• Hi Guest! Forum rules have been updated. All users please read here.

Merriman's Submarine Modelling

merriman

David Douglass Merriman lll
Joined
Mar 18, 2013
Messages
336
Reaction score
207
Today was spent machining the motor bulkheads to fit the Lexan cylinders that house the items within the model submarine that have to be maintained in a dry environment. Most of the work dedicated to mounting the outrunner type brushless motors within the MB's.





To facilitate motor testing and MB certification I assembled this test rig comprising a battery, electronic speed controller, switch, and servo-setter (pressed into service to act as a 'throttle').





Tomorrow I explore what it takes to assemble, test, and certify a MB equipped with the gear-splitter.



 

merriman

David Douglass Merriman lll
Joined
Mar 18, 2013
Messages
336
Reaction score
207
Old and new r/c submarine operating systems on display here. Each type embodies the means of control, propulsion, and ballast water management needed to make a scale model submarine work in a credible and reliable manner.
The system on top is the old, single cylinder type SubDriver (SD). The two bulkheads that divided the cylinder into three spaces are fixed in place with machine screws -- screw holes that sometimes resulted in cracks that would migrate over the seals causing water leaks into the dry spaces. And this type SD compelled me to select one diameter size cylinder for the entire length of the SD, this often not the ideal utilization of annular space between it and the interior of the model submarines hull. And the single cylinder system had just too many hoses and manifolds sitting proud of the cylinder, all potential points of failure.
Many of the SD shortcomings have been eliminated with the next step up the evolutionary ladder: the Modular SubDriver (MSD), seen at the bottom of the picture.
No mechanical fasteners to hold bulkheads in place. Instead, only O-ring friction holds three separate lengths of Lexan cylinder in place -- this innovation making access for repair, maintenance, and adjustment a much easier task. As an added benefit the MSD’s ballast water management sub-system has been consolidated into a tight, accessible package, eliminating most of the external plumbing which plagued the original SubDriver design.


The MSD contains the same devices as the earlier SD but does it within an envelope that can quickly and easily be changed in length and diameter to suit a specific application. As exemplified with this tear-drop shaped hull the arrangement of the three separate cylinders has been selected to make maximum use of the available space within this free-flooding model submarine model.

With few exceptions an r/c submarine makes use of the traditional devices as other r/c controlled vehicles. However, only air-ships and submarines require a means of changing the vehicles displacement within the fluid it operates; and only a submarine requires an assured means of autonomously sensing and correcting its pitch angle. The main destinguishing burden an r/c submarine has over all other vehicle types is the need to keep things dry at all times.


There are many ways to move water in and out of the ballast tank if the intent is to change the submarines displacement by taking on an amount of water weight equal to the weight of water the above waterline structures displace when immersed.
My SemiASperated (SAS) ballast water management sub-system pushes the water out of the ballast tank by displacing it with air. Air either scavenged from within the dry spaces of the system or from atmosphere. Pictured is an old SubDriver system employing the SAS cycle -- a Rube Goldberg delight, to be sure.


This better illustrates the SAS ballast sub-system.
A vent valve atop the ballast tank (not shown) opens, venting the air from within the ballast tank, allowing water to fill the tank and the submarine is totally under water. The formerly above waterline portions of the submarine, now fully immersed in water, produces a buoyant force equal to, but opposed to, the weight of the ballast water taken on. The boat assumes the state of ‘neutral buoyancy’.
To surface the water in the ballast tank is blown out with air compressed by the LPB. Air is initially scavenged from within the SubDrivers interior (the snorkel valve is closed). Once the sail broaches air is taken from atmosphere.
Internal air is only good for a partial blow of the ballast tank, but it’s enough to broach the sail above the surface. Once the snorkel head-valve opens the partial vacuum created within the dry spaces is back-filled with surface air and the blow continues with air from the surface.


Two-valve protection is an almost religious tenant within the submarine community – you always want a back-up stop to any line subjected to sea pressure. That philosophy has carried over to my model submarines as well. The ‘safety float-valve’ is the back-up valve within the induction side of the SAS ballast sub-system. The primary stop to water ingress to the induction line is the snorkel head-valve up within the sail. The safety float-valve is the backup, it prevents any water that gets past the snorkel from leaking into the SubDrivers dry spaces – it only closes if there is water in the line, otherwise it passes air going in or out of the systems dry spaces.
Here I’m testing a unit by injecting first air, then water. It must pass the air, but immediately block the flow of water.


Within the safety float-valve a float, with a rubber disc atop it and a weight within it will remain clear of the air passage between the nipple at the bottom and the nipple at the top of the device. However, should water get into the safety float-valve, the passage is blocked, keeping water from getting into the dry space of the SD/MSD.


The body of the safety float-valve is formed from a short length of copper pipe and two copper caps. The lower cap is permanently soldered in place; the top cap is removable for servicing and is secured and made watertight with RTV adhesive. Here I’m cleaning parts for soldering. The end-game: I’m holding a completed, ready-for-issue unit.




