Home Page
PRR Chartiers Branch
The Layout
My Library
Site Map

Peter's Model Railroading | The Layout | Stops | The Hazel Mine Complex | Hazel Mine Tipple
Tipple Support Construction 2

If you have been following along, you will have read that I wasn't happy with the results of the first version of the coal tipple support structure, the part of the building that holds the main sheathed building up in the air. Using Evergreen Styrene's delicate styrene parts made it look great, but it caused the model to warp as more and more pieces were attached to it. It is my hope that this second version is going to be "the one". While "scratchbuilding" may make some people think that it is cheaper, these kinds of projects are not cheap, so having to re-do a portion of it is an expensive exercise.

So, in reviewing what I have learned from the first version, I am going to make some changes to how I build the second one. First, I am going to build it in-place on the module. This makes it harder to construct (due to its overall size and complexity), but that guarantees that the parts fit the concrete pads I already have on my module, and since it will be directly mounted to the module, that should eliminate any warping and thus misalignments. Second, I am going to switch from Evergreen styrene to Plastruct's ABS strips. Plastruct's material is a bit thicker, or cruder, but I have used it before and it does not warp at all. When the whole support structure is painted in "weathered black", the thickness of the columns' webs should be all but unnoticeable. Third, I am going to create a CAD drawing of the whole structure before building it. Fourth, there will be a slight compromise in the construction materials as Plastruct doesn't make scale 10" C-channels in their ABS material, so I'll be using scale 12" H-columns instead for the main vertical and horizontal beams.

When I started the construction of the first version of the support structure, I had a 3D CAD program on my computer, but it was just too difficult to use, especially for such complex designs as this support structure. However, while I was building "version 1", I discovered OpenSCAD and really liked its approach. Instead of doing a bunch of heuristics with the mouse and the CAD program misinterpreting my intentions, or dealing with the complex user interface of the program itself, OpenSCAD makes it possible to simply type in the dimension you want along the X-, Y-, and Z-axis. There are a few commands to learn, but they are all well-documented. Using that CAD program, I built the design you see here.

This close-up shows the design for how the parts attach to each other. The vertical columns are a scale 12" square H-columns. I will be using the same material for the horizontal pieces going between the long sections (the ones going away from the view shown here). In between the H-columns (going left-to-right here) will be 4"x8" I-beams that fit within the webs of the H-column.

This screen capture shows some of the improvements made to the drawing itself. I changed the colors on some of the parts to make things easier to see. I also added the concrete footings, which are now based on the actual locations of where they are on the module upon which this structure will reside. Since those are already installed, I figured that it is best to have the design match those. You are looking at the west side of the structure, with the Chartiers creek and the mine's entrance being on the right. The front of the module is on the left. Note that the center vertical H-columns are taller than the outer ones. This is because I built the main building such that its floor is inset from the bottom of the building. So, these H-columns are taller to still make contact with the main building's floor.

The next step in the design was to add the cross braces (blue) down the length/depth of the structure. On the bottom level, only the back 4 bays have the cross bracing, as the other bays are used for rail and road traffic to pass under the coal dumps. For the braces I will use Plastruct's 4" wide L-beams (part #90002). The orientation of these screen captures is that the creek is on the left of the building, and the front of the layout is on the right.

The design has also been updated to include the "missing" back, left corner area, as it was in the prototype.

Next, the cross braces in between the vertical columns, going from left to right, are added (purple).

The first set of girder plates (small gray squares) tie the cross braces together. These are a bit harder to see, so I have zoomed in on a few of them. They are faced toward the interior of the structure, the same as how the L-beams of the cross braces are oriented. The prototype engineering drawing does not show these for the left-to-right cross braces, as those L-beams are not touching each other.

This view is of the very back of the structure, to show the large plate girders that provide support for the L-beams of the cross braces where they attach to the main horizontal and vertical beams and columns. The right two rows have plates facing toward the left, and the lest two rows (only one is shown) have the plates facing toward the right. That's how the prototype drawing has them.

This view shows all of the plate girders in place (gray). While time-consuming and tedious, building this design did help me think through the process of how I am actually going to build the model. I have also captured the desired sizes of the girders, so that should help me with mass-producing the quantities I need of each size, which should help speed up the construction phase of this model.

Drawn in red are the corner braces used only in the lower open bays.

This is the entire support structure's completed design.

And this diagram approximates what it might look like when it is painted.

This is the ground-level side view of the structure.

So, now that I have this great design made, reality sets in. I need to buy the raw materials to actually construct this thing. As I stated before, I am going to build this using Plastruct's ABS strips. After analyzing the CAD drawing, I concluded that I need the following:

90064, H-6, 24" strips; need 21 strips, i.e. 6 packages (for the vertical columns & left-to-right horizontal beams).
90023, B-4, 15" strips; need 19 strips, i.e. 3 packages (for the front-to-back horizontal beams).
90002, A-2, 10" strips; need 99 strips, i.e. 10 packages (for the many cross-braces).

