1:32 Animated Model of Battleship New Jersey's Steering System

[mmarcovitch]

If any of you read my Engine Room thread, you’ll remember that I started talking about the next model for the Battleship New Jersey Museum & Memorial; the Steering Gear System including the Rudder. When I delivered the Engine Room to the Ship las December, my nephew and I took many pictures of this space. And like the engine rorom project, the success of the project was predicated on obtaining accurate engineering drawings on which to base the model. John Miano, that wonder fellow who sent me the engine rooms drawings has done it again. Last night I got the drawings needed to start the project. He had already sent me a few including detail drawings of the inboard propeller, but I needed much more.

I now have the complete set of profiles for the rudder. The rudder—which I thought was a gigantic casting when I saw the ship in dry dock—is built like an airplane wing with formers and a welded skin. There is a very large hub forging that is the rudder mounting component. I plan on building the rudder as it in the 1:1 world with some part cutaway to show how it’s constructed.

John originally sent drawings he made of the constructed rudder. And then today he sent the entire set of formers and castings that make up the rudder, plus a nicely dimensioned drawing of the steering system. While there as dimensions missing—mainly those concerned with the equipment sizes themselves—there are enough meausurements of the overall siting and foundation that I can scale the rest nicely in SketchUp.

Here are some of detail drawings of the rudder assembly. I’m noodling how to create the rudder. If I made it out of styrene, I can cut all the rib profiless on my Silhouette vinyl cutter. It can’t cut through the styrene, but it can accurately scored for snapping or cutting with a knife. It could also be constructed out of ply ribs and balsa skinning a la a model airplane wing. I’m not sure how the cutter would work with ply, but it could work with balsa ribs. It will be time for experimentation. The rudder is essentially an airfoil the creates pressure differentials that help in creating steering pressure. The central hub is a fabricated part out of forgings and that would be 3D printed.

And here’s the main drawing.

Notice the different in angle of the starboard (bottom) and port side machines, and notice on the elevation drawing that the floor is not level in that part of the ship and the foundation accounts for that and levels it out.

This is what the machine looks like for real.

All the exterior walls are armored. The partition wall is not. The armored side walls are angled away at the top following a similar scheme as with the main armor. However, unlike the main armor, this is a structural part of the space. You can see the scalloped weld straps at the bottom edges.

With the drawings and hundreds of images I made during my December visit to ship, I believe I can do a respectable job. I still need some framing drawings from John Miano with the goal to frame that area of the ship as I did under the engine room. It’s more complicated in the aft due to the compound angularity of everything.

Here’s where we are now. I started working on the movind parts and planning on how to make them. I was originally going to again build in 1:48 scale, but it seemed too light for the moving parts. I can handle 1:32 on my printer as can be seen in this test fit of the massive 17’6" inboard propeller.

And here’s what’s done so far. I’m going to need help from DaveR in making the fillet junction between the hub and shaft. I can’t seem to figure out how to do a fillet between two curved surfaces.

For animation I’m going to use a heavy duty RC servo driven by an Arduion MicroController. I don’t know how to program these clever devices and was starting to look at YouTube tutorials. Then one of my genius grandkids suggested to just ask ChatGPT to write the program setting out the parameters I want it to meet. I want the rudder and machinery to turn to port 35º wait a 10 seconds, return to center wait 10 seconds, turn to starboard 35º and back to center and pause for a minute and then repeat.

It produced a perfect program in about 2 seconds including what terminals to connect the servo to on the Arfuino and offered to provide more instruction if you need it. It was my first application of AI and it was pretty darn good. There’s no reason I have to learn a programming language for an application I probably only going to use once.

I also am developing a walkthrough movie of the engine room based on photo realistically rendered screen prints of moving through the model describing the machinery and their functions. After writing it and getting Ryan Szimanski’s go ahead, I wasn’t happy about me narrating. So I sent the script to ChatGPT, and instructed to just one paragraph aloud as a test. I also told it what kind of voice I wanted and what the end use was for. I wanted something like Gregory Speak… i.e. deep and authoritative. It read it perfectly.

Now I have to be able to capture the audio output of my laptop for input to the movie soundtrack. My genius daughter-in-law had a solution for that in an app that captures audio on screen in a Mac. I’m all set to produce a perfectly narrated soundtrack for the program. It will be put up on YouTube and played on the monitor near the model in the ship.

Say what you will, this tech is astounding, and I’m not afraid to apply it.

[mmarcovitch]

Started some premliminary design work. My first decision is 1:48 is probably going to be too small for the operating features to be robust enough for museum usage, and the model is smaller than the others and can be done in a larger scale on my equipment. Since I’m not going to 3D print the largest part (the rudder) and build it like an RC model aircraft. I just did a check witrh the prop and it will fit on my 3D printer at 1:32. The model will be about 20" long at that scale with is doable.

The ram links have fillets where the shaft portion reaches the hub. I need some help in drawing that since they’re following a curved surface, so FOLLOW ME won’t work. I think I need DaveR’s help on this one.

