Model Railroading > ScaleTrains’ BNSF Frostline Reefer (WARP 7) – Do or Die Tryin

This is Part 7 of a series on creating the Weld Warpage on the Scale Trains’ BNSF 72’ Frostline Reefer.
(Parts 1 thru 6 were posted earlier.)
 
The 3D printed Dies are a critical part of this process.  Each one is slightly different.  After the surface of the Weld Warpage is defined in CAD (Computer Aided Design software), it needs to be made into a solid part to serve as a Die.   These Dies are 3D printed to have a “Negative” and “Positive” of the Weld Warpage for a specific 2.5’ sub-panel.  The Negative Die will be on the shell exterior, with the Positive Die on the interior.  In theory, heat + gentle pressure from the Dies will mold the Weld Warpage into the shell. 
 
Slice ‘n Dice
The CAD geometry starts as a flat, featureless Plane.  Lines are added to divide-up the surface, forming a Mesh.  These lines are pushed and pulled to render the shape desired.  The fewer lines used the better.  (More lines are only needed if the shape is to be complex.)  Since the Weld Warpage is a gentle ripple, a horizontal line for each peak and valley works.  A few vertical lines define the width of the peak or valley.
 
Initially the geometry looked “blurry” and was difficult to use.  It took a few days of trial and error to change the Appearance parameters to provide enough contrast to show the contours.  CAD has the ability to make the surface look like different materials (concrete, steel, glass, paint, etc.)  To my surprise, the Glass option didn’t provide a shiny surface to increase the contrast.  After using the Steel – Stain option, the appearance was much better.  The image rendered had high contrast, making the contours much more visible.
 
Do Me A Solid
Once the Plane is defined, it isn’t ready for the 3D printer since it is merely a very thin surface in virtual space.  Only solid bodies can be 3D printed.  In order to create a solid body from the Plane, the Extrude command is used.  It thickens the Plane into solid geometry that can now be 3D printed.
 
The contours are maintained when the surface is made thicker.  This means the top is the “Positive” image, and the bottom-side is the “Negative” image.  The bonus is both are needed to create the required Dies.  The upper Die presses down on the exterior shell surface (with the Negative image), while the lower Die presses up from the interior shell surface (with the Positive image). 
 
Bowled Over
Weld Warpage creates gentle peaks and valleys on the side panels.  For Weld Warpage to be visible, what should the distance be from the top of a peak to the bottom of a valley?  On the prototype the panel it was about 0.5”.  This was enough to cast the signature shadows.  Would this be enough on the model?  The 0.5” (prototype) would be 0.006” actual inches on the model.  An average sheet of paper is 0.004” (depending on the “weight” of the paper).  Would that be enough to highlight the Weld Warpage on the model?  It seemed like a good place to start.  It can always be increased.  The desired effect is for the panels to have a subtle waviness.  It is easy to over-do it and end-up looking like a bag of bowling balls.  :-0
 
New Fangled Angle to the Dangle
In a previous post I described the documentation convention for the Weld Warpage on a panel, which was the vertical position and angle of each peak.  This was tedious and time consuming.  Turns out, there is a better way.  An image of a BNSF Reefer is used to create a watermark “under” the CAD drawing.  It is used as a guideline for creating the CAD geometry.  Rather than document the location and angle of each peak, the more streamlined method superimposes lines onto the Reefer image.   Solid lines represent a peak, dashed lines are valleys.  This is much more visual, and easier to implement.
 
Disappearing Act
One curveball of using Forms is the contours can only be seen when there are large differences in the surface.  With the difference of 0.006” between the top of a peak and the bottom of a valley, the contours nearly disappear and the Form becomes largely featureless.  Slots need to be cut into the Die, which requires knowing where the peaks and valleys are located.  With the peak and valley features “missing”, it is impossible to determine where the Slots need to be positioned with any reasonable accuracy.  This is where the lines on the canvas image comes to the rescue.  Since those lines were used to position the peaks and valleys, they can also be used to locate the Slots.
 
