I’ve been busy 3D printing lately. So busy I’ve gone through nearly eight 1Kg spools of filament so far. That’s almost twenty pounds of plastic and I’m not finished yet! Why so much? Well, that’s a bit of a story, but I’ll try to keep it brief.
I’ve been struggling with optimizing the casting process, or rather, processes. Roadbed bricks. Buildings. Infrastructure, like retaining walls and culverts and such. You name it. I’m struggling with it, and how to marry the castings with pouring concrete using the castings as forms for the concrete.
It’s difficult to visualize how all the parts need to come together to make one final piece. How do all the puzzle pieces fit together? How to hold the castings in place? How much concrete and where? A lot of questions with no answers, even with many design drawings to help.
The photo shows the initial mock up for the culverts and retaining wall of the station siding along the patio. That passenger car looks mighty lonely sitting there all by itself. Why so much mockup? I’ll blame learning curve for that. Learning from the mistakes of the initial design, and iterations that followed, to reach the final design… For the mockup.
And that’s just the tangent (straight) part. The curved portions require their own designs, and one for every different curvature! While one design is printing behind me, I’m sitting at the computer working on the next design element, be it a different part or an iteration of an existing one.
The work bench is littered with inadequate parts from failed designs. Stacks of parts yet to be assembled await the remaining pieces necessary to put the next 8″ unit together. Be it tangent or curved, each unit is based on an 8″ length.
Design Constraints
Why 8″ and not some other length? There are two main reasons why that size was chosen. The first was a matter of print volume. The old 3D printer’s bed is roughly 8½” x 8½”, capable of printing to a height of a little over 9″. All my original designs had to fit within those limits.
The new 3D printer has a print volume of 12″ x 12″ x 15″, so my later designs don’t have that constraint, except for the second reason. The sectional curved track requires a certain number of sections to complete a circle. The smaller the diameter, the larger the track section can be, requiring fewer sections to make a circle.
For example, the 10′ diameter sections are roughly 32″ long, and require only 12 to complete a circle. By contrast, the 20′ diameter sections are nearly 48″ long, and require 16 to make up a circle. So here’s some math for you. What’s the lowest common denominator for those two sizes? That’s right, you guessed it, an 8″ length!
The next choice in the design may not seem so obvious, but it’s time to select the angle occupied by each roughly 8″ section. In the case of the above design capture, it takes three 10′ diameter curve sections to create that 90° arc. It takes four of those 8″ units for each curve section, by three sections, for a total of twelve segments. Each sweeps a 7.5° angle.
While the wood stringers and other structures made up of thin, “laminated” strips may be formed into continuous curves, we don’t have that luxury when it comes to concrete. Granted, continuous monolithic slabs can be formed.
But when multiple, separately cast parts are involved, not so much. The design has to be segmented to allow for many individual parts to be cast independent of one another. And at different times. These cast wall parts are made in assembly line fashion, one or two at a time.
Design Decisions
The whole point of 3D printing these mockups is to better visualize the overall casting approach. Let’s take a closer look at what we’re dealing with. Near the center and moving to the right are the casting molds for a cut stone retaining wall and 8′ diameter arch culvert. Note those additional pieces on either end.
These pieces can be fitted as necessary to create 7″, 7½”, and 8″ long castings, roughly ¼” thick, made using concrete patch. This is very similar to the process used to cast the Downtown Marketplace building faces. Those older castings were made using a brick pattern sheet and crude foam blocks. But that’s another story…
Another piece is precision fit to the arch of the culvert section, knowing the difficulty of slicing a piece of foam to fit that opening. The piece on the very right is meant to cast the “pillars” used to disguise the joints between each of the wall or culvert sections.
The top left portion shows how the various parts, including the casting mockups from the bottom half, fit together. This helps to visualize developing a plan around pouring concrete, the external mold parts, and the steps involved.
It doesn’t reveal all the pitfalls that await, like how to form that depression for the track or how to form the pillar in place over the joints, but it will help to reduce the number of avoidable mistakes before they become mistakes.
