A few reflections on yesterday.

I think it's important to distinguish between the failed solid aluminium 8414 tests undertaken by Kingspan Offsite and the tests that Kingspan Insulation and Celotex had done.

1/8

The latter were ‘worst case scenario’ tests aimed (ill-advisably) at creating the widest possible system scope of application.

So they were ‘best case’ attempts to pass – large sheet panels, narrow joints etc. These passed.

2/8

The Sotech tests looked more ‘project specific’. The panels were small (lots of gaps for oxygen) & cassettes (harder for cavity barriers), so not the 'best effort' at creating a reference ‘pass’. The first test failed badly even though the cladding was ‘non-combustible’.

3/8

This highlights:

http://a.You  couldn't predict these tests. Introducing Desktop studies as a means to bypass testing before the knowledge was established to justify it was reckless
b.Small detail changes can make large performance differences

4/8

The 1st point supports the use of large tests, but the 2nd is a stark warning. Test rigs are built perfectly with no time/cost pressure. They don’t have defects so don’t convey the dangers of ill-conditioned systems that pass if perfectly built but fail badly if not.

5/8

It is possible, (& Kingspan have shown) that solid aluminium cassette rainscreen can pass an 8414 with K15 but to most people the system wouldn’t look much different to the one that failed so badly with Sotech. That's a big challenge for certifiers and building control.

6/8

I’ve said before that the people that know most about how such systems behave are those who see these tests day-in day-out. The intelligence from these tests should have led to the desktop study route having been stamped out at an early stage and a warning sent to industry.

7/8

This is my main argument against combustible materials. It's ‘possible’ to build safely with them but if installed poorly, then the potential for danger is far greater than with non-combustible materials. We're talking about their application in construction, not aerospace.

8/8

I guess, one of my other practical concerns with combustible materials that rely on flame retardants, is that you just can't easily determine (e.g. see) what you've got.

9/8