Watertight design verification for 3D printingI

[vitaliy_kh]

My name is Vitaliy.

I am new to Fusion 360 and a bit familiar with 3D printing but did use many differet EE CAD tools.

Originally I created my design in a TinkerCAD, printed it 4 times with different settings but it always was leaking. Somebody told me I have to use Fusion 360. Watertight design could be achieved only with a right tools (I 100%+ agree with this statement). So, I redesign my object in the Fusion 360 using personal free license. By necked eye the design looks very good and solid.  I used a Section Analysis tool and it also looks good. Just in case below are pictures of my design.

My question is: How to make sure the design is watertight before attempting to 3D print it?

Any ideas, help and advice how to analyse a design are greatly appreciated.

I can provide more details but I am not sure what else could be useful.

Thank you in advance,

- Vitaliy

[Drewpan]

Hi,

You cannot 3D print a Fusion design directly. Usually you will use the Export tool

or you would turn it into or save it as a Mesh. Meshes are 3D printable and can

be saved as 3mf, obj, stl etc.

The usual method for creating a 3D printable object is like this.

Switch to the MESH tools.

Use the Tessellate command to turn your Body into a Mesh.

Most of the time this will create a "watertight" 3D printable body but for

various reasons not always. Typically the Repair command will fix these

issues and I normally use it anytime I tessellate something.

Then you simply use the Save as Mesh command by Right clicking on

the Body in the Browser Tree.

You set up the type of file you want.

And Save it.

This should create a file that can be imported into the Slicer of your Choice

and should freely print from there.

Cheers

Andrew

[TheCADWhisperer]

@vitaliy_kh 

Can you File>Export your *.f3d file to your local drive and then Attach it here to a Reply?

And Attach the resulting *.stl file as well/

[vitaliy_kh]

Hi Andrew,

Thank you for the advice. I went through the recommended steps including a REPAIR.

REPAIR has a multiple options. I used a default:

I created STL flles and imported in the Orca-Flashforge slicer.

Optin Gap Infill highlighted almost the entire body:

I am confused what I am seeing.

Does it mean the entire design is bunch of gaps?

[vitaliy_kh]

All the requested files are attached.

The entire design consist of two bodies/parts which will be glued together.

Unfortunately it will not fit into 3D printer as a single piece.

Thank you for your help,

- Vitaliy

[TrippyLighting]

---

@Drewpan wrote:

Hi,

You cannot 3D print a Fusion design directly.

---

While I would not recommend this for the OP, Fusion does have support for Additive Manufacturing.

---

@Drewpan wrote:

Most of the time this will create a "watertight" 3D printable body but for

various reasons not always. Typically the Repair command will fix these

issues and I normally use it anytime I tessellate something.

---

This only applies if the model is not available in the originating format it was designed in. 

If the user designed the part in Fusion, they should go and find the root cause in the CAD design!

[vitaliy_kh]

OK, did I get it right:

The REPAIR option should/must not be used if parts are natively created in Fusion 360.

But which tool/option should I used to make sure the design is watertight?

[TrippyLighting]

---

@vitaliy_kh wrote:

... but it always was leaking...

...

...

My question is: How to make sure the design is watertight...

---

I think there is a confusion of terms here 😉

If the physical artifact you printed is leaking water after you printed it and glued it together, that can have several reasons that are all related to how the part was manufactured. The clearances in the design (often mistakenly labeled tolerances) don't match the manufacturing method, or the porosity of the 3D printed materials, or insufficient glue, or the wrong glue etc.

That has nothing to do with how you modeled it in your CAD package.

What CAD and 3D modeling folks refer to as "watertight" is that the user creates models that fully enclose a volume, without openings. Your model consists o two solid bodies. Both of these solid bodies are "watertight" models. 

---

@vitaliy_kh wrote:
...

Somebody told me I have to use Fusion 360. Watertight design could be achieved only with a right tools (I 100%+ agree with this statement).

---

While it is easy to create "watertight" models with Fusion, watertight models can be created with ANY of the 3D modeling software I have worked with in approximately 35 years of use.
That includes, but isn't limited to Inventor, SolidWorks, ZW3D, Fusion, Blender, Meshmixer, SideFX Houdini.