The air-pump used to discharge the ballast water is this small diaphragm pump, modified to make it suitable for handling water as well as air – if, and when, water gets into the induction line (and it will!!) I don’t want any of it to get out of the pump and into the dry spaces of the SD/MSD. The elastic elements of a diaphragm pump prevent ‘water hammering’ of the mechanism should it encounter a non-compressible fluid. Though technically described as positive displacement type pumps, because of their slight ‘give’, the diaphragm type will move water or air with great enthusiasm and without hammering itself to death.
Each pump -- I revert to submarine-speak and call them Low Pressure Blowers (LPB) – had its rubber seal, which isolates the pump workings from the pumps surroundings, mashed tighter within its housing through a few modifications of the assembly. This work to insure no water leakage past the pump body and into the dry space.


Each modified LPB was then subjected to about 15 psig of water pressure at the discharge and induction sides of the pump and the pump body examined for water leakage past the seal.


Once a LPB had passed its leak-check it was then outfitted with two spark-suppression .01 micro Farad capacitors. Electronic ‘noise’ within the tight confines of an r/c vehicle has to be avoided; spark-suppression of brushed motors and switches is a necessity.


Final check of the LPB’s was to spin the motor under load (dead-header test), followed by an affirmation of correct discharge rate. At this point I declare the units, ready for issue.


The new MSD design has greatly streamlined the integration of the SAS elements. Here you see a typical MSD ‘after ballast tank bulkhead’, or ‘union’ along with the fasteners that hold things together, servo, linkages, LPB switch, safety float-valve, plumbing, and LPB.
Unlike the earlier SD with its many externally running hoses, nipples and manifolds, the new MSD’s SAS plumbing is all internal with only the flexible induction hose running from the system to the snorkel head valve located high up in the submarines sail.
The union is of two-piece construction which permits me to mix-and-match different diameter lengths of Lexan cylinder. This particular union provides interconnection between a 2.5” diameter after dry space cylinder and a 3” diameter ballast tank cylinder.


After assembling the two union halves the ballast sub-system servo – that opens and closes the ballast tank vent valve as well as activating the limit-switch that turns the LPB on and off – is strapped in place. The servo pushrod passes into the ballast tank through a watertight seal and works the linkage that opens/closes the vent valve atop the ballast tank.


The LPB and safety float-valve mount, as a unit, in front of the servo. Note that the LPB induction is split between the safety float-valve and nipple which connects to the snorkel head-valve through a long length of flexible hose. The LPB discharges directly into the ballast tank.

 

merriman

David Douglass Merriman lll
Joined
Mar 18, 2013
Messages
336
Reaction score
207
When I got into the r/c vehicle game, in the mid-60’s, the vehicles receiver, unless it was of the super-heterodyne type suffered from low selectivity; it was most susceptible to adjacent frequencies, ‘electrical noise’ and unwanted RF from other devices in close proximity to the receiver. Back in those days, when dinosaurs still roamed the Earth, all electrical and electronic devices within the model airplane or boat had to be well distanced and spark suppressed if any credible range was to be achieved between transmitter and receiver. The devices could not be packed in close proximity to one another; the inverse square law was (and still is) your friend.
Flash forward to today: We are now using receivers that not only feature very selective detectors, and the signal they are tuned for is ‘processed’ to weed out both external and internal RF energy not emanating from the controlling transmitter. And it is these advancements in receiver technology – and the introduction of brushless motors, servos that are suppressed at the factory and other device improvements -- that permits dense crowding of electrical and electronic devices within the tight confines of a SD or MSD.


Once I had settle on a rational placement of the devices within the after dry space I set about designing and proofing a means of mounting those devices within the cylinder. The eventual foundations would be fabricated from .031” thick aluminum sheet, in the form of trays and circular bulkheads. Sheet metal work 101. Of course, it did not go according to plan.
Good practice: Before committing to the metal one should first mocked-up the foundations using cardboard cut with knife and scissors – easy to work with and easily modified as problems of fit and placement were resolved. I started with an initial cardboard template, and from that marked out a cardboard mock-up; that mock-up to affirm fit within the after dry space.

I took advantage of the motor mounting studs, using their forward ends to make a four-point attachment to the after vertical face of the eventual device foundations.


Once I had constructed the cardboard foundation mock-up and worked it – along with the template – to fit the cylinder, I quickly shaped scrap pieces of 20 lbs. RenShape to stand in for the actual devices that would eventually populate production MSD’s. I make it a practice to slightly over-size stand-ins like this to account for mounting tape, leads, heat-shrink wrap, and other unaccounted for obstructions. In other words: if I can get the stand-ins to fit, I won’t have any trouble getting the actual devices to fit.
IDIOT!
Nothing revolutionary here in the design process; shipyards have been doing this ‘try it before you buy it’ mocking up for centuries. I don’t invent ideas. I steal ideas (but, only the good ones). Though, sometimes I don’t apply those ideas very well.


The MSD will accommodate five servos. On in the after ballast tank union, two at the forward end of the forward dry space, and two back in the after dry space – the space I’m working up the device foundation for.

As this size MSD is for the intermediate-to-large size r/c submarines we wanted those two after servos to have the ass to move substantially sized control surfaces, so I sized the servo stand-ins to represent ‘standard’ sized servos. Sure, they’re big bulky things with a substantial foot-print. But, what’s a guy to do? Once those stand-ins were in place there was precious little real-estate left for the other device stand-ins.
And this brings us full-circle back from my observation about the ability of today’s receivers to tolerate other electrical and electronic devices in close proximity without being swamped with RF ‘noise’. The packaging illustrated here would have been impossible back in the 60’s! Some things do improve with age.


But will the organized chaos fit?.... Hell no!


