Committee of Experts on the Transport of Dangerous Goods and on the Globally Harmonized System of Classification and Labelling of Chemicals

Sub-Committee of Experts on the Transport of Dangerous Goods

Forty-first session

Geneva, 25 June – 4 July 2012
Item 2 (d) of the provisional agenda

Explosives and related matters: DDT Test and criteria for flash composition

Proposed alternate flash composition test for fireworks classification using the default table

Transmitted by the expert from the United States of America

Background

1.  At the 37^th session, the expert from the United States of Americafirst presented an alternative test method (the “Deflagration to Detonation Transition” [DDT] Flash Composition Test) to the current HSL Flash Composition Test method for evaluating pyrotechnic mixtures (see ST/SG/AC.10/C.3/2010/31) supported by test data from ten different materials. At the 39^th session, the expert from the United Statesprovided a revised drawing of the DDT Flash Composition Test Fixture and test results for twelve additional materials using the methodology (see informal document INF.44, 39^th session). Other informal papers were also presented by the expert from Japan (INF.22) and the expert from Germany (INF.16).  All of the testing done indicated general agreement with the results obtained by the expert from the United States of America. 

2.  Since the proposed method is easier to perform and utilizes larger samples, it was considered by the Working Group on Explosives to be an attractive alternative test. There was group consensus that the DDT test proposed by the United States of Americawas a good way forward (see the report of the Working Group on Explosives in informal document INF.58, 39^th session). A number of comments were received that the United Stateshas worked to address in this revised proposal. For example, Germanypointed out some safety issues related to the size of the mortar that could be encountered in performing the test.  Germanyalso cautioned that the mass of the mortar could be an influencing factor on the outcome of the test and offered to investigate further.  Japanand the United Kingdomobserved that their work indicates that the degree of granularity of composition can affect results, and consideration should be given to expanding the method to include samples with granular material.  They agreed that the weight of the tube could be a safety or test outcome-influencing factor and support Germany’s further research.  Other experts such as the Netherlandsand Australiaalso expressed the opinion that the sample mass could influence the outcome of the test and recommended that this potential effect should be studied further.  The United Kingdomobserved that the test was really straightforward to perform and supported its continued development.  The Netherlandsobserved that the test only screens for detonation and that the criteria may not coincide with what would have been referred to as “flash powder” 15 years ago. AEISG requested additional time to review the proposal to try to identify any criteria that may have been over-prescribed. The United Kingdomobserved that acceptance of the test would be easier if the focus was fireworks rather than flash powder.

3. Taking note of the working group’s comments, the United States of America has continued to refine and prove the reliability of the test, particularly, concerns related to the reproducibility of test results, comparison of the test results with the HSL Flash Composition Test Method, effect of the weight of the steel confining tube on the sample mass, and preliminary results related to granulated material.  Five additional compositions have also been examined in addition to the twenty-two prior results, bringing the total DDT Flash Composition Test data base to twenty-nine different pyrotechnic substances. 

4. The experimental results with DDT Test Fixtures and with HSL Flash Composition Test Fixtures are given in annex I (English only). The revised proposals for adoption of the DDT Flash Composition test are given in annex II.

Annex I

A. Experimental results with DDT Test Fixtures

1. To address questions about the reproducibility of the DDT Flash Composition Test, the expert from the United Stateshas redone testing on all the twenty-two compositions previously reported plus added five new compositions.  All test conditions were identical to those used previously except that the steel confining sleeve had been slightly modified by the addition of a handle (See Figure 1.) which added approximately 200 grams to its weight.  The added handle made it easier to place the confining sleeve over the fiberboard sample tube and also made a convenient place to attach a 12 meter long steel cable for safety reasons. 

2.  The results of these re-tests, as compared to the original results are shown in Table I.   In twenty-one of twenty-two cases, there was no change in the outcome.  In almost every case, the results were quite reproducible in triplicate.  However, in the case of Sample No. 3, the original single positive ( + ) result could not be duplicated, even after eight re-tests.  It is believed that the first result, which was marginally positive (a very slight “tear” in the witness plate) was possibly the result of a non-uniform sample.  The composition contained an unusually wide particle size distribution of magnesium metal powder, and the one sample first tested might have been higher in the ultra-small particle size fraction than realized.

Figure 1.  Standard DDT test fixture

Table I

In addition to the original 22 formulations or compositions re-tested, the expert for the United Stateshas also tested five new compositions (23-27) which were included in Table I. Perhaps the most surprising test result was that for Formulation No. 27, a frequently used “Whistle Powder” mixture of ground potassium perchlorate and fine potassium benzoate (70/30 by weight). Even though this mixture contained NO metal particulate fuels, the results clearly show that the formulation met the proposed criteria for a “flash composition” as shown. Thus, it can be stated that “flash compositions” at least as defined by the proposed test method, need not always contain metal particulate fuels.

3. The question was asked by the expert from Germanywhether increasing the weight of the steel confining sleeve or the quantity of sample would shift the result outcomes. To fully answer the German expert’s question, a brief review of how the final Alternate Flash Composition Test Method evolved to its current state is appropriate here.