Discoveries
That’s not to say I’ll find every problem by constructing these mockups, but it sure does help to get “hands on” experience with parts before the first casting is created. It’s too late to find out the track won’t fit in the casting or the passenger car steps will hang up on the pillar capstones once the concrete’s already set.
I’ve already found that the track won’t fit the curved mockup, we’ll call it a “track trough”, so the design needed modification. Even then it’s a tight fit, but hopefully that will help hold the track from going places, especially with the pups pounding on it.
And while the Bachmann passenger car steps clear the capstones with no problem, I have yet to try it with the new USA Trains heavyweight passenger cars that are nearly twice as long. At almost three feet long, it will require two of those 32″ sections to accommodate testing even one car.
I suppose with a little “modification”, those early sections already assembled with the faulty “track troughs” could be reworked to allow the track to fit. The original straight section was divided into two straight segments at half the angle each.
In other words, if that design drawing above had 24 segments instead of 12, and each 8″ unit included two segments rather than one. Looking ahead at the next steps is much easier with mockups in hand.
Next Steps
Speaking of next steps, what are they? Now that mockups can be made in assembly line fashion, it’s time to address how things need to come together to produce the desired outcome, a more or less permanent concrete fixture.
First is to define the process, from start to finish, and the steps involved. This includes what forms and external rigging are needed, along with an estimate of the amount of concrete needed, optimizing for 60# or 80# bags if possible. Of ultimate importance is how to keep it puppy proof while curing.
Once the process is defined, it’s time to test how well it works, making changes as needed. But even then, many questions remain. This will most likely be an iterative process, experimenting with different approaches before deciding on the final definition.
- Will the new, “wetter” mix allow the track relief to be worked in place?
- If so, will a mold need left in place (to overcome slump)?
- If not, will it take a roadbed brick equivalent to be laid on top of the fresh pour?
- Will each 8″ unit need to be poured solid or can it be made hollow, like a concrete block?
- If solid, how to allow for segments and/or expansion joints?
- If hollow, could a sacrificial 3D printed insert be used and left in place if necessary? How will that affect simple footing?
- Will the retaining wall and culvert castings require a dedicated footing? The designs simply use a block of wood in place of a footing currently.
- No dedicated footing – Simple formed and allowed to run out the bottom as one monolithic pour.
- Dedicated footing – Needs to be poured first then built upon.
- Determine how capstones work. Cast as separate piece and placed atop the pour?
- Modify designs for “staircasing”, i.e. gradually build altitude above terrain or lose it as terrain rises.
- Measure and record terrain height map. See if it can be imported into SketchUp.
- What format for import?
And honestly, some of the next steps have nothing to do with mockups, or even casting at all. It’s been a balancing act between work, home improvement, and making progress toward these Barkyard goals. There’s always something else that needs done first!
Excuses, Excuses!
Work has been absolutely brutal lately, busier than it’s been since I started there more than two years ago. As for home improvement, let’s just say it’s not our goals, it’s our individual goals combined. Ann wanted the raised bed planters along the fence by the driveway removed.
I wanted to get all these 3D printed mockups and jigs complete enough to get the assembly line started, saving all that hard work of removing the planters until Fall when the weather finally cools down from the “feels like 107°” by eleven o’clock in the morning, already here at the start of Summer.
Ann removed one of the planters and even transplanted the ponytail palm herself. She did not ask for, nor want my help. Unfortunately, that left me with bent screws and the aftermath of just ripping everything loose. Not wanting to sound like sour grapes, but not the kind of work I prefer doing in the Barkyard either.
That left behind a large area of dirt in need of turf. I’ve had a 7′ x 13′ roll waiting for placement elsewhere that got “requisitioned” for this task. The hardest part was getting that old chunk of turf out of the way, now full of dirt, and twice as heavy.
Always Something Else That Needs Done First
That revealed the ragged, rotten bottom edge of the 6′ tall fence panels we put up not even five years ago. Add to that one of the 4×4 posts is rotted out right at the ground, allowing the fence to sway with the breeze, and it’s time to replace that fence before it blows over.
