12.A.I.L. = Twelfth Attempt In Learning

Welcome, dear reader. And prepare to ditch a few biases you may have about a “Cinderella Material” if ever there was one. Hell, if enough people would do that, then one day this material might actually break through.

?My next material failure lesson in this series started (where else?) at DAF Trucks. One fateful morning in the early ‘00s, there was a meeting between two Very Important People: Hans de Wit, Director of – at the time – Corus R&D, and Martien de Louw, head of DAF Product Development. Once these two VIPs were done comparing the relative merits of their respective automobiles, they decided it was high time to set up some kind of collaborative effort. The result? To find applications of something called “thin-walled ductile iron”, or TDI for short. And once the VIP dust had settled, this task was assigned to me.

?

Let’s talk physics first. When a metal goes from liquid to solid, it shrinks in volume by a few percent. If you want to cast anything more complex than – say – a flat plate, this “solidification shrinkage” isn’t on your side. This is because the thinner parts of your shape cool down faster than the thicker parts, and may turn solid while the rest is still liquid. The result is that once everything is solidified, your cast shape suffers from lots of internal stresses, or – choose your evil – comes out warped, or cracked, even. Yes, combinations of these defects can also be on the menu. Understandably, none of that is likely to make you any friends soon. If you’re smart, you will therefore design your shape with that shrinkage in mind (e.g. keep part thickness constant), but that restricts your form freedom. It helps to get a skilled foundry specialist who knows how to position “risers” and “feeders” properly, but even the best ones cannot beat physics.

?The fun thing with cast iron (or rather, one of several fun things) is that it has very small solidification shrinkage – close to zero in fact, or even negative, in some cases. This simple fact alone already facilitates the design of castings considerably. Now, ordinary cast iron parts tend to be a bit “chunky”, but not this TDI: that can be cast in sections just 3 mm thick, without endangering the material properties. Speaking of which, as the “D” implies, it’s tough stuff, with a strain-to-failure of 20% or more. It can also be surprisingly hard and strong, handles the heat like a pro, and with a lick of paint it won’t rust. Suddenly the value proposition begins to look really appealing, all the more because as a base material, cast iron is dirt-cheap, and there’s nothing in TDI that in principle makes it a lot more expensive.

?

That much I knew going into the project.

?

Hans de Wit had wasted no time keeping his end of the project bargain, and had assigned a gung-ho collective of R&D specialists to start designing TDI parts for DAF. I knew them well, thanks to a sort of internship I had done at their home base, the Product Application Centre, back in the early months of the year 2000. They were really, really good. At my end, backed up by Martien, I made sure this collective got to see an actual truck up close and personal (a first time for many of them, I recall), scrutinizing the various parts they were going to try and replace. Lots of welded sheet metal brackets, mainly: nothing super-fancy, but all fair game in the hunt for cost- and weight savings. Plus, I’d arranged for the various DAF engineers responsible for those parts to be on stand-by, ready to divulge specifications, load cases, assembly considerations and other vital information needed for redesign. It sure was a promising day, and by the end of it, the Corus collective went home satisfied.

?Over the next two weeks or so, I received lots of proposals for redesigned parts, all to be made from TDI. At least on paper (read: computer screen) it was all looking good, with colourful Funny Elephant Artwork to convince everyone of how strong those parts were going to be. The DAF engineers were as impressed as I was, and eager to see how this would pan out.

?

Meanwhile, back at the ranch, err, Central Laboratories…

?

At the time, Corus was not suppling cast parts to DAF. That of course needed to be fixed, so I got in touch with the Purchase department pronto. Supplier selection is something that the folks in Eindhoven took (and still take) very seriously – I hope to tell you more about this topic one day – but perhaps, I asked as sweetly as I could, we could speed things up here to appease our two VIPs? Sure thing, Purchase said, let’s ask Corus to supply a few samples so we can tell where they stand, and fast-track the paperwork. And of course, they added, let’s send out RfQ’s (that’s “requests for quotations”, you’re welcome) for those parts, so we can calculate how much money we are going to save.

?

Unfortunately, here I ran into a cast iron wall. And no, it was not a thin one.

?

As good as Corus was in supplying design proposals, they just did not have their TDI casting supply chain in order, at least, not to a point where a major OEM like DAF was convinced. Not just any cast iron supplier can deal with this stuff: for starters, you need to know how to make sand moulds with very tight tolerances, and the metallurgy isn’t exactly standard either: I recall something about inoculating the molten metal with the right amounts of salt and pepper. There was a lot of increasingly-frantic back-and-forth chatter between Corus and DAF, but sadly, we didn’t get further than those colourful FEA plots.

?

I say “sadly” because it really was that.

?

You see, TDI is wonder-stuff for structural engineers – not just in terms of cost, but surprisingly enough, also in terms of performance per unit weight. Sure, it has a high density, so TDI parts loaded in bending, buckling, or torsion will be way heavier than a properly-designed aluminium casting. But the form freedom provided by casting is such that these unfavourable load cases can usually be avoided, putting the material in pure tension and compression instead. Ideal for brackets and similar “node-shaped” structural elements! Plus, with proper design, you can avoid major stress concentrations, guaranteeing good fatigue properties as well. Like Cinderella, it may not look like much at first sight, but this cheap-ass material packs a surprisingly powerful punch. Oh, if only…

?

My team leader, the incomparable Jack Martens, was quick to offer consolation. “You know”, he began, “this was actually our third TDI project. The first took over one year before it stalled, the second roughly six months. You managed to prove it is still a dead end in just six weeks – that’s progress!”

?

To this day I am not entirely convinced that he wasn’t being [sic] iron-ic.

?

Lesson learned: if you want to get an OEM to use your new material, don’t just get yourself a gung-ho development collective – get your supply chain in a row, too.

?

PS the abbreviation “TDI” can of course refer to the Cinderella material I just told you about, but it is more commonly associated with the common-rail turbo-diesel engines that became popular in the 1990s. Prior to “Dieselgate”, they had a very good reputation. That’s why I used the TDI logo (shown above) in all project communication – but that too wasn’t enough. Sigh…

?