The test method was simply developed by a process of trial and error beginning with a small number of compositions which the expert from the United Statesknew were already being widely used in aerial fireworks “salute” shells. These included Formulation Sample Numbers 4-8 in Table 1. Initially, a 71 gram (3 ounce) sample size was selected, primarily because it was the approximate quantity of “Flash Compositions” commonly found in 80 mm (3 inch) diameter spherical aerial “salute” shells.  For initial tests, almost no confinement was used other than a paperboard container, which was placed on top of the witness plate.  A steel ring the same diameter as the support ring was also placed around the container, just to contain the horizontal flash effect.   The 71 gram sample formulations 4‑8 all easily and reproducibly punctured the witness plate when ignited with a Davey-Bickford electric match.  The formulations sample size was then halved to 35 grams, with the same light confinement in the same paperboard containers and the results were unchanged from the 71 gram sample sizes for Formulations 4-8. However, when the sample size was halved again to 17.5 grams, negative ( - ) results started to occur, either through poor ignition or inadequate sample depth or both. It was decided to replace the wider paperboard containers with convolute fiberboard wound tubing of approx. 25 mm in diameter and approximately 150 mm in height and thereby improve the sample depth. This replacement was partially successful, but the sample size had to be increased, first to 20 grams and then ultimately to 25 grams for optimal reproducibility.

4.  Because of the relative instability of a tall, thinner sample formulation holder under ambient wind conditions on a typical outdoor test site, the fiberboard tube was then enclosed in a 150 mm high section of inexpensive steel pipe having a diameter just slightly larger than the fiberboard.   This led to much more reproducible results, but the inexpensive steel pipe sleeves had to be frequently replaced because they bulged out from the force of the explosions.  The final solution was to custom design a confining sleeve, precisely bored from a solid billet of steel a size, weight and thickness which would prove rugged enough so as to only need very infrequent replacement.   It was decided at this point to also add top confinement to the steel sleeve design to make it more similar to other DDT apparatus.  Several prototype designs were tried for the steel confining sleeve, some lighter and some heavier, but the final dimensions of what became the “Standard”  3 Kilogram fixture  (shown in Figure 1) were chosen based on the fixture ruggedness and overall handling convenience for the technicians conducting  the tests. The confining sleeves were all machined out of non-stainless steels and will slowly corrode depending on ambient moisture, but their service life, due to continual end abuse (rising in air and falling back from achieved heights of up to 15 meters) is perhaps a few months to a year, well below any presumed long term corrosion effects that may occur.

5. However, it is logical to assume that the more the confining sleeve weighs, the longer the confinement will remain around the sample tube and therefore the more force might be directed downward on the witness plate. To prove this, a very heavy steel confining sleeve was fabricated which weighed approximately five to six times more than the “Standard” confining sleeve shown in Figure 1.   This heavy confining sleeve fixture is depicted in Figure 2 below.

Figure 2.  Heavy steel confining sleeve fixture

Table II.  Comparison of standard and heavy steel confining sleeve results

Results with the heavier confining “B” sleeve on six compositions that previously had given negative “( - )” results with the standard “A” sleeve, showed that all had been shifted to positive “( + )” outcomes except Goex 5FA black powder. 

6.  The question then evolves as to whether the heavier “B” sleeve should actually replace the standard “A” sleeve in the DDT Flash Composition test method?  There are two counter-arguments against taking that decision in the view of the expert from the United States. The first is that shifting more and more pyrotechnic compositions into the category of “flash compositions” which really do not have a potential for violent energetic release when accidentally ignited in a low state of confinement typically found in fireworks articles do not pose a significant risk in transport may “over-regulate” the fireworks industry.  The second is the operational concern of conducting the testing with a 17 kilogram steel sleeve that increases the risk of accidental ignition in the event of a sudden drop in its final positioning on the steel witness plate and puts the operator at risk of injury.  For these reasons, replacement of the Standard “A” confining sleeve with the heavier “B” confining sleeve is not preferred.

7.  The question was also raised by the expert from Japanas to whether granulating or “coating” of a composition which tested positive in the propose DDT Flash Composition would alter its performance.  Some preliminary trials were made using 25 grams of sample No. 6 ( a 70/30 potassium perchlorate / atomized aluminum mixture) coated onto rice hulls at two different volumetric ratios as show in Figure 3.  Results are shown in Table III below.

Figure 3

Table III. Comparison of Uncoated and Coated Compositions

8. These results, although limited, do show that the proposed alternate Flash Composition test method is amenable to testing substances which are non-uniform agglomerates such as rice hulls coated with 25 grams of a formulation known to have previously tested as a “flash composition”, which is another advantage when compared to the HSL Flash Composition Test fixture which is much more quantity limited. 

