Jump to content

3D Modeling; The Future


Kabutoni

Recommended Posts

With shaking hands I opened the package.

 

post-188-0-14024800-1429161977_thumb.jpg

 

Wrapped securely in pockets of air in a too large box.

 

post-188-0-71462200-1429161979_thumb.jpg

 

The result of two weeks of intense learning and modeling.

 

post-188-0-95381300-1429161982_thumb.jpg

 

There is still some improvement to be done I see. Warping is indeed an issue, but nothing that can't be overcome with correct design solutions.

 

The tools of the creator and the creation in harmony.

 

post-188-0-41254800-1429161984_thumb.jpg

 

It won't be until a month that I have the time to assemble it and coat the model with primer to exactly see what details and design issues need improvement, but it's here! My mind has now accepted the fact that my own 3D modeling is good enough for more investment.

 

This is the future.

 
  • Like 2
Link to comment

A month! How can that beautiful product not make you pull an all-nighter? :)

 

The product can, but my wife is MUCH more powerful. xD

Link to comment

Looks good. I can't really see any warping, but the bottom part of the sides do look a  bit too thin. I'm looking forward to the finished product!

Link to comment

Looks good. I can't really see any warping, but the bottom part of the sides do look a  bit too thin. I'm looking forward to the finished product!

 

The sides are fine, as they have a 'T' structure for side support on the Tomytec TM-08 power unit, which also counts as a anti-warping device (I anticipated that with the design). The roof has warping, as I forgot to incorporate a structure like this, since it doesn't have to support a secondary device. I'll have to incorporate this concept with future designs. Other than this, it is pretty much done design-wise and could only see improvement in detailing.

 

You can make flat surfaces super thin, as long as you have support beams to prevent warping. I have some parts in this structure (which are actually too thin) that are only fractions of a millimeter, but still hold up. 3D printing is pretty amazing in this regard.

Link to comment

I got a Shapeways order today, Pullman 12-1 sleeper, SP commuter car, 12-axle flatcar, and U33C conversion for the Kato U30C.

 

I'll post pictures later.  Now I have to go off to learn to operate a laser cutter, so I can cut Ttrak module pieces out of 1/4" plywood, with an accuracy of 0.05mm!!!

Link to comment

this is absolutely necessary as they require tolerances of 0.05mm or just plain won't work! don't even try if off by a tenth of a mmm!

 

tolerance issues I've had making ttrak boxes were never the saw, always the #$@% tolerances on the thickness of the ply! its gotten really bad the last few years, changing quite a bit across a piece of ply.

 

jeff

Link to comment

I couldn't resist and took the model from its frame. After some trying and fitting, there are big changes to be made to my basic design.

 

- Warping: all big surfaces, the roof and sides have warping in their structure. It's not bad, as this can be negated with some heating, but it's not supposed to happen. I'd like to have them rigid, like the smaller surfaces. The solution to this is to thicken the material, but this wouldn't guarantee less warping;

- Fitting details: if not for the warping, all parts fit perfectly of course, but there is the issue with the plug-and-play element I'd like to have incorporated in the design. 

Another few small details that I need to improve on are the bogies. These need holes for the protrusions of the axles. I oversaw these;

- Acrylic glue: stupid me was thinking that acrylic would be able to stick together with simple hobby glue. No. It doesn't...

 

Anyway, apart from the glue and so in, I'm thinking of reverting the whole model to a solid state 3D print, as it has a lot of benefits in the end:

- Less glueing;

- Less chance of warping in general;

- Saves time at the assembly.

 

Downside would be that it'd use more volume, thus possibly increasing the price. However, the extra parts, like lights, underbody equipment can be incorporated within the central structure, actually decreasing the volume. I think this would be the best solution...

Link to comment

Toni,

 

How about some thin ribbing on the inside to help brace on the warping? Perhaps larger interlocking Ls on the edges to mate to the sides to the roof.

 

Jeff

Link to comment

How about some thin ribbing on the inside to help brace on the warping? Perhaps larger interlocking Ls on the edges to mate to the sides to the roof.

 

I was thinking about that, but also about completely merging everything together in a super dense structure. Complete efficiency of usage of space with a minimum chance of warping anywhere like this:

 

post-188-0-00664100-1429248138_thumb.png

 

This will a little bit more usage volume in the printing machine and will probably increase the price a little, but will overall generate a more acceptable result. The final product will not be able to be converted into a mould, but that is not my concern anymore, as I've stepped off that idea a little bit. It's maybe more rewarding to offer direct data for printing instead of risking the investment in badly made moulds.

