CAD library (Some thoughts)


Richard Brennan <brennan8@...>
 

At 06:38 AM 8/14/2014, 'Al and Patricia
Westerfield' westerfieldalfred@frontier.com [S wrote:
Has anyone considered setting standards for
model railroad rapid prototyping? Here are some thoughts on the subject.
Way overdue! Ok - Some thoughts:

Select a simple CAD program for use of everyone on the list.
Noooooo!!! We do NOT need to select the CAD
program... only the exchangeable file type(s).
Possible choice would be a "mesh" file format; like .STL, .PLY, or .OBJ.

Establish a library of designs that can be
plugged into different prototypes – example:
reefer door hinges, rivets, scribing.
e.g. As a start; all the stuff in the GrandtLine,
PSC and similar parts catalogs.
However... with unlimited variation
possible, the constraints imposed by 'close but not quite' parts go away.

How about metadata? Maybe in an XML format...
with source information: measured in the field
from car No. xxxxxx; scaled from photo; derived from plans; etc...

In my dreams, GrandtLine would someday accept
bulk orders as STEP-files for CNC-cutting molds for styrene parts on-demand...
and PSC would offer 3D-to-Wax-to-Brass
printing+investment casting services over the web. <<sigh>>

Set up a system of free and for-profit downloads.
Free: Certainly!
For-profit: I doubt it...
AFAIK there is no DRM (Digital Rights
Management) capability on anything so intrinsically 'editable'
Additionally, if I modify the file to decrease
the head size on one rivet... now is it MY design... or Yours?

Establish standards for quality and material with perhaps qualified vendors.
Establish standards for thicknesses of major parts so kitbashing is simpler.
..because our Hobby has such a good track record
in setting relevant standards, updating, and enforcing them????
This really a downstream production issue... You
tend to get what you pay for, and technology
progress keeps the value increasing.

Sell downloads for major parts such as roofs and ends.
Well.. People do sell public domain stuff all the time;
So I guess selling canned designs is possible,
even if they are readily available elsewhere.
The user value found in the design is
small... the value of a perfect part out the end of the process is large.

Can we foresee the day of the local 'Starbucks 3D print Bureau'...
"Would you like a Vente Latte while that Deco End prints?..."

There are so many opportunities here I‘ve
only scratched the surface. Anyone want to add some ideas? – Al Westerfield
Al...
I believe it's ALL YOUR fault (...at least partially),
For leading us onto that long, steep, slippery slope of:
W-A-N-T M-O-R-E A-C-C-U-R-A-T-E M-O-D-E-LS !!!

Thanks!

--------------------
Richard Brennan - San Leandro CA
--------------------


Robert kirkham
 

As I am doing some 3d drafting and making parts for modelling, a few
comments.

- if there was a suitable way that one could design, for example, a reefer
door hinge 3d file, make it available for a reasonable royalty, share it
with only folks one trusted, the investment wouldn't be unreasonable and
getting parts made for other models would be a lot easier. I'd like it.
Ditto for things like other door hardware, power handbrake covers, etc etc.
Love to see shapes for various air brake components. Some of the metal
fittings on real freight equipment are very fiddly to get to print nicely,
and since each of those tiny lines has to be drawn to create a file, any one
draft person's output is limited by the same thing that limits all of us -
time. So a warehouse of some parts would help.
- for me, such shortcuts would have to work on the basis of clear agreements
that released all rights of the designer in terms of the iterations used in
models. But it would be on the basis a) of a trust relationship, and b) the
idea that the designer knows what the part will be used for and the part
isn't used for anything more than agreed. The fee fits the use.
- if I was designing a model for someone else, I'd only venture down this
path if they were fully informed and agreeable.
- regarding standards. Hmm. well part thickness is OK to talk about, but
if someone offered the 3d file for a hinge and I didn’t like the thickness
(or any other element of it), I'd modify it to suit. That too would be part
of the agreement. The guy who is going to make castings from the printed
part will have a say in what the standards are - but I find that input
surprisingly flexible.
- tougher issues are designing parts to work with current technology.
Flanges have to be thick enough to print (and then to be cast), so the idea
of, for example, printing a 3d version of drawings in the Cyc's doesn't lead
in a straight line to completed models. Parts are exaggerated in subtle
ways to work in HO scale. I've not really tried N, but it has similar and
more significant challenges to represent without making lumps. If I recall
correctly, I think Dennis (some years ago) talked about the issues of
designing a rivet that looks right. That was for injection molds. Lots of
experience and know how went into it. Yet in 3d printing I doubt the
lessons with other technologies provide exact answers - Instead they merely
suggest a direction the modeller needs to experiment with. The shape, the
proportions, etc have to be changed by the 3 d modeller. So far, the only
rivets and bolt heads I have printed that looked good on the printed part
were ugly exaggerations when viewed in the 3d drawing - but quite reasonable
when printed. A polling pocket I drew for a friend came through with
surprising detail (but flanges were grossly exaggerated in the drawing). So
I agree there is a need for standards for designing parts. But what kind
of standards? Today's technological limitations can easily change with
innovation over time. Flange thickness get's narrower with each change. So
if we set standards today, we are apt to repeat what happened with NMRA
wheel standards - too thick and uuuuugly! So it may be better for drawings
to be done to prototype specs after all. So long as I can take that file
and tweak it to suit the modelling scale and technology I am working with,
then it is a help in creating models. I see I've come full circle here. . . . Still thinking my way through it.

