Numerical and experimental study on the response characteristics of warhead in the fast cook-off process

Min Zhu,Sheng-ao Wang,Huang Huang,Gui Huang,Fei Wu,Shao-hua Sun,Biao Li,Zi-jian Xu

College of Nuclear Science and Technology,Naval University of Engineering,Wuhan,430033,China

Keywords: Warhead Explosives Fast cook-off Numerical calculations

ABSTRACT The response characteristics of the warhead under thermal stimuli conditions are important to the safety improvement.The goal of this study is to obtain data on the warhead in the fast cook-off process.In this paper,a numerical calculation method is proposed,whose reliability is supported by comparison with experimental results.Through the numerical calculation,the temperature distribution,temperature change,and ignition time are acquired.The numerical results show that the ignition time is 76 s after the warhead started to burn and that the maximum temperature of the explosive’s outer surface is 238.3°C at the ignition time.The fast cook-off experiment of the warhead is implemented so as to get the flame temperature and reaction grades that are not available through numerical calculation.The experimental results show that the overpressure fails to reach the preset minimum value which is equivalent to 6 kg of TNT and that the reaction grade is de flagration.The research results have reference value for the design of the warhead and the reduction of detonation risks.

1.Introduction

During the transportation and storage of ammunition,the abnormal thermal stimulation such as fire may cause serious accidents such as ignition and explosion.Therefore,it is critical to study the response characteristics of explosives under the cook-off condition.And the research results are of great importance to improve the thermal stability of the warhead.

Experimental and numerical calculation is the primary method to study the phenomenon of explosive cook-off.In the experimental aspect,Parker et al.firstly designed the small-scale cook-off experiment(SCB)[1].Scholtes et al.also established a small-scale experimental system that is similar to the SCB[2]and applied the device that can control the heating rate.Dai et al.studied the fast cook-off safety characteristics of large-scale penetrating projectiles through experiments[3].Liu et al.designed explosives based on CL-20/FOX-7 and tested their cook-off characteristics[4].The domestic explosive cook-off experiment is conducted based on the SCB method.Some scholars have likewise carried out explosive cook-off experiments[5-7].The cook-off experiment can directly and effectively evaluate the thermal stability of the explosive,but it has problems such as high cost,long cycle,fewer data.

The numerical calculation method can perform reliable numerical calculations on various characteristic parameters of explosives such as blast wave[8],detonation wavefront[9].This method has been recognized and used by most scholars in the field.The numerical calculation method can change the heating rate,the explosive size,the shell thickness,and so on.It is widely valued due to the ability concerning the comprehensive prediction of ignition characteristics.In 1991,Jones et al.firstly used HEAT software to simulate the small-scale cook-off experiment[10],but only based on the 1Dmodel.Chidester et al.improved the calculation model of the explosive cook-off process,but the model is relatively simple[11,12].Yoh et al.utilized the chemical kinetic model of multi-step reaction to simulate the cook-off process of LX-10 explosive[13,14]and analyzed the reaction intensity through the shell deformation and rupture condition.Zeng Jia et al.established a cook-off calculation model for explosives and simulated the roasting response at a heating rate of 0.055,0.5,1.0,2.0,3.0,5.0 K/min[15].Niu et al.used LS-DYNA to establish a simulation model to predict the ignition point and ignition time of PBX explosive in strong constrain condition[16].Liu et al.established a 3Dslow cook-off test model of DNAN-based RDX melt-cast mixed explosives and performed a numerical simulation at a heating rate of 0.055°C/min[17].Jing et al.carried out simulation calculations of explosive cook-off process domestically[18,19],and determined the ignition time and ignition temperature.Niu et al.established a dynamite cook-off thermal reaction model and compared the numerical calculation results with experimental results[20].Most research focuses on the response characteristics of explosives in the cook-off condition.However,in various application scenarios,explosives are loaded into warheads.Therefore,not only does this paper propose a credible numerical calculation method,but also it conducts a cookoff experiment for a certain type of warhead.Finally,we obtained the response characteristics of explosives that are closer to actual use environment.

