Thread Tools
Oct 09, 2020, 09:40 AM
Registered User
Thread OP
Build Log

A Trumpeter Bismarck 1/200


A Trumpeter Bismark 1/200

Introduction

I have long had an interest in the Bismarck as a great example of German engineering and decided to build an RC model using the Trumpeter 1/200 kit. I read up as many posts as I could find and in particular was inspired by “capricorn” with his Arduino control of turrets as well as other aspects.
The wealth of experience in this group and elsewhere gave me the confidence to proceed, this being the first powered RC model I have built.
By way of background I am a retired mechanical engineer whose hobbies include scale model engines and the restoration of old cars. I am fortunate to have a basement workshop with a lathe and mill as well as electronic tools that make me fairly self sufficient.
This build log starts by reporting on the various choices I made for the machinery, based heavily on all the experience that I have read about and then with my own spin on it. I am always interested in analyzing the physics of my models so you will see this in my comments below.
The first object is to get the ship powered and working, then build the superstructure, then work on the turret mechanisms.
I expect to use upgrade kit(s) with brass fittings to make a more robust model than the original kit and I will be seeking advice in due course on the best way to do this.
To date the vessel hull has been tested in a swimming pool and runs fine with lots of power. The turning circle going forward is quite large and in reverse, the rudders are ineffective.

MODEL SPEED
The maximum model speed should I suppose lie somewhere between the linear scale speed and the percentage of hull speed at full scale. Running the model at the same percentage of hull speed will give the same wave pattern as the full scale. The full scale speed is 30Kts (50ft/sec) which is about 80% of hull speed. 80% of model hull speed is 3.6 ft/sec so this will be the upper speed limit and design goal.

MOTORS
Brushed motors seemed to be the right choice and I like large motors that turn relatively slowly for reliability. My choice is MFA RE385LN whose diameter allows them to be mounted on one bracket and spaced the same as the prop shafts. A low noise version was chosen because “why not?”. at 12V, the max efficiency speed is 5950 RPM at an output of 6.7W each or 20W total.
The full size ship has a max power of 110MW so scaling this by the cube of 200 results in model power of 13.75W. Of course efficiency plays a great part in all this but it is an interesting comparison anyway.
The pictures below show the three motors fitted to a common bracket that is screwed to the hull using polystyrene bearers with #4-40 threaded inserts for ease of servicing.
The bracket was angled and mounted sufficiently far forward so that the motor axes were lined up closely with the prop shaft axes, thus allowing the use of a simple couplings that did not have to accommodate much angular travel and yet could allow axial play so that the motors would not have ant axial force applied.
The motors have 0.1mF suppression capacitors added.

PROPELLERS
Raboesch propellers (162-13 & 162 – 14) were chosen and while the scale size would be 25mm diameter, I went for 30mm just to have the best performance capability
The model prop pitch is 26.7mm so if there were no slip, at 1000 RPM, the boat speed would be 1.46 ft/sec. Let's assume a propeller slip of 50% so we should expect 0.73 ft/sec per 1000 RPM or 4900 RPM for 3.6 ft/sec which is a good speed for the motors chosen.
The props have right hand fixing threads which means that one of them has to run so as to try and unscrew from the shaft. This is the starboard prop and is secured with Loctite 242. The model is designed so that the prop shafts can be easily removed except that the center prop will interfere with the rudders so it must be unscrewed before the shaft can be withdrawn

PROPELLER SHAFTS
I wanted to make maintenance free shafting without the need for periodic oiling.
The maximum diameter of the prop shaft tubes is 5mm to fit reasonably in the model.. The prop end bearing is Delrin, water lubricated and the fore end has a sealed ball bearing to act as the thrust point and a stuffing box. The shaft diameter is 3/32” and has an adapter to 4mm thread brazed on to match the prop threads. The shafts are secured with a set screw in front of the ball bearing that allows easy removal of the shafts. At the rear, a hole is drilled in each tube to allow water into the bearing area.
Originally, I was using a 2mm shaft but my tests showed a 10” length whirled at about 6300 RPM so I increased the diameter to 3/32 to give a higher whirl speed. The bearings were originally held with an internal retainer ring but I found the bearing outer races would rotate due to the drag torque of the seals. They are now Loctited in and hopefully will not need to be changed. I did not aim for an interference fit since this was hard to achieve and would hinder removal should this be necessary. However this is now moot.
Machining the Delrin bearings was a challenge since this material is not rigid and moves away from the cutting tool. 3/16” Delrin was used and a 0.93” hole drilled after which the drill bit was turned round and acted as a steady so that the outer diameter could be turned to get a snug fit in the outer tube – see picture.
The rear struts are brass and are soldered to the outer tubes.
A sturdy bulkhead provided prop shaft location and also is used to mount a ballast weight of 2 lbs.
The shafts were epoxied in using JB Weld ( because I like it) but unfortunately there was leakage so silicon sealant was applied as well.