9. The issue of overall operational safety of the proposed DDT Flash Composition Test has been previously raised by the expert from Canada.  The chief concern to the operating personnel would be from the flying steel confining sleeve as it is propelled straight up into the air by the force of the burning 25 gram pyrotechnic sample material, but also from lateral hazard from the witness plate and supporting ring which can be thrown at high speeds as well. To address these concerns, the test methodology has now been revised to include a 12 meter long steel cable at least 1.2 cm in diameter to be attached to the standard “A” steel confining sleeve by looping it though a rugged handle which has been added along its side.   The other end of the steel cable is attached to a passenger car or truck tire, which is in turn chained to either a mooring post or to a 25 kilogram cement filled container.  The attached steel cable line, as it stretches during the flight of the confining sleeve, pulls it away from the witness plate and test ring, so that it cannot damage them further.    To protect operators from injury from a “flying” witness plate and supporting ring, the DDT Test Fixture could also be placed behind a heavy wire or heavy plywood enclosure at least 1.5-2.0 meters in height.

B. Experimental Results with HSL Flash Composition Test Fixtures

10. The expert from the United Statesalso sought to compare the DDT Flash Composition Method and the HSL Flash Composition Test Method results for the same formulations. Two HSL Time-Pressure Fixtures were constructed from the drawings and directions provided in Appendix 7 of the Manual of Tests and Criteria. Since the Chemring Energetics LLC “Vulcan” igniters were found to be unavailable in the United States, a study was first made of various available igniters to determine which might be the most suitable replacement for the “Vulcan” igniter. The results of this study are shown below in Table IV.

Table IV.  Comparison of ignition sources in the HSL flash composition fixture

Based on this comparison the Davey Fire “F” Igniter was chosen for use as having the best combination of shortest average response time and lowest average contributory pressures.

11. In attempting to gather data using the HSL Flash Composition Test Fixture on the same pyrotechnic compositions previously examined using the proposed DDT Flash Composition Test Fixture, the expert from the United Statesexperienced significant difficulties with the apparatus in its current design which slowed or hindered this effort.  Specifically, the inset screws which hold the igniter wire in position and make electric contact with the connecting pins to the firing circuit were very prone to fouling and breakage.  The minute plastic insulator that goes against one of the set screws was found to melt and seal the screw holes if the apparatus were not cooled down thoroughly between tests.  The plastic insulated leg (See Figure 4 below.) also frequently was damaged by heat and/or pressure and had to be constantly replaced.   As a result of these equipment design limitations, data gathering was extremely slow with frequent clean-outs and replacements of the set screws and plastic parts.  Often, only a few tests could be run before the whole apparatus had to be dis-assembled, cleaned or re-bored and re-threaded.

Figure 4.  HSL Fixture Plastic Insulated Component Damage

12. Twenty-twoof the pyrotechnic compositions that were previously tested with the DDT Flash Composition Test Method were re-tested using the HSL Flash Composition Test as given in Annex 7 of the UN Manual of Tests and Criteria and the results are shown in Table V below. All tests were run three times and the shortest interval of the three firings was used for classification per Paragraph 3.2 of Annex 7 in the UN Manual of Tests and Criteria (5^th Revision).

13. Good agreement between the HSL Flash Composition  Test Results shown in Table V and the proposed DDT Test Results in Table I was achieved for Formulations 4-8, Formulations 11, 12 and 22.

Table V.   HSL flash composition test results

14.  Anomalous correlation results between the HSL Flash Composition Test Results in Table V and the proposed DDT Flash Composition test results were seen with Formulations 1-3, 9, 10 and 13-21.  In all these cases but one (Formulation No. 9), the HSL Flash Composition Test gave a positive outcome whereas the DDT Flash Composition gave a negative outcome, indicating that the HSL Flash Composition was giving  a more “conservative” assessment of the hazard of that particular formulation than the proposed DDT Flash Composition Test.  Formulation No. 9 which gave a negative outcome in the HSL Flash Composition Test and a positive outcome in the proposed DDT Flash Composition Test, both initially and upon seven additional trials, represents the only unexplainable experimental result.  The specific potassium perchlorate/Magnalium mixture in Formulation No. 9 is known to have been used in “salute” shells for producing typical large aural effects, so the HSL Test results are of concern. 

15.  Moreover, the classification of traditional black powders like formulation No. 1 as “flash compositions” using the HSL Flash Composition Test Criteria raises the question of fireworks industry “over-regulation” because black powder, either a propellant or expellant composition, has historically not been considered a “flash composition”.  As noted in Table II, even with heavy steel confinement, Goex 5FA black powder did not damage the witness plate sufficiently to be considered a “flash composition” in the strictest sense.

Annex II

Revised proposals for adoption of the DDT Flash Composition Test

Proposal 1: Revise Note 2 of 2.1.3.5.5 of the Recommendations on the Transport of Dangerous Goods Model Regulations (UN Model Regulations Default Fireworks Classification Table) to read as follows:

“Flash Composition” in this table refers to any pyrotechnic substances in powder form or as pyrotechnic units as presented in fireworks that are used to produce an aural effect, or used as a bursting charge or lifting charge, unless the time taken for the pressure rise is demonstrated to be not more than 8 milliseconds is demonstrated to be more than 8 ms. For 0.5 g of pyrotechnic substance in the HSL Flash Composition Test in Appendix 7 of the Manual of Tests and Criteria, or unless at least two positive ( + ) results are achieved in up to ten (10) trials for a 25 g net weight of pyrotechnic substance in the DDT Flash Composition Test in Annex XX of the Manual of Tests and Criteria.