Fig.1.Finite element model of the experimental device.

The numerical calculation in this paper is based on the international general finite element software.A numerical calculation method for the explosive cook-off process was established.Its reliability was veri fied by comparison with the experiment results.Based on the numerical calculation method of the explosive cookoff process,a 3D model of the cook-off process was established for the typical warhead,and the reaction time under the fast cookoff condition was calculated.Besides,the response characteristics of the warhead under fast cook-off conditions were studied by the full-scale experiment.The numerical simulation results can provide technical support for the design of the subsequent fast cook-off experiment.In addition to enhance the understanding of warhead safety,the full-scale experimental results can also accumulate data for establishing full-scale experiment evaluation methods.The experimental results can also provide a basis for subsequent research on the technology,which can improve the safety of the warhead.

2.Method

2.1.Numerical calculation method of explosive cook-off process

2.1.1.Calculation model

The following assumptions are made once establishing the calculation model of the explosive cook-off process.

1)The explosives are homogeneous solid,without considering the phase transformation.

2)The explosives and the shell are isotropic,and its physical and chemical parameters are constants that do not vary with temperature.

3)There is no gap between the shell and explosives,without considering the in fluence of gas products on heat transfer.

Based on the above assumptions,the basic formula of the explosive cook-off process in the Cartesian coordinate system is:

Where:ρis the explosive density(kg/m3).Cis the speci fic heat capacity(J/kg·K).Tis the temperature(K).tis the time(s).λis thermal conductivity(w/m·K).Sis the heat source term of explosive self-exothermic reaction,and can be expressed by Frank-Kamenetskii reaction model of Arrhenius equation.

Where:Qis the reaction heat(J/kg).Zis the pre-exponential factor(s-1).Eis the activation energy(J/mol).Ris the universal gas constant(J/mol·K),generally 8.314.

2.1.2.Finite element model of the numerical calculation method

Finite element software ABAQUS was used to establish the 3D calculation model of explosive cook-off experiment.The heat source term of explosive self-exothermic reaction S was loaded into the main program by the user subroutine HETVAL,which was written by FORTRAN language[5-7].

A quarter model was set up according to the structure and boundary symmetry of the experimental device.In the calculation,the outer wall of the warhead’s shell was the heating boundary.The surface between the shell and the explosive was the coupled heat conduction interface.Thus,the temperature and heat flux of the contact surface between the shell and the explosive were continuous.End cover,shell,and explosive were simulated by 816,924 and 2548 hexahedral elements,respectively.Fig.1 shows the finite element model.

The shell and the end cover of the cook-off bomb are made of 45#steel,filled with typical PBX(s)explosive.Table 1 shows the thermophysical parameters of the material.

Table 1 Thermophysical parameters of materials in the reliability testing process.

2.1.3.Simulation of explosive self-exothermic reaction

The user subroutine HETVAL was used to de fine the source term of the explosive self-exothermic reaction.The user subroutine HETVAL was produced by FORTRAN language to de fine the source term of explosive self-exothermic reaction in the cook-off process.The subroutine HETVAL calculates and continuously updates the temperature of explosives.When the temperature suddenly rises rapidly,and the temperature gradient is in finity,this program terminates.At this time,the explosive reaches a state of thermal runaway,which is the ignition of explosives.Fig.2 shows the reaction process of explosive self-exothermic reaction simulated by subroutine HETVAL.

2.1.4.Reliability test of the numerical calculation method

Fig.2.Flow chart of calling subroutine HETVAL.

Fig.3 shows the experimental system includes the test bomb,furnace,and thermocouple.

Fig.3.Schematic diagram of heater and cook-off sample.

The size of the roasted bombshell isΦ46 mm×56 mm.The thickness of the wall is 3 mm.Both ends of the tested bomb were sealed with threaded end caps.

Temperature rise starts at room temperature of 22°C(295K),at a constant rate of 1K/min until the tested bomb reacted.The temperature of the pellet center(point A)and the outer wall of the shell(point C)were measured.

