Shapeways FM switchers: Cleaning FUD

[Alwyn Cutmore]

Chris,

Mark4 Designs does 3D pieces as well but they are only broken down for ease of pattern making and to set the item up to give the best resolution and finish. The great part is that they can then be cemented to-gether using liquid cement.

If a N Scale pattern would cost over $500.00 I would go he! Out here I believe the going rate is about $250.00 for a N Scale Loco Shell. I will try and take some picks to-morrow and put them up for you to see the type of quality that can been achieved.

The other way to achieve is to do what Bryan has done and insert brass overlays. The is the way Dick Billings got that wonderful effect on the PRR Round Roof Box Cars. The brass etch he did was paper thin. I do believe that some of the patterns for Kaslo shops were done by 3D prototyping that was polished.

I am afraid me and Shapeways will never shake hands. They may be great blokes but I do not believe what they are doing is right for current modelling standards. Some of what they produce have more lines than a teachers punishment.

Regards

Al

[Lemosteam]

All maybe I can help explain the overhang issue- as each layer is printed it is very thin, if there were nothing there to support it in its design intent state then it would just flop down the next layer would hve nothing to adhere to.  So the machine builds a layer of wax (or rib in a fluid RP) there tio support it.

Matthew, not criticizing, here, I applaud your work.  It may have been better design the model without the hood or the roof, at least from the model side's perspective, and print it with the walls vertical.  One  could add some break out cross bracing to keep the model from warping in the bed.  If there were any warp issues with the roof or hood, they woiuld be reset when the modeler glues them to the sides, like a cap.

The pictured bed stack example is visually missing one thing- in fud there will be wax surrounding every one of those surfaces on each designer's parts, in a traditional lazer/fliud RP process, the PR software would add thin ribbed support structure all the way up thru each part until any horizontal surface is supported.  These ribs are a B-itch to remove from the intended surface post printing, i've done it.

I'm afraid that until someone comes out with an affordale machine like Finelines uses, say in the $2000-$3000 range, where the user can control the orientation of the designs in the vat, manage the depth of the build, and not have to worry about other peoples designs, over time the investment would pay itself off.  it is still unclear to me how much post print cleanup that Finelines does or how clean the model is after it comes out of the machine.

[bbussey]

Not so much an alternative to Shapeways as an alternative to Projet (FUD). I haven't heard of another way to remove the wax (although the results might be better if the technicians were a bit more patient). That said I've received a handful of slightly 'melted' models; I requested reprints and then repaired the originals using 'old fashioned' modelling techniques (brass wire and styrene sheet). That's not always an option of course, but when it works out you end up with a free car/locomotive out of the deal...

Bestine (heptane) dissolves the wax.  I'm tempted to ask one of the contractors to send a test part "untreated" to see if using Bestine to dissolve all of the wax is a viable option.  That would solve the warping, but does nothing to alleviate the uneven finish on some of the surfaces.

[bbussey]

Bryan, what company do you use for prototyping parts?

Dan, the best thing to do is shop around for ProJet3000+ contractors to find one that will fit the need.  The FD and FUD Shapeways designations are really the HD and UHD designations, and XHD is a higher resolution than those.  Shapeways uses a contractor in the Netherlands that has a ProJet3000 (no plus), so they are unable to render at the XHD resolution.  But there are a ton of contractors in the States and Canada.  The key is finding one that will charge reasonably and render the parts in a timely manner - given that these types of model-rendering jobs would be the lowest priority juxtaposed with higher-volume time-sensitive requirements of higher-paying customers.  I currently utilize the services of one in Wisconsin, but I've had some recent re-do's that are occurring more regularly, so I may start exploring other contractors again.

FYI, there also are other PerFactory contractors in the States and Canada.  But they are expensive.  There is one three towns away from me, and I took some product there for them to review, but they were pricey.

[sirenwerks]

The other way to achieve is to do what Bryan has done and insert brass overlays. The is the way Dick Billings got that wonderful effect on the PRR Round Roof Box Cars. The brass etch he did was paper thin. I do believe that some of the patterns for Kaslo shops were done by 3D prototyping that was polished.

I was thinking about doing the brass overlay just last night, to model the Burlington's unique Pacific Railway Equipment-built Silver Pendulum coach.  It has side curvature similar, but not as drastic, to the Amtrak Amfleet cars and using Shapeways to create a skeleton to support the curve of brass sides and provide the smooth roof and support for etched ends seemed a great solution.  I'm hung up on how to secure the sides to the Shapeways material though; which adhesive would be enough to resist the tension created by the curvature and provide a tight seam along the horizontal edges to minimize the roof seam best.

