Continuing Saga …

The second lathe looks great!  I'm sure much more detail will show as you say!  My wife would say, 'Don't think about it' also!

I took advice and redid the paint job. I wouldn't have except I found that both long walls had the problem and I found exactly when it happened. My scribe was not tight enough in its holder and was moving. About a third of the way along on the first long wall it began moving and within a couple of bays was almost a 1/4" higher and I didn't notice it. I rescribed the walls ensuring that the scribe was tight this time. 

Now the corners match perfectly and I can sleep better at night. A friend of mine just nicknamed me, "Mr Never Good Enough".

I fixed the headstock on the lathe model where some material was missing. Bondic is exactly the same chemical composition as the 3D resin and cures at the same 405nm UV wavelength. I built a little masking tape dam around the perimeter and just filled it to the edge with Bondic, letting it extend past the bottom edge. I cured it first with the Bondic LED and then stuck it in my UV chamber to harden it all the way. A few strokes on my sand paper glued to a surface plate and the bottom was whole and the repair was entirely invisible.

The segmentation on the chuck is due to having the circle segment setting in SketchUp at 24. You can set it to any number, but the more segments the more CPU time and larger the files. I'm going to up it since the printer sees everything! Notice that you can see the hexagonal nut on the tailstock lock. You can also see space in the rotary tool post for a tool holder. If I would have put threads on the lead screw you would have seen them too.

With my increasing confidence, I'm "building" an EMD 567 in SketchUp that I'll 3D print. I'm going to print this as a model with multiple parts so if one print doesn't go so well, I don't have to spend the time reprinting the entire model. So it will have separate block, valve cover, heads, front accessory drives, exhaust manifolds, roots blowers, etc.

It's going to be one heck of a model. Walthers does have an HO plastic version of this baby, but no such luck for us O'scalers. I'm drawing it all from scratch based on a good cross-section drawing and photos. I don't know how long it is. I had to increase the length since the power heads didn't fit into the valve box. I've yet to put in the cam, rocker arms and fuel pipes. Don't what the printer will do with the flue pipes. That may be some thing that can be installed afterwards. Took quite a while to get the bridge rocker right as well as those access covers. Regardless, it's certainly easier drawing it, than trying to make it. I checked the block with the slicer and, at least, that will print. I'm 

I know I sound like a broken record… this thing changes everything.

That's really cool what you're doing with the printer!

And the engine house is coming along nicely too!

I drop in now and again and have a look great stuff I like what you are doing with the printer and the small machines.

Years ago when I was careless with money I bought every small machine kit three times over that I could find made out of metal I still have them all, never built, lot more work putting them together your method is better I guess machines are taking over I'm a dinosaur. The Engine House is coming along fine looks good thank you for posting. Roo.

I knew you would repaint the inside walls.  I’m glad you did. I lost some of my OCD tendencies over the years with kids growing up.  Now, I still get so little time that I would probably follow my advice thinking if I redid everything I would never get to run trains for several more years! 

The lathe looks great painted’

Reprinting the wall is for the best. For a little effort, you'll get peace of mind. I continue to be fastenated by the work you are doing with the printer.

Myles,

We have been at the beach for the last couple of weeks and I have had little success in getting downloads of the forum

I am now catching up

This is a great thread and your experiments are truly informative and really exciting for the builders in our hobby

Keep  it up and thank you

Jerry

Thanks for following and the recognition. I spent most of the day today drawing more details into the 567. I've got the basic block in good shape for printing and I've finished fully detailing 1/2 of the head area of one cylinder back. The details are comprehensive, but not 100% accurate. All of the stuff is there: Injectors, injector rocker, four exhaust valves per power pack, Valve bridge rocker and rockers, rocker shafts and supports, cam shaft and supports, injector rack, head to sleeve bolts and power pack clamps and bolts. I then shrunk it down .021% to make it O'scale. SketchUp doesn't like very detailed drawings in small sizes and it crashed a number of times. I finally was able to export it an .STL, and then into the slicer and it did slice. I put it onto the thumb drive and was going to take it down to the shop and print it, but my wife suggested maybe not such a good idea since we're experiencing thunder storms and the power skipped a while ago. I can wait 

I redrew the valve covers and put a lip on the open one to give it some gluing surface. I'm going to print the open one as a separate object. The heads will also be a separate print. I mentioned this before that I'm going to treat this almost as a model with separate pieces helping to reduce the print failure potential. As I gain more confidence in the process, I'll print more comprehensive models. The printed details will be a lot crisper than this PNG screen print. All that's functionally missing are the fuel lines. 