COUPLING SHAFTS
These connect the motors to the prop shafts and originally I tried some U-joints but these did provide adequate radial location resulting in a lot of vibration. After a light bulb moment I realized I could simply make end couplings that would have slight radial clearance and have slots to engage set screws to transmit torque. These end couplings have an 0.086” hole and are pushed onto 3/32” bar. They run very nicely. Each one has a piece of reflective tape so that the RPM can be measured with an IR tacho.

WATERLINE LOCATION
Scaling from the original, the draft on the model would be between 1.8 and 1.9”. Scaling from the drawing in the kit suggests 1.7” so that is what I used.

RUDDERS
It seems to be current wisdom that models need larger rudders than scale. I made rudders that extend down almost to the plane of the hull bottom which added about 30% area. They were made of brass shaped to a profile and had 3/32 brass shafts fitted.
The rudder bearing housings were made from 1/2” polystyrene with a central core of 1/4” Delrin drilled 3/32”. The tops of the rudder housings are located in a transverse beam to provide strength against any external load being applied to the rudder. They extend as far above the water line as possible.
After fitting I found that the shaft axes are not supposed to be vertical but inclined inwards. I just bent the shafts so that the rudder angle was correct in the straight ahead position.
Brass links are fixed with #2-56 set screws to the rudder shafts.
Initially the rudders swing +/- 35Ί but I will try greater since the turning circle is quite large.


BATTERY & POWER SYSTEM

SLA for weight and convenient size and simplicity of use. The largest battery that would fit is a Power Sonic PS-1238 of 3800mAH capacity. It seems from first tests that the current draw will be 1.5A maximum so, from the performance curves I should get long hours of operation.
As a result of the first tests which showed the boat to be overpowered, a 6V regulator is fitted in the battery line. This means that the battery can run down to at least 8V without affecting the performance.
The battery data show that I can expect at least one hour operation at full power and realistically several hours at normal speeds will be available.

The speed control is a 3 channel HYDRA from Robot Power Inc. who provided very helpful advice. This is a high frequency switcher so there is no associated motor noise and it has plenty of capability to run the three motors as well as having a BEC.

WEIGHT

3.5 lb battery and 2 lb of brass at the rear bulkhead gives a total weight of the hull as 9.4 lbs. The rear weight also acts as a heat sink for the 6V regulator.


INITIAL TEST RESULTS

An IR tacho was used to measure the prop shaft RPM's.
In air, the motor speeds versus drive voltage were measured and were linear up to 6500 RPM at 12.5V.
Holding the boat in water and repeating some test voltages showed almost no effect on motor speed of the water loading. Dry runs show that there is a significant drag from the bearing seals. For example at 6V drive the free motor current is 100mA but when connected to the bearings, this typically rises to over 300mA.
Several water runs were made at different motor voltages and the travel was videoed at 15 frames per second. By looking at the individual frames the time of travel for one boat length was estimated., hence the boat speed. As we know the RPM from the tests above we can calculate the propeller slip percnt.

The test results are shown in the picture below

You can see how the slip increases as the speed goes up, which is not surprising. It is gratifying to see reasonable numbers from this rather approximate test. Up to the hull speed of 3.6 ft/sec, the speed is essentially proportional to prop speed - 1.2 ft/sec per 1000RPM.which makes sense. Above that there is the effect of drag rise and probably cavitation as well.

I have decided to fit a voltage regulator in series with the battery to limit the maximum speed to about 4ft/sec. With a suitable transmitter I could set the travel but mine is but an elementary one.

For all tests the average speed of the three shafts and motor voltages was used.

SECOND TESTS

These runs were made with a series resistor in the battery line so that 6V would be the maximum applied to the motors.

The test results are shown in the picture below

I also did a bollard style test to see how much thrust with no boat speed. Used a small dynamometer seen in the Test Equipment picture below.

At an average prop speed of 2500RPM, the boat pull was about 0.26 lbs. - see picture.