[bbussey]

This entire conversation is still missing one more point.

As an example I'll use ESM's White Tower building (which is not done by Shapeways but the printing technique is similar).  I don't have time to take example photos so I'll just describe it.

It is true that bottom surfaces of parts which are supported by wax will have rough surface, but it is not only the bottoms. If any part of the object protrudes beyond the surface under it, the wax will have to be built up to support that part. For example if you have a flat vertical wall of a building but in certain areas there are window sills then the wax will need to be built up during printing to support the sills. That wax will also be in contact with the vertical wall surface and will cause the solid vertical wall to have a rough texture  under each sill and smooth elsewhere.   So there are more problem areas than just the bottoms of surfaces...

Just to clarify - while the ESM White Tower is not rendered by Shapeways, it is rendered using the exact same process and materials that Shapeways contracts for its FD/FUD products but at the highest resolution.  Shapeways' contractor uses a ProJet3000.  The White Towers are rendered on a ProJet3000+, which allows for a third more-fine rendering option that is not available to Shapeways.

Regarding the orientation - the interiors are built upright, as all the detail fits on the floor and the floor sits directly on the build-tray while being rendered, so there are no wax-related issues.  The exteriors are built upside-down, so the "rough" surface is the roof top and simulates a tarred-gravel finish.  The tower and roof trim are unaffected because they are small detail areas.  The "rough" effect only seems to affect large surface areas.  The orientation to the printhead also matters when it comes to the crispness of the lettering and window treatments.  I don't remember offhand whether the exteriors/interiors are positioned parallel to the X or Y axis, and which way the tower treatment has to face, but the contractor knows.  There also is a distortion factor that varies from ProJet to ProJet, depending on how well they are calibrated.  For the machine that renders my parts, parts have to be stretched 1.0055 on the X axis and 1.0025 on the Y axis.  That doesn't sound like much, but that works out to about .020" for an 85' model such as a passenger car or auto rack.  It becomes more critical when using the model as a core to laminate etched parts to, such as the KV caboose.  The photo I posted up-thread was the first attempt.  The core body was short by about .005" and the roofs by about .002", which was enough for the brass insets not to fit.

[bbussey]

I was thinking about doing the brass overlay just last night, to model the Burlington's unique Pacific Railway Equipment-built Silver Pendulum coach.  It has side curvature similar, but not as drastic, to the Amtrak Amfleet cars and using Shapeways to create a skeleton to support the curve of brass sides and provide the smooth roof and support for etched ends seemed a great solution.  I'm hung up on how to secure the sides to the Shapeways material though; which adhesive would be enough to resist the tension created by the curvature and provide a tight seam along the horizontal edges to minimize the roof seam best.

If you use Bestine to clean the ProJet core, any super glue should do.  I used the CA+ from Gorilla Glue, which supposedly has a little flex to it.  But don't design it to be your final model.  Design it to be a master from which to make polyurethane castings.  The key is to have the master survive long enough to make molds from.  With the urethane castings, you won't have to worry about the disparate materials separating over time, and you save a couple of the first-generation castings for future molds if so desired.

You also should pre-curve the sides before laminating them to the core, in your example, which would reduce tension.  That is what was done for the Keyser Valley caboose roof parts.  I used a large X-Acto knife blade as a rolling pin on the etchings until the curvature matched that of the RP part and then cemented them in place.

[mmyers05]

All maybe I can help explain the overhang issue- as each layer is printed it is very thin, if there were nothing there to support it in its design intent state then it would just flop down the next layer would hve nothing to adhere to.  So the machine builds a layer of wax (or rib in a fluid RP) there tio support it.

Matthew, not criticizing, here, I applaud your work.  It may have been better design the model without the hood or the roof, at least from the model side's perspective, and print it with the walls vertical.  One  could add some break out cross bracing to keep the model from warping in the bed.  If there were any warp issues with the roof or hood, they woiuld be reset when the modeler glues them to the sides, like a cap.

The pictured bed stack example is visually missing one thing- in fud there will be wax surrounding every one of those surfaces on each designer's parts, in a traditional lazer/fliud RP process, the PR software would add thin ribbed support structure all the way up thru each part until any horizontal surface is supported.  These ribs are a B-itch to remove from the intended surface post printing, i've done it.