The other scheme I’m noodling is how much fo the operating portion can be 3D printed resin and how much should I machine out of metal. For example: since the ram’s themselves are polished hydraulic pistons they should be duplicated in metal. The cylinders could also be machined. The wall thicknesses of the cylinder ends is thin and resin doesn’t fair well like that. Foundations and frames can be resin, also I can fabricate brass if I have to. I’m not afraid of machining since I was a metal shop teacher in another life a half century ago.

If I make the ram links out of resin, I will embed brass bearing into them and have metal spindles for all the rotating components. I’ve even been thinking about using miniature ball bearings that RC cars use to reduce any friction in the system.

Here’s how a 1:32 prop fits on the printer. At this scale the prop is almost as big as a real 1:1 outboard motor prop. It’s a shame I can’t make it out of bronze. I suppose someone could cast it for me, but that’s $$$ which I’m not spending. All those flat edges get hand shaped after printing.

So there’s a lot of thinking that will have to take place before this is done.

[mmarcovitch]

Design work is continuing even on Father’s Day (or becasue it is). Of course it’s never easy. The bias that the starboard side machinery sits compicates working in SketchUp and the slanted floor adds to the fun. The anti-corrosion protection covers are blocking some of the more esoteric details like how the ram crosshead is tied to the slides that flank the sides. It’s also difficult to see in pictures and in the drawings how the crosshead looks below the ram. The drawings I have are not about buidling the machine, but how the machine is situated within the space. I am making a big assumption that the draftsman that drew the machine was using accurate measurements to do it. i got the two-layer foundation bed most drawn.

This shows that angular bias that the starboard side machine has. The port side machine, on the other hand, is in line with the ship’s centerline. Don’t know why this is, but everything is the ship’s architecture is done deliberately, so I’m going to assume they knew what they were doing.

This is the work done so far. I don’t have an end view of the machinel; only top and side. For end views, I’m relying on pictures were took.

I will be printing the base in two parts to ensure that all of the geometry is faithfully captured.

I thickened the webbing on the lower base to give it more heft in the printed part. If I wasn’t going to animate this, the construction would be very simple. Once I decided that parts needed to move and, more importantly move for years and years without attention, design became more engineering and less art.

My youngest grandson, who just graduated as a mechanical engineer, is home for a month until he starts his real job in August. He’s going to “consult” with me on the design and he’s also an expert at programming Arduino micro-controllers. We’ll breadboard the operating mecahnisms to fine tune their design. He and his mechanical engineering brother both blame working with me in my shop for reasons why they chose this profession. Both were hired at good companies before graduating.

[mmarcovitch]

Design work continues. I’m working on several fronts simultaneously; the ram, rudder and operating mechanism. I have many decisions to make. While engineering the rudder stock bearing system under the base was easy, I don’t have a clue. The 3D printed rudder forging, rudder bearing support and rudder stock are complicated parts and there’s no apparent place to squeeze in a ball bearings. I may have to rely on sleeving with brass to prevent resin-to-resin surfaces.

The rudder forgings are hollow structures with is helpful in making resin parts. Save resin without sacrificing strength. You just have to have drainage holes to be able to evacuate the liquid, un-cured resin that resides within.

​The ram assembly now has crosshead and crosshead guides. The ram and crosshead guides are polished stainless steel rods. I’ve sized them with nominal dimensions that will allow me to use standard brass tubing with a sliding fit i.d. for crosshead bearings. There’s more detailing on the cylinders. I have to add hydraulic pipe flanges to their ends. And there’s a rudder position indicating device in the hydraulic compartment that’s driven by a rack and pinion arrangement. The rack moves with the cross head and the pinion motion is transmitted to the meter via a stout steel shaft with unviversal joints. I will model, but not animate this.

I don’t have good information about the bottom supports of the rudder top. I have a ragged 3D scan I took of it during my visitaton. Some fragments of views from the still pictures show a bit here and there, but not enough. I’ve asked Ryan to take some more specific pictures of that area. It’s probably where I can hide the upper bearing.

Here’s a progress shot of the rudder hub. I have drawings for four sections with three shown here. Those wings on the upper one are welding flanges to weld this structure into the ship’s framing. I’m going to try to include this feature to help tie the same part into the model. The rudder stock (shaft) scales to 8" and fits my printer (just barely) in 1:32. The lower extension is my added shafting to go below and be driven by the servo. The rudder is just wide enough at the bottom to hide it. If not, I will go to a smaller diameter shaft.

Underneath I’ve started specifying the operating components. I chosed the servo, the bearing pilow blocks that wil faciliate assembly, the micro-controller and power supply. The latter is already in the shop. Rudder travel is 35º in each direction.

There is a bottom pillow block that’s connected to a bottom plate. Having two bearings at the bottom should provide a stable shaft regardless of what I decide for the upper works.