The problem usually sits in front of the keyboard 😉

On that basis and with that background I wholeheartedley disargeree!

[vitaliy_kh]

Let me make things clear.

Yes, the enire design consist of two separate parts.

And yes, both parts have an opening on both sides.

So each part is not considered as a watertight. I am sorry, I did not use a correct terminology.

But here is a thing.

I printed and tested only horisontal portion. I pulugged the round pipe portion and

filled in with water. Unfortunately my printed part was leaking water in many places.

Most leaks were around transition portion from round to flat.

Now to my eyes both parts looks like a design with solid walls but slicer is trying

to clese many gaps. I am not sure if this is OK or this is still a problem.

And that is why I am looking for the advice ether how to make sure my design is

OK "as is" or if it is a problematic how to fix the problem.

PS.

I am experienced EE. I did created and printed few custom parts but this is my

first design which will be exposed to water with near zero pressure.

However my attempt to print this part did not survive the very simple test.

Printed part was designed in TinkerCAD.

I did not printed out the part redesigned in the Fusion 360 because I have to

make sure it will not be leaking (I hope so).

[hfcandrew]

haha omg this is the first time I've seen the literal term of watertight (doesn't physically leak out water) being mixed up with the colloquial CAD usage of watertight (is solid).

Anyways if its physically leaking water, design your male end bigger. Right now you have your male and female ends the same diameter, that will not make for a tight plastic fit. If its easy to put together then thats no good.

Some pictures of your prints of where it was leaking would help too.

[vitaliy_kh]

First - I am sorry, I used a wrong terminology (I am EE, not ME).

Here is a portion of the horizontal pipe where most (if not all) leaks occured

(a transition portion from round hose connector to the flat horizontal pipe):

The above picture is from Fussion desin but initiallyboth parts were designed in TinkerCAD.

Vertical port was OK and passed a leak test. But horizontal part was badly leaking.

For the test I plugged round pipe with cork and filled part with a water. 

I did not print out a part redesigned in a Fusion because I don't want to waste material and time.

That is why I am looking for the way to test parts in a Fusion (or whatever other SW if this is required).

I would like to make sure parts will not leak before printing them out and test in real life.

Just in case, here is a picture of the entire assembly:

Here is a horizontal part:

And here is a vertical part:

For the fitting the offset of the eliptical hole on the bottom of the horizontal pipe is 0.1mm

Finally these 2 parts will be glued together with epoxy.

I am planning to use ASA filament for 3D printing.

This is a custom drain for the washer machine.

The water pressure will be near atmosferic (it could be a bit higher because

the pump is really strong).

Per request fro @TheCADWhisperer the entire Fusion design file and both STL

files are attached in my responce above.

[TrippyLighting]

---

@vitaliy_kh wrote:

...

Printed part was designed in TinkerCAD.

I did not printed out the part redesigned in the Fusion 360 because I have to

make sure it will not be leaking (I hope so).

---

If the design printed without problems, then it doesn't matter what software was used to design it!

What matters is that those areas where the 3D printed parts interface with the dish washer are designed with the right clearances/tolerances.

It also matters that the clearances between the two 3D printed parts are correct for assembly and gluing (you need a bond gap for epoxies) and that there is enough surface area between the parts, so the glue can do it's work.

There is no way I am aware of to verify physical water tightness using just software. That can definitely NOT be done with Fusion. Print & Test!

Nothing beats real physical validation, physical prototype work with real materials. Any simulation work is just a precursor to that and might serve as a way to reduce physical prototype testing, but it is not going to replace it (yet).
At least not in my almost 40 years of professional experience in manufacturing a with 34 of those as an engineer!

Print & Test!

[TheCADWhisperer]

@vitaliy_kh 

You have done a fairly good job - I would model differently and I think I need to provide some information about the Ellipse curves, but this does not relate the leaking issue.

Not having seen the actual leak - my first guess would be print layer orientation.

And,

I doubt I would trust a FDM printer to print watertight layers.

Show a picture of the leaking.

Edit: Just went back through the wall of text in this discussion and found this that I missed earlier.

