Continuing Saga …

No images again, Myles. 

Something's screwy. I've been uploading to this thread for over 6 years and basically do it the same way. I'll reload them. There were a ton of them!

Despite being no where near size limits, anything more than six files gives me issues at times over the years. Over ten, chances of failure doubled. The OS doesn't seem to matter (Win./ Droid anyhow. i stuff?.. ?)

Okay… I reloaded all of those pictures. It almost takes as long to reload as it does to write the post in the first place.

Today I have a ton more. Like I said, working on many aspects at once creates a lot of pictures. Each time I glue something with Titebond, I have to let it sit about 1/2 hour before de-clamping.

Today's production included finishing the pilasters and crane rail support installation on the left inner wall. I also cut, shaped and installed the upper supports for the roof trusses.  I assembled the very slick laser cut railing for the gantry. I started sorting massive numbers of windows and figuring out my process. I pulled the clamps off the first gantry rail assembly and J-B Welded the other one, which will be cured tomorrow. And I tried the left wall's fit next to the floor and found two problems, both which offer some challenges.

I finished up adding the pilasters to the left inner wall, then glued on the 8 gantry rail supports. To make sure that these were level with each other I used a straight edge and periodically checked a reference location between the straight edge and the top of the 2nd level window openings. It didn't very more than a few thousandths.

The pilasters were keyed to the reference line I drew at the bottom accounting for floor thickness. BTW: I re-glued the end pilasters with their correct 1-wall-thickness offset.

While this was drying, using the new piece of 3/4" pine I bought last night, I cut pieces that conformed to the contour on the template pieces that I had cut. These templates came in handy! The bottom side is tapered for clearance of the gantry, and I put that new belt/disc sander to work. I was this particular part that I bought that machine for. At first I thought I would have to rough the stock off with the big belt, but the disc did just fine. It made a mess and I work a dust mask and goggles to keep from ingesting all that fine sawdust. 

I set the miter gauge to match the template's angle and in a couple of passes took off all the stock.

After this I used the shop vac on the machine, the room and me. I don't have a dust collector system.

I used the template itself to set the gluing distance for the 14 truss supports. Notice that the gantry supports skip every other pilaster.

After all were glued I weighted them all at once with a big level and some gravity inducing items.

I will flip this over tomorrow and add the pilasters on the outside and that wall will be completed.

While this was drying I assembled the 3-part gantry railing. When I drew this I wasn't sure if it would be too delicate to cut, but Stephen used a good laser board and the details came out perfectly. It is a 3-layer sandwich. The center layer has the lattice and the outer two are duplicates that give it depth. This was cut on self-stick material and it went together very nicely. This will be painted "equipment yellow" after the gantry is assembled.

I set this aside for later...

I opened up all the window frets and will spray paint them while still in the frets. These too are self-stick. Lots and lots of windows, but assembly is really slick and Stephen has included pre-cut acetate too. Much less work than my previous buildings.

When the wall was dried enough to handle I set in in place adjacent to the finished floor and found two areas of concern. The most egregious is the floor ending up being 1/8" too short. I hate when stuff is TOO SHORT. That 1//8" number is suspicious telling me that I took one wall thickness to much off the part design. Remember, the laser never lies. Any measurement error is all mine. Darn! I will splice material on and it will be okay.

This is one end in its proper position.

Here's the shortfall.

The second problem is the gaps that were under the gantry supports. Something ain't right! But my gluing job was almost perfect as seen in this image.

But look at the alignment on the truss supports… Notice, I've tried on the gantry track rail. It lined up nicely.

Here's the problem… The ground ain't level… just like the real world. There is a dip in the platform under the yard tracks in this location.

I can't bring the floor up to join the buttresses since it raises it too far above track height and the big O'gauge flanges will ride up on the floor. I can shim under the buttresses so the gluing surface will be intact and let the floor follow the terrain. There's one other option which I will explore. This is an L-girder layout. I may be able to adjust the cleats and risers holding up the yard platform. I don't know how much give there is since there's some plastering in the area, but it's worth a try.

On another topic, my Kalmbach special issue on the resurrection of UP 4014 arrived yesterday and it's a terrific book. Seeing all the things that UP's steam crew did to make that engine alive again was nothing short of miraculous. A note of interest. UP's shop is world class, but to turn and quarter Big Boy's massive drivers they shipped them off to the Strassburg RR's steam shop who has a wheel lathe large enough to service 1940's loco's drivers.

I also did one more thing. I've volunteered to do a soldering demonstration for the Military Modelers Club of Louisville. I've been soldering photo-etched parts which is not something that many folks do. I also have perfected hand soldering those tiny surface mount LEDs. These are both topics of interest to plastic modelers.

Myles, the project is looking great!  I think the platforms on the outside would be sufficient to give the right look!  I suppose if you don't like it and feel energetic later on, you could add ones between the tracks later.  I was wondering if you can't raise the table top sufficiently, you could just put a sill under all those supports.  Once in and painted, you will probably be the only one who would notice.

Thanks Mark! Yup! Could just dress up the buttresses, or just add a shim. I'm interested to find out if the raising the risers about 3/16" could take out the dip. The center section of the layout is the basically paved with OSB sitting on risers. In the center of the yard, the joint between two panels actually created a little hump. Not a real hump yard, just a teensy little bump, but I know it's there. Dips in the tracks in a yard don't matter. My main line track is pretty consistent.

Myles, yes I agree if you can raise the risers 3/16” that is best!