I can see a 3d warehouse having some value, on the right terms, with the
right people. I draw for fun; not for profit. I know others are investing
their own real money to run a business and don't want the risks that might
happen if warehoused drawings are included in their parts without proper
legal protection. I respect that. But I think there are ways to accomplish
both goals.

Rob Kirkham


Dennis Storzek
 

---In STMFC@..., <brennan8@...> wrote :
"In my dreams, GrandtLine would someday accept bulk orders as STEP-files for CNC-cutting molds for styrene parts on-demand...and PSC would offer 3D-to-Wax-to-Brass printing+investment casting services over the web. <<sigh>>"Are you suggesting that someone with no experience in designing plastic parts can just whip up a design for a mold cavity that will:1. Fill2. Solidify without objectionable sinks3. Pull across (stay on the proper side of the mold to be ejected)4. Eject without distortion.I've been doing this stuff for twenty five years now, and have just gone through eight weeks of tooling revisions where my "gut feeling and educated guesses" obviously weren't good enough. Always firmly in the back of my mind is the fact that it IS possible to spend serious coin on a mold that will never make an acceptable part.Even doing this for a living, when I wanted Grandt to cut a mold for some sill steps, I sent Dave a sketch and let him design the cavities... he knows his machining capabilities, he knows what worked in the past, no reason not to take advantage of his expertiseDennis Storzek.
 


Richard Brennan <brennan8@...>
 

...awakened from my dream by the harsh cold water of reality!

Dennis;
Good points... I'm not suggesting this is a near-term possibility.
There are obviously large-scale prototyping service bureaus that do what I describe... and for smallish parts the required tooling modifications may be quite do-able, given editable 3D files.
In the long run, I think it may be possible to automate SOME of the needed tooling/mold cavity accommodations with post-processing software, similar to the ones used in 3D printing support, but in the final analysis, a sentient entity is going to have to intervene.

Richard

At 07:24 AM 8/15/2014, destorzek@mchsi.com [STMFC] wrote:
---In STMFC@yahoogroups.com, <brennan8@...> wrote :
"In my dreams, GrandtLine would someday accept bulk orders as STEP-files for CNC-cutting molds for styrene parts on-demand...and PSC would offer 3D-to-Wax-to-Brass printing+investment casting services over the web. <<sigh>>"
Are you suggesting that someone with no experience in designing plastic parts can just whip up a design for a mold cavity that will:1. Fill2. Solidify without objectionable sinks3. Pull across (stay on the proper side of the mold to be ejected)4. Eject without distortion.I've been doing this stuff for twenty five years now, and have just gone through eight weeks of tooling revisions where my "gut feeling and educated guesses" obviously weren't good enough. Always firmly in the back of my mind is the fact that it IS possible to spend serious coin on a mold that will never make an acceptable part.Even doing this for a living, when I wanted Grandt to cut a mold for some sill steps, I sent Dave a sketch and let him design the cavities... he knows his machining capabilities, he knows what worked in the past, no reason not to take advantage of his expertise
Dennis Storzek


