Best Of ACL MEC low side gondola , interest ? ESM please read .

[lock4244]

So Bryan, if someone had a nice HO model of a freight car, it'd be easier to scale down to N than from drawings alone? That being the case, how necessary are drawings of the prototype?

[bbussey]

So Bryan, if someone had a nice HO model of a freight car, it'd be easier to scale down to N than from drawings alone? That being the case, how necessary are drawings of the prototype?

Official engineering drawings are preferred because they are more likely to be correct to the prototype than anything else.  Copying an existing model in another scale leaves you at the mercy of the original model designer regarding the "correctness" of the model, and it opens you up to copyright infringement.  In this case, the HO model is being used to get the basic dimensions only, and the decals will be a combination of the ACL and MEC prototypes based on available drawings.

What file format?  I work in AutoCAD; can do dxf and dwg easily.

DXF is fine.

[Iain]

Allright, I will get on it as soon as I can; what's the deadline?

[bbussey]

You have time.  Sometime over the summer.

[Chris333]

Iain, Any pics of the NS car?

[bbussey]

There is one here, about a third down the page:
http://www.rmweb.co.uk/community/index.php?/topic/46392-from-the-workbench-for-everyone-to-contribute/page-1

Upon further review, the NS car is a 14-panel prototype, while the MEC/ACL car is a 9-panel prototype.  I don't see how it's feasible to include decals for any schemes not on 9-panel cars ... as it opens a Pandora's Box regarding all low-side gons, in addition to exceeding the boundaries of ESM philosophy regarding prototypical accuracy.

[Iain]

Crap

[wcfn100]

Crap

Have faith.  SAL G-10 14 panel gondola general arrangement drawings have been acquired.   

Extra time,  ehhh.


Jason

[bbussey]

I had a spare hour or so tonight, so I took a set of digital calipers to Rich's model and created the digital template in HO scale.  I already see some issues that will have to be addressed when it is reduced - the thickness (or more accurately, the thinness) of the walls to start.  The integrated coupler pockets and brake appliances will be added after the model is reduced to N scale size.  No need for rivet detail or the rib flanges to be rendered, as that will be done on etched brass insets for the sides and etched flanges respectively.  Unlike the HO model, the inside walls will have rivet detail.  The model appears to be at the correct height over the rail, so the digital model is being built at the same height over the Y plane.  This way, it will be at the correct height when scaled down and the bolsters are modified to accept BLMA trucks.











The dreadnaught ends will be adapted from the XIH boxcar model once the gon is shrunk and split into flat parts.  The cavities for the insets on the outside and inside faces of the sides will be done after the model is reduced.  The centersill width and shape will be determined by the MTL fishbelly sill part, and the bolsters will be adjusted to conform to the BLMA trucks.  The provisions for the OEM parts obviously occur after the model is reduced.  And no need to include a floor since it will be fabricated from metal components.

[up1950s]

Wow , more technical then I imagined .

[sirenwerks]

Bryan, I really appreciate the walk-through of your process.  It both confirms and contradicts some of my beliefs about the production side.  Additionally, your opening up of the secret society of model makers' processes, while dangerous in the respect that others will now know how to DIY, is fodder for less bashing and more useful critical feedback. 

I wish more manufacturers would take this tact. While N scalers can be complacent little babies at times (myself included), knowing the process can allow us the possibility of offering improvements to manufacturer processes culled from our own knowledge.  I feel the urge to hold hands with everyone and sing Kumbaya, but I will refrain.  This is just one of the aspects that sets Railwire apart from the rest.

Now, back to my regularly scheduled prickly, persnickety, curmudgeoness...

[bbussey]

I think it goes without saying that the Railwire community in general, among other higher characteristics, is the most - I'll say couched in reality - of all the model railroading boards.  So it's appropriate to log this type of thread in this venue - not only where it would be understood, but where it would be appreciated for what it is: an effort to enlighten.  An enlightened community is better positioned to present requests to manufacturers that would be considered - as well as constructively critique products already present in a way the manufacturers will respect - if it is better versed in the process.  I've lobbied many a manufacturer on bringing certain models and certain features to market, and have had a fairly good success rate in doing so.  The N scale industry shift to body-mounted couplers on rolling stock is one example.

Knowing the process also may spur community members to dabble in the arena.  More people knowing the "how-to" and offering the fruits of their talents and skills to the masses is never a bad thing.  There remain plenty of untapped prototypes to go around.

[wcfn100]

So Bryan, as it sits in the rendering you showed, is RP still the most cost effective way to produce that part vs. something like CNC?

I know that rivets create problems with RP rendering, but if not for the rivets, I don't see RP as being that great of a technology especially if your going to get the dreadnaught ends from a previous car.


Jason

[bbussey]

It is for a low-yield low-cost (relatively) polyurethane-resin-kit type project such as this.  If planning for a mass production injection-molded product, there would be differences but the process (at this stage) is the same in general.  You have to recreate the prototype, then make the adjustments for whatever production process will be used to render the model.  For rapid prototype masters for making resin castings, that means insuring that all visible surfaces have as little visible stepping as possible.  For injection molded and diecast items, that means adding all your surface detail and draft angles, and figuring out the best location to gate the part to the runner. 

In both circumstances, you must account for shrinkage and distortion issues:  Plastic, diecast, resin and rubber all shrink; metal etchings over-etch by half the metal thickness; and RP machines all have some adjustment factor on their XYZ axis.  Those shrinkage factors are adjusted when the solid model parts are complete.  For example - on the ProJet 3000+ machine I contract to, we rendered a 100mm x 100mm square part, then measured it to get the adjustment factors on the XY axis (the Z axis has no distortion to account for).  For injection-molded and diecast items - since the tooling most likely is being contracted out, it's best to give them the solid models as is and let them apply their own shrinkage factors.

This gondola model will have etched brass insets for the side panels, and etched bass surface detail for the rib flanges.  The goal is to create a master for casting.  We will hope the master survives the casting process but that isn't the primary goal.  You see how that process evolves within the next three weeks or so.

Instead of building a new dreadnaught end, it is far easier and far less time-consuming to extract the dreadnaught A end from the XIH, discard everything above the second rib, mirror the outside face detail to the inside face, and stretch the part to match the gondola dimensions.  Especially if it gives me 95% of what I want.  Dreadnaught ends and stamped roofs are very labor intensive.  And that includes taking into consideration that you create just one stamped section and duplicate it to get the length or height you need.  If this was for an injection-molded mass-market model, I would have to create the part from scratch so that it was precise.

[skm]

I add myself to the group that are fascinated by the design and production process! Watching the steps to retail would be fantastic and if you have the time to post as you go that would be a real learning experience for me!

You mentioned Railwire interaction, so I have a 'request' if you will.

Is it at all possible to make the bolsters so that other than BLMA trucks could be used? (say, a detachable pad?)

The coupler pockets/pad so that Z scale couplers could be substituted?

And in your discussion of dreadnaught ends, isn't the reinforcing rib on the end simply a stamped part riveted on?
Or are you planning a different end entirely?