Too many images could be the problem… I generally don't have this many images, but usually don't have so many topics I'm working simultaneously.

Tried to raise the platform to remove some of the dip. Was marginally successful. I had to secure a block on the L-girder to give enough space for my Quickie Clamp to fit. The clamp's fixed jaw is reversible so it can function as a spreader, but it's minimum width was large than the space between the girder and the OSB. I loosened two risers in the vicinity  and exerted as much pressure as I could on the clamp. It did move a bit. I reattached the risers in this new position, but it's still about 1/8" low in the wall's center. I'll shim the buttresses.

Here's the clamp doing its work. Spax star-head screws are terrific. Even though they don't need a clearance hole, I use them since it insures that the piece you're screwing to pulls up tight.

I flipped the wall over and started laying in the pilasters on the outer face. The end two are wider designed to extend out one MDF thickness to wrap over the pilasters that will go on the end wall. I used a piece of the same material to ensure that the offset was, indeed, perfect.

I started laying in the remaining pilasters and wanted to find a way to streamline aligning them to the wall and each other. I centered one of them as best as I could and glued it in. I then placed the next one and measure the distance which I transferred to a piece of the 3/4" stock and cut two gauges. These came out pretty close and I was able to use them to space the pilasters to that well-placed first piece.

After doing a couple of them, I noticed that I was getting some off-center drift and corrected alignment before clamping them down. My gauges may have been a few thou off the exact number and I was getting some stacking error, or there could be some minor variations between window spacing. Either way, I was able get the first half into position. Notice that the pilasters aren't reaching the bottom. This WAS NOT supposed to happen. It's due to using "one size fits all" mentality. For the inside pilasters, I needed that clearance to accommodate the floor. I also needed the same clearance for the machine shop floor. But I DIDN'T NEED THE CLEARANCE for all the rest of them. Easy fix… I have a scrap wall which Stephen cut a bit short. I measured the piece I needed and cut it off on the scroll saw. I don't have a circular saw. For long rip cuts I really miss that. The scroll saw doesn't cut a very true line. I made the piece a bit oversize and will sand it down flush after the glue dries. My mantra: "It's not if you make a mistake… it's how you recover!"

While this was drying, I prepared and painted all the windows while still in their frets. Since I was going to blast them with Tamiya lacquer-based primer, AND it actually wasn't raining outside, I taped the windows to some cardboard and went outside and painted them. I had to be careful to not tape over parts of the windows that would be seen.

This is Tamiya's rattle can Red Oxide primer. I'm thinking to just leave them at that.  If I spray acrylic I can do it in the basement. My RR is a merger of the Pennsy and UP. That leaves Tuscan Red, Dark Green, Harbor Mist Gray or Armor Yellow as appropriate colors. I need to complete all the painting before assembling them.

Back to the Left Wall. The after part of the wall intersects with the Machine Shop. I needed to be sure that the pilasters didn't foul the mounting tab slot. I used the machine shop left wall as a positive stop to put in that pilaster. The fit is superb.

Then it occurred to me that the Machine Shop roof is slotted to clear the pilasters and therefore can be used as a positioning device. By doing so ensured that the roof would slip over the pilasters perfectly without any fussing. With the pilaster tops properly aligned all I had to do was get them nice and perpendicular. Those longer "window-looking" openings are going to be open spaces connecting the machine shop to the main floor.

This was the last thing that was done today. The engine house is going to be quite heavy since just this one wall has some weight. There's over 60 pieces just in this wall. I want to populate (eventually) the machine shop with some appropriate equipment. Western Scale Models used to offer a whole line of locomotive repair machinery. 

Already, I love the way this is going. Nice job, Myles. 

Myles, it is coming along nicely!  It's hard to believe you had a day without rain so you could paint the windows!!

Thanks Mark! It was really nice earlier today too so the roofer was able to put sealant on our brick chimney that was weeping moisture into our sunroom and another spot on a skylight that needed caulking. Tonight, on the other hand, the boomers have been booming.

Didn't post last night so there's two days of work to report on. Got all the pilasters on the the right side wall (inner and outer) and discovered that I had measured the pilasters correctly and actually had two different lengths cut: long ones that reached the bottom of the wall for the outside and shorter ones by 1 MDF thickness for clearance of the floor. Stephen packed the long ones with bubble wrap underneath the shorter ones. I had forgotten that there were two lengths and just started using them as they were in the box. It only affected that front portion of the left wall. The right wall has them going all the way down.

For the inside pilasters I used an 18" ruler to set the top position. I just had to push them up the ruler and make sure they were square. The Titebond is very slippery and even clamped, they can slide out of position, so I had to go back and check a couple of times. I would glue about five at a time which was the ruler's span. 

I blew another small dimensional error. The upper cap strips should have been one MDF width longer since they don't properly overlap the end caps at the corner. The end pieces are just clamped for this fitting. They'll be assembled next week. I'll have to graft a trim piece in the corner with a little filler. The building is a beast! With all the pilasters, trim and windows, it's not really model making… it's more like construction. What I'm actually building is sort of a small shed. It's carpentry. And it's going to weigh a ton, kind of like an IKEA flat pack.

I started painting the windows Harbor Mist Gray (UP gray), but ran out and switched to US NAVY Haze Gray. Haze is a tad darker. I'm going to try and keep the darker and lighter windows on different walls so the colors won't be so easy to compare. They're really close. I used Model Flex by Badger by airbrush. Their bottles have the same thread as the Badger bottles so it's very convenient. It almost has no odor and covers well. Being water-based acrylic, it does clog the airbrush eventually and I had to do major cleaning twice during all the window painting. I use acetone to clean the congealed paint. If you buy one of those inexpensive Chinese airbrushes, don't use aggressive solvents. They have O-rings that dissolve in acetone (as me how I know). Badger, on the other hand, uses Teflon seals that don't react to lacquers.

