Carina by Hellmut1956 - scale 1:20 - long keel yacht around 1900

[Hellmut1956]

Dear friends

 

I have started to build a sail ship model based on a french plan for which it was impossible to get additional information. The reason was that I lost my job in 2000 when the internet hype broke and never ever since than been employed again. Since than I have worked on all kind of issues and after years my old degree as a technical translator has proven to be the best source of income until in April 2012 I was hit by a stroke which has degraded some how my abilities. But this project has proven over the years to be a source for having every day an agenda that has no empty space and have my life organized, my mood up and my grey cells active  I am telling you this, for one reason because this might help others to find in this hobby a way to stay on top in hard times, but also because this explains my main objective. I am not building this model to finish it, but to have a platform that keeps me studying, give a red line to keep touching fascinating areas of competence into which to dwell and stay modest and try to help anyone that might benefit from my knowledge, as I have benefitted from many who helped around the globe.

 

My objective is to build a sailship modell able to navigate along a regatta course autonomously.

 

Lets start with the description of the Carina. A project that started in 2008 as a "light version of the Sabrina", the first hull that I started in 2002. The Sabrina proved to be too demanding to start and so, because my son Andreas had to engage into a 1 year project for the school, he decided to do this with his father as Mentor and I hoped that the love for modelism would jump to the next generation, in which I did not succeed.

 

 

This plan, a bit in bad shape because I did not take it into account to care more, was all I had to start with. As a consequence I decided to make the first step to digitize the content of the plan. Goal was to correct errors in the original plan that I found during the construction of the hull for the Sabrina.

 

 

So Andreas started by pasing the image of the views of the hull onto milimeter paper by oiling the paper of the plan and pinching with needles to pass the lines onto the the paper with 1 mm squares on it, 3/64", sorry we are metric outside the USA.

 

 

Then we passed the digitized values into a CAD SW. We did digitized and passed the following points of each frame along the longitudinal axis of the view of the hull. As a base line we took the waterline, assigning it the value "0" and moved upwards in steps of 20 mm or 60/64" documenting it by increasing the digit next to the horizontal line. So frame 3 is already above the waterline. Traditionally we recorded the deck line to document where the frame changes from the hull to the deck.

 

We than passed the values into an excel sheet to get visually the information where the digitized values were wrong by one or more reasons. The aim is to get continues curves. Today I would do it differently.

 

After making sure the smooth curves in the CAD images of the frames did result in smooth curves we did print the pictures of the frames onto a thicker paper, about 160 gr paper, 5,65 oz, to prevent the glue with which we fixed them onto the wood, would modify the shape by it making it wet.

 

 

The would choosen is a kind of triplex which is used to form the boxes were concrete is put into in construction. 5 mm thick, or 45/64". The reason is the wood is very brittle and so it can be removed easier later. I choose the method of putting the frames onto a table, upside-down in a way the waterline of all frames was in one level. You can see the tabs at the edges and how they were mounted on the table in the folowing picture that shows how I am removing the finished hull from the table.

[Hellmut1956]

 

 

 

The hull was made out of wooden laths of pine of 3x10 mm, 9/64" by 30/64". Those dimensions have proven to result in a hull that needs no smoothing stuff and as a result allows to keep the wood visible. The work is much less complicated than it is usually assumed. As you can appreciate from the foto, a hull with a long keel results in a beautiful hull.

 

Of course to get to this result heavy work sanding is required. I remember and still make jokes about this to my son, how often he claimed that nor it was fine! I have kept him sanding the hull one side at a time. When he claimed a side was fine I asked him to check against the other side and so we kept sanding for 2 full weeks! may be that is why I could not win him for the hobby. But he was very proud of the result.

 

Next a few pictures that show the hull and its frames from different perspectives:

 

 

Here a view from within the hull with the frames still covered with the paper from the printing out of the CAD SW.

 

 

Here a view from the top of the hull with all the tags still on the frames.

 

I then decided to have all the frames removed by just taking a sete of  pliers and twisting the wood  to have it break thanks to its brittle tendency. As you can see the wood of the frames could be removed to close to none residuals left inside the hull:

 

 

You can also see in the bottom of the hull the lead I just poured into it melted. If you begin carefully just putting little amount of molted lead into the hull, the wood does not suffer from the heat. The more lead is already in it takes the heat from the lead and cools it. This way you can be adding more lead at a time.

 

Here a picture of the complete hull. I love the sight!

 

 

 

 

[Hellmut1956]

To protect the would from water y apply a polyurethane coating with a 3:1 mixture with the appropriate dissolvent to have it penetrate the wood deeply and i apply this 3-4 times, sanding it carefully in between. The result is an hardened wood that water cannot impact. This I do consider important as it is always possible that a later coating suffers by some mechanical impact making it possible for water to get below the coating. here one more picture of the complete hull:

 

 

 

 

Here you can see my son Andreas applying fiberglass and epoxy to the hull internally. It is important to laminate symmetrically from within and outside, to ensure the tendency of the wood to "work" and make the structure visible later, even having it sanded to hide all.

 

At this point we decided that the complete deck of the sailboat would be removable. As a sailboat, by definition, will have water flowing over the deck, preventing water from penetrating the hull is key. by laminating it, it was made sure that no water would come into the hull through the hull. For the removable deck i decided to have us build a labyrinth seal additionally to other methods we would be applying. The next picture shows how we put a "U" form aluminium profile, open to the top all around the hull:

 

 

On this photo you can also see how we build a structure to connect the shrouds to the hull. A large aluminium piece with "L" profile was glued on top of a flat 5 mm strong aluminum piece in which threaded holes were cut, M3. by combing this 2 aluminium profiles i did achieve 2 things. One, the forces from the shrouds, which would come from the mast and the sails would be introduced into the hull over a large surface enabling an otherwise eggshell thick hull to take large forces. On the other side i could screw the aluminum strip, that defines the position of the mast and holds it, were it goes through the deck and being able to lead those forces into the hull by the same principle.

 

water that might pass the sealing I will still provide would get into the aluminium "U" profile and collected there. At the rear you can see the large servo to control the rudder.

 

Next a later picture of the hull. The deck of this sail ship has 4 constructions, each of the will also be removable and as a consequence the labyrinth sealing also takes care about them, surrounding the locations of those.

 

 

You can see one of those removable deck construction in place.