[mmarcovitch]

The upper bearing is now engineered. It’s fitting into socket at the top of the rudder stock. The rudder stock itself will be non-moving. The motion will be carried with a 12.7 mm (1/2") rod fitted to the bore of the upper and lower ball bearings. This means that the rudder will be connected to the inner rod, not the stock. Complicates things a bit, but there was no way I cuold get the stock to work with the space for the bearings. The real thing uses sleeve bearings in the rudder hub.

I was able to figure out what the rudder hub base looked like from a few glimpses in the images I had plus a very ragged 3D scan I make with Scaniverse on my iPhone 16 Pro. There is a hand wheel driving a worm gear to a large pinion on the rudder stock. Supposedly, you can turn the rudder manually through this mechanism. I am dubious…

I found some ball bearings that will fit inside the rudder stock. They’re 3/4" o.d. X 1/2" i.d. and 5/32" thick. Perfect! The outer race will be captivated by using a small amount of epoxy. Also shown are the two machined brass spindles that hold the inner races of the ram rod bearings. Again, the outer races will be somehow captivated in the ram rod heads and the inner race will be a light press fit onto the spindles. There are two more brass spindles to support the ram ends of the rods. These will pass entirely thru the rams serving the mount for the ram rod bearing and locking the cross head to the ram. I will be machining these. They don’t rotate, only the bearings and the ram rods do.

I’ve started buying materials for the mechanics. I need to have them in hand to ensure that the printed parts conform to the sizes I’ve chosen. This model is costing more out-of-pocket than the previous ones due to the animation, but it will be worth it if I can pull it off.

I’m procrastinating designing the rudder hub and upper forging due to their complexity in drawing and printing challenges.

[mmarcovitch]

My procrastination was correct. The rudder hub has taken hours of design work and I still don’t have it done. Two things, so far, are complicating things. The weld flanges are not square with the part. Instead they are slanted according to the lay of the ship mold lines. I had to chop the top off the upper section to a slight angle corresponding to the floor angle of the machine room. The lower plate is slanted even more with the line of the hull at the bottom. I drew them square, removed them and then rotated them to the correct angle. The second problem was worse. How to skin the sides of this complex part? My first attempt took a few hours, but it was a bust. It wouldn’t/couldn’t print. The skin had no depth. That breaks rule number 2 of the three rules about designing for 3D printing. To add depth to the walls was an exercise in futility. I’m going back to old school. I’m going to skin the contours with 0.040" styrene sheet glued into gluing lands I’m designing in the part.

Here’s the upper part showing the angularity of the upper surface and weld flanges. The walls on this part are thick and will print. Note: This part has no taper making it easy to just extrude the walls to the desired height.

Here’s the comparison of the first attempt and the second. When the first attempt is exported as a printable STL file, the side walls, without thickness, disappear. Something that has no depth can’t exist in our 3 dimensional world. Everything that can exist has thickness unless we’re talking about an electromagnetic wave. All those zig-zag faces has to be hand woven and it took a lot of patience and care.

This part is not yet complete. It has a center section with more taper and it has a fairwater leading edge that has to be drawn and printed OR made model RC Plane style by carving a block of balsa or carving foam. I can even make the shape out of clay and cast a resin shape to do it. I can also skin the shape with balsa. I have experience with balsa and styrene. I do whatever is least expensive. This part supports the rudder, but the rudder itself will be skinned with balsa and be very light weight.

On another tack…

I got all the materials for the machinery either in hand or ordered. The stainless steel rod stock arrived from Amazon and I picked up the KS Metals brass tubing that will serve as the sleeve bearings. The fit is, by design, perfect.

All I have to do is cut the shafts to length. There will be no polishing or finishing needed, and being stainless, I won’t have to worry about corrosion occuring over the years the model will be on display.

I’ve got the bearings in hand and will design and run a test article to find the right diameters for the various sleeve and bearing mounting holes. Just telling the printer I want a hole of X size doesn’t guarantee that it will be that. Most likely it will not. The resin expands and changes during printing and post curing. For most applications this usually does’t matter. But here I’m trying to print holes that are going to be tight fits on bearings and machined pins. For the pins that I’m machining, I adjust the size based on the actual hole, but in the case of finished bearing and sleeves, I have to print the holes the right size. For through holes, I can open them up if they’re too small. Not so for blind holes like the cups that will hold the tiny bearings at the ends of the ram rods. These will have to be right.

So I’m creating a test article with a series of holes representing those needed in the model. I have converted all the key dimensions to full-size 1:1 to draw them in SketchUp and will reduce to the 1:32 scale. I will start with the actual recorded size of the bearing/sleeve o.d.s and then enlarge by small increments. I will note their sizes on the print itself. After printing, the sizes that work best will be incorporated into the model’s design. It will be much easier (and cheaper) to print on piece with a bunch of holes than to print the real complex parts of the model only to find that the bearings don’t fit.

[mmarcovitch]

As per my usual MO, I’m starting to print the rudder project with—what I considered—the most complicated part to design and print in the project; the rudder hub. After getting a successful print—Success as defined by the part forming properly—but was unsuccessful because the STL file I created from the drawing lacked part of it AND after examining the finished print I found things I needed to modify. This is a BIG part!