Dennis Storzek
 

One of the problems with Shapeways is their automated system takes all the control out of the process, and Shapeways keeps their price down by packing as many parts (from different customers) in the machines "build envelope" as they possibly can. The problem with this is their highest resolution process uses a second material to build support structures under overhanging features, and the surface which is built against this support material is considerably rougher than we would like.Tom Madden has written about this a bit, either here or on the Passenger Car List, and has posted some photos that illustrate. Tom has made the comment that in order to get a shot of having his parts built in the orientation he would prefer, he has taken to grouping three parts oriented differently into a single assembly, essencially paying Shapeways 3X their normal charge, to run his parts. Given the fact that Tom consults for another service bureau and has access to machine time at below market rates, this should tell you something about what you can expect to pay for industrial grade work. If you want your part built all by itself in the preferred prientation, you are going to pay dearly for it.My own personal test part has been an HO scale type L passenger car triple valve, to be added to the head of a PSC passenger car brake cylinder. Actually, I'd like to model the valve, cylinder, and slack adjuster and associated piping as one piece, but the larger part would be way too costly for a test. This is a part that really has no preferred orientation; like many of our parts, all surfaces have some level of detail, and all are "appearance" surfaces, so it is an excellent test of the process. If and when the process progresses to the point where adequate quality parts can be made at a reasonable price, all sorts of brake parts, slack adjusters, truck sideframes, and the like could be made.The first time I sent this part file out was to Print-A-Part in January of 2010, then later to Shapeways in October of 2012 Each time it's cost me around a hundred bucks, so far for naught, and I won't be sending it out again until someone claims they've increased the printing resolution.This link should take you to a series of pix on Flickr:L Triple ValveDennis Storzek

 


Dennis Storzek
 

I apologize for the total lack of formatting in my recent messages... I've been replying via the Whahoo "REPLY" text editor, and it insists in stripping out spaces and carriage returns today. Thanks, Whahoo.....Dennis


Rod Miller
 

I wonder if standards and RPs on this would be appropriate
for the NMRA to address. Consider where DCC might be if it
were not standardized.

NMRA haters please refrain from venting.

I asked Charlie Getz, NMRA President, if the NMRA would be
interested in this as a project. His response was yes but
it won't work if no one volunteers to help. So if there is
consensus that this warrants consideration, are any of you
interested in volunteering? If so, please let Charlie know
at: pres at hq d o t nmra dot o r g

Thx

Rod

On 8/14/14, 8:37 PM, Richard Brennan brennan8@earthlink.net [STMFC] wrote:
At 06:38 AM 8/14/2014, 'Al and Patricia
Westerfield' westerfieldalfred@frontier.com [S wrote:
Has anyone considered setting standards for
model railroad rapid prototyping? Here are some thoughts on the subject.
Way overdue! Ok - Some thoughts:

Select a simple CAD program for use of everyone on the list.
Noooooo!!! We do NOT need to select the CAD
program... only the exchangeable file type(s).
Possible choice would be a "mesh" file format; like .STL, .PLY, or .OBJ.

Establish a library of designs that can be
plugged into different prototypes � example:
reefer door hinges, rivets, scribing.
e.g. As a start; all the stuff in the GrandtLine,
PSC and similar parts catalogs.
However... with unlimited variation
possible, the constraints imposed by 'close but not quite' parts go away.

How about metadata? Maybe in an XML format...
with source information: measured in the field
from car No. xxxxxx; scaled from photo; derived from plans; etc...

In my dreams, GrandtLine would someday accept
bulk orders as STEP-files for CNC-cutting molds for styrene parts on-demand...
and PSC would offer 3D-to-Wax-to-Brass
printing+investment casting services over the web. <<sigh>>

Set up a system of free and for-profit downloads.
Free: Certainly!
For-profit: I doubt it...
AFAIK there is no DRM (Digital Rights
Management) capability on anything so intrinsically 'editable'
Additionally, if I modify the file to decrease
the head size on one rivet... now is it MY design... or Yours?

Establish standards for quality and material with perhaps qualified vendors.
Establish standards for thicknesses of major parts so kitbashing is simpler.
..because our Hobby has such a good track record
in setting relevant standards, updating, and enforcing them????
This really a downstream production issue... You
tend to get what you pay for, and technology
progress keeps the value increasing.

Sell downloads for major parts such as roofs and ends.
Well.. People do sell public domain stuff all the time;
So I guess selling canned designs is possible,
even if they are readily available elsewhere.
The user value found in the design is
small... the value of a perfect part out the end of the process is large.

Can we foresee the day of the local 'Starbucks 3D print Bureau'...
"Would you like a Vente Latte while that Deco End prints?..."

There are so many opportunities here I‘ve
only scratched the surface. Anyone want to add some ideas? � Al Westerfield
Al...
I believe it's ALL YOUR fault (...at least partially),
For leading us onto that long, steep, slippery slope of:
W-A-N-T M-O-R-E A-C-C-U-R-A-T-E M-O-D-E-LS !!!