 

Next i started to work on constructing the removable Deck. 

 

 

As you can appreciate at this phase, the deck construction is very filigree. Remember, over a length of 165 cm, 5.42' the "U" profile of the deck that just fits into the "U" profile of the hull opened to the top, requires extreme precision! Also the deck line is not flat. It is higher at the rear, and the front and follows a smooth curve along the length of the hull.

 

[Hellmut1956]

Above you see the front of the deck, at the bottom, the rear.

 

Here a detail of the front:

 

 

You can see the cross profile and the end of the 2 lateral length profiles are prepared to be fixed but not yet screwed together. I used the the method to screw all parts together, as neither welding nor gluing did work for me.

 

You can also see the 2 lateral elements for the front most deck construction and you can see the cross aluminium bands used to fix the elements. Those will be applied holes to much later to reduce the weight.

 

Next a detail view of the rear:

 

 

To make a long story short, here a much more advanced view of the deck:

 

 

The way to get a proper solution to fix the deck to the hull proved to be quiet complex as some methods did prove not to be adequate.

 

The basic principle is shown in the following graphic:

 

 

The graphic shows the "U" profile fixed to the hull with the open side to the top. Within this one the "U" profile from the removable deck. The horizontal element on the top is the aluminium structure shown in a previous image, to which the "U" profile from the deck is screwed to.Within the "U" profile of the hull is a device which presents a threaded hole in which a M3 screw is screwed to press the deck onto the hull.

 

 

This bad photo shows how the boarder of the hull, i had just glued the veneer to the deckline, its width is just 1/3 of what you see in the image. On the Aluminium side of the deck a seal strip will be glued onto which by the screw pressing the deck onto the deck line of the hull, will be pressed against the veneer and so seal the connection between the deck and the hull. Should, what I do not believe, a drop of water pass through this sealing it will drop into the "U" profile of the hull. So nothing should get by this way into the hull either.

[Hellmut1956]

Next pictures about the mechanics for the rudder:

 

 

Here the picture about the rudder control within the hull:

 

 

The way the axis of the rudder is fixed at the bottom much later proved to be too weak and broke by accident. I did much later replace it by inserting a piece of steel, so that the final 29 kg weight of the model can be resisted by that part.

 

Following pictures that show how i have implemented the curvature of the deck and details about how the labyrinth sealing for the deck constructions work:

 

 

 

 

[Hellmut1956]

 

next image shows a top view how the labyrinth sealing around the deck constructions work:

 

 

I f you look in detail, you can see the "U" profile open to the top of the hull. On the deck you can see the wall that prevents water from flowing into the hull. The hull construction has a part that fits within and has an aluminium sheet that will end within the "U" profile of the desk and another parts that gets down to the deck outside this wall. On top of this "wall there is again a sealing strip against which the construction inserted and then screwed into position will press to prevent water from penetrating the hull on this way. Should a drop of water pass the sealing on top of the "wall" it will follow the aluminium sheet and drop into the "U" profile of the deck.

 

Well I will stop here to see if there is any interest in this description. have a nice day!

Edited February 16, 2013 by Hellmut1956

[Jim Lad]

A fascinating project, mate - and a huge model.

 

Please continue to update your log for us so we can follow the build of this beautiful yacht.

 

John

[michael mott]

What a lovely hull shape, I will be following along.

 

Michael

[Hellmut1956]

Thanks for your responses, I do continue. Yes Michael, I am in love with the shape of the hull. Some times I wish I would have left the hull as shown in the picture. But my son had to deliver its school work, so we colored the hull and after this it was impossible to keep it this way so I decided to laminate it with mahogany veneer. I have not completed the lamination yet, as there is always the chance to ruin the surface and as a consequence this will be done in the future. Here pictures from the colored hull. My son did a jokeand let it look like a dolphin.

 

 

 

Edited February 17, 2013 by Hellmut1956

[Hellmut1956]

What also resulted in a huge challenge was the choice of the method to press the deck against the hull. First i tried using the spacers as used to mount a motherboard of a PC into its case. Having an hexagonal shape I thought that by fixing it with a security nut and applying Epoxy to it would prevent them from turning reliably. It proved not to be the case and I had to drill one of them that got loose to be able to remove the deck. So I decided to use rivets together with an stainless-steel parts milled into a shape that made it possible for water in the "U" profile to move past those.

 

 

In this picture you see elements of the methods tried.First you see the hexagonal spacer that proved to be inadequate. Then you can see the stainless-steel bar, were every 3 holes make up one connection element. The next picture shows the shape of this bar milled  You can also see one rivet in place. The middle hole has an M3 thread cut into it, the two next to it a 2,5 mm hole for the rivets. This war while the rivets keep the connection bar in place it is impossible for it to turn has the spacer did. Unfortunately the rivets with this diameter proved to be too weak.

 

 

So I did cut M3 threads into all 3 holes, fixing the bar in its place with 2 M3 screws. But getting the M3 thread done was a mayor problem and so I learned a new rule I didn´t know before. A thread in a stainless-steel material cannot be deeper than its diameter. I had many taps break. So I had to drill the holes partly with an 3 mm diameter, to reduce the length of the thread to approx. 3 mm. of course I used best cutting oil and highest quality taps. But one more thing I learned is to use a hand drill to cut the thread. This has many benefits. One is it makes the task much easier, but the second reason is that it ensures a steady cutting speed and so this cutting benefits from the difference gliding resistance to the initial resistance once cutting. The first being much lower and so reducing the stress on the cutter.

 

 

This picture shows a connection element milled out of a stainless-steel bar with a a rivet, once on the desk and once inserted to document the method. As I wrote the rivet was later replaced by a m3 screw. To the right is an element made in my lathe with a screw next to it. The concept is to embed each screw used to press the desk against the hull with one of this pieces. Many of you, I am sure have experienced how a screw driver can slide away and this would scratch the surface of the finished deck. By embedding the screw into one of this brass parts the head of the screw would be surrounded by the brass part and so prevent the screw driver to scratch the surface of the desk.

