TECHNICAL REPORT 4704.CO.Y.NO... . ..... , ..

SUBSTITUTION OF ALUMINUM FOR MAGNESIUMI;

II 'V '

AS A FUEL IN FLARES

B. JACKSON, JR.F. R. TAYLOR ~

R. MOITOS. Md. KAYE V12U

JANUARY1975

APPROVED FOR PUBIUC RELEASE; DISTRIBUTION UNLIMITED.

PICATINNY ARSENAL,"DOVER, NEW JERSEY

The findings in this report are not to be construed

as an officlial Department of t'be Army Position.

DISPOSITION

Destroy this report when no longer needed. Do notreturn it to the ocihlgntor.

M=01 forOTIS illi #0100

I....................--

$131 AT -dC. U 4 '4f

UNICLASSIFIEDSECURITY CLASSIFICATION OF THIS PACE Men.. Dat 'fttor4-I ~REP~jRT DOCUMENTATION PAGE SWORE COMPLKEnqapORlg

I.REPORT MUMUER 2.JT 0 - GOVT ACCtSSI , rr!hT.

7 Technical ef;t_ 04 -VolI - ,,

SUBSTITUTION OF4,,LUMINUM FOR.kMAGNESI UM"# AS AFUEL IN FLARES, 0. PERFORMING ORG. REPORT HUMSER

7. AUT ~R a S. CONTRACT ON GRANT NUMBER(03

165 8.,Jackson, Jr., F.R/Taylor, R./(MottoS.ML.Kaye

NAME AND ADDRESS I

Feltman Research Laboratory NIVMCV 101.43.Picatinny Arsenalt, Dover, NJ ACVS4'1.643.

11. CONTROLLING OFFICE NAME AN4D ADDRESS 1.EOTDT

ICMONITORING AGENCY NAME 0 ADORESSIf different frmm Control md Office) IS. SECURITY CLASS. (of this report)

Unclassified1114. DECL AISI F1CATION/ DOWNGRADING

SCHEDUJLE

IS. DISTRIBUTION STATEMENT (of this Report)

Approved for public release; distribution unlimited.

17. DIST RIS1UTION ST ATEMN NT (of the abstroct entered In Block 20, It dilfwavn-A from. Report)

IS. SUPPLEMENTARY MOTES

IS. KEY WORDS (Conthna. on roew.. aide it necessar and Identify by block number)

Flare Magnesium Self-heating Transition metal compoundsAluminum Laminac Trip flare FuelSodium nitrat Viton Hydrogen 4Exotherm Gassing Tungsten

t0. ADSTRACT (ConilaýVan 'overea aidb It noeeoaivy and Icdenttiy by bleek number)

A very broad Investigation has been conducted on the* .~utilization of powdered aluminum as a substitute fuel for atomized

magnesium in pyrotecnnic Illuminant compositions developed primarily

for the M49AI trip flare.~

O)D 1473 EDITION OF I NO0V 69 5 OISOSOLElE UCASFE2 SECURITY CLASSIFICATION OF THISl PAGE (Whon Date Entered) 1

UNCLASSIFIEDSECURITY CLASIFICATION OF THIS PAGaM,, 90- O...

20. Abstract (Continued)

Studies were conducted on the combustion and propagativecharacteristics of powdered aluminum/sodium nitrate mixtures.These studies included the effects of particle size, fuel-oxidant ratio,binder concentration, additives, fuel coating, flare case materials,flare case coatings and moisture.

Thermal analytical studies of aluminum/sodium nitratesystems provided conclusive evidence of a low temperature exotherm(70'C-135 C) which produces self-heating of the compositions in thepresence of moisture.

A formulation utilizing powdered aluminum as a substitute fuelfor atomized magnesium has been developed for use in the Army M49A1trip flare. The burning characteristics of the composition containing35% six micron powdered aluminum, 53% sodium nitrate, 7% sevenmicron tungsten powder and 5% Laminac 4116 polyester are adequateto meet the requirements of the M49A1 trip flare. By careful exclusionof moisture during processing, this composition demonstrated long termstorage stability at elevated, ambient and low temperatures.

UNCLASSIFIEDSECURITY CLASSIFICATION OF THIS PAE(IRWM. DMOa EnaImW

N '-

-.---

--

The citation in this report of the names of commercial firms orcommercially available products or services does not constitute officialendorsement or approval of such commercial firms, products, or servicesby the US Government.

FoewrdTABLE OF CONTENTS

i i •.i:Page .No.

