cross-section dimensions along all channels is roughly similar, the variation in rate of flow among the channels will appear only as minor irregularities in the surface which marks the average distance the fluid has penetrated throughout the fibrous structure. (In general , this is what is actually observed when coating fabric. ) This approximately uniform rate of flow dx/dt can be determined by equating the volume rate of flow corresponding to a uniform rate in every channel, i.e.,

dV _ dx " dt i=0 dt
to that given by Equation (34).

r=0

7rr2NT^(r,l)drdl

(35)

37

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 15 of 18

W. D. Freeston, Jr.

dx

1 8µ'

oplx

dt

[l0
f='0

(36)

f
r =O

r2 NT^(rl)drdl

I [

,^I ^

r 0 NT r

r4

drdl I

Separating variables and integrating the left hand side of Equation (36) with respect to x and the right hand with respect to time, the following expression is obtained for the distance the fluid penetrates the fibrous structure as a function of time, pressure differential, fluid viscosity, the total number of capillaries per unit volume and the capillary dimensions.

Apt
x -

J

4^^O,r_O r2NTD ( r,l)drdl ]

1=0 r f ,^_o NT

drdl]
14

(37)
This expression is identical in form to Equation (30) and reduces to Equation (30) for the case of flow through a single uniform capillary of radius R. For knife coating of fabric, the degree of penetration of the coating material into the fabric is given by Equation (37) with the Ap replaced by the pressure generated under the knife, Equation (19). Carrying out this substitution and noting that t is the length of time the fabric is under the knife, and therefore l/t . - U, the running speed of the fabric, the following expression is obtained.

x= 02

[f -o

00

` Jr-O r2 NT^ (r,l)drdl ]

'V

('
[" I=0

f _0 NT

(r,l )

(38)

drdl

This equation indicates that the degree of penetration of the coating material is independent of the speed of the fabric, thereby verifying the validity of extrapolating the results of investigations on laboratory coaters to high_...m.spe^dprod^rcti.on ma^hYbres: -_^.,_._..^...^...^....^.....^Y...,,......,._ .:.^..^.. __,^.,_._....._...^ .._...._^.,^.._.......,.._s ..... Equation (38) also shows that for a given fabric the variable that has the largest effect on the degree of penetration of coating material achieved in knife coating is the angle 0 the blade makes with the surface of the fibrous structure. The smaller the angle, the greater the penetration due to the increased pressure generated under the knife. The presence of the viscosity ratio C viscosity under the knife in the viscosity in the pores numerator of Equation (38) indicates that the penetration is dependent upon the shearing rate dependency of the viscosity, and not upon its absolute
38

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 16 of 18

Knife Coating of Porous Subsirah'.

level. Clearly some minimum viscosity limit must be imposed upon this statement, since the assumption that the pressure generated tinder the knife was significantly greater than that due to the action of any other external force such as gravity was implicit in the derivation. This condition is met by most normal coating materials. Thus, it is in general a valid and most significant observation that it is the pseudo-plasticity of the coating material which plays an important role in determining the degree of penetration, and not the absolute value of the viscosity at any given shearing rate.

Newton's law of friction says T µ du dy (39)

where T denotes shear stress and the other terms are as previously defined. From the analogy between this law and Hooke's law for an elastic solid body, the velocity gradient du/dy can be thought of as the rate of change of the shear strain (or shearing rate) in the fluid. . The shearing rate under the coating blade is given by the first differential of Equation (7) with respect to y. Carrying out this operation and substituting Equation (9)

du _ dy

U h

U 2h2

Q (6h -12y) 113 f

(40)

where Q is given by Equation (12). This expression shows that the shearing rate is a maximum at the fabric surface and decreases with increasing distance from the fabric. Similarly, the shearing rate in the fabric pores is given by the first differential of Equation (26) with respect to r
du dr 1 op r (41)

2µ' x

where op is given by Equation (18), since the pressure at the entrance to the pores is the same as that under the knife and the pressure at the exit, i.e., the back side of the fabric, atmospheric. Equation (41) shows that the shearing rate in the pores is zero at the center and a maximum at the wall. Using the values for U, S, 0 and l given previously and letting x equal a fabric thickness of 0.007 inch (typical value) and the pore radius R - 0.0020.004 inch, i.e., the distance between yarns in typical fabrics, it can be shown that the shearing rate under the knife is two to three orders of magnitude greater than in the fabric pores.
39

