angusgbishop - Personally I'm running an SLA machine rather than FDM, the biggest problem I think I have is the constant maintenance of my resin tank, plus there is always that problem solving process and inevitable doubt after a failed print.
Most of my issues and frustrations are probably about repeatability, my resin is prohibitively expensive and I don't usually charge enough to risk a failed print, therefore when a part fails and I cant immediately identify the cause, not only am I usually out of pocket, I'm also concerned about my 3000 USD printer not functioning.

elpulpo - Interesting question! To be honest, the biggest frustration is my own limitations when it comes to design ability. I'd love to be able to design more organic shapes and you see wonderful designs online but I haven't a hope of freeing up enough time to get to know the software platforms sufficiently well. Sketchup, Tinkercad etc are great for rectilinear designs but I really need to get up to speed with Rhino, Blender etc and these babies take some serious time.

Far as the printing itself is concerned, my biggest frustration is the limitation of a proprietary printer. Limited temperature control and difficulty in printing some new materials. That'll all be cured with my new extruder! :-D

tanju-b - I have an FDM reprap.
Definitely echo the sentiment of designing. Designing is still hard, and a mix of art and science.

With an open source printer, all of the print variables are open to me compared to proprietary printers which means that I have to spend time tweaking and calibrating for different prints. It can become consuming at times.

With 3D printing all of the elements/limits of the physical sciences come into play: Newtonian physics about mass, velocity, acceleration, etc; then you have heat transfer and thermodynamics to contend with as well; lastly is material science. And then on top of all that is computer science and the tool chain: the modeling, the slicing, and the controller.

It's a lot. To make matters more challenging, we've all been conditioned to think of cheap plastic mass produced parts as being really cheap and low cost. 3D printing is not for mass production, it's for customization. So when someone wants a customized plastic part and expects it to cost $5-$10 and it's gonna cost you $25 just to break even on a print..... That's a huge disconnect

vinnie90 - First off I would have to agree with everyone above. You hit the nail on the head. Just to add to the frustration list is ABS printing. It warps too much and you wont know until you are 30 min into a print and see it popping up. I am able to tame it some what but some models are long in length and those are the worst. Another one is when you are printing more than one part at a time and one part messes up or comes off the build plate. Then you have to restart over.

defcon3d - There are many... :)
The biggest is when I'm into hour 5 of a 7 hour non supervised print. I get busy on the phone or with a client and the filament on the spool decides to stop cooperating.
Of course at that point I don't here the noise coming from the spool. The printer starts to air-print for the next 10 minutes before I get back to realize what has just transpired.

The great news is that I've got a new job lined up from another client. Now to go back and start the print all over again.

Although, it doesn't happen very often, when it does it is always seems to be during a long run part.

The cost of doing business, yah got to love it.. Haha!

toybuilder - On the point about people's expectations for low cost printing -- I've generally found professional customers to not have problems, while the "average joe" with trinkets think the prints should be priced like a low-cost mass-produced item.

Cynical as it may be, you may need to adjust your expectations based on the type of order coming in, and adjust your quoting process to match. For smaller "trinket" items, you might be better off using a quick-and-dirty estimate based on the object volume and quote the upper end of a reasonable range that is likely to apply for the print.

For more complex prints, quote more carefully and spend a little more time working with the customer to refine the job requirements and to discuss the technology.

Also, I don't try to print jobs "overnight" anymore unless the client has expressed a strong desire for it, and is willing to pay for that. If it's important enough for you to lose sleep to make sure the job is done, it should be important enough for the client to pay the premium.

A 2 to 3 day turnaround with the possibility of earlier delivery is, imho, reasonable. That way, if a print fails or the printer goes down (which does happen, eventually!), there's time to catch up.