 

EDIT: I have had the new data checked and it's actually roughly ¥200 cheaper, as I use less material!

 

EDIT 2: A cost-offset test with a 2-car print proves to be about 155% more expensive, instead of 200%! I expected the price to be a bit lower than double, but a 45% bonus is surprisingly good.

 

post-188-0-98736300-1429255972_thumb.png

Edited by Toni Babelony
Link to comment

I always leave one part of the model separate, usually the floor.  That allows access to the interior, if I want to install windows or other details, or to paint the inside walls.  If I model the floor so that it fits inside the lower edge of the body exactly, I don't have to glue it in.  The material is just rough enough to hold the pieces together.

 

As I mentioned before, long thin shapes tend to warp more than short thick ones.  It's the opposite of thermoplastic, which warps because of shrinkage, and thick parts shrink more than thin ones.

Link to comment

If you pack the requested amount of cars (number of orders) into one batch it will be even more cheaper. Anyway i think the base price is already acceptable. So please tell us when we can order...

Link to comment

+one more thing: The quality might suffer a bit if you don't print all sides horizontally flat since the layer boundaries will show. This can be avoided with completly flat train sides and flat kits. Curved surfaces suffer both ways.

Link to comment

This will a little bit more usage volume in the printing machine and will probably increase the price a little, but will overall generate a more acceptable result.

 

EDIT: I have had the new data checked and it's actually roughly ¥200 cheaper, as I use less material!

 

EDIT 2: A cost-offset test with a 2-car print proves to be about 155% more expensive, instead of 200%! I expected the price to be a bit lower than double, but a 45% bonus is surprisingly good.

Getting rid of the "runners" probably saved you a lot of material.  The printing company will combine several user's jobs into a single machine run, so the arrangement of the parts in the file you submit doesn't really matter much. 

 

Once you clean and paint the model, you'll never notice the layer lines, no matter what orientation it got printed.

 

Be sure you prep the parts correctly before painting.  The printing process leaves the material saturated with various substances (sometimes vegetable oil) that can prevent a good paint job.  Many people who have experience painting these models recommend a short soak in a solvent called "Bestine" (sold in arts & crafts stores in the US).  Others have recommended a cleaner called "Goo Gone".

 

I don't know what equivalents you can find in Japan, so you'll need to look around.

Link to comment

Bestine is actually heptane, a certain type of gasoline. I'm not 100% sure this is safe for the plastics. The other one is baking soda and gasoline. I would use something a bit more gentle.

 

The type of residue depends on the material in use. Extruded abs should be pretty clean, while polystyrene can have solvents in it. Some printing companies clean the parts before shipping, some don't.

 

 

Once you clean and paint the model, you'll never notice the layer lines, no matter what orientation it got printed.

This depends on the technology and the thickness of the printed layers. Also the cleaning can damage smaller details, so imho the less cleaning is needed, the more small details will remain (if they are there after printing). Although the one piece shell has the benefit fitting properly together without filing and filling.

 

 

It's maybe more rewarding to offer direct data for printing instead of risking the investment in badly made moulds.

The quality of those are much less that what is available with 3D printing. So i would say 3D print every order in aggregated batches (to make it more economical to print) or invest in a proper copper injection mold. (it's worth it above around 1000 pieces, in this case above 500 sets) Flat molds are much cheaper and more simple to make, so for cheaper kits this is the only alternative. For high volume production, a proper mold can be used, similar to what Tomytec uses. (they can be cnc machined very quickly if the producing company has the right design files)

Link to comment

Bestine is pretty safe on acrylic as you use it to clean off the paper film goo when you use sheet acrylic plex. i cleaned a lot of clear acrylic plex with it and never any issues. not sure about soaking for a long period, but short should be fine. all part of the testing to see what it will work. it is basically hexane. it will strip paint pretty well! its rubber cement thinner as well (bestine is the brand name).

 

goo gone is very safe stuff. its basically some oils that are good at disrupting most tacky glue bits adhesion. doesn't total solubilize it but i think just interrupts the stickyness. I've used it on all sorts of plastics and not found it attacking anything and even on some painted things and again no issues for me so far. it seems to evaporate pretty clean. one of our members uses it on his tracks when they are really dirty. after 15 min it seems to be all gone, no detectable residue with the finger and tissue wipe. in any case a good wash with soapy water would help get any oils and such off.

