MHG Plastic Industries Pty Limited v Australian Competition and Consumer Commission

Federal Court of Australia|2000-08-09|Before: Mansfield JJ

1 This is an appeal from the decision of a judge of the Court who declared that three models of motorcycle helmets manufactured by the appellant since 1 July 1996 - nearly 51,000 in all - do not comply with Australian Standard AS 1698-1988 "Protective Helmets for Vehicle Users" in that they do not satisfy certain of the performance requirements specified in clauses 6.1 and 6.3 of that Standard. His Honour further declared that the appellant, by supplying those helmets to purchasers after 1 July 1996, contravened s 65C(1) the Trade Practices Act 1974 (Cth). In pronouncing judgment, his Honour made a number of orders consequential upon these two declarations, including orders that obliged the appellant to give public notice of the Court's determination and to refund to purchasers of the helmets the price they paid for them and to use its best endeavours to recover all the helmets distributed by it. 2 The Australian Competition and Consumer Commission's case, as pleaded, was that the appellant contravened s 65C the Trade Practices Act 1974 (Cth) because the helmets sold by it did not comply with AS 1698-1988 in respect of "resistance to penetration" and "strength of retention system". The learned primary judge held that the Commission made out its case in respect of resistance to penetration. 3 Since 1986, MHG has produced three models of helmet, an open-faced EXR model in six sizes; an MXR model in six sizes identical to the EXR but with a chin guard and a closed-face RXR model in five sizes. All EXR, MXR and RXR models have the same size outer shell, though the RXR model is of a different shape from that of the EXR and MXR models. The variations in sizes of all helmets are achieved by varying the thickness of the polystyrene helmet liner and the comfort padding. 4 The appellant has long operated its own testing facility in which it has routinely tested many samples of the helmets manufactured by it for compliance with AS 1698-1988 by way of quality control of its products. Its testing facility has at all relevant times been accredited by the National Association of Testing Authorities Australia ("NATA") which conducts annual reviews of the facility. It has also long held a licence from Quality Assurance Services Pty Ltd (QAS), an arm of the Standards Association of Australia, to mark its helmets to show they comply with the QAS testing regime. (This includes, but is not limited to testing in accordance with this Australian product standard. Under the QAS regime, there is provision for "type" testing of new models of helmet for initial certification of the new model, batch testing of samples of production runs of particular models of helmet and annual audit testing of the manufacturer's own batch testing laboratory). 5 In early 1999 the respondent had a number of the appellant's helmets tested by Crashlab, a division of the Roads and Traffic Authority of New South Wales, which the respondent supplied to Crashlab. (This series of tests was not performed as part of the type or batch or audit testing processes conducted on MHG helmets under the QAS "StandardsMark" licensing scheme). 6 Penetration tests carried out by MHG over the years were performed only at two sites on each helmet it tested, viz, site E and site F, which are both located on the centre line of the crown of the helmet. MHG used a size C headform in all its testing. There were no failures though many hundreds of helmets were tested. Over the years, Crashlab has performed many tests on helmets at these same two sites. But in the case of some of the MHG helmets it tested in early 1999 at the request of the respondent, it also tested at two other sites, G (on the left front of the helmet) and H (on the right rear). The evidence was that until it performed the tests on the MHG helmets off the centre line of the crown in March 1999, Crashlab, like MHG, had only ever previously tested helmets along the centre line. The learned primary judge said that there was no evidence to explain what prompted Crashlab to depart from centre line testing in March 1999. 7 As the Crashlab test report of 3 March 1999 (TR99/135) notes, four EXR model helmets in size M were tested at sites E and F on a size B headform (in respect of resistance to penetration) (which is smaller than the size C headform used by MHG and previously, by Crashlab). All helmets passed, save for one, which failed at site F, and for another, which failed at both sites E and F. As the test report of 19 March 1999 (TR99/146) records, four helmets of the same size and model were similarly tested, again on a B headform (in respect of resistance to penetration). All passed, save for one, which failed at site F. As is recorded in the first test report of 23 March 1999 (TR99/147), four helmets of the EXR model size L were all tested at sites E and F, with one of them alone being tested at site G. All were tested on a size C headform. All passed, save that the one tested at site G failed there. His Honour expressed no criticism as to the reliability of any of these three lots of tests. In the second lot of tests performed on 23 March 1999 (TR99/148) on four RXR model helmets, all size L, all four were tested at sites E and F, while two were also tested at site G. Again a C headform was used. All passed, save that one of the two tested at site G failed at that location. His Honour regarded this series of tests as "not entirely satisfactory". In the series of tests performed on 23 and 27 April 1999 (TR99/189), tests were performed, firstly, on four EXR models, all size XS, this time using a size B headform. Each was tested at sites E, F, G and H; all four helmets failed at site G but otherwise passed these tests. (It was upon this series of tests that his Honour based his decision.) Tests were next performed on two EXR models size L. They were tested only at sites E and F with a C headform, and both helmets passed all tests. Four EXR models size XXL were next tested only at sites E and F with C headform; all passed at site F, but three of the four failed at site E and no result was obtained in respect of the tested site E for the fourth helmet. His Honour did not regard this series of tests as satisfactory. Next, one MXR model size XXL was tested at sites E, F and G on a C headform, passing at the two former sites but failing at the last site. Next, one MXR model size XL was tested only at sites E and F on a C headform and passed. Finally, two MXR models size S were tested only at sites E and F on a B headform and passed. His Honour did not comment on whether he found these tests on the MXR helmets satisfactory. But it is apparent from his reasons, to which we will refer later, that he placed no reliance on them in concluding that all MXR helmets "would fail the penetration test applied at site G". 8 His Honour concluded that helmets of the three models in question (EXR, MXR and RXR) "satisfy the performance requirements as to resistance to penetration when tested at sites E and F" because he considered "on the balance of probabilities, that each of the models will not fail the penetration test when applied at site E or site F". His Honour stated this conclusion even though (as Mr Simmons noted in his report), seven of the sixty-four penetration tests done by Crashlab in early 1999 at sites E and F failed. It appears that three of these failures occurred in tests which his Honour accepted were properly performed, while the other four occurred in the tests on the four EXR size XXL helmets on 23 and 27 April 1999, which his Honour was not prepared to accept as reliable. Three failures at sites E and F out of a total of sixty-four tests, coupled with the long history of tests by MHG itself at these same two sites, without any failures, appears to have led his Honour to the conclusion that each of the three models satisfied the performance requirements to resistance to penetration when tested at sites E and F. 9 His Honour plainly took the view that a single failure or even a small number of failures on penetration testing was not necessarily of itself sufficient to establish that a helmet of the type tested did not comply with the Product Standard, AS 1698-1988. He said:

"[53] The testing of any particular helmet results in destruction of the helmet. In order to determine whether a helmet which is offered for sale … does not comply with a standard such as the Standard [AS 1698-1988], it is necessary to make an extrapolation from the results of tests conducted on other helmets. There is an underlying assumption in a provision such as s 65C(1)(a) that every helmet manufactured in accordance with a particular specification will perform in the same way as, or will have the same performance characteristics as, all other helmets manufactured in accordance with the same specification."

"[58] The Commission has the onus of establishing, on the balance of probabilities, that any particular helmet that MHG threatens to supply in trade or commerce does not comply with the Standard. That question can only be answered by an extrapolation from the test results. The question is whether, on the basis of the evidence, including the test results of a particular model, it is more likely than not that a given helmet will fail the tests in the Standard."

11 Product Standard AS 1698-1988 and the associated Standards (in contrast to the QAS test regime) do not require the random selection of helmets to be tested from a defined population of helmets; nor do they provide for the evaluation of test results by reference to statistical techniques. The respondent did not contend that a single test failure necessarily meant that the particular model did not conform to AS 1698-1988. It was alert to the problems thrown up by the limited nature of the testing scheme provided for by AS 1698-1988. In pleading that the appellant's helmets failed to comply with this Standard it gave the following particulars:

"A. Clause 6.3 of the Australian Standard requires that when the helmet is tested in accordance with AS 2512.4 there shall be no contact between the striker and the surface of the test headform ('the resistance to penetration test').

B. (i) On the proper construction of the Australian Standard any statistically significant failure of a sampled product to pass the resistance to penetration test constitutes non-compliance of the product with clause 6.3; and

(ii) a statistically significant number of helmets in each Model fail the resistance to penetration test.

C. Alternatively, the motor cycle helmets do not comply with clause 6.3 of the Australian Standard in that it is more likely than not that each helmet in each Model will fail the resistance to penetration test."

12 The respondent succeeded at trial on its alternative case pleaded in paragraph C. It is understandable that his Honour took the view that the question whether a helmet of a particular model does not conform to the product standard involves a matter of judgment as to whether the available test results show that it is more likely than not that all examples of the particular model will meet the product standard. The testing scheme in AS 1698-1988 is not based on statistical techniques. His Honour's approach makes an allowance, necessarily rough, for the fact that one (or a small number) of tests may fail, but that should not automatically require the conclusion that the whole population of helmets, of which only a relatively small number may be tested, should be condemned as not complying with the Standard. 13 However, his Honour, after observing that the evidence as to the application of the penetration test at site G was "not conclusive", concluded that "it is more likely than not that size XS of the EXR model will fail the penetration test applied at site G". The only tests done by Crashlab on helmets of that size and model are those recorded in report TR99/189: all four of these helmets failed the penetration test at site G. 14 How his Honour went on to reach his conclusion that all other sizes of the EXR model and all sizes of the other two models are likely to fail the penetration test when applied at site G appears from par [86] and following of his reasons. Having found that size XS of EXR model will fail the test at site G, his Honour continued:

"[86] Since the EXR model and the MXR model have identical shells, I am also satisfied that the XS size of the MXR model is also more likely than not to fail the penetration test applied at site G. Further, in the absence of any evidence indicating that the thickness of the liner has any effect on the penetration test, I am satisfied that, on the balance of probabilities, all sizes of models EXR and MXR would fail the penetration test applied at site G.