I hollowed out the block a lot so it will use less resin and drain well. With resin printing you need a way to get rid of uncured resin. I opened four cylinders to key into the power pack heads of the four that I've detailed.

Next I'm working on the front and read accessory drives. The front one is is just a serious of cylindrical objects and connecting stuff, but the rear has the two Roots blowers and I'm going to have learn how to draw the waffled surface detail. I put a question on the SU Forum and see if they can get me started. The other complication is the soft curves in the air transfer manifolds. SketchUp is not particularly happy with soft curves. You can do it, but it's not one of my strong suits.

Trainman2001 posted:

I redrew the valve covers and put a lip on the open one to give it some gluing surface. I'm going to print the open one as a separate object. The heads will also be a separate print. I mentioned this before that I'm going to treat this almost as a model with separate pieces helping to reduce the print failure potential. 

 

So, this raises the question of the feasibility of printing functioning components. With the resolution of these devices, one can start to think of printing working gear chains, crankshafts, rods and cylinders that move, etc. I wonder how small a nut and bolt you could print that would screw together?

Some day, I guess we will be printing working locomotive drive gear.

Printing the engine is a great idea. Not only is it an interesting detail for the shop but by printing multiple copies you can have interesting flatcar loads. Picture an engine in the shop with some of the pieces not connected and a flatcar on the siding outside the shop with a new engine awaiting instillation.

Trainman2001 posted:

I took advice and redid the paint job. I wouldn't have except I found that both long walls had the problem and I found exactly when it happened. My scribe was not tight enough in its holder and was moving. About a third of the way along on the first long wall it began moving and within a couple of bays was almost a 1/4" higher and I didn't notice it. I rescribed the walls ensuring that the scribe was tight this time. 

Now the corners match perfectly and I can sleep better at night. A friend of mine just nicknamed me, "Mr Never Good Enough".

I fixed the headstock on the lathe model where some material was missing. Bondic is exactly the same chemical composition as the 3D resin and cures at the same 405nm UV wavelength. I built a little masking tape dam around the perimeter and just filled it to the edge with Bondic, letting it extend past the bottom edge. I cured it first with the Bondic LED and then stuck it in my UV chamber to harden it all the way. A few strokes on my sand paper glued to a surface plate and the bottom was whole and the repair was entirely invisible.

The segmentation on the chuck is due to having the circle segment setting in SketchUp at 24. You can set it to any number, but the more segments the more CPU time and larger the files. I'm going to up it since the printer sees everything! Notice that you can see the hexagonal nut on the tailstock lock. You can also see space in the rotary tool post for a tool holder. If I would have put threads on the lead screw you would have seen them too.

With my increasing confidence, I'm "building" an EMD 567 in SketchUp that I'll 3D print. I'm going to print this as a model with multiple parts so if one print doesn't go so well, I don't have to spend the time reprinting the entire model. So it will have separate block, valve cover, heads, front accessory drives, exhaust manifolds, roots blowers, etc.

It's going to be one heck of a model. Walthers does have an HO plastic version of this baby, but no such luck for us O'scalers. I'm drawing it all from scratch based on a good cross-section drawing and photos. I don't know how long it is. I had to increase the length since the power heads didn't fit into the valve box. I've yet to put in the cam, rocker arms and fuel pipes. Don't what the printer will do with the flue pipes. That may be some thing that can be installed afterwards. Took quite a while to get the bridge rocker right as well as those access covers. Regardless, it's certainly easier drawing it, than trying to make it. I checked the block with the slicer and, at least, that will print. I'm 

I know I sound like a broken record… this thing changes everything.

Will you be making the engine available to others?

Let me take them one at a time:

Functional parts are a possibility. The UV curing resin is pretty tough, but I'd have to do more research. I will also have to see just how close the tolerances work out for parts that have to work together.

Multiple engines is a distinct possibility. Once the drawings work, you can make a million of them.

