ISO/DIS 24499

ISO/TC 86/SC 8

Secretariat: ANSI

Date: 2025-02-14

Method of test for burning velocity measurement of A2L flammable gases

Méthode d'essai pour mesurer la vitesse d'inflammabilité des gaz inflammables A2L

DIS stage

Warning for WD’s and CD’s

This document is not an ISO International Standard. It is distributed for review and comment. It is subject to change without notice and may not be referred to as an International Standard.

Recipients of this draft are invited to submit, with their comments, notification of any relevant patent rights of which they are aware and to provide supporting documentation.

© ISO 2025

All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication may be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying, or posting on the internet or an intranet, without prior written permission. Permission can be requested from either ISO at the address below or ISO’s member body in the country of the requester.

ISO copyright office

CP 401 • Ch. de Blandonnet 8

CH-1214 Vernier, Geneva

Phone: + 41 22 749 01 11

E-mail: copyright@iso.org

Website: www.iso.org

Published in Switzerland

Contents

Foreword

Introduction

Scope

Normative references

Terms and definitions

General/principle of the test

General

Principle of the test method

Measurement parameters

General

Flame propagation velocity

Flame surface area

Cross-sectional area of the flame base

Test method

General

Gas handling and mixtures preparation

The test tube

Ignition

Flame front visualization

Purge, exhaust and gas treatment systems

Test temperature setting

Experimental protocol for mixtures prepared using partial pressure technique

Evaluation and expression of results

General

Uncertainty

Safety precautions

Overview on flame shape, propagation regimes and stability

Flame shape

Flame propagation regimes

Flame stability in tubes

Observations of flames in tubes

Flame quenching in circular tubes

Flame propagation velocity and tube diameter

Flame area calculation

Bibliography

Foreword

ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies (ISO member bodies). The work of preparing International Standards is normally carried out through ISO technical committees. Each member body interested in a subject for which a technical committee has been established has the right to be represented on that committee. International organizations, governmental and non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.

The procedures used to develop this document and those intended for its further maintenance are described in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the different types of ISO document should be noted. This document was drafted in accordance with the editorial rules of the ISO/IEC Directives, Part 2 (see www.iso.org/directives).

ISO draws attention to the possibility that the implementation of this document may involve the use of (a) patent(s). ISO takes no position concerning the evidence, validity or applicability of any claimed patent rights in respect thereof. As of the date of publication of this document, ISO had not received notice of (a) patent(s) which may be required to implement this document. However, implementers are cautioned that this may not represent the latest information, which may be obtained from the patent database available at www.iso.org/patents. ISO shall not be held responsible for identifying any or all such patent rights.

Any trade name used in this document is information given for the convenience of users and does not constitute an endorsement.

For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and expressions related to conformity assessment, as well as information about ISO's adherence to the World Trade Organization (WTO) principles in the Technical Barriers to Trade (TBT), see www.iso.org/iso/foreword.html.

This document was prepared by Technical Committee ISO/TC 86, Refrigeration and air-conditioning, Subcommittee SC 8, Refrigerants and refrigeration lubricants.

Any feedback or questions on this document should be directed to the user’s national standards body. A complete listing of these bodies can be found at www.iso.org/members.html.

Introduction

Safety classification and relative flammability properties of refrigerants are a critical part of ISO 817. The flammability limits of refrigerant gas in air, as described in Annex B.1 of ISO 817^[1]1) give a partial measure of the relative flammability. Another dimension of flammability is how fast a substance burns, releases energy and spreads a flame. One can measure the rate at which a flame front moves through a cloud of refrigerant gas in air, or its burning velocity (BV). This document describes one method that can be useful for BV measurement, and thereby better quantify and compare relative flame fronts for some refrigerant classes. In this test, a flame is allowed to propagate upward (vertically) through a well-mixed, quiescent column of a refrigerant-air mixture enclosed in an open-ended glass tube. Optical systems are used to measure the upward velocity of the flame front.

The measurement of BV has been widely used in the past to compare highly energetic fluids, such as motor fuels and rocket propellants. The BV measurement of slow burning fluids, such as ammonia and fluorinated refrigerants can be more difficult to measure due to the inherent instability of a slow flame. The low rate of energy evolution from a slow flame makes it susceptible to quenching from a variety of sources. For slow burning refrigerants, turbulence and convection currents, can break the flame front, and hence quench the flame. In addition, the test chamber surface can quench free radical flame intermediates as well as extract some of the heat necessary for flame propagation. Gas-phase thermal radiation is also important for flames with low burning velocity. These effects are important to note as they tend to diminish and sometimes quench a weak flame.

The use of the vertical tube method for BV characterization of slow burning refrigerants was the subject of doctoral research which was used in Annex C of ISO 817:— and is the basis for this document^[1]. In addition, ASHRAE research notes the use and limitations of the vertical tube technique^[2][3]. While the basic framework of the method is relatively simple, some sophisticated imaging instrumentation and mathematics are necessary to extract an average local burning velocity separate from the bulk burning speed as the flame progresses up the tube. Since 2004 other laboratories have used the basic principle of vertical tube method and have shown acceptable results for reproducing the measurement of R-32, at 6,7 +/- 0,7 cm/s. Slower burning velocities (i.e. <4 cm/s) become more difficult to measure reproducibility, so variability can increase as flame instability is increasing. The lower burning velocity limit of this method, as described, is between 3 cm/s and 4 cm/s, depending on the actual design and geometry of the apparatus being used. The uncertainty of the measurement of flames that burn more slowly than R-32 has not yet been determined in any multi lab comparative testing. The appealing aspects of this test are the relative simplicity and low cost of its implementation.