 

 

Finally here a connection element screwed into position:

 

Edited February 17, 2013 by Hellmut1956

[Hellmut1956]

the next series of activities needs some explanation. The size of the sail boat model and the size of the largest rig, here a picture that shows the size of the sails:

 

 

To control the sails a commercial hoist is exorbitant. A fellow model ship builder paid approx. 400.- Euros or more than 500.- USD for such a piece of equipment made this choice forbidden for me. On the other side investigating along many lines made me come to the decision to make my own one based on a stepper motor. A stepper motor has the advantage to offer the largest torque when not moving and to have its torque being reduced as it turns faster.  Another aspect is that you can increase the torque by increasing the voltage applied. The electronics involved take care about limiting the amount of current  The technique is called PWM, or pulse width modulation. by the way this is the same technique used to control the servos by the receiver output. Every 20 ms a pulse is delivered in a length between 0.5 ms and 1.5 ms. depending on the length of this pulse the servo takes a related position. Well this technique makes sure that the sum of the current flowing is not exceeding the maximum tolerated by the motor by cutting the current in intervals to reduce the total amount of current flowing. The amount of current flowing is defined by the law of Ohm which is:

 

I [A] = U [V] / R [Ohm]

 

The nominal values of a motor give an indication of its internal resistance, i.e. 2 VDC and 3 A:

 

R [Ohm] = U [V] / I [A = 2/3 Ohm

 

But on a motor you have rotating coils which add a alternate current element to the equation. This means that there is a resistance part were R is replaced by I (inductance). Simply said, the faster the coils rotate and so the motor, an induced electrical tension builds up with a polarity inverted to the one applied  Lets say at a certain rotation speed you have applied 40 VDC and the induced tension is -20 VDC. Adding the 2 tensions results in just 20 VDC left and in consequence the torque reduced by 50%. This can play to our benefit, as a drum for the sheet controlling the sails having twice the diameter will reduce the need for rotating speed to achieve the same result by 50%. This means the larger the diameter of a drum, the slower the speed of the stepper motor, the larger the available torque. This works to compensate the higher torque demand resulting from a larger drum diameter, were the force issued to the drum from the sail and passed to the drum by the sheet, is multiplied with the radius of the drum to calculate the torque demand generated. As a consequence I have decided to use a large diameter drum. When I started studying this issues i did not know all I do know and so I had to decide the characteristics of the battery pack to be used in my model. So as in this issue are very many aspects, lets start step by step every issue of it.

 

For the control of the sails in small sail boat models the use of a servo can be an adequate and economical way to address this:

 

 

I think even as the text is in German, the picture is self explanatory. Should somebody wish it, just ask and i add the translations.

 

The limitation is the length of the rudder arm and the torque available from the servo. I would say that sail boats with a hull length up to 1 yard can live with this elegant solution.

 

A simple version of the solution using a hoist is shown in the next picture:

 

 

A rubber band is used to keep the sheet tensioned and so prevent it from making knots if the tension disappears when the sail moves towards the center axis of the hull demanding less sheet.

 

The most popular technique used in sail boats is shown in the next 2 pictures:

 

 

 

 

 

Here and endless sheet in the hull is moved by the hoist and the sheets that go to the main sail and the head sail are knotted to it. The available movement of the sheet is limited by the available length within the hull, or you have to pay with a loss i.e. of 50% of the available torque by using a block that doubles the available displacement length at that price.

[Hellmut1956]

Another important contribution came from Klaus Prystaz, who invented another technique he called the "magic box".

 

 

Here a schematic of the principle:

 

 

On the top of the second picture is the boom of the main. As it is done in reality in large sail ships, the sheet goes a few times between the boom and the deck to reduce the force on the control but multiplies the amount of sheet that requires being moved. His method overcomes this limitation by using two drums in its magic box. The difference of the radius of the two concentric drums achieves the increase of the displacement of the sheet to the boom. The mechanics at the bottom of the schematic shows the endless sheet, but doubling the amount of sheet displaced by using a block that is moved along a bar by the hoist.

 

Important to highlight is the role of the hoist and the endless sheet at the bottom. This is were the commands given by the user from its transmitter come into the equation. By moving the hoist connected to the receiver the user defines how much sheet can be made available at any single moment to the booms of the 2 sails and in this way limit how much the sails can open.

Edited February 17, 2013 by Hellmut1956

[Hellmut1956]

This proven method is as reliable as the most popular solution with the endless sheet shown above. What i did not like were a couple of issues. Once the endless sheet. This involves a mechanic that can lead to problems and demands quite a bit of complexity when building it. The second aspect I didn´t like was the loss of torque resulting from the 2 drums of different radius.

 

The stepper motor has another characteristic that can be taken into the equation when using a stepper opposed to using a hoist. by definition the position of a stepper motor can be known to the electronics and the software running on it. A short explanation. The most common high performance stepper motors do 200 steps for one 360° turn, or 1.8° per step. So the stepper motor can go to 200 known positions every turn! The next issue are the micro steps. Micro stepping is a technique, where every single one of those 200 steps can be divided into partial steps in increments defined by the power-of-two, said in numbers, by 1, 2, 4, 8, 16, 32, 64, 128 or 256 partial steps per single steps. lets take as an extreme using 256 micro steps per steps. A stepper makes 200 steps per 360° turn, if at each of those steps he makes 256 micro steps a total of 200 * 256 = 51,200 micro steps take place per one 360° turn of the motor or 51,200 positions are available per turn. I have decide to make a drum that has circumference of 400 mm, so the length of the sheet can be defined in 400 mm divided by 51,200 positions = 0.00781125 mm. As every one of you knows this is far in excess of the precision required to control the length of the sheet, this is beyond the resolution we can control from our transmitter. My demand for change of the sheet length is 12 times 400 mm = 4800 mm or 4.8 meters or 5.25 yards. As you can check, this change of length of the sheet cannot be implemented with the usual techniques or demands very special hoists and conic shaped drums as offered by 500.- USD plus commercial offerings! Here a picture of the sheet for the main as it was on the Endeavour, the way I want to implement it:

 

 

One of the implicit benefits of having such a high resolution, 51,200 x 12 turns = 614,400 positions at 256 micro steps is that I can reduce the resolution to achieve stability in the position information. 256 microsteps means 8 bits are used to describe each of those positions within a single step of the stepper. by ignoring the lowest value bit I just get 128 bits or 25.600 positions, or the lowest 2 bits, still 12,800 positions per step, still much more than needed. You should know, that the higher the number of micro steps, ideally 256, the smoother the stepper moves resulting in a very silent movement.