Thanks!

--------------------
Richard Brennan - San Leandro CA
--------------------



------------------------------------
Posted by: Richard Brennan <brennan8@earthlink.net>
------------------------------------
--
Rod Miller
Handcraftsman
===
Custom 2-rail O Scale Models: Drives, | O Scale West / S West
Repairs, Steam Loco Building, More | 2015 Meet is Feb 5 - 7
http://www.rodmiller.com | http://www.oscalewest.com


arved_grass
 

I'll try not to vent, but I'd rather not have the NMRA involved. Otherwise, you could expect everything in HO scale to end up S scaled, such as the NMRA coupler pocket spec (RP-22). I prefer the prototype to be my standard, rather than some arbitrary rules that impede progress.

While DCC should be standardized, the correct standards body should be the IEEE, and not the NMRA. The NMRA completely overstepped it's bounds assuming an electrical standard. Where would DCC be without the NMRA? Under the IEEE, my best guess would be that we'd already have "The Internet of Everything" with full IP addresses instead of a hodgepodge of 2 and 4 digit addressing, and numerous other advantages.

I would hate to see the crippling effect of NMRA standards and recommended practices on what could and should be a major leap forward for the model builder. If there is a standards body that should be involved, I would suggest the International Plastic Modelers Association. The technology has a lot more market share to the entirety of plastic model building, than just steam era freight car modelers.

Let's learn from our mistakes and keep the NMRA out of it.
 
------------------------
Arved Grass
Fleming Island, Florida


From: "Rod Miller rod@... [STMFC]"
To: STMFC@...
Cc: pres@...
Sent: Friday, August 15, 2014 3:32 PM
Subject: Re: [STMFC] CAD library (Some thoughts)

I wonder if standards and RPs on this would be appropriate
for the NMRA to address. Consider where DCC might be if it
were not standardized.

NMRA haters please refrain from venting.

I asked Charlie Getz, NMRA President, if the NMRA would be
interested in this as a project. His response was yes but
it won't work if no one volunteers to help. So if there is
consensus that this warrants consideration, are any of you
interested in volunteering? If so, please let Charlie know
at: pres at hq d o t nmra dot o r g

Thx

Rod



On 8/14/14, 8:37 PM, Richard Brennan brennan8@... [STMFC] wrote:
> At 06:38 AM 8/14/2014, 'Al and Patricia
> Westerfield' westerfieldalfred@... [S wrote:
>> Has anyone considered setting standards for
>> model railroad rapid prototyping?  Here are some thoughts on the subject.
>
> Way overdue! Ok - Some thoughts:
>
>>  Select a simple CAD program for use of everyone on the list.
>
> Noooooo!!!  We do NOT need to select the CAD
> program... only the exchangeable file type(s).
> Possible choice would be a "mesh" file format; like  .STL, .PLY, or .OBJ.
>
>> Establish a library of designs that can be
>> plugged into different prototypes – example:
>> reefer door hinges, rivets, scribing.
>
> e.g. As a start; all the stuff in the GrandtLine,
> PSC and similar parts catalogs.
> However... with unlimited variation
> possible,  the constraints imposed by 'close but not quite' parts go away.
>
> How about metadata?  Maybe in an XML format...
> with source information: measured in the field
> from car No. xxxxxx; scaled from photo; derived from plans; etc...
>
> In my dreams, GrandtLine would someday accept
> bulk orders as STEP-files for CNC-cutting molds for styrene parts on-demand...
> and PSC would offer 3D-to-Wax-to-Brass
> printing+investment casting services over the web.  <>
>
>> Set up a system of free and for-profit downloads.
>
> Free: Certainly!
> For-profit: I doubt it...
>    AFAIK there is no DRM (Digital Rights
> Management) capability on anything so intrinsically 'editable'
>    Additionally, if I modify the file to decrease
> the head size on one rivet... now is it MY design... or Yours?
>
>> Establish standards for quality and material with perhaps qualified vendors.
>> Establish standards for thicknesses of major parts so kitbashing is simpler.
>
> ..because our Hobby has such a good track record
> in setting relevant standards, updating, and enforcing them????
> This really a downstream production issue... You
> tend to get what you pay for, and technology
> progress keeps the value increasing.
>
>> Sell downloads for major parts such as roofs and ends.
>
> Well.. People do sell public domain stuff all the time;
> So I guess selling canned designs is possible,
> even if they are readily available elsewhere.
> The user value found in the design is
> small...  the value of a perfect part out the end of the process is large.
>
> Can we foresee the day of the local 'Starbucks 3D print Bureau'...
> "Would you like a Vente Latte while that Deco End prints?..."
>
>>  There are so many opportunities here I‘ve
>> only scratched the surface.  Anyone want to add some ideas? – Al Westerfield
>
> Al...
> I believe it's ALL YOUR fault (...at least partially),
> For leading us onto that long, steep, slippery slope of:
> W-A-N-T  M-O-R-E  A-C-C-U-R-A-T-E  M-O-D-E-LS !!!
>
> Thanks!
>
> --------------------
> Richard Brennan - San Leandro CA
> --------------------
>
>
>
> ------------------------------------
> Posted by: Richard Brennan <brennan8@...
>
> ------------------------------------