Method of test for burning velocity measurement of A2L flammable gases

This document specifies a method of measuring the burning velocity (BV) of slowing burning refrigerants (< 10 cm/s) for use with other standards that utilize the BV for determining safety classification of refrigerants (e.g. ISO 817) or that use the BV in establishing requirements on the use of slow burning refrigerants (e.g. ISO 5149).

There are no normative references in this document.

For the purposes of this document, the following terms and definitions apply.

ISO and IEC maintain terminology databases for use in standardization at the following addresses:

• ISO Online browsing platform: available at https://www.iso.org/obp
• IEC Electropedia: available at https://www.electropedia.org/

blends

mixtures composed of two or more refrigerants

burning velocity

S_u

maximum velocity at which a laminar flame (3.5) propagates in a normal direction relative to the unburned gas ahead of it

Note 1 to entry: This value is expressed in centimetres per second.

combustion

exothermic reaction between an oxidant (e.g. air) and a combustible fuel

compound

substance composed of two or more atoms chemically bonded in definite proportions

flame

space where combustion takes place, resulting in a temperature increase and light emission

flame propagation

combustion, causing a continuous flame (3.5) which moves upward and outward from the point of ignition without the influence of the ignition source

flame propagation velocity

S_s

velocity at which the continuous flame (3.5) moves upward and outward from the point of ignition without the point of ignition and without the influence of the ignition source

flame surface area

A_f

surface area of the flame (3.5) generated during the combustion of the flammable (3.9) gas

flammable

property of a mixture in which a flame (3.5) is capable of self-propagating

quenching

effect of extinction of the flame near a surface due to heat conduction losses, absorption of active chemical species and viscous effects of the surface

refrigerant

fluid used for heat transfer in a refrigerating system, which absorbs heat at a low temperature and a low pressure of the fluid and rejects it at a higher temperature and a higher pressure of the fluid usually involving changes of the phase of the fluid

stoichiometric concentration

C_st

concentration of a fuel in a fuel-air mixture that contains exactly the necessary quantity of air (approximately 21 % O₂ / 79 % N₂ by volume) needed for the complete oxidation of all the compounds (3.1.4) present

The test method is based on:

the initiation of the combustion of the gas, or blends of gases, in a stagnant homogeneous mixture with air contained in a vertical cylindrical tube;

the observation and the recording of the flame propagation;

determining the surface area of the flame.

The burning velocity is a function of the flammable gas concentration in the total mixture with air. The burning velocity reaches a maximum in the vicinity of the stoichiometric concentration.

This test method involves the use of hazardous substances and therefore requires, for a safe handling and testing, the knowledge of safety parameters and prevention measures. These measures shall be the user’s responsibility. However, general safety precautions are given in Clause 8.

The test method consists of initiating the combustion of a homogeneous mixture of a flammable gas (or a flammable mixture of gases) and air, contained in a vertical tube opened at the lower ignition end, and propagating a flame upwardly to the upper closed end; see Figure 1. In the early stages of this propagation, there is a phase of uniform movement during which the shape and the size of the flame are constant.

Taking into account the mass and species balance through the flame front, the burning velocity, S_u, is calculated from the knowledge of the flame propagation velocity, S_s, in the tube and the ratio of the flame surface area to its base cross-sectional area. The volume of burned gas per second and per unit area, or the burning velocity, S_u, is obtained by dividing the mixture volume which is consumed per second, at the test temperature and pressure, by the flame surface area, A_f (the subscript “f” denotes the flame). The volume consumption of the mix per second is the volume swept by a cross-sectional area of the flame base, a_f, with a velocity equal to the flame propagation velocity S_s. Formula (1) is used to determine volume consumption per unit time.

(1)

where

NOTE The cross-sectional area of the flame base is equal to the tube cross-section reduced by the quenching area (the area between the edge of the flame and the tube wall).

At a given temperature and pressure, the burning velocity is only a function of the type of flammable substance and its concentration with the oxidant and is dependent to a limited extent on the experimental apparatus.

Key

Figure 1 — Schematic of the flame propagation in a vertical tube

The measurement of the burning velocity requires the knowledge of the following three parameters of Formula (1):

1. the flame propagation velocity, S_s;
2. the flame surface area A_f;
3. the cross-sectional area of the flame base a_f.

The flame propagation velocity in the tube is required for the measurement of the burning velocity. As a condition to the derivation Formula (1), only parts of uniform flame propagation shall be considered in the measurements (constant S_s).

The linear propagation velocity of the flame is obtained from the direct measurement of the flame front displacement determined by two successive images with a known time interval (30 Hz to 50 Hz) of the camera acquisition frequency. More than one succession of images shall be used to check that the flame propagation is uniform. An image treatment is necessary in order to enhance the flame front shape and to locate on both images an identical luminous spot (pixels with equal brightness level) that corresponds to the same location on the front and deduce the flame front displacement. This procedure is proved necessary with low luminosity flames since any uncertainty in the flame front displacement leads to an uncertainty in the flame propagation velocity and thus on the burning velocity.