See that big hole in the bottom portion. That hole should be 1/2" diamter to accept the “real” rudder stock—a piece of 1/2"dia brass rod. The sleeve I drew to properly size the hole wasn’t grouped with the rest of the part. When I selected the part to be converted to an STL, it left the sleeve behind. The design problem was the little raised flange on the edges of the open areas. Those flanges are supposed to be the surface to which the balsa or styrene sheeting is supposed to join. The sleeves are just too small. The upper portion will remain open as this is encased within the hull. The lower portions are in the water and needed to be faired in. There is also a leading edge, called a fairwater, has to be applied that serves a similar purpose as the leading edge on an airplane wing… reduce drag and direct flow. I’m either going to print bulkheads and skin this portion like the rest or make it more or less solid. Bulkheads is prototypically correct and uses much less resin.

This is the redesigned rudder hub showing the enlarged flanges and properly sized rudder stock hole. I’m going to use the weld flanges on the part just as they are on the real ship; to provide more surface area for support.

I put this new part on the printer late yesterday, but when I checked it after an hour and a half found that the print was failing. The base raft was delaminating. This was occurring in an area where the part was not even forming yet. The only load on the raft was the supports themselves. I’m going to change the PFA film at the vat bottom. There’s too much adhesive to the film. It should be releasing each layer without that much tension. When the raft fails, everthing that portion supports will be distorted, warped or worse. You can see the part beginning to form in the background.

Meanwhile, I now have all the mechanical parts in my hands with the bottom main bearings arriving yesterday. These will fasten to the wood base with M4 screws.

My test article to determine the right sized holes worked as designed. Here was the ram rod bearing fitting the 3rd sized hole perfectly. The nominal sized hole printed too small. That number is the 1:1 size that I needed to draw on the plans and then reduced to 3.12% in the slicer to create the hole size needed for the bearings. I was asked on another place I post “why can’t I just use a drill of the correct size?” Valid question, but there’s a reason. There’s only one place where an undersized hole can be correctly opened with a drill. All the rest are either blind holes or in parts with wall thickness that couldn’t stand the abuse caused by drilling. And the blind holes also make sanding to size hard to do. Creating the openings right the first time was the best approach.

Here’s the rest of the rotating parts sitting in their correct hole sizes. Note, none of them are the first measured size.

I continued designing the rest of the moving machinery. I got the cylinders drawn including their properly-sized opening for the bear sleeves and the hydraulic fittings for the piping and the mechanism that tracks rudder position for the indicator on the control panel.

The indicator mechanism is my invention. I have no good images or drawings of this contraption (yet) and had to use my imagination. I know what it’s supposed to do and understand where the rack should be, but what the actual gearing is or it’s sizes is a wild ass guess.

Just for laughs, I’m going to see if the printer can resolve the gears hidden within. If it doesn’t work, I’ll print without the cutaway.

[mmarcovitch]

Yesterday, I took the rejected 1st attempt at a rudder hub and designated it as a test article and, in case the next print also fails, Plan B. It gave me the opportunity to see how the skinning technique should work. I used CA, but it wasn’t holding well and that’s a good thing to find out before the model’s sitting in a museum.

I used 0.020 styrene sheet, but the actual depth of the rabbets are 0.030" I will be getting some stock of that size. There’s not much of a radius to bend the sheet so even with the slightly thicker material, it should still hold. I also planning on using some small screws to retain the material, at least until the glue cures.

The Upper Rudder Stock hole wasn’t. It was a solid plug, so very careful drilling was done using a 1/2" brad-point drill while holding gently in my wood workers vise.

I’m using a 1/2 brass rod as the actual rudder stock, so the fit in the 1/2 hole was too tight. Because this shaft is actually going to be supported in ball bearings top and bottom, the holes through the resin part need to be very sloppy. It turns out that openning the hole with the largest metric drill I had—13.5mm—was the perfect clearance for the hole.

In this reject part, the bottom hole is way out of whack because the cylindrical center didn’t transfer to the slicer. I’ve drawn and will print a filler piece so this hole will be in spec as well.

This is a massive part and was just about the largest piece I can produce in my printer. Even will all the openings and hollow parts it used almost $9.00 worth of resin. The other large part will be the mating forging that forms the upper part of the rudder proper. I haven’t drawn that yet.

I will attempt to print the corrected part again on Monday, but if it doesn’t work, I can make this piece do the job.

I stil have to create the leading edge shape that goes on the angular flat side of the lower portion.

[mmarcovitch]

Not much visually to report today, but genius grandson #2 came over and brought me his Arduino Uno set. He spent about an hour with me in the shop and programmed the micro-controller to acuate a servo to move the rudder. Very slick!

I realized when awakening that with the smaller servo arm on the servo end and wanting to move the rudder in a 70º arc, the servo arm would have to move something more than 70º. I Drew a scale diagram in CorelDraw to visually calculate what that angle which would be the swing programmed into the Arduino.