2.2.The fast cook-off model of the warhead

Based on the credibility veri fication of the numerical calculation method for the fast cook-off process,the study on response characteristics of the warhead in the fast cook-off process was undertaken.Fig.4 shows the quarter model established according to the symmetry of the warhead’s structure and boundary,a.In the calculation,the shell and the outer wall of the simulated cabin were heating boundaries.The interfaces between the shell and the explosive,the shell and the liner,the shell and the simulated cabin,and the shell and the e were coupled heat conduction interfaces.Thus,the temperature and heat flux between the interfaces were continuous.

Fig.4.FEM model of the cook-off experiment.

The material of the warhead’s shell is a high-strength cast aluminium alloy(ZA1Si7MgY).The material of the liner is copper,and the material of the network plate,fuse,and simulated cabin are aluminium alloy 2A12.The principal charge of the warhead is fused cast explosive RHT-906.The parameters are given in Table 2.

Table 2 Thermophysical parameters of materials in the fast cook-off process.

2.3.Experimental method for the fast cook-off process of the warhead

Fig.5 shows that a fixed test bomb and a fixed bracket are applied in the fast cook-off experiment,and the fixed bracket does not affect the heating of the test bomb.During the experiment,the test bomb was positioned in the fire.The warhead’s reaction was registered over time.The experimental equipment mainly includes a test bomb,combustion pool,fuel oil,bracket,camera,photography system,temperature measurement system,and overpressure measurement system.The combustion pool should be large enough,and the fuel should be enough.The combustion flame should completely cover the test bomb.The layout of the experiment shall be implemented in accordance with NATO Standardization Agreement STANAG 4240 and Appendix A.

Fragment information(fragment size,distribution,etc.)generated during the test bomb’s reaction is collected by the witness board.The time history of the flame temperature is registered by the thermocouple.The overpressure value of shock waves after the reaction is monitored by the measurement system.The entire process is recorded by high-speed photography.Finally,the conditions of the test bomb and the scenes before and after the experiment are recorded by static photography.

Fig.5.Schematic diagram of the fast cook-off experiment.

Fig.6 displays the layout of the temperature sensors and overpressure sensors around the projectile.Four thermocouple sensors were arranged around the warhead to measure the flame temperature.To obtain the shock wave overpressure,four overpressure sensors(PCB-137)were arranged around the warhead.The fast cook-off experiment cannot determine the reaction time and reaction grade.It can only be passively triggered.Therefore,according to the charge equivalent of the warhead,overpressure thresholds were set into four different TNT equivalents of 10%,20%,50%,and 75%.

Fig.6.Layout of measurement point of temperature and overpressure sensors.

Fig.7.Temperature change curve at measurement point A.

A pit(5 m×4 m×2 m)was dug in the experimental site.The combustion pool was positioned in the pit.The warhead bracket was welded in the combustion pool.The warhead was set on the bracket with wire to prevent the warhead from moving.The distance from the oil level to the bottom of the warhead was 0.7 m.Two protective plates were arranged along the two sides of the warhead’s course to reduce the natural wind’s effect on the flame’s stability.At the same time,it can also intercept the fragments of the warhead and scattered fuel. A veri fication board(1 m×1 m×0.005 m)was arranged at the warhead’s front end to measure the power of the liner after the explosion.

A video surveillance system was established with two highde finition cameras to observe the fast cook-off process of the warhead.The two cameras were set to the warhead’s both ends.One was given to the protective mound,which was 12 m from the center of the warhead.The other was located in the rear end,which was 11.7 m from the center of the warhead.The image transmission distance of the camera was not less than 150 m.Besides,a highspeed camera was arranged to record the scene of the instant reaction of the test bomb.During the experiment,experimenters observed and recorded the response of the warhead shell during flame torch on a computer.Throughout the fast cook-off experiment,video images were an essential basis for analyzing potential safety hazards in this experiment.