EV, Foreword I

Introduction 3

Results and Discussion 4

Gassing and Self-heating of AI/NaNO3 compositions 8

Conclusions 11

Experimental Procedure 11

Blending 11Loading 11Testing 12

List of Materials Used 12

References 14

Distribution List 37

Tables

1 Performance characteristics obtained March 1968for AI/NaNO3 additive compositions 15

2 Effects of manganese carb(,nate and Laminac onperformance characteristics of aluminum/sodiumnitrate compositions 16

3 Performance characteristics of AI/NaNO3 compo-

sitions in various coated flare cases

a Paraffin coated Kraft paper cases 17

b Viton A coated Al trip flare cases 18

ii ,

-I

'V. ~1

c Laminac 4116 coated Al trip flare cases 19

d Laminac/asbestos coated Al trip flare cases 20

4 Comparison of past and present performancecharacteristics of AI/NaNO3 ,/2% MnCO3 /5% Laminaccompositions 21

5 Effect of case material and case liners on per-formance characteristics of aluminum/sodiumnitrate/manganese carbonate/Laminac 4116(35/58/2/5 composition 22

6 Effect of various additives on performancecharacteristics of aluminum/sodium nitrate/Laminac4116 in M49AI trip flare case 23

7 Effect of aluminum content on performance charac-tc.ristics of aluminum/sodium nitrate/Laminac 4116with optimum additives in M49A1 trip flare case 24

8 Effect of ferric oxide on the performance character-istics of aluminum/sodium nitrate/Laminac 4116composition (38/52/5) in M49A1 trip flz,-e case 25

9 Performance characteristics of aluminum (Alccaand Alcan)/sodium nitrate compositions 26

10 Effect of magnesium as an additive on performancecharacteristics of aluminum/sodium nitrate/tungsten/Laminac 4116 composition in M49A1 trip flare case 27

11 Sixteen-hour temperature conditioning test on"performance characteristics of uptimum aluminumfueled composition in assembled M49A1 trip flare 28

12 Four-month stor.age surveillance of performancecharacteristics of optimum aluminum fueled compositionin assembled M49A1 trip flares 29

.. iii j lk.

13 Amount of hydrogen generated by powdered"aluminum 30

14 Effect of isostearic acid coated aluminum powderf"', on performance characteristics in M49/\1 trip

flare case 31

Figures

1 DTA showing early appearance of low temperatureexotherm in AI/NaNO3 composition after exposureto 52% RH atmosphere, ignition temperature - 915SC 32

2 DTA showin-g low temperature exotherm developedby AI/NaNO3 composition after exposure to 79% RHatmosphere 33

"3 Development of 135 0C exotherm in AI/NaNO3 compo-sitions exposed to increasing amounts of moisture.All DTA's employed 2 g samples of 45/55 AI/NaNO3

* at heating rate of 9.6 0C/min, X--axis = 5.0 mv/in.,Y-axis = 0.5 mv/in. 34

S4 DTA showing the thermal behavior of a complexAI/NaNO 3 system conditioned at 50% RH 35

5 DTA showing the appearance of a low temperatureexotherm (95 1251C) in a co.iaplex AI/NaNO 3composition after conditioning at 90% RH 36

iv

ii

U

FOREWORD

Shidlovsky in his memorable treatise entitled Fundamentals ofPyrotechnics (Ref 1) appearing in the Russian book Osnovy Piroteckhniki(1964) states:

"The combustible substances used in pyrotechnic compositionsmust satisfy the following basic requirements:

1. They must have a heat of combustion providing forthe best special effect of the composition;

2. They must oxidize adequately by means of the oxygenin the oxidizer c " by means of the oxygen in the air.

3. They must give products in combustion that providefor obtaining the best special effect of the composition,

4. They must require the minimum quantity of oxygenfor their combustion;

5. They must he chemically and physically stable in thetemperature interval from -60 to +601C, and be as stable as possible tothe effect of weak solutions of acids and alkalis;

6. They must be nonhygroscopic;

7. 1 hey must be easily pulverized;

ognsC. They must not have any toxic effect on the humani •,•iorganism;

9. They must be easily obtainable materials that arenot in critically short supply." p

If we examine the chemical and physical properties of the variousmetals available, we arrive at the fact, as also pointed out by Shidlovsky,that the following six substances constitute high energy combustibles:lithium, beryllium, magnesium, calcium, aluminum, titanium, andzirconium. If we apply Shidlovsky's criteria, lithium and calcium are

4i 1 -

& F

unusable pyrotechnic fuels because of their energetic reactivity to themoist jre and oxygen of the air. Beryllium, as always, Is avoided becauseof its extreme toxicity. Titanium and zirconium are used in igniter anddelay applications ra'her than flares because of their high cost. Conse-quently we are IEft with magnesium and aluminum.

Classically, magnesium is superior to aluminum as a pyrotechnicfuel for two reasons. First, magnesium does not have a protective oxidecoating but aluminum does. Second, magnesium has a higher vaporpressure than aluminum, as shown below, which allows it to vaporizerapidly ance to be oxidized by the oxygen of the air. Aluminum vaporizeswith difficulty and is generally spewed out of the flare as incandescentparticles.