Proposal 2:Add the following procedure as a new Appendix XX to the Manual of Tests and Criteria:

DDT Flash Composition Test

Introduction

The DDT Flash Composition Test may be used to determine if a pyrotechnic substances in powder form or as pyrotechnic units as presented in fireworks that are used to produce an aural effect or used as a bursting charge or lifting charge, may be considered a “Flash Composition” for the purposes of the UN Model Regulations Default Fireworks Classification Table in Section 2.1.3.5.5 of the Model Regulations.

Apparatus and materials

The experimental set up for the DDT Flash Composition Test consists of a heavy-wall cardboard or fiberboard convolute sample tube with an inside diameter of 25.4 mm and height 150 mm with a maximum wall thickness of 3.8 mm, closed at the base with a thin cardboard or paperboard plug or cap just sufficient to retain the sample.  This cardboard or fibreboard sample tube is centred on a square rolled mild steel witness plate of the same type used in the UN 5a Cap Sensitivity Test - 1 mm in thickness and 160 mm on edge. The ignition source is provided by an electric igniter inserted in the top of the pyrotechnic sample in the tube.  Any suitable electric igniter may be used for this purpose, provided the lead wires are at least 30 cm in length. The lead wires are bent so the electric igniter match-head is approximately in the center of the interior pyrotechnic sample column and to a depth of approximately 10 mm. Another paperboard or cardboard plug or cap is then inserted into the top of the fibreboard tube to retain the positioning of the igniter wires.  The lead wires of the igniter outside the fibreboard sample tube are then bent down and travel along the outside of the sample tube to the bottom.

A mild steel confinement sleeve which is bored from a solid billet approx 1 mm deeper than the overall sample tube length and having an inside diameter of 38 mm, an outside diameter of 100 mm and a height of 165 mm with a rugged steel handle attached and a notch or groove cut into one radius of the open end sufficient to allow the igniter lead wires to pass through and weighing approx. 3 kilograms is then placed over the sample tube and also rests on the witness plate.

Supporting the sample tube and its surrounding steel confining sleeve is the square shaped steel witness plate which is itself supported on a steel ring of approximately 51 mm height with an inner diameter of approx. 93.5 mm, an outer diameter of approx. 100 mm having a 6.5 mm wall thickness. The entire apparatus is placed onto a square shaped steel base plate of approx. 25mm thickness and 152 mm on edge.  (See Figure YY below.).  The handle of the steel sleeve can be attached securely to a steel safety cable or chain to restrain its travel by any means appropriate.  As a precautionary measure, the test stand fixture may be enclosed on three sides with heavy chain-link fencing, concrete block or plywood at least 18 mm thickness up to 1.5 meters in height.

Procedure

Prior to testing, all pyrotechnic substances whose sensitivity could be moisture dependent should be stored for at least 24 hours in desiccators at a temperature of 28 - 30 °C. Twenty-five (25) grams net weight of the pure pyrotechnic substance to be tested, as a loose powder or granulated or coated onto any substrate, is pre-weighed and then poured carefully into a fiberboard sample tube with the bottom end closed with one of the cardboard or paperboard caps or plugs. After filling, the top cardboard or paperboard cap or plug can be inserted lightly to protect the sample from spillage during transport to the test stand. The height of the sample substance in the tube will vary, depending on its density. The sample should be first consolidated by lightly tapping the tube on a non-sparking surface. The final density of the pyrotechnic substance in the tube should be as close as possible to the density achieved when it contained in a fireworks device. The sample tube is placed in the center of the steel confining sleeve fixture shown in the diagram in Figure YY. which rests on the witness plate, steel ring and steel base plate.

The steel base plate, supporting ring and witness plate are pre-positioned on the test stand.   If present, the paperboard or cardboard to plug or cap of the fibreboard sample tube is removed and the electric igniter is inserted into the top of the pyrotechnic composition to be tested and visually positioned to an approximate depth of 10 mm. The paperboard or cardboard top cap or plug is then inserted or re-inserted, fixing the igniter’s position in the fibreboard sample tube and the depth of its match-head. The lead wires are bent over and down along the sidewall and bent away at the bottom.  The sample tube is placed vertically and centred exactly on the steel witness plate.  The 3 kilogram “tethered” steel confining sleeve is placed over the fibreboard sample tube and re-centred. The igniter lead wires are positioned to pass through the slotted groove in the bottom edge of the steel confining sleeve and will be ready to attach to the firing circuit apparatus.

The entire apparatus is made secure so nothing will shift or change orientation.  As a safety precaution, the fourth side of the test bay could then be closed with a portable section of chain-link fence or 18 mm plywood sheets to a height of at least 1.5 meters.  The electric igniter is then initiated from a safe position from the test bay. After initiation and a suitable interval to allow for falling debris, if any, the witness plate is recovered and examined. The test should be repeated at least twice, and as many as ten (10) times if necessary.  