My attempt to raise the layout platform was marginally successful. In order to use a Quickie Clamp as a spreader, I reversed the jaws and had to screw on a separate block since the clamp was too wide to fit onto the girder itself. I loosened the adjacent risers, applied as much pressure as a I could and maybe got a 1/16" movement. I reattached the risers to this new height. I'll shim the buttresses.

Today I continued gluing things on, waiting for it to set while doing something else. Today, that something else was building windows and running some trains.

Windows are pretty simple and slick. Outer, self-stick frame, acetate glazing laser cut to the exact same size, and an inner, self-stick frame that goes on the other side of the glazing. The inner frame has a wider flange that fastens the window to the inner building wall. When I peeled the inner sash backing paper it exposes the very sticky flange. The windows aren't going in for a while. The building needs painting inside and out before they do. I was worried that the sticky would stick to things it wasn't supposed to so plan A was cutting the outside frames perimeter on the inner sash backing sheet leaving the outer edge in place to protect it. It was a slow process and I have over 60 to build, so I needed to find a better way.

Here was the Plan A window with the perimeter backing strip.

Plan B was much simpler. I just cut up some wax paper and put it between the finished windows to protect the exposed adhesive. I almost got all the small size done, and got all the mediums done. 

Last thing I did was start installing the all the inside buttresses. I finished the gantry supports and will put the truss supports in on Monday.

It's starting to take shape nicely.

Thanks Ken.

I just ordered my 3D LCD resin printer. I was thinking this morning, during my creative just-waking-up process, about what I'd use it for. Then it hits me. The engine house alone has opportunities for a high-res printer. I want to put HVAC units on the roof. These don't exist in the aftermarket in our scale. I have drawings of them. If I can draw them, I can print them. I have a machine shop on the building, but no machines. Western Scale Models sold a line of white metal machine tools for a narrow gauge. Again, I can find drawings on the SketchUp of machine tools and can draw them. If I can draw them, I can print them. This is just one building that could almost pay for the 3D printer. I'm now stepping firmly into the 21st Century!

On Thursday, I had the pleasure of making a soldering demonstration and clinic at the monthly Military Modelers Club of Louisville's plenary session. One of the members is the administrator of the University of Louisville's Advanced Manufacturing Lab (one of the most comprehensive in the country), and he has 3 of the small, low cost resin printers like the one I'm looking at. He says they're terrific. They're not particularly fast, but they have great resolution, can use any resin that cures at 405 nm wavelength and he would highly recommend it. He also offered to print things for me. That's cool too, but I'd like to have one next to my elbow. Not bad for a guy in his mid-70s.

Myles,

you are amazing.

I have always been fascinated with 3D printers and believe they are changing the modeling and sampling world. I would love one but I can't justify it. I wouldn't use it enough to justify the cost. On the other hand, I expect that you will make good use of it and will justify the cost in short order. This is another reason that this thread is one of my favorite. Please include your foray into 3D printing in your posts.

It is interesting to think about the relationship between modeling as we have known it and the new kind that is rapidly emerging as a result of 3-D printers, laser cutters, and various other new technologies.

My first thought is that is it kind of like the relationship between painting and photography. In some sense, they do much the same thing, but in other ways they couldn't be more different. Photography retains many of the requisite skills of the painter -- composition, framing, color management, etc. But it almost completely removes the manual skills that good painters must cultivate, and replaces them with other, more technical skill requirements. Automated modeling does much the same, I think -- retaining a lot of the "deepest" skills, but totally eliminating some others (while creating new and challenging skills to learn).

It is worth noting that, contrary to some early predictions, photography in no way eliminated painting--the two art forms coexist very nicely, even to the extent that serious painters often use photography as a tool in studying their subject matter. I kind of think that we will come to think of manual modeling and automated work as being two separate, equally-interesting activities, each judged by its own criteria.

Myles,  Yes I can see the enginehouse will be a beast!  The best part of your posts is where you tell us about your mistakes and times where you think of a better method to continue with the rest of the project!  All this thinking helps you and any of us young!

Pete, you comments comparing painting and photography, and traditional modeling and 3D modeling are very well put!  I couldn't agree more!

If you get anxious, and farm some "shopwork" out, don't forget about Andre's shop tools at River Leaf Models.( His 1:48 operating bench vice comes to mind everytime I look at one of mine) 

 HVAC; while newer ones are more similar to each other and many much smaller than they used to be; the older, the more imagination you can use...talk about variation! Everthing had more style years ago, even hvac 😄  On a modern one, an open panel with a keyboard& computer screen running Linux wouldn't be beyond you . 

  The best choice of system size itself taylored to the space inside, expected air travel and operation costs are a science in themselves anymore. Older systems often began with an office or two with other units being added much later to expand the treated area. Large and high traffic buildings go sar far as to read air pressure drops when specific doors open and compesate in an effient manner. That ability may or may not have heat/cool ability and may be integrated or a stand alone unit. ...bla bla bla... You can ask, but the sky is the limit on a big box with vents I've replaced "bomber propellers", installed "indoor Pacifics" and worked larger , improved paintbooth filter and flow for RR shops in 1:1 scale, piped diesels exhaust away from our First Responders vehicles and worked in the one of the best air scrubbers any surgery wing ever had. But nothing beat running in a 12foot squirrel cage fan because I could 😁... 🐿️🥜🍨.  🤔🗓️..😜

I just blew 20 minutes of writing a thread when switched to another site, but forgot to open another window. DOH!