[87] The conduct of the penetration test at site G in relation to RXR model helmets is not entirely satisfactory. As I have indicated, the helmets that failed that test appear to have been subjected to the penetration test at sites E and F beforehand. There are only two helmets that have been subjected to the tests. One had been conditioned to ambient temperature and one had been conditioned to high temperature.

[88] While the design of the outer shell of the RXR model differs from that of the EXR and MXR models, they are all manufactured from the same substance. There is no evidence to indicate that the different shape has any bearing on the conduct of the penetration tests at site G. Accordingly, I am satisfied, on the balance of probabilities, that the RXR model would also fail the penetration test applied at site G."

15 The only test results his Honour relied on to hold that all three helmet models, EXR, MXR and RXR, "would fail the penetration test applied at site 'G'" were the results in report TR99/189 for the four EXR helmets, size XS, which were tested on a B headform and which all failed at site G. His Honour did not place any reliance, in reaching his ultimate conclusion with respect to all sizes of all three models, on the one failure at site G of the EXR helmets in report TR99/147: it was a size L. Nor did he rely on the failure at site G of the one MXR helmet tested at that site in report TR99/189. This is so even though both helmets were tested on a C headform and he expressed no criticism of the tests. His Honour also noted that clause 7 of the Product Standard AS 1698-1988 required "at least four helmets of the same size shall be submitted for test" and that penetration test sites must be "at least 76 mm apart" and that "at least two penetration sites shall be tested", requirements that were not satisfied by some of the test series. This may explain why his Honour did not place any reliance on certain of the Crashlab tests.

16 The first challenge to the learned primary judge's decision contained in grounds 1 to 3 of the Amended Notice of Appeal raises, as counsel explained, two questions: firstly, whether the testing apparatus used by Crashlab on which it mounted the test helmet when testing at site G satisfied the requirements of the Standard. Secondly, it was said that the learned primary judge erred in his construction of the phrases "rigidly mounted test headform" and "rigid mount" in the Standard and erred in concluding that even if Crashlab's cantilever arm testing apparatus otherwise satisfied the requirements of the Standard, it did not constitute (or incorporate) a "rigid mount" as required by the Standard. 17 When testing on the crown of the helmet (at sites E and F), Crashlab, like MHG, used a test rig which involved positioning the helmet with its sides vertical on the test headform, which was in turn mounted on a column fixed to a solid base. But when Crashlab tested at site G (and site H), it used a test mount on which the helmet was held on a horizontal arm that was fixed to a vertical post, fixed in turn to a solid base, with the sides of the helmet parallel to the ground. It was said that the Standard only permitted the use of a vertically standing columnar mount on which the helmet is held crown uppermost with its sides vertical to the ground. As to the second point, the appellant contended that the rigidity requirement could only be satisfied if the cantilever arm met a particular quantitative resonance criterion for which it had not been evaluated. 18 The next challenge, contained in ground 5, raises the question whether the learned primary judge was right in concluding that there was no evidence, at the time of promulgation of the Standard, of an industry practice of limiting the test points, when testing for resistance to penetration, to points along the centre line of the crown of the helmets. It was said that such was then in fact the industry practice and that the Standard should accordingly be read as permitting resistance of helmets to penetration testing only in that limited area. 19 The next challenge contained in grounds 6 and 7 of the Amended Notice of Appeal was based on Crashlab's failure to produce reports of its tests resulting in penetration resistance failures in the form prescribed by one of the Standards incorporated by reference in AS 1698-1988. The challenge contained in ground 8 of the Amended Notice of Appeal turns on the fact that Crashlab tested the four EXR size XS helmets that failed the resistance to penetration test the subject of report TR99/189 on a size B test headform, when they should have been tested on a larger size C test headform. The final challenge contained in ground 10 of the Amended Notice of Appeal attacks the use by the learned primary judge of the fail results for the EXR model in report TR99/189 as a proper basis for his finding that the RXR model helmet failed to meet the relevant provisions of the Standard. 20 The case was presented in this Court on the basis that though the appeal challenges findings of fact made by the learned primary judge, they are ultimate findings based on primary facts not in dispute and not based on findings as to credibility.

21 Section 65C(1)(a) the Trade Practices Act 1974 (Cth) relevantly provides:

"(1) A corporation shall not, in trade or commerce, supply goods that are intended to be used, or are of a kind likely to be used, by a consumer if the goods are of a kind:

(a) in respect of which there is a prescribed consumer product safety standard and which do not comply with that standard;

22 The case, including the appeal, was fought on the basis that the helmets are goods within s 65C the Trade Practices Act 1974 (Cth) and that AS 1698-1988 is a safety standard prescribed for those helmets. (Pursuant to s 65E, by Consumer Protection Notice No 9 of 10 December 1990, the Minister declared that, in respect of protective helmets for motor cyclists, AS 1698-1988 is a Consumer Product Safety Standard for the purposes of s 65C.) 23 The following provisions of AS 1698-1988 are relevant:

"1 SCOPE. This Standard specifies requirements for protective headgear for vehicle users, as designed to mitigate the adverse effect of a blow on the head. The Standard is written with particular reference to motor cyclists, but is equally applicable to users of other types of vehicle.

2 REFERENCED DOCUMENTS. The documents below are referred to in this Standard.

Method 2: General requirements for the conditioning and preparation of test specimens and laboratory conditions (AS 2512.2)

Method 4: Determination of penetration resistance (AS 2512.4)

3 DEFINITIONS. For the purpose of this Standard, the definitions given in AS 2512.1 apply.

4.1 General. The helmet shall consist of a shell with a hard smooth outer surface capable of resisting penetration, a means of absorbing impact energy, and a retention system.

5 MATERIALS. Except as specifically provided for herein, the characteristics of the materials used in the manufacture of helmets shall be established by the manufacturer as being suitable for the purpose, having regard to the provisions of Appendix A.

6.1 General. The tests specified in Clauses 6.2 and 6.3 shall be applied at any points above the test line, other than specifically excluded within that Clause.

This sequence ensures that tests which require the helmet to be intact are carried out before any cutting of the helmet, as such alteration may reduce the rigidity of the helmet.

To enable the tests to be correctly applied, it is necessary for the helmet manufacturer to specify the helmet positioning index (see Clause 10). It is a requirement of AS 2512.5 that the helmet positioning index be known and applied by the tester, and recorded on the test report.

It is expected that impact tests will be specified for regions below the basic plane and in the chin region at a future date.

Conditioning, including conditioning by solvent, and preparation of test specimens for testing shall be in accordance with AS 2512.2.

6.3 Resistance to penetration. When the helmet is tested in accordance with AS 2512.4, there shall be no contact between the striker and the surface of the test headform.

The penetration test site(s) shall be at a point above the test line but not on a fastener or other rigid projection. Sites shall be at least 76 mm apart, and at least 76 mm from the centres of any impacts applied during the impact energy attenuation test. At least two penetration sites shall be tested.

The height of the guided free fall shall be 3000 ±15 mm.

7 TYPE TESTING. At least four helmets of the same size shall be submitted for test. The helmets shall be in the condition in which they are offered for sale …

NOTE: The certifying body may waive tests on some helmets within a range of helmets on the basis of engineering evaluation.

9 LABELLING. … each helmet shall be accompanied by an informative brochure or label which shall include the following information:

(f) The helmet is designed to absorb shock by partial destruction of the shell and liner. …

10 INFORMATION TO BE SUPPLIED BY MANUFACTURER. The helmet positioning index shall be given by the manufacturer to any person requesting that information, for a helmet identified by at least the detail of manufacturer, model designation, and size.

NOTE: This information is necessary for test purposes, see Clause 6.1."

24 Clause 7 of AS 1698-1988, which requires "at least four helmets of the same size" to be submitted for testing, is, we think, limited to "type" or initial certification testing, for bodies such as QAS, of new models of helmet by a testing laboratory independent of the helmet manufacturer. The Note to this clause suggests as much. Mr Wainohu of Crashlab explained what is described in this Note as "engineering evaluation" as well as "type testing" and "batch testing". Type testing is to be contrasted with batch testing of helmets taken from a particular production run which, under the QAS scheme, can be performed by the manufacturer in its own laboratory, if it is suitably accredited (as was MHG's). But clause 2 of AS 2512.2 shows that when helmets are tested for compliance with the performance requirements of AS 1698-1988, four samples must be tested, with one being conditioned to be at ambient temperature, another to be at low temperature, a third to be at high temperature and another by having been immersed for a period in water. We think these four samples must be of the same model and size: that is the requirement of clause 7 for type testing and we can see no justification for relaxing that requirement when testing to AS 1698-1988 for other purposes. Clause 4 of AS 2512.2 and clause 6 of AS 1698-1988 show that the intent of the latter Standard is that the same four test helmets are to be subjected, in turn, to each of the tests in clauses 6.2, 6.3, 6.4 and 6.5 of that Standard. 25 The following provisions of AS 2512.1 referred to in clause 2 of the main Standard are also relevant:

"1 SCOPE. This standard sets out general information relating to the methods of testing protective helmets.

