3D printed switch repair parts (#P31F1)

Some more details on the Bosch oven repair. These are the replacement parts that I ordered from Sculpteo, which, according to its own website, is the market leader in online ordered printed parts. Sculpteo also happened to be bought by BASF this spring, and they’re mainly printing in France but offer support in German (with offices in Germany, I think). They offer laser cutting as well but charge additional shipping cost for those, which made me skip parts of my order…

The part in question was already featured in WHL #61 (crappy automated model generation) after which I recreated it by hand. The exchange of these parts and the successful fix of the two oven units is also a while back now, so I’d like to get this over with and share some insight on the parts and also the order process.

Link to the parts: RSDOC, exported STL, thingiverse (this is the exact STL that I ordered – yes, thingiverse is currently a train wreck).

So after weighing my options on metal prints and dismissing Materialise (which I used for a business print a couple weeks after – superb quality on aluminium parts!), I ordered those four different options for the very same design:

  • Rough stainless steel at 14.12 €
  • Alumide at 6.50 €
  • SLS (Kunststoff), Nylon 3200 Glaskugelgefüllt (Weiß) at 5.50 € (white glass fibre filled Nylon in SLS print)
  • Multijet Fusion, PA12 (Grau) at 5.50 € (grey)
  • SLS (Kunststoff) Polypropylen (Weiß) (white) *see below*

These are somewhat in order of expected parts strength, starting with a real metal print, going to aluminium and GF filled options and one polyamide (not-quite-Nylon) part as reference. Of course I also have a preprint from a colleague to see if my design will fit and work as intended, which was printed in ABS (according to the smell). I can’t really tell what the original material was, I only know it is not ABS (also according to smell). It does not deform with a lighter applied, but once it lights it has a greenish flame and smells charred, not that much of a chemical smell as ABS and polycarbonate do. It does not support combustion, but that could be an additive since it was part of oven switchgear. Who knows which injection moulding plastics were en vogue in the late 1980s, early 90s.

Before I can finally start the comparison, I need to rant a little about the order process.

After I ordered on 23rd of September, 00:29, I got a mail from tech support (non-personalized sales@ email address but fixed service rep) on the same day at 16:20. I was asked if I wanted to pursue the GF Nylon and metal prints on my own risk, as they were flagged automatically as fragile. One minute later, the order for the alumide part was cancelled “automatically” and a refund was issued. Well, what a coincidence. After a lot of back-and-forth in no more than 10 emails over the next couple of days, these are the things that I will remember before the next order:

  • auto-cancelled parts can not be re-added by customer service regardless of current order status/progress
  • ordering auto-cancelled parts (you got a refund!) in a compatible material will of course set you back by the shipping cost (10.00€ as they print in France) even if your order isn’t even in the factory yet and the re-order will ship before the original order
  • ordering single parts might also set you back X amount of money to reach their minimum order quantity (this would have been 8.50€ in my case)
  • and of course all of those things have VAT slapped onto them, which is another 16% (soon back at 19%)

In the end I accepted a “one time” “free sample” print of the thing in white polypropylene – which would have been the exact same amount as the cancelled Alumide print that still sits as refund balance in my Sculpteo account. Oh well. That’s Blue Chips flexibility and customer support for you.

And there’s another thing: Probably due to the cancellation, addition and order shuffle (the additional part was shipped with all of the others, which is not possible with two orders by the customer), there’s, to this day, this list on the order view page:

They still expect to ship the metal print in about five days time and the free PP part is nowhere to be found. I’ve created a ghost part that is constantly due next week and keeps my order active indefinitely. Every production plant management should review all delayed and overdue orders on at least a weekly basis and decide what to do, but this one has been in a delay loop for two entire months now :unsure:


The parts. I wanted to talk about the parts.

Here’s an overview of the ones that are still here. It’s (left to right) the original part, glass fibre filled Nylon, brass filled stainless steel and the preproduction, non-commercial ABS – from top and bottom.

The GF part is really thin and flimsy. Again, these are all created from the very same STL file, so the GF Nylon process must have vastly different print properties compared to all the others, including the metal one. One can feel the coarse structure despite the fine layer thickness (100µm), and the part broke before it arrived – unknown if it happened during transport in their well padded box, or even during production.

It is also interesting to note that the GF part is really flexible in general, but has some very brittle spots in between. After taking the images I bent it around, and it can do 90° and 180° bends on some stretches of the outer band just fine (flexing back after a quick while), while on other pieces it just breaks like raw spaghetti. This is just a speculation, but maybe the fibre part isn’t all that well dispersed in the polyamide base material, and whenever there are clumps there’s a major change in material properties? Photo of the bent piece is attached below.

So GF Nylon 3200 wasn’t suitable for obvious reasons. Stainless steel, on the other hand, could have been. I’ve got two issues with that: a) It’s coarse metal on plastic, so long-term abrasive properties instead of a quick break-in for plastic on plastic, b) it’s slightly warped on one end.

While it is hard to say how hard it would file down the catches on the end of the switches (how often do you switch modes on your oven), I didn’t want to take risks. Plus, it’s slightly out of whack, so I would have needed to file it down or to bend it in place, with the side effect of moving the 90° alignment of the end piece to something oblique. Here’s a cropped version from the above image to illustrate the warpage on top, compared to the exactly straight band on the bottom:

I still think it is mightily impressive that one can get such detailed metal parts made for 14€ + VAT! This ability will be used in future projects, for sure.

The other two parts are already built into fully working ovens, so I only got the pictures from the switch repair and unfortunately no overview picture of them all. Here’s the polypropylene unit next to another broken original part.

It was slightly too large all around, so I filed it down on all outside edges. Insides were alright and the catches work fine, not perfect, but certainly catching most of the button presses. I’d like to add that this part was really dusty for some reason, while all of the others were clean. Maybe a side effect from the special free sample thing, where someone picked this up by hand and transferred it to the others.

And this is the PA12 part, the one that I originally threw in the shopping cart as a backup part in case everything else fails. Well, I think it turned out great and was the easiest to work with. The slightly rough (as in “grippy, textured”) surface is great, sizing was exact, details were perfect and this one went into the switch for my oven unit at home.

So for those small and complex parts, my order preference would be PA12 first, then polypropylene. If abrasion was no issue or structural strength was very important, stainless steel could be an option, depending on vendor/pricing – aluminium prints would be preferred, though. I would not order GF filled parts again, as I don’t see any advantage over PA12 (same price) and ABS (homebrew). I bet they have their niche application as well, but even with that delicate part I wasn’t able to exceed the properties of somewhat standard, pure 3D printing materials. Those who do need GF filled plastics will know they need it, everybody else can just opt for the PA12 stuff that yields really nice parts without print optimizations.

Hope that helps – next post will feature the Christmas tree again :cool:

Leave a Reply

Your email address will not be published. Required fields are marked *