This thread was about a nice Saturday session. I finished the right side wall, built more windows, downloaded some machine tool drawings from the SketchUp 3D Warehouse, and fixed some more problems.

I decided to fit the ends to the left side to see how square they came out to be. They were nicely square.

Then I tried the newly completed right side wall. Oh no! I got the offset wrong again! What happened this time? My wife always wants to know what I was thinking when I have these lapses. I explained that I had reversed the order in building the second wall. I did all the outside pilasters first and then did the inside. I had already offset the corner pilaster to overlap the end's corner. That offset confused me so when I placed the inside corner pilaster I already thought it was offset for the end wall, but I was mistaken. I needed to offset it one more time. The glue was cured. I needed to remove the parts using the plane blade. Titebond is stronger than the substrate and when you try and separate it, the MDF fails. Luckily most of the failure was on the pilaster and not the wall itself. I have many, many extra pilasters. 

While this was drying I was making more windows… lots of windows.

Having so many windows to produce gave me ample opportunity to find the best way to assemble. I found out that holding the window and bringing it in contact with the adhesive frame was easier to align than how I first was doing it; holding the acetate and bringing it to the frame.

I still have a few more big windows to build. I then had another faux panic! I have 29 small windows (13 for each side and 3 on the back), but I only had 24 big windows on the frets. I started writing an eMail to Rail Scale. During the typing it hit me. The left side only has 5 big windows due to the machine shop, so that adds up to 21 (13 + 5+ 3 = 21). Instead of being short, I have 3 extra. I deleted the half-written eMail.

Then I found another problem. My window arrangement on the right side was not symmetrical…it was supposed to be, but wasn't. Not enough space on one end to clear the wide inner window frame. The result was trimming the pilaster and then the buttress.

I trimmed the pilaster on the scroll saw, but the buttress, at 3/4" thick, was too thick for that machine, so I sawed the relief cut with a hand jig saw. I smoothed out the cut with the 1" belt sander. 

I removed the foam under the machine shop and prepared the site for that part of the building. It's being supported by the foundation blocks as I did on the main floor. I'm still deciding if I should fasten this floor to the main, or leave it just connected to the machine shop walls.

The other thing I was doing while stuff was drying was download a bunch of great machine tool drawings. These are all drawn 1:1 scale and needed to be shrunk to 1:48. Then they all need to be modified for 3D printing. That will take some time since all aspects must be read as "solids" by the slicing software. A "solid" in this instance is a shape that has all edges and faces as a contiguous thing. Any non-connection renders the object "not solid" and not printable. Then I have to evaluate them based on how small of a feature size is capable of printing. I'll do that experimentally. The printer arrives tomorrow and I'll have a pretty steep learning curve, but it's going to be fun.

I looked for larger machines like vertical turret lathes or Wheel Lathes, but came up empty. If I want to include those, I will have to draw them myself. I will also need a crane system for the machine shop. The parts are heavy.

I'm going to do the same thing on the HVAC units. 

My special Kalmbach book on the 4014 came and was great. It really showed the beautiful work that was done on that remarkable engine. The drivers were turned at Strassburg RR's Steam Shop. Funny that this little railroad with only small 1900s small-drivered steam engines in their stable, has the only capability in the country to turn monster drivers. They chose 4014 out of the eight extant Big Boys since it was in the best condition and the group that owned it was amenable to make a trade for it. Being in Southern Cal for the past 60 years kept it from rusting away. The pressure vessel was in very good shape, and required little in the way of rebuilding to make it functional, mostly around changing out staybolts.

I will fully document the learning curve on the 3D project so the rest of you can decide on whether it's for you.

Myles,

Thought you might get a kick out of this. An REA structure on the Carbondale. Hardware not installed yet. I am building a large Locomotive repair shop this fall along the lines of what you're doing, but it actually will serve as a divider between the loco yard and curved mainline.

Myles, Isn't any engineering project fraught with mental lapses, mistakes, and other surprises?  Anything I have been involved with is.  The best, or most experienced of us is still only human!  I will look forward to seeing what you document about the 3D process.  It intrigues me, because my small parts skills are poor.  Also, the sky is the limit with what can be accomplished.

Ron, your engine house looks great!

I agree, nice engine house!

All sorts of stuff is happening. after correcting the latest screwup I tested the fixed right wall and it fits very nicely.

I was pleased to see that the walls were dead straight.

Here's another view showing all those buttresses.

I glued the four segments of the roof together using a model airplane technique of placing the parts together, tape them on one side, bend the joint open, apply glue, push it flat and tape the other side. This is how you glue wing skins together. 

I'm concerned that this will not be strong enough since the trusses run parallel to the roof and not across the joint. Again, I'm going to use another RC plane technique by reinforcing the joint using graphite tape. The tape is held with thin CA and makes the joint very strong. That being said, I came to another decision to glue all the trusses to the roof itself and not the buttresses. This way, all the trusses and light assemblies will come up with the roof giving me completely clear access to the entire interior. I am absolutely certain that there will be derailments in that building especially when I run into the end-of-track bumper instead of stopping.