Vapor PressureMetal 1 Torr 10 Torr 760 Torr

Mg 621 0C 7113 0C 1 I07Cc

Al 1284PC 14870C 20560C

In the following report, the possible use of atomized aluminum powder-additive systems as a potential alternate fuel for magnesium is discussed.This study shws that a substitute aluminum based formulation has beendeveloped which exceeds the minimum performance requirements of35,000 candles for 55 seconds in the M49A1 trip flare. This compositioncontains 35% six micron atomized aluminum, 53% sodium nitrate, 7% sevenmicron tungsten powder and 5% Lamlnac 4116 polyester resin.

The two basic advantages of aluminum over magnesium are cost

(38 cents per pound as compared to 85 cents for 20/50 mesh atomizedmagnesium) and ready commercial availability.

The disadvantages of the aluminum system, when exposed to moisture,are significant gassing, possible self-ignition. By careful exclusion ofmoisture during processing, this phenomenon can be eliminated, as wasdemonstrated by long-term storage stability testing.

2

INTRODUCTION

F The present study was conducted to determine if a compositioncould be developed for the Army M49A1 trip flare containing aluminumas a substitute for magnesium. The primary justification for this programwas based on cost and availability of aluminum as opposed to magnesium.Atomized magnesium powder is primarily produced for pyrotechnic appli-cations, while powdered aluminum is mass-produced for a host of applica-tions ranging from pigments in paints to energy constituents in propellantsand explosiveb.

Although the study was directed primarily toward the trip flare, itdelves into all aspects of flare technology. For example investigationswere made of the effects of particle size, fuel-oxidant ratio, bind:r con-centration, additives, fuel coating, flare case materials, flare case coatings,moisture, and short and long-term stability.

Previously, aluminum was not considered to be a good fuel for flareapplications since mixtures of atomized aluminum (diameter > 15l.) andsodium nitrate ignite poorly, propagate poorly, and burn very erraticallyand inefficiently. One of the reasons for the inefficiency of the AI/NaNO3

system is that it produces a profusion of incandescent particles. ThisRoman Candle effect makes the flare an incendiary device as well as anillumination source. This incendiary aspect is very undesirable, possiblycausing harm to users and starting ground fires.

A few years ago, however, it was demonstrated that powdered

aluminum in combination with sodium nitrate (Ref 2,3) will burn pro-pagatively when small quantities of various transition metals or metal

compounds are incorporated into the mixture. These studies showed thatthe transition metal compounds affecting the AI/NaNO3 reaction fell intofive classes. Class I compounds are those which increase the efficiencyof the system by decreasing the thermal conductivity of the basic binary.Class 2 compounds, consicting of manganese oxides, catalyze the normaldecomposition of sodium nitrate to evolve oxygen at low temperatures.Compounds which cause ignition of aluminum at about 700 0C (a value well)below its normal ig ifition temperature of 10000C) are in Class 3. Class 4embraces transition metals which ignite with sodium nitrate at lowertemperatures than aluminum. Finally, Class 5 compounds are those whichalter the normal decomposition pattern of sodium nitrate, causing theevolution of the oxides of nitrogen at the melting point of the nitrate.

3

Wk.

I IN I I I IQM

wiv '

RESULTS AND DISCUSSION

As stated in the Introduction, previous studies showed that thecombustion properties of AI/NaNO 3 compositions can be dramaticallyimproved by the addition of certain additives (Raf 2,3). Since aninefficient Mg/NaNO 3 /Laminac composition* is presently employed in theM49A1 trip flare (higher candlepower apparently not required), it wasnot difficult to meet the basic trip flare requirements of 35,000 candles

and minimum burning time of 55 seconds.

The performance characteristics of the basic Al /NaNO 3 /2% MnCO 3(previously found to be the best additive system) and the total systemAI/NaNO3 /2%, MnC03/5% Laminac binder 4116 were gathered from Reference

2 and are presented In Table 1. To simplify the analysis of this data, acomparison of the absolute values of the various performance character-istics is presented in Table 2 together with percentage changes in theparameters v it- respect to the basic binary system. It is seen that theadditive MnCJ 3 produced marked reductions in burning rates and causedsizeable-i-tcreases in light output and efficiencies. One interesting andunfortunate finding was that the addition of Laminac binder reduced thebeneficial effects of the MnCO 3 .

A series of Ai/NaNO3 /2% MnCO 3 /5% Laminac compositions wereloaded into paper and aluminum trip flare cases. The cases were coatedinternally with a variety of case coatings. The purpose of these experi-ments was threefold. First, to see if the performance characteristicsobtained were comparable to those previously obtained with flares madein 1968 (Ref 2). Second, to determine the effect of case material on per-formance characteristics. Finally, to determine if the Laminac/asbestoscoating is the optimum insulator for use with the mnetal trip flare case.The performance characteristics obtained in these tests are presented inTables 3a through 3d.