The flame front shape cannot be generated by the revolution of a parabola nor by the approximation by an ellipsoid segment, even though in many cases this shape is symmetrical. An accurate method is needed to calculate the flame surface area A_f. For an upward propagation, the flame usually shows a symmetrical front surface referred to the tube axis. For a uniform propagation, the shape of the flame front remains constant. Fast moving flames are almost hemispherical, the slower flames are somewhat elongated.

9.7 describes a mathematical and geometrical model to calculate the flame surface area. In summary, the flame front profile is marked with fitting points (20 to 40 fitting points) then divided into two or more horizontal sections. The fitting points shall be selected on the rim of the most luminous zone on the flame front.

For each section a polynomial fit equation of appropriate order is made in order to give the best fit curve to the points selected on that section. The best fit gives the minimum deviation of the fit curve to the fitting experimental points. The area of each section shall then be calculated separately by dividing it into many small elementary sections. The area of each elementary section is then calculated from the assumption of a revolution shape., taking into account the bottom edge of the flame not being horizontal.

The cross-sectional area, a_f, of the flame base shall be calculated from knowledge of the diameter d measured at the base of the flame as illustrated in 9.7. In that case, use Formula (2):

(2)

where

Measuring the burning velocity in a tube consists of

1. propagating a flame in a vertical transparent tube, opened at the lower ignition end, closed at the other upper end, and filled with the flammable mixture,
2. measuring the velocity of the flame propagating along the tube, and
3. recording the flame surface area with a camera.

Measurements are performed at atmospheric pressure.

The test bench layout is shown in Figure 2. The main elements of the bench are

• mixing vessel,
• ignition system,
• camera,
• test temperature control, and
• gas treatment systems.

NOTE To minimize pressure feedback effects, typically the scrubber system is not attached during the ignition and burn portion of the testing (seeFigure 8).

Key

Figure 2 — Schematic of the test bench

The gas mixture preparation is described in ISO 817:—, 6.1.3. If used, the scrubbing system described in 6.6 should be disconnected so that the expansion volume is not filled with a flammable concentration. The constant composition blend is then caused to flow through the tube until the gas mixture displaces at least thirteen times the air volume of the tube. Care should be taken to ensure that the gas mixture exiting the bottom of the tube is properly vented. Once the desired mixture has been achieved in the tube, the mixing vessel shall be isolated from the tube before ignition to prevent ignition of the gas in the vessel. It is good laboratory practice to measure the concentration of the gas mixture in the tube to ensure the methods employed adequately accomplish this objective. A paramagnetic oxygen analyser is effective for this determination.

It is recommended that all the components, connections and parts of the test bench be resistant to their use with corrosive gases, such as ammonia and copper, or other oxidation reactions. Stainless steel may be used, or any other material identified to be adequate for use with the substances to test.

The test tube shall be designed to ease the flame propagation with less possible disturbances, especially at the ignition level and the first stage of flame propagation; see Figure 3. The design of the test tube should look into the following points:

1. the ignition system, the quenching screen, and the damping orifice should be designed as close as possible to the outlet of the tube;
2. the outlet of the tube (at the lower end) should be designed to facilitate its connection to the extraction and gas treatment systems;
3. the tube should be fixed on a vertical support and at a level below the ignition system to prevent the fixing support from disturbing the flame propagation (excessive cooling) or any obstruction of the flame photography;
4. technical limitation with glass design and work should be considered as well.

Dimensions in millimetres

Key

Figure 3 — Test tube design and main dimensions

The tube shall be made of glass, 1,2 m long with a 40 mm internal diameter. The diameter has been chosen as a compromise between narrower tubes that increase the quenching effect but allow more stable propagation regimes, and larger tubes in which the losses to the walls are smaller but associated with an increase of instabilities^[1][4]. The choice of the 40 mm diameter has been shown^[5] to be the most convenient for measurement of burning velocities below 40 cm/s. It withstands a pressure of 100 kPa above the atmospheric pressure even if the overpressure is very limited, the bottom end of the tube being the open end.

NOTE Unstable regimes are frequent with fast propagating flames; see 9.3. The tube length is based on dimensions from previous research. Any great change in that length affects the flame propagation regimes and its stability only when working with high burning velocity compounds.

The tube should be placed in a vertical position to reduce possible deformations of the flame front from buoyant effect and to ensure a more symmetrical shape. In this position the flame propagates upwardly, the ignition occurring at the lower end of the tube.

The bottom end of the tube shall be open to the atmosphere. At this end are located the ignition system and the damping orifices. A GL45 cap can be used to maintain the system in place (see Figure 4 and Figure 5). With harmful components present in the combustion products (toxic or corrosive, e.g. HF, HCl, NH₃), the lower end should be connected to a gas post-treatment system (see 6.6). This design does not allow excessive pressure build-up and the combustion products can freely exit the tube or expand in a 125 l tank if the gas treatment system is used.

The upper end of the tube should be connected to the mixing vessel. The mixture flows out from mixing vessel into the tube and out of its bottom end. A GL45 cap shall be used to fix the inlet system. This end shall be closed before the ignition and until the end of flame propagation.

