Lessons Learned

[TobyCWood]

I am reposting this updated and as a sticky.

I do make my own things in VIACAD (which BTW to me is the biggest thrill of all the fun aspects in 3D printing), but the majority of stuff I have printed is a large variety of things from the Thingiverse. According to my db I am now at about 340 things that I have printed many of which I printed multiple times. I also do a lot of printing for paying clients. This all makes for a very wide variety of types of things to print. This variety made for some tough lessons:

1. A sample of 1 or 2 is not enough to form an opinion on something. You need to output quite a bit of stuff with a lot of controls before you are on solid ground. For example, printing on to acrylic without any tape worked great for me the first 4 or 5 prints but by the 6th or 7th the PLA bonds to the acrylic. or... Any unknown or unexpected change in the PLA even within the same spool will make for nondeterminism. Such as diameter changes, temp range changes, density changes, etc. or... it took me lots of experimentation before I learned how the hair spray coated, dried and released in the freezer.

2. What you are printing is a VERY big determinant. So you control all sorts of stuff; leveled bed, extruder is upgraded, everything is fine and you can easily print your design you made in your CAD program. A design which is mostly contiguous, flat or curved surfaces...  then, for fun, you try to print something with a complex mesh with completely different toolpaths and all your deterministic assumptions evaporate. Yes, you need to re-examine all your settings. 

3. The issue of tolerances is pretty difficult to nail down. Yes, getting the bed level in relationship to the hot end is pretty important, as is that distance... but the distance from the previous layer as the build progresses changes. That distance from the nozzle to the previous layerl at the beginning may differ as a result of the build 7 hours later. Also, there IS a sag dead center of the X and Y gantry rails... but it is within the acceptable tolerance range as long as everything else is set within that range. To learn that takes time, patience and perseverance and even after 10 months I still can not be 100% spot on for everything I try to print.

4. You cannot care AT ALL about the print or the PLA. If you do and it fails you will get discouraged. After a 10 hour print a failure can really get to you. You simply can't let it.

 

Given the above... you print something big and complex successfully one day... a week later... possibly not. That does not mean your device is faulty or you did something wrong. There is a very large set of variability one must understand and then control... even after that there are uncontrollable variables. I believe it is that lack of complete control that make this so addictive!

[James Holmes-Siedle]

Good post Toby - I feel everything you say - except I use Autodesk Revit to design and am going to explore Inventor.

James

[TobyCWood]

I have a family member who just switched to Revit. I like Inventor.

[PrintedSolid]

#4 is an excellent recommendation for your mental health.  It should be engraved on the front of every machine :)

[James Holmes-Siedle]

The problem with Revit is it does not like itty bitty - i.e any lone or segment under 1mm

Plus it is not great at voids/combinations - which is fairly basic stuff, but then that is not what it is made for :) (It is really good at buildings!)

Will be trying inventor in the new year as apparently it will talk to revit.

James

[John Borlaug]

Tim Olson sure has the marbles in the bag on the ViaCAD doesn't he! I can hardly wait for my recycler- then I will be totally 'unattached' from any more stresses the screwin' with 2X has produced me!

[3D printing Co Ltd]

Recycle?    I’ll pay for the shipping to NZ.

Toby’s leading post is on the money. Out of the five printers we have which includes a Stratasys Eden 333 capable of 16 microns, the two MB 2x we continually use just rule. Gr8 bang for the buck. One 2x has just tipped 3,000 hours and still going strong with a few mods and good calibration maintained they are fantastic machines. Note: - It must be said that a well-constructed model using a good CAD program is essential. We use Inventor Pro, I believe a must for controlling the design/model prior to print.

If one studies certain idiosyncrasy long enough one will conker. There is no substitute for that of experience, one cannot buy it, you can’t borrow it, and you can’t even rent it, you must earn it. That is time!!!

Cheers,

www.3dprintingco.co.nz

[Joe Paladin]

I am a mechanical engineer, and very familiar with scientific method. At the same time, if you are given pot metal to work with I do not care how much you try you are not going to make Gold jewelry.

If you are printing with PLA plastic that is just not a professional thermoplastic more for toys.  After buying my Makerbot 5th gen and realizing it was only intended to print PLA I did a lot of research and testing.

The end result in my opinion is PLA sucks and ESun PETG is amazing using the right machine settings.  Fortunately the 5th gen is capable of printing PETG perfectly without a heated bed, I print just on blue painters tape.

The PETG filament comes on a spool that does not fit into the Makerbot, and you need a well designed filament roll holder not something generic on ebay sitting on roller bearings with no filament tube guide.

Also you need the right machine settings to print PETG and you probably do not want to go through all the trial and error I went through even if you are a mechanical engineer like myself.

It takes 11 hours to print the components in PETG.  In PETG you can squeeze the locking tabs together 1,000s of times with no change, printed in PLA the first time I squeezed them together one side snapped off. PLA is to brittle and not strong.

Mine design is a similar design to what is on the machine sliding the Esun PETG spool over flexing plastic that locks the spool of filament in place in a rectangular wood frame and filament tube guide.

The machine settings are on a USB drive.  My email is joepa...@hotmail.com   For $99 plus $20 shipping I could make one for you.