I wanted to make the trusses a little stiffer in the lateral direction so I added some balsa strips on the lower longeron. With the trusses on the roof, I can also add some inter-truss X-bracing to prevent tipping. If the trusses were attached to the building, this kind of bracing would completely block interior access.

I finished the outside pilasters on the end wall making doubly sure that I had correctly accounted for any overlap.

Tomorrow, I'll add the inside pilasters. 

I ended up with literally dozens of extra pilasters, both long and short versions. I couldn't understand why I asked Stephen to cut so many. Then I went back and studied my SU drawing and realized that I originally wanted them to protrude a scale foot, so I envisioned doubling each 1/8" strip. As it worked out, I didn't do this and found the single depth conveys the bulk of the building successfully, so I now have a full extra set of both kinds of pilasters.

And then… tonight the Amazon delivery man gave me a present. I ordered the 3D printer yesterday and it arrived today. I have a huge list of things this amazing machine is going to do. Remember those hand made telephone pole transformers. Yup! Gonna print them. I've been wanting to scratch-build a diesel prime mover. I've done some drawings of both an in-line Baldwin, and an EMD 567, but really couldn't get my arms around how to scratch-built them. Yup! Gonna print it. These models can also act as masters for resin casting. So do I see a cottage industry in my future… hmmm. This is all in addition to the parts I'm going to produce for the engine house.

I started working on the first machine tool to print. While the 3D model looked terrific, it was a terrible SketchUp drawing. The developer violated all the SU rules. The drawn object must be a "solid" in 3D drawing terms or it won't print. This means all faces and edges must be connected. Otherwise, the printer wouldn't understand where the object began and where it ended. I have software extensions that can analyze the drawing to determine if it's a "solid" and if not, why not. This drawing was a mess. 

The biggest failure was the drawing was created as a single object. That is an SU 101 no-no. Instead, after you draw even the most simple rectangle, you  should immediately turn it into a group. In SU, any ungrouped object touching any other ungrouped object immediately "glues" to that object. When you attempt to move it, it drags whatever face or edges it's contacting creating some kind of Cubist image. It is impossible to edit such an object, and, as in this case, can't turn it into a sold, and thus it's unprintable. 

So I had to de-construct the lathe by highlighting some parts and copy/pasting them to another part of the page. I then broke that apart and re-drew most of the component parts from scratch ensuring that each component was solid in its own right, and then combining them back into a compound object.

The disembowled headstock on the left is what remains of the original drawing. I still have to remove all the controls on it and redraw them too so they will be solids. All the rest of what you see is now printable. I'm basically building a lathe. This took almost three hours. I suspect that a lot of the machines I downloaded will also be drawn improperly.

It only took five minutes to uncrate the printer and set it up to print. Tomorrow we'll see what this baby can do. They've included some test files and I have the drawings that created the appliances and the hoist mechanism too. I'll keep y'all posted.

Looks real good, Myles. Thanks for the glueing tip on on the roof. I will use that one of these days. I'm anxious to watch your progress on the 3D printing.

Looks great Myles. You are indeed a true modeler in scale.

I do have a question....how are you planning to paint the inside of the structure with it already assembled?

Thanks. I do not post a lot but I follow your posts everyday.

Donald

The enginehouse is really coming along!  That is fast delivery on the 3D printer!  I'll be interested in seeing how it does.

Thanks Mark and everyone! Yes… it's coming along. Today was a mixed bag of picking up the ancillary items I need for the printer and doing some engine house work. 

This morning the exterminator came for our quarterly visit, so I was downstairs with him just puttering around. I decided to see how the completed roof fit with the walls. Remember, I was worried that those butt joints would be weak. Two of the four pieces discombobulated before I even got it to the building from its own weight. Later today I added carbon fiber tape which helped a lot, but even that isn't really strong enough, so the last thing I did was add a piece of thin ply. Really all the joints need this form of strengthening. When the roof is on the building supported all over the place, the weakness won't matter, but it does when you try and life it.

I finished up both ends. So the interior is now ready for paint. So to answer Donald's question, I'm painting the interior when the walls are separate. It will be a two-tone affair with white up about 6 scale feet and the rest a light gray, both using Rust-O-Leum rattle can enamel. Weather tomorrow should be dry. I will have to mask up and down all those wall protrusions.

Tonight I just put all the trusses on the buttresses to see how it looks. The trusses will be glued to the roof.

I finished the lathe redraw, exported it to an STL file and then imported it into the Elegoo ChiTuBox Slicer software. I processed very fast, but I think the object has translated as being very small, I mean infinitesimally small! I say this because using a 0.02mm layer thickness, it only seemed to have 20 layers. In fact, when I imported to the program, the platen field appeared to be empty. I had to zoom in and in until this little white speck showed up and then with continued zooming the lathe appeared. I need to do some more research to find out what's going on. The piece was so small that it would be produced with $0.00 worth of resin.

The extras I bought today are aluminum roasting pans to surround the unit to prevent resin spills on the workbench, disposal Rubbermaid containers for the alcohol rinses, a bigger funnel, and a new thumb drive. Generally, you convey the files to the printer via the thumb drive so you're not tying up your computer. I bought 190 proof grain alcohol at the liquor store that will be the resin rinse solvent. You can either use Iso or Ethanol, but I couldn't find Iso in 95% concentration, but 190 grain alcohol is 95%. Beside, if I get bored, I can mix a little in some fresh squeezed orange juice and have a screwdriver.

Last thing I did was make a dust/light cover for the unit out of the remains of the Amazon shipping carton. If you can keep all light off the resin tank, you can leave un-cured resin in the tank for the next time.

