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

[bbussey]

Love them Bryan , let me know when you want the money for my 10 cars .

My guess is late summer / early autumn.  This is roughly the sequence of steps that remain to be done:

• Adding the remaining underbody details to the digital model.
• Creating the cavities on the digital models for the insets (body sides inside and outside faces, flat ends inside and outside faces).
• Tracing the insets and using that to create the half-etch artwork for the brass inset panels.
• Drawing the no-etch patterns (rivets) and full-etch patterns (holes for grab irons) of the etched brass insets.
• Splitting the digital body into flats for RP rendering (to insure no visible stepping on the ribs) and adding the mold separation membranes.
• Subcontracting the ProJet XHD body and ends.
• Subcontracting the etched brass panels and rib flanges.
• Assembling the casting masters.
• Subcontracting the molding of the masters.
• Subcontracting the casting of the resin components.
• Subcontracting the fabrication of the steel sub-floor.
• Procuring the OEM components (trucks, couplers, detail parts).
• Drawing the patterns for the etched stainless steel parts (floor tops and brakewheel step).
• Subcontracting the etched stainless steel parts.
• Subcontracting the fabrication of the foam nests for the jewel box.
• Ordering additional jewel boxes from the factory.
• Drawing the decals artwork.
• Subcontracting the decal printing.
• Subcontracting the printing, cutting and scoring of the paper inserts for the jewel box.

Got a ways to go.

I'll return the Ertl model after the drawings for the etched brass insets are complete.

Love the step-by-step, and jealous of the 3D 

You're a master at creating etching patterns.  You could learn solid modeling within a week.
 

[up1950s]

No rush at all on the Ertl Bryan , hang on to it till the finished models come out and ship it all together if you like .

[bbussey]

Fifth session.

The brake appliances are imported from the XIH boxcar model and positioned accordingly.  The top faces are increased to the level of the eventual floor.



The appliances are integrated into the gondola body, and the holes for MTL/BLMA bolster pins are added.  The ESM branding and country of origin are placed so that they are not visible when the model is upright.



Next is to prepare for the brass insets.  The body and flat ends are scaled up temporarily to compensate for the shrinkage of the rubber mold and casting resin, which in this case will be 1.00750065.

.10mm of material is removed from all the flat surfaces that will be represented by the brass.  First, the lower side panels ...



... then, the main side panels ...



... which leaves the outer faces of the digital sides brass-ready.



The inside face of the sides also must be prepared.



Now the insets must be traced to create patterns for the brass insets, as well as proper alignment of the rivets.



The tracings for the inner and outer faces of the sides and flat ends are consolidated into a single drawing, which is exported to a DXF file.



The DXF file, in turn, is imported into CorelDraw, where the artwork is created that will etch .15mm brass stock.  These won't be commissioned until the rapid prototype parts arrive, so that the measurements can be tweaked if necessary.  Calipers will compare the rapid prototype to the drawing, and the drawing will be re-scaled if necessary to fit.  The magenta areas are what remains after the bottom etch, the black areas are what remains after the top etch.  In this scenario, all areas under the black are magenta, so the magenta serves as the half-etch, the black serves as no-etch, and the white serves as fully-etched.  The diameter of the rivets is half the thickness of the brass, which insures the etching process will create rounded rivet heads. 



This is not the final artwork, as holes for the grab irons and vertical brake staff mounts remain to be added.  The grab irons will be the BLMA .008" diameter stainless steel wire scale 18" width straight version, which means the diameter of the holes will be .009" x 1.00750065 and the distance between the hole centers will be 18" / 160 x 1.00750065.  The vertical distance between the holes will be known after measuring the Ertl model.  That distance will be multiplied by .544375 and then by 1.00750065.  The vertical corner posts and the push-poll pockets must be added as well.  All of these details will be added to the rapid prototype master prior to casting, to insure glass-smooth surfaces with rivet detail in the large flat areas as well as to provide drilling guides for the grab irons and MTL vertical brake staff pins.

On the solid parts, the shrinkage factor is removed for now.  It will be reapplied prior to the STL file being generated for the rendering of the rapid prototype parts, along with the XY scale adjustment factor to compensate for the variance in the ProJet 3000+ that will render the parts.

Total hours invested to date:  16
Total material cost expended to date:  $0

Still all labor and no material cost to this point, but that will change shortly.  Next update next weekend, which will consist of preparing the digital solid model for rapid prototyping.

[up1950s]

I am getting the feeling that these are going to be the most researched and created non-all-brass gondolas in N .

[Chris333]

How thin are those rivet strips on each side of the ribs    might have to barrow David's reading glasses!

[bbussey]

How thin are those rivet strips on each side of the ribs    might have to barrow David's reading glasses!

No need for worry.  They will be part of the master, which will be an amalgamation of rapid prototype and etched components.  The result will be highly detailed castings.  See past threads on the construction of the Keyser Valley caboose masters.
 

[Chris333]

I know. I ment you'd have to borrow the glasses 

[bbussey]

Ah.  I'm very near-sighted.  Just have to take the eyeglasses off to accomplish the same thing. 

