My first print for 3D Hubs, an impeller for Jacuzzi pump.
Cost for injection molded part: 200 euro.
Cost for 3D printed part: 36 euro.
Printed on Makerbot Replicator 2x. ABS, with supprorts.
How large is it? 36 euro seems like a) a LOT of plastic, b) a lot of iterations ( 
) or c) You’re paying a lot for your filament!
@Mike_Miller Same as the price for the injected part does not include only the price of the raw material, also the price of the 3D printed part is mainly composed of other factors like the price for the machine, skills needed, working time…
Gotcha
Not to be a downer, but I suspect that it won’t last long. Quality Injection molded parts often have a difficult time surviving in chlorinated/brominated water. Chemical attack is likely the cause of the original failure. A 3D printed part has so many more porous surfaces for the water to penetrate into. Not to mention that the layer bonding is significantly weaker and easy for the chlorine to attack.
Not to get on a rant or that you did anything wrong, but this is a classic example of why 3D printing in its current state will not replace traditional manufacturing methods in the near future. This is also similar to people who believe they can replace friction fit knobs that have broke with 3D printed parts.
Source: I had a stint as a plastics engineer performing failure analysis. Many of the failures I saw were plumbing parts.
What about resin impregnation? There is a steel laser sintering method where the final part is dipped in molten bronze for strength. Could we not do the same with 3d printed parts, but use resin instead?
To some extent yes, but the way that metals form bonds is a little different from the way plastics form bonds. Especially when talking about bonding thermosets to thermoplsatics. Further, dipping in a resin really only takes care of the outer surface, and not the bonds further into the part.
Not all thermoset plastics bond well to thermoplastics either. Thermoset plastics are still organic materials and can suffer similar chemical effects that thermoplastics do.
A major porblem in plastic design is creep, which often gets coupled with other issues when in use. I’m not talking about the short term creep where you visibly see the material stretch over a relatively short period of time. I’m talking about long term creep at relatively low average stress. Plastics become very sensitive to this type of creep failure when stress concentrations are present. 3D printed parts (namely FDM) are filled with stress concentrators. On top of that the stress concentrators coincide with the already weaker areas of the parts. So while a part may not fail today or in a month, it is much more prone to failing sooner than an original molded part would be.
@Ben_Malcheski I don’t know how long did it take for the original part to fail, however I think that if we are talking roughly 1 tenth of the price for the 3D printed part it worth giving it a chance and see how long will it hang on.
Didn’t mean to sound like it wasn’t worth trying. By all means experiment. My comments were more pointed at the media hype that everyone will be printing 1-to-1 replacement parts in their homes. The other thing about this prospect that scares me is that it is possible for people who do not fully understand the principles of engineering to make parts that could potentially be critical to safety. This article does a pretty good job of bringing up the discussion point: http://www.inside3dp.com/better-oversight-3d-printed-products/
I have a 1960’s vintage Lathe, with a 1HP three phase 220v motor. It has no legal warnings, has no warrantee (express or implied), has no safety interlocks, and can kill me in a very graphic way. In 8 years I’ve never been injured, I respect the hell out of the tool, and it will outlive my children. It is possible for things to exist without warning labels in all languages where the device might be used.
There will come a time in 3d printing where the question will be asked “Who’s liable? who can I sue?” and I will be very disappointed when that happens. In the article’s example, it’s her own damn fault for not knowing any better, but that’s not the right answer, now, is it?
@Ben_Malcheski I do agree with about the thing that currently the technology requires a great amount of knowledge and skill, however the link describes a situation where obviously Dafna is the one to blame.
@jenia_zelikman You might want to think about vapor smoothing this print. It should help to reduce some of the issues that @Ben_Malcheski mentioned by reducing the surface area as well as increasing the bond between the layers (at least on the surface).
Thanks @Jason_Barnett it crossed my mind.