Making available? There's a problem with that. I use SketchUp Make, which is the free version. If you gain any commercial value at all from the software you've violated the agreement and have to buy the Pro version. Pro version is $800.00. I've been reluctant, for obvious reasons, to do anything like that. That being said, If I could break even to cover the Pro version, and if the engine is successful, they could be available. Those are two big ifs I realize. 

I could also make the .STL files available for folks that want to buy this kind of machine. The .STL files will not print, but are the precursors to any slicing software. Details on this model are too fine for a filament machine.

Let's first see how one engine comes out before I attempt to make more.

I'm test printing the heads with all the valve mechanisms. In four hours or so, I'll be able to tell if resolves as it should. Meanwhile, I have to take an entire load of LED replacement tubes back to Home Depot. I bought Phillips tubes thinking they were the same as I used in our master walk-in closet. Only one fixture lit, one lit momentarily and then turned off. Another didn't turn on at all. I assumed that fixtures made by the same manufacturer had similar ballasts. I'm also going to try and get the engine house walls glued together. We're heading back East tomorrow for a week away.

Myles,

You are really turning these part in record time. Does printing more than one of the same part increase print time?

Gerry

...and then there is the potential to print negative molds for bulk casting. I know that there is a flexible resin available--don't know how flexible, though.

At this level of detail, you can even keep the rivet-counters happy. 

SGMret posted:

Myles,

You are really turning these part in record time. Does printing more than one of the same part increase print time?

Gerry

Anything that fits alongside each other on the build plate takes the same amount of time if they are the same height.

That's correct! What controls print time in a resin machine is layer thickness, exposure time per layer and the height of the model. Since each layer is displayed and exposed in one shot, no matter how many things get illuminated makes no difference. I really didn't get much done today. I was thwarted by taking back a box of Phillips LED replacement tubes to THD. I mentioned that they weren't working except for two tubes in one fixture, so I wanted to return the box with 8 tubes. I was willing to pay the $16 for the two that I was keeping, but they said that since it was purchased on line, they had to accept the entire box and had no control over the transaction. They're apparently two separate companies. So I was forced to go back home, remove the remaining LEDs and re-box them, and then bring the whole deal back to the store. I got the full refund on my PayPal.

So I didn't glue any engine house walls together, nor will I until we return in two weeks.

I did make a test print of the 567 head. It was a total failure. I believe the cause was a damage teflon sheet at the reservoir's bottom. A little bit of material was stuck to it and I had seen a video on how to clean it off. I used the plastic scraper and took off the little bit, but then I made a mistake. I cleaned the sheet with alcohol and dried it off with a paper towel. The towel scratched the film and the fine scratches made the resin stick to it and pull off all of the supports. When it was done, all I had hanging on the platen was the supports. Anything resembling a model was stuck firmly to the teflon. It looked like something that was caught in a Star Trek Transportor that didn't work quite right.

The machine came with an additional sheet. The film holder is a two-part affair. You first remove a number of screws that attach the frame to the reservoir bottom. Then you turn it over and remove 24 smaller screws that hold the two halves together. I watched how this is done in a vid and you're supposed to leave slack in the film since it is ultimately pulled drum tight when the frame is tighten over an elevated lip on the reservoir rim. To keep the slack I put a couple of thicknesses of heavy cardboard when lighting the 24 frame screws.

With new film in place, I'm making another print of the same part with the same settings just to eliminate the slicing scheme as a problem. I know now to not use anything harsh to wipe the film. After cleaning with alcohol, I'm just going to blow it dry with an airbrush with no paint. 

You can pause the print at anytime and the platen goes up to the home position so you examine progress. I did this and saw that the cylinder head was actually sticking to the supports and not the teflon. I'm taking that as good news and will see the finished product at around 9:00p.m. tonight.

Meanwhile, I got some support on how to create the texture on the Roots blowers, and have started defining the front end detailing including piping, governor, low-pressure fuel pump, and water pumps. The tail end is basically the blowers with their ducting into the airbox and the big flywheel.

I'm going to buy more film since this is clearly a maintenance item with this kind of machine. On filament machines it's the hot-end that gives you the most grief. Everything has its Achilles Heel.