Update: Apparently due to the supply-chain issues, Plastruct does not have the 90023, B-4 strips anymore. Oh well, such is life! The parts in orange have been replaced by using the same H-column pieces used for the vertical columns and other horizontal beams. That doesn't quite match the original design of the tipple, but it'll have to do.

There will be a total of 460 individual pieces that I need to cut. This does not include the fabrication of the angle braces in the top corners of the open bays, nor the thin styrene pieces that I will make to represent the gussets. To be able to make that many repetitive cuts accurately, I have decided to make the investment of a top-quality mini table saw. Scratchbuilding is a major part of the enjoyment of this hobby for me, and so a tool such as this will be put to use very often. It is a very well-built unit, and it weighs a ton (~30lbs). I ordered a number of accessories to complete the saw (not shown in the photo, which just shows the base saw only), so that I can use it for this project. Learn more about this machine by visiting my page dedicated to it.

The next step in the construction phase is to prepare the module. Due to the open nature of the support structure, it is hard to reinforce the model, as I learned the hard way when I built "version 1" of it, which wound up warping. At that time, I wasn't worried about "finishing" the scenery base layer as the building would be removable. However, with version 2, I have decided to build the model in-place, meaning that at some point in time certain areas of the scenery base layer are going to be hard, if not impossible, to reach anymore. So, I took the time to add several very thin layers of various scenery materials to build up a coal tipple yard that looks like it is covered in coal bits and coal dust. On the far left, where local automobile traffic comes by to fill up their trucks with coal, has been smoothed over and a bit of dirt is visible through the coal dust. In the back on the right, where we are starting to be some distance away from the tipple itself and closer to the creek, you see some more dirt and some basic greenery. The area under the tipple is critical now; the rest of the module can continue to be enhanced in the future.

The actual construction phase of the support structure, version 2, of the coal tipple is, finally, beginning. Using my new mini table saw, I cut 24 pieces of Plastruct H-column (12" square) that are a scale 33 feet long. These are used, vertically, to rest the outside of the main building on.

I then cut 96 pieces that are 9 feet long, which go horizontally between the vertical columns, from left to right (when facing the front of the building).

Next, I cut 18 pieces that are 38'5" long (background, center of photo) which are the central vertical columns to support the interior of the main building (the floor of which is higher than the bottom of the exterior walls).

So, with these 138 pieces, I can start building the left-to-right framing. This photo shows my set-up for gluing the various parts together in one session. The wooden popsicle sticks are used to make sure the horizontal pieces are all going to be the correct distance from the top of the framing. The squares are used to make sure everything stays square, and the weights are used to make sure nothing moves while I carefully apply the Plastruct "Plastic Weld" to the joints.

The majority of these frames have two taller center columns, so I had to change my set-up a bit. I cut three scale 1-foot-wide strip wood pieces to act as the bottom spacers. So, this is the glue-up jig. The top horizontal pieces aren't supported, but by being careful when applying the glue, they should not move. The 3-2-1 block on the left puts pressure on them to stay put. The triangle at the top is just to verify that the two center columns are even. The metal square in the lower right-hand corner is just to give some weight to the builder's square, as it can move easily on the glass.

Here is one of the center framework pieces put in its position. It is just loosely placed there, held up by a metal square, just to show what it looks like; there is more work to be done to these before they are ready to be mounted to the module permanently. The Plastruct "Plastic Weld" that I am using as the glue is working well. I only had one of these delicate structures fall apart on me, but that was likely due to me not putting enough glue on a couple of its joints.

All eleven of the framework pieces that go from left-to-right are now constructed. The one at the top of the pile is the one for the very back, where the fourth vertical column is not needed. The one on the right of the photo is where I have already started to install the gusset plates. These are used to provide support for the diagonal crossbraces, which come later. However, in the meantime, these plates also function as reinforcements of the joints. I am making these gusset plates out of 0.020" sheet styrene, which comes out to 1-1/2" scale inches thick. Depending on where the gusset plate is positioned, it is either 3'x3', 3'x2', or 2'x'2 in scale dimensions (the latter still needs to be added to the outside corners).

The framework sub-assemblies have now been covered with the gussets. The two in the front are the back one and the front one. In all, there were 129 gussets cut and installed.

I have gotten started with installing the crossbraces that go in between the gussets. These are made by cutting strips of Plastruct ABS angles (part #90002) that measure 6 scale inches. I cut them to length using the mini table saw, which went OK, but it did rip some of the ends to pieces (it went better when I ran them through the blade quickly). In the end, when the whole model is finished and painted, you are likely not to see that, so I'm not too worried about that.