The flame propagation velocity and the flame shape vary with the type of flammable substance and the composition of its mixture with the oxidant. Adjusting the exit diameter at the lower open end by insertion of calibrated orifices helps stabilize the flame front shape by reducing the instabilities and damping the acoustic effects^[6][7][8][9] and therefore helps to reproduce a better shape of the flame front. The diameters of the damping orifices for a tube of 40 mm internal diameter vary from 9 mm to 11 mm (see Reference ^[9] for detailed calculation). The damping orifices are recommended with relatively high burning velocities (i.e. higher than 25 cm/s).

Dimensions in millimetres

Key

Figure 4 — Drawing of the lower end of the tube showing the ignition electrodes and the damping (fitting) orifice

Quenching screens shall be mounted at both ends of the tube, resistant to the reaction with HF and NH₃, to prevent any hazard to the surroundings. The quenching screens shall have a mesh size of 1 (+ 0,5 - 0,1) mm.

The spectral emissions of most flames are presumed to be in the range of 250 nm to 600 nm. To prevent excessive losses, it is important to compare the glass transmission profiles before selecting the type of glass (e.g. silica glass, borosilicate glass).

The test tube shall be purged by the mixture under test with a continuous flow from the mixing vessel with an equivalent volume flow rate which represents at least 13 times the internal tube volume. The gas mixture shall enter the upper end of the tube and exit from its lower end. The lower end may be closed after purging to avoid any possible concentration variation by dilution in the neighbourhood of the electrodes. This end is opened to the atmosphere just before ignition.

The presence of substances such as hydrogen fluoride (HF) or hydrogen chloride (HCl) with water residues in the combustion products of HFCs or HCFCs results in tube etching so that after several tests (30 to 50 depending on the cleaning process) the tube turns opaque with an almost white colour (see Figure 5).

For this reason, the tube shall be purged immediately after the end of the flame propagation with a stream of dry nitrogen. Afterwards, a wet wiper may be introduced inside the tube to clear all deposits on the inner wall. A stream of nitrogen may be again circulated inside the tube to remove water deposits from the wiper.

With this cleaning technique the same tube may be used for a larger number of tests before the etching effect becomes noticeable and the tube has to be discarded.

Figure 5 — Tube etching due to hydrogen fluoride

The ignition source can affect not only the flammability limit results but also the flame propagation regime. Analyses of spark ignition have been made by many researchers (Reference ^[10] gives a survey) and deal with the electrode arrangement, type (flange electrodes for instance), material and size, the electrode gap, the spark duration and the breakdown voltage as well as the effect of these on the minimum ignition energy.

The ignition system described in this test method has the same characteristics as the ignition system used in the ASTM E681 flammability test method in terms of the electrode dimensions, the gap distance, the ignition time and the power supply. This similarity helps to ensure that the vertical tube burning velocity method and complements the ASTM flammability method.

NOTE These ignition specifications are also very similar to those specified in DIN 51649-1 (which is meant by the flammability limits)^[9].

The mixture is ignited with an electrical spark produced by two electrodes.

The ignition occurs at the bottom end of the tube. The electrodes are fixed diametrically opposite on the tube, centred on its axis and positioned 5 mm to 10 mm above the upper surface of the interchangeable orifices. The electrodes shall be fixed using RIN 10/19 PTFE stoppers lodged in specially conceived RIN 10/19 housing (see Figure 4).

The electrodes are made of tungsten with 1 mm diameter. The gap between the electrodes is 6,4 mm. When necessary, a special calibrating cylinder can be inserted inside the tube and in-between the electrodes in order to verify their eccentricity and to ensure a correct gap distance.

To ensure good ignition conditions, especially near the lower and upper propagation limits, the electrodes shall be repeatedly cleaned of any deposit.

Power to the ignition electrodes shall be supplied by a transformer with an output of 15 kV, 30 mA. Usually, such high voltage is not required except with compounds having a high breakdown potential. The power supply system is connected to the electrodes using insulation rated for at least 15 kV to avoid short circuits and poor contacts avoiding overheating.

The ignition time shall be set at (0,3 ±0,05) s by adjusting the spark duration with a timer. This time duration has been proved to be the most appropriate for flammability limits measurements^[9].

Ignition should not be made immediately after filling the tube with the corresponding mixture, but 5 s to 10 s later, permitting the turbulence to cease in the tube.

NOTE The excessive energy release from this ignition system might be responsible for emitting waves inducing turbulence in the flame front and the mixture ahead of it. Flame propagation is not steady close to the ignition source.

Direct photography is used to record the flame front images. These images are used for the calculation of the flame propagation velocity as well as its surface area.

The burning velocity measurement Formula (1) is based on the calculation of the flame surface area at the preheat zone layer. With direct photography, the luminous zones of the flame are revealed. Therefore, any measurement made with this photography technique shall be based on the zone of the flame of most intense illumination. This zone corresponds to the region of the flame between the point whose temperature is equal to the ignition temperature and the point at the end of reaction (see Figure 6). The relative uncertainty in the burning velocity assessed with the flame surface area calculation based on flame profiles from direct photography is 6,5 %.

NOTE The 6,5 % relative uncertainty can be reduced and the correct surface position can be better approached if the profile of the outer edge of the luminous zone is shifted outwards by a distance equivalent to the luminous zone width.