Amazing what it can do! Now if you were to put these into production for resale, you would need to have 45 prepaid orders at least $20  each just to cover the software fee and materials

The second print failed with the new film. I changed the orientation (leveled it out) and increased the diameter and contact point of the supports, and will test it later tonight. Meanwhile, I downloaded a resin 3D test part that shows any calibration changes that may be needed. It's only a 1.5 hour print and I should get the engine head on the machine in time to run both. I'd really like to have a good print before heading out tomorrow so I don't have to think about it. I did a "pause check" of the current piece and it is adhering to the platen. That's a good sign.

For a complete O'scale 567 I would have to charge at least $50. It doesn't use much resin, but it's taking a ton of development time and amortizing the machine. At this moment, I really just want to get one good one for myself. While it's fun to think about, I really don't want to develop a cottage industry. I don't want the responsibility. If I give the drawings away for free, I don't have to worry about the SketchUp license.

The test article came out perfectly. So the new film is okay.

Once again, the translucent resin makes it difficult to see just what's going on with this tiny model. They are slots and holes and tiny probes. There a field of needless in increasing length and thinning diameters. They were all resolved correctly. The machine is capable of doing some very fine work. With this assurance, I've started printing the 567 head, only this time, I'm orienting it horizontally, with stouter supports. If that doesn't work, I'm going to attach directly to the plate with no supports. It's a long flat piece mainly in 2 dimensions without much of a z axis.

Wish me luck...

The head printed half good. The valve area printed decently, but the other end… the plain end, separated from the support web and didn't form properly. I'm going to keep experimenting with the orientation and maybe increase exposure timing just a tad. 

The good: All the bolts are recognizable. The camshaft is completely separate from the deck below. Without a magnifying glass you really can't see some of it. 

The bad: besides the right side deformation, the fuel rack levers are deformed and missing on the 3rd cylinder. The rack levers are missing and probably weren't on the drawing. I'll check it out. Head clamp bolts need to be taller to stand out more. I'll figure it all out.

Back from Philly on Thursday, finally got a bit of work in the shop on Sunday, but was also working with the 3D machine. I'm about halfway up the learning curve. The biggest challenge is learning how to create the best support system. The software will do it for you, but there's a ton of variables that you can adjust. Any failures I had so far are due to the supports letting go. When the supports let go, the suction force of the previous cured layer on the Teflon surface exceeds the strength of the support and the cured piece falls off or bends and deforms. When that happens any subsequent layers in that part of the job are ruined.

But I did get some interesting things done, the best of which is the first iteration of the 567's cylinder block.

This end view shows the hollow construction. I used the "Hollow" feature of the slicing software. To help drain the entrapped un-cured resin, I used the "Dig Holes" feature to open up strategically placed drain holes. I want to draw you attention to how perfectly formed the bottom T-shaped members are formed. The ability of the machine to recreate tiny details is amazing. 

Due to the translucency, you can really see the details of the access hatches, but they're there. I have some cleanup to do since the slicing software added supports leading under the overhangs of the access hatches. The resin is hard and I'll use diamond burrs to carefully remove any nubs.

'

This picture shows a problem I'm having. One of the things I learned about the resin is that it shrinks upon curing. It's very similar to casting metal. The shrinkage puts stresses on the support system and caused the extremis of the top surfaces (which were on the bottom) to not fully form. I had the same problem on the lathe. Bondic will successfully resurface it, but not without more work. I'll experimenting with different approaches to solve this including using the "Heavy" setting on the support system. The block is bare on the ends since these are going the extra pieces which I'll be adding. I'm drawing them now.

This view shows the access ports better including the details on the hand clamps in their centers. You can also make out the remnants of the supports at the bottom of the ports.

I

I also printer the two heads with their valve covers. They too had the same support problem at their ends. I even printed a very thing open lid that I'll CA to the head after assembly.

While all this was being done, I did some weathering and accenting on the engine house floor and started putting it together. After applying the low-viscosity, mineral spirit based stain, you let it dry a bit and go back with a Q-tip lightly dampened with low-odor mineral spirits and remove the excess leaving the engraved expansion joints filled.

I then went back with some grimy gray pastel weathering powder and finally added oil spills with some Tamiya Clear Smoke. It dries glossy and looks very oily.

I glued up alternate corners first and let them dry overnight. Today, I glued the full building together.