Here is a close-up of the crossbraces, mainly to show how they are oriented. I followed the prototype design diagram to position these. I am not going to bother installing NBW or rivets on these as they are most likely not going to be visible, and it would take a very long time to do this many of them.

All 11 sub-assemblies are now complete, with the installation of the crossbraces at the bottom level. There were a total of 128 crossbrace pieces cut and installed.

Thinking through the process of how I am going to install these vertical sub-assemblies and attach them to the foundation blocks already installed on the module, I decided that I'd want some sort of angle bracket at the bottom, on both sides of each of the vertical columns. My first thought was to form these out of brass, but I'd have to cut those from a sheet of brass, and then shape them to be a perfect 90-degree angle. The odds of getting that right, especially having to do 86 of them, were slim. Then, one day it dawned on me that I could just use the angle strip styrene. This is the route I chose. I cut these from the strip at the widths approximately matching the interior space of the vertical H-columns. I then put the sub-assembly vertically on my glass plate, put lots of squares and weights against it to ensure that it was at a perfect 90-degree angle to the glass plate, and then glued these small angle brackets to the bottom of each of the vertical H-columns, one on each side. This does two things: one, it makes it easier to balance the sub-assembly when it is time to permanently install it on the foundation blocks. And, two, it gives me just a bit more glue surface to which the 5-minute epoxy can adhere when it is time to glue them in place.

Next was one of those two-steps-forward, one-step-back kind of situations. Having attached the styrene angles to the bottom of the Plastruct ABS vertical columns of the coal tipple support structure's frame sections, I tried to do a "dry-run" of how I am going to glue those into position on the module. In the process several of the angles simply dropped off. The lesson learned is that styrene and ABS cannot be glued together using Plastruct's "Plastic Weld". I am now reinforcing each of them with superglue, and then drilling a hole through the two angles and the H-column, and supergluing a small section of brass wire in that hole, to act as a "bolt". This took a while to do.

With the angles secured to the frame sub-assemblies, I can now actually focus on installing those. The difficulty in doing this is, of course, to make sure that all 11 sub-assemblies are even and square to one-another. After thinking this process through, I came up with the jig shown in this photo. It is made out of strip wood. I only had pieces that were 24" long, but since the building spans 29", I had to glue to of them together to obtain the desired length. I then used two aluminum bars and used 5-minute epoxy to glue those to the wood. This was because the wood was flexing too much. The bars are actually ones that are used in filing cabinets upon which the file folders hang! We got rid of the cabinets, but I salvaged the metal bars from them, figuring that they'd come in handy some day! I then installed 5 lateral pieces of wood that were attached in a perfect 90-degree angle. Note that the photo distorts the shape; the jig is perfectly straight and square.

And this photo shows its intended use. I built it such that the frame sub-assemblies tightly fit inside the jig, with the front and back ones resting up against the front and back lateral pieces of wood. Since the jig is straight and square that means the sub-assemblies will be installed straight and square. When they are all installed, I will cut away the jig.

And finally, after many, many months, the first sub-assembly is being glued to the foundation blocks. I put several weights on the jig after getting it perfectly aligned with the foundation blocks. I then put a 3-2-1 block on a set of foundation blocks, to make sure that it is level, and then put a machinist square on top of that. Its long edge fits inside the groove of the H-column of the sub-assembly. After a dry-fit test, I mixed up some 5-minute epoxy, applied it to the bottom of the H-columns, and carefully positioned it on the foundation blocks and up against the lateral piece of wood of the jig. Then, I slid the machinist square into position and stepped away to let it cure.

Here are ten of the sub-assemblies permanently glued to their "concrete" foundation blocks. While not difficult to do, it was stressful as each needed to be as square and straight as possible. Unfortunately, when I very carefully put the main building on top of them, the eleventh sub-assembly did not match up with the back end of the building. I have no clue as to how I could have made such a mistake. However, the back side of this structure will need lots of work down the road, so I will deal with the error at that point in time. I am leaving the alignment jig in place for now, until I get all the horizontal pieces installed between these vertical frames.

And this is the start of installing those horizontal H-column pieces. I needed some solution that would keep these pieces even and level while the glue set. I found this brass bar that was straight and so I positioned it atop the vertical frames, cut and shaped the H-column piece, and clamped it in place with these yellow plastic clothespins (they are great for when you don't want a lot of pressure, and they are lightweight). After getting the part lined up correctly, I could then apply the glue and walk away to let it cure. In this photo, the left side of the H-column is in contact with a styrene gusset plate, so I needed to use superglue for that connection. On the right-hand side, though, it is ABS to ABS, so I used the Plastruct Plastic Weld. This went well.