Test criteria and method of assessing results

The result of any one test shall be considered positive, ( + ) if the pyrotechnic substance being tested is found to have undergone a “deflagration to detonation transition” (DDT) event.  The proof of a DDT event shall be if the witness plate is torn, perforated, pierced or otherwise penetrated (i.e. light is visible through the plate). NOTE: Bulges or folds in the witness plate shall NOT be considered to be proof of DDT event and those results shall be considered “( - )”.

For any pyrotechnic substance to meet the definition of a “Flash Composition” by the DDT Flash Composition Test, there must be at least two (2) positive ( + )  DDT events in up to ten (10) sequential trials.  For example, if there are two consecutive positive ( + ) DDT events in the first two trials, the pyrotechnic substance can be considered a “Flash Composition” without further testing.   If however, there is only one positive  ( + ) DDT event occurs in the first two trials, and there are intervening negative ( - ) results, testing should be continued until either another positive ( + ) result is obtained or until ten (10) trials have been conducted, whichever comes first.  If after ten consecutive trials there is still only one positive ( + ) result, the pyrotechnic substance shall NOT be considered as having met the definition of a “Flash Composition.

Figure YY

联合国危险货物运输专家和全球化学品统一分类和标签制度专家委员会

危险货物运输专家分委员会

第四十一次会议

日内瓦, 2012年6月25日 – 2012年7月4日
议程第2 (d) 项

爆炸物和相关内容：闪光成分的DDT试验和标准

使用默认烟花爆竹分类表进行分类时，提议的替代闪光成分试验方法

由美国专家提交

背景

1. 在37次会议上，美国专家首次提出对现行用于评估烟火药混合物闪光成分的HSL试验的替代试验方法（爆燃转爆轰[DDT]闪光成分试验），该提案通过10种不同物质的试验数据来说明替代试验方法的可行性(见ST/SG/AC.10/C.3/2010/31)。在39次会议上，美国专家再次提供了改进后的DDT闪光成分试验设备图纸，以及根据该方法进行的另外12种物质的试验结果（见39次会议非正式文件，INF.44）。同时，来自日本（INF.22）和德国（INF.16）的专家也提交了非正式文件。总体上，这些试验结果和美国专家获取的试验结果是一致的。

2.  由于推荐方法使用的样品量较大且比较容易实现，爆炸品工作小组认为该方法是一种可行的替代方法，该小组一致同意继续推进美国提交的DDT试验方法（见39次会议，爆炸品工作小组的报告文件，INF.58）。对于以前收到的各方评论，本提案都已做过相关工作并进行了说明。如，德国专家指出在试验过程中，钢套筒的尺寸可能会引发一些安全问题；同时，钢套筒重量可能是一个影响试验结果的因素，并且建议进行进一步的研究。而日本和英国的研究说明成分的粒度会影响试验结果，因此他们认为试验方法内应当进一步考虑颗粒状物质的试验；他们也认为钢套筒的重量可能是一个安全或影响试验结果的因素，并支持德国方面提出的应进一步研究的建议。其他专家，如荷兰和澳大利亚专家认为样品量可能会影响试验结果，并建议对这一潜在的影响应当进行进一步的研究。英国方面认为该试验操作直观可行，支持其进一步的发展。荷兰专家认为，该方法只是用于筛选是否发生爆轰，因此和15年前 “闪光药剂”定义的标准并不冲突。AEISG 建议花更多的时间对提案进行进一步审阅，以便确定是否存在任何规定过严的现象。英国专家认为，若将关注的焦点放在烟花爆竹，而非闪光药剂上，则试验方法可能更容易被接受。

3. 基于工作组的评论，美国专家一直在改善试验和证明试验方法的可靠性，特别是在试验结果的重复性方面，和HSL闪光成分试验方法进行了比较，进行了钢约束套筒重量对样品量的影响评估，并给出了颗粒状物质相关的初步试验结果。在原先22中试验结果基础上，新增了5种组分的试验，这样，就使DDT闪光成分的试验数据库包含了29种不同烟火类物质的试验结果。

4. 使用DDT试验装置和HSL试验装置进行试验的结果比对见附件I。采纳DDT闪光成分试验方法的提案修订稿见附件II。

附件I

A. 使用DDT试验装置得出的试验结果

1. 为了说明有关DDT闪光成分试验重复性的问题，美国专家对先前已有的22种组分以及新增的5种组分重新进行了试验。钢约束套筒进行过小幅改进，新增了一个把手（见图1），大约增加了200g的重量，除此之外，其他试验条件都和之前的相同。新增把手使约束套筒可以更为轻易的套在纤维样品管上，也是一个方便的地方可以系上12m长的钢索来增加安全性。

2.  重新试验的结果和原始结果的比较见表1。22个组分里边的21个组分的试验结论都未发生改变。几乎每一个组分的三个试验系列的结果重复性都很好。但是，对于3号样品，即使进行了8次重复试验，却也没有再现原先已有的一次阳性（+）结果。可能是由于样品不规则的原因导致了第一次试验出现临界阳性结果（验证屏只有轻微的“撕裂”）。该组分包含了不常见的大粒度分布的金属镁粉，而第一次试验样品含有的极细小粒径的比例可能高于预期。