Without a roof of floor to keep the sides straight, I made a couple of temporary spreaders to keep the long (and very flexible) sides parallel while gluing it all together. 

Being impatient, I let the monster dry for about a half hour I de-clamped, took it to the railroad and placed it on the floor, and then tried to fit the roof. It's about 1/16" to long, put too much stress on the corners and one popped open. When I examined it, I saw that the Titebond was not dry in the joint. I re-clamps and added some CA. I think I may add some small screws into the corner blocks to give it more security. It's a big, unwieldy structure.

I designed and started printing the AC systems for the engine house roof. The first print was failing early, so I aborted it and redesigned how the object was oriented. It's now downstairs working away and I'll get it tomorrow morning. Unlike the string printers, these resin units can run unattended without fear of filament clogs. 

The building is coming along nicely, Myles. Your quickly becoming an expert on the printer. This is going to increase the amount of detail your models will have. Great tool.

Myles, I was just thinking last evening wondering when you were getting back, and then you posted after I had signed off!  

Half way up the learning curve already!  You are doing great!  Yes those T shaped supports look perfect!!  The weathering work on the enginehouse floor looks very nice indeed!

Thanks guys!

Today I was wrestling with the 3D Printer. I'm breaking the AC project into smaller pieces to improve printing success. Here's the entire unit. Remember, the main driver of print time is the number of layers = height so shorter pieces print faster.

I did get a perfect print of the intake section of the roof AC units that I'm going to print for the Engine House. All the surfaces were flat and edges perfectly square. 

I was having trouble with the touch panel with the touch points being significantly out of registration (almost 1/4" low). It was so low that icons on the bottom row were not responding at all. I contacted the company in China via eMail and they sent me a link to a downloadable file that could be used to calibrate the touch pad. They said if it didn't work they would send me a new pad.

After fussing with it, I determined that the touch pad was, indeed, defective and asked them to execute Plan B. It will take between 5 and 12 days to get a new one. I don't know why this one failed.

Meanwhile, I can print because the print and file controls are accessible although out of position. I then attempted printing the main AC unit. Yesterday I tried with the legs down and it failed. Today I re-sliced it so the fan was facing down and the legs up. This made a huge flat surface that had to be held by all the supports. They didn't! The fan printed, but the flat surface disengaged and was stuck on the bottom teflon window. I had to scrape it off and it damaged the surface enough so I have to replace it. My supply of bulk FEP film arrived just in time from McMaster-Carr.

I'm going to redesign the AC so the fan will print sepeartely, and then I'll print the body on a bias. I can see why angling is better since it narrows the cross-section of material that has to be constantly pulled away from the teflon. There's a lot of variables at work here.

Meanwhile, the four walls of the engine house are fully glued. I then permanently added one 3/4" square cross brace to keep the walls aligned properly. I then also added some corner braces since the size of the building made it still a little too flexible for my liking.

I removed about 3/32" from the length of the roof so it now is a nice slip-fit into the building. I put the whole deal onto the floor plate and found to my dismay that the floor is too short by too much to fudge. I'm going to have to fix this since there's nothing anchoring the thin separaters between the doors. I'm also thinking that the floor actually doesn't need to be glued to the walls. The building is heavy enough to hold itself in place without gluing it. I'm also concerned about the flexibility of the middle floor pieces. They're 40" long, only a few inches wide and how now cross bracing in the middle since the tracks interfere with doing so. But… if I don't glue the floor into the building there will be no way to secure those front verticals.

I also added some roof stops to the machine shop roof so it's removable and I can inside easily to add the machines as I print them.

I traced the inner machine shop walls to the roof, glued the pieces in place, and then put the building back on it again and carefully slid the stops up against the walls so they were properly placed.

I then removed the building and weighed down the pieces so they dried. 

I did get a successful print of the first smoke jack. I made it so the conical top was actually connected and open to the pipe. In the act of shaking it to get rid of the excess alcohol, I broke off the top. It seems that even though the printer can print anything, it still responds to the laws of physics. My supports connecting the cone to the pipe were just too tiny and not attached to enough meat on the tube. I redrew and re-sliced it and will print it again.

Good progress,Myles. Too bad about the panel. Your imbedded pictures didn't come out but the pics on the bottom of your post are viewable.