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 17 of 18

W. D. Freeston, Jr.

CONCLUSIONS
This analysis of the fluid dynamics of knife coating of porous substrates was based on an idealized geometric model of the knife-coating-substrate interface, and on several simplifying assumptions with respect to the fluid properties of the coating material and the nature of the flow. Thus, the results would not be expected to quantitatively predict coating behavior; this must be left for future refinements both in the analysis and in our understanding of the process itself. It is not too much to hope, however, that the results can at least provide significant qualitative insights into the effect of the critical parameters of a knife coating operation. These can be summarized as follows : (a) The dynamic pressure which develops in the fluid coating between the knife and the substrate increases with increasing coating viscosity, increasing velocity of the substrate relative to the knife, decreasing knife thickness and, most importantly, decreasing angle between the bottom surface of the knife and the surface of the substrate. (b) This pressure plays a major role in forcing the coating into the pores of the substrate. The depth of penetration is independent of substrate velocity but increases with increasing ratio of coating viscosity at the shear rates existing under the knife to the coating viscosity at the shear rates within the pores of the substrate, that is, with the degree of coating pseudo-plasticity. It also increases as the angle between the blade and the substrate surfaces decreases and as the distribution of pore sizes in the substrate tends towards shorter and/or larger diameter pores.
These conclusions, particularly that relating penetration to the shear rate dependency of coating viscosity, explain some of the perplexing observations which have been made in the course of coating trials, and for which no satisfactory explanation could hitherto be given. The scope of usefulness of such a simplified analysis in a practical sense, however, -can only be determined by relating the trends from carefully controlled experiments to those indicated by the analysis. There is little doubt that other considerations will have* to be introduced before predictable control over a knife coating operation canbeachieved.Limitingvauesof nl e^ Lilic1Cness, coaingvise OSi and substrate velocity must be considered, The importance of capillary effects and their relation to the surface energies of substrate and coating over the range of temperatures which will be experienced must be analyzed. .And, of course, the analysis should be extended to cover other coating techniques, such as roller coating, calender coating, and transfer coating. In spite of the fact that the analysis has been so limited, the value of such an approach is clearly indicated. It is hoped that future opportunities will arise to extend and -refine the work, in order to provide a sound theoretical basis 40

Case 1:06-cv-00751-MPT

Document 52-2

Filed 07/22/2008

Page 18 of 18

Knife Coating of Porous Substrates

for modifying coating procedures to obtain better productivity and improved products. REFERENCES
1, H. Schlichting, "Boundary Layer Theory," Pergamon Press, N. Y. (1955), 2. W. Windle and K. M. Beazley, "The Mechanics of Blade Coating," Tappi, Vol. 50 (1967), pp. 1-7. 3. W. J. Follette and R. W. Fowells, "Operating Variables of a Blade Coater," Tappi, Vol. 43 (1960), pp. 953-957. 4. E. B¢hmer and Jan Lute, "Adhesive Migration and Water Retention with Reference .,. . ..,..t4 Blade fCryating;:-SvenskXapperstifining,Vol.-.69,^ 19664,lal)_B10T6,1& 5. J. T. Lazor, "How to Relate Plastisol Viscosity to Processing Problems," Modern Plastics, Vol, 42 (May, 1965), pp. 149-152, 154, 157, 182. 6. O. Kallmes and H. Corte, "The Structure of Paper I. The Statistical Geometry of an Ideal Two Dimensional Fiber Network," Tappi, Vol. 43, No. 9 (1960), pp. 737-752. 7. R. L. Peek- and D. A.- McLean, "Capillary Penetration of Fibrous Materials," Incl. and Eng. Chem., Analytical Edition, Vol. 6, No, 2 (1934), pp. 85-90.

41

Case 1:06-cv-00751-MPT

Document 52-3

Filed 07/22/2008

Page 1 of 11

EXHIBIT 2

Case 1:06-cv-00751-MPT

Document 52-3

Filed 07/22/2008

Page 2 of 11

a

W

COATED FABRICS TECHNOLOGY

Sponsored by the COATED FABRICS TECHNOLOGY COMMITTEE of the

AMERICAN ASSOCIATION OF TEXTILE CHEMISTS AND COLORISTS
March 28-29, 1973 i at the

, TS1512

Marriott Motor Inn Newton, Mass.

I _- J

Case 1:06-cv-00751-MPT

Document 52-3

Filed 07/22/2008

Page 3 of 11

® 1973 by the American Association of Textile ChemColorists. All rights reserved . This book, or parts nay not be reproduced in any form without the pert the publisher.