*A composition employing Mg/NaNO 3 /Laminac ir, respective proportionsof 65/25/10 will provide 154,000 candles for essentially 40 seconds durationand an efficiency of 46,000 candleseconds/gram.

4

As previousty stated, the first goal was to determine how thecharacteristics of tt.e present flares compared to those made six yearsago. Table 4 makes this comparison and shows that the performancecharacteristics of the flare are quite comparable, especially in their

* luminous efficiencies.

The second goal of these experiments was to determine the effectsof the two different flare cases. Table 5 shows the performance character-istics obtained using Kraft paper cases versus aluminum cases. Themetal cases used are those employed to make M49A1 production items.These cases are extruded and have a wall thickness of 1/16 inch. Thedata in Table 5 shows two interesting facts. First, changing the casefrom paper to aluminum results in only a 15% reduction in the luminousoutput of the flares. In fact, the luminous output was only minimallyaffected using aluminum cases, regardless of case liner material.Second, and more important, nearly equivalent burning rates wererealized in both paper and metal casesprovided that an excellent in-sulating liner such as Laminac/asbestos was employed. When eitherViton A or Laminac was used, the total achievable time was reduced from2 100 to 2 59 seconds. This demonstrates that without an insulator, heatis transferred down the metal tube causing preheating and rapid burningof the composition. Furthermore, the data of Table 5 presents an unusualcase in flare behavior. Normally, an increase of nearly 50% in the burningrate of a flare, such as that occurring when changing from Kraft paper toLamirnac coated aluminum, would have resulL.d in a significant increasein luminous output. In the present case, however, this did not occur.The luminous output remained approximately the same.

The next step in the program was to determine if MnCO 3 was thebest additive for the Ai/NaNO3 system when using an aluminum case asit was for a paper case. Using the efficient 40/48 AI/NaNO3 system as abase, a number of formulations was made and loaded containing thefollowing additives (based on previous studies, Ref 2 and 3): 5% ferricoxide (red), 5% chromium sesquioxide, 7% tungsten powder, and 3%magnanese carbonate. The performance characteristics obtained withthese formulations are presented in Table 6. The data obtained showsthat all of the compositions produced candlepower values in excess of35,000 candles, the minimum requirement. During these tests, asignificant observation was made. The flares containing ferric oxide andtungsten exhibited a marked reduction in the amount of incandescent

5

particles ejected. In view of this result, as well as the candlepowersproduced, ferric oxide. tungsten and chromium sesquioxide wereselected for further evaluation., New additive formulations were madeusing reduced percentages of aluminum powder in order to increase theburning time of the systems. The performance characteristics obtainedfor these new formulations are presented in Table 7 and show that severalof the systems give outputs meeting the minimum requirements of theM49A1 flare. Based upon these results as well as visual evaluation, thecompositions employing 7% tungsten powder (FY-1639) and 5% ferric oxide(FY-1634) were chosen as prime candidates for the M49A1 flare. Recon-firmation tests of these two systems showed that the tungsten system gavethe same results but the ferric oxide one did not. it was discovered thatthe change in the performance characteristics of the ferric o',ide systemwas due to the use of a new lot of material. From a purely cost advantage,it would be more desirable to use ferric oxide than tungsten. Consequently,some efforts were expended to characterize the old and new ferric oxides.An X-ray analysis of the original material showed the presence cf 10%silicon dioxide (alpha quartz form), while the new material containad nosilicon dioxide.

Table 8 presents the results of experiments to determine if otherforms of ferric oxide or ferric oxide plus silicon dioxide (Superflois,Johns-Manville Company) could be employed to produce desirable per-formance chr racteristics. It was found that the jeweler's rouge form offerric oxide, as well as silicon dioxide (Superfloss), gave performancecharacteristics approaching that of the composition containing unknowngrade ferric oxide. However, these systems were abandoned becauseboth of them did not reduce the profuse production of incandescentaluminum particles. It was decided to continue the investigation using thetungsten formulation because of its superior performance characteristics.

A comparative investigation was next conducted using a lower costatomized aluminum obtained from the Alcoa Corporation. This materialcosts $0.38/lb per 450 pounds whereas the Alcan material previouslyemployed was in excess of $1.00/lb. The results of this comparison arepresented in Table 9 and show that the lower cost Alcoa aluminum gaveperformance characteristics comparable to the more expensive 8V Alcanaluminum.

"6Ii

II

Before subjecting the tungsten additive system to the necessarystorage and safety tests, a final attempt was made to further improvethe luminous efficiency of the AI/NaNO3 /W system by adding smallpercentages of 20/50 mesh atomized magnesium. As Table 10 shows, theaddition of 5% or 10% atom!zed Mg did not produce any increase in theefficiency of the system while resulting in slightly increased burningrates. This data did not indicate any advantage in utilizing Mg in theAI/NaNO3 /W system.

