Something to be said for a bit of experience.

I moved to London on the opening weekend of the Olympics 2012 (don't do that, stupid idea) to start a new role as an in-house engineer at a CLT design/supplier. I was only there 16 months in the end - felt longer, but for all the right reasons, we got a lot done and loved my time and the people there.

When I joined roofs like this one were just beginning to turn up on proposed drawings and people were exploring what could be done with CLT. They used to scare the hell out of me and I used to stick a ridge beam or bespoke ties in to make the thing work at the time rather than play about with a FE model that I wasn't convinced by.

This one also still scared the hell out of me, but I've had 10 years of playing about with the numbers since I first saw these and now my screw connection calculation is about 10 pages long, my CLT panel design is a good 5 pages and I chased all the forces through by hand - and it all.... felt right?!

Sounds daft for an engineer to say the numbers felt right, but let me explain.

On a roof like this, it can only really act one way, the roof wants to spread like playing cards when you prop them against each other. So to stop that happening we add a ridge beam, or tension rods - each method slightly changes how the roof panels behave (in bending, compression and at the reactions for instance) each calculation check varies ever so slightly so you're almost looking for worst-case. And as an engineer I might play about with the height of the tension rods to understand just how much force is required at each position to then compare that to what's realistic and what's not from my connection capacities.

It was the Design Manager on this job who told me they actually prefer to detail the ridge a certain way here to make installation easier and as soon as they mentioned it, I knew I had a third method of looking at how this roof would behave.

So, ran the numbers again. First step, the roof spread forces - I tried looking at in as many different ways as I could - all by hand. Each method should be giving the same number more or less which they were (I'm a big fan of the KISS method). Looking at those forces I knew they weren't far off what an engineered timber screw was capable of carrying - or right ballpark - and they were fine.

As it happens, the permanent designer shared their connection forces with me after this exercise that they'd derived from an FE model and I was bob on with each of them which was nice to see.

A younger me with 9 years solid timber experience behind him at the time didn't fancy this roof - he definitely would have had questions. Me +10 years on (with very little timber design in that time to be honest) is dare I say... a better timber engineer than what I was?

And I can only put that down to experience. Understanding the numbers and getting a better 'feel' for what works better and what doesn't. Knowing the devil is in the detail and playing about with it all, running numbers on different configurations, challenging FE model outputs and connection supplier permissible forces to be able to understand the big picture and see what's critical and what's not.

As much as you've designed, checked and re-checked a design, its still a heart in the mouth experience when they remove the temporary works and everything works as it should do (and any engineer who says otherwise is a big fat liar). It was a nice moment to see.

I think this means that I'm an experienced engineer now, only took 17 years.

.

.

.

.

#refurbishment #design #london #structures #structuralengineering #construction #building #structuraldesign #structure