Published by American Association of Textile Chemists and Colorists

Case 1:06-cv-00751-MPT

Document 52-3

Filed 07/22/2008

Page 4 of 11

PAPERS
Title and Author INTRODUCTORY REMARKS .................................................................................................................... Stanley J. Shurtleft, US Army Natick Laboratories SUBSTRATES FOR COATED APPAREL APPLICATIONS .................................................................. F. J. Beaulieu and M . D. Troxler, West Point Pepperell COATING BASE FOR SYNTHETIC LEATHER MANUFACTURE .................................................... Gustav Sinn and Wolfram von Langenthal , Bayer A.G. Page 1

2

5

PERFORMANCE OF PLASTICIZED VINYL FILMS IN DRYCLEANING SOLVENTS ................ 15 Manfred Wentz and Ivan J. Andrasik, International Fabricare Institute THE CONSUMER AND COATED FABRICS .......................................................................................... Ronald J. Pacheco, J. C. Penney Co. THE ROLE OF THE COMMISSION FINISHER AND COATER IN TODAYS APPAREL INDUSTRY ..................................................................................................................... Brian W. Curtis, Kenyon Piece Dyeworks 24

... 27

REACTIVE FUNCTIONALITY IN DISPERSION RESINS-ADHESION OF CARBOXYL-MODIFIED PLASTISOLS TO FABRICS ..........................................................................

30

J. R. Russell and D. W. Ward, B. F. Goodrich Chemical Co.
EDGE-ABRASION CHARACTERISTICS OF COATED APPAREL FABRICS .... ............ _.._.......... J. H. Dusenbury and H. D. Lyons, Deering Milliken Research Corp. ELECTROLESS PLATING OF METALLIC COATINGS ON FABRICS.......................................... D. P. Thibodeaux and A. Baril, Jr., Southern Regional Research Labs, USDA DIVERSIFICATION OF COATED FABRIC TECHNOLOGY ................................................................ Donald K. Klumpp , Aldan Rubber Co. POLYURETHANE COATINGS: CHEMISTRY AND THE CONSUMER .......................................... James H. Badger and Michael Lee Rosin, Soluol Chemical Co., Inc. AQUEOUS DISPERSIONS OF POLYURETHANE IONOMERS FOR COATING AND LAMINATING ................................................................................................................................................ .. , H. H. Neumaier, Bayer A. G. URETHANE COATINGS FOR TEXTILE SUBSTRATES .................................................................... R. T. McSweeney and R . J. Athey, E. I. du Pont de Nemours & Co., Inc. HIGH-PERFORMANCE URETHANE FABRIC COATINGS FROM POLYCAPROLACTONE POLYOLS ............................................................................................................ L. R. Comstock, R. M. Gerkin, F. E. Critchfield, Union Carbide Corp. APPAREL MARKETS FOR URETHANE COATED FABRICS .......................................................... George J . Zahringer, Jr., National Credit Office, Div. Dun & Bradstreet 40

42

50

54

62

71

83

96

DEVELOPMENT OF A 100% SOLIDS URETHANE FABRIC COATING PROCESS .................. 101 John C. Zemlin, Bostik Division, USM Corp. Biographies ................................................................................................................................_.................. 108 Manuscripts Not Received Prior to Publication COMPARISON OF THE ATTRIBUTES OF THE BEMA AND THE ZIMMER HOT MELT COATING TECHNIQUES

Allan Mann , Incopa Industries, Inc.

y 1 blirhe:i ^n JOURNAL. OF COA'TEA FA^^' =s. Vci.3 no,I, July !1^

f^.3! 37,

Case 1:06-cv-00751-MPT

Document 52-3

Filed 07/22/2008

Page 5 of 11

DEVELOPMENT OF A 100% SOLIDS URETHANE FABRIC COATING PROCESS

by
John C. Zemlin

BACKGROUND

Fabric coating today is a very large and very diverse industry and a brief description of it may be helpful in delineating those areas where our process may find applicaIt has been estimated by one market researcher that in 1972, 60 million square tion. yards of urethane coated fabric were produced in this pountry. Approximately one quartet of this production was "direct coated" fabric while three quarters were prepared by the so-called " transfer " technique. A more complete breakdown is shown in Table I. Essentially all of the urethane coating today is ing resins . most of the urethane coating resins used Some weight, fully reacted, thermoplastic urethanes. of the two-part urethanes which were the basis of the a few years ago. done with solvent solutions of coatin this country are high molecular use, however, is still being made first "wet look" coated fabrics of

In direct coating, the resin solution, a viscous liquid, is applied by knifing directly to the fabric; usually a floating knife, a knife over blanket or a knife over roll-coating technique is used. These coating techniques are illustrated in Figure 1.