The draft was about 1.6”.

DESIGN COMMENTS

Clearly i have designed a more powerful system than I really needed. The incremental motor and propeller costs are really not so much and I do have the benefit of a pretty relaxed drive train running well with its capabilities.
I do wish the rudder performance was better but there does not seem to be much to be done here.

CONTINUATION

1) Painting the hull and building the kit. No reason to blog this since so may have done this already. I have been advised to use brass fittings for the superstructure to withstand the rigors of playing with model boats. I would like comments on peoples' experiences with the various upgrade kits available.

2) Set up the turret and gun elevation control. This will need lots of design help from other users including capricorn.
Sign up now
to remove ads between posts
Oct 09, 2020, 02:22 PM
Registered User
Welcome BESM, Nice work. Am looking forward to seeing it go together

I think I can assume the main deck will be fixed with access through the deck openings only? Operating elevation of the guns in those small turrets will be a trick, interested to hear how that will be done.

You came to the right place, there are many extremely skilled model builders here who can provide much helpful info. Thanks for posting this. Cap
Oct 18, 2020, 10:17 PM
Registered User
Any progress on this? I was interested if it will have the animated turrets, and how that works. I recall this ship has three main gun directors as opposed to the usual single director, does anyone know how that system functioned? How was the pointing info from the director communicated to the turrets: gears and shafts, chains and sprockets, electrically, radio? Cap
Oct 21, 2020, 03:14 AM
Registered User
Quote:
The props have right hand fixing threads which means that one of them has to run so as to try and unscrew from the shaft. This is the starboard prop and is secured with Loctite 242. The model is designed so that the prop shafts can be easily removed except that the center prop will interfere with the rudders so it must be unscrewed before the shaft can be withdrawn
Loctite them all. A prop that unscews itself when reverse is used is annoying.
Oct 21, 2020, 06:44 AM
Registered User
Welcome.

It is good to see test data but I am confused by your test results. As I understand it the only difference between test 1 and test 2 was to reduce the voltage via a resistor from max 12v to max 6v. Looking at the results for 6v (available in both tests).

The results for Test 1 on 6.0v show 3120rpm for 3.94 knots and a slippage rate of 14%
The results for Test 2 on 5.8v show 3193rpm for 3.23 knots and a slippage rate of 31%

How can slightly less voltage in test 2 result in more rpm but a nearly 20% reduction in speed and a more than doubled slippage rate when neither the prop nor the model changed?
Last edited by ChrisE; Oct 21, 2020 at 07:14 AM.
Oct 22, 2020, 01:59 PM
Registered User
Thread OP
Hello ChrisE

Thank you for reviewing the test results and your comments and questions. I ran two sets of runs, the first being rather quick and dirty just to get the measure of the beast and a second to obtain reasonably accurate readings.

The "RPM per Volt" readings are fairly consistent between the tests but the "boat speed per volt" are not and these first test results should be ignored: I shouldn't have included them

The first tests were done using a low res phone camera to measure the speed. The speed readings obtained on the first tests are thus rather approximate since the picture quality was bad, fuzzy images and with a low frame rate.
The second test readings used a camera on a tripod and are reasonably accurate I think.

You are correct that in the second tests I had a resistor in the supply line to limit the maximum speed. After the test, I then fitted a 6.9V voltage regulator in place of the resistor since sometimes during the test the voltage became too low for the BCE to drive the receiver.

I have found that the major torque loading is due to the drag of the bearing seals and this will change with time, as I have measured. Water loading seems to have a relatively small impact on the prop speeds, only about 3%. As a general point, at 6V drive each motor alone drew about 100mA but when connected to the bearings the current draw increased to 300mA or thereabouts. Adding the prop shafts did not increase the current draw to any extent.
The speed readings previously shown were the average of all three shafts at a particular supply voltage.
In view of your comments, I have therefore expanded the test results and these are attached. I did not factor in the assumed speed drop due to water loading.
You will see some numerical inconsistencies between the runs, nevertheless I think they give a fair idea what's going on.
I need to devote my time to getting the Arduino turret system understood built and working.

The bearings I used were very cheap and i wondered how well they would seal so I have now fitted "high quality" bearings at $4 ea!

The takeaway is that at 6 V one can expect about 3200RPM and a prop slip of about 30%

The attached pictures may clarify the situation but I am certainly happy to discuss them further.

No more tests until the spring - the pools are closed.