The finalized 35/53/7/5 A!/NaNO3/W/Laminac system was loadedinto M49A1 flare cases and then incorporated into final end items. Twogroups of these end items wc-e subjected to two temperature extremes

for 16 hours. One group was maintained at 75'C (hot), while the otherk was kept at -51PC (cold). After conditioning at these temperatures, the

flares were examined visually and their performance characteristicsdetermined. Table 11 shows that no deleterious results occurred due toconditioning at these extreme temperatures.

Standard sensitivity characteristics were determined for the

AI/NaNO 3/W/Laminac system. The values obtained are as follows.

Impact test, PA apparatus -+ 21 inches

Friction pendulum, steel shoo -* no action

Electrostatic initiation, minimum joules -* > 1 .025

Ignition temperature, five second value -* 564PC

The sensitivity values obtained indicate that only normal safeguards arerequired during blending and loading operations.

A series of M49A1 trip flares were loaded using FY-1639 composition(35% atomized Al, 53% NaNOs, 7% W, and 5% Laminac resin binder). Thisgroup of flares was placed in storage for four months at both ambient and750 C. Table 12 shows the results obtained at monthly intervals over thefour-month period. The data obtained clearly indicates that no degradationoccurred in the performance characteristics of the Al composition duringthis extended storage period.

7

.77•

rr

GASSING AND SELF-HEATING OF AI/NaNO3 COMPOSITIONS

One of the basic properties of pyrotechnic compositions leading toinstability in storage is their reactivity to moisture. The troublesomegassing occurring in Mg systems is attributable to the well known reaction:

"5Mg+NaNO3 +8H 2 O-'5Mg (OH) 24 H2t+NH3 t+NaOH (1)

This reaction is the summation of the followiI ig two fundamental ones:

Mg + 2H-2 -0 _+" t + Mg (OH) 2 4 (2)

and 4 Mg + NaNO 3 + 7 H2 0 -) 4 Mg (OH) 2 -:- NH4OH + NaOH (3)

NHt + H2 0

The generation of NH3 and H2 gases by Equation 1, of course, is greatlyaccelerated with increasing temperature. Mun•.ion items in extendedstorage often suffer damage such as ruptured closure seals and splitcases due to this unwanted generation of gases.

The gaseous as well as thermal output of reaction 1 is moderatedby the insolubility of the magnesium oxide in the increasing alkalinity ofthe medium resulting from the products of NaOH. This growing insolubilityfollows from the rudimentary equilibrium process shown by the followingequation.

-~++ I&Mg (OH) 2- Mg + 20H (4)

As the concentration of the OH'ion increases, the equilibrium shifts tothe left favoring the formation of the protective coating or at least satur-ating the solution with Mg(OH) 2 , thereby preventing dissolution of themetal coating.

S Consider now the case of aluminum. The reaction of Al withmoisture and sodium nitrate is shown in the following equation.

3NaNO3 + 8 Al + 18 OH2 -*3NaOH + 8AI(OH).3 +-3NH, 3 t (5)

8

I

r7-v

Unlike the Mg reaction, the aluminum oxide coating and AI(OH) 3 formed

are quite soluble in the resulting alkaline medium. Consequently,

reaction 5 which produces NH3 (probably also some H2 by the reaction2AI + 6H 2 0 -* 2AI (OH) 3 + 3H2 t ) and heat proceeds at an accelerating rate.This latter process eventually causes extensive self-heating and, under

certain conditions, spontaneous ignition of the composition.

The Al composition developed in this prroram, FY-1639, exposed"to an uncontrolled ambient environment for a few days, generated(11 +)ml of gas in five minutes duration at 120 0C temperature using thestandard vacuum stability test. When another sample of this compositionwas dried at 105 0C for 16 hours prior to conducting the vacuum stabilitytest, only 0.35 ml of gas was produced after 40 hours at 120 0C. Thisresult demonstrates that a stable Al composition can be processed andloaded by rigorous exclusion of moisture.

I - ? Table 13 summarizes data obtained from the Alcoa Corporation

which shows the amount of gas produced by refluxing powdered aluminumin a 20% ammonium nitrate solution. This data shows the effect ofchanging the particle size on gassing and how this gassing can be signifi-cantly reduced by coating the metal with isostearic acid. When isostearicacid coated aluminum was employed in the FY-1639 formulation, however,it was found, as shown in Table 14, that a marked reduction occurred inboth candlepower and efficiency. For example, the candlepower wasreduced from 50,000 to 31,000 candles and the efficiency went from 17,000to 12,000 candleseconds/gram. Consequently, the more feasible methodto eliminate gassing is rigorous drying rather than coating the metal.

Although the gassing of the aluminum/sodium nitrate compositionis an annoying problem, the more serious danger is that of self-heating

;; l leading to spontaneous ignition. A good discussion of this phenomenonis given by Johansson and Persson in the Swedish journal Pyrotechnikdogen(Ref 4).