EST19Aro Q VW -AW%WW c a= AIsm -1172
34'M huu us of WA

St CTS
war,4^

!sI CQAT110
1m -gtt

Dina O*TING · TOtAL

16

2

1.3

TQANSctR CAA1 * TOTAL
APM2tL BOOTS

44
22 6

2
L5 3.0
1.3

1.3
1.0 1.3 .

SHOE

UP= HANDBAGS

3
7

2.0

2.0
3.3

1.0

UPHOLSTERY

6

3.0

+^ ?Aiw^ywyr.^ ur^,a dn/,r, ^,w/j' ·.^.^,: !/^' ,, ^t . x: Figure 1

Table I

Usually several light coatings, followed by passage each time through a drying oven, are applied in sequence . This is necessary to prevent excessive strike through which would produce a very "boardy" hand in the fabric and also to permit complete solvent evaporation from each layer before the next is applied. Depending on the thickness of coating needed , four, six, ten and even more individual applications are used to build up to the final desired coating thickness. Oven temperatures are limited to those where boiling and bubbling in the coating will not occur. Most frequently, solvents such as DMF, tetrahydrofuran, methyl ethyl ketone and the like are necessary to obtain reasonable resin concentrations in the coating solution.

101

Case 1:06-cv-00751-MPT

Document 52-3

Filed 07/22/2008

Page 6 of 11

*5 i649 C^k Opir jim WQ{1na U10 .NO SOvi* Uss

The major disadvantage of our coating process is the requirement for minimum coating thickness. This makes it essentially impossible to prepare the very lightly coating nylon fabrics used for many athletic outerwear garments and also for lightweight tenting and A problem also exists in transfer tarpaulins. coating in duplicating leather grains with deep profiles. Alligator skin is a typical example. In this instance , the contours of the pattern paper vary by 4 or 5 mils, and since the applied coating must be about 2 mils thicker than the highest contour, a very heavy coating results.

ad IN

40 eximsal+o Air Loss wfaa 01 0&- Rssm C 'cw ' DI$ADVANTAGE3

PROCESS DESIGN CONSIDERATIONS

Conk.g Thidcrwss 2k miis rfa,, 2-Rut Urethane.. Aoppen.ar'

The use of a solvent -free liquid urethane in fabric coating both limits the type of equipment and necessitates rather precise control of several physical and chemical characteristics of the liquid system. Almost all solvent-free urethane systems have high viscosities, measuring in the thousands of centipoises. Viscosities such as these, combined with a curing mix of limited "open time" preclude almost all coating techniques except knife coating, extrusion coating and variations of these. Our work has been essentially limited to knife coating using equipment as indicated in Figure 2.

Table II

A second restriction we have placed on our process is the need for operation at room temperature. Many urethane systems are based on materials which crystallize or congeal Although it is possible to provide heating to the coating section of at room temperature. a coating line, the resulting complications such as difficulty of access to the knife, and shortened "open time" seemed sufficiently troublesome to cause us to avoid them by using systems which are liquid at ambient temperatures. As we proceeded in our development we found a number of practical problems which had to be met. These include , as we proceed from the first application of mixed material to the "web" or "substrate"; ready mixing in the bank before the knife, easy knifing and freedom from strike through or from drawing back into droplets (direct coating and transfer coating respectively). These problems will be considered in some detail in the following. · Since many of the above problems relate to viscosity of the mixed urethane system, let us first examine the viscosity of one system which has proven to be useful in our process . The viscosity of this material, which we call System A, has been determined on a Haake Rotovisco, RV 3 using their PK Viscosity Sensors.* The PK units are plate and cone sensors , the cone having an angle of 0.3°. The spring loaded plate fits closely into a massive metal block which is jacketed for heating and cooling. The unit is of particular advantage in that only about 0.1 gms of mixed urethane is needed for a viscosity determination - thus avoiding exotherm problems which make viscosity determinations difficult with most other viscosimeters. The data obtained is shown in Figure 3 and indicates the viscosity as measured at 20°C. at several shear rates as the mixed material ages and gradually gels. It is quite Br en er International, P.O. Box 128, Rochelle Park, N.J. 07662