Terms are defined and the dimensions and contours of reference and test headforms are given.

2 GENERAL PRINCIPLES. The tests relate to the capacity of the helmet to absorb impact energy and to resist penetration, and to the strength of the retention system. A number of shock attenuation tests cover the diverse activities for which head protection is required.

3 DEFINITIONS. For the purposes of these tests, the following definitions apply:

3.1 Protective helmet - a device worn on the head, designed to mitigate the adverse effects of a blow to the head within a specified area. Hereinafter referred to as a 'helmet'.

3.9 Reference headform - a measuring device contoured to the dimensions shown in Figs 4, 5, 6 and 7 with surface markings indicating the locations of the basic, mid-sagittal and reference planes, and the centres of the external ear openings.

3.10 Test headform - a test device contoured to the dimensions shown in Figs 4, 5, 6 and 7 for all surface areas that contact the helmet, with surface markings indicating the locations of the basic, mid-sagittal and reference planes.

3.11 Helmet positioning index - the distance, as specified by the manufacturer, from the lowest point of the brow opening at the lateral midpoint of the helmet to the basic plane of a reference headform, when the helmet is firmly and properly positioned on the reference headform.

3.12 Test line - a line drawn on the outer surface of a helmet coinciding with portions of the intersection of that surface with the following planes, as shown in Fig. 2:

(a) A plane 25 mm above and parallel to the reference plane in the anterior portion of the reference headform.

(b) A vertical transverse plane 65 mm behind the point on the anterior surface of the reference headform at the intersection of the mid-sagittal and reference planes.

(d) A vertical transverse plane 65 mm behind the centre of the external ear opening in a side view.

(e) A plane 25 mm below and parallel to the reference plane in the posterior portion of the reference headform.

4 HEADFORMS. The reference and test headforms shall conform nominally to the dimensions and contours given in Figs 4, 5, 6 and 7."

26 The following clauses of AS 2512.2 are also relevant:

"1 SCOPE. This standard sets out conditioning procedures for protective helmets that are to be tested, the way in which the helmets are prepared for test, and the laboratory conditions for the test.

3 DETERMINATION OF THE TEST LINE OF THE HELMET. The test line of the helmet shall be determined as follows:

(a) Place the complete helmet to be tested on a reference headform of the largest size which it fits when the headband is adjusted to its largest setting or, if no headband is provided, the largest size corresponding to the interior surface of the helmet. Ensure that the reference headform is firmly seated with the basic and reference planes horizontal.

(b) Apply a static force of 45.0 + 0, -0.5N to the apex of the helmet. Centre the helmet laterally and seat it firmly on the reference headform according to its helmet positioning index.

(c) Maintaining the above force and position, draw on the outer surface of the helmet the test line as defined in Clause 3.12 of AS 2512.1.

4 POSITIONING OF THE HELMET FOR TEST. The helmet shall be placed on a test headform of the same size designation as the reference headform used for determining the test line, in a position that conforms to its helmet positioning index. The helmet shall be secured so that it does not shift position prior to test.

The retention system shall be positioned so that it does not interfere with free fall, impact or penetration."

27 The Standard dealing with "Determination of Penetration Resistance" is AS 2512.4, which provides:

"1 SCOPE. This standard sets out a method for determining the penetration resistance of a protective helmet.

2 PRINCIPLE. A penetration test striker is dropped onto the outer surface of a rigidly mounted helmet positioned on a rigidly mounted test headform in a direction essentially normal to the outer surface of the helmet.

3 APPARATUS. The following test apparatus is required:

(a) Headform of the dimensions and design specified in AS 2512.1 and made of magnesium alloy, e.g. K-1A, and exhibiting no resonant frequencies below 3000 Hz. The contactable surfaces of the test headform shall be constructed of a metal or metallic alloy having a Brinell hardness number not greater than 55, which will readily permit detection should contact by the striker occur. The surface shall be re-finished if necessary prior to each penetration test blow to permit detection of contact by the striker.

NOTE: The composition of magnesium alloy K-1A is 0.7 percent zirconium, balance magnesium.

(b) Penetration striker complying with the following requirements:

(i) The mass of the test striker shall be 3 +0.045, -0 kg.

(ii) The point of the striker shall have an included angle of 60 ±0.5 degrees and a cone height of not less than 38 mm.

(iii) The radius of the striking point shall be 0.5 ±0.1 mm.

(iv) The striking tip shall have a hardness of at least 60 Rockwell (Scale C).

(d) A means to control the direction of the free fall.

(a) Condition and prepare the helmet(s) in accordance with AS 2512.2, Clauses 2 and 3.

(b) Ensure that the laboratory conditions are as specified in AS 2512.2, Clause 5.

(d) Perform the penetration test(s) as specified in the product standard. With its axis aligned vertically, drop the penetration test striker from the height specified in the product standard onto the outer surface of the helmet in the direction essentially normal to that surface.

NOTE: The height is measured from the striker point to the impact point on the outer surface of the helmet.

(d) The number of this Australian standard, i.e. AS 2512.4."

28 The "product standard" referred to in clause 4(d) of AS 2512.4 and elsewhere is AS 1698-1988 itself. The height referred to in clause 4(d) of AS 2512.4 from which the striker must be dropped onto the surface of the test helmet is that specified in clause 6.3 of that product standard. Clause 4(c) of AS 2512.3, which is referred to in clause 3(c) of AS 2512.4, provides:

"4 APPARATUS. The following test apparatus is required. Typical apparatus is shown in Fig. 1:

(a) Headform of the dimensions and design specified in AS 2512.1 …

(b) Anvils. A flat steel anvil the diameter of which shall be not less than 127 mm and a hemispherical steel anvil the diameter of which shall be 48 mm.

(c) Mount for anvils consisting of a solid mass of at least 130 kg faced with a steel plate of at least 25 mm thickness, the lateral dimension of which shall be not less than 300 mm.

Grounds 1 to 3 - Crashlab's cantilever arm test rig

29 His Honour rejected MHG's contention that Crashlab's cantilever arm testing apparatus was of a configuration that did not conform to the Product Standard, saying:

"[23] AS 2512.4 requires the dropping of a striker on to a 'rigidly mounted helmet', being a helmet which is positioned on a 'rigidly mounted test headform'. It is apparent from the description of the apparatus in clause 3 of AS 2512.4 that the test headform must be rigidly mounted to a 'rigid mount', being the apparatus specified in paragraph 4(c) of AS 2512.3. That paragraph refers to a 'mount for anvils' having certain specifications. It is clear from the description of the apparatus in AS 2512.3 that a mount for an anvil is different from an anvil.

[24] The apparatus employed by Crashlab in its penetration resistance tests at site G consisted of a test headform mounted on a ball socket which is screwed into a horizontal post protruding from a 100 mm square section of steel which is 320 mm in height. The horizontal post passes through the section of steel and is welded at both sides. The section of steel, which is hollow, is welded onto a solid base of steel approximately 460 mm x 150 mm x 50 mm high. The section of steel is also welded to the solid base by means of a gusset. The solid base is clamped to a rigid mount that satisfies the description in clause 4(c) of AS 2512.3.

[25] MHG contends that such apparatus does not satisfy the requirements of AS 2512.4. Rather, it is contended, AS 2512.4 contemplates a mechanism whereby the striker is dropped onto the surface of the helmet which is fixed onto the headform which is fixed on an anvil which is then fixed on the rigid base. I do not consider there is any basis in the language of AS 2512.4 for that contention. The description of the apparatus refers only to the rigid mount specified in clause 4(c) of AS 2512.3. No mention is made of the anvil. I consider that the apparatus employed by Crashlab satisfies the description of the apparatus in AS 2512.4."