Well … what worked in mock-up, did not work when I started to mount the actual hardware. Servo leads got in the way; the receiver pin array stood too proud and would not fit the narrow slot I had initially assigned for it. Little things like that, not accounted for in mock-up spilled all the beans. A re-think of how things would be arranged was in order, on the fly. Chaos management 101.
And this, boys and girls, is why you test fit before committing to a permanent install.


This is as far as I got with the real-deal install: the two big ‘standard’ sized servos and that rather chunky Mtroniks brushless motor ESC. I found that the forward end of the aluminum foundation was not getting it done. The vertical attachment to the motor mounting studs was good, as was the long running horizontal base. But the forward, starboard plate, where I planned to mount the receiver was too tight a fit. So the forward end of the foundation needed a re-work, and I had to find a new home for the receiver – I elected to raft it over the ESC.


A little sketching to skull out the new forward area of the foundation and in no time I had my plan-B. A little sheet-metal-thinking-on-paper and it was worked out that a single piece of sheet could be bent to produce the receiver raft, as well as two vertical faces to mount the smaller devices. Origami for idiots!


From brain, to sketch, to template, to laid-out sheet aluminum, to band saw, drill press, and mini-break.

 

merriman

David Douglass Merriman lll
Joined
Mar 18, 2013
Messages
336
Reaction score
207
The electronic devices within the cylinder are commercial products with leads usually longer than required to reach the receiver. Aboard the Modular SubDriver the receiver is the nexus from which most propulsion, control, and ballast sub-systems receive their intelligence.

To get everything to work as a system its good practice to first arrange all the electrical and electronic devices outside of the very cramped MSD and get things operational. Problems are identified and corrected easily at this stage and some of the setup protocols performed. Some setup tasks have to be differed until the devices are installed within the MSD.



Longer than necessary leads within the tight confines of the cylinder not only makes for a messy arrangement, they also act as antennas that capture spurious RF energy and pump it into the receiver where that noise could swamp out the transmitted signal.

Long leads bad.

Short leads good.

Each lead is shortened but for a little slack to alleviate any strain on the wires, plug and PCB. Getting rid of all that spaghetti makes for a much tighter assembly within the cylinder and greatly reduces the possibility of RF noise causing self-glitching.



Dykes, wire-stripers, solder, 25-Watt soldering iron, non-acid flux, heat-shrink tubing, and patience, outside door secured so that screams of rage don’t disturb the neighbors, and the leads are sized to suit the devices distance from the receiver.



Gathering, testing, programming, and integrating the electrical and electronic devices – wrangling all the magic gizmos that make the damned thing work, are tasks I loathe doing; these aspects of r/c model submarine building and operation interest me not in the slightest!!!!

I’m competent enough when it comes to donkey-work like hooking up receiver to servo or ESC, that basic stuff I can handle, any moron can do that! BUT, what frosts my butt; what drives me nuts; what sets my hair on fire, is the task of ‘setup’ of the two devices unique to r/c submarines: the Battery Link Monitor (BLM) and Angle Driver (AD2).

I’ve read and re-read the instructions and simply cannot get a setup to work right the first time. To be fair to the product, I can think of no better word description than what Kevin has authored in the instructions.

Apparently I’m not wired to translate written instructions into the perfectly choreographed button-pushing, and transmitter stick twiddling actions needed to get the devices to talk to one another in a civil manner. I NEVER get it right the first time. But, I’m a special type of hard-head; I eventually get the damned things working right. Given a choice between chewing broken glass and setting up these devices I would have to think about it a few moments.

I hate this shit!

Don’t get me wrong. No one on this planet appreciates the availability and utility of these devices more than me; particularly the units produced by Kevin McLeod of KMH. His devices present small foot-prints, are rock solid, and consume very little current. But, because of the sophistication of their operation and enhanced capabilities, these devices -- specifically the Battery Link Monitor (BLM) and Angle Driver (AD2) -- demand full attention as you attempt to follow the instructions. Programming is specific to the model, r/c system, and battery type. One size does not fit all.



Only r/c submarines require a device to autonomously drive the stern planes to keep the model horizontal when running underwater; and a fail-safe device to blow ballast water if the signal is lost (not a unique requirement in itself to r/c submarines), and also actuates if the battery voltage drops to a dangerous level.
(The ADF2 pictured below is an older type that featured an integrated fail-safe circuit, but that chore is now handled by the BLM).



Setup of the BLM can be done before putting it into the cylinder, but if care is taken to make it assessable once mounted – you have to get at it to push the ‘set-button’ – there’s no problem programming it in situ. As you can see I’ve mounted this device on the side of the starboard (stern plane) servo.



A simpler setup routine is employed to get the Depth Commander (DC) device up and running. This optional piece of gear drives the bow/fairwater planes to maintain the last commanded depth setting. It can be commanded off and on from the transmitter. Slick! This device is not vital, but something I’ve come to embrace.

You see, I’m a bit of a cow-boy when it comes to mixing it up with surface craft (targets) at the lake. The DC greatly reduces operator work-load as one weaves in and out and under the surface pukes. Much good fun to be had busting up the regimentation of a nicely arrayed battle-group.

“What? Something spooked you guys? There were collisions? You all should work on your group discipline … just say’n”.
RHIPMF’s



The AD2 has to be set once mounted on the MSD’s device foundation. This is because the devices reference plane is gravitational force and the basic setup operation tells the devise which-way-is-up.