                                                                          图1  标准的DDT试验装置

表I

除了重新测试原有的22种配方或组分外，美国专家也增加了5种新组分（23-27）的试验，这些结果也已包含在表1内。这其中，最出乎意料之外的结果恐怕就是27号配方了，该配方是通常用作“笛音药剂”的混合物，含研磨过的高氯酸钾和精细苯甲酸钾粉末（70/30，按质量比）。尽管该混合物并没有金属颗粒燃料，但其结果却很明确的表明该配方符合“闪光成分”的定义标准。由此，按照所提试验方法进行定义的话，“闪光成分”并不总是含金属颗粒燃料。

3.  德国专家曾问过，增加钢约束套筒的重量或样品量，是否会改变试验结果。为了更好的回答德国专家的这个问题，首先，有必要对于此闪光成分替代试验方法如何演变为现行状况进行一个简单的回顾。

该试验方法的形成是一个不断试验和纠错的过程，刚开始时所用的小量成分据美国专家了解已经广泛应用在高空烟花“礼花弹”中，即表1中的4-8号配方。起初只用了71g（3 盎司）的样品量，主要原因就是这一样品量是80mm（3 英寸）直径的球形高空“礼花弹”所使用“闪光成分”的大致数量。最初的试验除了纸板容器几乎没有用其它的约束，纸板容器是直接放在验证屏上面的。和支撑环同样直径的钢环放在纸板容器四周，只是为了限制水平喷射效果。使用Davey-Bickford 电点火头进行点火时，71g的4-8号配方样品都重现了对验证屏的穿透结果。于是将4-8号的样品量减半至35g，使用的同样的纸板容器、施加轻微的约束，和71g样品量相比，结果并无改变。但是，当样品量再次减半至17.5g时，阴性（-）结果开始出现，不是难以点火就是样品深度不够，或者两者都有。于是决定使用直径约25mm，高约150mm的卷曲纤维板管代替宽大的纸板容器，以此增加样品深度。这一替换取得了部分成效，但必须增加样品量，起先是20g，最终增至25g时取得了最佳的重复性。

4.  高瘦的样品容器受野外试验场的风速影响较大，因此用了一个比纤维板容器稍微大点的钢管将其罩在里边。这样一来，结果的重复性更好，但是便宜的钢管套筒在爆炸力作用下会胀大，因此不得不进行频繁的更换。最终的解决办法是在钢块上进行精确挖孔，形成约束套筒，该套筒的尺寸、重量和厚度都设计成能承受足够的膨胀力，从而不需要进行频繁的更换。同时，也在钢套筒上增加了顶部约束，这样使其和其他DDT设备更为类似。钢约束套筒的设计有几套方案，质量有重有轻，但根据装置的强度和试验时总体的操作便利性，最终的尺寸定为“标准的”3kg（如图1）。约束套筒都不是用不锈钢制作而成，因此会因环境湿度发生缓慢的锈蚀，但由于试验时的极端使用（被抛向空中，然后从高达15m处坠落），其使用期限大概为几个月到一年，远远小于其锈蚀作用可能发生时间。

5. 然而，逻辑上若约束钢套筒越重，那么在样品周围产生的约束就越久，因此验证屏就承受更多向下的力量。为了验证这一点，做了一个大约是“标准”约束套筒（图1）的5到6倍重的钢约束套筒。这一重的约束套筒装置如下图2。

图2 重约束套筒装置图表

表II  标准约束套筒和重约束套筒的试验结果比对

除Goex 5FA黑火药外，之前使用标准“A”套筒得到阴性（-）结果的6种组分，在使用质量更重的“B”套筒时，都变成了阳性（+）结果。

6.  那么问题又来了，在DDT闪光成分试验方法中，更重的“B”套筒是否应该用来替换标准的“A”套筒呢？在美国专家看来，以下两方面可以反驳这种观点。其一，若将越来越多的烟火药剂归到“闪光成分”，而这其中很多成分在低约束条件下被意外点燃时确实不存在剧烈的能量释放，特别是烟花类产品在运输中并不呈现出明显的危险性，因此这样对烟花行业就有可能是“控制过头”了；其二，从操作角度考虑，将重达17kg的钢套筒突然间放置在验证屏上时，增加了意外点燃的风险，这样势必增加操作人员受伤害的危险。因此，不建议使用重的“B”约束套筒来代替标准的“A”约束套筒。

7.  日本专家提到，对在DDT试验中呈阳性的成分，在造粒或者“涂层”后，是否会改变其试验结果。针对此，将25g的6号样品（70/30的高氯酸钾/雾化铝混合物）按不同体积比例（图3）涂在稻谷壳上，进行了一些初步的尝试。试验结果见表III。

图3

表III.  涂层样和不涂层样的比对

8.  虽然试验不多，但也说明所提议的替代闪光成分试验方法对于不规则的块状物仍然是有效的，如涂有25g已知是“闪光成分”配方的稻谷壳涂层。与样品量限制较大的HSL闪光成分试验相比，这又是一个优点。