 

 

Sorry, this my video is in Spanish and it is recorded to demonstrate the operation the stepper I plan to use, but just fed with 12 VDC, much less than the up to 40 VDC i will be supplying at full batteries. At the beginning I thought i would need to isolate the stepper from the desk to prevent the desk surface to work as an amplifier, but it proved unnecessary. Still so you can appreciate how silent a stepper can be operated and how fast it can turn. As i just need a maximum of 12 turns of the stepper, the available speed is much higher than I do need what translate into slower resolutions and as a consequence higher available torque. As the applied tension will be between over 24 VDC with batteries needing to be recharged, twice to what is used at the video and a bit more than 40 VDC with full batteries, the available torque will be twice to more than 3 times what is available at the video.

 

What also can be seen at the video is that at certain speeds the stepper runs making more noise and trembling and that by lowering the number of micro steps and increasing the speed the stepper can stop rotating. This means that in the model I have to make sure the stepper runs and such speeds were it moves smooth and silent. Different from using servos or hoists, you will hear nothing when my stepper is operated in the model!

 

But here comes another issue! By being able to turn so fast it can destroy the mechanics and the sheet by accelerating too fast! Well the controller electronic component I plan to use and were I am using a small evaluation card in this video called StepRocker from a company named Trinamic, In my model I will put a card developed by myself. using those components there are a couple of parameters that allow to control this and to prevent this.  One is that the speed profile of the stepper motor follows a trapezoidal shape. It will start form 0 and accelerate to a target speed an decelerate as it approaches its target position resulting in the trapezoidal speed profile. But there is another feature called S-ramp, which smooths the trapezoidal speed profile by accelerating smoothly changing the lineal shapes of the ramps into an S-form shape. There is a video that demonstrates how a glass of water is once accelerated with a trapezoidal speed profile and once with the S-form ramps:

 

 

By accelerating the drum this way I can prevent the mechanics and the sheet in the model to be over stressed by controlling the steepness of the ramp and using this S-form speed profile. A benefit of the component is it computes by itself when the deceleration has to be started to stop at the target position.

 

having a stepper with 3 Nm torque, means a hell of a lot of power, if something goes wrong it can kill the complete model. for this the components from Trinamic have an additional feature that is called "stallguard". The following video from Trinamic, I have the written permission to use their videos, shows how stallguard prevents the potential hazard:

 

 

I don´t know why the last link is not working! all that comes after the second "=" sign identifies the video. It shows how if an obstacle is encountered it is possible to react to it by software.

Edited February 17, 2013 by Hellmut1956

[Hellmut1956]

Lets continue with background information to explain why I have made the decisions that I made to implement the sail control. Up to know I have explained how the sheet is moved to control the position of the boom of a sail and the pros and cons of the difference approaches and have started to explain my concept by describing and explaining the elements involved. So far i believe to have explained why a stepper is an option and presented many elements that document my choice. Key are 2 abilities. To know the position and to have a higher torque the slower it rotates, demonstrating how fast this can be and what consequences have to be taken into account. lets now move to my choice of batteries and why I choose the type and number and connection of the batteries.

 

I said the higher the voltage applied to a stepper, the higher the torque they make available, possible because the control electronics limit the amount of current flowing through the coils of the stepper by using a technique called PWM. I ant to add that the highest amount of energy would flow through the stepper when it is holding its position. as the stepper holding position is what will happen most of the time, this is a critical element to take into account.

 

For preventing this high level of energy consumption during the time it holds the stepper in position, I decided to use an electrical break. This breaks require 24 VDC to be applied to the break to release the shaft of the stepper. When no tension is applied the break holds the shaft in position and so the current flowing through the stepper can be reduced to "0" and increased to the required level prior to the next release of the break which happens prior to rotating the stepper motor.

 

I decided to go for batteries using the chemistry LiFePO4. This lithium based batteries are known to be the most reliable, most stable and most robust lithium based battery technology offering a still very high energy density. Headway produces some that have provisions at its poles to screw M6 screws into it. The fully loaded batteries offer a tension of 3.65 VDC and have to be charged again no later when the voltage of a cell reaches 2.1 VDC. I do need 24 VDC to release the breaks of the steppers, even when batteries are discharged to a level were recharging becomes mandatory. As a consequence the use of 12 of this batteries connected in series: 12 x 2.1 VDC = 25.2 VDC or more than 24 VDC required for the breaks. Fully charged means they supply 12 x 3.65 VDC = 43.8 VDC. This voltage is below 48 VDC, a voltage used in the telecommunication market place and so it is a value that electronic components can handle. So I go with 12 of this cells in series:

 

 

now the next question that came was to decide what capacity to use:

 

The lowest capacity of this kind of cells was 6 Ah which have a size of 38 m diameter and 138 mm length, the highest capacity, 16 Ah, 40 mm diameter by 163 mm length. So the difference in size in irrelevant for the model and as I did not and do not yet understand what the real power consumption will be i decided to go the safe path and went for the largest capacity of 16 Ah. It is easy to live with extra capacity and no way to fix if capacity is short!

 

The next issue was the time it would take until I really needed all those 12 batteries, at about 45 USD each. having batteries in inventory over a long time is not necessarily the way to go, specially when 45,- by 12 = 540.- USD are involved for a full set! Another aspect to worry about is the possibility of short circuiting them during the work of making the model and the danger that arises from 12 batteries, each 16 Ah capacity with a short circuit current of 160 A! So I decided to buy just one to have an original in my hands and make fake batteries out of wood and lead that would have the same dimensions and the same weight!

 

Here the manufacturing of those 12 fake batteries using my mill and a rotating table:

 

First I cut 12 pieces of beech wood.

 

 

You see the original battery cell next to the wooden pieces. next you see a single fake battery cell before lead is introduced:

 

 

Edited February 17, 2013 by Hellmut1956

[Hellmut1956]

Lets go to the manufacturing of the fake c ells:

 

 

This is the rotation table I own. next you see how I mill 6 long cavities into the sides of the wooden pieces. The size of those cavities was computed such, that filled with lead the fake batteries approx. would reach the same weight as the original battery.

 

 

 

And the result is shown next:

 

 

Here you see a single fake battery with the lead in the cavities and trimmed to meat exactly the weight of the original battery cell:

 

[Hellmut1956]

Next you can have a glimps about how the batteries will fit into the hull, placed into 2 boxes with 6 battery cells each.