The Battery Link Monitor is a device that, in the fail-safe modes, autonomously commands a ballast tank blow if there is a Loss Of Signal (LOS) between transmitter and receiver; and/or the battery runs down to a preset low-voltage point.

The monitor mode logs the number of LOS events occurred during the last sortie. Informing you to what degree the body of water you are operating in is attenuating the transmitted signal. Good to know stuff as you prepare for the next patrol. It’s been my experience that any body of water, be it pool or lake, has its radio ‘dead spots’ which when identified should be avoided when operating the model submarine submerged. The BLM’s monitor is a very useful feature in that it helps you survey the patrol area for these no-go locations.

 

Grey Havoc

The path not taken.
Senior Member
Joined
Oct 9, 2009
Messages
9,701
Reaction score
495
You see, I’m a bit of a cow-boy when it comes to mixing it up with surface craft (targets) at the lake. The DC greatly reduces operator work-load as one weaves in and out and under the surface pukes. Much good fun to be had busting up the regimentation of a nicely arrayed battle-group.

“What? Something spooked you guys? There were collisions? You all should work on your group discipline … just say’n”.
RHIPMF’s
:D
 

merriman

David Douglass Merriman lll
Joined
Mar 18, 2013
Messages
336
Reaction score
207
Mocking up of the Modular SubDriver is over, I’ve just finished populating the proof MSD with the devices needed to make it operational and tomorrow I set about the tasks of certifying the unit for operation (leak-check and proper operation of the SAS ballast sub-system). And, with that, the time has come to integrate the MSD with this old 1/72 THRESHER class r/c submarine hull and check this puppy out in the water.



The rudder and stern plane linkage was a straightforward affair. The control surfaces connecting through a ‘yoke’ that not only served as bell-crank but also the means of providing clearance for the centrally running propeller shaft.

The sail planes differed in that the bell-crank was formed from two ‘floating’ magnetic couplers that translated axial motion to rotational motion; a bell-crank in function but not requiring making up fittings within the tight confines of the THRESHER’s very narrow sail.



Arraying two, even three different diameter cylinder sections into a MSD presents its own special problems over a constant diameter SubDriver. At least three support saddles are required, each sized to fit the cylinder over it. And there are no mechanical fasteners holding the three cylinders together, only O-ring friction retains each cylinder in place on its accommodating radial flanges. Care has to be taken to not accidently twist or bend the assembled cylinders out of alignment with one another during handling.



This shot well demonstrates how much annular space is made available by changing diameters of the three Lexan cylinders. As you can see, there is plenty of room for buoyant foam (to counter the weight of the fixed lead ballast low in the hull needed to produce static roll stability) between hull and MSD.

You can just make out the two pushrods extending forward from the face of the forward bulkhead. One has already been outfitted with a magnetic coupler and will actuate the sail-planes; the other pushrod will eventually operate the four torpedo launchers through an escapement sequencer – but I won’t pour time into that till I first validate the MSD. The model here is, after all, a test-mule and the primary mission is to weed out problems not yet identified.



A close look at the stern plane and rudder yokes. Note how they are shaped to permit unobstructed passage of the intermediate propeller drive shaft. These are cast from white-metal (Tin and Antimony) in a two-part, disc shaped, rubber centrifugal tool.



My first attempt to work a linkage between the servo, located within the forward bulkhead of the MSD, and the planes set up high on the sail. It just did not work out. What should have been axial motion instead, because of the magnetic couplers propensity to shift laterally, resulted in severe binding and loss of motion; the entire exercise an example of bad design from beginning to end and the lack of good sense to find an alternative solution right away. Hard-headedness can sometimes be a virtue. But, usually not.



No matter how much lipstick I smeared on this pig, it stubbornly refused to be anything other than a pig. It refused to work. A half-day’s work went straight into the shit-can. Sometimes that three-pointer effort, launched at extremely high velocity, from across the shop is damned good therapy!



The sail is held down onto the hull with two machine screws. Not only is the SAS snorkel head-valve housed within the tight confines of the sail, so too is the linkage that operates the sail-planes. A tight fit, but it all works ... NOW!



Members of the Captain’s-conference (a body experienced submarine officers who formulated much of the desired characteristics as a new submarine design took shape) must have been away when Portsmouth designed the THRESHER class boats. That or they put a premium on ships speed and maneuverability over optical and electronic sensors.

The initial boats of the THRESHER-PERMIT class were a step-back from the electronic information gathering capabilities of the earlier SSN’s – the sail was short in beam and length in an effort to not only reduce friction, but also to mitigate the dreaded sail induces, ‘sail-roll’ of boats with larger sails that engage in high rate underwater turns at speed. The first boats featured the diminutive sail. One thing in particular must have bugged the hell out of the ship control party was the single Type-2 periscope with no radar ranging or warning antenna atop it, and minimal light-gathering ability. The distinctive small sails would be replaced on follow on boats of the class with a more comprehensive antenna and scope array housed within lengthened sails.



Not the current project, but one I got operational a few years back using the ‘old’ constant diameter SubDriver. Just a tease to show you the end-game.