That’s a neat way to clamp the corners.  I hope to remember it.  I forget too quickly any more.  LOL. I see what you mean about making the 3D parts in sections.  Great tip!

The panel calibration I did helped. The buttons are much less out of registration. The Chinese said instructed me to disassemble the panel from the unit and reconnect it. If that doesn't work, they'll send a new one, but it will probably take about two weeks. Meanwhile it still prints well as long as your fingers touch the right spots.

Mark, you should take notes… Maybe I'll give you an assignment. Or you could periodically re-read the whole thread. I do occasionally just to remember all the stuff that I've done.

I'm getting better at the printing with the refinement of using the supports correctly. It's all about the combination of angle that the piece is situated and the type of the supports. I re-printed the engine block with total success. I put it at a 45 degree angle and used the "heavy" support setting. It didn't budge and all the surfaces were square and flat. I did some post-printing damage when I was a little rough in cutting off the thicker supports and took out a chunk on both sides of that thin lower rail. If I put this engine outside the building as a scrapper, then a rusted and damaged part will look just fine. Here it is in the finished position. This version was completely open in the middle it worked very well.

This shows the labyrinth of supports. Each one of those sticks is interconnected with all the other with angular webbing. The thing is quite strong. Going with the thicker supports eliminated separation and no warpage.

Lastly, here's the engine block with the supports gone. You can see that damage to that lower rail. There's similar damage on the other side. I'll be more careful next time. With each print I'm learning more and more about what you need to consider. So, you can't actually print anything. The last thing I attempted was just the fan assembly for the air conditioner. I was concerned about those cross guard bars being unsupported. After about an hour of printing I checked it  and, sure enough, the print was failing since those bars were just hanging there, being unsupported. 

I will print it without the bars and do them as a separate scratch-build activity with some brass wire through some drilled holes. Just because you can print something doesn't mean you have to print everything.

While the printer was humming away I did work on the engine house. First of all, when I have a 50/50 chance to get something wrong, the odds are 100% that I will. This time it was the roof stops I glued in yesterday. I forgot to check which side of the roof I was gluing them to. The roof was not symmetrical. I glued them to the top side. I had to rip them off with the plane iron, then use some spackle to fill the gouges made in one part. I re-measured and glued new strips to the correct side and tomorrow I'll sand the filler and it will be done.

I did decide to glue in the floor. I had to add some pieces to fill the gaps and then make some angle gussets to tie the floor to the flimsy front pieces. The last time I checked it, the floor structure is sound except for the terrible flexing in the middle of the mid pieces. These will be correctly supported when the building is on the layout since they nestle down next to the tracks. I may use some silicone adhesive in strategic locations to keep it nice and flat.

Here's a close up of the gussets holding the front wall to the floor. It really stiffened the front wall.

I ran into a slight problem from my corner braces. They interfered with the trusses. I trimmed the end trusses so they slipped over the corner brace. Can't see it, and now the roof fits perfectly and is removable.

To help the mid piece gussets dry properly I artificially raised those pieces so they didn't torque the glue joint. With the glued floor, the walls no longer flex. It's heavy and I have to be careful carrying it from the layout area to the shop.

I bought my own roll of Glad Press-n-Seal which I'm going to use to mask the windows from the inside. I can then spray the outsides without wrecking the interior paint. I chose to buy my own roll instead of using the kitchen roll for obvious reasons. It's going to be very hot and humid for the next couple of days and since I have to paint outside, it may delay construction. I have a few more windows to construct, and I'm slowly going to print the stacks, AC units and machines for the shop. At the speed the printer works, I have to print parts when I'm not in the shop. Once I'm sure about how to do it without failures, I can start a print in the eventing and get the job in the morning. Doing it this way, you can do three to four prints a day depending on the time to print.

I'm beginning to understand why you orient the 3D part at 45 degrees. In addition to adding strength to thin parts, it minimizes the attachment stresses on the teflon bottom sheet. My AC print yesterday failed for two reasons, I didn't use strong enough supports and had it horizontal. When the almost 4" long and 2" wide flat surface attempted to pull off the bottom, the supports didn't have a chance. When I was asking Walt Gillespie at Rusty Stumps to print the appliances he was giving me some resistance saying that he had to redraw them and they all needed supports. I didn't know what he was talking about. I really didn't understand how the process worked. I didn't realize that the forming part had to be consistently tied to the platen so it could be raised enough to build the next layer. I now know.