30 Counsel for MHG criticised his Honour's reasoning as follows: clause 2 of AS 2512.4 required the test helmet to be positioned "on a rigidly mounted test headform". Clause 3(c) identified the rigid mount of the test headform by reference to clause 4(c) of AS 2512.3. This specified a "mount for anvils" having certain dimensions. Though clause 4(c) of AS 2512.3 contained the specifications only for a mount and not for the anvils required to be fixed to that mount by that particular Standard, the requirements of clause 2 of AS 2512.4 could only be satisfied if the testing apparatus was such as to comprise a mount incorporating some form of anvil on which the test headform was placed. 31 This submission is untenable. Though clause 4(c) of AS 2512.3 is incorporated by reference in AS 2512.4, AS 2512.3 deals with a test different from that for testing resistance to penetration (which alone is the subject of AS 2512.4). The test the subject of AS 2512.3 involves dropping the test helmet onto a flat anvil and then onto a hemispherical anvil, with each anvil fixed to a solid mount. The penetration resistance test, in contrast, involves dropping a striker onto a helmet in a fixed position. Clause 4(c) of AS 2512.3 specifies only a mount of a minimum mass and thickness (with a centre hole to take the stems of the anvils). Further, the only anvil dimensions specified in clause 4(b) of AS 2512.3 is the minimum diameter of the flat anvil and the exact diameter (48 mm) of the hemispherical anvil there referred to. It is apparent from this and from clause 4 and Figure 1 in AS 2512.3 that the drop test is intended to be performed on anvils that could not serve as a mount for the test headform for the purposes of clause 2 of AS 2512.4 ("penetration resistance test"), because neither stands high enough above the surface of the mount: the helmet penetration test specified in AS 2512.4 cannot be carried out at any site on the helmet crown unless the test headform is mounted on a column or a base at a point sufficiently high above the surface of the base to allow the sides of the test helmet to stand clear of the top surface of the base. 32 Although the wording of clause 3(c) of AS 2512.4 is less than felicitous, it is, we think, clear enough that the apparatus specified in that clause requires only that the rigid mount for the test headform incorporate a base of the kind specified in clause 4(c) of AS 2512.3. There is no requirement that that mount also incorporate an anvil. There is no dispute that Crashlab's cantilever arm apparatus did incorporate such a base. 33 The appellant's submission is also in conflict with the fact that the relevant Standards permit penetration tests to be performed anywhere on the curved surface of a test helmet, so long as the test sites are located above the test line for the particular helmet. Clause 2 of AS 2512.4 requires the test striker to be dropped onto the outer surface of the test helmet in a direction essentially "normal", ie, perpendicular to the outer surface of the helmet, and clause 4(d) of that same Standard also requires the penetration test specified in AS 1689-1988 to be performed by dropping the striker from the height specified in that product Standard onto the helmet surface "with its [the striker's] axis aligned vertically". Clause 6.3 of AS 1698-1988 requires that the penetration test site shall be at a point above the "test line". How the "test line" of a test helmet is to be determined is provided for by clause 3 of AS 2512.2 and it is defined by clause 3.12 of AS 2512.1. It is apparent from this definition and from Figure 2 in this lastmentioned Standard that the "test line" marks out the whole of the roughly hemispherical surface of the helmet above that line, ie, approximately the whole of the upper half of the test helmet. 34 If a test site is chosen along the line of the crown of the helmet, as is perfectly permissible, then, to ensure that the striker drops vertically onto that site, the helmet will necessarily have to be mounted on a test headform so that the crown of the helmet is uppermost and its sides are vertical, ie, it will necessarily have to be mounted on a vertical post extending above a rigid mount or base of the kind referred to in clause 4(c) of AS 2512.3. This is the form of test rig used by both MHG and Crashlab when conducting the penetration test on helmets at sites E and F. But if test sites anywhere above the test line other than along the line of the crown of the helmet are selected, the helmet must be rotated to a greater or lesser extent away from its position when vertically mounted to ensure that the striker can be dropped vertically, in free fall, onto such other sites. 35 The test apparatus used by Crashlab, which provides a means of holding the test helmet in a position out of the vertical, therefore meets the requirements of clause 3 of AS2512.4, (provided it can be said that the test helmet attached to it is "positioned on a rigidly mounted test headform"). 36 The appellant's second point is that the cantilever arm rig used in the Crashlab tests was not shown to be sufficiently rigid for it to be said that the test helmets were all positioned "on a rigidly mounted test headform" as required by clause 2 of AS 2512.4, when tested at site G. 37 The learned primary judge rejected this argument. It is convenient to set out the passage from his reasons in which he explained his conclusion:

"[26] MHG contends that the horizontal post that forms part of the Crashlab apparatus, which was sometimes referred to as a cantilever arm or support, is not 'rigidly mounted' as required by AS 2512.4. Mr John C. Simmons, who was called as a witness by MHG, is an expert in the field of structural dynamics, amongst other things. Mr Simmons said that, besides the commonly understood notion of rigidity 'in the sense of it being very stiff', there is a more quantitative expression of rigidity used widely in 'the dynamics literature'. Mr Simmons said that a structure or object is regarded as 'rigid' if its lowest natural frequency is significantly higher than the frequency of a relevant excitation force, i.e. its shortest natural period is significantly less than the period of the excitation force. It is common ground that such a definition of rigidity is widely accepted in the dynamics literature.

[27] Mr Simmons considered that a cantilever type of mount, such as that employed by Crashlab, cannot be regarded as 'rigid' without detailed tests of its resonant frequencies and associated mode shapes and comparison of those results with the relevant characteristic impact periods. He was of the opinion that, without further data regarding the Crashlab cantilever support, he could not determine whether that type of support would cause a penetration result to be higher or lower than a truly rigid anvil mount. He was of the opinion that it may be either higher or lower but that it is very likely to be different.

[28] The Commission accepted that, if Mr Simmons definition of 'rigidity' is appropriate for the expression 'rigidly mounted', then the requirements of AS 2512.4 would not be satisfied. That is to say, if the headform on which a helmet is mounted must be rigidly mounted on the 'rigid mount' such that the lowest natural frequency of the cantilever arm is significantly higher than the frequency of the relevant excitation force, the cantilever arm employed by Crashlab was not rigidly mounted.

[29] I do not consider that there is any justification for adopting Mr Simmons' (sic) definition of 'rigidity' in the present context. There is no definition of the term 'rigidly' or 'rigid' in the Standard or AS 2512.1, which contains definitions for use in the standards that specify methods of testing protective helmets. I do not consider that there is any justification for construing those terms otherwise than in accordance with their normal meaning. When the author of AS 2512.4 considered that it was desirable to specify the resonant frequencies of any part of the apparatus, that was done (as is apparent from the definition of 'headform' in both AS 2512.3 and AS 2512.4).

[30] A text book, Mechanical Engineering Design, by J.E. Shigley, was accepted by both parties as containing an appropriate definition of 'rigid' for the purposes of general usage by mechanical engineers. The definition is as follows:

'A structure or mechanical element is said to be rigid when it does not bend, or deflect, or twist too much when an external force, moment, or torque is applied. But if the movement due to the external disturbance is large, the member is said to be flexible. The words rigidity and flexibility are qualitative terms which depend upon the situation. Thus, the floor of a building which bends only 0.1 inches due to the weight of a machine placed upon it would be considered very rigid if the machine were heavy. But a surface plate which bends 0.1 inches due to its own weight would be considered too flexible.'

Further, the term 'rigid' is defined in the Macquarie Dictionary in the following terms:

Stiff or unyielding; not pliant or flexible; hard.

[31] The description of the cantilever arm, which I have set out above, indicates, in my view, that it is rigidly attached to the rigid mount as that terminology would be understood in ordinary usage. I consider that the apparatus employed by Crashlab satisfies the requirement that a rigidly mounted helmet is positioned on a rigidly mounted test headform, in that the headform is rigidly attached to the cantilever arm which in turn is rigidly mounted on the rigid mount."