My grandson (starting his new job at Seargent and Lundy Engineering Consultants in Chicago in two weeks, knows Arduino programming language fluently and did the program for me, saving me money on the Arduino and tons of time.

The top diagram was with the smaller servo arm. Then I found a longer arm in the servo kit and used that. I will be making the rudder arm out of some thick sheet brass and machine the hub out of brass round stock. I just received a set of Metric set screws that I will use for the clamp bolts to hold the rudder arm to the rudder stock. I also have to made a brass thrust bearing to support the lower end of the stock in the bottom bearing.

Meanwhile, had another dramatic failure of the raft on the fourth attempt to print the rudder hub. I made some adjustments, but then the printer kept giving a “Auto-leveling Failure” warning. I tried to fix it once, but ran out of shop time. I will get back to it tomorrow or Wednesday. Never a dull moment.

[mmarcovitch]

Two steps forward, one step back…

The repair of the first Rudder Hub print continues successfully. And I machined the center piece of the manufactured servo horn that’s going on the rudder stock. Those were the good things. My 4th and last attempt to print a new Rudder Hub failed miserably with almost a complete delamination of the base layers. I never had this problem before with this printer and I did print the repair sleeve and the top assembly for the rudder stock in between these failures.

Let’s see the good stuff first.

My sizing on the repair sleeve was dead on. It slide nicely onto the 1/2" brass rudder stock and fit loosely in the rudder hub, but good enough in the printed part.

Here it is after epoxying the sleeve into position. I used the rudder stock to hold the part in alignment while the epoxy set. With this being the actual part I’m going to use, it will need some more cleanup.

For the Servo Arm, I had some 1" round brass stock. First I had to hacksaw a chunk that would fit onn my tiny lathe. If I had a “real” lathe with a reasonably-sized hole through the arbor, I would have just chuck the piece in and got to work. I had to first get my hacksaw at my daughter’s house. My son in law borrowed it and forget to give it back. I rarely use this, but when you need it, there is no substitute.

I turned the lug end to just fit into the hole I cut in the brass arm. It didn’t have to be a tight fit. The whole deal is going to be soldered together.

I turned the part around and turned the o.d. of the body so it coincided with the edges of the arm. I couldn’t cut it all the way back to the smaller diameter, and didn’t trust the relatively light 3-jaw chuck. I also have a very robust 4-jaw independent chuck and put that to use to ensure that the piece is concetric and true. I use a Starrett Last Word dial indicator to get the runout to about 001".

After turning the o.d. I needed to chop off a whole lot and again used the hacksaw to do this. I have cutoff tools, but the lathe isn’t powerful enough to do very deep cutoff operations.

After cutting the excess I faced that end, center-drilled a starting point and drilled through with a smallish diamter pilot drill. I followed this up with the largest drill my machine can handle—1/4"—and then used a small boring bar to open the hole to a clearance fit of the rudder stock.

I turned it around again and shortened the narrow end. I only need a lip on that end for the soldering. I need all the stock on the fat end to handle the set screw, which I will do tormorrow.

Here it is cleaned, but not yet soldered.

I also got reasonably successful prints of the upper rudder support, but there was a drawing error in it. I could still make it work and my upper bearing just about fits. Need to relieve the hole slightly. I have modified the drawings and will attempt to reprint.

Now to the horrow show.
This is what a severe delamination looks like.

Notice that the initial part of the raft is firmly attached to the build plate, but the transition layers are not. I did a Google search to find out causes for this and made modifications to the printing parameters concerning the raft. Those changes made things worse, not better. I’m at a loss since the first print, while incorrect, did actually print. Everyone I attempted after that had some degree of delamination, from mild to wild. The first five layers cure for 8 to 12 times the working layers exposure time. I was curring my rafts at 35 seconds, but after reading about the problem reduced it to 26 seconds. My exposure time is 2.1 seconds. Then for the next five or six layers, the exposure time is reduced by a linear progression until it reaches the working time of 2.1. This is where the trouble is and I don’t yet how to fix it.

It’s why that reject part is so important.

Because of the misshapen supports, there’s a good chance that there is some resin stuck to the barrier film. If there’s even a very tiny bump on the film, the auto-leveling will report an error. This means, removing and emptying the vat through a fine mesh filter to catch anything floating in the resin, removing any bumps on the film and replacing and refilling the vat. It’s a messy part of the process that I don’t relish.

[mmarcovitch]

It happened one more time—which makes 5 attempts to print a successful rudder hub—so it’s time to rethink the problem. I was told that probably breaking it into sections and printing that way would solve the problem, and that’s what I’m doing. I 3/4 through redesigning the rudder hub in three sections. They’ll print on the machine better and much faster since the printer is only concerned how high something is (number of layers) to determine printing time.

Meanwhile, the reprinted rudder top frame, rudder crank and ram cylinders printed well and will be used. So it’s really a function of the large size of the raft that seems to be causing the trouble.

With just running a 1/4 drill by hand through the crank’s pivot hole, the 1/4 brass rod, that’s going to ultimately form the pivot points for the ram rods, fit perfectly.