Best regards


Bruce
Dec 12, 2020, 02:44 AM
Registered User
Thread OP

Second Installment


This second installment covers the preparation of the main deck for the installation of the turret stepper motors, and the magnetic clamping of the removable superstructure together with measures to impede the entry of water.
The design of the turret motors is shown and the way that they are removable if repair is ever needed.
Finally, the Arduino coding for the turret aiming system is described.
I claim no originality for any of the above, just my own interpretation of what others have done.
“Capricorn” provided the ideas for the turret aiming concept and coding which I much appreciate, my being a newbie to Arduino .

The hull is in process of painting and this has been a difficult process. Mainly I had wrinkling due to second coating being applied before the first was completely dry. However things are looking good at the moment. I copied the way Ron Calverly (on YouTube) supported the hull while painting.
Picture below before the grey paint was applied.
Finally ended up using Humbrol Crimson Enamel #20 and Dark Admiralty Grey #5. Not sure about the black stripe at the waterline. The original did not appear to have this and I have had enough trouble painting already. If I do add the black, I will probably use auto pin striping tape.

MAIN DECK
The picture below shows the things I did to the main deck.
This was modified so that the Anton and Dora turrets could be fitted after the deck was glued to the hull. Rare earth magnets 0.32” dia, 0.25” long were fixed to the underside of the superstructure allowing its easy removal. The magnet targets are #10-24 flathead screws threaded into blocks for ease of adjustment and fitted with nuts to lock them in position. It was hard to fix the magnets to the plastic and so they are held with straps of 0.020” styrene instead.
Although the magnets provide a good closing force, it is quite possible that deck water could migrate into the hull so I added an extra barrier beyond the basic surround. The barrier is made of 0.020” styrene about 1/2” high.

Wire clips are fitted to guide the wiring and and neaten the under-deck layout.

TURRET MOTORS
These are the canonical 28BYJ-48's, converted to bi-polar operation so that I could use A4988 drivers fitted to a Protoneer CNC shield on the Uno. This makes for a fairly compact arrangement.
The motor spindles have a brass extension with an O-ring that provides friction drive to a plastic cylinder that is glued to the turret base. This provides slippage in case of a program failure that causes structural impact
I decided that I wanted to have absolute positioning of the steppers from start up, so turrets Anton and Dora have Hall sensors that are used for the homing as does the Director. The only manual adjustment required is to align the front and rear turrets to their respective partners since Bruno and Casar do not have homing.
The director stepper is mounted on a square plate that is screwed into bosses on the forward superstructure. It has a homing magnet on a brass arm and its shaft is connected via a 3/16” styrene tube to the director, topside. A 1/4” hole is drilled through the superstructure for the tube passage.
Some pictures below give a general idea.

COMPASS
The compass is a Pololu LSM303 module and will lie on the small platform at the bottom of the bow. It will be fitted to a small plate on the end of a styrene tube. This will be inserted after the deck is fitted and fixed with two screws to the frame pieces on the deck.

TURRET AIMING SYSTEM
This is based on Joe Cain's work and functions similarly although my code is based on AccelStepper.h
It seems to be functioning OK. One nice feature of AccelStepper is that it allows accelerations as well as steady speed rotations. I also liked it easy position control.
The basic operation is as follows:-
1) Turn on and reset Arduino
2) All steppers will then home to their neutral positions.
3) The director will be commanded to an angle with ref to the ship axis, this command being the difference between the desired target heading and the ship compass. The desired heading command comes from the R/C receiver thus allowing vectoring as desired.
As the ship turns the target heading is maintained.
4) The director angle is then copied by the turret pairs in turn, except for the 90 degree arc which would cause structural interference.

A picture of the breadboard is below and the only difference now is that the slider pot has been changed for the R/C receiver. I must here also acknowledge Ryan Boland who wrote the code for using the R/C receiver output in place of the slider pot. I don't really understand some of the subtleties in the interrupt routines which may be evident in my coding but thank goodness it works.

A video of the aiming system test is shown on YouTube at
Gun Turret Control for 1/200 scale Bismarck Battleship (3 min 25 sec)


I plan to add an R/C controlled on/off for the turret aiming and perhaps speed things up by using faster turret rotations that occur in sequence and not concurrently.

Rather than take up space here, I can send anyone who wishes it a copy of the code listing, via PM.