Most leaks were around transition portion from round to flat.

Beginning to reinforce my original guess that this is a layering (fusing) issue. 

Check the slicer too.  I would want to see the toolpath of the printer.

[vitaliy_kh]

I am new to the Fusion (and mechanical) designs but I am very experienced EE.

I tryed hard to design my parts fully parametric and I guess, I achieved this goal.

I cannot provide a photo of leaks (all that leaked parts already trashed).

Print orientation is this (this is the only way how part could be printed because

of size. and this is why I need two parts eventually gluid together):

This picture is from an Orca-Flashforge slicer with only Gap Fill option is turned on for visability.

To my eyes it looks like the design has gazilions of gaps whic slicer is trying to close.

I mifgt be wrong but this does not look right.

Is my asumption correct (very unfortunately if "yes")?

As I already mentioned, the majoryti (if not all) leaks happened around transithin

from round pipe to flat pipe:

Strange, I read about a lot of successful storied when people 3D prinded

very solid parts with FDM priners which did hold water very well.

(different vases and other things).

@TrippyLighting 

So, what my options are?

Should I try to print a horizontal portion one more time and see what happens?

Is it anything I have to tune up in the Slicer SW before printing this part one more time?

Note: The leak happened just with single solid (I though it was solid) part.

Not where two parts supposed to be glued together (unfortunately I did not get to this point).

[TheCADWhisperer]

@vitaliy_kh 

I am not familiar with that slicer, but the image does not look right at all.

[vitaliy_kh]

This is a Flashforge Slicer wich is a branch from very the very popular Orca Slicer.

(My 3D Printer is a Flashforge Advanture 5M Pro).

Well, do you think this is a design problem or problem with slicer settings?

If this is a design problem what I did wrong (I think you examined my design)?

Could you please advice how to fix a design?

My design approach was to create all User Defined values, use a 2D sketches

as much as possible and finally Extrude them into 3D model.

I tryed to create a single Body instead of creating just differen Components

and the join them together. This design approach came from really good

Youtube lessons how to use Fusion in a most efficien way. 

If this is a problem with Slicer configuration could you please advice

what parameters are most critical for my case and how to tune up a slicer?

[vitaliy_kh]

I found a Meshmixer tool from Autodesk and checked my design.

It did not find any problems at all (promisible).

But the suggested printing orientaion is this:

So, which orientation should I use?

(I was not even thinking this way.)

If I turn this part 180 degree I might be able to print both

parts as a single body. Is this something to consider or

should I print bot part separately (my original intention)

and glue them together?

[vitaliy_kh]

OK, I played with the Slicer settings.

Changing Wall Generator from Classic to Arachne drasically chaged number of Gap Infills (from 19% to <0.1%).

Changing other parameters almost did not change (unfortunately) a picture.

I would like to get rid of Gap Infills all together but don't see any good way how to remove them.

Any ideas are very welcome.

There are few Seams.

The location of this Seams makes no sense at all.

Here are few Seams shown in a Slicer view:

And here is the same place from CAD (with related sketches visability turned on):

The shape of the biggest Seam follows the shape of the elliptical connector

but the location of this Seam is in the middle of nowhere.

What is the reason for Slicer for generating this unwanted Seam?

Is it anything in CAD and/or Slicer could be done to prevent this Seam generation?

[TheCADWhisperer]

---

@vitaliy_kh wrote:

There are few Seams.

The location of this Seams makes no sense at all.

The shape of the biggest Seam follows the shape of the elliptical connector

but the location of this Seam is in the middle of nowhere.

---

@vitaliy_kh 

I am sure it is obvious to you - but I don't know what you are referring to.

[vitaliy_kh]

These dark lines on the green image (first one) is a Seams locations

reported by Orca-Flashforge slicer. Second dark gray image is the same

area from Fusion. The location of these Seams are in the long flat area.

The shape of the biggest Seam seems to follow elliptical portion of the

round pipe connector (but this could be just a visual effect).

Anyway Seams are on the very long (210mm) strech of the flat pipe.

I wonder what the reason for these Seams in this speciffic area.

To my eyes it makes no sense at all.