I have to turn the ends down to the snug fit inside the small ram rod ball bearings. Will do on Monday. I hadn’t even post-hardened the crank nor did final finishing when I took the pic. Also shown on the above is the turned thrust bearing that will help support the weight of the rudder stock and focus its weight on the lower bearing’s rotating inner race.

I printed extras along with the rudder upper frame and redid the frame to make the handwheel more robust and did better job of supporting it. So it’s perfectly formed in this interation.

The cylinders are okay, but there will be some post-print trimming needed. There is some unwanted extra resin in the bore opening that I will have to remove. Something about the support scheme promoted this malformation. I can always remove stuff. Much easier than adding anything that’s missing.

After cleaning out the vat for the 5th time, I reset the printer to print the cylinders. We went out to the movies and when I got home saw the job was showed “COMPLETE” on the LCD panel. When I lifted the lid, I saw this.

In my haste to get something accomplished on the printer, I started the job without installing the vat hold down screws. The suction on the barrier film is so high, that after printing a few raft layers the entire vat was now rising and lowering with the build plate, and it lifted the whole deal to the finished postion while printing absolutely nothing! Since nothing was actually printed other than some prelimenary raft layers, nothing was stuck to the film. I just popped the errant layers off the build plate and started the job again… this time with the hold down screws holding the vat down. This was probably the most bone-headed stunt I ever did with the printer. I will not repeat it.

Work up yesterday morning at 5:00 am, and laid in bed for a couple of hours building the project in my head. I came up with 13 decisions on moving forward. I got up at 7:30 and wrote them down so I wouldn’t forget them.

I discovered that if I printed the upper foundation frames as I drew it, the crosshead and lower ram rods would have no where to go. The frame needs to be open on top at least to foreward cylinder.

I also decided to print the rudder framing complete and do it in sections like I’m now doing on the rudder hub. Still haven’t decided on sheathing with styrene or balsa. Both have their pros and cons. Styrene won’t require filling to get a smooth metal-like finish, but balsa responds better to gluing with epoxy or CA.

I also with need some metallic thrust washers wherever there’s UV resin-on-resin contact.

And I did some research on adding a sound module driven by the Arduino so I can have hydraulic motor sounds sychronized to the rudder’s motion. I was able find and download hydraulic machinery sounds and found that with some clever programming, the Arduino can control two things at once. The little Mini-SD Card player is less than $10 for two of them. And I have a speaker that would work. So it will not be expensive and adds another dimension of interest to the project. I have no idea what the steering gear actually sounded like. There are no operating battleships to hear it. I suppose other large ships with massive rudders would have similar sounds, but how do I capture them…. Hmmm.

The real fun of these kinds of never-before-done models is the imagination I have to generate to pull it off. And I’m turning 81 next Thursday. Age is just a number!

[mmarcovitch]

I posted my printing screwup on a FaceBook affinity group for Elegoo Saturn users. I got a lot of funny comments and many from folks who have done the exact same thing. Then today, I attempted to run a job with the build plate off the machine. I was working away on other stuff. When I checked back I got a “vat level too high” error which stops the machine. When I opened the lid, I see no build plate. It was minding its own business on the table next to the printer. I said I wasn’t going to make the last mistake again. But that didn’t preclude making a completely different mistake.

Got reasonable prints of the cylinders however, there’s some craft work needed to make them workable. I had to run a #22 drill through the crosshead guide holes so the polished stainless rods slipped in. I also had to do a Bondic repair on a small area on the same crosshead guide lug that didn’t have adequate support. I will sand this area smooth next session and when painted will no longer be seen.

I also have to clear out the cylinder bores to remove that excees resin that formed there.

I then machined the crank pins that will hold the small ball bearings for the ram rods. This is for the rudder crank end. After I print the crosshead itself, I will machine two more of these.

The upper rudder frame printed perfectly and the ball bearing fit the bore perfectly. The handwheel and worm gear also printed well and, so far, hasn’t broken.

Here’s the rudder crank with the machined pins. I used thin CA to secure them. I will mask them when painting. I chose to install in the unpainted part because the fits are so tight that any paint in the bores would cause a problem.

The bearing fit on the pins is a slip fit, not press. The lower ram rods are hanging down and the could slip off the pins so I’m going to secure the bearing to the pin using some Loctite thread lock anerobic adhesive. I will also have glue the outer race into the ram rod for the same reason.

Here’s how the upper rudder pieces stack when installed in the ship.

The last thing was machining another thrust collar. I made this one a bit thinner than the first. Fun turning project!

Set screws I’m using are Metric M4.

[joshua_studiojdb]

Miles, these are some phenomenal drawings that John Miano found for you that indeed look like a massive airplane wing. As someone who spent most of his days in architecture, I rarely get to see these kinds of drawings and it’s amazing to see the amount of detail that goes into the dimensioning.

My brother and I used to work at a generator company in the Netherlands, which would provide electricity for ships traveling on the Western European rivers. Seeing some of the engine room photographs brings me back some memories of my childhood.