9.  加拿大专家提到DDT闪光成分试验的总体操作安全事宜。对于操作人员而言，主要考虑的是飞起的钢约束套筒，因为受25g烟火样品燃烧的冲击力影响，套筒是直接推向空中的，但同时也要考虑由于验证屏和支撑环高速抛射后产生的侧向危险。为了解决这些问题，现在试验方法已经包含了一个12m长、直径至少1.2cm的钢索，钢索一端系在标准钢约束套筒A的把手上，另一端系在一客车或货车轮胎上，该轮胎和抛锚桩或者灌满25kg水泥的容器链接在一起。相连的钢索，在约束套筒飞行期间也随之伸展，并将其从验证屏和支撑环上拉开，避免对他们产生破坏。为保护操作者不受“飞行”的验证屏和支撑环的伤害，DDT试验装置也可以安装在巨大的金属网后面或者至少1.5-2.0m高的重胶合板围墙内。

B.  HSL闪光成分试验结果

10. 同时，美国专家也寻求使用同样的配方进行DDT闪光成分试验和HSL闪光成分试验的比较。专家按照试验和标准手册的附录7提供的图纸和说明，配置了2台HSL时间-压力设备。由于Chemring Energetics LLC “Vulcan”点火头在美国无法获取，因此，首先进行的研究是比较可获取的各种点火头，确定哪个点火最适合于替代“Vulcan”点火头。该研究的结果如下表IV。

表IV. HSL闪光成分试验装置点火源比对

基于上述比较，由于Davey Fire “F”点火头在最短平均响应时间和最低平均贡献压力上吻合的最好，因此被选用为试验用点火头。

11. 在使用之前已有DDT试验数据的几种组分进行HSL闪光成分试验过程中，该设备的设计使美国专家遇到了不小的麻烦，以至于明显减缓或阻碍了试验进程。特别是用于安装点火头和连接放电回路的螺丝非常容易弄脏和破损。而且在试验过程中，如果设备没有完全彻底的冷却下来，螺丝钉另外一头的小塑料绝缘部件容易熔化并堵住螺丝孔。塑料绝缘腿（见下图4）也经常在高热和/或者压力作用下破损，不得不经常更换。正是由于这些设计上的限制，需要不停的清洗和更换螺丝和塑料部分，使得数据获取过程极其缓慢。往往进行了只几次试验后，整个试验设备就需要进行重新拆解、清洗或重新打螺纹孔。

图4.  HSL装置塑料绝缘部件损坏程度

12. 对于之前使用DDT闪光成分试验方法进行试验的22种烟火组分，美国专家又使用UN试验和标准手册附录7给出的HSL闪光成分试验方法进行了重新试验，试验结果见下表V。按照UN 试验和标准手册（第5修订版）附件7 第3.2章，所有组分的试验都进行3次，取3次试验的最小值作为判定数据。

13. 对于4-8号配方，11，12和22号配方，表V所示HSL闪光成分试验结果和表I所示DDT试验结果一致性很好。

表V.  HSL闪光成分试验结果

14.  对于1-3号，9号，10号和13-21号配方，表V中的HSL闪光成分试验结果和提议的DDT闪光成分试验结果之间的相关性出现了不一致。除9号外，所有其他组分的HSL方法给出的是阳性结论，而DDT方法给出的是阴性结论，说明对于这些配方，HSL方法给出了比提议的DDT方法更为“保守”的危险性评估。9号组分，在初始试验和后续的7次附加试验中，HSL试验给出的都是阴性结论，而提议的DDT方法给出的都是阳性结论，这成了其中唯一不能解释的试验结果。9号组分中的特定高氯酸钾/镁铝混合物，是被用于制作产生巨大听觉效果的“礼花弹”，因此，HSL的试验结果更有意义。

15.  此外，使用HSL方法进行试验，像1号配方这样的传统黑火药都被归为“闪光成分”，会引起烟花行业的“过度控制”问题，因为无论是用作推进剂还是驱动组分，在历史上黑火药都不被认为是一种“闪光成分”。从最严格角度来看，如表II所示，即使使用重的钢约束，Goex 5FA黑火药都没有将验证屏破损至应归类为“闪光成分”的程度。

附件II

推荐采用的DDT闪光成分试验方法的提案修订稿

提议 1：修改危险货物运输建议书 规章范本（UN规章范本，默认的烟花爆竹分类表）2.1.3.5.5的注释2如下：

本表中的“闪光成分”是指在烟花爆竹中用于产生听觉效果，或用作爆炸药，或升空药的粉末状烟火药物质，除非使用0.5g烟火药物质按试验和标准手册附录7的HSL闪光成分试验得出的压力上升时间大于8ms。或者，除非使用净重25g的烟火药物质按照试验和标准手册附录XX DDT闪光成分试验进行至多10次试验时，至少得到两个阳性（+）结果。