I'm afraid that until someone comes out with an affordale machine like Finelines uses, say in the $2000-$3000 range, where the user can control the orientation of the designs in the vat, manage the depth of the build, and not have to worry about other peoples designs, over time the investment would pay itself off.  it is still unclear to me how much post print cleanup that Finelines does or how clean the model is after it comes out of the machine.

Well said.

To be honest your idea of removing the roof probably makes sense. What always holds me back is this: since I can't specify the orientation, I am always "playing the odds" so to speak. Consequently, it's very difficult to justify designing to a specific orientation only to have the model rendered in another (to end up with the problem that bbussey described earlier). Basically I've prioritized ease of assembly (minimizing extra parts), structural integrity and reducing volume over improving surface finish (since it's so difficult to control surface finish with the tools that Shapeways currently offers). 

In addition, in my experience, people (not necessarily the people reading this thread) generally react much more negatively the big flat surfaces being rough and/or warped than they do to 'roughness' in limited places (even if the aforementioned problems are relatively easy to correct). Generally, in my experience, when the roughness is superimposed against details, many people find it less noticeable. However, when the nice shiny roof looks like it has been made of tar paper, or the components are  warped - even if the sides are rendered darn near perfectly- I get angry emails from customers complaining about their 'crummy shells.' When only a few surfaces are rough (like in this case), I get "do you have any tricks for cleaning up the surface?" responses instead. I have to say that the latter are much more pleasant to answer...

You guys also have to remember: I never intended for these to be high quality craftsman kits (with many etched components, etc.) to begin with. As many people have mentioned - there are much better ways of achieving a near perfect finish if that was the goal. I make my designs because I enjoy the digital design work. I place them up for sale because people ask me to do so. If someone came to me with a suitable bankroll and said: "make a craftsman kit for me" then I'd spend more time worrying about etching/Finelines etc. As it stands, I made these in my spare time because I happen to like the look of the H-10-44 and because a few were owned by the Rio Grande. I bankrolled the project myself and put as much work into it as I found enjoyable. Frankly, that is what the hobby is all about for me...     

Bestine (heptane) dissolves the wax.  I'm tempted to ask one of the contractors to send a test part "untreated" to see if using Bestine to dissolve all of the wax is a viable option.  That would solve the warping, but does nothing to alleviate the uneven finish on some of the surfaces.

Hmm, I suppose I phrased myself poorly: I meant methods for removing the "bulk wax" so to speak. Do please let us know if Bestine is sufficient  to do the entire job.

I'm not sure, but I think what Mark uses might require the 3D model to be broken down into small pieces. Not saying it couldn't be done, but the drawings would probably need to be re-worked.

Ever think of asking the designer to send the file to Finelines and get a quote for micro green. (I'm guessing over $200 tho)

Bingo - EnvisionTEC (Mark) has its own pitfalls. They aren't insurmountable necessarily but I would need to do some considerable redesign work...

[sirenwerks]

If you use Bestine to clean the ProJet core, any super glue should do.  I used the CA+ from Gorilla Glue, which supposedly has a little flex to it.  But don't design it to be your final model.  Design it to be a master from which to make polyurethane castings.  The key is to have the master survive long enough to make molds from.  With the urethane castings, you won't have to worry about the disparate materials separating over time, and you save a couple of the first-generation castings for future molds if so desired.

You also should pre-curve the sides before laminating them to the core, in your example, which would reduce tension.  That is what was done for the Keyser Valley caboose roof parts.  I used a large X-Acto knife blade as a rolling pin on the etchings until the curvature matched that of the RP part and then cemented them in place.

Excellent, Bryan!  Thanks for the problem solving.  I didn't even think about molding them.  Not that it would be a hot seller, only three prototype cars were made; one each to the Burlington, Great Northern, and Santa Fe.  I don't know of any existing photos of the latter two to know how they were painted.