 

merriman

David Douglass Merriman lll
Joined
Mar 18, 2013
Messages
336
Reaction score
207
At long last, after months of consultation with my Boss, Bob Martin (who sells my stuff) I got underway with prototype work. Masters; tools; trial assembly and evaluation; re-design; master and tool modification; assembly, evaluation, and tentative approval; and finally this pre-production Modular SubDriver assembled, tested, certified, and ready to be proofed in an honest-to-god, real-life, deep-sea-wonder r/c submarine. This has taken much time, significant material, and neglect/deferral of other responsibilities. I’m confident it’s all been worth it.

At this point I’m just about done integrating the MSD with the hull; doing the 101 little things it takes to make the two comfortably compatible: working out the saddles (foundations) the MSD sits on; installing and dialing-in the control surface linkages; making the propeller intermediate drive shaft that fits between the MSD and propeller shaft; and installing fixed ballast weight in the bottom of the hull so as to establish the vehicles center of gravity at the longitudinal center of the ballast tank (itself located half-way along the length of the submarine) and low to the hulls longitudinal centerline so as to produce static stability about the roll axis. CB high, CG low; the greater that moment the more statically stable becomes the submarine.



Unlike the earlier SubDriver (SD), which employed a single constant diameter length of Lexan cylinder, the modular SubDriver (MSD) presents the opportunity to integrate different lengths and diameters of cylinder in order to get a more conformal fit of system to hull. The separable cylinders afford quicker and easier access to the devices within the system.



Driving home Darren Scannell’s recent cautionary posting to the Warships Models Underway forum, is the problem I encountered after splicing servo lead wires together – I wound up with thickened leads which were very hard to pass through the narrow confines of the ballast tank conduit (a 5/16” brass tube) interconnecting the forward and after dry spaces.



The magnetic mission-switch greatly simplifies system start-up and shut-down. With this very useful device there is no need for a boot-seal over a mechanical switch toggle, or need to pop the forward bulkhead on and off its cylinder to access an internal switch.

Though making for a difficult cable-run through the conduit, there is much to recommend placing servos at the extreme front end of the MSD. No need for long external pushrods running from the SD’s motor-bulkhead to the front of the boat; less clutter, in the form of pushrods in the annular space between SD and hull; and by reducing elements of the linkages you eliminate stiffness, non-linear response, and back-lash.

(This feature of the MSD – the placement of the two servos up front almost did not happen, and resulted only at the insistence of Bob Martin. He’s built up and got more r/c submarines in the water than anyone I know, so when he makes a ‘suggestion’ like this, I snap too and get to work making it a practical, user-friendly ‘thing’. I did, and was pleasantly surprised to find the space it saved in the after dry space as well as the simplification of linkages needed to animate things near the bow of the model submarine of great benefit. We’re never too old to learn new tricks … never equate age with wisdom!)



As it is magnetic influence that turns the mission-switch on and off I took care to place the device up high against the inside of the forward cylinder – this done by mounting it on a block atop the battery and securing everything together with a few wraps of Electrician’s tape. Now, without the need to flip a switch or access the forward bulkhead, I can simply turn the system on and off by waving a magnet over the MSD.

A neat feature to the magnetic mission-switch is it’s built in fuse: Should the current draw rise above 10 Amperes (there is a 25A version of the switch) it will open the circuit, shutting everything down. After a brief ‘off’ period it closes, restoring power to get the boat back.



It took considerable time, care, and … always in short supply … patience to reeve all the wires through the conduit. This is the forward section of 2.5” diameter Lexan cylinder that houses the battery, two servos (fair-water planes, and torpedo launch) and magnetically actuated mission-switch. The output wires from the mission switch, and the two servo leads run through the ballast tanks conduit. A tight fit!

 

aeroengineer1

CLEARANCE: Confidential
Joined
Jan 27, 2008
Messages
80
Reaction score
4
David is a very wonderful craftsman. I have been able to personally witness his work and learn from him. Salty at times for sure, but has always been willing to share his love of modeling with any that truly wants to learn.

Adam
 

merriman

David Douglass Merriman lll
Joined
Mar 18, 2013
Messages
336
Reaction score
207
David is a very wonderful craftsman. I have been able to personally witness his work and learn from him. Salty at times for sure, but has always been willing to share his love of modeling with any that truly wants to learn.

Adam
No one owns these techniques, they are only temporary custodians of knowledge and skills.

Adam, like all the greats, and soon to be greats, is a Student and always will be. The breadth of his technical know-how and achievements has been equaled by few. And he's still a young man. He will be a force to be reckoned with. Here's some shots of the boy when he understudied with me:





















 

merriman

David Douglass Merriman lll
Joined
Mar 18, 2013
Messages
336
Reaction score
207
Flu! For three weeks! I’m almost over it. However, in those few moments when I could drag my fat ass into the shop I was able to piddle away on three 1/96 scale submarine models representing ‘modern’ types. When completed and made operational they will, this October, join the rest of my sub fleet at this year’s big Fleet-Run event; a wonderful three days of constant scale (1/96 and 1/100 only) ship and submarine running at the City Lake, Rocky Mount, North Carolina.



The smallest of the three, next to the wall, is the USS ALBACORE, that model representing the ground-breaking non-combatant research submarine that investigated the virtues of the tear-drop shaped hull, various control surface arrangements, propeller and battery types, as well as the systems slatted for use in the next generation of attack and missile submarines. Middle, is a model representing a unit of the SKIPJACK class attack submarine, the first American submarines to capitalize on the ALBACORE findings. The model will be completed as the ill fated USS SCORPION. And foreground a model of a Soviet ALFA class. These were the fastest attack submarines ever made and were marvels of engineering and innovation.