With the platen size of 2" X 4" I can print multiple parts in one go. Unlike a filament machine when the more stuff on the project, the slower the printer, with this concept, the entire layer is exposed simultaneously. Therefore, the only variable affecting speed is the layer thickness and the height of the object (how many layers are being printed). It is much faster than a filament machine. If I'm not mistaken, this technique first appeared in the press by a chemist who developed a light curing resin where it was all exposed in a full layer. His chemistry was more complicated involving oxygenation, and the units listed for $30,000. He was leasing the machines. They were 10X faster than machines that produced objects by writing single lines  inscribing the object's parameters.

It's really the same technical leap that occurred when the dot matrix printer gave way to the laser printer, where the laser defined a single page in one go instead of writing a letter at a time. Even before that was the plotter where the pen had to literally draw every line in serial fashion.

I'm worried about the scale. The imported STL into the slicer seems to be tiny. The blue boundary is the platen and that tiny blue speck is the lathe. It's looks like a lathe for Z-gauge. I have to find out more about how all this works. Ironically, the printer will actually print that tiny thing.

The building is coming along nicely. I expect lining up the trusses to glue to the roof will be a tedious job. It's easier to glue them to the building but glueing them to the roof will be better in the long run. I am still fascinated wit the printer. There is always bugs to work out with new technology like that. It's complicated by the fact that there is no experienced person locally to consult. Is there a web page or phone line for technical support?

Thanks Pat… big progress is being made every day. As for the printer… let me discuss that first since it really is "earth shaking". I loaded my lathe file, but inadvertently loaded the teeny tiny one. Let me put it this way. It printed in a couple of minutes and was so small… "How small was it?" It was so small that it was smaller than an ant or piece of rice. It was a lathe which you could barely make out with an unaided eye. There was no way to successfully scrape it off the platen, it simply fell apart. It was so small that it was a for a machine shop in Hooville, and only a Hoo machinist could use it.

Then I loaded the correctly sized one which I seemed to have solved by changing the units the model was exported in from inches to millimeters. I even used the auto support system that puts all the legs on it that usually you have add singling in the software.

But it didn't print. The file loaded in the machine and the image showed up in the display window. The platen went down to zero level, sat there for a few moments and then rose saying, "Print complete." I had forgotten that the printer came with its own thumb drive that has a sample file on it. I loaded that tonight and it's now printing as it should. It shows elapsed time, time to completion and what layer it's on out of the total number. This print will take about 3 hours and I'll show you it. It's magic to watch. Each layer is taking about 10 seconds, then the Z axis rises a bit to break the connection with the build zone and goes back down to establish the new layer. Each layer is exposed entirely in one go.

My last part of the 3D printing "lab" arrived: the UV curing light. I'm going to make a box line with aluminum foil to get the UV distributed all over. I honestly never believed that a completely affordable desk-top Hi-res printer was in my stars, and now it's in the basement.

Today was a painting day on the engine house. First up was painting the trusses. I used a Tamiya Rattle Can Gray Primer followed by Coral Blue. It's a nice structural steel color. I did all the painting outside which was 90° and humidity wasn't so bad. Paint dried quickly.

I then masked all the gluing surfaces on the various wall sections. I don't want to glue painted surfaces since Titebond needs porosity to really develop strength. I then took all of this outside and used Rust-O-Leum Gray Primer followed by their Gloss White. I was just painting up the bottom 2 inches for a two-tone paint job. 

Tomorrow, when fully dry, I'll use a surface gauge to mark off 6 scale feet, and then mask the whole deal so I can paint the light gray wall remainder.

I glued together the machine shop since I wanted to paint its interior at the same time. I used my picture frame clamp and corner clamps to hold it square and used Titebond.

When this was set I glue in the floor since it wasn't very stable with just the walls. I also added 1/4" square bass wood as corner reinforcement. The floor was just a tad oversized about 1/64" which I removed with some hand sanding on my sandpaper-glued-to-a-surface-plate sanding device.

I then laid out the roof to accept the trusses. I used the sides to pinpoint where the trusses will go, and then draw lines, square to the edges across the entire roof. I used narrow Tamiya tape to mask these line areas so the trusses would glue to native surface.

I took the roof outside and painted it entirely Tamiya Light Gray Primer which is going to be the final color. When dry (quickly) I pulled the tape and glued the trusses. I used two angle blocks to hold the truss vertical, then a heavy weight to press it down until the glue set a bit, then moved on to the next one. I used some thin CA to tie done each end so I could take the weights off without anything shifting.

Here is the entire roof with all the trusses. I'm going to run some longitudinal members wide enough so the LED light patches can stick to them. The lights already have mounting tape on their backs. The roof is still not stiff over its length and it needs to be boxed in a bit.

To make the roof easy to lift off, I'm going to have to make the attachments strong enough so they can double as lifting devices.

Oh… and one more boneheaded thing. I had Stephen cut the strips for the between-the-rails space. There was only two things I could get wrong with this: width and quantity. I blew it in both directions. I forgot to add relief for the flange way so the strips are about 1/16" too wide. And I have three tracks, so I'm four short. Each piece is 24" and the length is 40". I'm not getting more cut, I'll do something by hand. DOH!

While I wasn't (yet) able to print my lathe, the machine comes with a sample file which did print, beautifully. In fact, it's magic! To see a fully formed object rising slowing out of a 1/2" deep bath of clear liquid is nothing short of science fiction. The object, two 2" high rooks, were so intricate as to be hard to describe in words, so here're pictures.