Key

Figure 6 — Temperature profile along a combustion flame and luminous zone

The spectra peaks from combustion depend on the type of substance combusted and the radicals formed such as OH, HCO, CH, C₂ and C₃. From a qualitative point of view, it can be stated that the typical peaks for maximum emission, and even sometimes a higher-level continuum, are in the range of 250 nm to 600 nm for HC and HFC flames.

A digital camera shall be used to visualize the flame propagation. The flame front images shall be recorded and saved for further treatment (flame propagation velocity measurement and flame surface area calculation).

When identifying the camera to run the tests, the characteristics of exposure time and acquisition rate shall be selected as a function of the velocity range being measured. With very fast flames, a high acquisition rate and small exposure time are needed (i.e. <1 ms). The spectral response of the camera shall be also taken into account and the higher efficiency of the quantum efficiency curve shall cover the range of typical wavelength of the flames being visualized.

NOTE A set of adjustments and different operating modes, such as the resolution, image enhancements, image rate, exposure time, number of frames during record, pre-/post-trigger and parameters for image output, performed via an appropriate interface, can help in adapting the images to the type of flame front being recorded. A set of lenses may also be used to zoom and focus the optimized photography frame.

Setting the exposure time is necessary before starting the photography of the flame propagation to best reproduce the flame front shape and increase the precision of its area measurement.

Since there is no defined relationship between the flame propagation velocity and its more or less luminous aspect, for fast and low luminous flames the tester has to find a compromise for setting the exposure time. A higher exposure time compensates the low luminosity but results in an imprecise shape of the flame front due to its displacement during the exposure time.

For measurements around the stoichiometry, the recommended exposure times are of 1 ms or less. This value is determined by practical experience and depends on the camera.

The camera recording field shall be adjusted to the appropriate position and height of the tube where the flame movement is known to be uniform. Only images taken at the same level of the lens axis shall be used to calculate the flame surface area and reduce the imprecision on the flame front dimensions.

Scaling of the flame images to the real flame front dimensions can be achieved by taking a photo of a graduated ruler placed along the tube in order that the graduations coincide with a layer crossing the centre of the tube and at right angles to the camera axis.

NOTE The optical deformation due to the tube wall geometry is negligible.

A mirror at 45° is placed beside the tube in order to identify the irregularities of the flame surface and to increase the accuracy of the tests by ensuring the correct assumptions for the flame front area calculation. With the camera facing both the tube and the mirror, recorded images give both the front view and the side view of the flame front (see Figure 7). Note that the plane in which the mirror image is placed is located behind that of the direct photography. If used for calculation, these images shall be scaled to the same vertical layer crossing the centre of the tube and at right angles to the camera axis.

Key

Figure 7 — Schematic of the face and side view image as received by the camera

The essential uncertainty in the measurement of the burning velocity with the tube method is related to the image resolution for the flame propagation velocity, the scaling factor, and the flame surface area. An increase in the resolution provides more accurate results, but to the extent where the points fitting on the flame front image becomes independent of the pixels of small dimensions.

The test apparatus shall be cleaned thoroughly after each test to remove the remaining combustion products and effluents from the previous test and to make sure that the combustion products shall not harm the personnel or damage the environment. An appropriate treatment process of the combustion products, especially with fluorinated gases, shall be put in place: the extraction of the combustion products shall be performed rapidly at the end of the test in order to neutralize the HF or HCl, which in some cases constitutes more than 30 % of the combustion products, and reduce the corrosion of the test tube (a small amount of moisture makes it very corrosive). For this purpose, an ad hoc treatment system and exhaust gas clean-up is designed and installed at the outlet of the test apparatus, permitting the removal of the corrosive substances by splashing the exhaust gas into a basic water solution (e.g. NaOH). The treatment system also includes an expansion tank connected between the gas treatment system and the lower end of the tube to simulate a constant pressure expansion condition (see Figure 8).

An extraction fan evacuates the combustion products from the tube, through the expansion tank and basic water solution and then the clean gas to the hood.

If needed, a water scrubbing system may be used instead of a splashing system. In that case a cone nozzle may be used for example to spray water onto the upward exhaust gas flow. The acid water is collected at the bottom and drained into a larger tank filled with a basic water solution to neutralize the acid water.

The water solution tank should be emptied regularly, and the acidity of the water solution checked to guarantee safe handling.

NOTE If necessary, two treatment systems may be placed in series to achieve a more complete removal.

8 a) — With splashing

Key

8 b) — With scrubbing

Figure 8 — Gas treatment system

The test should be conducted at 23 °C. A test temperature of 23 °C may simply be achieved by controlling the temperature in the test room. Since the flame propagation is temperature sensitive, the temperature gradient along the tube should minimized (e.g. to less than 1 K). This may be achieved by an appropriate circulation and control of a temperature-controlled air stream.

The following protocol is applied for running the tests when the partial pressure technique is used to prepare the mixtures.