The ALBACORE is assembled from a GRP, resin, and metal kit produced by Scott Terrey and myself; the SKIPJACK is assembled from a GRP, resin, and metal kit I produced back in the 80’s; and the ALFA GRP hull was purchased from The Scale Shipyard (an excellent GRP lay-up, by the way), and I produced all the masters, tools, and parts for the appendages.

These type subjects present few challenges to the model-builder other than the few deck fittings and their sail mounted retractable optical, air induction and electronic masts. So, that’s where a lot of my scratch-building skills are lavished, as demonstrated in this shot of the masts projecting from the tops of the ALBACORE, SCORPION, and ALFA models.

I’ve been working these three, off-and-on, the past two years, with a plan to get them to the stage where I could address them all, as a group, when it came time to paint, detail, mark, weather, and clear-coat.



It’s a good practice to pace your work sessions on a model – space out your flurry of tasks with a ‘cooling-off’ period. As it’s during those periods of rest that your mind fills with ever more objective thoughts of what-I-did-right-and-what-I-did-wrong, which often leads into a plan to undo or alter recently completed work. Such was the case with the little 1/96 ALBACORE model when I rejoined that project recently.

Initially I scribed in a longitudinally running line along each side of the hull. The trouble was instead of a line of slight, graceful curvature; it wound up being a little ‘wavy’ at points along its length. It was during the most recent hiatus (actually, sitting on the toilet … is there a linkage between anus and brain?) that I formulated a method of straightening that engraved line. As it turned out a simple solution in practice, but something that took me several weeks of casual thinking as I pondered how I was going to execute that fix. And it was so easy: tape down a piece of styrene sheet to the hull and re-scribe those portions of the engraved line that were out of true. The task took only minutes. A little touch-up putty over the ‘fixed’ areas of the engraved line, one final pass of the engraving tool, and the two engraved lines were straight and true.

Problem identification plus time and thought, stir in experience, mix well, and you eventually get the best solution to the problem.



The three models were pulled off the wall and surveyed for sloppy engraving, scratches and tool-marks that required touch-up putty and sanding. Case in point is the ALFA’s bridge closures, represented by these engravings: their outlines were too deep and wide.



So, I filled the engravings with putty, and while it was still wet I chased out the excess putty with a very narrow scribing tool. Note the use of masking tape to limit the spread of the putty to only those areas needing repair work.



Again, working the three boats at the same sitting, I wet-sanded back applied touch-up putty. Each boat was wiped down with a damp cloth and blow-dried with low pressure air. All scribed lines were lightly engraved to remove any sanding dust. Stiff sanding tools were made by taking a strip of .032” thick brass sheet and CA’ing a piece of #400 grit sandpaper to ones side, and #600 grit sandpaper to the other. These are perfect little sanding blocks for knocking down dried putty. I used tightly wound sandpaper twists to abrade into and over tight compound curves such as the ALFA condenser scoop intakes and stabilizer fillets.



Fine steel-wool is used to lightly abrade tight-radius areas of the model, such as the fillets between ALFA stabilizers and hull. Before using steel-wool on a models surface the steel-wool has to be rinsed in lacquer thinner to remove the preservation oils that impregnate it – that oil there to keep the fine steel fibers from rusting. Failure to degrease the steel-wool can result in poor coating adhesion.



Once all touch-up putty had been applied, sanded smooth, and engraved lines chased out with a scribe, the three hulls and they’re appendages were given spot-coats of primer on the worked areas.



The demarcation lines between stabilizer and hull fillets had to be enhanced after all that sanding and abrasion with steel-wool. So, to re-enhance the fillet plates, I masked the stabilizers and hull with adhesive tape and spray painted on two heavy coats of primer. Removing the mask revealed the desired raised edge of the fillet plates that fair the stabilizers to the hull.

Note the use of s ‘swivel-knife’ and plastic sheet stencil used to produce the uniquely shaped self-adhesive masks.



Special sanding tools were made and used to refine the shape of the many different sized and shapped limber, flood-drain, and vent holes along the length of the ALFA’s hull.



Scratch-building would be a joy if multi-view orthographic drawings of the subject are at hand. But, in the real world the smaller items of a subject presented as a plan are done so with few if any enlarged auxiliary views that would reveal the intricacies of these smaller items. So, by necessity, photographic interpretation is a big part of documentation enhancement that has to be done by the careful modeler. A valuable tool in your inventory of skills is the ability to generate your own ‘shop sketches’, based on study of still and video images of the prototype. Such renderings helping you identify the geometry of the object being rendered as a three-dimensional model.

Just such a drawing used to assist me as I turned a length of machine brass round-stock to form the ALFA’s only periscope – a big ugly thing definitely not intended for close-in attack duty. Most of this item was lathe turned, but the scope head itself had to be carefully carved shaped by hand using riffler files, knife, and moto-tool burrs.



Something the Russian submarine design bureaus did little about, until recent years, was the use of streamlined mast fairings to streamline the cylindrical retractable masts to the water flow. As a submarine travels at speed the drag of the water passing around the cylindrical mast will, at some critical speed, go turbulent and will, like clock-work, induce a vibration onto the mast. Mast vibration will interfere with the optics of a periscope, or the wave-guide efficiency of a high-frequency antenna system. That same mechanical vibration also presents a strain on mast seals and shears. The inevitable vibration problem addressed to some degree by the Russian practice of making their masts of (by American standard) very large diameter steel tube.