It took me a relaxing Sunday to get all 9 of these pieces cut and installed. Since each required setting up the mini table saw to make the cut, I cut two of them, because the odds are good that the same-length piece is needed in the middle level (this turned out to be mostly true).

The middle level's installation of the horizontal H-column pieces was a bit more tricky, in that there is no way to put a brass bar across them. I spent some time thinking about how to do this and tried a couple of ideas, until this simple idea popped in my head. That is, I clamped two popsicle sticks to the bottom of the inside horizontal pieces, so that the new between-the-frames horizontal piece could just rest on the sticks, ensuring that it would be level and even on both ends with the interior framing. Not shown in this photo is that I also put another yellow clamp where the two popsicle sticks meet, so that they were more even.

As I got toward the back side of the structure, the gaps between the frame sections is smaller, and the popsicle sticks were too long. I also discovered that the popsicle sticks have a slight bow to them, which affected the positioning of the H-column piece. I then found two leftover pieces of strip wood, and this is the method I am now using for all of these types of installations. The shorter length makes them easier to position.

And so here we are after a week's worth of work. I have more time to devote to the hobby on most weekends, but during the week I was able to glue two of these horizontal pieces in position every day, so that helps. I have to do this several more times...

One of the things I have to do quite often is measure the inside distance between two spots on this structure. The way I am doing is by tightly squeezing two pieces of wood to fit within the opening and holding them together with a pair of tweezers, the kind that clamp all the time unless squeezed (in woodworking terms these boards are called "story sticks").

I took this photo when I completed the left two front-to-back sections (left to right in the photo). Each of these sections requires 72 individual pieces to be cut, shaped, and glued into place, so this takes a while.

This photo shows the middle-right front-to-back section (left to right in the photo) completed. It takes me about two weeks of modeling time to do one front-to-back section.

When I put the first top row of horizontal H-beams on the left-most front-to-back sections, I used a 2-foot brass bar. Now that I am working on the last one, it dawned on me that I have a 4-foot straightedge in the garage, and it has a nice lip to which I can clip these plastic "clamps" to hold the ABS piece in place while I apply the glue. This approach makes sure that all of the tops of these H-beams are even, to offer a flat surface upon which the main building can sit.

I was then able to complete the cross-bracing in that last, right-most top sections.

Here's a full side-on view. You can see the styrene-based support structure I had built last year, in the background on the top shelf. Also in the background on the next shelf down, you can see a partial view of the main building that I finished building last year.

The back three bottom sections also need the crossbracing installed, so I've added the horizontal beams at about two feet above the ground.

In my enthusiasm for completing the top level's crossbracing, I didn't plan far enough ahead for the last three bottom sections that also need the crossbracing. This prevented me from having gussets that could support the crossbraces. The gussets at the very bottom I could just install as before (although they are incredibly difficult to reach, as I used the bottom access I had for doing the same thing at the top level, which I don't have at the bottom level). For the top gussets, I added a scale 1' x 3' strip of styrene in the upper left corner, and attempted to make them look like extensions of the upper level's gussets. They, too, are near impossible to reach and get in there. I only did this for the left side of the opening, because the right side at the top corner of the opening literally has nothing meaningful to be glued to. So, I just made sure that the brace piece was long enough to just attach to the top level's gusset. While these are definitely "hacks" that I am not proud of, it is my hope that once painted and the fact that these are at the very back of the structure, these work-arounds will not be visible from the normal viewing angles in the future.

I have finally completed the construction phase of the main tipple support structure, which has been the topic of this page. This shows the front, side view.

This view shows the support structure viewing it toward the west (the right side of the photo is the front of the module).

Just for fun, I temporarily placed the main building on top of the support structure, and added a two-bay hopper to the photo for perspective.

The entire model is 12 inches (30.5cm) tall. The support structure itself is 26 inches (66cm) long and 6 inches (15.25cm) wide. I made the decision to scrap my first attempt (version 1) back in early October 2021. While I was designing this new version in a 3D CAD program, I did some other projects to take a break from this one. I also needed to buy the mini tablesaw before I could get started on this project. That tool has been the key to getting the work done and for all the parts to come out with nice straight cuts. Well worth the investment. The first vertical frame section was built on April 23, 2022, and this photo was taken on September 17, 2022. So, it has taken me just shy of 5 months of hobby time to build this tipple support structure.

The final task on the tipple support structure was to install the upper corner braces on each of the six open bays. I had intended on using flat brass stock, but I didn't have any on hand, so I decided to just use scale 4"x4" styrene strips. It may not be a perfect match to the prototype's drawings, but it looks about right, size-wise.

There are a total of 890 pieces installed to construct just this support structure!