Got a week of testing coming up next week in Scotland, which I’m looking forwards to, good crew to spend time with, in the Highlands, which is always beautiful – and let’s not talk about the weather forecast. I’ll be writing about it on the blog as we progress…
Thinking about this got me pondering testing.
Testing implies something that is complex and somehow mysterious. It’s not – or at least it needn’t be. As a kid I loved Heath Robinson’s drawings (see above) – still do, come to that – where he would develop a solution to a seemingly easy task by means of some ridiculously complex contraption held together by bits of string, nails and springs (I suppose the modern-day equivalent to that would be zip-ties and Duck tape). This is one way to do you testing, but there is a different approach, which – as in any systems design – is to keep it as simple as possible.
First off you want to decide what you want to find out about: maybe it was a question that came up during a lunch break and that lead to a heated debate. One way to resolve this is to define a test set- up and to run some tests. An example? One year I finished off a presentation with a slide with a picture of a compact steel karabiner choked onto a lanyard. The discussion that year had been on configuration of connectors and one of the summaries was a suggestion or a question whether it might be an idea to use a compact steel connector in certain applications. Thinking about this, observing my own use of karabiners and discussing it with Chris raised further questions, so we decided to do some testing on choked karabiners.
We defined a test methodology, met in our yard on a Saturday, set up and got going… The test set-up involved a mechanical advantage system, three steel drums of different diameters to simulate a range of branch diameters, a load cell and a Dyneema line.
We defined a three minute 15kN pull on three different diameters of “branch” and documented each sample. This resulted either in a pass or fail. The summary was that all karabiners struggled on the small diameter, with one complete failure (karabiner broke), whilst on the larger diameters, the shorter the karabiner, the better it handled this unusual loading – unusual in the sense that the connectors are not designed for this. No manufacturer would ok this kind of loading, the design merely foresees loads along major and minor axis and with the gate open. Still, we felt this was a configuration used fairly frequently in tree care and we weren’t able to find any answers to our questions, so by doing this testing we improved our knowledge – and were actually able to back up a gut feeling with facts.
The other testing I mentioned in a post a while ago was the continuous long descents testing. For this we went to the Musical Theatre here in Basel, as that offered a height of 20m plus, sufficient to do the testing in. Next problem was measuring the descent speed, 1, 2 and 3m/second. The way we sorted this was by building a rig consisting of a wooden base that we could load with weights and a small 20″ kid’s bike wheel with a speedometer on it. The line ran round this and gave us the speed km/ h speed that we converted into m/ sec. Obviously not super-accurate, but good enough to get the descent speeds roughly right.
Not brilliant quality image, sorry about that, lighting left something to be desired… still, gets the point across. Actually looks quite Heath Robinson-ish, come to think of it.
The point I’m getting to here is that if you have a question about a technical issue, go beyond a web search on Wikipedia, Youtube and the forums, get together with some friends, work out a replicable test methodology and go for it. Don’t let yourself be intimidated, testing is not something reserved for boffins in labs in white coats, it can be hands-on, exciting and can expand your comprehension of tools and systems you are working with.
It doesn’t even have to look very sexy, as long as it does the trick! Just look at the Russian space program – basically tractors sitting atop big rockets, but they get the job done.