Figure 1 shows the DTA of a simple 45/55 Al (6gp)/NaNO3 compositionwhich had been conditioned at 52% RH for six days. As is seen, the pattern

' Ais quite complex having three primary exotherms at approximately 135sm,690*C, and 915'C. The last exotherm resulted in ignition. When the sameAI/NaNO3 composition was conditioned at a higher RH of 79%6 for the same

9

j iI

length of time, six days, a different DTA pattern resulted as shown inFigure 2. Here the small exotherm at 1350C grew tremendously in intensity,and ignition occurred at the second exotherm at 6900C. With increased massand confinement, ignition occurs easily in :he region of the first exotherm,at temperatures as low as 60 0C to 70 0C.

Figure 3 shows that the exotherm at 1350 C is solely dependentupon moisture content. In the DTA (a), no exotherm appears for aperfectly dry AI/NaNO2 sample. Some activity begins to appear in run(b) when the same AI/NaNO3 composition is exposed to a 79% RHatmosphere for 22 hours. Finally, the DTA in (c) clearly shows thedramatic growth of the exotherm using a sample which had been exposedto a 79% RII atmosphere for 66 hours.

In another program design to furnish a highly exothermic compo-sition for rocket assisted projectiles, an Al composition was developedcontaining the additive MnCO 3 . This highly metalized composition con-tained 85/10/2/5 Al (61.t)/NaNO 3iMnCO 3/Viton A. Figure 4 presents theDTA behavior for the basic composition which had been conditioned at50% RH for five weeks. This co.;lposition exhibited no low temperatureexotherm and ignited immediately following the melting of NaNO3 atapproximately 3000C. This ignition behavior is peculiar to aluminum/NaNO3 systems containing additives such as inorganic fluorides (e.g.,NaF) and organic fluorides (e.g. Viton). It has been hypothesized bymany that at the melting point of NaNO 3 , the additive either removesand/or makes the aluminum oxide protective coating permeable so thatan immediate reaction occurs between the clean Al surface and themolten oxidant.

When this additive composition was exposed to a 90% RH atmospherefor five weeks, the same low temperature exotherm occurred as with theAI/NaNO 3 binary. This phenomenon is clearly seen in Figure 5. Con-sequently, it appears that the presence of moisture sensitizes simple aswell as complex AI/NaNO3 systems. Sudies are presently being con-ducted to determh-e the nature of the intermediate which causes the lowtemperature ignition of AI/NaNO3 systems. This intermediate is probablyNaOH which reacts exothermally with both Al an"! its protective coating.

10

AA

CONCLUSIONS

A formulation has been developed for use in the U.S. Army TripFlare whic'i utilizes aluminum instead of magnesium for a fuel. Thiscompositioti contains 35% six micron atomized aluminum, 53% sodium

nitrate, 7% seven micron tungsten powder and 5% Laminac 4116 polyesterres i n.

Two basic advantages of the new aluminum system are cost (38 centsper pound for aluminum as compared to 85 cents a pound for 20/50 meshatomized magnesium) and reat"y commercial availability.

The disadvantages of the aluminum systern are gassing and self-heating which occur when it is exposed to excess mcisture. By carefulexclusion of moisture during processing, however, these effects can beeliminated as was demonstrated by long-term storage stability testing.

EXPERIMENTAL PROCEDURE

t ~Blending!

The compositions cited in this report were blended in a Lancastercounter-current mixer, model PC, which imparts a mulling-type actionto the constituents of the blend. The sodium nitrate was dried for 24hours at 110 0 C prior to use. The resultant flare compositions were driedin trays containing a one-inch thickness )f composition for a minimum of16 hours at 105IC.

Loading

Each composition was loaded in four equal increments at a pressureof 10,000 psi. Ten grams of DP-1886 igniter composition (65% five microntungsten powder, 24% barium chromate, 10% potassium perchlorate and1% Vinyl Alcohol Acetate Resin MA-28-180 were used as the first fire.

Ten grams of Laminac coated fireclay were used as an inert charge at thebottom of the Kraft paper cased flares. The interior walls of the flarecases were coated with the materials indicated in the tables.

4k 11

I •, q• -: .. .. . . .. . ... .. i

Testing

Light measurements of the flares were determined in accordancewith MIL-P-20464 except that a black background was used instead of awhite one.

