3d print job any takers?

Hi all!

I'm working on a hypothetical extension of Questacon in Canberra, and plan on printing a model of questacon and the extension. Would anyone be interested in printing my models?

I'll give the job to the best quote
i'll attached a 1:1000 model of questacon to give an idea of the model's volume; attached skp file because stl was too big. I know how much it costs to print locally on a z-printer which is a lot! so i'd be interested to hear some other quotes

P.S i need these models by Wednesday, the 20th, so probably need it sent by the 25th. however im desperate enough to have it sent by overnight delivery on the 28th so it arrives on the 29th.

hope to hear some replies

its a sketchup file because stl was too big.

I imported the model and it scales as 86 mm x 56 mm x 26 mm high. (See attached). How big do you actually want it?

It looks like it should mostly print OK - I'm a bit worried about the posts that support the walkway canopies etc - they could be so fine that they may break off - especially as there will need to be some support material to hold up the overhanging elements, and it could be tricky to remove the support material without damaging the columns.

Edited 1 time(s). Last edit at 11/19/2011 06:51PM by julianh72.

---

Follow my Mendel Prusa build here: [julianh72.blogspot.com]

those are the right dimensions. It's at 1:1000 in real life scale. is it possible to scale up or down from a fully sized model in sketchcup? to print it locally they need it scaled down so its 1:1000 in real life. the point note about the column thickness and overhang is noted, I'll remodel detract the the overhang a bit and thicken up the columns. is it OK that it's a shell?

a picture of the bottom

here's the edited model, thicker columns, overhangs are separate from model so no 'scaffolding' required.

thanks guys, and thanks Julian for your speedy reply

Julian I just realized the dimensions you listed are slightly scaled up, were are originally 52.8mm by 82.2mm by 25.7mm.

Overhangs are not necessarily a problem per se, as the software will automatically add "support" material where necessary, which can be broken away very easily with your fingers. The problem is with very fine isolated parts (like 1 mm slender columns), which the printer will try to build, but is right at the limit of the technology.

I was rounding off the imported dimensions a bit. Yes, the software can scale an imported part; I just wanted to know how big it SHOULD be. I didn't want to replicate the infamous Spinal Tap "Stonehenge" set debacle! [www.metacafe.com]

When doing architectural models, for example, it can be easiest to just export a full-scale model, and then apply the required build scaling factor when it is time to print. A very common problem is parts which are saved in dimensions of inches (or feet), but imported in units of mm, and then you have a scaling factor of 25.4 (or 304.8) to deal with.

However, your newest mutli-part model looks very printable - I'll have a go tonight, and let you know how it goes. The parts should be able to be glued with plastic glue, Super Glue, etc. The model has a net volume of about 11 cc - which is about 50 cents to $1.00 of PLA (allowing for wastage, support, etc). At that rate, I'm prepared to have a go for my own entertainment. If the parts come out OK, I'll post a picture, and you can decide if you want to pay for me to mail them to you.

---

Follow my Mendel Prusa build here: [julianh72.blogspot.com]

hahaha omg! do you know how much that costs on a z-printer, try $80. that's made my day. although does it matter that the current model i sent is a hollow shell? should it be solid?
thanks julian!

Julian of course i'll pay you the cost to make a tester and for your time even if it doesn't work out.

Sketchy,

Is this a student project, or a real commercial commission? I am not sure my machine is up to commercial expectations for quality yet.

In general, hollow shells are probablyy better than solids for items like this - much faster to print, and much less material required. Basically, I would only use solid when you NEED it - e.g. parts which have to carry real loads, etc.

My machine is still "under development", so I am not charging people for small private / student jobs yet. Basically, I am happy to tackle some of these small "real world" commissions as part of my own learning and trouble-shooting process.