 

 

The process to decide to split the battery box in 2 came from the objective to achieve the lowest possible point of gravity in the hull, which is of special importance in a sail boat with long keel, as it does not have a sword with a lead bomb at the bottom end.

 

 

 

First I tried to put all into one box and the result shown in the CAD picture made evident, that many of the cells would be placed above the floating level, which is highly undesirable. Another aspect related to it was the question as to where to put the stepper motor and where to put the drum.

 

As result of a hand made drawing shown next i did come to the decision of not only implementing 2 battery boxes, but to put a stepper motor in each and for each its own sheet drum, so that the sails could be controlled independently:

 

 

 

 

 

 

This picture shows the 2 stepper motors mounted on top of the 2 battery boxes which each an attached drum which is into a case milled out of 2 aluminium plates, 10 mm thick each. The distance between the walls of the battery boxes has to be at least 60 mm, so that the drum can be taken off the shaft of a stepper motor and removed at any time. It is assumed that water will come into the case of the drum brought into by wet sheets, so a need to remove the water is given!

 

This picture of one of the walls of the battery boxes shows the position of the battery cells in each of the boxes: The boxes are different, as the width of the hull is different at each of their locations:

 

 

 

This last picture can be used to present some of the concepts. You can see 6 circles of 40 mm diameter that represent the location were those batteries will be in the box and the shape of the wall shows the shape of the hull at the place it belongs. In this wooden version of the wall the holes were drilled with 6 mm diameter, so the M6 screws fit tightly into them holding the fake cells and the original cells later in position. Next to the 6 large circles you see 4 small circles here each 10 mm diameter. The holes drilled there have 3 mm diameter.

 

You can further see that I have made the effort to put the batteries as low as possible into the available space to ensure a lowest possible center of gravity of the hull. Those small circles are there to show the location of aluminium rods will go from one wall to the other one. Over those and around the outer batteries. I will bend a 0.75 mm thick aluminium plate to build a receptacle for the batteries that will be 100% sealed to prevent water from penetrating it and getting to the batteries.

Edited February 17, 2013 by Hellmut1956

[Hellmut1956]

The wooden version of the wall seen above is just a provisional one. A couple of reasons for this. One is, that they are made out of cheap wood, so replacing them when a conceptual error is found, represents no problem. Another reason is that being made out of wood no short circuit can happen through them. Further there will be the need to adapt them exactly to the shape of the hull when placed where they belong, so I will have to work on them, so they will need to be disposed anyhow. The final version will be made using an aluminium sheet, 10 mm thick. As aluminium is electrically conductive, there is the need to ensure no short circuit happens. For this I made plastic parts on my lathe and next you see the schematic made on my CAD software:

 

 

By now this parts were made, so here you get pictures of them:

 

 

To the left the part that goes on the inner side of the wall of the battery box. just 2 mm thick, but in the middle it has a noose 8 mm outer diameter, 6 mm inn er diameter, in which the screw that fits into the pole of the battery just fits. Next you see how they fit together:

 

 

The reason why embedding this parts into the aluminium wall of the box is so important results when viewing the location of the box for the drums, as you can see on the CAD drawing above. The case is mounted directly on the surface of the wall of the box. Another reason is that all interconnections, those connecting the battery cells in series and those needed for balancing the cells will be embedded in the wall. This way no accidental short can happen, as all will be sealed in Epoxy within that wall.

 

Here you can see how I milled the cavities to have those plastic parts fit into the wall using the rotating table.

 

 

Next you see those plastic parts fitted into those cavities:

 

[michael mott]

Very interesting Helmut I will have to study your design carefully so that i understand how it works.

 

Michael

[Hellmut1956]

Please, do not hesitate to ask. it is difficult for me to pass the big picture of my concept to control the displacement of the sheet, not knowing how much upfront knowledge a reader has. I do assume little knowledge, so I started explaining the methods.

 

Basically what I do is while still using conventional radio control in the 40 MHz range, is to decode the pulse length from every one of the 8 channels and translating it into a digital value by use of a micro controller  This info tells me what the command from the user is. Lets say it is a value between 400 and 800 between a stick fully pulled to a stick fully pushed.

 

In case of the control of the sail this is translated into a position for the drum, and consequentially for the stepper motor. It knows it may not move beyond this position as this is the limit the user has set through its transmitter.

 

I have the boom connected to an axis which rotates as the boom rotates. At the lower end of this axis, it touches the deck surface nearly, there is a special magnet. In the deck surface is at this place a small plastic sheet. It prevents water to penetrate into the hull but does not block the magnetic lines of the magnetic field. Below deck, exactly below the magnet there is an magnetic encoder, that supplies digitally 14 bits of angle encoding and an index pulse once per rotation. This gives a microcontroller below deck with an encoder decoding module in hardware integrated the position data of the boom. The software knows into which position it has the stepper motor to move so that enough sheet is made available for the boom. The software is programmed in such a way that it allows the boom to move freely supplying the amount of sheet required in its actual position of the boom. So never too much sheet is made available and so always a certain tension of the sheet is maintained. As soon as the boom tries to move beyond the position the user has limited its move to, it stops the boom. As a consequence no endless sheet mechanics are required simplifying the hardware under the deck dramatically.

[mtaylor]

This is a fascinating project, Hellmut.  An interesting mix of metal and wood and electronics.  I'm following this.

[druxey]

This is a fascinating build, Helmut. The technique you used for casting lead into the keel is an interesting one!

[Hellmut1956]

@druxley: casting lead into a wooden frame is no problem, as long as either only little quantities at a time are poured into, or if enough lead is already there to diffuse the heat. A different challenge arises as i will be using the technique that translated literally means lost wax. Lead shrinks by 2% when changing from liquid state to solid. When pouring it into the hull as I have done i.e. Epoxy has to be added to fill the gap of the shrunken lead. This of course only after having all lead desired.When I need to fill the empty spaces between the batteries or the empty spaces between the battery box and the hull. I first will fill those spaces with wax. Than, like making a pizza,  form a flat sheet of wax that has the thickness equivalent to 2% the lead will shrink and apply that on top of the wax form. The goal is that after shrinking of the lead it will have the right size! Lets see if it will work! 