Very cool. This is like watching a double feature. The design and assembly of an engine shed and the trials and tribulations of using a 3D printer. Thanks for sharing.

Been far too long since I got my nose in here. Everything looks amazing even though it is still in the ruff.

Myles, I take notes, then I forget where they are or if I ever took them.  I do know where your topic is, so I can remember that and can look back here!  It is good to reread, and I'll bet you find it interesting.  I could see myself actually wondering later how did I do such and such!  

More progress looks great!!

Well then keep reading and I'll keep writing.

It's the 50th Anniversary of Apollo 11. We were married just a year when it happened. I took a picture of the TV with Armstrong stepping off the LEM. Today, I read in Aviation Week (of which I am a 45 year subscriber) that Lockheed Martin is on their 5th iteration of a compact fusion reactor being design by the Skunk Works. They were getting tired of the delays and lack of progress in the international efforts such as ITER. It's way over budget (billions) and way behind schedule (years). Lockheed thought there might be another approach that doesn't require building something as a large as a big building. The initiative started in 2014, and while it's going a little slower than they would have liked, they're optimistic that a self-sustaining power reactor is within reach. They're looking to us it for aircraft propulsion and small utility use. American ingenuity!

I didn't post last night, so this is yesterday's post. I'll journal today's work later. 

Yesterday was one of the those potpourri days, with a little of this and a little of that. I fixed the little corner bits on the main building and then used joint compound to fill all the gaps. I masked and painted the machine shop floor and added a wood strip around the bottom of the roof edge to give it more thickness and a place to attach the rain gutters. I finished all the windows (61) and they're ready for installation as soon as the exterior is painted. I added some rivet decals to the splice plates on the main gantry rail. I cleaned up the shop! And I 3D printed some more parts.

Didn't take much filler to close the tiny gaps. It will just make the outside look more put together.

The floor glued in tightly and really stiffened the walls. Here's a view looking down the length. I'll be brush painting those remaining wood structural members.

To mask the machine shop walls I tried using a combination of masking tape and Glad Press-n-Seal (PNS). The PNS (don't read that acronym phonetically or editor is going to zap my post) didn't work as well as I thought it should. The  spray can's pressure was lifting it a bit. I was going to use it to seal all the main building window openings, but now I'm not so sure. It's cheaper than masking tape. It's really hot out today, but it's not raining so I may attempt to spray the building.

I have some MicroMark (Archer Fine Transfers) rivet decals left over from the truss bridge project of many years ago and used them to dress up the splice plates on the long main gantry rails. I shot them with DullCoat after applying to seal them. I'm painting the rails assemblies red primer color. I didn't have many decals left so I just did the exposed side.

I had one more medium-sized and seven large-sized windows left to assemble. All of them (61) are now complete. The window flanges are adhesive backed and having a glossy surface on which to stick them will be helpful.

As for the printing. Here's the primed parts. The new heads came out pretty well. I'm still having some trouble with bits of the details not appearing, and I awoke this morning and figured out why. I did a mental experiment and layered the part as the machine would in my head and discovered that when there are small overhangs that are facing in the direction of the build, the exposure is trying to form them in free space. In 3D printing that doesn't work. So the tiny layer forms, but not being attached to the main body or a support, it just floats away. The next layer forms and the same thing happens. This goes on until the overhanging bit is actually attached to the part and it starts forming properly. It's why the corner of the access hatches isn't there. I drew it as it is on the real thing, with an extended lip. That extended lip started to form in free space. Lesson No. 1: watch out for unsupported overhangs and fix them in the drawings. I'll go back and revise the drawing to get rid of the overhang and the problem will be resolved. There's also a similar situation in the head detail where the clamp bolts on one side aren't forming. I believe the same mechanism is at work. Otherwise, this engine looks pretty darn good! 

I made another fan, and this time, the fan blades, while being just as drawn, are simply too thin after printing to maintain their structural integrity. Lesson no. 2: just because the printer will resolve it, doesn't mean it will work in the real world. The resin is only so strong and extremely thin cross-sections will break. I'm also going to thicken that lower flange on the engine mount so it won't break away when I remove the supports. I believe I'll have all this 3D printing stuff figured out in a few more weeks...