When you model this geometry, do you model everything as groups and components or do you end up exploding everything so that it becomes a solid form, making sure that it is handled well in the 3D printer? I rarely draw things to be 3D printed but on every odd weekend job there are some parts and I tend to struggle every now and then to make sure that it is a solid shape.

Apart from posting on this forum, do you also have a YouTube channel where you showed the process of making this? Your screenshots and photographs do the detailing justice but I would love to see there the process in action. Thanks for sharing.

[mmarcovitch]

Joshua, let me answer your questions before I post today’s report. The rules are “Group everything as soon as possible” so geometry doesn’t stick to other surfaces where you may need to reposition things. I try to do that as much as possible, BUT, often I find myself in the design mode adding shapes and modifying things and forgetting to group. I also use components as often as possible, but many times, I find that I’m using “Make Unique” way too often telling me I probably shouldn’t have made it a component in the first place. When this happens, I pay for it later.

Neither the slicer nor the STL converter cares about groups, only normal faces, and bona fide solids (no leaking if filled with water). Grouping is strictly a SketchUp issue.

Today saw more successes. The scheme to create realistic and robust ram rods was flawless. The groove to receive the 1/16" reinforcing brass rod was sized perfectly. the rod required a little coaxing to slide down to the bottom. And the bore size to receive the ball bearings was aslo dead on. It took just a bit of pressure from my parallel jaw pliers to seat them to the bottom. It was very rewarding.

I used Bondic to fill and level the gap. You apply it in several layers, hitting each with the little UV LED, it cures in a few seconds.

Here’s what it looks like after sanding, but without paint.

I’m glad I spent the extra time creating that “Hole Sizing” test article. The printed bore diameter was a perfect press fit. All this work was done before post-curing. I wanted the resin to still be pretty flexible so any pressure would be accommodated and not forcing the ends to crack when the bearings were pressed home.

I finished all the finishing work on the ram cylinders. I made extra and used one to experiment on how to best remove that excess plastic what was semi-blocking the bore. At first I tried using the Dremel with a carbide routed, but it was way too rough and there was no way I could maintain concentricity. Chucking it in the lathe seemed like it could work if I was very careful.

It did work. When I did the first piece, I creeped up on the final i.d., but found that if I left the boring bar at the same position, carefully re-chucking each piece in the same orientation and slowly feeding the bar into the bore, I was able to get the finished i.d. in one cut. I repeated this for the remaining cylinders.

Here’s a movie showing the cut being made. The feed had to be very slow, otherwise, the tool would dig in an either rip it out of the chuck or disintegrate it. I’m having trouble getting this movie to run on YouTube. See if it works better for you…

SG Cyl Bore Repair SD 480p

The brass tubing cylinder liners are a slip fit. This is okay since I’m going to secure them with epoxy. The cylinder walls aren’t that thick and a press fit would have invited big trouble.

The ram is going to slide nicely. I will add some very nice grease to make it slider even better.

The foundation print was okay, but had some weird areas that were porous and spongy. Never saw that before. Using the part fails or doesn’t. It doesn’t decompose. I filled all the areas as best as I could with Bondic and it will be okay. The areas are not easily visible.

I placed the cylinders in their respective positions to show how they will be fixed.

Last thing I did was cut the four cylinder bearing liners. Cutting thin wall tubing is difficult. My mini-cutoff saw’s abrasive wheel has been wearing so the diameter is no longer able to cut more than about 1/4" diameter stock. MicroMark doesn’t seem to have any in stock.

Starting tomorrow, work will stop while we take a mid-year trip back East to visit with family and friends. Work will commence after Aug 1.

[joshua_studiojdb]

@mmarcovitch

Joshua, let me answer your questions before I post today’s report. The rules are “Group everything as soon as possible” so geometry doesn’t stick to other surfaces where you may need to reposition things. I try to do that as much as possible, BUT, often I find myself in the design mode adding shapes and modifying things and forgetting to group. I also use components as often as possible, but many times, I find that I’m using “Make Unique” way too often telling me I probably shouldn’t have made it a component in the first place. When this happens, I pay for it later.

Neither the slicer nor the STL converter cares about groups, only normal faces, and bona fide solids (no leaking if filled with water). Grouping is strictly a SketchUp issue.

Thank you sir. Yes the golden rule of groups. It’s something that a lot of students that I meet at colleges struggle with when they come from a Revit background and they get to experience SketchUp for the first time. The ease of drawing gives way to an abundance of drawing - forgetting to group any of the geometry. Glad to see more of the project. Thank you.

[mmarcovitch]

Got back from the trip back East to visit family and friends. It was my 81st birthday so we had some celebratiobns. I gained four pounds and my wife and I ended up with puffy ankles due to all the eating, sitting around and the 11 hour car ride to and from. It will take some time to get things back on an even keel.

And I got into the shop today and completed and important and challenging step: finishing up the oh-so-critical rams.