1. The following items shall be checked before starting a new test:
2. recording the reference scale of the camera to that of the real flame dimension (pixels/m, or equivalent);
3. clean the tube by dry gas purging (air or nitrogen);
4. check the electrodes gap distance and eccentricity;
5. select the appropriate exit orifice diameter.
6. The mixing vessel and all connecting pipes and tubes shall be first evacuated to a pressure of 10 Pa abs or less.
7. The mixing vessel shall be then filled with the different mixture components, each at its corresponding partial pressure. The connections shall be evacuated each time a new gas is introduced into the mixing vessel. The magnetic stirrer should be turned on at the start of the process and for at least 5 min after the end of the filling process.
8. The mixture may be then allowed to leave the mixing vessel, circulate through the tube, out from its lower exit end and to the extraction hood. An equivalent volume of at least 13 times the internal volume of the tube is circulated.
9. The upper end of the vertical tube shall be closed first and then the lower end immediately after it to prevent any possible dilution or concentration changing within the electrodes region.
10. 5 s to 10 s should be given for the mixture inside the vertical tube to become quiescent.
11. The lower end shall be opened gently to avoid any perturbation or concentration changes around the ignition region, and then ignition is made. Just before ignition, the camera should be activated and the images are recorded.
12. After the end of the flame propagation, quenched at the upper end of the tube, the combustion products are driven out by a stream of nitrogen or air circulated inside the tube. In the case of harmful combustion products, the gas treatment system shall be installed and used.

Test burning velocity is validated by Formula (3):

(3)

where

In a few cases, Formula (3) does not achieve the best fit to the burning velocity experimental results. Other adequate fitting equations should then be elaborated. The experimental points may also be split into two parts and for each, a separate fitting equation may be used.

NOTE In general, the maximum burning velocity is met at an equivalence ratio between 1,00 and 1,15.

The total relative uncertainty of the burning velocity measurements as described in this International Standard is estimated between 7 % and 10 % and is due.

1. primarily to uncertainties in the flame surface area calculation (65 % of the total uncertainty), and
2. the flame propagation speed measurement (35 % of the total uncertainty).

NOTE As described by Takizawa et al.,^[2][3] the method is not usable for burning velocities below 4 cm/s because the rising, hot, burned gas bubble controls the flame propagation rate, not the burning velocity of the mixture.

The concentrations of the mixtures as prepared with the partial pressure method are subjected to an uncertainty arising mainly from:

1. pressure transducer measurement;
2. ideal gas law used to derive the densities from the pressure and temperature. With air being the major component in the mixtures, the mixture state at pressures between 300 kPa and 400 kPa abs is not very far from the ideal state. In that case the relative uncertainty in the density is evaluated to 2 % whereas it may be neglected for pressures below 100 kPa abs.

NOTE The greater part of the uncertainty in the concentration arises from the ideal gas law assumption. For concentrations as high as 30 % volume fraction, the absolute uncertainty may be estimated at 0,6 % volume fraction or 2 % relative. For concentrations as low as 2 % volume fraction, the absolute uncertainty may be estimated at 0,08 % volume fraction and the relative uncertainty at 4 %.

The uncertainties in the burning velocity measurements and the mixture concentrations shall be determined specifically for each test bench.

8.1 The safety recommendations are made for a proportion of oxygen in the air no higher than 21 % volume fraction.

8.2 For high burning velocities (>30 cm/s), it is recommended to make the tests starting from the lowest LFL of the components and gradually increasing the concentration. This is necessary to avoid explosions with fast propagating flames near the stoichiometric concentration. An excessive restriction of the tube exit end with the interchangeable orifices shall be avoided.

8.3 Appropriate personal protective equipment (PPE) shall be employed by the test bench user (e.g. gloves, eye and head protection, etc.). Best practices suggest having personnel training in potential system hazards and HF safety kits available. HF compatible PPE should be used following safety program protocols.

8.4 The ad hoc gas treatment system shall be manipulated with caution because of its content of corrosive substances. The treatment system shall be sealed, and an extraction fan should be used to avoid any inhalation of the combustion products.

8.5 The high voltage ignition system, connections and electrodes should be handled with care and protected to prevent any direct contact. For safety reasons, potential ignition sources other than that intended for the testing should be avoided (e.g. switches, electrical contacts).

8.6 Quenching screens shall be mounted on both ends of the test tube. An additional quenching screen shall be placed at the entrance to the expansion volume to prevent any ignition hazard within that volume.

When considering a combustion wave propagating from the open to the closed end of the tube, the unburned gas ahead of the wave is contained by the tube wall so that it forms a stationary column. The thermal expansion within the wave generates a continuous flow of burned gas towards the open end. The main parameters contributing to the establishment of the flame front shape are the following:

1. viscous drag responsible for flow retarding at the wall and acceleration in the centre of the tube (higher pressure thrust in the centre than near the wall);
2. unburned gas flow away from the flame front at the centre and towards it at the edges (the unburned gas ahead of the flame is pushed towards a closed end);
3. convection effect of burned gas resulting in an elongated front with slower propagating flames whereas the front takes an almost spherical shape with fast propagating flames;
4. constant burning rate in a direction normal to the flame front.

The balance between all of the above-mentioned effects seems to be requirements for maintaining the stability of the flame front during uniform movement (see Figure 9).

Key

Figure 9 — Direction of flow and particle velocity for laminar combustion wave propagation from open to closed end of the tube

With high burning velocities, the propagation usually develops in three distinct stages or regimes with two possible types of movements. The three stages can be distinguished either by flame structure or the amplitude of pressure and flame oscillations as well. The flame movement consists of one of the two following movements:

1. uniform movement;
2. vibratory movement.