However, these simple mast cylinders are a god-send to the lazy model-builder, like me: a length of aluminum tube, cut to appropriate length, topped by a cast scope, snorkel induction, or antenna and I’m done!

Most Russian designed boats feature these simple tubular masts. Our Russian counterparts did not employ mast fairings till later versions of the KILO class boats came off the building ways. Nowaday’s Russian boats employ streamlined fairings on most of their retractable masts… welcome to the 21st century, Ivan!



Once the RenShape mast foundation piece is glued within the top of the sail the upper hull is positioned onto the bed of the drill press and the appropriately sized bit is used to drill a perfectly aligned hole through the foundation, forming a interference fit between it and the base of a mast.



I have yet to make the tool and cast parts for the optical, induction, and electronic items that top the masts that project atop the ALFA’s sail. What you are looking at are the brass masters that will eventually be used to make the rubber tools from which white-metal production parts will be cast.

What I’m handling in this shot is what I assume to be a combined snorkel induction head-valve-antenna mast – I have not been able to get a definitive answer as to the function of this mast, but that is my best guess. Anyway, it does demonstrate how the masts make a friction fit to the internal mast foundation piece.



An obvious cheat on my part is the non-scale representation of extended masts through closed fairing hatches. Though I’ve produced parts representing opened hatches I elected to simply poke holes through the centers of the engraved representation of those hatches through which the masts passed. Most of the time the operational model will only have the scope in place, the other masts left back on the table as I drive the model – the open holes serving a practical function: venting the free-flooding hull as the model submarine submerges and surfaces. All masts are in place only for display out of the water.



Presented here are the ALFA, ALBACORE, and SCORPION propellers at various states of finish. From left to right: the ALFA propeller demonstrates the initial work needed to take a cast white-metal propeller from the raw state to a point where it’s ready for pickling. Here I’m filing back the inevitable flash imparted to the casting from the tool.

The two-propeller ALBACORE propulsor featured two concentric shafts, each swinging a counter-rotating propeller – identifying this as a phase-4 arrangement. The filing, pickling, and initial primer work has been done.

The completed and painted SCORPION wheel shows the end-game of propeller manufacture.



Many non-ferrous metals tend to shrug off primer and paint if not first oxidized to pit the surface of the metal in order to secure a tight mechanical bond between surface and coating. The white-meal, (Tin-Antimony alloy) part is soaked in acid to oxidize its surface. The oxidation creates zillions of little pits that key with the coating applied over it. This ‘pickling’ process is simple: dunk the work in acid and while its immersed brush the surface of the part vigorously with (duh!) an acid-brush to insure complete oxidation of the parts surface, rinse in fresh water to get the pH back to normal, dry, and prime.



Here, demonstrating the difference between a raw whit-metal casting and a worked and oxidized propeller. Once so pickled, the substrate is most receptive to the primer, producing a tight bond between metal and coating.

 

merriman

David Douglass Merriman lll
Joined
Mar 18, 2013
Messages
336
Reaction score
207
Last installment I presented the pretty side of model-building; kit assembly and the joyous and challenging tasks of scribing, detailing, and part integration.

Well, this isn’t that!

This installment I present to you the dark underbelly of professional model-making: dumb-ass, boring, repetitive, no-talent, soul-crushing production work.

Exhibit-A: casting bulkheads and unions for our current line of Modular SubDrivers (MSD); a look at how raw resin parts are cast and machined into useful components that fit together into a rational system.

Pictured here is about two-gallon worth of polyurethane casting resin in the solid state, fresh out of the rubber tools that gave them form. Note the attached sprues and risers hanging off of the castings.

Casting resin parts goes like this: Rubber tools (molds), usually of the two-part type, are prepared; polyurethane resin is mixed with its hardener (catalyst), and quickly poured into the tool; the tool is placed in a pressure pot and subjected to at least one-atmosphere of pressure (14 psig) which is maintained until the liquid polymerizes, i.e. changes state to a solid; the tool is removed from the pot and opened up, and the part popped out; repeat till you crave death!



I hate production work!

Most of the MSD tooling is of the two-part type, the halves held together with rubber bands backed up by strongbacks – typically plywood or chip-board ‘shelving’ sock. But the tools are not assembled until their cavities and flange faces have been treated with a silicon part-release spray and a good dusting of talc or corn starch.



My initial tooling for the MSD work was formed from masters that – as we found out after the first MSD’s were being evaluated – featured ‘stops’ that were too thin and, on the cast resin parts, would break easily when pried against, such as when attempting to pry a cylinder loose during disassembly. Instead of wasting the initial set of tooling after only a few shots (RTV rubber is expensive!)I elected to modify some of them by cutting away portions of the rubber, this resulting in castings with fatter stops. A crude solution, but one that worked well enough to justify the extra machining needed to true up the fatter stops.



To slit the three different diameters of tools, I made three specific slitting tools; one for the 2.5” diameter union tools, one for the 3” diameter union tools, and another for the 3.75” diameter union tools. The work went surprisingly fast.



The operation was simple enough: hold the semi-circular edge of the slitting tool in the existing grove of the tool, and rotate. Follow that with a 90-degree slit from the top with a hand-held X-Acto knife, and I’m done.