 



Edited 1 time(s). Last edit at 04/15/24 08:49 by tmotor.

---

---

 
Slots
On the interior (Positive) Die, each peak and valley needed to have a Slot opening to allow heat to pass-thru.  Fortunately, the (appropriately named) Slot command (in CAD) is just the ticket.  All parameters (width, length, and angle) of the Slot are adjustable.  Following the lines on the canvas Reefer image, the Slots were positioned.
 
The amount of heat that reaches the surface of the shell is determined by the width of the Slot, the temperature of the hot air, and the amount of time the hot air is applied.  During the R&D process, each of these parameters will be adjusted to determine its effects on the shell.  The goal is to find the parameters that will repeatedly produce the desired Weld Warpage.
 
Mirror, Mirror…
The concept of using the same set of Dies on the opposite side of the shell was conceptually easy, but harder to execute than first thought.  The easiest method would be to take the entire set of images on one side, and mirror it along the B-end (or A-end) to match the panel pattern on the opposite side of the shell.  The problem is this also flips the direction of the diagonals 180-degrees.  The vast majority of the Weld Warpage (that I have seen) is angled upwards at about 45-degrees.  If it were mirrored, the diagonals would be pointing downwards at about -45-degrees.  Though this might look plausible, it wouldn’t be prototypical.  The solution is to rearrange the sub-panels to best fit on the opposite side of the shell.  The order was basically the same, with the diagonals kept at the same positive angle. 
 
This complicated the logistics of keeping track of which sub-panel was used at each shell location.  To help reduce the confusion factor, a new naming convention was needed.  Instead of trying to use the same Dies on both sides, Dies will be dedicated to a single side and NOT “reused”.   The amount of extra resin needed to 3D print them was negligible compared to trying to keep track of all of the puzzle pieces. 
 
Into The Void
There are some details molded into the shell that will get in the way of the external (Negative) Dies making full contact with the shell surface.  (Removing those details is an option.  They could be replaced with 3D printed parts.  But if this can be avoided it would save a lot of time.)  The Negative Die (that goes on the outside of the shell) needs to have a void where a detail is located on the shell.  The details that are the most troublesome are the brackets that serve as door-stops.  Their location within the panel is different for each side of the shell, so trying to use that Die on both sides of the shell would not work.  This need for customization is the main reason to have the Dies dedicated to a specific side of the shell.
 
The new naming convention will look like this:
-  The 12 road numbers will be given a letter designation (A, B, C, D … M),
-  When looking at the shell with the HVAC unit to the left, those sub-panels will be numerically incremented starting at 100 (100, 101, 102…). 
-  If the B-end is to the left (HVAC reefer unit is to the right), then the sub-panels will start incrementing with 200. 
 
This Alpha-Numeric code (X999) will be molded into the Die.  By reading the Die’s code, it indicates the road number of the Reefer, the side of the shell assigned to the Die, and its position on that side of the shell.  So, even if all of the Dies were all mixed-up in a huge pile, I can pick one up and will know which road number it is assigned to, the side of the shell it does on, and where to place it along that side.
 
My Good Side
If the prototype image has the HVAC unit on the left, then the image of the model should have the same orientation.  After the matching side was completely documented, then the other side needs to be done as well.  This was accomplished by moving most of the panels to similar positions. Then a bit of a shell game was played with the oddball panels on the ends.  They were moved around for a best fit, that kept the diagonal pattern consistent.
 
A Positive Impression
Each sub-panel will have a pair of Dies, one Positive, and the other Negative.  Each will be given the same code, since they are a matching pair.  The Positive Die will have Slots in it, the Negative Die will not have any Slots.  With the shell sandwiched between them, they will press from both the exterior and interior.  It will take a bit of trial and error to determine how much pressure is needed.  Too much pressure from the Dies will squish material out the sides, which will cause the shell to be out of square.  This is to be avoided.  The desired Weld Warpage is merely cosmetic.  The goal is to have the shell remain square and true. 
 
 
Dave
 



Edited 4 time(s). Last edit at 04/17/24 10:47 by tmotor.

---

---