The regime of flame propagation can develop into the following respective stages.

• After ignition, the flame propagates smoothly across the first part (first stage) of the tube at a uniform velocity which depends on the mixture and the tube length.
• An oscillatory motion of the flame can superimpose after the first stage. These oscillations begin with the appearance of a cellular structure in the flame front. The onset of oscillation and vibrations during propagation is the result of a coupling between flame and pressure oscillations in the gas.
• As the flame progresses, the tube becomes increasingly filled by hot combustion products and hence the basic frequency of the oscillation rises. The flame front can be subject to violent reciprocating motions and would accelerate steadily until the appearance of the turbulent propagation regime, leading to a large instability level, which persists until combustion is complete.

The termination of the uniform movement by a vibratory motion of flame always occurs when ignition is done at the opened end of the tube, the only exception being with slow burning mixtures, in which combustion can proceed at a uniform rate over a major part of the tube length.

NOTE 1 Measurements in a 40 mm tube of burning velocities below 23 cm/s^[9] showed almost no flame acceleration. Compounds having burning velocities below 10 cm/s have been witnessed to propagate with velocities not exceeding 25 cm/s.

NOTE 2 Cellular form of flames is associated with high propagation velocity flames (high burning velocities) and rich mixtures. The flame surface area of cellular flame cannot be measured with acceptable precision. For this reason, the tube method is limited to comparatively low burning velocity measurements.

The main and possible reasons behind turbulence generation should be identified in order to understand particular behaviours of flame propagation. In general, the turbulence of a premixed flame can be attributed to one of the following reasons:

1. the initial gas flow turbulence in the flammable mixture can induce disturbances in the flame front shape and behaviour;
2. non-uniformity of the flammable mixture concentration, pressure, and temperature. In that case the flame propagates in a mixture in which conditions are changing and thus its propagation is affected, too;
3. gas flow turbulence in the flammable mixture at the shear-flow region between the wall or the obstacles and the gas flow induced by propagation of the flame;
4. in an accelerating flow field, gas flow turbulence is generated near the flame front;
5. flame front disturbances caused by thermal and mass diffusion mechanisms;
6. interaction of the flame front with the acoustic waves emitted by the flame propagation.

The stage of uniform movement of the flame can start when the disturbances in the unburned mixture are damped and the flame propagates in a medium at rest.

Particular attention should be given to acoustic instabilities identified with fast propagating flames. As a potential turbulence generator, acoustic instabilities turn out to be the most difficult to eliminate but can be controlled by a proper design. Damping of acoustic vibrations can be achieved by inducing a resistance to the flow of burned gas to increase the corresponding pressure above a certain level^[8]. This may be done by reducing the opened end of the tube by means of an orifice. This reduction will increase the uniform movement of the flame.

The acoustic instabilities can hence be attenuated by reducing the opened end of the tube by means of an orifice whose diameter is defined by Formula (4):

(4)

where

This operation is necessary only with fast propagating flames, as experienced during experimental tests, and high burning velocities. However, the acoustic damping can be insufficient to eliminate the fast flame disturbances in its initial propagation phase. The disturbances in the unburned mixture ahead of the flame shall be prevented as well. This may be achieved by placing fibreglass pads some distance from the closed downstream end of the tube.

1. The fastest flames lose their uniform motion after some distance, and the slower flames travel a longer distance before the flame front loses its regular shape. The motion becomes non-uniform because of a vibration motion, which is sometimes so severe in the slower burning mixtures that the flame is extinguished.
2. The faster flames remain upright until the non-uniform motion. The slower flames are sometimes slanting during the uniform motion. The intermediate velocity flames (particularly on the rich side of the concentrations) seem to be able to adopt either an upright or a slanting form, or the flame can change gradually or abruptly from the upright to the slanting form. The point where this change occurs seems to be quite arbitrary, and some flames which can adopt the slanting form remain upright until the onset of the non-uniform motion. The slanting form seems to be the more stable, since the change never occurs in the reverse direction. Both forms of the motion seem to be quite uniform, with the result that the flame can have first a slow uniform velocity, followed by a faster uniform velocity, before the non-uniform motion starts.
3. Some flames can show some unusual behaviour from all the others corresponding mixtures. Those flames proceed rapidly, upright and emit considerably more light than the other observed flames, while expected to be rather slow and probably to adopt a slanting form. It is possible that this anomaly is connected in some way with the change in reaction mechanism noticed for some mixture compositions. The general structure of the flame might also appear to be different in some of these cases. There is no explanation for such exceptional flames, but it does seem that slight conditions can cause a marked change in flame velocity, and perhaps even in the mode of flame propagation in tubes.

When a flame is propagating near a surface, effects on the flame configuration shall not be neglected. The flame appears to be quenched and to be standing a distance of few millimetres away from the wall. This distance is called the quenching distance. This quenching phenomenon is governed by three different mechanisms that can be summarized in the following:

1. heat conduction from the flame to the wall reduces the energy available to preheat the gas ahead of the flame;
2. absorption, by the wall, of active chemical species that are important in the chain of chemical reactions that create a continuous propagation;
3. viscous effects at the wall.

In most cases the first mechanism seems to be the most significant. This unavoidable effect of flame cooling by the tube wall lowers the apparent burning velocity obtained by this method below its value in free space.