To all, I was concerned if those parts of the project having to do with electronics and with motor control would be of interest. Fact is that by some relatively simple basics electronics can become a part of the tool set of a ship modeler. A friend and myself started this in Germany and now it is amazing what people do. I have in lieu of this experience done some evangelism in a forum in Spain and recently started in another one in Mexico with similar effects. Some basics about how to make a circuit board, how to get a micro controller running and the how the micro controller can decode the signals from the receptor is the first step to open that black box that electronics is for many model ship builder. i a bit like combing Lego stones.

[Hellmut1956]

Lets go back to describe the construction of the battery boxes. being by far the largest element within the hull and its position within the hull is most critical, I needed to have a reference point within the hull to build the following parts.

 

 

This reference point is a step clearly defined by lead within the hull Here you get a top view of this place with a wooden version of the wall put in place.

 

 

 

This picture shows the same provisional wall looking within the hull to the front. To explain the reason for this 2 aluminium bars glued longitudinal into the hull you need to know that finding the right way to implement things is not straight forward as it might look right now, as I am taking years long of experimenting, studying and thinking about how to do something. my best moments to think about issues are when I am either waling outside with mi dog, this gives him and me the physical exercise we both need, or being in the bath tub. At bot occasions I visualize what I am planing to do do and I come to recognize things that otherwise would have noticed only after implementing it. What also can be seen in this picture what huge problems it makes to me to adapt the frames to the inside geometries of the hull. You can see that the outline does not fit, neither at the bottom, nor at the right side. But as it is of utmost importance not only to get a perfect fit, but also to have it 100% aligned with the hull being 100% parallel to the waterline level, a lot of tuning is required.

 

By the way I want to introduce to you 2 aluminium raw parts that will play an important role.

 

 

This are the 2 aluminium parts from which I will make the drums for the sheet using the lathe and the mill.

 

 

This are the 3 aluminium plates, 10 mm thick I will use to make the walls of the 2 battery boxes.

[Hellmut1956]

 

Here you can see how my efforts to achieve an improved quality of a wooden wall, it is still the same one as before, has led to a much better but yet not good enough quality. You can also see that I have started to work on the second wall too. But doing so I did realize that the required precision to define the exact location were the second wall would be placed, as this has an immediate repercussion on the geometries of this wall.

 

Here a picture from the top showing the 2 walls as they are with the original of the battery installed between them:

 

 

 

The 2 pictures give you the opportunity to appreciate even more details. lets start with the first of the 2:

 

A view from within the hull you can see the screws i used to connect the parts of the aluminium structure of the hull that received its counter part, the deck. it also shows the effort I took to make sure the forces from the mast and the shrouds could be as large as They could be and that those forces are passed to en eggshell thick hull without endangering it. Behind the brass colored "L" profiles on both sides there are bars of aluminium, one on each side, 5 mm thick and made out of specially hard aluminium glued to the hull. As you can see in the second picture I have led the threaded holes in this aluminium bars to the deck level to connect the shrouds and its spanners to them. You can see that brass tubes lead those holes to the deck level in the middle bottom of the second picture.

The second picture shows how i glued to 5 mm wood plates together to reflect the 10 mm aluminium plates that will be used for those walls.

 

A critical element that I became aware around this time and were i was simply lucky, was the location of the rear wall of this front battery box and the height of the battery boxes. Next to this rear wall the drums and their cases will be placed. First those cases just fit into the available space and second it will be possible to place the and remove them from their future location. The same became true for the 2 stepper motors they just fit in place! Lucky me.

 

 

But it also shows how the batteries will be placed as close to the bottom as possible. You can see on this last pictures how I did make all 4 walls for the 2 battery boxes and how they are placed next to each other. I did make he two cases for the 2 drums provisionally in wood to verify this kind of things and to visualize the construction so i could better understand the implications of a design decision. each case is milled out of 2 10 mm thick aluminium plates, being the 2 halves in between which the drum would be rotating. To be able to remove them the length of the axis of the drums had to be short enough so I could remove the drum and in consequence the case in which it is, but all had to fit into an empty space between the 2 cases for the drum. As a result the distance between the back wall of the from front battery box and the front wall of the rear battery box have to be 6 cm, or 60 mm apart from each other. What i was lucky in, is that the  case fitted exactly in front of the cross "U" profile aluminium structure of the hull, as you can see on the image. This really was just luck, or God has had is eyes on me! Seeing this I did verify with putting the deck onto the hull, that those cases would not get in conflict with structures of the deck. This also indicated to me, that extreme caution and attention had to be placed on the exact location of the boxes within the hull. As a consequence I decided to make the final aluminium version of the rear wall of the front battery box which is placed at the step formed by the lead in the keel of the hull and start building the rest starting from this clear reference point!

 

 

In the light of this I repeat publishing this picture. it shows the 2 stepper motors with its correspondent cases for the drums and the drums. The distance between the drums are those 20 mm I wrote about earlier. The platform on which the 2 stepper motors are mounted are the top cover of its respective battery boxes. Getting to the point to build the real final version of the rear wall of the front battery box I decided and fortunately I had saved the money to buy the 11 original batteries to complete the 12 needed. Shortly after buying those Headway discontinued the 16 Ah capacity version of those batteries, and now only offers a 15 Ah version. First, this will not impact me until I need to replace a cell, but second, it proved the design decision to be right. Those 15 Ah batteries have the same diameter, but are a bit shorter. So when replacing the current battery cells I can make steel elements that compensate this change in length with threaded holes drilled into these steel parts that would compensate the weight and length of the new cells and could be screwed.

 

 

This picture shows in detail the positive pole of one of those battery cells with its threaded hole in the center. Going to build the real final version of the rear wall of the front battery box confronted me with the potential hazard resulting from the energy stored in those batteries. A short circuit would generate a flow of 160 A. So special attention had to be placed on not letting this happen.

Edited February 18, 2013 by Hellmut1956

[Hellmut1956]

Lets go to have a view of the electrical circuit of one of the 2 battery boxes:

 

 

The thin blue lines are those used by the balancer of the charger to measure the tension of each cell in the pack. So for 6 batteries you need 7 lines. Those lines have to be embedded into the aluminium wall of the battery box.