提议 2：将下述程序作为新的附录XX加入试验和标准手册内：

 DDT闪光成分试验方法

介绍

使用规章范本2.1.3.5.5节中的默认烟花爆竹分类表时，DDT闪光成分试验可以用于确定烟花爆竹中用于产生听觉效果、爆炸效果，升空效果的粉末状烟火药物质或烟火药单元，是否应当归为“闪光成分”。

设备和材料

DDT闪光成分试验装置包括：

由厚的硬纸板或纤维板卷曲成的样品盛放管，该管内径25.4mm，高150mm，最大壁厚3.8mm，使用刚刚好能够保持住样品的薄硬纸板或纸板塞或盖子封闭底部。硬纸板或纤维板样品管放置于正方形轧制软钢验证屏中央，验证屏类型和UN 5a雷管敏感度试验中的一致，即1mm厚，边长160mm。从管内烟火药样品顶部插入电点火头充当点火源；只要导线至少有30cm长，任何合适的电点火头都可以用作点火源。弯曲导线以使点火头放置位置大约保持在烟火药样品柱的中间，且深度大约为10mm。为保持点火引线的位置，另外使用一个纸板或硬纸板塞子或盖子插入纤维板管的顶部。弯曲纤维板样品管外的点火头导线使其沿着纤维管外管壁至底部。

由钢块钻削而来的软钢约束套筒，管内长度比样品管总长深约1mm，内径38mm，外径100mm，高165mm，带弯曲的把手，开口端挖一个足于使点火头导线通过的凹痕或沟槽，重约3kg，也直立于验证屏，放置在样品管之上。

四方形钢验证屏用于放置样品管及其周围的钢约束套筒，而钢验证屏又由钢环支撑，钢环高约51mm，内径约93.5mm，外径约100mm，壁厚6.5mm。整个试验装置放在厚约25mm，边长152mm的四方形钢座上（见下图YY）。钢套筒的把手可以系上安全钢索或链条，以通过合适的方式限制其运动轨迹。作为一个预防措施，可以使用厚重的链环栅栏，混凝土块或胶合板在三面隔离试验装置，厚度至少18mm，高至少1.5m。

程序

试验前，对于水分含量会影响其敏感度的烟火药物质，应当在温度28 - 30 °C条件下的干燥器内至少储存24h。无论是以松散的粉末还是颗粒状物质亦或是涂于某一种附属物上的形式进行试验，都应预先称取净重25g的烟火药物质，然后小心倒入已经用纸板或硬纸板盖子或塞子封住底端的纤维板样品管内。填充完毕后，可以轻轻的塞入顶部的硬纸板或纸板盖子或塞子，以防在搬运至试验台过程中洒出。因密度不一，样品在管中的高度各异，应将样品管道在不会产生火花的表面上轻轻的撞击以压实样品。试验管中烟火药物质的最终密度应尽量接近于烟花爆竹产品内含物的密度。如图YY所示，样品管放在钢约束套筒中间，然后立在验证屏、钢环和钢座上。

预先将钢底座、支撑环及验证屏放置于试验场地。样品放置好后，移除纤维板样品管上的纸板或硬纸板塞子或盖子，然后将电点火头从顶部插入，目测放置在大约10mm深处。随后插入或重新插入纸板或硬纸板塞子或盖子，固定点火头在样品中的位置及其头部的深度。导线弯曲后沿着外壁面向下然后再在底部弯曲离开。样品管垂直放置并且中心正好保持在钢验证屏上，3kg带把手的钢约束套筒放在纤维板样品管上，并重新对中。点火头导线穿过钢约束套筒底部挖好的小沟槽，准备和点火回路设备相连。

整套设备的制作已考虑安全因素，使用时不需任何的变换或方向改变。不过，作为一个安全预防措施，可以再使用高至少为1.5m的链锁墙或18mm厚胶合板制作成一段可移动的防护墙，从而形成一个四面包围的试验堡。在试验堡外的安全位置开启电点火头，实验触发，待碎片掉落完全后，重新找出验证屏（如有），并对其进行鉴别。试验至少进行2次，如果需要可达到10次。

试验标准和结果的评估方法

若发现被测烟火药物有发生过“爆燃转爆轰”（DDT）现象，则结果应为阳性（+）。可以通过验证屏是否撕裂，穿孔，刺穿或其他一些穿透现象（如，透过验证屏可以看见光亮），来判定是否发生DDT现象。

注：验证屏的鼓胀或折叠现象不应作为DDT现象发生的证据，这时结果应为阴性”(-)”。

某一种烟花药物质在进行DDT闪光成分试验时，若在至多10次连续的试验中，至少有2次是阳性（+）结果，则确定该物质符合“闪光成分”的定义。举例来说，若在前两个序列试验中已经得到连续两次的阳性（+）结果，则该物质就可以直接归为“闪光成分”，而不用进一步的试验。若前两次序列试验中，只出现一次阳性（+）结果，其中有一次是阴性（-）结果，则应该继续进行试验，直到再次出现阳性（+）结果为止，或者一直进行至10个序列试验结束，以较早发生者为准。若10个序列试验完成后，只有一次阳性（+）结果，则该烟火药不应认为符合“闪光成分”定义。

图 YY