3.Results and discussions

3.1.The credibility of the numerical calculation method

Fig.7 shows the comparison of numerical calculation results with experimental results and literature values on measurement point A.The comparison of ignition time,the temperature of point C and A is listed in Table 3.According to Table 3,there is a certain deviation between the numerical calculation results and the experimental results over a period of time after heating.This is caused by the neglect of the explosive’s heterogeneity and phase transition in the numerical calculation method.With the increase in temperature of explosives,the in fluence of phase transition on the thermal conduction is negligible.Subsequently,the numerical calculation results agree well with experimental results and literature.According to the research of Niu Yulei and others,there is only a small part of the phase transition in an authentic cook-off process[20].The ignition time and temperature of the explosive are concerned by the engineering veri fication.The ignition time and the temperature of characteristic points obtained by the numerical calculation method are consistent with the experimental results.Thus,it shows that reasonable results can be achieved by the numerical calculation method proposed in this paper.

Table 3 Comparison of numerical calculation results with experimental results and literature values.

3.2.Numerical calculation for the fast cook-off process of warhead

3.2.1.Initial and boundary conditions

The ambient temperature was set to 27°C.The flame temperature around the warhead’s shell and the outer wall of the simulated cabin reached 800°C within 1 min.Then the flame temperature was maintained at 800°C.Fig.8 shows the loading curve of the flame temperature.The subroutine HETVAL also realized the heat source term of the explosive self-exothermic reaction during the fast cook-off process.

3.2.2.Temperature distribution and ignition time

Fig.8.Loading T-t curve of the flame temperature.

Fig.9 shows the temperature distribution of the warhead at the ignition time.Fig.9 indicates that with the exception of explosives that were in contact with the warhead shell,the temperature of most explosives at the moment of ignition is low.Therefore,less heat was moved to the interior of the charge during the fire.The ignition position is on the outer surface of the charge.

Fig.9.Temperature distribution at ignition time.

Fig.10.T-t curve of ignition position obtained by numerical calculation.

At the time of ignition,the temperature of the shell’s outer surface is 238.7°C.The maximum temperature of the explosive’s outer surface is 238.3°C.The temperature difference between the two is minimal.This phenomenon can be ascribed to two reasons.First,the material of the shell is an aluminium alloy,which has good thermal conductivity.Besides,in the numerical calculation,it is assumed that the temperature of the outer surface of the explosive and the inner surface of the shell are continuous,indicating that the temperature gradient at the interface between the shell and explosive is neglected.

When it comes to the judgement whether the explosive is ignited,the self-heating reaction calori fic value of the explosive is calculated according to the temperature of the explosive,using the Frank-Kamenetskii reaction model.The temperature of the explosive is calculated and continuously updated by the subroutine HETVAL.When the temperature gradient rises to in finity,the subroutine HETVAL terminates.It is determined that the explosive has reached a thermal runaway state,that is,the explosive has been ignited.

According to Fig.10,the thermal runaway occurred at 76 s after the warhead started to burn,indicating short ignition time.This phenomenon can be ascribed to two reasons.1)In the numerical simulation,the phase transition of the explosive is not taken into account.In the actual process,due to the melting of TNT in the charge(melting point:80.9°C),the explosive has a sharp endothermic process.2)In numerical calculation,whether the explosive is ignited is determined by the Frank-Kamenetskii reaction model,and the values of related activation energy,pre-exponential factor,and reaction heat have apparent effects on ignition time.

3.3.Results and analysis for the fast cook-off experiment of warhead

3.3.1.Flame temperature

The change of temperature around the warhead was recorded by temperature sensors.Fig.11 shows the change curves of temperature around the warhead were recorded by temperature sensors.The highest flame temperatures measured by the four sensors before the warhead’s reaction time(300 s)are 587°C,711°C,819°C,and 620°C,respectively.The 1#and 4#sensors are located at the front and left of the warhead.After the warhead’s violent reaction,a high-temperature signal appears,which coincides with the intense combustion of scattered explosives at this position.

Fig.11.Change curves of temperature around the warhead.

According to the comparison between Figs.8 and 11,There are differences between the numerical calculation values and the fast cook-off experiment results.In the numerical calculation,the outer surface of the warhead shell and the simulated cabin are heated uniformly.In the fast cook-off experiment,there also exists a heating process.However,due to the size of the projectile and the wind,there is not any ideal temperature stabilization stage,and the flame temperature fluctuates greatly.