The 1:32 ram scales out to 4.25", I painted some black Sharpie on the approximate place of each ram that represented the cut line, and then hand hacksawed them a little bit long. I used some cutting oil to make the hacksawing of stainless steel a bit easier. Stainless is tough stuff to cut and machine.

I then chucked each piece in my Taig Lathe and faced the end down to the line. I supported the overhanging lenth with a steady rest. I then reversed the ends and faced off the other end. Both ends had a slight chamfer added to eliminate any burr that would snag the cylinder bearings.

The crosshead hole lies in the exact center of the ram. I needed a hole that was 1/4" and directly through the center and plumb. I needed to create a small flat so the drills wouldn’t skid off the surface. I used a flat diamond coated disc with Dremel to create a small flat. I then center-punched the location of the drilled hole.

I used a #33 drill as the pilot hole—only because it was sitting at the drill press from the tapping operation last performed—and then followed up with the 1/4". It was a painstacking operation to ensure that the hole was being drilled down the center of the cylindrical shape and supported. I then used a countersink to chamfer the edges so the brass crank pin would slide in unimpeded.

Here’s the end result.

And after trying on a crosshead, I was rewarded with a nice fit. Tomorrow I will cut the crank pin to length and then machine these ends just like I did for the ones in the rudder crank.

As much as I like to visit with friends and family, my happy place is creating in the shop.

[ateliernab]

@mmarcovitch

As much as I like to visit with friends and family, my happy place is creating in the shop.

we should print that on a teeshirt…

[mmarcovitch]

Be my quest… It kind of rings true.

Due to waiting for our HVAC tech to arrive, got very little accomplished today. Whenever you go on a trip, when you return something’s not working. This time it was the 2nd story HVAC unit. It’s quite old, but only needed a motor start capacitor.

I did start the machining of the crosshead pins. Got the first end of the first one finished.

Tomorrow, these will be finished and I will start painting and assembling this part of the job.

I want to add sound and Arduino just came out with a 2 channel unit the Arduino Uno-Q. While I’m not certain, I think I can control the motion AND play the synchronized sound track with it.

[mmarcovitch]

All of the upper motion parts machining is done. I’m now determining the assembly/paint processes to do it most effectively and efficiently.

After measuring the proper length of the crosshead pins, I cut off the excess with a cutoff tool in the lathe. My MicroLux Mini Cutoff Saw needs an abrasive wheel replacement and MicroMark is out of stock so I had to do it in the lathe. That said, I just did an Amazon search and found many 2" wheels with a 3/8" hole on sale. I’m going that route and will have them in my hand soon.

I machined both and they work well.

Next up was cutting the four bushings for the cross head slides. Again, without the cutoff saw, I did it in the lathe.

So all the motion for the machine itself is machined and ready for paint/assembly.

On another front… I printed the first two parts of the disassembled rudder hub that caused so much trouble printing. Those parts are done and draining in the machine. I’ll finish them up tomorrow and print the main upper portion.

As seen by this image, the lower portion print failed. Probably a support failure, but I’m not sure. I awoke this morning thinking of redesigning… again. I don’t have to make the part as in the prototype. The insides don’t need to be displayed.

The difficulty is the taper. The whole part that extends out of the hull is tapered towards the bottom. Making assymetrical tapers in SketchUp isn’t easy. I’ll keep you posted…

[mmarcovitch]

The previous post was Friday’s. Today i cut the four cross head guides after changing the abrasive cutoff wheel on my MicroLux Mini-Cutoff Saw. MicroMark sells this, but the unit and all the wheels were out of stock. I found many on Amazon. They arrived yesterday so I was able to cut the parts today.

I completely redid the drawings for the rudder hub, both simplifying and skinning the parts to eliminate the skinning process afterwards. I also eliminated all the interior cavities shown on the real forging and instead, hollowed out a solid part with plenty of drain holes to remove any liquid uncured resin remaining within. I uploaded both parts: the lower and middle sections are now a single part, and the upper another.

I uploaded both as STL files to my slicer and prepared them for print. Today, I was able to start the job, but was concerned about the failures and partial failures I was having and, using a technique I recently read in an Elegoo Saturn Group on FB, I checked the LCD screen. It’s been in service over 2 years and has done many jobs.

The test showed dead pixels.

Where there are dead pixels there is unexposed resin which is a failure. Those spongy sections I was seeing in my latest prints on some parts was due to these dead pixels. And the failure of the last print could also be from it. If the dead spots happen to fall on supports that are being formed, those supports will either be missing or flawed so they can’t perform.

I’ve ordered a new LCD from Amazon and will rebuild the machine in a couple of days. I’ve replaced LCDs on both of my previous Elegoo printers. It is considered a wear item with a life of 2,000 hours. I don’t know if I’ve reached that. If do know it’s been in operation since June of 2024. Screens are killed by several factors including: long exposure to strong UV light and excessive pressures created when failed print resin that’s sticking to the barrier film applies high spot loads onto the LCD. The new printers have strong glass shields to help protect them from the latter.

Stay tuned…