Release of the missing Reynobond 33 BS 476-6 test report: the Zerobond question deepens

In my last article, I discussed the implications of Ian Abley's 2019 'Zerobond' tests, in which a sample of Reynobond PE ACM failed to achieve Class 0, because of the escape of molten material through the air inlet of the combustion chamber of the BS 476-6 Fire Propagation test:

The temperature data for all three specimens exhibited a late rise in the differential between specimen temperature and calibration temperature. For example, for Specimen 1, for which we have Laboratory Record confirmation of the escape of molten material before the end of the test, the first observation of a positive differential was at 18 minutes:

I argued that the only plausible explanation for the temperature rise was combustion of molten PE that had escaped the specimen core and entered the chamber. The chamber floor is horizontal and is flush with the lower edge of the air inlet, so that there is no lip that could retain the material in the chamber. Abley's photographs show that it remains molten at the chamber temperatures. The only factors I can see that would prevent the material escaping from the air inlet and invalidating the test are a) its viscosity and b) loss of material through combustion.

At that time (16 May 2024), I had BS 476-6 temperature data for three specimens each of three PE ACM products: Reynobond 160 (May 1997), Reynobond 33 (3.4 mm core, Sep 2006), and Reynobond 33 (1.4 mm core, Sep 2006). In six and perhaps seven cases, the data indicated that molten PE had entered the chamber and combusted. Specimen 2 of the 3.4 mm core product was of particular interest since it showed definite evidence of combustion in the chamber at 16 minutes:

earlier than for 'Zerobond' Specimen 1. I asked in my last article whether it was plausible that there was no escape of molten material from the air inlet in this test, and suggested that some experimental work could help provide an answer.

I also drew attention to the apparent failure of the Inquiry to obtain test reports for the 2.4 mm core version of Reynobond 33 which also obtained a Class 0 classification in September 2006. Its BS 476-6 Fire Propagation Index was 2.0, higher than the value for the other versions and for Reynobond 160 (1997), and higher than the value that would have been derived from 'Zerobond's temperature data if the tests had not been invalidated. I argued that a higher index seems to indicate either an earlier entry of flaming molten material into the chamber, or a greater amount of it, or both, and that any combination of these would seem to increase the probability of the escape of material from the air inlet and invalidation.

The Inquiry published this test report on 18 June 2024:

and the temperature data, as expected, is of great interest. Specimen 1 had positive differentials of 86o C, 122o C and 100o C at 16, 18 and 20 minutes respectively:

Specimen 2 had a positive differential of 2o C at 14 minutes. Given that there had been negative differentials of -6o C, -5o C, -4o C and -4o C at 8, 9, 10 and 12 minutes, this could well represent the beginning of combustion in the chamber rather than normal experimental variation. There were positive differentials of 81o C, 103o C and 94o C at 16, 18 and 20 minutes respectively:

Specimen 3 is the most remarkable of all, with positive differentials of 13o C, 105o C, 169o C and 147o C at 14, 16, 18 and 20 minutes respectively:

In this case, I think we can say with some confidence that there was combustion of molten PE in the chamber at 14 minutes, six minutes before the end of the test.

Let us now compare this newly released data with the 'Zerobond' data. The table shows the differential (Specimen temperature minus Calibration temperature) in oC from time 10.00 to 20.00 for the three 'Zerobond' specimens and three 2.4 mm core Reynobond 33 specimens:

Is it plausible that there was no escape of molten PE from the air inlet in any of the 2.4 mm core Reynobond 33 tests? Some experimental testing of a range of PE ACM products should help provide an answer. The tests should be video recorded, and it strikes me that a high temperature camera looking into the chamber through the air inlet would give extra understanding of how the molten PE flows across the chamber floor.

Were the 2.4 mm core Reynobond 33 test reports withheld deliberately?

At the time of my last article, there were publicly available classification reports for all three versions of Reynobond 33, but test reports for only two of them. I thought this was odd in itself, and perhaps even suspicious given that the Fire Propagation sub-index i(3), which pertains to the final ten minutes of the test, was more than double that of either of the other two versions:

Now that we have the test reports for all three versions, we can compare the fire behaviour of the three versions in more detail (Specimen temperature minus Calibration temperature from 10.00 to 20.00 in oC):

The 2.4 mm core version is the only one showing evidence of flaming molten PE in the chamber at 14 minutes, and at 16 minutes, the highest temperature differential at 105o C is almost three times the highest differential in either of the other two versions at 37o C. Was this report deliberately withheld from the Inquiry in order to avoid drawing attention to the (to my mind) puzzling phenomenon of flaming molten PE in the chamber for at least six minutes without any egress from the air inlet?

It would appear that the Inquiry did not have the 2.4 mm Reynobond 33 test reports at the time (24 October 2018) that Barbara Lane published her Phase 1 Report (supplemental), Appendix O (p. O-7) showing only the 1.4 mm and 3.4 mm core test reports:

The recently published 2.4 mm core BS 476-6 test report has Inquiry identifier EXO00001951, indicating presumably that it was obtained from Exova rather than Arconic, and it was read into the Inquiry record by Inquiry Counsel Richard Millett on 10 March 2021 (with 2.4 mm core corresponding to 3 mm panel thickness):

I understand from an email from the Inquiry dated 20 May 2024 that not all documents read into the record had been published:

and I doubt that the Inquiry would have had any reason deliberately to withhold the reports from publication. There remains the question as to whether Arconic deliberately withheld them from the Inquiry.

Andrew Chapman