Why can't Shapeways print straight?

Hello all,

Below is a picture of a print I received today next to a spray can. As you can see it is very bent on one side, towards the top.



The column that is bent is 1cm thick so it has plenty of support. I have no idea why it has done this.

I noticed the side that is bend is smooth, whereas the other sides are rough.

Is this a print error, or does Shapeways sand down models (and have gone overboard on my print!) ?

Philip

Hi, Philip.

It looks to be Frosted Ultra Detail. Take a look in the Post Production Techniques section of this forum about how to correct bending of this material when it's cleaned with bestine or acetone.

My experience with this FUD material is a little short, but I have good customers that works and clean this material and they have found answers to this type of bending here.

I think is warm water inmersion that correct it, but is best if you look in the forum.

Regards

Luis

Thanks Luis, and yes it's Frosted Ultra Detail

At first I thought it was simply bent too, but if you look at the photo: the left side is pretty straight, while the right side is curved. Looking at the post from the top I noticed the square cross section (or what is supposed to be a square cross section) is distorted into a rectangular shape. It looks like the printer started off with square and then changed the cross section as it worked its way up the post.

This is not a simple bend, but a printing or finishing error!

Philip

Hi, Philip.

Yes, It couldn't watch well in the picture.

If you haven't done yet, send a complaint to Shappy Team service@shapeways.com with some photos, number of order and wide description.

My experience with the S-Team in solution of this type of issues is an 11 on 10. And they are very kind.

Good luck!

Regards

Luis

Thank again. I've contacted the Shapeways team and they asked for some photos, which I've just sent.

Seems like a very odd error, perhaps I had a dodgy machine make mine?

I'm also curious why quality control didn't catch it?

Philip

It would be rather an exceptional case for a model to be distorted that way by the printer.
I suppose it's possible, but other items in the batch would have been affected the same way.
Can you provide some additional photos at other angles and/or on a different backdrop?
I think the curvature of the paint can is complicating analysis of how the model itself looks.

The two prints side by side;



The posts (on the left of each print) are distorted, but also the flat surfaces are a little bit out too.

When I hold them up to the light, and look down the post (to get the light to reflect off the surface so it is easier to see), you can see the curve of the post is far more on one side that the other. The post tapers towards the top and is skewed to one side.

Hi Philip,

Are these parts hollow? The second photo has some odd "crystal growth" in the parts that looks like residue from wax support that wasn't cleaned out completely - suggesting it is trapped in a narrow cavity. If that is the case, your distortion could be an artefact of the cleaning process, where they bake the parts to get the wax out. I'm not sure how exactly, but it could be your parts softened a bit in the heat, and with the internal void "pulling" on the sides it caused the squared post to shrink inwards. If this is the case, you might need to add more material inside to stiffen things, since I doubt the existing parts could be "unshrunk". I've had simple box shapes come with inwardly warped sides, but these were easily straightened with a dunk in hot water followed by pressure against a flat surface. The thicker the walls, the more resistant things are to distortion.

You might want to lay the parts on some graph paper and hold the camera directly above them at a reasonable distance (to remove any close-up "fisheye" and "perspective" distortion from the pic) to get an accurate image of the curvature when you contact SW. (Your second pic has perspective distortion because the camera was off to the side instead of directly above the parts...)

Hope this helps!

They are not hollow. The crystal growth is *in* the material, it's like 3D crystals, or multiple layer crystals.

I'll take some more photos when I get back later.

If they aren't hollow, then they should have been completely transparent. My guess is that indeed something went wrong during printing, which also caused these 'crystals' to grow.

The models from the last picture have what we call "Spider web" .
That happen when a model changes temperature too quick
I would advise you to heat up your model (70C max) until the model becomes completely white and let it cool down with room temperature.
This should solve the problem!


Thanks

Wow - solid? Then I think you need to contact SW before you do anything else to them. (Another recent post about some trains mentioned that SW were unable to help because he'd already "processed" his parts by painting them before complaining...)

Warped resin is fairly common in the model-building fraternity, and heat is the usual cure. I believe the DG for FUD says it is only temperature-safe up to 80 degrees C, so if parts get warm they soften and can deform. Trouble with de-warping by heating is you can easily make things worse!

That kind of warping I think you can't solve it.
In the first picture you can see that the left side of your model is straight but the right side is not.
This means is not a warping because of post-process (temperature problems) but a printing problem.
About the "spider-web" you can easily remove it with heating up your model like i explained in my last post.

Cheers!

Yeah, I'm not touching them until Shapeways gets back to me.

Here is a shot looking down the post;



And here is one showing the amount of missing material on the larger flat surface. The area ringed in red is supposed to line up!



The grey part was also printed by Shapeways (sprayed grey) and came out pretty good. Not perfectly straight but good enough to test out the concept and get a feel for the product.

I'm sorry but I forgot to say that I work at Shapeways!


Thanks!

Shapeways have concluded their investigation and;

"This should never have been shipped out.
The bending is because the platform was not that clean
The parts being cloudy are a result of putting too warm models in a water"

They are shipping out replacements.

When I receive the replacements I'll post pictures.

I too have seen this issue with FD/FUD. In fact, I had two prints of the same part in the same order- one straight, the other bent as shown with the original post.
The problem? The bent part was set on end in the printer, while the straight part was laying flat. Simple if you think about a stack of pancakes... a tall stack of small pancakes is unstable, while a short stack of wide pancakes sets quite well.

I'm sure it's also affected by heat during build. If I understand correctly, FD/FUD material is heated when sprayed out, so there must be issues as it cools on the bed plate.

But regarding the print orientation-- we have no control over how Shapeways orients parts, and the only recourse is to complain and get a reprint. Bad for creators, even worse for our unwitting customers who have no idea of the process randomness they're being subjected to.

Thanks for the info patmat2350. The model is standing up in the file, perhaps next time I'll lay it flat and see if that results in a better print (I didn't know Shapeways can rearrange things?)

Replacements are being printed as we speak...

Best of luck!
There's no guarantee that the technician won't reposition the model in random ways to fit with the other parts in the job-- sort of like 3D Tetris.

But from laziness, it's likelier that if you draw it horizontal, it might stay that way...

Nope! My most recent design for a "flat" thing was stood on end in the printer! (The layer lines are a dead give-away...) Until such time as they give us the option to specify orientation, the only way to guarantee it is to design a VERY expensive model that exactly fills the print tray! (And that only works if the tray is different dimensions in the three axes...)