If you look down inside you see a spiral staircase and a twisted support that extends from the top platform to the base.

You simply cannot see the layer lines. There are 1,000 layers in this print. They were .04 mm each. The printer is able to go down to .01 which is ridiculous. The resin cures hard as ceramic and appears quite tough. I put it under the UV light to finish it up for about 10 minutes.

Parts cleaning is done in 190 proof grain alcohol in two baths. I also cleaned the platen with the same solvent.

I'm in communication with the Chinese to troubleshoot my design, by I think it's probably something quite simple since the machine works perfectly. There are many less things that go wrong than with the filament machines.

I've sent a request to RJ Corman Rail Services to have a visit to their engine servicing facility in Lexington. I needed to get some first hand information about what I should do for my interior detailing. I'm also hoping that I could get some direct measurements off of a EMD prime mover if there's one sitting outside of an engine. I have good pictures and cutaways, but would like some real-world measures to scale the drawings.

I woke this morning thinking about just how many significant technological innovations had to come together to create a magic machine that turns drawings into beautiful objects. Here's my list:

1. LED (in this case UV LEDs) - although based on much earlier research, first commercial availability was 1968
2. Servo Stepper Motors - Servo systems were well developed during WW II
3. Micro-computers - 1980s
4. Photopolymers - 1960s
5. High-speed USB connection - 1990s
6. Thumb Drives (solid state memory chips) - Late 80s, early 90s
7. Liquid Crystal Display - 1970s
8. High Resolution LCD - 1990s
9. SketchUp 3D Design Software for the masses - Early 1990s
10. Slicer software to drive 3D printers - 2000s
11. Stereolithography (first 3D printing tech) - Patented in 1986

That's all I can come up with from the top of my head. Feel free to add to the list.

They often say that "Any technology sufficiently sophisticated can be viewed as magic." After seeing this little thing work last night, it certainly seemed magical.

Frankly, the hardest thing to master is the 3D drawing and design. The machine is a snap. There are drawings available through outside sources. RCScalebuilder.com (a scale RC plane builder forum) has an entire library of free designs for scale planes. The scales are too big for us railroaders at 1:16, 1:9, 1:6 and 1:5.

That chess piece is amazing.

That machine is incredible, especially for the price! I'm very interested to see how the detail parts turn out, because the obvious layer lines has always been a big turn off for me.

That is incredible.  And at that low price, no less!  I agree, mastering the drawing will be the hard part!

This 3D thing has made my posting more complicated. It's really two significant projects intertwined. As I did yesterday, let me give you a 3D print update and then go into the build progress. 

With the success of the rooks, I was assured that the machine works. I found that the machine didn't recognize my file when it was in the thumb drive's root directory. After I moved the file to the finder containing the rook file, it read it, opened it and ran it. The print took over five hours. The Chinese man who was helping me solve the problem suggested setting it on a bias which helps give strength to fine details. Here's a WIP shot of the "thing" being created. Like I said, it's almost science fiction; a solid shaped object coming out an amorphous bed of liquid. 

The slicing software created a forrest of trussed support strands to hold the object steady while it's being created.

While it was printing I built a light box to cure the semi-finished pieces. I took a cardboard box, made a hole for the light, strapped it in with some Hot Glued cardboard strips and then spray-glued some foil all around. 

After five hours, it was done and hanging there.

The lathe, unfortunately, did not come out as well as it should. It appears that it stopped printing on the bias angle. For example: the bed didn't complete and the tailstock bench stand is missing entirely. Before I wrote another letter to China, I decided to the drawings. When I looked at the sliced drawing I saw that the bed didn't connect to the headstock. Then I checked the .STL file, and finally my SketchUp drawing. And of course, it was about 3 scale inches away. I hadn't extended it far enough. I fixed the SU drawing, exported it again, and sliced it. It's now down on the machine being made. I also put it on straight instead of angled. It's a steep learning curve.

You can see how the missing material goes off on the angle that it was sitting when printing. I suspect that the bed not attached to one end may have contributed to the errors. We'll see in a couple of hours. Even with the defect, you can see that detail. Even the chuck jaws and tailstock hand wheel are rendered.

The control panel is slick. It shows elapsed time and time to completion, plus a count-down register showing the layer total and number completed. If this next test fails, I'm going to have to do more research. Meanwhile, I'm working on a plan to print the entire gantry hoist, carriage/wheels and all as one single piece. That too should be interesting.

Engine House:

Using the surface gauge on the Corian work surface I scribed the white/gray line. I then used Frog tape to mask all the walls. 

Went outside with my safety goggles a full vapor filter mask and sprayed all the Rust-O-Leum light gray. It was a very hot day and the paint set up quickly.

I brought it all inside to dry and displayed them in their relative locations.

And now a mystery… for some goofy reason, the white portion of the foreground wall is almost a 1/4" higher than the other three. I, frankly, don't know how this could happen since I don't remember resetting the surface gauge during the entire scribing operation. Now I have a decision.. is this a big deal or since it's buried inside, leave well enough alone. Feedback is requested.

I was thinking that tomorrow I would erect the four walls and realized I couldn't do this until the floor was painted. In fact, I had come upstairs around 2:30, changed out of my shop clothes and was going to do some more drawing. I changed back and tackled the floor.