38 His Honour was, in our opinion, correct, for the reasons he gave, in holding that the terms "rigidly" and "rigid" where used in clauses 2 and 3(c) of AS 2512.4 are used according to their ordinary meaning. His Honour was therefore right to reject the appellant's argument that the expression "rigidly mounted" should be understood as referring to a precisely ascertainable characteristic, ie, as importing into the Standard the technical definition of "rigid" to which Mr Simmons (the appellant's consultant engineer) referred. 39 But to correctly hold that these terms bear their ordinary meaning does not, in our opinion, resolve the question whether the Crashlab apparatus incorporating the cantilever arm was an apparatus on which the test headform could be said to be "rigidly mounted" for the purposes of clause 2 of that Standard. 40 Though the expression "rigid" is used in the Standard in its ordinary sense, it takes its precise connotation according to that meaning, from its context. Few non-technical words have a fixed connotation, irrespective of the context in which they are used. The dictionary definition referred to by the learned primary judge lists a range of connotations that the term "rigid" is capable of bearing in ordinary usage, according to the context in which it is used. In recognition of this, it is a long established rule that a term used in a statute or other instrument in its ordinary English sense should be construed in its context: see Busby v Australian Telecommunications Commission (1988) 20 FCR 463 at 468. In CIC Insurance Ltd v Bankstown Football Club Ltd (1997) 187 CLR 384, it was said at 408 that "the modern approach to statutory interpretation insists that the context be considered in the first instance, not merely at some later stage when ambiguity might be thought to arise." 41 That the Crashlab rig might be described as rigid by an observer not concerned to read and appreciate the function the rig serves on penetration testing does not mean that it must be regarded as providing a rigid mount for the test headform within the meaning of that expression in clause 2 of AS 2512.4. The requirement of clause 2 that the test headform be "rigidly mounted" is a requirement for a test that involves applying the force generated by a test striker falling from a set height onto the outer shell of a helmet in order to determine whether the helmet possesses sufficient capacity to resist penetration by the striker. The usefulness of that test as an indicator of helmet capacity will be compromised if the test rig in which the helmet is held itself behaves in a way which affects the behaviour of the helmet in resisting penetration on striker impact. We would therefore read the term, "rigidly" in the phrase "a rigidly mounted test headform" in clause 2 of AS 2512.4 as requiring a mount for the test headform that is sufficiently rigid so as not to influence, in a material way, the results of the penetration testing of the helmet. 42 Mr Simmons and Mr Emslie and Mr Gibson (the respondent's consultant engineers) were all in agreement. They all said that the requirement of AS 2512.4 for a rigid mount for the test headform was that it should be sufficiently rigid so as not to influence to any significant extent, the energy absorption mechanisms provided by the structure of the helmet during the penetration test. We regard their view as confirming how the expression "rigidly" is to be understood here: though the word is used in its ordinary meaning it is also relevant in construing the Standard in which it appears to have regard to the fact that the Standard is addressed to "practical people skilled in the particular trade or industry …". See Trade Practices Commission v BMW Australia Ltd (1985) ATPR 40-620 at 47,004-5; see also The Queen v Aird; Ex parte the Australian Workers' Union (1973) 129 CLR 654 at 659. 43 Mr Simmons in his report also said that he considered the resonance specification for the headform in clause 3(a) of AS 2512.4, which was directed to ensuring that the headform itself did not lead to an under-measure of the energy absorption properties of the helmet in the penetration test, was an indication that his more precise concept of rigidity was that which was used throughout AS 2512.4. In reliance on this evidence, appellant's counsel submitted that, since clause 3(a) of AS 2512.4 specified a headform that exhibited "no resonant frequencies below 3,000 Hz", it should be inferred that this criterion was to be satisfied by the headform in situ on its mount and not "in the abstract", ie, when the headform is not part of the test apparatus. In terms, clause 3(a) requires only that the headform meet this criterion. It is difficult to give clause 3(a) the quite tortured reading of prescribing a resonance criterion not just for the headform, but for the combination of headform and mount, for which counsel contends. But there is no reason to doubt that the resonance criterion prescribed for the test headform in clause 3(a) is directed to ensuring that the metal test headform itself does not have any significant influence on the penetration test as a measure of the energy absorption characteristics of the helmet. And it makes little sense to fix such a criterion for the test headform unless the requirement that it be "rigidly mounted" is a requirement that the test rig on which the headform is mounted is also sufficiently rigid to prevent it having any significant influence on the energy absorption characteristics of the helmet when subjected to the penetration test. 44 There was a substantial body of evidence to the effect that the Crashlab cantilever arm rig possessed more flexibility under the penetration test at site G than does a columnar mount for a test headform of the kind used by both Crashlab and MHG in its own tests, at sites E and F. The evidence on this issue is highly technical. But it does point to the prospect of the cantilever arm having characteristics which may have polluted, in a material way, the Crashlab penetration test results recorded in test report TR99/189 at site G on which the learned primary judge relied to determine the case against the appellant. A major issue at the trial was whether this difference was sufficient to cast doubt on the reliability of that Crashlab test report as an indicator of the penetration resistance capacity of the test helmets. 45 His Honour could not therefore stop at concluding that in the expression "rigidly mounted" in clause 2 of AS 2512.4 the term "rigid" was used in its ordinary English meaning. We think his Honour, of necessity, had to go on and evaluate the conflicting technical evidence to determine whether the cantilever arm apparatus possessed so much flexibility as to influence, in a significant way, the results of the penetration tests in report TR99/189 that he treated as being of critical importance to his ultimate conclusion that all the helmet models did not conform to the product standard. He did not do this. 46 But the appellant made no complaint in its notice of appeal or in argument to this Court about his Honour's failure to give the reasons necessary to determine whether the cantilever arm rig was sufficiently rigid, within the ordinary meaning of that term in its context, to entitle him to rely on those particular test results. It did not argue that if the expression "rigidity" has its ordinary meaning, but coloured by its context, the trial judge should have found that the cantilever arm rig nevertheless did not provide for a rigidly mounted test headform. Instead, it has only urged this Court, unsuccessfully, to construe the Standards as requiring a test rig that conforms to the technical meaning of rigidity put forward by Mr Simmons. 47 It is doubtful if we are in a position to make our own assessment on that issue, even if it were proper to make the attempt, having regard to the way the appellant chose to run its case on the question of the rigidity of the cantilever arm. The evidence is of a highly technical nature and assessments of the reliability of the relevant witnesses may well need to be made. 48 However, it is unnecessary to consider grounds 1 to 3 further: as will appear, we think the appeal should be upheld on another ground.

Ground 5 - Industry practice in testing only at sites E and F

49 This ground can be quickly disposed of. His Honour accepted that the evidence, including that as to Crashlab's own long-standing practice to conduct penetration tests only at points on the helmet centre line, showed that industry practice was to conduct such tests only on helmet centre lines. But he said that the practice adopted by industry could not affect the true construction of the Standard, whose meaning could not be changed by practices adopted after its promulgation. That is plainly correct. His Honour accepted that practices in vogue at the time of promulgation of the Standard could have a bearing on its construction, but observed that there was no evidence as to whether there were any such practices. It is difficult to see the relevance of such practices, even if their existence was established. There appears to be no basis for assuming that the relevant Standard merely codified existing practices. Moreover, the wording of the Standard is, in presently relevant respects, unambiguous: it expressly permits testing at any site above the "test-line" of a helmet, ie, at a wide range of sites away from the centre line of the helmet crown. The Standard therefore cannot be given the restricted operation suggested, even if a practice may have existed at the time the Standard was adopted that involved testing at a significantly narrower range of sites than is now permitted by the Standard. That the Standard should be so drafted is understandable in view of the evidence that, in accident situations, motor cyclists are much more likely to suffer impacts to their head at locations away from, rather than on, the centre line of the crown. 50 We would reject this challenge to the decision.

51 The Product Standard AS 1698-1988 specifies, as one of the performance requirements, "resistance to penetration", in clause 6.3 in this way: "When the helmet is tested in accordance with AS 2512.4 there shall be no contact between the striker and the surface of the test headform". Clause 5 of AS 2512.4 requires the report of the penetration resistance test to include, inter alia: "Degree of penetration and whether or not the striker contacted the surface of the headform". The critical tests of the four EXR model size XS helmets in the report TR99/189, upon which his Honour relied to reach his ultimate conclusions adverse to the appellant, were said by the appellant not to comply with this requirement. 52 In the column of each of the Crashlab test reports headed "Assessment", there is a notation against the particular helmet tested, for each of the sites at which it was tested, of either the word "Pass" or "Fail". In none of its reports is there any notation listing the degree of penetration, whether the particular test is described as a "Pass" or a "Fail". 53 However, it is report TR99/189 that is of critical importance to the judgment at trial differs from all the other reports. As his Honour observed, in the "Comments" section, it does report "contact between the striker and the headform when tested at site G", and it also identifies the five helmets where that occurred: four model EXR size XS and one model MXR size XXL. His Honour concluded by saying, in respect of this particular report that:

"[64] I do not consider that the failure to comply would affect a conclusion that the helmets do not comply with the Standard in so far as they fail the penetration resistance test."

54 To state in a report of a penetration test on a particular helmet that there was contact between the striker and the headform necessarily implies full penetration of helmet shell and liner. This report thus states all the information required to be set out in a penetration test report. For this reason, we would regard the critical test report, TR99/189, as complying with clause 5 of AS 2512.4 in relation to the four EXR helmets size XS that failed at site G. 55 It is therefore unnecessary to express any opinion on whether a test result can be used as evidencing compliance or non-compliance by a helmet with the particular performance requirement the subject of the test if the report does not comply with the requirements of clause 5 of AS 2512.4 (in respect of the penetration test), clause 7 of AS 2512.3 (in respect of the helmet drop test) and clause 6 of AS 2512.5 (in respect of the retention test). 56 These grounds of appeal are not made out.

57 The respondent submitted that the only question raised as to whether Crashlab used a proper headform size was in respect of irrelevant tests ie, tests upon which the learned trial judge did not rely in reaching his ultimate conclusions that all three models of MHG's helmets did not comply with the product standard. That submission is wrong. The Crashlab test report, TR99/189, of tests conducted on 23 and 27 April 1999 on the EXR model size XS helmets which, on penetration testing, resulted in the four failures at site G on which the trial judge relied, records that each helmet was tested on a headform B. It is therefore critical to whether the judgment can stand to determine if the judge was correct in finding that Crashlab had a discretion to test these four helmets with the headform B it in fact used, or whether it was essential to use a headform C. Another issue canvassed at the hearing of the appeal is whether Crashlab properly carried out these penetration tests, even if it was entitled to use a B size headform. 58 The learned primary judge acknowledged that the use of an incorrect reference headform to establish the test line will distort the test area of a helmet. He said this was so because large air gaps will occur at the helmet crown if too large a headform is chosen while a smaller headform may give a test result more typical of a user with a well-fitted helmet. His Honour held:

"[69] Due to the need for compromise between those two conflicting requirements, the reference headform size is at the discretion of the test agency. …

[70] Mr Gibson was of the opinion that the more appropriate test headform for conducting the penetration test on a size M helmet [of model EXR] was the C size test headform. However, he was of the view that the choice of the appropriate test headform was always at the discretion of a testing agency. He considered that it was open to Crashlab to determine that the B size was the largest size headform corresponding to the interior surface of a size M helmet, having regard to the air gap observed with the C size test headform.

[71] It would be unsatisfactory if the question of compliance with the Standard depended upon the exercise of a judgment involving selection of one of two possible headforms, where each choice was a reasonable one to be made. However, having regard to the negative nature of the test, as described above, I consider that for there to be compliance with the Standard, it would be necessary to demonstrate that the helmet passed the test on the basis of any headform which it was open to the testing authority to select. I consider that on the evidence before me, it was open to Crashlab to select the size B test headform, albeit that it was also open to select the size C test headform."