--
Rod Miller
Handcraftsman
===
Custom 2-rail O Scale Models: Drives,  |  O Scale West / S West
Repairs, Steam Loco Building, More    |  2015 Meet is Feb 5 - 7
http://www.rodmiller.com               |  http://www.oscalewest.com


------------------------------------

------------------------------------


------------------------------------

Yahoo Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/STMFC/

<*> Your email settings:
    Individual Email | Traditional

<*> To change settings online go to:
    http://groups.yahoo.com/group/STMFC/join
    (Yahoo! ID required)

<*> To change settings via email:
    STMFC-digest@...
    STMFC-fullfeatured@...

<*> To unsubscribe from this group, send an email to:
    STMFC-unsubscribe@...

<*> Your use of Yahoo Groups is subject to:
    https://info.yahoo.com/legal/us/yahoo/utos/terms/





Tim O'Connor
 

It was the US government that funded, and created most of the key
standards that govern the network layer of the Internet. The standards
function was stewarded for a time by keepers of the flame, before the
IETF standards body took over.

Anyway there is a HUGE difference between STANDARDS and DESIGNS. I can
imagine a Shapeways "standard" that encompasses CAD software conventions,
notations, required levels of resolution, accuracy, etc.

But once someone invests a hundred or a thousand hours of their time in
a project, are they just going to give that away? I guess they could (an
example would be Linux which has thousands of unpaid contributors) but it
seems unlikely to amount to much.

So I think CAD library Standards are a good idea. Free sharing of designed
parts? Not so much.

For models, a good standard might be to design a single way to attach the
end of a steel box car to the body (sides and roof and floor) -- perhaps the
standard would include variations for different car heights.

This way I could buy ends, sides, roof and floor that all conformed to the
standard, and build myself a freight car out of parts from several designers
(or vendors)

Tim O'Connor

I'll try not to vent, but I'd rather not have the NMRA involved. Otherwise, you could expect everything in HO scale to end up S scaled, such as the NMRA coupler pocket spec (RP-22). I prefer the prototype to be my standard, rather than some arbitrary rules that impede progress.

While DCC should be standardized, the correct standards body should be the IEEE, and not the NMRA. The NMRA completely overstepped it's bounds assuming an electrical standard. Where would DCC be without the NMRA? Under the IEEE, my best guess would be that we'd already have "The Internet of Everything" with full IP addresses instead of a hodgepodge of 2 and 4 digit addressing, and numerous other advantages.

I would hate to see the crippling effect of NMRA standards and recommended practices on what could and should be a major leap forward for the model builder. If there is a standards body that should be involved, I would suggest the International Plastic Modelers Association. The technology has a lot more market share to the entirety of plastic model building, than just steam era freight car modelers.

Let's learn from our mistakes and keep the NMRA out of it.

------------------------
Arved Grass
Fleming Island, Florida


Carl Gustafson
 

On Fri, Aug 15, 2014 at 11:12:27PM -0400, Tim O'Connor timboconnor@comcast.net [STMFC] wrote:

But once someone invests a hundred or a thousand hours of their time in
a project, are they just going to give that away? I guess they could (an
example would be Linux which has thousands of unpaid contributors) but it
seems unlikely to amount to much.
As a point of interest, many of the biggest contributors to the linux kernel are employees of
companies who benefit from the system, and are paid by their employers to work on the software. Linus
Torvalds himself benefitted in this way, his main function in several organizations being to shepherd
the Linux project.

The real example here is that numerous for-profit organizations saw the benefit of contributing to a
community project, and paid their employees to do so. I could see the same thing happening here. Need
a particular roof or end for a box-car project? Someone's already done that, but you will need to
contribute a door, or something like that.

Carl "Open Source Believer" Gustafson