I I I I I I I I I I I
I I I t

102

Case 1:06-cv-00751-MPT

Document 52-3

Filed 07/22/2008

Page 7 of 11

I I
I

I i
I

Transfer coating requires similar equipment to that needed for direct coatin Transfer coating differs in that the coating is applied to a pre-patterned or emb_ release paper . Since the "strike through " is not a problem , heavier coatings can be applied to the paper, although again the amount of coating is limited by the necessity of obtaining complete release of solvent before recoating. And the thicker the coating, the slower the solvent release. With light-weight goods for the apparel market, typically a heavy coating is applied to the release paper such that a dry thickness of 1 mil is obtained. After passing the release paper through the drying oven, it is returned to the coating head and a second coating is applied. This is an adhesive coating which may or may not have the same composition as the first coating. Before substantial drying occurs, the fabric is "nipped" into the adhesive layer. The combination of release paper, coating and fabric then passes through the heating tunnel again, and at the end, the fabric with the coating adhering to it is rolled up separately from the release paper. Thin coatings of 1 1/2 mils may readily be applied by this process. As the coating thickness goes up, however, so does the difficulty in carrying out the transfer coating operation. Multiple coats must be applied and considerable care given to complete solvent removal. There are additional problems associated with present techniques for direct and transfer coating and among them are the high cost of'·the solvents used, the hazards associated with the solvent use, and the enormous amount of heated air needed to safely exhaust the solvents . For example , a single coating line can easily consume 3 ,000 lbs. of solvent iri eight hours of operation. The existence of these problems, in the coating industry, was recognized some time ago by USM Corporation ' s BOSTIK DIVISION and we were asked to develop a fabric .coating process which would obviate them. DESCRIPTION OF PROCESS The following is a summary of part of our work, in response to the above request, on a fabric coating process which uses solvent-free, two-part liquid urethane systems. The process has been carried through the pilot scale of development using equipment as diagramed in Figure 2. The advantages and disadvantages of the process as compared to the present solvent coating technique can be summarized approximately as in Table II. In almost all cases our process can provide the desired coating in a single coating operation. The only exceptions would be in instances where a coating is required on both sides of the fabric or where a top coating is required having different properties, say higher hardness or different color, than those of the bottom coating. In general, most of·the present urethane coating lines for fabric can be utilized for our coating operation with the addition of a urethane dispenser . Since no solvent is used, the very large amounts of heated air which were formerly vented to the atmosphere need not be used , and only a small amount of - heated air recirculation is required to-control the curing temperature.
Figure 2

i
i
I

I I I

An additional theoretical advantage is the possibility of obtaining novel affects as by the use of serated blades, air jets and the like.

103

Case 1:06-cv-00751-MPT

Document 52-3

Filed 07/22/2008

Page 8 of 11

100,000
9o,c0o

, Viscosity Change of
Compound A at

obvious that the system is non-Newtonian in At low shear rates it has a very character. high viscosity which decreases sharply at higher shear rates.

moon

taco. Szc"

Sono sr G a FIFO Vc-

During operation of our coating process, a freshly mixed material such as System A is applied to the moving web in front of the knife. The knife then permits an amount of material approximately equal to the gap between the tip of the blade and web passing below to flow on with the web. It is necessary, however, that a reservoir of material be present behind the knife in order that the knife can meter an even coating. This reservoir, called the "bank", is formed by supplying a momentary excess of material over that needed for metering. In order that the bank be maintained at a fixed size, of course, the average feed to the bank must equal the amount metered out to the web. The steadily increasing viscosity of our systems requires that good mixing be obtained in the bank since for best operation the metering knife must have a reservoir of uniform vkscosity material. Mixing does not occur unless a shear rate of reasonable magnitude exists. The shearing force in the bank is imparted by the web as the bank is restrained by the knife. A very approximate shear rate in the bank can be calculated by assuming the bank to be a cylinder rotating about a fixed center with the speed of rotation equal to the web speed and that all of the shearing force is due to the fixed knife. The equation becomes:

10o

200

300

400

700

boo

700

EIOo

Tina Afker Mixing -SEC
Figure 3

Shear Rate

Difference in velocity over sheared layer _ Web speed Thickness of layer Banc radius

Although the actual situation is more complex, with highest shear rates occurring at the outside of the bank, this approximation does show patterns of change and is probably
of the right order of magnitude. tion has been made for web speeds yards/minute using bank diameters and 2". The results are shown in The calculaup to 20 of 1/2", 1" Figure 4.