59 His Honour accordingly did not accept the appellant's contention that he should reject the Crashlab results, including those of special importance in report TR99/189 as unreliable, because the incorrect test headform size was selected by Crashlab, viz, a size B headform. 60 Mr Gibson acknowledged the necessary consequence of his view that there was a discretion as to the choice of headform, viz, that a helmet may pass or fail the penetration test (or one of the other tests in clause 6 of AS 1698-1988), according to the choice of headform made by the particular testing authority as that which, in its opinion, will give the more accurate result. He went so far as to express the opinion that he personally considered a size C headform was the more appropriate one to use when testing the EXR model, medium size, helmets, but nevertheless was prepared to support Crashlab's choice of a size B headform for testing that same helmet as one within its discretion. 61 His Honour agreed with Mr Gibson. He accepted that MHG's choice of a size C headform in its testing was permitted by the Standard but, because of the discretion he considered each testing authority had under the standard, he accepted that Crashlab's choice of the different size B headform for its testing was also permitted by the Standard. His Honour resolved the problem created by recognition of the existence of such a discretion in the testing authorities, viz, that a properly performed test by one tester could yield a "Pass" result, while another properly performed test on the same type of helmet by another tester could yield a "Fail" result, by his holding in [71] of his reasons:

"… I consider that for there to be compliance with the Standard, it would be necessary to demonstrate that the helmet passed the test on the basis of any headform which it was open to the testing authority to select."

62 In our opinion, the possibility that the draftsman of the product and associated Standards may have intended that the same model helmet could both pass and fail the same test, according to the differing views of two testing authorities as to which size headform would yield the more accurate result, is so unlikely that the clearest evidence would need to be found in wording expressing that intent in those Standards. In our opinion, no such intent is suggested. In view of the wording of the relevant provisions of the Standards prescribing the test procedures, we cannot accept that there is any discretion as to which size headform can be used in testing to the Product Standard AS 1698-1988. 63 Reference and test headforms must conform nominally to the dimensions and contours of the four headforms depicted in figures 4 to 7 of AS 2512.1 - clause 4. As the learned primary judge noted, figure 4 (headform A) depicts the smallest of the four headforms and figure 7 (headform D) the largest. A reference headform is the headform used to establish the test line of a helmet to be tested. (Clauses 3.9 and 3.12 of AS 2512.1 and clause 3 of AS 2512.2.) The test line for each helmet to be tested is established by placing the helmet "on a reference headform of the largest size which it fits when the headband is adjusted to its largest setting or, if no headband is provided, the largest size corresponding to the interior surface of the helmet" (clause 3(a) of AS 2512.2). (The evidence is that motor cycle helmets are not fitted with headbands.) When a helmet is to be tested, it "shall be placed on a test headform of the same size designation as the reference headform used for determining the test line" (clause 4 of AS 2512.2). 64 The Standards state explicitly that the helmet penetration test is to be performed on that one of the four test headforms (whose shape and dimension are each precisely specified) which best answers the requirements of clause 4(a) of AS 2512.4 and clauses 3(a) and 4 of AS 2512.2. The wording of these Standards is, in our opinion, unambiguous. There can only be one correct size of headform to use in testing a particular model and size of helmet. These provisions of the Standards do not confer any discretion on a testing authority to select a different size headform from that so specified, even though the authority may consider it will give a more accurate result for one or more of the tests in clause 6 of AS 1698-1988. 65 There is no doubt room for disagreement as to whether one headform rather than another of different size answers the criteria for selection of the correct test headform in clauses 3(a) and 4 of AS 2512.2, ie, whether one, rather than another, is "the largest size" etc. In the event of a dispute, it is the Court which will ultimately resolve the disagreement and identify the correct headform. But that there can be such disagreement is an entirely different situation from holding that the Standards confer a discretion on testing authorities, so that two authorities can choose different sizes of headforms for the same type of helmet and both carry out tests that comply with AS 1698-1988, though they yield conflicting results. 66 His Honour was therefore in error in construing the product and associated Standards as leaving it to the discretion of each testing authority to choose what it considered was the most appropriate test headform. 67 MHG had only ever batch tested its EXR helmets of all sizes on a size C headform. It is likely that when Crashlab tested MHG's helmets it, too, only ever used a size C headform, until Crashlab performed the tests now in question at the request of the respondent in early 1999. 68 It was Mr Small, a technical officer in Crashlab responsible for performing the tests in question, who decided to use the smaller headform B instead of a headform C, initially when testing the EXR model size M helmets for resistance to penetration on 3 March 1999. He made this decision after discussion with his superior, Mr Wood, because he considered there was too large an air gap between the headform C and the interior of the crown of the medium size helmet. There was better contact there if headform B was used. His concern was that he might not get a reliable result with headform C because of this gap at the crown: a striker might penetrate the helmet shell and liner completely, but not make contact with headform C, whereas that would be less likely to occur with headform B. When he made this decision, he had in mind testing at sites E and F on top of the crown. However, it appears that it was for this same reason that Mr Small continued to use headform B for some of his subsequent tests, including those performed at sites G and H and reported on in report TR99/189. 69 The Crashlab test report TR99/135 of 3 March 1999 on EXR size M helmets explains that "the penetration test was conducted on a 'B' sized headform, because at the crown of the 'C' sized headform there was a void (approximately 1 cm) between the headform and the inside surface of the helmet". It goes on to say that:

"AS 1698-1988 Clause 6.3 states that when tested in accordance with AS 2512.4 there shall be no contact between the striker and the surface of the test headform. Since AS 2512.4 sets out a method for determining the penetration resistance of a protective helmet, it is assumed that the contactable surface of the headform beneath the impact site would be in contact with the internal surface of the test headform."

70 The assumption is wrong. Mr Small acknowledged that there is nothing in the standard which requires there to be contact between the top of the headform and the helmet. The assumption ignores the limitations that are imposed on the accuracy of the tests in clause 6 of AS 1698-1988 by the fact that only a small range of test headforms of precisely defined dimensions can be used in those tests. 71 Mr Gibson, in evidence accepted by the learned primary judge explaining why he considered the testing authority has a discretion to select what it considers to be the headform size that would give the most accurate test results, referred to the limited range of four headforms and to how that creates difficulties for accuracy in testing, given the greater range of helmet sizes and the poor approximation of the available headforms to the range of variation in the human head. He said:

"… the choice of the headform has a critical role in ensuring the protective ability of the helmet. Hence the headform size is at the discretion of the test agency not the manufacturer. The test agency is obliged to test to the worst case within the limits allowed by the requirements of the standard. This judgment is made more difficult by the poor anthropometric shape of the headforms specified for use in AS 1698-1988. This poor shape is one of the reasons the change is being made to the ISO headforms in the newest standards for motorcycle helmets." (emphasis added)

72 So far as we can see, the test agency is not obliged to test to the worst case. So long as it tests within the tolerances prescribed by AS 1698-1988 and the associated Standards for the relevant test, that is all that it is required to do on any particular occasion. It is true that QAS requires testing to the worst case as part of its own initial certification procedure, ie, for type testing (though it accepts any testing within the tolerances allowed by clause 6 of AS 1698-1988 for batch testing). But that has nothing to do with what is required by the Product Standard AS 1698-1988. As the note to clause 8 of AS 1698-1988 shows, "the presence of the StandardsMark on a product is an assurance that the goods have been produced under a system of supervision, control and testing applied during manufacture and including periodical inspections at the manufacturer's works in accordance with the certification mark scheme of the SAA". However, as this note also makes clear, this is a scheme for independent assurance: although it makes use of various Australian Standards including AS 1698-1988, it requires a manufacturer to comply with a more detailed regime of certification, audit and quality control testing than the testing provided for by AS 1698-1988. 73 Mr Gibson illustrated the concerns he shared with Mr Small about the use of an inappropriate size test headform by reference to experiments he performed on the EXR model size XS, recorded in photographs put in evidence, prepared for testing with a headform C and a headform B. He observed that with the headform C, there was a large space at the crown while this model helmet fitted onto headform B with "far less gapping at the crown". He concluded:

"Fitting the wrong headform to the helmet for testing may affect the test results, for both the attenuation test and the penetration resistance test. The selection of too large a headform may leave gaps between the liner and the headform; it will also reduce the area available for testing. The existence of the gap may allow the helmet to pass the penetration test for reasons due to the fit. Both these effects may lead to a reduced level of safety for the typical wearer. For this reason the test headform used is at the discretion of the test agency. In this case the B size headform was appropriate for the smaller helmets, XS and S. It more accurately reflects the fit of the helmet on a typical user."

74 But selecting a size B headform in preference to a size C headform to reduce the air gap between headform and helmet liner at the crown of a helmet creates other problems. 75 Mr Hadanich of MHG, in criticising Crashlab's use of a B size headform in its tests, said:

"In my opinion, a 'B' size headform is too small for MHG's helmets and does not fit properly at the sides of helmets and around the circumference of the 'hatband' of the helmets. It is further my opinion that if a test headform is used which does not fit properly at the sides of the helmet, the helmet may not be adequately supported during testing."

76 He was speaking here about the use of the "B" headform in Crashlab's retention tests. But the same size headform must be used in each helmet that is subjected to the four tests in clause 6 of AS 1698-1988: only one of the four possible headforms will answer the criteria for headform selection to which we have referred. 77 In criticising Crashlab's use of the B headform in its penetration tests, he said:

"In my opinion, the use of a 'B' headform will distort Penetration Test results on MHG helmets. This distortion is related to the performance of the liner. I refer to Clause 9(f) of AS1698-1988 which indicates that the helmet is designed to absorb shock by partial destruction of the shell and liner and clause 9(g) of AS1698-1988 which indicates that the liner is essential to the intended performance of the helmet. The proper size of headform should support the helmet around the circumference of the hatband and the sides of the helmet and enable the liner to absorb energy over a greater area. If a smaller headform is used than is appropriate, where there is contact between the helmet and the headform, impact energy is dissipated at the point of contact, rather than being dissipated over the helmet as a whole. This will result in excessive crushing and cracking of the liner at the localised points of contact, and will distort results of Penetration Testing on the shell and liner."