Fig.12.Relation of projection kinetic energy and reaction grade.

Based on the differences between the numerical calculation and the actual experimental process in the boundary conditions and the supposed conditions,the numerical calculation results are more conservative than experimental results.Therefore,the numerical calculation method can not only be used to estimate the ignition time but also provide a reference for the fuel quantity and other parameters on the design of the experiment.

3.3.2.Reaction grades

There are no signals from the four overpressure sensors,indicating that the overpressure after the warhead reaction did not reach the preset minimum value,which is has the same power as 6 kg of TNT.According to the size and number of fragments formed by the test bomb and the no signal state of the sensors,it can be determined that the test bomb’s reaction grade does not reach the detonation grade.To determine the reaction grade of the test warhead as accurately as possible,the experiment refers to the MILSTD-2105D standard[21].Compared to the explosion and combustion reactions,the reaction of the test bomb matches the description of the de flagration reaction grade better.The detailed standard bases are as follows:

1)Most explosives were burned after being thrown,and only a small amount of unreacted explosive debris was recovered,satisfying the requirement that partial or full explosive has burned.

2)After the reaction of the warhead,the structure of the liner,network plate,and fuse were complete.The shell at both ends of the warhead ruptured into 3-5 larger fragments,which meet the requirement that the shell ruptures and several large fragments(including accessories and connectors)are formed.

3)According to the calculation basis of fragment kinetic energy in MIL-STD-2105D,when the fragment flight distanceDand fragment mass m satisfyD=4.212×m(-1.103-0.0788×ln(m)),the corresponding kinetic energy is 20 J.Fig.12 shows the relationship between fragment kinetic energy and reaction grade.In the experiment,the weight of simulated cabin at the rear end of warhead is 2.03 kg.After hitting the support rod of the video surveillance at the rear end of the warhead,the flight distance reached 11.7 m,and its kinetic energy was far higher than 20 J,as shown in Fig.12.The kinetic energy of the simulated cabin section is considered far higher than 20 J.The situation of hitting the video surveillance support rod is also considered.It is assumed that this experiment satis fies the requirement that,if there is no evident criterion that a more severe reaction has occurred,but there is evidence that the fragment can be pushed out more than 15 m,this reaction is also determined as de flagration.

In summary,the reaction grade of the warhead is de flagration.

4.Conclusions

(1)A numerical method to simulate the fast cook off process of warhead is proposed in this paper.The credibility of this method is con firmed by the comparison with the literature values.

(2)Through the numerical calculation for the fast cook-off process,the temperature distribution and ignition time of the warhead are obtained.The thermal runaway occurs at 76 s after the warhead started to burn,and the maximum temperature of the explosive’s outer surface is 238.3°C at the ignition time.

(3)The layout of the experimental site is designed in this paper.According to the experimental results,the overpressure value measured in the fast cook-off process is lower than that of the explosion of 6 kg TNT,and the reaction grade is determined as de flagration referring to MIL-STD-2105D.

(4)The numerical and experimental results demonstrate that the design of the warhead does not meet the ammunition safety standards.Therefore,it is necessary to improve safety technology.The combination of numerical calculation and experiment proposed in this paper can be employed to warhead safety analysis.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to in fluence the work reported in this paper.

Acknowledgements

Thanks to the equipment pre-research fund(41426030107)for supporting the completion of this project.

Appendix

SPECIFICATION FOR STANDARD LIQUID FUEL/EXTERNAL FIRE TEST

1.Guidance for the use of this Annex is given in paragraphs 7-19 of the STANAG.

2.Hearth Requirements.The hearth shall be large enough to allow at least 1m clearance on each side of the test item,and designed to provide a volume of flame which completely engulfs the test item throughout the trial.