The loss of heat to the wall cannot be eliminated entirely, but it is confined to the region behind the flame where high temperatures, high gas velocities and disruption of the boundary layer on the tube walls by the passage of the flame serve to increase the rate of heat transfer between the wall and the burned products. If the tube is closed at both ends, the heat loss would be noticed by a decrease in pressure, but in the current measurements the pressure is maintained by the surrounding atmosphere, and hence the cooling of the burned products has no influence on the progress of the flame.

One effect of this layer of unburned gas on the wall surface is to reduce the effective cross-sectional area of the tube. This reduction shall be taken into account for the correct measurement of the burning velocity, otherwise it will introduce an error. The level of this error is proportional to the ratio of tube diameter to flame area as can be deduced from Formula (1). It is therefore desirable to use a tube of a cross-sectional area as large as possible. But in practice, a critical size exists above which the flame surface becomes distorted by convection effects in the burnt products. Consequently, for high velocity measurements, the tube diameter should not be larger than a certain critical diameter to avoid turbulent flame fronts. For slow burning mixtures, the tube diameter should not be less than a critical diameter to avoid excessive surface effects. A 40 mm internal diameter can achieve this compromise (see 6.3.2).

The flame velocity of uniform propagation in a tube increases with increasing tube diameter.

Based on experimental results, Formula (5) is derived to estimate the flame propagation velocity at a level of burning velocity below 50 cm/s measured in a tube of diameter d between 25 mm and 60 mm.

(5)

where

An accurate method is needed to calculate the flame surface area, A_f. Even though in many cases the flame front shape is symmetrical, this shape cannot be generated by the revolution of a parabola nor by approximation by an ellipsoid segment. For an upward propagation, the flame shows a symmetrical front surface referred to the tube axis. For a uniform propagation, the shape of the flame front remains constant. Fast moving flames are almost hemispherical, and the slower flames are somewhat elongated.

To measure A_f, the flame front profile shall be marked with fitting points (20 to 40 fitting points) then divided into two or more horizontal sections. The fitting points shall be selected on the rim of the most luminous zone on the flame front (the outer layer towards the unburned gas; see 6.5.2). For each section a polynomial fit equation of appropriate order is made in order to give the best fit curve to the points selected on that section. The best fit is the one that gives the minimum deviation of the fit curve from the fitting points. Usually, polynomial equations of the 6th order and lower are sufficient to achieve the exact fitting. The area of each section is then calculated separately by dividing it into a large number of small elementary sections. The area of each elementary section is then calculated from the assumption of a revolution shape (a representation is illustrated in 10 a)).

Since the lower edges of the flame front are not necessarily at the same level, the case of tipped flames, a special mathematical and geometrical means taking into account the level difference between the two edges are used to estimate the area of that part of the flame front (a representation is illustrated in 10 b)). Tests with very tipped flames should not be used in the measurement of the flame area. Tipped flames are mostly seen on the rich side of the concentrations and rarely around the stoichiometry or even at the maximum burning velocity. In that case more than one measurement is repeated in order to have a better representation of the correct value. Usually, the measurement that represents more than 5 % to 10 % difference from the average should not be considered.

Key

Figure 10 — Elementary non-tipped and tipped section in flame surface area calculation

A_f is obtained by summing all the surface areas of the sections as per Formula (6):

(6)

With a non-tipped section, A_i is given by 10 a) Formula (7):

(7)

With a tipped section, A_i can be estimated by 10 b) Formula (8):

(8)

where M_i are the segment midpoints.

Bibliography

[1] Jabbour, T. Flammable Refrigerant Classification Based on the Burning Velocity, Thesis, Ecole des Mines de Paris, France, June 2004

[2] Takizawa, K., Igarashi, N., Tokuhashi, K., Kondo, S., Mamiya, M., Nagai, H. ASHRAE, Assessment of Burning Velocity Test Methods for Mildly Flammable Refrigerants, Part 2: Vertical-Tube Method, in: ASHRAE Transactions 2013, Vol 119, Pt 2, 119, 2013, 255–264

[3] Takizawa, K., Igarashi, N., Tokuhashi, K., Kondo, S., Mamiya, M., Nagai, H. ASHRAE Research Project Report RP-1583, ASHRAE, Atlanta, GA, 2017, 74

[4] Guénoche H. Laffitte P. Comptes Rendus de l’Académie des Sciences 222. 12 Juin 1946, 1394–1396

[5] Guénoche H., Manson N., Monnot G. Comptes Rendus de l’Académie des Sciences 226. 5 Jan. 1948, 69–71

[6] Guénoche H., Manson N., Monnot G. Comptes Rendus de l’Académie des Sciences 226. 12 Jan. 1948, 163–164

[7] Guénoche H. Non-steady flame propagation (Ed. G. H. Markstein), Chapter E. AGARD, Pergamon, 1964

[8] Babrauskas V. Ignition Handbook. Fire Science Publishers, 2003

[9]

[10] Kondo S., Urano Y., Takahashi A., Tokuhashi K. Reinvestigation of flammability limits measurement of methane by the conventional vessel method with AC discharge ignition. Combust. Sci. Technol. 1999, 145, 1–15

1. 1) Under preparation. Stage at the time of publication: ISO/DIS 817:2023. ↑