Low speeds and large banks yield very low shear rates which are associated with high viscosities.
A second factor peculiar to our situation is the need to flood the bank when feeding rather than supplying uniformly thin layers equal to the metering thickness. This is necessary because fresh material has a relatively lower viscosity and if present in a thin film will tend to flow under the bank and be metered out under the knife without actually mixing into the bank. Flooding can be accomplished 6y the slow traversing of the mixing head or by single point deposition of mixed material.

EFFECT C M SPEED AND MNK SIZE OR $NEAR RATE 1EASE Of PNXMIG) IN SANK

10

20

AVG. SHEAK RATE -SECT 30 40

Figure 4

It is important for good mixing as discussed above, and, for easy knifing, that the material in the bank be at a relatively law viscosity level.

104

Case 1:06-cv-00751-MPT

Document 52-3

Filed 07/22/2008

Page 9 of 11

i

Lowest viscosities are associated with A calculation of average freshly mixed material. residence time in the bank can readily be made if we assume equal input and output at the bank along with perfect mixing, i.e. no "slippage" of freshly mixed product under the bank and out under the knife., The equation is: Average residence time Volume of bank Rate of a ing new material Cross sectional area of bank x unit lien th Web velocity x knife opening x unit width (As mentioned, the input of new material must be balanced by an equal loss of material under the knife if the bank is to remain at the same size.) The results of these calculations are shown in Figure 5.

I I
M

11
s

The matter of bank size is thus quite i.mportant, since it affects residence time, and thus bank viscosity and because it directly affects bank shear rate or bank mixing. Further, exothermic polymerization is greatly facilitated by larger bank sizes. These factors of bank size , web speed and knife opening actually exert a "multiplier effect" and if not controlled can cause an "avalanche" type of gelling reaction in the bank. We have found with System A that a bank size of about 1 1/2" is the maximum possible at a web speed of 5 yards/minute. At slower speeds or with larger banks at this speed, mixing will not occur in center of the bank and this portion of the bank will tend to gel and solidify.

50 100 IS0 700 250 303

350 400

Avlua Rvsidewr Time -SEC

I I I I
I

Figure 5

Easy knifing is also obtained if the material in the bank has a low viscosity under the high shear condition of knifing. Figure 3 shows that System A has a relatively low viscosity under shear rates of 1,000 sec.-1 or more , and that this condition exists for an
appreciable length of time after mixing. The approximate shear rate under the knife can be calculated in a manner identical to that above, i.e. Web s peed Shear rate =

Gap between knife and web

SMEAR SATE PEVEIOPEp AT VARYING V O SPEEPS AND 1(14IWE SETTINGS

16 14

W1e SPEED 11 APIN

The calculated shear rates for various web speeds and knife settings are shown in Figure 6. Shear rates of the order of 1,000 sec.-1 are obtained at higher web speeds and at smaller settings. Since commercial speeds will be at the high end of this scale most knife clearance settings will produce shear rates in the 2,0003,000 sec.- range. Thus, quite low viscosities will result during knifing. Fabric "strike through" in coating is a function of time and viscosity at the applied shearing force for any given fabric and coating. It becomes a potential problem as soon as mixed material is deposited on the moving fabric before the knife. We have found that no appreciable strike through of our systems occurs through