LIST OF MATERIALS USED

Aluminum powder, atomized, Lot No. 1072, average particle diameter6.2 microns, Aluminum Company of Canada

Aluminum powder, atomized, Lot No. 1401, average particle d;ameter8.0 microns, Aluminum Company of America

Sodium nitrate, double refined, USP grade, average particle diameter35 microns, Davies Nitrate Company

Tungsten powder, Lot No. 6154D, average part le diameter 6.7 microns,General Electric Compan',

Magnesium powder, atorrized, 20/50 mesh, Specification MIL-P-14067B,Hart Metals Incorporated

Ferric oxide, red, unknown grade, lot, company

Ferric oxide, jeweler's rouge, catalogue No. I-115, average particlediameter 0.6 microns, Fisher Scientific Company

Ferric oxide, reagent grade, catalogue No. 1-116, average particlediameter 0.5 microns, Fisher Scientific Company

Chromium sesquioxide, Lot 32194, average particle diameter 2 microns,J .T. Baker Chemical Company

Sodium fluoride, S-299, average particle diameter 8.6 microns, FisherScientific Company

12

Superfloss (SiO2 ), Johns-Manville Company

('ab-()-Sil (S;O), MS-7, Cabot Corporation

Magnesium, atomized, 30/50 mesh, Specification MIL-P-l4067B3, HartMetals Incorporated

Laminac polyester resin 4116, American Cyanamid Company

I ~ Vinyl alcohol acetate resin, MA--28-18, Palmer Cement Company

Viton A fluorocarbon resin, Dupont Cheu.':,":a.l Company

Asbestos powder, technical grade, Lot N(H A1971, Fisher ScientificCompany

It

13

REFERENCES

1. Shidlovsky, A., Fundamentals of Pyrotechnics, (OsnovyPiroteckhniki), Picatinny Arsenal Technical Memorandum1615, May 1965

2. Jackson, B., Kaye, S.M., and Taylor, F R. ,Aluminum Powderas an Alternate Fuel for Magnesium in Illuminant Compositions,Picatinny Arsenal Technical Report 3713, November 1968

3. Leader, P.J., Westerdahl, R.R., and Taylor, F.R., The Effectsof Some Transition Metal Compounds on the Performance Character-isticc of Aluminum/Sodium Nitrate Compositions, Picatinny Arsenal

Technical Report 3846, May 1969

4. Johansson, S.R. and Persson, K.G., Explosion Hazards ofPyrotechnic Aluminum Compositions, Pyroteknikdogen, Stockholm,

10 May 1971, pp 74-90

14

V ,

LM cc C4 LM 4Tm

LM In

0 nn co4 C4 - OO

Urn n

#A In - r-4r

4 n M ~

QJLn In

6 Lnn

.C

b II'10 I

IICL

0 1. u

U)U#o

SUC

E E

m to 00 C ~

E + + ++

M~b InO C4 -r C

C4~

+

OA Crr 4 f- -omC

.0

(U~ E4 0

0 dP dP 0Z IJI I+

m

In 2 M in

m++ #++ In+ +

16

1W~C _ .,

- rn m1"

mm~o

z.L< L.co 4 L

M >

00

CN (

LiA

-- 0C--

'.0M u

.~ c

a ~m :33:

17

C ~ N r4 C

::0. MULAZTc

M~ I..n C

0.

1..

EUl

U Ul 0j c S..

E r u

CE$5A

0 L0

2 tV

LM .

*1 E~ 06

E mLC LL E.

CO CO_

owl",

44 Fm0 4Ln %

4), L nC 4L n% nr(CU m 41 IfLrIU C14~r

*~ ~ 1 u a. mi

441

> 0

E1 u E20 r.4IwS4- E cUAD 0

* - E LL -I

U IA~

*~'0 ~c0 4 C.

'oC 0V .. CL

*-c L..c c

19

1. CoC A LAC .* n mfr.Ll Ll O

to u

(ftf

4)4L

En

Cr0L

EE 13

N c~r 4 In.*-- CA o

CE ~'C

200

In

0 a).-

L.t'1aO~f

"T tD 0.

U c4c

E -- >V

4L -1C -

-~~d c c ,.

S0

C4c C

0n L 0 'o~

LnL0~

toL. 0 r4

Wu.0

LL

22

L.n I I I

Go r- LnCýU

IA

4L ID If.

v 0 C4 0 .L A

4. E

0 W~

L.041E

~~0

E IA

c E0 = E

4'r to'5 4

E p~ 0 ,w-

8. c u M

~Li rIu

MAQ I I~- 1 0; 0

en~ I I LM-4 -

cm .

.0M

'..

U)U

EcEu2

4I cE C4 0

'n Lno

-C rýIflCD

U~uSU

4)24

ElL-

EIV

i. E.

0 tuV

ul UULf IA 0 WA

> 25

fj -

FAL

i Lo I ~~m ~fl n 42LMt

u0

0u

-0'

U. dP dP

E ~E ' -E uv

IV r-d

Lan ;;c

26

j~~~~~ i _________

O*O Ln~

U ~ ~ Or4Ov

U C4 i6I A 04) .n I l-L

E3E

EOC L.LM ~

27

C4 0 -

U L n

UU

Cn r4 W-r

I.m

c LC%

LMM - M.. ~ n

L 0'

ul~ 0

-E

ti EU,

.P,8s a

EvC 0

X cx I2 2 t 28

do

co LM c

'o CfI- A n n

E V

IM,

LI.M

29

.0 0)C

0 .0

00

( U0.~F' 1> 2LL

.4 4

Cc

0 0 E -.UA

30

-?