 

 

Here I encountered another of those potential errors that could have proved very critical and expensive if not discovered that early. my first intention was to place the circuitry of the top of the 2 graphs in the rear wall of the front battery box and the same circuitry into the front of the rear battery box. The reason for this thinking was, that so I could interconnect the two battery boxes by just placing a bridge between the positive and the negative pole of one of the boxes, evidently the top circuitry needs to be implemented mirrored, so that in the front box would have i.e. on the left side the positive pole, as shown in the graph, and the front wall of the rear box the negative pole on this side. I will connect both poles using a 3 mm diameter solid copper cable, which would connect to an automotive melting fuse for 16 A for protection purposes. This copper cable would get a powder coating and a heat shrinkable tube to protect it against accidental short circuiting. Now, walking through the country side with my dog I came to visualize that this is an error! The reason being, That doing it this way this would put me in conflict with the cases for the drums and for the bearing holding element for the axis of the drum, and if led to an outside location, than there would be no space available for the balancer lines. Remember the balancer lines have to go through the 10 mm thick cover of the battery boxes made out of aluminium and so have to be mechanically stable and protected not to cause a short circuit while removing or applying the cover of the boxes. The second electrical circuit picture shows the routing of the balancer lines through the respective walls of the battery boxes.  So I came to the conclusion, that those circuits that let the power connections of the battery boxes out had to be placed at the respective other side of the boxes, the front wall of the front battery box and the rear wall of the rear battery box.

 

It is this kind of issues I am encountering often during my design and construction phases that do require a lot of time to visualize and analyze to find this potential conflicts. On some of the partial projects with this model this kind of issues have taken me more than a year of research and studies. This things make this project so rewarding to me!

 

Each battery box consists of 6 individual batteries. They have to be connected in series as shown in the graph by the thick green lines. At the end of this series you have the output of the positive and negative pole shown in the picture as thick red and black lines. The power of the supply of the battery pack would flow through this interconnects. This, in case of a short circuit could be up to 160 A! heat is generated electrically and can be described using the Ohm formula and the formula for Energy:

 

R [Ohm] = U [V] / I [A]

 

P [W] = I [A] * U [V]

 

The energy 584 W would be flowing through the interconnection. The resistance is proportional to the heat generated while the current flows through the interconnections. As a consequence the circuit will heat up where the highest resistance is encountered. I did have an accident that made a short circuit for a fraction of a second. It proved that the point where the screws used to connect the batteries and the bridges that interconnect the poles of the battery were this point. The batteries did not suffer from this event and at that point the heat generated a yellow/white light. You can see in the next picture how the heads of the screws got a bit of black color, center top and middle left. On the other wall you can see the place were that light was visible.

 

 

 

And the next 2 pictures show what happened to the plastic parts during this short shortcircuit period of time:

 

 

 

You can see that the provisions taken did do their work and proved to be adequate. The rest of the data in this pictures, please ignore them as i will come back to them while reporting the construction of those elements.

Edited February 18, 2013 by Hellmut1956

[Hellmut1956]

Lets go ahead with the construction of the rear wall of the battery box in aluminium and explain some of the intentions i have and why I have those. The benefit in this report is that I have pictures of later stages of the project that allow to document my text with the adequate pictures.

 

 

This picture documents on the left side of the front battery box that the wall is composed of 2 10 mm thick aluminium plates. in this picture they appear to have a certain distance between them. Ignore right now the rest of what you see in this picture. The reason for this gap between the 2 plates are the heads of the screws used to screw the aluminium bars you see at the inner side of the box. I still had to mill cavities, 6 mm diameter, to have the heads of the screws for Allen screwdrivers embedded.

 

Here a picture of the "inner wall " at the rear of this battery box:

 

 

You see large holes, 8 mm diameter, drilled into it, those are the ones that correspond to the screws to connect the poles of the batteries and small, 3 mm diameter holes which are for the screws that fix the aluminium bars visible at the first image. This inner wall will have its inside milled away, so that by having a 0.75 mm aluminium plate following this line, and banded around the battery cells and those aluminium bars will result in a box were the batteries will rest upon.

 

 

 

her you have a view from within the hull from outside the box onto the rear wall of the battery box. The screw inserted in the middle top hole goes through the equivalent, still existing hole in the inner wall. You can see how well this wall fits to the hull curvature, but I will come back to this later.

 

 

here you can see how I am milling away the inner part of the inner wall so it takes the function planned for it. This inner wall, after having been covered with powder coating except where it will be glued to the hull, will ensure that perfect placement of the real rear wall of front battery box. The real battery box will be screwed to this inner wall which is glued to the hull and exactly that place, were the real wall has to be, ensuring the perfect placement of this first reference wall.

 

 

 

here you can see the end result of the inner wall, now with the cavities milled into it, so the heads of the Allen screws are embedded into the wall allowing for the real wall to be placed immediately adjacent to this one! You can now also appreciate that the removed parts of this wall allow for the room required by the battery cells and the aluminium sheet screwed and glued to it.

[Hellmut1956]

Lets go now to details how this inner wall was made. Again, I used my mill and my rotating table introduced to you earlier.

 

Here a picture of my mill, an Optimum BF20L with a digital indication of the position of the coordinate table and bench vise. This mill has a specially long coordinate table, X-axis, that allows very good work on it. I wish the Y-axis was longer! All three axis, X, Y and Z are displayed digitally, something that has proven very useful!

 

 

Next the rotating table plate:

 

 

This is a picture from when I received it and first did prepare it for work. I had to clean it and apply special oil to protect it against humidity and wear. But this picture shows that at the center of this plate there is just that part in black screwed into its position. You have to know that for milling round cavities you have to make sure the part being milled is placed in such a way onto this plate that the center rotating axis of the rotating table is were the center axis of that round cavity i.e. has to be and this has to be at the exact place were the center of the rotating mill element is. For the second you use the coordinate table and some sort of devices or methods, those depending on the part to be processed, to find the correct location for the rotating table mounted onto the coordinate table. When found this position is defined in the digital indicator as the position "0" of both the X and the Y axis. The first of the two requirements i would have expected the supplier to offer a solution, which was not the case. I have found, that as a hobbyist and amateur in using this kind of equipment we encounter problems that a professional worker would never encounter, as he has learned how to do it!

 

So what i did to mill the cavities into the wooden version of the wall was to make a piece of plastic that fits into the hole as shown in the next picture:

 

 

You can see that the hole in the middle of this flat table of the rotating table at the top has a cylindrical portion and below a thread. The plastic part I made on my lathe fits exactly into this hole and has a noose in the center on which I could place the holes drilled into the wall that defined the center of the cavities.

