To further this point, 95% of hubs and 99% of customers do not have sufficient metrology to measure the things that are being printed. Bounding box dimensions are a great tool to get the relative size of a part in a digital space but are useless in the real world. Most prints are not rectilinear and are impossible to measure short of scanning them back into the digital space and comparing them to their source file. Taking a photograph of a print next to a ruler or with a caliper are both useless verification methods unless focal length of camera and a bunch of other easily manipulated factors are taken into account. To validate a dimension the measuring technique has to be objective.
How about instead of holding hubs to a ridiculous standard of quality which are unenforceable and will undoubtedly increase the cost of every FDM print and reduce overall ease of use for this whole system we add a step to the verification process where the hubs have to verify the dimensional consistency of their machine. Hubs could print several calibration cubes and either send them direct to 3D Hubs headquarters or measure them, similar to the initial marvin. This could even be a periodic check. This would be like a drivers license, you need to prove you can do it once or twice and then you’re left alone. Asking for every single print to be measured increases the cost and significantly reduces the appeal of 3d hubs in general for both the customers and the hubs. I truly understand the spirit of what it is you guys are trying to do but implementing blanket rules that cant be enforced without tons of effort not the way to do it.
I think its mainly because they dont have enough feedback internally on the practicality and implementation of things they propose. This very much feels like a marketing driven company with very little engineering or maker feedback, which quickly turns into a circle jerk of corporate values and quarterly earnings oriented business moves. The customers needs will slowly get belittled until they are forgotten, all the customers will feel alienated. A new startup will pop up and cater aggressively to the core user needs and 3D hubs will fail as a business. Unless they start listening and discussing the issues at hand with us.
I understand your concerns as a final user, however wouldn’t be more easy for you and the hubs that print your cabinets to have a list with the specifications that you want your cabinets to be printed (i.e. I want them with so much infill and so much thick walls)?
That way both you and the hub will know what you want? To blame each hub that they haven’t print your part the way that you want when you didn’t gave them any specifications is a little unfair. And no 3DHUBS they should not make standard the specifications that you want your part to be printed because it is not standard for millions of other parts. So my suggestion is to submit your specifications when you place the order.
Hi all, newbie here, been lurking as I’ve got a cheap CTC machine I’ve had for a week and been playing with/upgrading, anyway, couldn’t resist weighing in on this one - I used to run a CNC mill for prototyping parts, and there’s no way in hell I’d agree to a 1% tolerance on most small plastic parts, especially ones made hot on the machine, even though I could hold far closer tolerances than that on the machine, why?
Well not only do you have shrinkage to deal with, but if the part relaxes over a few days whilst in shipping, absorbs some moisture whilst it’s sat in their workshop for a week, or you live in a place much colder/hotter than the customer, the plastic parts could have moved out of tolerance just through natural growth/reshaping. You could have a 1% change in dimensions just through normal room temperature water absorption rates alone on such as Nylon going from dry printed back to it’s more normal 2-6%
If you’re going to specify a 1% tolerance you really need to spec out whether that is relative or absolute, and at what temperature and humidity, and explain to the customers that their parts are going need to sit in a room with a controlled temperature and moisture level for a couple of days to stabilise so you can measure them accurately. Otherwise you’re going to have some real issues guaranteeing those tolerances on FF production parts which have all manner of internal stresses just from the way they’re formed, not to mention varying crystallinity not just from surface to interior but also on the filament skin to core. That’s not even taking into account any creep that might occur on the part later on depending on how and where it’s stored.