I I I
I

1000

3000 2000 Shear cat: · sec,,

4000

Figure 6

105

Case 1:06-cv-00751-MPT

Document 52-3

Filed 07/22/2008

Page 10 of 11

2 1/2 oz., 210 denier, plain weave, nylon fabric when the viscosity is about 20,000 cps. and the contact time is less than 30 sec. At 10,000 cps. strike through occurs in about Since the fabric is usually in contact with mixed material for only a fraction 15 sec. of a second before reaching the coating knife, initial strike through is not a problem. Under the high shear of the coating knife, a substantial reduction in viscosity At 20°C. viscosities of about 10,000 to 15,000 cps. exist. This reduction peroccurs. mits easy knifing and again owing to the brief exposure to high shear conditions, strike through is rarely a problem. At higher ambient temperatures, still lower viscosities- at 40°C.,a little below 10,000 cps. with System A - are encountered as shown in Figure 7. Even, here and even on quite open fabrics, strike through does not usually become a problem with the proper knife selection. One consideration in all fabric coating is anchorage of the coating to the fabric. Good anchorage is rarely obtained without considerable penetration into the interstices of the fabric, that is without surrounding or enveloping some of the yarns with the coat ing. Such penetration does not affect the flexibility or drape of the fabric to the same degree as penetration into the yarns. The novel character of our coating systems does permit penetration into Fe fabric interstices under high shear but without significant penetration into the yarn.
The amount of penetration into the fabric can also be controlled by the position of the knife blade above the fabric. Thus, a square edged blade set at 90° to fabric will' cause only modest penetration into the fabric. If the leading edge of the blade is tapered to a 45° angle so that the blade exerts a wedging or "pumping" effect on the coating passing beneath, more penetration is obtained. In general, we have found a tapered blade of one type or another to be desirable in most types of coating, giving better coating anchorage on fabrics and better pattern reproduction on transfer coating.

1oopo° Week of Tsmpsraurs on 9opo0 Viscosity hnn9s} of Compound A at Twohaar Kotss

gg00o eoaca,

so,aoo

Comparison^v%costty Chan;¢ of L_ .1 Two Uu ibis Compounds A Low And Nigh JWbigri Ta^npen }tvrrs^.

20'

aoo

Tim. After Mixing -seC
Figure 7

Time After Mixing -sec
Figure 8

106

I
I

Case 1:06-cv-00751-MPT

Document 52-3

Filed 07/22/2008

Page 11 of 11

I I I I I i I I I I t
I

The most trouble we encountered in our coating development was in viscosity control Curing ovens are large expensive objects and they are usually the limitduring heating . Since urethanes cure more rapidly ing factor as far as coating line output is concerned . at high temperatures , it is obviously desirable to have the oven at as high a temperature This method of operation means, however, that the coating applied at room as practical . temperature in a smooth thin film is suddenly exposed to hot air or to radiated heat so that its temperature rises to perhaps 100°C. in as little as 30 sec. The difficulty is most noticeable on the silicone treated transfer paper where if any substantial reduction in viscosity occurs, the coating will tend to pinhole or draw back into droplets . Our present coatings do not decrease i n viscosity to the extent that this becomes a problem. Although only one compound, our System A, has been discussed so far , we have actually developed a series of useful compounds for our process. Extensive viscosity data has not been obtained for most of them since rapid screening for viscosity " fit" is normally all that is needed to assure their performance in our coating process. Viscosity curves comparing System A, a Shore A 55 compound with a second useful compound , a Shore A 85 material labeled System B are presented in Figure 8. The close similarity in viscosity/time behavior is evident and is typical of all of the compounds useable in our process.
This process is still considered to be in development status by USM BOSTIK DIVISION and neither samples nor chemical details of the process are available . Patent applications have been filed.

Acknowledgement
I wish to thank Robert H. Miller of USM's BOSTIK DIVISION for assistance in much of the experimental work ; E. C. Stanley of Chemicals and Pigments Equipment Co., Newton, Mass. for advice and assistance in operating the Haake Rotovisco , RV-3; and R . Battersby of USM's BOSTIK DIVISION for advice in the rheological area.

I I
I

I I

107

Case 1:06-cv-00751-MPT

Document 52-4

Filed 07/22/2008

Page 1 of 9

EXHIB I T 3

Case 1:06-cv-00751-MPT

Document 52-4

Filed 07/22/2008

Page 2 of 9

=+x#
=r at/
J

s-iiT 021

rtheaster'n University Libr·ar4 Seria 0 Hunt i nQton Ave

oton MA O2i i5-5005

L!ill

T"fl,t

S1

{,

(A

tli &

1

*kil

l 1 !
1 ;

4

w9m

Case 1:06-cv-00751-MPT

Document 52-4

Filed 07/22/2008

Page 3 of 9

85

Predicting In-Service Weatherability of Automotive Coatings: A New Approach-D.R. Bauer (1996 Mattiello Memorial Lecture)
The author presents a formalism for estimating in-service failure rates in coatings. The formalism involves developing an analytical model for time-to-failure in terms of measurable material, process, and exposure variables.