 

again a small ultrasonic bath would really help to clean off this stuff w/o having to do any rubbing or scrubbing. I'm amazed how well it pops paint right out from cracks and details after soaking and scrubbing has not gotten it. it really is a handy tool to have laying around (plus can score you points cleaning jewelry, glasses, etc for the better half!)

 

cheers,

 

jeff

Link to comment

Tomytec just announced the release of the first modern Kantō Railway DMU, two versions of the 2-car KIHA 2100 types in September 2015. This means an end to my quest. These models come painted, assembled and at a fraction of the price of 3D printing. In the future, Tomytec will probably release other versions, like the related single car versions.

 

Anyway, I don't feel beaten, as it was good practice and I've already stopped development a few weeks ago on these models. Rather than that, I've ventured into creating the trams used by the Utrecht Sneltram before the modernisation:

 

post-188-0-20498800-1431654557_thumb.png

 

As you can see, I haven't finished the model yet, as the roof details haven't been added yet, as well as the bogie designs. The 3D model is also being made at roughly 1/110, so it can be reduced to 1/120, 1/150 and 1/160 easily. Upscaling it to 1/87 shouldn't be much of an issue either, but wouldn't do the detailing much good (although it is at 0.1mm level). xD

 

I'm planning to have part of this train sit on a Tomytec TM-10 power chassis, as this comes closest to the dimensions of the real thing. If the TM-10 proves to be difficult to find, I'm opting for the Tomytec hakotetsu HM-01 power unit (to be located in the jacobs bogie part) or look for an alternative with flywheel (preferably). An alternative would be the TM-11, but this one has the bogies apart by 5mm more than the TM-10, which I don't like at all.

  • Like 1
Link to comment

Nice! When finished you maybe should let the Dutch railway modelling community know. I think there will be some interest at least.

Link to comment

Imho a half Tomix TM-TR01 frame in each car and one of the unpowered bogies mounted as a jackobs bogie might match the wheel and bogie size of the original units better.

Link to comment

Nice! When finished you maybe should let the Dutch railway modelling community know. I think there will be some interest at least.

 

 

Maybe I should make an account on Beneluxrail for that ヽ( ̄▽ ̄)ノ(;_;)

 

 

Imho a half Tomix TM-TR01 frame in each car and one of the unpowered bogies mounted as a jackobs bogie might match the wheel and bogie size of the original units better.

The goal is to keep costs low, as well as minimise cutting and modding of 3rd party parts. Two motors is too much. The HM-01 (bought one today) can be modded by replacing the body carrying part, which seems universal with other Tomytec units. I'll look into that later.

Link to comment

It's actually called "Beneluxspoor" nowadays. ;)

 

I'm actually pretty interested in these Utrecht trams myself, but the thought of having to paint it myself scares me off a bit. And I don't think my wallet will agree anytime soon.

 

I think we need more Dutch trams. ^^; Being from the Hague myself, these are perfect TM-01 candidates: http://nl.wikipedia.org/wiki/PCC-trams_in_Den_Haag & http://nl.wikipedia.org/wiki/Gele_Tram

The "Buitenlijner" trams can be prototypically run together with the SUNIJ trams as well :grin:

http://www.htmfoto.net/index.php?page=tramhistorie/UtrechtsNieuwsblad1988.htm

https://www.flickr.com/photos/tramwegstichting/9771358102/

https://www.flickr.com/photos/tramwegstichting/9771550144/

https://htmfoto.nl/algemeen/tonen2.php?pagina=1353873334

Link to comment

Toni, do you plan to make models based on comissions? There are a few japanese trains that can not be found in N scale and there would be a market for them. For example there is the classic JNR 12 series: kumoha, moha, kuha, saha and saro variants, with and without a wide cab (please correct me if i'm wrong).

 

Personally, i would like to have an 4+4 car train with this type. (like kumoha-saha-moha-kuha+kumoha-saro-moha-kuha) If you think is is a viable product or there are similar ones, we could start a preorder round here on this forum to see if making them would be economically viable for you. (i choose the 12 series as this type used the same frame and general body design for all variants and they can be spray painted plain brown, which is relatively easy and they can be run from the pre ww2 period up to the jr era)

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...