[Chris333]

Not that it matters now, but I just got ACL decals in the mail from Jerry Glow...

[bbussey]

Session #6

Time to multi-task.  First, it's time to render the rapid prototype model for verifying the ride-height and coupler height are correct.  Additional solid models are derived from the one created so far.  The reason for creating derived multiples is so that if any corrections are needed, they are made to the original which then affects all of the derived models.

For the first derived model, the body is reassembled, with one end being the flat style and the other being the corrugated style.  The build-adjustment factors for the specific ProJet machine the RP will be rendered by are added in, which in this case is 1.0055% on the X axis and 1.0025% on the Y axis.



The body is converted into an STL file, which is a text file consisting of a collection of coordinates that form triangles.  Each triangle is formed from three points in a 3D matrix, each point with an XYZ coordinate in relation to the origin (which in this case, is the rail height + the center of the carbody).  Curved surfaces are converted into a series of triangles.  The more fine the resolution, the less likely that is is visible to the human eye.  In other words - all curved complex surfaces of injection-molded models are an illusion, as those surfaces are comprised of a collection of very tiny triangles.  Note the near solid-black areas on the model, which represent "curved" surfaces on the model.



Obviously, the more triangles, the larger the STL file.  This particular STL file is 2GB.  To put that into perspective - the STL file for the XIH pilot model is 69GB; the STL file for the as-built X58 pilot is 52GB.

Also important is the build orientation of the model, which matters since the build-compensation factors are different on the X and Y axis of the ProJet.  The arrow represents the path of the print head.



The STL file and the above jpeg file are sent to the contractor for rendering, which will cost $100 and be the first material cost attached to the project.  The part will arrive in 7-10 days.

The second derived solid model is the one that is prepared for the resin castings.  The sides are separated from the floor, because if possible all outer surfaces should be on the Z or "Top" plane (face-up) when being rendered on the ProJet to avoid visible layering.
 


At first, the sides were being separated only for rendering purposes.  But since this is a craftsman kit, the modeler should be capable enough to attach sides to a floor.  So keys and keyholes are added to the bolster areas on the floor and sides.  Designing the model so that the final result is a flat kit will make it far easier to cast.



All the parts are rotated and arranged so that they are on the same plane, and so that all top surfaces will face the print head.  There will still be some minor roughness on the bottom edges of the parts that will need to be smoothed with jewelers files or fine sandpaper, but they will not be visible when the model is on the track.

The mold separation membrane is designed.  It also will link all the parts together on one component.





Next step is to add thickeners to the areas of the membrane approximately 1mm away from the parts.  While this results in more waste resin being used to create the castings, it also insures the parts both remain undamaged and are easily separated from the "fret".





There is relief on the underside of the parts as well.



The brass insets will be added to the rapid prototype rendered from this solid model prior to casting.  But this wont be sent to the contractor until the first model returns and everything is verified to be correct.

The straight centersill inadvertently was omitted, and is now added to the fret.







In the mean time, RFQs have been solicited from the various manufacturers on the OEM parts.

Next will be the testing of the first model to see if all the measurements are correct and how all the OEM components fit, as well as how it operates in a freight consist.

Total hours invested to date:  17½
Total material cost expended to date:  $100

[daniel_leavitt2000]

Bryan, is there any way of casting the straght centerrail in pewter? This would add a lot of weight to a lightweight model. This isn't as big an issue with the MTL fishbelly as it is already metal.

[Chris333]

I'm just talkin' here. But what would be the cost to have a block of aluminum or wax milled using your 3D drawings to make a mold for resin casting? Not that I see anything wrong with the way it's being done. Just seems you could skip right to casting.

[bbussey]

Bryan, is there any way of casting the straght centerrail in pewter? This would add a lot of weight to a lightweight model. This isn't as big an issue with the MTL fishbelly as it is already metal.

Sure, but then you would have a different set of shrinkage parameters to deal with, in addition to compensating for the (most likely) distortion of the spin casting method.  Much more costly also.  The steel floor slab on its own will provide adequate weight.  The MTL diecast fishbelly really is sauce for the goose and actually doesn't weigh much.  The advantage of the MTL part is that it exists in quantity and eliminates the need to design/produce an additional component.  If it was plastic instead of zinc, I still would have included it.

I'm just talkin' here. But what would be the cost to have a block of aluminum or wax milled using your 3D drawings to make a mold for resin casting? Not that I see anything wrong with the way it's being done. Just seems you could skip right to casting.

It's doable.  But much easier to do when you don't have two mold halves to match up, essentially a one-cavity mold.  Also, more costly and better suited to a handful of castings rather than a few dozen.

[JoeD]

Looking real good Bryan.  Love seeing the in progress stuff, you learn tons about the process.

Cheers

Joe
Not a Micro Trains Representative on Railwire N Scale Forum.   

[bbussey]

Not that it matters now, but I just got ACL decals in the mail from Jerry Glow...

If you can scan them at your highest resolution and send them to me, it will help.