NEXT STAGES

Finish painting hull
Fit rudder system
Install main deck on hull
Install propeller shafts
Fit interior wiring and all components
Run functional tests
Install wooden decking from Scaledecks.com :-)
Dec 14, 2020, 07:25 AM
Registered User
Bruce, that's pretty slick. I'm surprised there isn't more interest in it, maybe a bit too technical, I guess the stepper motors and arduino are a bit ominous for most.

Looking forward to seeing more. Joe
Dec 14, 2020, 01:47 PM
Big Boats Rule!
boater_dave's Avatar
Watching with great interest, as this was how I envision controlling the turrets on my Fletcher. The great debate between steppers and servos continues, as both have certain advantages and disadvantages. Motivation to spend the winter getting my destroyer on the workbench!

Your work is fabulous, by the way. I'm sure your model will be the center of attention at any club meet you attend.


Dave
Dec 15, 2020, 02:11 AM
Registered User
maxthepiper's Avatar
Fascinating

That turret aiming system is exactly what I'm looking for my SMS Brandenburg.
Keep up the good work

Cheers, Max
Dec 26, 2020, 01:17 PM
Registered User
Thread OP

Third Installment


This third installment describes the fitting of the main deck and installation of the electro-mechanical parts. The R/C transmitter was also modified to make it more user-friendly for model boating.

HULL WORK
The hull exterior and interior was painted prior to fitting the main deck. I did not put the black line around the hull at the waterline since I had had enough of painting besides which the launch picture in 1939 seemed to show that such a line was not applied, see below.

Taking Ron Calverly's approach shown on YouTube, I used a thick CA glue that takes several minutes to set up allowing plenty of time to run a thin bead around the whole glue line. Some clamping was needed to maintain snug fitting.

WIRING
As mentioned earlier the use of a CNC shield on the Arduino kept the wiring compact. Connectors were constructed for ease of wiring to the motors fitted to the superstructure. These were made from pin headers soldered onto perf board and screwed into mounting locations on the hull.

COMPASS
The compass board was attached to a triangular plate that nestled at the bow, resting on the convenient platform there. A plastic tube provided wire guidance and allowed the compass to be taken in or out if needed.

WOODEN DECK
Wooden deck pieces from scaledecks.com were glued using 3M spray adhesive 77. This is a great produce and has fast adhesion at the same time as allowing relocation if needed. The precision of the Scale Decks product was such that all the parts went down in exactly the right place with no need to lift and reapply. Prior to fitting these pieces, they were coated with a matt finish varnish for staining protection.

TRANSMITTER
I find that joysticks are unfriendly to use for model boats. They are inappropriate for the sort of low speed control of ships and it is very easy to apply inadvertent control inputs. I therefore changed out the gimbals on my DX5E and fitted two knobs and two slider controls instead. One knob is for the helm and the other is for the gun director steering. One slider is the engine throttle and the second is unused for now until I come up with another feature to add.
Dec 26, 2020, 05:37 PM
Registered User
Wonderful. I have seen a similar system shown on Model Mayhem.
One question, When your director is aligned with the bow, 0 degrees, without command input from radio, does the director and the turrets remain so aligned?
Dec 27, 2020, 01:37 AM
Registered User
Nice build. Just feel like the prop shaft seals a bit of an overkill. Would have gone with a simple stuffing tube with simple bearings at the end and an oiling tube.
Dec 27, 2020, 09:03 AM
Registered User
Thread OP
Jeno

The radio link to the director is always turned on. When the knob is rotated fully CCW, the program loop stops and the director and turrets assume a neutral position with respect to the ship. A small CW rotation then starts the target aiming. Sort of like a volume control switch on a radio.
This means that
Dec 27, 2020, 09:08 AM
Registered User
Thread OP
Jeno
Sorry was too quick with the send button.
Meant to add that the turn off stops the guns being aimed when in port.


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Build Log Trumpeter 1/200 Bismarck RC with Arduino HeavyMetal7734 Scale Boats 32 Nov 19, 2019 08:35 AM
Discussion trumpeter 1/200 H.M.S. HOOD from Deans Marine rondean1 Scale Boats 15 Aug 22, 2016 08:52 PM
Discussion Trumpeter 1/200 Scale HMS Hood laser110 Scale Boats 9 Jun 12, 2016 04:47 PM
Discussion Trumpeter 1:200 Bismarck patmat2350 Scale Boats 57 May 31, 2016 09:04 AM
Discussion 1:200 HMS Hood from Trumpeter patmat2350 Scale Boats 2 Dec 02, 2015 02:18 PM