SM ":> •

r g-g

in UI CON Ln

41

0 o

u 0c E

2- O..'-

W U % m

U U N m--•E• ' E . UL

"4- c' v• C,-M m1 4.;••Sm m 4,n

4.'.-*-to m :0

"" E L E l Ui ri4

V) Jc

31

EE'.o:2 -

E

Sy m

~.12

M V4

Ln

LOL

Ln CD

o to

32

Ii7177-

CL

0

0)

*0

-00

10

~I -%..I

-J x

LA X;mi w .C 43; ý = CLM~

CD LC

VWI3HIOX3*!a*WH3HOGN3

33

040E2E

QE

L4 )0Ln~

z C4

E4)

xý<

u 0

SL

34 cL-

In

z

zz<0

0 0

a

C.)Mf

-o (a)

is Ls

*q

0 0

Ln)

-77-

x

LC

LU

04-

0.

C'I'

LmI~±x Ci 4bH10CLf

611 cj~~LA

DISTRIBUTION LIST

Copy No.

CommanderPicatinny Arsenal

ATTN: SARPA-TS-S 1-5Mr. S. Fleischnick, ADED 6-9Mr. H. D. Rutkovsky, ADED 10-12Mr.* 0. Katz, ADED 13-15Mr. E. Belitsky, ADED 16-18

SMr. J. Matura, ADED 19-21Dover, New Jersey 07801

CommanderU.S. Army Materiel CommandATTN: Scientific and Technical Info Div 22

AMCRD-Q 23!AMCRD-RS 24

5001 Eisenhower AveAlexandria, Virginia 22364

CommanderI U.S. Army Armament CommandATTN: Scientific and Technical Info Div 25

AMSMU-RE 26Rock Isiand, Illinois 61201

CommanderAberdeen Proving GroundATTN: STEAP-TL (Bldg 305) 27Aberdeen Proving Ground, Maryland 21005

DirectorU.S. Army Ballistic Research LaboratoriesATTN: Scientific and Technical Info Div 28

Dr. J. Richard Ward 29Aberdeen Proving Ground, Maryland 21005

37

77j

CommanderHarry Diamond LaboratoriesATTN: Library, Room 211, Bldg 92 30Connecticut Ave and VanNess St, NWWashington, D.C. 20438

CommanderU.S. Army Mobility Equipment R&D CenterATTN: Scientific and Technical Info Div 31Fort Beivoir, Virginia 22060

U.S. Army Foreign Science and Technology CenterATTN: AMXST-ISI 32220 Seventh St, NECharlottesville, Virginia 22901

Institute for Defense AnalysesATTN: Library, Documents 33400 Army-Navy DriveArlington, Virginia 22202

U.S. Army Research OfficeBox CM, Duke StationInformation Processing Office 34Durham, North Carolina 27706

Defense Documentation Center 35-46Cameron StationAlexandria, Virginia 22314

Central Intelligence AgencyATTN: CRS/ADD/Standard Distribution 47Washington, D.C. 20505

CommanderFrankford ArsenalATTN: Scientific and Technical Info DIv 48

Pitman-Dunn Laboratory 49Philadelphia, Pennsylvania 19137

38

"'.lI

CommanderU.S. Naval Ammunition Depot

*ATTN: Scientific and Technical Info Div 50Chemical Sciences Division 51

Crane, Indiana 47522

CommanderNaval Weaipons Center (Code 533)ATTN: Scientific and Technical Info Div 52China Lake, California 93555

Naval Ordnance Systems CommandDepartment of the NavyATTN: Code ORD-0632 53Washington, D.C. 20360

CommanderU.S. Naval Ordnance LaboratoryATTN: Scientific and Technical Info Div 54Silver Spring, Maryland 20910

CommanderAU.S. Naval Ordnance Station

OL ATTN: Scientific and Technical Info Div 55Indian Head, Maryland 20640

CommanderNaval Air Systems CommandDept of the NavyATTN: Code AIR-53221 56Washington, D.C. 20360

CommanderU.S. Air ForceATTN: Scientific and Technical Info Div 57

AFATL/ DLI P 58Eglin Air Force Base, Florida 32541

* 39

USAF/DCSResearch and DevelopmentPentagon 59Washington, D.C. 20330

Foreign Technology Division (DPXA)Wright-Patterson AFB, Ohio 45433 60

CommanderLowry Technical Training Center (TTOX) 61Lowry AFB, Colorado 80230

NASAScientific and Technical Info Div 62P.O. Box 33College Park, Maryland 20740

Denver Research InstituteUniversity of DenverATTN: Mr. Robert Blunt 63Denver Colorado 80210

I

Battelle Memorial InstituteColumbus LaboratoriesTACTEC 64505 King AveColumbus, Ohio 43201

40I

,, 4I•:i