And while replacing the fence doesn’t necessarily mean the other raised bed planter has to go, it does mean the dirt has to go somewhere while the fence panel gets replaced. The fence used to be only 4′ tall, but was replaced with 6′ tall fence to keep Brigel from jumping over it and chasing the neighbor’s cats into their yard.
Now that Brigel has crossed the Rainbow Bridge and the neighbor cats are seldom out during the day, we’d like to go back to the 4′ fence. Well, guess what Lowe’s no longer carries in stock and must be special ordered and delivered, to the tune of an added $75? All for three fence panels!
Plan “B”
Time for plan B. B, as in cut off the rotted Bottoms of the 6′ panels to make them 4′ panels! I got set up with the makeshift bench on saw horses, ready to make the cuts and coat them with wood preservative, while Ann and Nick handled wrangling the panels and posts.
We got ahead of ourselves on the first panel. It went up before I could cut the post down. Try as I may, I couldn’t get a clean cut, even with Nick removing the panel out of the way. It wasn’t certain that the rotted post was rotted off until the last screw holding the panel to it came out and the post toppled over.
Nick was able to fish the rest of the old post out of the ground. What was left of the post was just tall enough to go back in the same hole, once I gave the bottom a coat of preservative that is. Cutting those last two panels went quickly, but Nick was having a time of it, getting those two gate posts separated.
That gave me time to cut the gate down as well. One of the gate posts needed cut to length, and both needed a coat of preservative before they went back in the ground. Things kind of went downhill from there.
All Downhill
The drill bit broke off when it hit a screw in the post while Nick was making a pilot hole for the new hinge location. That meant even more work as the entire hinge needed relocated on the gate to avoid the bit still stuck in the post. Then one of the carriage bolt’s threads were stripped when it hit another screw in the post.
But even with all the trouble at the end, we managed to knock it out in just two and a half hours! The only things that remains is installing a new gate latch. I mentioned I may have one in the garage we could use, but Ann was having none of that. She’ll go pick one out tomorrow.
Tomorrow came and Ann went to Lowe’s, only to be disappointed they didn’t have the latch she wanted. Turns out I had the exact slide bolt latch she wanted sitting right there in the bottom drawer of the desk in the garage the night before. Oh well, all’s well that ends well.
If you thought that was the last of the distractions, the relentless things that needs done first, think again. Ann decided to remove the other planter and transplant the other ponytail palm a bit further away from the fence.
Then came the inevitable new turf to replace the old, now inadequate to cover the area required of it. The new roll is sitting outside the fence, waiting on me to cut up the old piece into strips more easily handled than the entire chunk all at once.
But wait, there’s more!
The split unit A/C in the garage decided it was going to flood the shelves, workbench, and table saw beneath it when the condensate line clogged up. I had just spent the last few weekends getting things squared away enough to start using the casting and trestle workbenches again.
Nope. Not this weekend. Now I get to move everything out of the way, everything I just moved out of the way of the casting bench by putting it away! And now I get to climb up and down a ladder with my bum knee just to get thing apart enough to work on it.
In the end the fix was to vacuum out the clog in the drain line. Sounds easy enough, but when the vacuum is on the other side of everything that had to be moved out of the way, now in the way of getting the vacuum out… Can’t win for losing sometimes.
The good news is everything is put back together and working again, without flooding the garage. Again, not so much sour grapes as comic relief for others. Life is what happens when you’re busy making other plans indeed!
It’s easy to see why it takes me so long to get anything accomplished on the Barkyard.
Other 3D Printing
And the 3D printing hasn’t stopped with casting mockups. I’ve even designed multiple trestle jigs for assembling the massive curved trestle from the deck to the new bridges and beyond.
Beyond that, it’s time to do something with that PET-G I’ve had for over a year now. Everything I’ve printed so far has been PLA. I bought so much PLA, and in so many different colors, that I didn’t notice my favorite maker, 3D Solutech, went out of business!
They are the only filament manufacturer I’d found that had such a wide variety of color selections. All the other have red, green, blue, black, white, gray, and that’s about it. Maybe yellow and orange. None of them have denim blue or steel blue or wheat or even brown for that matter.
Thankfully, white and gray will do for now, and I have plenty of it. At least another 8Kg anyway.