Resolving print quality issues with Shapeways

[Quarryscapes]

Thanks - can I ask how well the nylon takes paint?

 

Quite porous so the paint doesn't spread far. Finish not as good as FUD, but more consistent (i.e. the whole thing has the same patternation, not some bits being great and others horrendous). 

[Quarryscapes]

When I started to using 3D printing I was told that the quality of the finished item is directly to 3 factors

1. production drawing (needs to be built from a solid block)
2. CAD system (we used AutoCAD we were told that the we might have problems with turboCAD)
3. step value of the printer.

The chap who does our printing uses a machine a step value of 0.1mm which gives a very satisfactory finish for 7mm.

 

we draw every thing as 3D .dwg file then convert them to .stl files within AutoCAD 

 

Marc

 

The software has no bearing on the finish of a 3d print (unless there's an error in the file which screws things up, and yes Turbo CAD will give you plenty of those!) STL files simply store the surface, they don't store any solid data like what would be needed if you were to have something produced by CNC or other CAM process, for example injection moulding. 

 

Depending on the machine type, the layer thickness doesn't matter much either  to use Shapeways as an example. 

 

CWRail is 16microns, same as shapeways FXD. FUD is 30ish (variously reported). BHDA is 50microns, but smoother than all the previous. 

 

FUD, cleaned, tidied up, primed and several coats of livery colour, transfers and varnish:

 

 

BHDA, Supports removed and Mist coat of primer - this is the worst side too, where all the supports were, all roughness here is entirely due to my poor surgery to remove them. 

 

[billbedford]

The chap who does my prints uses nylon  

 

 

Not according to his website. He says he has an Objet Eden printer which utilises a photo acrylic resin.

[MarcD]

i have not had anything done for a while.

 

Marc

[SpoorObjecten]

Argh We reprinted the models within 48 hours and shipped them yesterday.

Unfortunately UPS needs 2 days to ship and thus won't have it delivered till Monday.

 

Hope the parts will turn out better!

[cornelius]

Bathtub: (sorry, i didn't add any visuals on the side of this one!)

Doesn't this image just illustrate that the model itself is not really suitable for FUD?

 

Not a criticism of Shapeways or the OP - with a few changes to the detail by the designer most of the need for support wax up the sides could be eliminated. There is then a bit of detailing / scratchbuilding work for the purchaser but easier than putting the effort in to rescuing a bad job.

 

Sure there may be times when for whatever reason the performance of the printer isn't up to scratch, but understanding and working with the limitations of the material is the first step.

[Guy Rixon]

It would probably print better if the roof were left off the print. 

[SpoorObjecten]

Definitely a good point cornelius!

The model itself is great, however as you point out, with the technique used to print FUD you will notice that by making a few changes, it will come out of the printer way better.

 

Next to keeping an eye on the guidelines for thickness, the printing technique itself is very important to keep an eye on too!

This is why a steam locomotive with a lot of rivets (they stick out and need to be supported) won't turn out as good as a streamlined train.

However when you print a steam locomotive using High Definition Acrylate, those rivets are suddenly not a problem anymore (different printing technique)

[Talltim]

Not a criticism of the OP, to my mind most affordable 3D printing isn't really suitable yet for models where acuracy in vital in scales smaller than H0 (and then only just). There seem to be a lot of N scale models on Shapeways and obviously their small size and volume make them relatively cheap, but I'm not convinced that most of them will print that well.:

[2mmMark]

I'm pleased to report that the reprinted items did arrive earlier this week and look very well printed.  All I've been able to do so far is clean them. There was a tiny bit of flash (not sure if that's the correct term form 3D printing) but it was not a problem. I was able to flick it off the print with the point of a pin.

 

When I get some primer on them, I'll post some photos.

 

Mark

[rue_d_etropal]

Tim, I would challenge what you say. I have only seen photos of the Nm models in CM, but they look very good, far better than most people could scratchbuild.

 

Those who can buld to that quality should build the models , but for everyone else, 3D printing is best method, unless someone is prepared to finance the cost of moulds etc for injection moulding.

[Dazzler Fan]

Wierd things can happen!

With FXD I have had one part printed through

another part, obviously where the support material

was not printed and missed by their QC.

 

Shapeways have always been good about re-prints.

 

Noel

[2mmMark]

Well, it's happened again. Not to same extent as before but it's not acceptable. The top half of the print is grainy and has a powdery finish while the bottom half is nice and crisp.

 

I don't know if this is a QC issue or checking issue. The loco body I got at the same time is much better.

 

This time I haven't primed the print so the photos aren't as clear as I'd like but hopefully they illustrate the problem. It has been throughly cleaned to remove residual wax.

 

I shall be requesting a reprint.

 

:-(

 

Mark

 

[MikeTrice]

I think your problem is all to do with overhang as SpoorObjecten explained earlier. From the pattern of support material on the step end the model has been printed roof down. which should in your case give a better print than the other way round. Personally I would seriously consider separating the roof from the body which would help considerably or better still break the model down into 5 parts (roof and four sides) as a kit of parts which could all be printed detail side upwards with minimum overhang.

[2mmMark]

I think your problem is all to do with overhang as SpoorObjecten explained earlier. From the pattern of support material on the step end the model has been printed roof down. which should in your case give a better print than the other way round. Personally I would seriously consider separating the roof from the body which would help considerably or better still break the model down into 5 parts (roof and four sides) as a kit of parts which could all be printed detail side upwards with minimum overhang.

Not my design, guv'nor but I take your point. I've done flat-backed resin coach castings as seperate sides and ends with successul results.

 

However, the reprints of my earlier problem items are excellent, which shows that an integral print can be to an acceptable standard. If the design is the problem, then Shapeways ought to advise that the results might be variable.

 

Mark

[rue_d_etropal]

just wondering,is this a new problem or one that has not been noticed by some people.I could ask someone I know who has some of these models, but as they have appeared in magazines I assume they are OK.

As I don't get prints done in FUD, I can not comment, but I have noticed that on the whole my WSF prints have got better. I have only used default orientation, but there has been a niggling problem since day 1. The stone detail on my inset track system is not very thick, but I did notice that sometimes it was missing a very thin top layer. It was not always, and sometimes this was on the base so was not so obvious. It did result in the stone pattern being not so defined, with some stones being joined. I put in a design fix which reduced effect of this, but with some new designs with some fine detail which printed OK first time, then in latest order some had fine detail, some did not. It is only a missing groove, which I can do myself.

When there was talk of having ability of print orientation, I considered this but it was not obvious which way I should set the print, and even then I was not sure if this was the problem. My feeling have aways been that maybe my items were near edge of machine box, but as it did not always happen it was very diffucult to monitor. Just wondering if anyone else has noticed something similar. I have only noticed it on my inset track sections. As these are thin objects they obvously don't waste space so possibly get squeezed into spaces. I may put in a small change, just in case.

On the whole with my new designs such as coaches I am keeping fine detail to a minumum. It is easier to add detail afterwards, and makes sanding down a lot easier.This does not take very long. Given what has been said above, it might mean that if I resize the models down to N scale, printing in FUD would be good option.

[Knuckles]

.........I have noticed that on the whole my WSF prints have got better....

 

Same here.  I haven't tinkered with the print orientation due to lack of knowledge.  If anyone here can offer advice that'd be good.

 

The last few WSF prints I have had were a lot better than older ones too.