Once my machine has proven itself capable of producing "marketable" items (I think I'm I'm pretty close now), I'll think about charging a "reasonable" price for materials and my time. E.g. a mate of mine has asked me to design and produce some items for him. He'll get the first one or two for free, to satisfy himself that I can deliver what he wants. Once I go into "production mode", I'll charge him a fair price (still to be negotiated) for materials and my design time and my machine time. (All the time it is printing things for other people, I can't use it for printing my own stuff!) I will probably charge more for items which are to be sold as part of a commercial proposition than I will for personal / student work.

I don't expect to make any real money - this is a hobby for me, not a business.

Cheers!

Edited 1 time(s). Last edit at 11/19/2011 09:53PM by julianh72.

---

Follow my Mendel Prusa build here: [julianh72.blogspot.com]

It's a student project

Sketchy,

Just starting to take a close look at your model - the finest features are still VERY fine - 1.0 mm shells, 1.0 mm column stubs, etc. I'll have a go at printing it, but I think for best results using RepRap printing, try to keep the following parameters in mind:

1. 1 mm shells are certainly possible, but the model will be a LOT stronger if you can limit shells to say 2 to 3 mm thickness

2. The smallest single entity I can reliably produce is approximately a cube or similar about 3 mm to a side. (RepRappers with better tuned machines and more experience than me might be able to do better.) I can get finer surface texture (0.3 mm surface detail say) as long as it is part of a larger piece, but components which are smaller than about 3 mm tend to end up as just small shapeless lumps. You might do better using other fabrication techniques for these very fine components.

Anyway, my first attempt at printing is underway. I'll post some pictures in a couple of hours (as long as my machine behaves itself).

---

Follow my Mendel Prusa build here: [julianh72.blogspot.com]

I'd suggest making a 'double-sized' model - 1:500. At 50x80x25, you can double that and still fit on your Prusa. You won't have to worry as much about the fine detail as well.

Please post heaps of photos!

Edited 1 time(s). Last edit at 11/20/2011 12:25AM by Shibboleth.

---

___________________________________________________________________________
systematictechnology.net

Well, I had a bit of a catastrophe - I had a momentary power black-out with about 5 minutes to go! Anyway, I have posted some photos, so you can see how it was going.

Two things to note:

1. My "raft" lifted off the "Blue Tape" at three of the corners, deforming the base of the print. (Note to self: always replace the "Blue Tape" before a major print job!)

2. As I feared, the fine detail parts are just too fine to be printed on my humble machine. A better-tuned machine may be able to do a better job on fine detail like this, but I suspect that detail this fine is really pushing beyond what is practical with current RepRap technology.

Sketchy - another point to note - it looks like you took the top ring off the tower to avoid overhangs where you want little “windows”; don't worry about this - RepRap should be able to handle small overhangs like you had in your first model with no real trouble.

More photos in following e-mail.

---

Follow my Mendel Prusa build here: [julianh72.blogspot.com]

Some more photos.

I'm not sure that these printed parts are of "merchant quality" yet!

---

Follow my Mendel Prusa build here: [julianh72.blogspot.com]

nice pics! the main building looks good, but like you said the finer details didn't make it. the top of the drum wasn't finished because of the blackout? i think 1:500 might be too small for the reprap. however in addition to a 1:1000 model of questacon and my extension i also need a larger models to show greater detail.

the scale of these can either be 1:500 or 1:250, my choice.

@ 1:250 the columns would be 4mm do you think that's thick enough too be accurately printed?

actually i think a 1:500 model that is a simple mass with little details would contrast nicely with a larger detailed model...

i'll make a revised questacon with less detail for 1:500 at full size, ie 1:1 in sketchup. and a more detailed one as well.

just to clarify, 4 models in total, 1 questacon and extension @ 1:1000
and 1 questacon and extension @ 1:250.

i'm still working on the extension so won't have that finished for a few days. the questacon models i'll have done by 3pm tomorrow. also would 1:250 be too large to fit on your build platform?