 

 

You can see that the main body of this plastic part remains below the surface of the table and so not interfering with the part to be milled mounted on top. The problem with this method is, that I milled i.e. the deeper cavity for the 5 mm thick plastic part and by doing so I shortened the noose of that plastic part rendering it useless for the future. The other problem is that I would need to make one of this parts for every drilled hole diameter.

 

Here a picture of the wooden version of a wall with the work to mill those cavities in progress:

 

[Hellmut1956]

So when preparing to do the aluminium version of the wall i started to think about another more flexible and adequate solution for this challenge! The solution that proved to work fine, was to make a part that fits into that center hole of the flat table of the rotating table and have an M3 threaded hole drilled into its center axis. Now i only need to make short cylinders of every diameter required with a center M3 threaded hole drilled into it. Passing a M3 screw through this center threaded hole allows me to screw that cylinder with the adequate diameter into the ground plate in the center hole, place the part to be milled such that the hole drilled at that center of the round cavity takes the cylinder noose and so automatically places the part to be milled at the right place. Once the orientation of the part milled is fixed by fixing the part to the flat plate I can unscrew this center noose and so the drilling will not affect it. I tried to make this part from steel, but failed to do this with my equipment. Here a picture of where I left:

 

 

 

The reasons why I left it at this stake were that the large flex I used to cut that piece from a longer bar used up a substantial part of the cutting disk and i have no money to replace it and that the cutting edge was so poor that milling it proved to go at the expense of the milling cutter, even more expensive. With the lathe the results were similar disastrous. The cutter of the lathe with cutting plate at the front proved to break due to the noncontinuous contact with the iron as it still was rough from cutting with the flex and with the unclean outer surface of the raw steel bars. That was an expensive try I stopped. using a normal cutter was better but i had problems and so I decided to use a plastic to make that ground disk. I know it will not last as long as it could an iron part, but I could move ahead!

 

This picture shows the result with an 8 mm diameter noose!

 

 

 

By the way this picture shows the screws with Allen heads I use to do all screwing connections in relation to the battery boxes. The tool used to screw and unscrew the screws is small enough to fit into the 2 cm space between the cases of the 2 drums. I remember an eye opening experience during my first large job at the university having to design and draw a gear. We had to write an instruction booklet how to mount and unmount the gear. Writing it I found out that this was actually impossible to accomplish so i did do a dirty trick to hide it and it was not discovered by the reviewers! Luck me! So with this experience in mind i do check this early in the design phase!

 

In this picture you can see how i mill the cavities into the aluminium wall of the battery box!

 

 

The same technique was used to mill away the inner wall described earlier. I am a bit in the hen and egg problem. I will continue my report describing how y got the aluminium wall made such that it fits as nicely as you have seen it into the hull.

Edited February 18, 2013 by Hellmut1956

[Hellmut1956]

So now I will describe how i came form the wooden wal to the aluminium one:

 

this picture shows how much room for improvement was left to adapt the form of the frame to the hull:

 

 

To improve it, besides working by sanding the wooden version to adapt it to the hull I went and put cellophane foil between the hull and the frame:

 

 

You can clearly see how bad the adaptation of the frame to the hull shape was!

 

 

Here you can see the outside view of this frame with the spaces between the hull and the frame filled, from the back looking forward!

 

 

Here the same frame but viewed from the other side!

 

 

Finally the frame sanded and now fitting perfectly!

 

This was then used to generate the aluminium version you have seen earlier. I cannot add that picture as I have reached the maximum number of pictures in an reply!

[Hellmut1956]

I will now continue to show my progress on making the connectors of the poles which will be integrated in the aluminium wall as written earlier.  For that purpose i purchased a 25 x 3 mm copper band. I did cut those into pieces long enough to make those connectors. All diagonal ones have a distance between center of holes drilled for the screws that makes the connection to the pole of a battery of 43 mm. Here a picture of the connector with just one hole drilled:

 

 

You can also appreciate a number of important steps that proved justified when I had the accidental short circuit described earlier. For having a maximum size of the contact surface between the head of the screw, you see its raw form next to the copper band, the hole drilled was receiving a phase of 90°, the same as the one the head of the screw has. Also, because the head of the screw needs to be embedded in the wall to prevent contact that could cause a short, the diameter of the cutter of the mill with this shape had a diameter of just 10 mm, so there is a step between the surface of the connector an the beginning of the phase. The screw is being worked on the lathe so that his head has just a diameter of a bit less than 10 mm and lowered to fit into the hole without reaching the surface level.

 

 

here you can see the 3 connectors with the worked on screws just mounted outside the front wall of the battery box. As I have written above, this was a wrong assumption, as this is the configuration that needs to be integrated into the rear wall of the front battery box, the one I am making first! i am proud as to how nice those connectors did result. I know, no mayor accomplishment, but I believe a model ship builder and specially one that makes it from scratch and not from a kit, needs those self motivation moments to keep working!

 

In this next picture you can see the current status of this first aluminium wall:

 

 

 

I have to rework the cavities for the copper connectors. As you can see from the left one, the distance between the aluminium of the wall and the copper connector is far too small and would have a high probability to have a short circuit happen at one of this places, when a short makes 160 A flow through the connectors. But the accident with the short proved, that this copper connectors did hot heat up! Evidently their size leads to a very low resistance which leads to very little energy being generated even at a 160 A of current flowing and their relatively big size helps the heat to be dissipated. What is missing before I can put this connectors in their places within the wall, besides reworking the cavities and fill those cavities with Epoxy is the fixing of the cavity dimensions for the connectors and integrating the channels for the balancing lines as shown in the graph further up.

 

 

I have received a sample of a textile that can resist up to 600°C to 700°C. I plan to put small strip of those between the copper connector an the aluminium of the wall to ensure that it is 100% impossible for the copper connectors to get in direct contact with the aluminium. I am further waiting to see if i get also a sample of the Epoxy capable to resist up to 1200°C of heat. I know I exaggerate, but that is in line with my way of proceeding. Only when that is solved and when I have made the final decision about how to place the the sealing strip on the wall and decided where to put the holes for screwing the wall to the inner wall, then I will apply powder coating to the aluminium part of the wall. For this I hope to get a transformer to change the 220 VAC we have here in Germany to the 110 VAC that the powder gun requires. This is typical for this project, I often have to get many lines of action to a common point to continue!