78 Mr Wood of Crashlab, in explaining why Crashlab used a B rather than a C size headform in its initial tests on the EXR size M helmets, acknowledged that the choice of a size B headform introduced problems for testing at sites other than along the centre line of the helmet crown:

"Headform size 'C' did not correspond to top the interior surface (sic) of the MHG helmets but headform size 'B' did. However, headform size 'B' did not correspond to the sides of the rest of the interior surface of the helmet in that there was lateral movement. In my opinion, the 'B' size headform was the more appropriate size to carry out the Penetration Test at sites E and F of the helmet test sites as depicted in TR99/135 …at the top of the helmet, because of the gap."

79 Mr Gibson, in oral evidence, made the same acknowledgment:

"Now again, I think it stands to reason, that the smaller the head form, the greater the prospect of contact with the top of the helmet, is that correct?---Yes definitely.

But it also correspondingly means that the smaller the size of the head form, the greater is the prospect for bowing at the side?---Yes, there has to (sic) more definite motion before the helmet clamps on to the head form.

So the larger the head form - the price you pay for that may be potentially, a gap at the top, but a tighter fit at the side, is that correct?---Yes."

80 It is apparent from appendix 3 to the judgment (Mr Gibson's Figure 1 "Diagrammatic effect of headform size on position in shell", adopted by his Honour) that with headform B, the gap between the top of the headform and the inside of the liner of the crown is considerably less than the corresponding gap when headform C is used. But adoption of headform B in preference to headform C results in there being a poorer fit, ie, a larger gap at the rear of the helmet above the test line in the area of test site H, than if headform C is used. 81 The use of a size B headform reduces the "problem" that the Crashlab people considered was presented by the air gap at the crown when testing MHG smaller size helmets at sites E and F. But it creates the same "problem" when testing at site H that does not exist if a size C headform is used. 82 There was much evidence before his Honour that indicated that the size C headform, rather than the size B headform was the appropriate one to use in testing the EXR size XS helmets the subject of Crashlab's tests in its report TR99/189, and that accordingly, Crashlab, by adopting headform B, made an impermissible deviation from the test specification prescribed by the relevant Standard. 83 His Honour considered that, because there was a discretion, it was open to Crashlab to select a size B headform for testing certain sizes of helmet. But that does not detract from his finding that it was also open to a testing laboratory to select a size C headform to test MHG's helmets, including those in report TR99/198. Mr Hadanich of MHG said he believed that headform C was the largest that fitted its various helmets, as required by clause 3 of AS 2512.2. And his Honour found that MHG was entitled to test the EXR helmet size XS on a C size headform. Neither Mr Gibson nor Mr Small suggest that headform C did not answer this description. Rather did they say the B size headform was more appropriate for penetration testing of smaller size helmets because it gave what they considered to be a more accurate result than the C size headform, by reducing the air gaps between headform crown and helmet liner. 84 Mr Hadanich said that the appellant tests all its helmets on a headform C for three reasons:

"(a) the Licence issued by QAS specifies a C headform and does not permit MHG to test other than on a C headform;

(b) MHG's NATA accreditation specifies a C headform;

(c) I believe that the standards require MHG to use a C size headform. Clause 4(c) of AS 2512.2 (sic 2512.4) and clauses 2 and 5(a) of AS 2512.5 require the use of the largest reference headform that fits within a helmet. The test headform must be the same size as the reference headform. I believe, for the reasons referred to below, that the C size headform is the largest reference headform, and therefore test headform, that fits within MHG's helmets."

85 As to Mr Hadanich's first reason for MHG using a C size headform, there is nothing that we can see in the licence MHG received from QAS to apply the QAS "StandardsMark" to its helmets or in the QAS Rules, Quality Assurance Program or Technical Schedule AS 1698-1988 incorporated in this licence, that limits MHG to using a size C headform on a model EXR helmet size XS when subjecting it to penetration testing. Mr Lee, Quality Assurance Auditor with Quality Assurance Services Pty Ltd, said it remained the appellant's responsibility at all times to ensure that its own testing was performed in accordance with AS 1698-1988, and that all its products produced under the StandardsMark licence comply with that product Standard; he also said that "QAS denies that by recording any particular details in the Set-Up Data the certifying body has in any way waived any requirements of AS 1698-1988 …". Mr Hadanich overstated the position in saying its QAS licence did not permit MHG to test other than on a C size headform. But the appendix to the schedule to MHG's licence number 437 from QAS, in listing the model identifications of the goods on which the StandardsMark of the Association may be used, identifies the MHG helmets on which the mark can be used by reference to, among other things, its model designation, eg, "EXR", its size, eg, "XS", the test headform, eg, "C", and the positioning index, eg, "35 millimetres". (These schedules list various models and sizes of MHG helmet, but for each only a headform C and a positioning index of 35 mm is referred to.) The fact remains that QAS accepted, from 1986 at least until December 1998, that it was appropriate to include in the identification of the helmets upon which MHG was licensed to apply the StandardsMark of model EXR size XS reference to a size C headform as the appropriate test headform for the particular helmet. 86 As to Mr Hadanich's second reason for MHG using a C size headform, NATA for over a decade before the litigation commenced, annually accredited MHG's laboratory as one approved for testing helmets to AS 1698-1988 clauses 4.3, 6.2, 6.3 and 6.4 "using headform C" only. (Crashlab was accredited by NATA to test helmets to this same Standard "using AS 2512 headforms A, B, C, D and ISO headforms A and J".) Mr Hadanich was correct in saying that MHG could only test helmets to AS 1698-1988 in a manner that complied with NATA's requirements by using a C size headform for all sizes and models of helmets. It is a necessary inference that NATA's personnel involved in the annual accreditation of MHG's testing laboratory themselves considered that a size C headform was the proper test headform for all MHG's helmets, including the EXR size XS the subject of Crashlab's report TR99/189. 87 Mr Wood of Crashlab, in commenting on Mr Hadanich's statement that Crashlab should have used a size C headform "for the retention test" on the EXR model size XS of helmet, said that its internal circumference of 52 - 54 cm: "combined with its shape meant that in my opinion the 'B' headform was more suited for testing. The headform 'C' suits a helmet with an internal circumference of at least 57 cm." He was not cross-examined on this particular statement. In dealing with Crashlab's use of the B headform for penetration testing, he also said, in a passage already set out, that it was "the more appropriate size to carry out the Penetration Test at sites E and F of the helmet test sites as depicted in TR99/135 … at the top of the helmet, because of the gap", although it was a poor fit for the EXR size M helmet at the sides. But he did not say that a size C headform did not meet the requirements of clauses 3 and 4 of AS 2512.2 for use in testing the EXR model XS helmet. In addition to being employed by Crashlab, Mr Wood is a technical assessor for NATA. He was involved in assessing the adequacy of MHG's testing laboratory for the purposes of its NATA accreditation, at least in 1995, an accreditation which limited MHG in testing to AS 1698-1988 to using only a C headform for all its various helmet types and sizes. He understandably acknowledged in cross-examination that he was satisfied with the performance of its testing staff and the quality of its test equipment when he was involved in accrediting MHG's laboratory. 88 Mr Wainohu, Crashlab's Test Laboratory Manager, the person who supervised the testing of certain MHG helmets in July 1997, in testing an EXR model size S helmet, used a size C headform. The evidence is that the EXR model size S and size XS (the subject of the critical tests in report TR99/189) have identical shells and liners and differ only in that the XS size has more "comfort padding", ie, soft material that does not affect penetration testing. 89 His Honour should not have held that the Standard conferred a discretion on testing authorities to select different headforms for a particular model and size of helmet. On the evidence before him, he should have found that a size C headform was the one headform that satisfied the requirements of the Standards for testing EXR size XS helmets, notwithstanding the concerns of the respondent's witnesses about whether a B size headform would generate more accurate results, at least for penetration testing at certain sites. Accordingly, he should not have placed any reliance on the Crashlab test results for that particular helmet in report TR99/189 because of Crashlab's use of a B size headform. 90 However, even if it is assumed that headform B, rather than headform C, is the particular headform mandated by AS 2512.2 clauses 3 and 4 for the EXR size XS helmet, the learned primary judge was, in our opinion, in error in treating the Crashlab results of penetration testing in report TR99/189 on the four EXR model helmet size XS as results upon which he was entitled to rely in reaching a conclusion adverse to the appellant. Those tests were not properly performed by Crashlab. 91 Critical to the proper performance of the tests stipulated by the product standard for both penetration and energy attenuation is the establishment of the test line for each helmet tested. See clause 4(a) of AS 2512.4 and 6(a) of AS 2512.3: a depiction of the test line is shown in Figure 2 to AS 2512.1. This procedure requires the helmet to be placed on a reference headform "of the largest size which it fits when the headband is adjusted to its largest setting or, if no headband is provided, the largest size corresponding to the interior surface of the helmet": clause 3(a) of AS 2512.2. This same clause 3(a) also requires the reference headform to be "firmly seated with the basic and reference planes horizontal". These planes are defined in clause 3.6 and 3.8 of AS 2512.1; the vertical distance between them varies from 54 mm for headform A to 64 mm for headform D. See figs 4, 5, 6 and 7 in AS 2512.1. Clause 3(b) of AS 2512.2 then requires the helmet to be firmly seated on the reference headform "according to its helmet positioning index", ie, "the distance, as specified by the manufacturer, from the lowest point of the brow opening at the lateral midpoint of the helmet to the basic plane of a reference headform, when the helmet is firmly and properly positioned on the reference headform": clause 3.11 of AS 2512.1. The test line is then drawn on the helmet surface in accordance with clause 3.12 of AS 2512.1. 92 In addition, AS 2512.2 clause 4 requires the helmet to be positioned for each of the tests in clause 6 of AS 1698-1988 by being "placed on a test headform of the same size designation as the reference headform used for determining the test line in a position that conforms to its helmet positioning index". 93 Not only is AS 2512.2 quite specific in requiring that this be done by way of conditioning and preparation of test specimens for testing, but AS 1698-1988 itself, in Note 2 to clause 6, expressly states that this must be done "to enable the tests to be correctly applied". 94 The helmet manufacturer must specify the helmet positioning index (HPI) for each model of helmet: the dimension of the HPI is controlled by the distance between the lowest point of the brow opening of the particular type of helmet and the basic plane of the reference headform that is the correct one of the four available sizes of headform for use with that helmet. To position a helmet for testing, it must be so placed on a test headform of the same size as the correct reference headform that the distance between the lowest point of the brow opening of the helmet and the basic plane of the correct headform equates to the HPI determined by the manufacturer for the model of helmet. 95 It is apparent from the definition of "Helmet Positioning Index" in clause 3.11 of AS 2512.1 and from clause 3 ("Determination of the Test Line of the Helmet") in AS 2512.2 that knowledge of the HPI for the particular model of helmet to be tested is essential to the establishment of that helmet's test line. It is also apparent from this definition of HPI and from clause 4 ("Positioning of the Helmet for Test") of AS 2512.2 that knowledge of the HPI for the particular model of helmet to be tested is essential to the proper positioning of the helmet on the test headform for testing under clauses 6.2, 6.3 and 6.4 of AS 1698-1988. 96 When he conducted the tests on 23 and 27 April 1999 with the size B headform, Mr Small did not mark any test line on the four EXR size XS helmets; nor did he attempt to position any of them by reference to the HPI for a size B headform. On 10 May 1999, he came back to the EXR helmets that he had tested on 23 and 27 April 1999 and set about marking on each of the tested helmets a test line. As a result of doing this he satisfied himself that all of the test sites, including those at site G where the four EXR size XS helmets failed on 23 and 27 April 1999, were above the test line which he considered to be appropriate for headform B. However, he acknowledges that, when he conducted his tests on 23 and 27 April 1999 and that when he drew the test lines on these helmets using a headform B, he "did not know the helmet positioning index for" such a headform. So he used instead, the HPI for the different sized headform C. Such an approach, departing as it does from the explicit instructions in the Standards, is without justification. Moreover, Mr Small's belated marking of the test line on the EXR helmets on 10 May 1999 does not provide any ground for thinking that his penetration test results, though not obtained from tests complying with the test specifications in the relevant Standards, are nevertheless reliable indicators of the capacity of EXR model size XS to resist penetration to the level required by AS 1698-1988. 97 Mr Gibson acknowledges that the HPI "ensures that the helmet is placed in a repeatable manner on the headform by the test agency in the position intended by the manufacturer". Figure 1 to Mr Gibson's report which his Honour annexed to his judgment depicts, according to Mr Gibson, the variation of the headform position with headform size: the B headform fits more deeply into the helmet than does a C headform, when the B headform planes are kept parallel to those for the C headform. He says:

"The position of the test line on the helmet is related to the geometry of the wearer's head and is intended to ensure that an adequate area of the head is protected. Two things control its position: the size of the headform, as discussed above and the helmet positioning index or HPI, which is specified by the manufacturer for the specific headform".

"Figure 1 illustrates diagrammatically the effect of changing headform size from C to B. The test area based on the headform C is larger than that based on headform B, and the positioning index [for headform C] is obviously not correct for the B headform. …" (emphasis added)

99 Mr Gibson states what is obvious from a reading of the Standards: if headform B is used, the test helmet must be set up for testing by reference to the HPI for the particular helmet and that headform. 100 It is because the HPI for headform B is ignored that that headform can be positioned so snugly in the helmet with a lesser air gap between the top of that headform and the inside of the helmet liner at the crown than the air gap there with a headform C as shown in Mr Gibson's Figure 1. Mr Gibson illustrates, by reference to his Figure 2, what happens if the HPI for headform C is maintained for headform B when the helmet is positioned for testing on headform B: "the helmet must be rotated on the headform, the headform planes no longer remaining parallel". As is apparent from Figure 2, headform B fits even more snugly into the crown of the helmet. But there is an even larger air gap between the rear of headform B and the rear inner helmet liner than exists when headform B is placed in the helmet, with the HPI for a headform C maintained (incorrectly) for headform B also. It appears from Mr Small's description of how he positioned the EXR size XS helmets on the B size headform on 23 and 27 April 1999, that he tested them in the position shown diagrammatically in Mr Gibson's Figure 2: he set them up for testing on a B headform according to the HPI for the different C headform. 101 Even if, contrary to what we think is the proper construction of the relevant Standards specifying how penetration testing is to be carried out, Crashlab had a discretion to choose a B headform instead of a C headform equally appropriate for that test, there is no justification for Crashlab to do what it then did, viz, use a headform B with an incorrect HPI. 102 Counsel for the respondent submitted that the inference from the evidence was that the manufacturer (MHG) had not specified any HPI for the B headform and the EXR size XS helmets. But even if that is so, on a proper construction of the relevant Standards, Mr Small was not entitled to conduct penetration tests on helmets using the B headform without knowing the HPI for it. The question being one of construction, it is open to the appellant to rely on this argument at this stage of the litigation. 103 We would allow the appeal on the basis that ground 8 has been made out.

104 The appellant contended that the learned primary judge could not properly reach his conclusion that the RXR model helmet would be likely to fail the penetration test at site G, on the ground on which he based this conclusion and which is set out above in [86] to [88] of his reasons. This ground assumes the reliability of the critical tests on the EXR helmets in report TR99/189. 105 The appellant did not complain about the use by the judge of the tests on the EXR model size XS helmets in that test report to find that all the MXR models did not comply with the performance requirement of resistance to penetration. It may well be appropriate, however, to treat the MXR model as identical to the EXR model, given that the only difference is that the MXR has a chin guard. But the RXR model is different. 106 His Honour considered he was able to use the test results on the EXR model size XS that failed at site G (TR99/189) as a basis for concluding that the RXR models would also fail if tested at site G, even though the design of the outer shell of the RXR model differs from that of the EXR (and MXR models). He arrived at this conclusion because the shell and liners of all three models were manufactured from the same substance and "there is no evidence to indicate that the different shape has any bearing on the conduct of the penetration tests at site G". The appellant contends that given the gravity of the consequences of an adverse finding for the appellant, and given that the ACCC bore the persuasive burden of proving that the RXR model did not conform to the product standard, the learned primary judge was not entitled to draw the inference adverse to the appellant in the absence of the evidence that the difference in helmet shape would not have a material effect on the capacity of the RXR model to resist penetration of the kind provided for by the penetration test. 107 Mr Hadanich's evidence was that the components of each model of MHG's helmets were made from identical materials. He also said that the shell of each of EXR, MXR and RXR models was made from "hard ABS plastic capable of resisting penetration over its entire area". Each shell is also "designed to spread the load of an impact over as large an area as possible". 108 The case was fought at trial on an artificial basis: as the learned trial judge observed, in the light of the evidence that the liner is made from a material which is capable of absorbing impact energy, and that size variations in helmet models are achieved by varying the thickness of the liner, he would have expected that that could have an effect on the penetration test. Clause 4.1 of AS 1698-1988 suggests that is the position. Mr Hadanich pointed out that clause 9(f) of that Standard describes the helmets conforming to it as "designed to absorb shock by partial destruction of the shell and liner" (emphasis added). As we have explained, when testing to this Standard, four helmets of the same size must be tested, a further indication that testing, including penetration testing, is not just of the capacity of the shell, but of shell and liner, to resist penetration. However, the parties proceeded on the basis that those matters were not material and a ground of appeal inconsistent with the way the trial was so conducted was abandoned. 109 Given the artificial basis on which the case was run, we think that the evidence as to the RXR shell being of the same materials and having the same design objective as the EXR (and MXR) shells was sufficient to enable the learned trial judge to reach the conclusion the subject of present challenge. 110 It remains to note that no argument was put in support of this last ground of appeal based on that fact that the performance requirements specified in clause 6 of AS 1698-1988 take the form of statements of what must not happen when a helmet of a particular type is tested by each of the tests in clauses 6.2 to 6.4. It might be thought that, in consequence, it is only if that model and size of helmet has been subjected to the relevant test that any question can arise as to whether that particular kind of helmet does or does not have the particular performance requirement specified by the Standard. 111 This ground of appeal is not made out.

112 The appeal is allowed. The orders of the learned primary judge are set aside and the respondent's application is dismissed.

I certify that the preceding one hundred and twelve (112) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Court.

Counsel for the Appellant: DEJ Ryan SC with C Champion

Solicitor for the Appellant: Freehill Hollingdale & Page

Counsel for the Respondent: SJ Gageler with JR Clarke

Solicitor for the Respondent: Australian Government Solicitor

Trade Practices Act 1974(Cth)s 65C

At a glance

Court

Decision date

Before

Source

Judgment (145 paragraphs)

Parties

Legislation Cited (1)

Cases Cited (3)

Cited By (3)