Of note here is where the blade of this slitting tool is projecting – this cavity is both sprue and riser; it’s where the liquid resin is introduced into the molds cavity and were make-up resin comes from to make up any liquid lost as air-bubbles within the mix are crushed into solution during pressurization.



I’ve poured several parts from these modified tools and I had no problem smoothing out the rough ‘stops’ on the lathe while turning the radial flanges to form tight fits to the Lexan cylinders they support. This is a stop-gap solution though; I’m re-working the masters with fattened stops and will produce ‘production tools’ that will incorporate that and other changes – to correct problems identified in the initial batch of MSD resin parts.



Tool preparation starts with a heavy spray coating of part-release. This silicon oil forms a barrier between the tools rubber and the polyurethane casting resin. I go through the Mann 200 part-release by the case! Good stuff.

Note the brass inserts in some of the tool halves. These ‘cores’ suspend an o-ring that will be encapsulated in the unions; that o-ring making a watertight seal between the unions internal bulkhead and brass tube conduit that runs the length of the ballast tank. After a casting is made, the tool is opened up and the part extracted, I then yank the brass core out which leaves the conduit bore and o-ring that slightly projects into that bore. Slick! I learned this trick by hanging around the rubber and mold shop aboard the USS YOSEMITY back when I was a snot-nosed Diver.



To enhance the ability of the liquid casting resin to fill all voids within the tool a thick coating of talc or corn starch is applied within the tool cavities – the excess powder shaken out onto the floor before the tools are assembled. The powder coating held in place by the sticky part-release previously applied. The powder works to further isolate the rubber from the crazing effect of the casting resin, extending tool life. The powder also works to wick resin into the tight areas of the cavity, contributing to a better fill.

Nowadays I can expect at least eighty cycles from a tool. Forty years ago I was lucky to get twenty.

The marvels of modern chemistry.



Artifacts of the resin casting process are the sprue and riser elements of the casting. These are the result of the cavities that introduce the resin to the tools cavities, permit the escape of displaced air, and provide make-up resin as entrapped air not vented away is crushed into solution during pressurization. These appendages to the casting proper have to be snipped or sawed off the resin part before any serious machining can begin.



Some castings after clean-up and sizing are further worked with the installation of Oilite bearings, such as these ‘gear-splitters’ used to produce two counter-rotating shaft outputs from a single shaft input. Surprisingly CA adhesive works to permanently glue these oily flanged shaft bearings in place.



Exemplified here a motor-bulkhead casting has not only been equipped with a gear-splitter unit, but also shows off the post casting insertion of three watertight seals that pass control surface pushrods. You can just make out support studs, spider motor-mount, and out-runner type brushless motor that outputs into the gear-splitter through its own watertight seal.

Casting MSD parts is only the start. There is much machining and part integration needed to make these parts useful elements of the system.



As commercially available Lexan cylinder has a wide variance of diameters (very, very sloppy industry tolerance) -- I don’t know from buy to buy just what actual diameter I have in the racks -- I produce my bulkhead and union masters substantially over-sized in diameter. This not only accounts for the inevitable shrinkage of the room temperature vulcanizing (RTV) tool rubber and shrinkage of resin as it changes state from liquid to solid, it also assures that no matter how out of specification my Lexan cylinder is, I would be able to reduce the diameter of the cast resin bulkhead or union to fit to the cylinder at hand. One size DOES NOT fit all!



In foreground is a raw cast resin union, mounted on a lathe holding fixture ready to be machined to the desired diameter. I’m holding a machined example of the same item, ready for insertion into the cylinder it has been sized to fit. The terribly flawed stop you see here, butted against the face of the holding fixture, is a consequence of the somewhat less than perfect slitting job I did on its tool to fatten up the stop in the eventual casting. Once faced and shaved to correct diameter the casting will be just like the one in hand.



Shaving the outside diameter of the radial flange; as well as facing the god-awful ragged ‘stop’, is a simple lathe job. However, a big side-step away from ‘shop safety’ is the use of a hand-held cutting tool to gouge out the two o-ring grooves. The depth of the groove set to be about two-thirds the wall thickness of the o-ring used to make watertight the union between ah … the union and Lexan cylinder.

Note how I use (most improperly) the lathes cross-slide mounted cutting bit as a tool-rest as I fine-tune, by hand, the depth of the o-ring grooves in this casting. Very bad shop practice to bare-fist a tool like this. No problem. It’s my frig’n shop, and I’ll do what I bloody well please, thank you very much!

OSHA and the other Federal and local Regulatory agencies can kiss my nonunion ass!



The o-ring groove cutting tool is simple enough, a tool-steel blank is mounted in an acrylic handle and its cutting edge ground and honed, and a modified wheel-collar used to act as a stop that insures the depth of o-ring groove cutting stops at the correct depth.



A little math and I set the distance of the wheel-collar stop to the tip of the tool using the appropriate leafs of a feeler-gauge.



The stop is secured to the tool with a set-screw.



A cheat I employ if I find that the o-ring is either too tight or loose when the cylinder is installed over the union or bulkhead radial flange is to go to an under or over-sized o-ring. The end-game is to get a good mashing of the o-ring between flange and cylinder, easy to see through the transparent Lexan.

Torpedoman A-School was not wasted on me!

 
Top