3.Hearth.Liquid fuel is normally contained within a speci fically prepared,polyethylene-lined,pan or hearth.The simplest design of pan or hearth can consist of a shallow,level pit,or a leveled area surrounded by a wall or embankment.The depth of the pit,or height of the wall or embankment,shall be sufficient to contain the required amounts of fuel and water as determined from paragraphs 5 and 6 of this Annex.The depth of the pit shall not block fragment projection.

4.Fuel.Suitable liquid hydrocarbon fuels are:JP-4,JP-5,Jet A-1,AVCAT(NATO F-34 or F-44),or commercial kerosene(Class C2/NATO F-58).

5.Fuel Quantity.The quantity of fuel should be sufficient to maintain a fully developed fire for the speci fied period,which shall be 150%of the estimated time to reaction.Water(from a low pressure hose)may be added,as required,to raise the fuel level to the correct distance below the test item,but the quantity of fuel over the water must be greater than 15 mm at all times during the test to prevent boiling of the water due to radiation from the fire.

6.Calculation.As a general guide in calculating quantities of fuel,the rate of fuel surface regression due to combustion for all the required fuels and all sizes of hearth can be taken as 7mm/minute.

7.Position and Mounting of the Test Item.Unless otherwise speci fied by the Design or Test Authority,the test item shall be centred within the hearth area with its major axis horizontal.The lower surface of the test item should be high enough above the initial fuel surface to

a.allow full combustion bellow the test item;

b.not unduly increase the chance of occasional emergence of the test item from the flame envelope.

In order to ensure the test item is not positioned in a cooler,fuelrich area of the flame,the position shall satisfy the temperature requirements of paragraph 12 of this Annex.As a guide,the initial height of the bottom of the test item above the fuel surface shall be no less than 0.3 meters at the start of the test.Note:For Hazard Classi fication the test item must be tested in its logistical con figuration.

8.Suspension and Support of Store.The methods used to position and hold the test itemwithin the fire hearth could have a very signi ficant effect on its response to the fire environment.Unless otherwise speci fied,the test item shall be realistically supported such that when positioning as in paragraph 7 of this Annex,any sagging would represent that which would occur in an actual incident.

9.Support Stands.Any additional support stands or props should make only minimum line contact with the test item and must not screen it from the enveloping fire.The number of such extra supporting points must be kept to a minimum and should,where practical,be con fined to positions where the casing of the test item is thickest.

10.Constraints.for test items that may become propulsive and compromise range safety,consider adequate means of constraint.Where necessary the restraining device should not unduly screen the test item from incoming heat radiation from the flames.The Design or Test Authority must approve details of any such restraining attachments.

11.Support Tray.If required,a perforated metal tray or grid may be arranged below the test item and extending sufficiently(1 m)on all sides,so that if the test item collapses,or its contents fall out.such items will be held to remain partially exposed to the fire.The design,construction and position of such a tray is at the discretion of the Design or Test Authority,but must be adequate to support the weight and impact of falling items.Preferably,the positioning of such a tray or grid prior to initiation of the fire should be about 50mm below the fuel surface so that it retains its strength and does not affect the combustion of the fuel.

12.Test(Flame)Requirements.An average flame temperature of at least 800°C,as measured by all valid thermocouples(sample rate>0.2 Hz)at the test-item without contribution of the burning munition,will be considered a valid test.This temperature is determined by averaging the temperature from the time the flame reaches 550°C until all munition reactions are completed.Any deviation from this shall be recorded with appropriate time versus temperature date.The flame temperature shell reach 550°C in the order of 30 seconds after ignition as measured by any two of four flame thermocouples.The time(over a 30-second period)until flame temperature,as measured by the two thermocouples,reaches 550°C shall be subtracted from the time of reaction.

13.Ignition.To ensure a rapid,consistent build-up of flame area,the fuel should be ignited at the mid-point of one or more of the sides of the hearth by means of suitable remotely operated flame-producing devices fired simultaneously.