The engine looks amazing. You already have more detail than can be discerned from outside the shop looking in. Great work Myles.

I agree, but since I'm on the learning curve, I'm using this as my test article. If I can get these little details right, it will make going forward must more predictable.

I am a little surprised that the slicer doesn't detect models that are unprintable due to floating features. Guess there is still a lot of room for growth in this amazing new technology. Are there separate validation programs available?

Thanks for taking the time to provide these incredibly instructive reports. You are tempting me to jump in.

Both projects look great!!  I'll keep reading!!  

one lesson i learned is dont try to build everything in one build. Break out the small parts  into small parts with attachment posts and holes.

also easier to paint.

also the 3D blender program should allow you preview each slice or groups of slices so you can see floaters before you commit the print

Good advice and I'm learning quickly. The slicer does show a pretty detailed rendering of the layers being added (or removed depending on which direction the animation is going). I also am paying more time to looking at the STL file first since it shows non-solid objects pretty clearly once I learned what to look for. I'm making the AC units in four parts, but just for fun, I'm going to try and print the entire gantry hoist machinery in one go. I had to redraw a lot of the downloaded components since they were not solids.

Today continued the multi-tasking. I masked and painted the engine house exterior, painted the gantry rails, printed the AC box and attempted to print the Bridgeport directly from the SketchUp warehouse model. It didn't work. It also needed some more support since the base was deforming so I stopped the print. It's good that you can pause and look at the work at any time without missing a beat. if you see early that the supports aren't working, you just kill the job and you're out a couple grams of resin.

The ac came out pretty well, but there was some distortion on the sides so I modified the drawing with some internal bracing. I also redrew the fan in the fan housing making it much thicker so the parts will hold together. What blows me away is that the door handle is actually sticking out from the door. The door louvers came out nice too. I'm still having that problem of material missing at the corners. It's just a little bit and no big deal to fix, but I want to know how to stop it from happening at all.

Right now (11:30 p.m.) the machine is running downstairs printing the modified fan assembly. It's a pleasant sounding machine, almost as if it's breathing. 

I glued the machine shop to the building, sanding all the spackle and then masking. As I noted this morning, I chose not to use the Press-n-Seal since it doesn't hold well under pressure.

Getting this beast up the cellar steps and out to the driveway was mildly difficult since not only is it a bit heavy, it's very unwieldy. I didn't break anything. To mask the big doors I used one piece of cardboard. I used the machine shop roof to mask its inside. I just sealed the junction to the building.

Here's what I looked like getting ready to paint this thing. The wind was shifting constantly and the super mask kept vapors and spray out of me. It was in the 90s and hotter then heck. 

It was so hot that some of the paint was arriving at the building as dust. It was drying before it reached the target. Get too close and it ran, stay back and it flashed. Not ideal paint conditions. The main roof served as the main interior mask.

Here it is back in the shop and, yes, I got it back in one piece.

The Rust-o-Leum Matte Camouflage tan was the perfect color for the 'concrete' exterior. I would work well for sidewalks too.

On Monday I'll do some minor sanding and maybe some touch up and the start adding windows.

Here's how the rivets showed up after painting. They're a pretty good alternative to punching your own.

Here's a good example of why printing on a 45 degree bias. This was the valve cover while it was printing. Instead of a broad surface that has to be pulled off the Teflon, you just have some small surface area that needs to release. The printer shows what each layer looks like as it progresses. Meanwhile, the Chinese guys have already shipped the new touch panel, but mine seems to be working better too. I have never had some much fun with a new tool!

Addendum to last night's post. My machine gave me a present this morning; a perfectly formed fan and housing for the AC unit. Thickening the blade's cross-section did the trick. There's round pin on the bottom. I could have had the hole made in printing, but decided, this time, I just drill it in post production. Just because I can print something doesn't mean I can't do more work on it… right? I have to drill and install the wires for the fan guard. One blade's trailing edge didn't quite form right. Don't know why that happened. Right now the machine's working on the leg assembly. I added more cross bracing to the leg frame to help resist any warpage.

That is one amazing machine. Thanks for sharing.