97

Mechanistic Considerations of Particle Size Effects on Film Properties of Hard/Soft Latex BlendsS.T. Eckersley and B.J. Helmer (First Place Winner in 1996 Roon Awards Competition)
This paper is a review of percolation theory as related to latex film formation. Effects of particle size and modulus are shown to be dependent variables contributing to block resistance of formulated films.

hi" 109 Rheology of Waterborne Coatings-R.D. Hester and D.R. Squire, Jr.
Presented is a survey of the rheology and viscosity control mechanisms of latex paints with emphasis on application properties and film formation. The mechanism of cellulosic and associative thickeners are discussed.

10 12 13 25 51

1997 Roon Awards Competition Underway: Entries Invited FSCT Invites Nominees for 1998 Mattiello Lecturer; and Student Papers Invited for 1997 A.L. Hendry Awards Competition Interview with FSCT President Jay Austin 1996 Annual Meeting and International Coatings Expo and Technology Conference Wrap-Up Fall 1996 FSCT Board of Directors Report

4 JCT Guide for Authors ............................................... 7 Comment ................................................................. 8 Technical Abstracts ................................................... 16 Industry News . ......................................................... 19 Meetings Update ...................................................... 49 Regulatory Update .................................................... Product Overview-Production Equipment ................... 122 Society Reports ... ............................................ ......... 125

Future Society Meetings ............................................. 130

New FSCT Members ................................................. 131
People in the News . .................................................. Obiturary ................................................................ New Products - ........................................................ Calendar of Events ... ............................. ................ _ Humbug from Hillman ............................................... 133 133 135 137 140

' ON OF S C TIES OR COATINt' S CHNOLOG!
3'` RNA ti s T " o 9 stage o s. LOG ISS 6 .8773) 5 p iblIs ed on l y Federation pf Societies f of ' I tG 61 Q) 940.0292 S ^# l P' 1 422 50 Pit ne 6-I 0 9 0-4 N orr s " B ^ an 9 ad t^ 1^ lingo ices . POST ER Senn 'address changes to a au l 1 year, '$9o,;.2years

$ 70. Europe )Ai o Via an, c ' a-^r, $60 yea $ 3Q 3 years. s year "fr y; i %ears, 5140; 3 years, $2 f ` 0 ter yntrtes ea s

Case 1:06-cv-00751-MPT

Document 52-4

Filed 07/22/2008

Page 4 of 9

F

Flame - Retardant Brominated Styrene-Based Polymers. X. Dibromostyrene Grafted Latexes
N.A. Favstritsky and J.-L. Wang-Great Lakes Chemical Corp.'

INTRODUCTION
atural and synthetic latexes are commercially used in a variety of coatings. In many coatings applications, the latexes with flame-retarding properties are desired. This applies in particular where latexes are used in textiles, carpetings, paints, clt'arcoatings, adhesives, sealants, caulks, nonwoven binders, etc. Until recently, coatings users have generally had to use flame retardants that were developed for other applications. All of these products, either liquid or solid, have had an adverse effect upon the quality of the coatingg. The usual method by which flame-retardant (FR) properties are imparted to latexes is the blending-in of flame retardant additives, but disadvantages, such as phase separation upon aging, plasticization and migration to the surface, coagulation and sedimentation, generation of strong white pigmenting and settling out effect, toxicity resulting from the presence of antimony trioxide, have been associated with this approach. The best solution of these problems appears to be the elimination of additive flame retardants by incorporating flame retardancy into the polymer used for coating. We have previously reported'-12 the synthesis of a variety of dibromostyrene (DBS) -based FR latexes. These latexes were prepared from DBS and other monomers such as (meth)acrylic acids and their esters, butadiene, vinyl acetate, and styrene. Some of the challenges have been cost, flame retardancy, coating properties, latex stability, and monomer reactivity. However, if a practical and useful FR latex can be produced by simply grafting the DBS monomers onto a commercial latex or mixture of latexes, then the cost of producing such an FR latex can he significantly reduced. N Although a vast number of grafted polymer latex compositions has been described, 3',-` none of them have been identified with polymer latex modification with bromine for flame retardance. Some of the grafted latexes are claimed to have value in improving bacteria resistance for use in paints or other coating compounds,'improving adhesion of the modified latex to silicate glass, aluminum foil, cellophane, and textiles zt7 increasing the
Presented of the 23rd Annual Woler'oorne. High-Sotds. and Powder <: clings Symposium. on lebruory 14-16. 1990. in New Orleans, LA.

This paper describes the synthesis, characterizatiort, and some applications of f1aine-retardent dibromostyrene grafted latexes. These latexes are synthesized by an. emulsion polipnerization techniqu