The tail floor joint is in a terrible place. All the leverage on that very long piece is focused right on that joint. I also wanted to scribe concrete expansion joints. I laid these out with a dividers, pencil and straightedge. I then used the same straightedge to control the path of the razor saw I used to cut the grooves.

I wanted to reinforce the floor just to carry it outside and taped some strips across the open end so they wouldn't swing around independently. For some reason the tape didn't hold at all, and sure enough, when attempting the carry it back after painting, the piece swung a bit and the joint separated. I regaled it with thick CA and added a ply doubler glued with Titebond which, while unsightly, will keep it together. It in the rearmost portion of the building and will have a end-of-track bumper on top of it, so no harm no foul.

A quick touch-up spray covered it up. I'll treat the expansion grooves and weather the floor. It's not a car showroom.

That's where I left it. Tomorrow, I'm going to put the four sides together and then on Monday, I'll spray paint the exterior. Meanwhile, I better figure out the 3D printing because I have ton of things to print for this project.

As to the 1/4" difference on the one wall versus the others, I don't think it will be that noticeable when the building is assembled.  I can see the difference with the parts sitting flat on the table, but I would thing you will have to point it out to someone looking at the finished project before they would notice.

I can see that getting the drawing exactly right will take some practice.  I think it is a pretty good run for such a complicated part as a lathe for your first try.  

You're very understanding, Mark. This is my second post this evening since I did get a pretty successful print of the lathe. I corrected the drawing and repositioned the lathe on the build plate, making it just horizontal. 

My first look was good since I saw both stands printed and a complete bed.

After removing it from the platen and washing it in two different baths of ethanol, I clipped all those supports and got a good look at it. It's remarkable! The ways are perfect V shaped. The drive screws are actually separate from the rest of the model. I actually think that the T-slots in the cross slide are really T-shaped, but there's no way to tell with the naked eye. You can't see it, but all the knobs and levers are there. The tool post actually has multiple layers with a space between them for the tool holders. The detail is so intense that it makes me want to spend the time to really make the drawings as good as I can.

If I had made the cross slide separate from the bed, I believe it would have actually slid. The model had two defects, and neither will keep it from getting painted and used in the machine shop. The rear of the headstock stand is not fully formed, but doesn't keep it from setting flat. I could easily fill it with Bondic which is also a UV cured polymer of high viscosity. When primed and painted, the detail will really pop.

And the second is the missing lower portion of the chuck. Again, not a show stopper. The bottom of the bed also is a bit misshapen.

As a proof of concept the Elegoo Mars UV LCD Mask printer passes with flying colors.  There's a ton of variables that can be controlled in the slicing program including not only how many supports and locating them, but their thickness, the shape of the connection to the model, the depth they penetrate into the model, etc. There are other variables regarding how long the exposures are at the bottom levels and all succeeding levels. Little by little I'll figure all this other stuff out.

Meanwhile, I just created a beautiful O'scale engine lathe that when painted will be believable and unattainable in any other way. I'm not seriously working to print the entire gantry hoist. While, it's not as "crafty" to make a single piece that would replace hours and hours of scratch-building, it makes it possible. Some of these ideas aren't doable without this technology.

And I had another thought. My brother-in-law's nephew is quite an entrepreneur. He started a company that processes bank statements for banks relieving them of this onerous clerical task. He started by buying a raft of Kyosera desk top laser printers and had them all running in parallel thereby producing what would have been output by a commercial, large volume printer that cost many times as much. It got the business going and eventually they did by a major production machine that did more than just print.

With a $360 3D hi-res, you could buy 10 of them equalling the cost of one machine just several years ago. The speed is ruled by physics. It takes so many seconds to expose and harden the resin. With 10 machines running in parallel you could actually have a business that could produce enough output to earn a living. When I suggested this to my wife this morning, she told me to not think about it.

 

Trainman2001 posted:

 

With a $360 3D hi-res, you could buy 10 of them equalling the cost of one machine just several years ago. The speed is ruled by physics. It takes so many seconds to expose and harden the resin. With 10 machines running in parallel you could actually have a business that could produce enough output to earn a living. When I suggested this to my wife this morning, she told me to not think about it.

You really nailed it, Miles. Congrats.  I can't imagine what you will be doing in a few months!

The "10 parallel printers" idea is interesting.  As I understand it, the big production print services (such as Shapeways) take a different approach. They use printers with very large platforms and have optimization programs that fit as many jobs as possible onto a single run. Space on the platform is the precious commodity, due to setup costs. They even print small items inside of large hollow ones!

Anyway, your pioneering this technique and sharing your experiences so generously are just awesome.

First, the paint mismatch. No one but you will notice. If your the type that will be bothered every time you look at it then fix it. If not, like the Beatles said, "Let it be".

The lathe looks fantastic.

Pat Kn posted:

First, the paint mismatch. No one but you will notice. If your the type that will be bothered every time you look at it then fix it. If not, like the Beatles said, "Let it be".

I have to say the other side of this. If it bothers you now, it always will. You should fix it.

I threw my lighting up carelessly on my building. I said who's going to bend down low enough to look up there? Besides me knowing how bad it looks, my first guest bent over to look in the building and I know he saw the sloppy work. The grandkids have the exact view of the roof because of their height. So I know they also see it. There are heavy Christmas light wires wrapped and hanging all over my roof structure. I should have spent the extra time to do it right!

Beyond that, all I can say is the work here is fantastic. I was too afraid to go down the wrong path with the new printing technology still evolving. So I did nothing. When I was younger, I normally would strive ahead and fix wrong paths faster than others would even get started. So now I'm that other guy!