14.Flame Spread Rate.To enhance the rate of spread of the flame area,particularly in conditions of low ambient temperature,20-30 litres of petrol(gasoline)are to be floated on the fuel at each of the ignition points.The time delay between placing of the petrol(gasoline)and ignition of the fire should be kept to a minimum to avoid excessive loss by evaporation and dispersion in the fuel.Initiation of the flame producing devices should ensure the best"all fire"probability.Suitable methods of remote ignition are:

a.A flame producing system that has been demonstrated to be effective and reliable consists of an electric igniter inserted into a small bagged power charge.

b.A bundle of cotton waste(above 1-2 kg)is placed in the fuel at the midpoint of each side of the hearth.The petrol(gasoline)is poured over the waste,and the charge/igniter unit is place on the top of the bundles.The charges will still function even when completely soaked in petrol(gasoline)or kerosene,however do not allow rain to saturate the bundles.

15.Test Environment.Fire tests should not be conducted in the rain(which will cause heating problems),or when wind velocities in the test area(or inside wind barriers if such barriers are used)exceed 10 km/h,because this condition prevents full engulfment of the test item even when a wind barrier is used.

16.Thermal Insulation of Support and Restraining Rigs.Even substantial components of the support and restraining will lose a major portion of their strength within a few minutes of the full fire developing(components of wall thickness of 6 mm could reach 700°C within 2 min).To avoid unduly massive constructions,some form of thermal insulation should be applied to the structural members.A suitable material is a mineral wood fibre(density 80-100 kg/m3).This is obtainable in the form of preformed sections and slabs of 25 mm thickness,which can readily be shaped as required.Glass fibre is not suitable as it melts at the temperature obtained within the fire.Where lagging is used,it must be kept dry until the last practical moment.

17.Instrumentation.To provide a consistent,remote indication of the full development of the fire,a minimum of 4 thermocouple elements are required.These thermocouples shall be mounted 40-60 mm from the surface of the test item at positions fore,aft,starboard and port along a horizontal plane through the test-item.Data must be recorded every five-seconds or less(0.2 Hz).Additional thermocouples may be positioned at the discretion of the Trial Authority.

18.Thermocouples.Type K thermocouples(nickel-chromium/nickel-aluminum conductors),sheathed in inert hermetically sealed insulation and capable of withstanding 1200°C,have been successfully used to measure test temperatures.

19.Timing.The zero for the timing of hazardous events is the instant that any two of the temperature sensors reach 550°C.(This is an indication the test item has been enveloped by the fire.)Veri fication of the time and degree of envelopment can be obtained from a combination of cine-film or video,and a timing device.

20.The most satisfactory position for a camera/video is upwind of the fire on a tower.

21.Observations and Records.Unless noted as"optional"for IM and/or HC determination,the following minimum observations are to be made and records kept:

a.Test item identi fication(model,serial numbers,number of test items,etc.);

b.A record of events versus time,from ignition of the fuel to the end of the trial;

c.The nature of any reactions by the test item;

d.The nature and distribution of residue and debris;

e.Wind velocities and direction inside and outside the enclosure before the trial,and any signi ficant change in velocity/direction outside the enclosure(preferably well clear of the enclosure)during the trial;

f.Type of fuel for the test;

g.Thermocouple identi fication and locations;

h.Listing of environmental preconditioning test performed;

i.Type of energetic material and weight;

j.The geographic orientation of the test item’s longitudinal axis;

k.Thermocouple data(versus time)for all sensors;

l.Indication of propulsion(video or other suitable means);

m.The temperatures at the recording points need to be continuously recorded.For the purpose of the test,simple direct-indicating meters are sufficient;

n.A microphone or other suitable listening device should be placed near the trial site to record audible events.The audio record shall be a sound track on the motion picture film,or on the videotape to enable correlation with visible events and indicated times;

o.Suitable blast or pressure gauges should be positioned around the test item;

p.Thermal flux measurements;(Optional for all but candidate HD 1.3 and 1.4 articles.)

q.Fragment recovery and mapping;

s.Witness screens as a measure of severity(optional).

22.The following photographic records and videos are to be made:

a.Still photographs of the test item before and after each trial;

b.Still photographs of any other residues arising as a result of the trial;

c.Colour cine-film or video for the duration of each trial.