[Alwyn Cutmore]

John,

John, I know what the problems are with RP and where the problem with the shapeways models stem from but until modellers understand  what RP was designed for there will always be a problem with lines. I can understand where shapeways is coming from, they just want to sell stuff, they will not give a hoot  what it looks like just so long as it sells. Their prime objective is to control what is coming out of their machine and get paid for it. They are not the only ones out there doing this. My prime objective is to buy a model that can be used without too much effort. A painted injected body for an E8 on the bay this morning is $35.00 therefore I would estimate that a good!!!!!!!!!!! craftsman kit with etches of a PRR specific E8 would cost me around $55.00 to $65.00.  The Kaslo Shops models are produced off RP patterns, and then pressure cast. They are around $75.00 CDN and the only thing you have to do there is remove the flash. FNS wagons I believe are pressure cast and are boxed at around $13.00. Basically what I am saying is that shapeways should do what they are there to do and that is produce a rapid prototype and then let the industry (craftsman kit manufacturer) get on with doing their job. That is what happens here and the resulting models are a lot better. Because of the size of the market here we pay just on $100.00 per kit. Understanding that I do not mind. I have been casting in resins now for just on 36 years and I understand what this side of the industry is about.

Matthew, I understand where you are coming from but I think you would do yourself a much better service to go that one step further and produce from what I can only perceive is a lot of work and skill which in the end deserves a better outcome. I was genuinely looking at the models for my PRR but the finish put me right off. The Mark4designs items are not done en-masse like what  has been shown but as individual units. I have seen his operation and they still have some striation but if it is for a pattern they are polished out. Warpage can be prevented by a matrix inside the model that is removed after all prototyping and polishing is finished. One more step is the difference. I think you said it all with

What always holds me back is this: since I can't specify the orientation, I am always "playing the odds" so to speak. Consequently, it's very difficult to justify designing to a specific orientation only to have the model rendered in another (to end up with the problem that bbussey described earlier).

. That is the Mark4design difference.

Regards

Al

[bbussey]

Excellent, Bryan!  Thanks for the problem solving.  I didn't even think about molding them.  Not that it would be a hot seller, only three prototype cars were made; one each to the Burlington, Great Northern, and Santa Fe.  I don't know of any existing photos of the latter two to know how they were painted.

I wasn't necessarily thinking along the lines of creating an item to be sold, although that is an option.  It's definitely a viable method for personal use as well, even if you only need a couple of them.  And you can always offset costs by offering extras to fellow modelers who may want some.

[highway70]

There is a product called Mr Surfacer, which is used by many plastic modelers:     "Gunze Sangyo’s Mr. Surfacer is a primer, that's available in both spray can and bottle. What is Mr. Surfacer? It is a thin filler, almost the consistency of paint. It works great for filling small gaps, seams and scratches. It is easy to
apply and quickly dries to a smooth surface without bubbles."

I have never used it, but wonder if it might be useful to clean up the imperfections on FUD models.

fortcrookipms.com/ckfinder/userfiles/files/Mr%20Surfacer.pdf

[katfudgirl]

Hey guys, I want to introduce myself.  My name is Kat, and I do all of the post processing for FUD/FD models at shapeways (for US, Canada and like 50% europe). Basically I'm the girl who unloads your model from the printer, removes the wax, sorts, cleans (q-tips and everything) and inspects EVERY single model before it leaves. I also do some of the tray planning and model orientation.

I'm here (off the record from SW) to answer questions if you have them.

oh, and the tray view from page 1 is not a FUD build. We never stack things in the FUD/FD  trays because of the effect the support material has on the crystal.

we/I do apply heat to remove the wax, but at a temperature low enough that it doesn't effect the plastic. if you ever receive a model that is warped because of this, it should be reprinted for you. On that note, I have been doing a ton of research on alternative wax removal processes, but I have not yet found a successful solution that doesn't degrade the quality/transparency of the plastic. I'm working on it though, and up for suggestions!

[daniel_leavitt2000]

Thanks for joining!
I have used Motsenbockers Lift Off to remove the wax. It works well and does not mar the surface of the plastic. Others have used similar products with great results.

Now for the questions:
1. Can we request specific orientation in the future?
2. Will you be offering higher resolution than the current FUD in the future?
3. Can we request the item be shipped with the wax in place so that we may remove the wax ourselves?
4. We have had issues with warping on flat parts. Creating a flat kit would eliminate many if the issues we have encountered. What is causing the warping issues? Is it an inherent problem with the printing process itself or with the heat applied after? Would leaving it in the wax carrier for a period of time fix the warping issues?

[Freight Train]

Question?
        Don't mean to sound stupid (really!!) but reading the beginning of this thread about the Shapeways FM shell and how it's produce and the final out come is one of it's side being bad makes me wonder why the shell can't be produced bottom placed first?? This to me would be the most logical solution providing that this product and manufacturing can be done in this way. The bottom would be only seen when handled and not during display and/or operation. Am I nuts to think this????