P.S thanks Julian,

goodnight,

Andrew, AKA Sketchy

Andrew,

Yes 4 mm columns would probably print fine - the surface might be a bit rough looking, but they should be OK otherwise.

My biggest possible print is about 165 mm x 165 mm x 150 mm high. Height shouldn’t be a problem, but the 1:500 m model will just about max out my horizontal size limits. 1:250 would probably need to be split into several smaller models.

Also, please not that doubling the scale of a model multiplies the amount of material and the time to print by a factor of 8 (2^3 = 8). The 1:1,000 model was close to finishing after about 2 hours print time; a 1:500 version might be a 12 hour print or longer! (I might have to seriously think about charging for print runs that tie up my machine that long!)

---

Follow my Mendel Prusa build here: [julianh72.blogspot.com]

It's hard to tell from the photos, but is that object solid?

---

___________________________________________________________________________
systematictechnology.net

@Shibboleth, my understanding is that it's a shell, with the material on the inside being support material.

12hrs sounds like a bit of a marathon! and it would be 24hrs since it would be 2 models. i'm happy to pay the material/time costs associated with that, which i'd assume to be in the hundreds... But would you be happy doing it? i don't want to place undue stress on you.

here's my email: u3039182@uni.canberra.edu.au

Andrew,

To be honest, I don’t think I am ready to take on the responsibility for promising to print such large parts for you. My machine is very new, and I am still calibrating and upgrading it. I can get errors from time to time even in short prints (2 – 3 hours duration); the likelihood of getting an error-free 12-hour print of a part that fills the print bed on my machine right now is probably only 20% – 30% I would guess. (I’m not sure, as I haven't printed anything that big yet!) I don’t want you to commit to expecting me to deliver large parts for you, and me failing to deliver, leaving you with nothing to show!

Shibboleth,

The part printed as a hollow shell, but for some reason that I don’t understand, the skeining process inserted a full "diaphragm floor” about 10 mm above the base of the print (see photo 13), and then filled the space between that floor and the roof with 30% “honeycomb fill” material. There is no such “suspended floor” in the STL file, so I’m not sure what happened. I suspect it is related to the apparent “glitch” that I see in the roof (see attached screen shot in EasyFIT), which I suspect lead to the following error messages which were generated during the skeining process:

...
Slice layer count 16...
This should never happen, there is a hole in the triangle mesh, each edge should have two faces.
6749 [4366] [2373, 2374]
Something will still be printed, but there is no guarantee that it will be the correct shape.
Once the gcode is saved, you should check over the layer with a z of:
-10.9499689438
Slice layer count 17...
This should never happen, there is a hole in the triangle mesh, each edge should have two faces.
6749 [4366] [2373, 2374]
Something will still be printed, but there is no guarantee that it will be the correct shape.
Once the gcode is saved, you should check over the layer with a z of:
-10.6499689438
...

(There were a whole bunch of these errors.)

In my previous dealings with imported CAD files of complex geometry, “glitches" in STL files are not all that uncommon. It is generally preferable to work with native CAD files until the last possible moment, and then produce STL files which use parameters which are “tuned” for the actual fabrication machine. Generic STL files can be coarsely faceted, or can have zero-thickness surfaces, missing facets, etc, any of which can corrupt the desired outcome.

Edited 1 time(s). Last edit at 11/20/2011 10:34PM by julianh72.

---

Follow my Mendel Prusa build here: [julianh72.blogspot.com]

It might be worthwhile putting the STL file into Netfabb's cloud service - http://cloud.netfabb.com/. They have a free STL cleanup service which might remove some of the problems and help generate a better plot.

---

___________________________________________________________________________
systematictechnology.net

Julian thanks for trying it out anyway! i think the errors might have occurred because it was just a sketch up model. usually i resurface the model using cad span to make it watertight but this time i didn't. if i thought of doing this sooner i could have made a better printable model.

thanks!
Andrew